06th International Conference on Permafrost - 2 ... - IARC Research

PERMAFROSTSixth <strong>International</strong> <strong>Conference</strong>PRQCEEDINGS (Vol. 2)July 5-9, 1993Beijing ChinaOrganized byLanzhou Institute of Glaciology & Geocryology,Chinese Academy of Sciences & Chinese Society ofGlaciology and GeocryologySouth China University of Technology Press

Sixth <strong>International</strong> <strong>Conference</strong> Proceedings on Permafrost (Vol. 2)Published by South China University of Technology PressI(Wushan Guangzhou China)First Published 1993ISBN 7-5623-0484-X / P 1

PERMAFROSTSixth <strong>International</strong> <strong>Conference</strong>PROCEEDINGS (Vol. 2)

GENERALISED DlSTRlBUTlCOF PERMAFROST IN THENORTHERN HEMISPHEREJ.A. HeginbottomGeological Survey of Canada1992LEGENDExtensive discontinupermafrost -ISporadic andmountain permafrost

PREFACEAbout one-fifth of the land area of the earth, .A total of 189 contributed papers are includedis underlain by perennially frozen ground, or ' in first volume of the proceedings which conpermafrost.It affects many human activities, tains almost all the papers accepted for prcsencausingunique problems in the environment, tation at the paper sessions. A total number ofecosystem, resource development and construc- 98 papers, reports and abstracts are included intions in cold regions. Since permafrost is a this second volume of the proceedings, amongthermal condition, it is very sensitive towhich one is the reviewed oral presentationchanges in climate. Global warming could resultin permafrost degradation, causing resultantecological andsocioeconomic consequences. Thus,permafrost has become more and more Important inpaper writen by G.C. Lewis et al, 9 are invitedpapers presented at the special sessions, 77are poster papers, 5 are poster abstracts. Theposter papers submitted by Chinese authors andthe development of polar and high altitude Mongulian authors are reviewed. Many scientificregions which occupy key positions in the global and engineering disciplines were represented,system.It is necessary to give scientists andengineers an opportunity to meet regularly inorder to discuss the state of the art of scienceand technology in their fields, and to gain theimpetus for further work, as well as to comparepermafrost conditions with other regions of theworld, particularly regions where only seasonallyfrozen soils currently exist. The Inrernational<strong>Conference</strong>s on Permafrost are organizedto serve this purpose.The First <strong>International</strong> <strong>Conference</strong> on Permafrostwas therefore held in the United Statesat Purdue University. in 1963; the Second inYakutsk, Siberia, 1973; the Third in Edmonton,Canada, 1978; the Fourth in Fairbanks, Alaska,1983; and the Fifth in Trondheim, Norway,1988.The Sixth <strong>International</strong> <strong>Conference</strong> onPefmafroat (VI ICOP) was co-sponsored by severalnational scientific and technical organizations,and was held under the auspices of the ChineseSociety of Gleciology and Geocryology (CSGG),which is in the Adhering National Body of the<strong>International</strong> Permafrost Association ( IFA,, andwas organized by the Lanzhou Institute of Glaciologyand Geocryology (LIGG), Chinese Academyof Sciences, with the collaboration of theState Key Laboratory of Frozen Soil Engineering.LIGG. The support and the guidance of the<strong>International</strong> Permafrost Association wereextremely important for us in preparation forthe conference.The VI ICOP was successfully held on 5-9July 1993 at Beijfng. China. About 274 parricipants(including 26 accompanyings) from 21countries attended this conference. Among them,27 and 22 attendants participated 'in the fieldtrips from Beijing to Lhasa and from Beijing toTianshan Mountain, respectively.The eighth IPA council meeting was heldduring this conference on 5 July 1993, from1950 to 2145, at which new IPA officers wereelected and appointed. They are: President,Cheng Guodong (China): Vice President, Hugh M.French (Canada): Vice-president, Nikolai N.Romanoukkii (Russia), and Secretary General,Jerry Brown (USA).including physics, chemistry and mechanics offrozen soil, geophysics, periglacial geomorphology,soil science, climatology, hydrology,ecology, civil and mechanical engineering. Thehigh quality of the papers was the result ofhard work by the authors, as well as from theassistance given by the Editorial Committee ofthe <strong>International</strong> Permafrost Association, andby the numerous reviewers in the membercountries.Finally, the Chinese Organizing Committeewishes to acknowledge all of you that haveparticipated in the preparation for thisconference: the authors of the papers, thereviewers, the sponsors, the publisher, and thestaff of many institutions that have beenworking to make it a successful conference.Cheng GuodongChairmanChinese Organizing CommitteeSixth <strong>International</strong><strong>Conference</strong> on permafrostV

CONTRIBUTING SPONSORSThe Financial Support to the Sixth <strong>International</strong> <strong>Conference</strong> on Permafrostis mainly contributed by the following organizations.GovernmentalCommission of National Natural Science Foundation of China;Chinese Academy of Sciences;Geography Society of China:South-South Cooperation Fellowship Program, Academia Sinica:China <strong>International</strong> Centre for Yconomic and Technical Exchanges, Ministryof Foreign Economic Relations and Trade.Nongovernmental, ,The First Highway Survey and Design Institute, Ministry .of Communication;Cold Regions Development and Research Society of China:Central Coal Mining Research Institute. Ministry of Coal Mining of China:Heilongjiang Provincial Institute of Water Conservancy Science:The First Survey and Design Instittue, Ministry of Railway of China;Heilongjhang Provincial Institute of Cold Region Construction Science;Heilongjiang Provincial Institute of Communication Science:Northwest Institute of Railway Science, Ministry of Railway of China:Gansu Provincial Institute of Water Conservancy Science:Northeast Survey and Design Instittue, Ministry of Water Conservancy:Mining Industry University of China;Inner Monggulia Institute of Water Conservancy Science;Jilin Provincial Institute of Water Conservancy Science;Heilongjiang Provincial Institute of Low Temp. Construction;<strong>International</strong> Science Foundation.

CHINESE HONORARY COMMITTEE.Chairman: Zhou GuangzhaoPresident of Chinese Academy of Sciences,Academician.Vice-Chairmen: Sun Honglie President of Natural Resources ExpeditionCommission of China, Academician.Liu Dongsheng President of <strong>International</strong> Union for QuaternaryResearch, Academician.Sun Shu Vice-president of the National Natural ScienceFoundation of China, Academician.Zhu Lilan Vice-president of the State Science andTechnology Commission of China. Academician.Liu Shu Professor, the State Science and TechnologyAssociation of China.Members: Li JiejunZheng DuLi Yusheng Wang SijingXu ShaoxingCONSULTATIVE COMMITTEE' Chairman: Shi YafengAcademicianVice-Chalrrnen: Dai Moan Vice-Governor of Heilongjiang Province.Yang Shengfu Director of Engineering Administration,Department of the Ministry of Communicationsof China,Zhang Xiangong Professor, Geology University of China.Members: Zhu XuanDai Dingzong Xu RonglieOuyang Ziyuan Xue Shiying Weng ShidaZhang Jiazhen Zhao Chunian Cui ZijiuORGANIZING COMMITTEEChairman: Cheng GuodongVice-Chacrmen: Wu Ziwang Zhou Youvu Yu XiangGe Qihua Zhang JieSeceetary-General: Zhu YuanlinAssociate Secretary-Generals: Xu Xiaozu Chen Xiaobai Tong BoliangHuang Yizhi Qiu Guoqing Tong ChangjiangEu ZhongweiMembers: Wu Jingming Yu Qun , Liu HongxuHuang Xiaoming Jia Jianhua Zhu QiangDai Huiming Xie Yingqi Xu BomongLu GuoweiOverseas Members: J. Brown H.M. French N.A. Grave

The reviewof abstracts andmanuscripts forthispre-<strong>Conference</strong> publication was conducted underthe supervision of the Chinese Organizing Cornmitteeand the IPA Editorial Committee. Approximately450 abstracts were received from 24countries. It was necessary to limit the numberof papers from Russia and China. Thesecountries were asked to invite a’ more limitednumber of appropriate papers. A review form wasagreed to and each paper received two or morereviews. Some papors were rejected; in othercases the authors simply did not submit manuscripts.Individuals without papers were encouragedto submit posters and,have their abstractspublished in a post-<strong>Conference</strong> volume.In order to save time and to employ nativelanguages in the reviews, all Chinese andMongolianpapers were reviewed in China and allRussian papers were reviewed in Russia, employingthe standard review fonn. All Russian andChinese reviewed papers were available inEnglish in August 1992 when members of the EditorialCommittee met in Washington, D.C., duringthe IPA Council meeting. The papers andreview forms were examined and discussed at thattime, and members of the IPA Council wefe askedto assist with additional reviews.Reviewofthe 100 non-Chinese or-Russianpapersinvolved reviewers from many of the IPA membercountries. This review process was conducted bythe Chair, IPA Editorial Committee, in coasultationwith members of the Cormnittee. Below isa list of all individuals who provided thesereviews. The IPA Editorial Committee and theChinese Organizing Committee express their appreciationtoallthosewhodevotedtheirvaluabletime and expertise to this process. Deservingparticular thanks for their assistance inselecting reviewers and following up with manyof them are Alan Heginbottom, Geological Surveyof Canada; Nikolai Grave and Valery Volgina,Russian Academy of Sciences; Eugene Marvin,Cold Regions Research and Engineering Laboratory,representing the American Society ofCivil Engineers; and John Zarling, Universityof Alaska, representing the Amerkcan Society ofMechanicalEngineers. TheColdRegions Researchand Engineering Laboratory is gratefully acknowledgedfor furnishing instructions, samplesand layout sheets for preparation of finalcamera Copy for the proceedings volumes. Membersof the IPii Editorial Committee are:YerEy Brown, Chair, USAH.M. French, CanadaN.A. Grave, RussiaCheng Guodong, ChinaL. King, GermanyE.A. Koster. The NetherlandsT .L. P&wb, Ex Officio, IPA ExecutiveCommitteeH. Jonas Akerman, University of Lund, Lund,SwedenDonald Albert, Cold Regions Research andEngineering Laboratory, Hanover, NewHampshire, USABernard AZkire, Michigan TechnologicalInstitute, Houghton, Michigan, USADuwayne Anderson, Texas A&M University,College Station, Texas, USA.Ronald Atkins, West Lebanon, New Hampshire,USAAbdul Aziz, Gonzaga University, Spokane,Washington, USAT.H.W. Baker, National Research Council ofCanada, Ottawa, Ontario, CanadaRichard Berg, Cold Regions Research andEngineering Laboratory, Hanover, NewHampshire, USAPatrick Black, Cold Regions Research andEngineering Laboratory, Hanover, NewHampshire, USAGeorqe BlaisdeJ.1, Cold Regions Research andEngineering Laboratory, Hanover, NewHampshire, USASteven Blasco, Geological Survey of Canada,Dartmouth, Nova Scotia, CanadaJerry Brown, Arlington, Virginia, USAMargo M. Burgess, Geological Survey ofCanada, Ottawa, Ontario, CanadaChris Eurn, Carleton Univsksity, Ottawa,Ontario, CanadaNe1 Caine, University of Colorado, Boulder,Colorado, USAL. David Carter, U.S. Geological Survey,Anchorage, Alaska, USAEdward Chacho, Yr., Cold Regions Researchand Engineering Laboratory, Hanover, NewHampshire, USAEdward Chamberlain, Cold Regions Researchbnd Engineering Laboratory, Iianover, NewHampshire, USAIan D. Clark, univer6lty of Ottawa, Ottawa,Ontlrio, CanadaGary Clow, U.S - Geological Sumey, MenloPark, California, USABill Connor, Alaska Department of Transportationand Public Facilities, Fairbanks,Alaska, USAScott Crowther, Crowther Associates, Anchorage,Alaska, USAScott Dallimore, Geological Survey ofCanada, Ottawa, Ontario, CanadaLarry Dingman, University of New Bampshire,Durham, New Hampshire, USAJean-Claude Dionne, Universitb Laval, QuibecCity, Quhbec, CanadaFrancesco Dramis, Universita di Camerino,Camerino, ItalyLarry Dyke, Geological Survey of Canada,Ottawa, Ontario, CanadaRobert Eaton, Cold Regions Research andEngineering Laboratory, Hanovex,New Hampshire, USAP.A. Egginton, Geological Survey of Canada,Ottawa, Ontario, Canadavm

, I fDavid Esch, Alaska Department of Transportationand Public Facilities, Juneau,Alaska, USAK,R. Everett, Ohio State University, Columbus,Ohio, USAOscar Ferrians, U.S. Geological Survey,Anchorage, Alaska, USAKaare Flaate, Norwegian Road Administration,Oslo, NorwayD.C. Foxd, McMaster University, Hamilton,Ontario, CanadaStephen Forman, Ohio State University,Columbus, Ohio, USAHugh M. French, University of Ottawa, Ottawa,Ontario, CanadaMasarni Fukuda, Hokkaido University, Sapporo,JapanJohn R. Giardino, Texas A&M University,College Station, Texas, USAOdd Gregersen, Norwegian Geotechnical Insti-' tute, Oslo, NorwayGeorge Gryc, U.S. Geological Survey, MenloPark, California, USAWilfried Haeberli, Versuchsanstalt furWasserbau, Hydrologie und Glaziologie,Zurich, SwitzerlandBernard Hallet, University of Washington,Seattle, Washington, USAStuart A. Harris, University of Calgary,Calgary, Alberta, CanadaWilliam D, Harrison, University oE Alaska,Fairbanks, Alaska, USADavid G. Harry, Enezgy, Mines and Resources,Ottawa, Ontario, CanadaDonald W. Nayley, EEA Engineering Consultants,Ltd., Edmonton, Alberta, CanadaBeez Haten, Northern Engineering and Scientific,Anchoxage, Alaska, USAJ. Alan Heginbottom, Geological Survey ofCanada, Ottawa, Ontario, CanadaKaren Henry, Cold Regions Research andEngineering Laboratory, Hanover,New Hampshire, USAChristopher E. Heuer, Exxon ProductionResearch Co., Houston, TBX~S, USAKen Hinkel, University of Cincinnati, Cincinnati,Ohio, USALarry Hinzrnan, University o€ Alaska,Fairbanks, Alaska, USAFieter Hoekstra, Blackhawk Geosciences, Inc.,Golden, Colorado, USAVincent Janoo, Cold Regions Research andEngineering Laboratory, Hanovef,New Hampshire, USAThomas Jenkins, Cold Regions RasearchaandEngineering Laboratory, Hanover,New Hampshire, USANil9 Johansen, University of Alaska,Fairbanks, Alaska, USAG.H. Johnston, Ottawa, Ontario. CanadaAlan S. Judge, Geological Survey of Canada,Ottawa, Ontario, CanadaDouglas Kanc, University ot Alaska,Fairbanks, Alaska, .USAB.D. Kay, University of Guelph, Guelph,Ontario, CanadaG. Peter Kershaw, University of Alberta,Edmonton, Alberta, CanadaStephen Ketcham, Cold Regions Research andEngineering Laboratory, nanover,New Hampshire, USAJohn Kimble, Soil Conservation Service,Lincoln, Ygbraska, USALoren2 Kinq, Justus &iebiq Unlversitat,Gieasen. GermanyEduard Koster, University of Utrecht,Utrecht,, Tho NetherlandsWilliam B. Krantz, University of Colorado,Boulder, Colorado, USARaymond A. Kreig, RA Kreig & Associates,Anchorage, Alaska, USAPave1 J. Kurfurst, Geological Survey ofCaRada, Ottawa, Ontario, CanadaArthur H. Lachenbruch, U.S. GeologicalSurvey, Menlo Park, California, USABranko Ladanyi, Universitb de Montrbal,Quibec, CanadaY.P. Lautridou. Centre de Gbomorphologie,Caen, FranceAntoni G. Lewkowicz, Erindale College,University of Toronto, Mississauga.Ontario, CanadaB.H. Luckman, University of Western Ontario,London, Ontario, CanadaVirgil Lunardini, Cold Regions Research andEngineering Laboratory, Hanover,New Hampshire, USAPhilip Mafsh, National Hydrology ResearchInstitute, Saskatoon, Saskatchewan,CanadaTerry McFadden, University of Alaska, Fairbanks,Alaska, USAJ.D. McKendrick, University of Alaska,Palmer, Alaska, USABrainerd Mears, University of Wyoming, Laramie,Wyoming, USAMichael Metz, GeoTech Services, InC .,Golden, Colorado, USABruce Molnia, U.S. Geological Sumay,Reston, Virginia, USAYoshisuke Nakano, Cold Regions Resaarch andEngineering Laboratory, Hanover,New Hampshire, USAE.E. Nelson, Rutgers University, NewBrunswick, New Jersey, USAJ.F, Nixon, Nixon Geotech, Calgary, Alberta,CanadaWalter Oechel, San Diego State University,San Diego, California, USAKevin O'Nsill, Cold Regions Research andEngineering Laboratory, Hanover,New Hampshire, USAThomas Osterkamp, University of Alaska,Fairbanks, Alaska, USASamuel 1. htealt, University of Michigan,Ann Arbor, Michigan, USAKim Peterson, University of Alaska, Anchorage,Alaaka, USATroy L* P&w&, Arizona Stace University,Tempe, Arizona, USAJ.A. Pilon, Gaolopical Survey of Canada,Ottawa, Ontario, CanadaWayne Pollard, McGill University, Monttlsl,QuBbec , CanadaVarn Rdmpton, Tarrain Analysis and MappingServices Ltd., Carp, Ontaxio, CanadaW.R. Rouse, McMaster University, Hamilton,Ontario, CanadaFrank Sayles, Cold Regions Research andEngineering Laboratory, Hanover,New Hampshire, USAD.C. Sego, University oil Alberta, Edmonton,Alberta, CanadaPaul V. Sellmann, Cold Regiona Research andEnginearinq Laboratory, Hanover,New Hampshire, USAEaius Shaver, Marine Biological LabOXatQEy,Woods Hole, Massachusetts, USASally Shoop, Cold Regions Research andEngineering Laboratory, Hanover,New Hampshire, USAXugi Shur, RA~Kroig h Asaociater. Anchorage,Alaska, w?n -

Michael Smith, Carleton University, Ottawa,' Ontario, CanadaC. Tarnocai, Agriculture Canada, Ottawa, Ontario,CanadaRupert G. Tarr, Jr., Golden Associates, Inc.,Anchorage, Alaska, USAJ.C.F. Tedrow, Hutgers University, New Drunswick,New Jersey, USAlloward 'l'homas, America North Inc., Anchorage,AI aska, USAClement Tremblay, Ministry of Trans.port, St. Foy,Qubhec, CnnadaRein Vaikmae, Estonian Academy of Sciences,'l'allir~n. EstoniaRohert 0 . Van Everdingen, Arctic Institute ofNorth America, Calgary, Alherta, CanadaRrigittc van VLiet-Lanoe, Centre deGbomorphologie, Caen, FranceTheodore Vinson, Oregon State University,Corvallis, Oregon, USAJohn Vitek, Oklahoma State University,Stillwater, Oklahoma, USA1)onald A. Walker, INSTAAR, University ufColurado, Boulder, Colorado, USAJames Walters, University of Northern Iowa,Cerlnr Falls, Iowa, USAI3aolai Wang, University of Ottawa, Ottawa,Ontario, CanadaA. 1,i11coln Washburn, University of Washington,Seattle, Washing~on,.USAKathleen D. White, Cold Regions Research endEngineering Laboratory, Aanover, New Hampshire,US4Sidrrey White, Ohio State Univcrsity, Columbus,Ohio, USAPeter J , Willinms, Carleton University. Ottawa,Ontar io, CnnatlnMing-Ko Woo, McMaster tlniversity, llomilton,Ontario, CanadaJohn Zarling, Iltliversity of Alaska, Fnirbanks,Alaska,. USAHai Chongyuan, LlCG, Chinese Academy of Sciences,C h i ir nChctr Xiaohai, TJGG, Chinese Academy of Sciences,Ch I riaCIrcng Guotlotlg, LIGG, Chinese Academy ofSciences, ChinaCu Zhongwci, LIGG. Chinese Academy of Sciences,ChinaGun Dongxing LIGG, Chinese Acnrlrmy of Sciences,ChinaI, lnng Yizhi, LIGG, Chinese Academy of Sciences,ChinaKang Erst, L I GG, Chinese Academy of Sciences, 'ChinaLi Cuoliang, Northwest Institute of RailwayScience, M i nistry of Railway, ChinsLi Shude, LI c: G, Chinese Academy of Sciences,Chi naLiu l'ieliang, Northwest Institute of RailwayScience, Ministry of Railway, Chinayiu Guoqing, LIGG, Chinese Academy of Sciences,ChinaSheng Zhongyan, LIGG, Chinese Academy ofSciences, ChinaTong Boliang, LIGG, Chinese Academy af Sciences,ChinaTong' Changjiang, LIGG, Chinese Academy ofSciences, ChinaVu Rangjun, LIGG, Chinese Academy of Sciences,ChinaWu Ziwang, LIGG, Chinese Academy of Sciences,ChinaXu Xiaosu, LIGG, Chinese Academy of Sciences,ChinaZong Zhonggong, LIGG, Chinese Academy ofSciences, ChinaZhang Changqing. LIGG, Chinese Academy ofSciences, ChinaZhou Youwu, LIGG, Chinese Academy of Sciences,ChinaZhu Qiang, Gansu Provincal Ins~itute of Water'Conservancy Science, Ministry of WaterConservancy, ChinaZhu Linnan, LIGG, Chinese Academy of Sciences,ChinaZhn Yuanlin, LtFG, Chinese Academy of Sciences,ChinaV'.T. Ralobaev, Permafrost InstituLe, Siberiannranch of the Russian Academy of Sciences,Yakutsk, RussiaL.N. Chrustalev, Moscow State University,Moscow, RussiaV.P. Chernjadiev, Institute of EngineeringConstruction Survey, Moscow, RussiaG.T. Dubikov, Institute of engineeringConstruction Survey, Moscow, RussiaA.D. Frolov, Russian llumanities University,Moscow, RussiaR.U, Genadinnlk, Institute of Cryosphere of theSiberinn Branch of the Russian Academy ofSciences, Tyumen. RussiaN.A. Grave, Russian National PermafrostCommittee, Russian Academy of Sciences,Moscow, RussiaS.E. Grechishev. Insti~dte of ltyrlrogeology andEngineering Geology, Moscow, RussiaI.E. Gurianov, Permafrost Institute, SiberianRranch of the Russian Academy of Sciences,Ynkutak, RussiaI.V. Klimovsky. Permafrost Institute, SiberianBranch of llle Russian Academy of Sciences,Yakutsk, RussiaA.A. Msndarov, Permafrost Institute, SiberianRranch of the Russian Academy of Sciences.Yakutsk. RussiaN.G. Moskalenko, Institute of Hyrlrogeo ogy andEngineering Geology, Moscow, RussiaA.V. Pavlov. Tnstitutc of Hydrogeology andEngineering Geology, Moscow, Russia0.P. Pavlova, Institute of EngineeringConstruction Survey, Moscow, RussiaN.N. Romanovsky, Moscow State Un+versiMoscow. RussiaG,E. Roaenbaum, Moscow State University, Moscow,Russin4.V. Sadovsky, Institute of Basements andUnderground Constructions, Moscow, RussiaN.V. Tumel, Moscow State University, Moscow,RussiaK.F. Voitkovsky. Moscow State University,Moscow, RussiaS.S. Vyalov, Moscow Ensineerlng ConstructionInstitute, Moscow, RussiaYu.K. Zatetsky, Institute "liydroproject."Mo'scow, RuasiaJerry BrownChairman of theIPA Editorial CommitteeZhu YuanlinSecretary-General of theChinese Organizing CommitteeN

Opening Plenary SessionClosing Plenary SessionField Trip A-1 from Lanzhou to Lhasa (July 12-22,1993)Tian Shan Field Trip, A-2 (July 11-18,1993)CONTENTS-<strong>International</strong> Permafrost Association Standing Committees and Working GroupsKcsolution965971914979983986Special SessionPermafrost and Changing ClimateNelson F.E., A.H. Lachenbruch, M.K. Woo, E.A. KosterT.E. Osterkamp and M.K, GavrilowPresent Human Induced Climatic Change and CryoecologyGavrilova Maria K.Recent Permafrost Degradation along the Qinghai-Tibet HighwayCheng Guodong, Huang Xiaoming and Kang XingchengResearch on Permafrost and Periglacial Processes in MountainAreas-Status and PerspectivesHaeberli WilfriedPermafrost in the Mountain Ranges of North AmericaHarris Stuart A. and John R. GiardinoMountain Permafrost in EuropeKing Lorenz and Jonas AkermanStudies on Mountain Permafrost in AsiaQiu GuoqingLinear Construction in Cold Regions-Paved Roads and AirfieldsVinson Ted S.Current Development on Prevention of Canal from Frost Damage in PRCChen XiaobaiContributed SessionA Model €or the Initiation of Patterned Ground,owing to DifferentialSecondary Frost HeaveLewis G.C., W.S. Krantz and N. CainePoster SessionInitiation of Segregation Freezing Observed,,in PorousSoft Rock during Melting ProcessAkagawa SatoshiThe Ensuring and Ecological Safety on the Gam PipelinesOperating in the Permafrost ZoneAntonov-Druzhinin Vitaly P.Engineering and Geocryological Studies of the Central Part OfYamal Peninsula Caused by its DevelopmentRaulin V.V., A.L. Chekhovsky and I.I ChamanovaMethods of Large Scale Ecological and GeocryologicalClassification of the Northern Part of Western SiberiaChehovsky A.L. and I.I Shamanova98710061010X0141019102210281031103710441050105410601062/I

Experimental Studies on Ice Segregation and the Modesof Frost HeavingChen Ruijie and Kaoru HoriquchiThe Relationship between Ice Intrusion Temperature andConfined PressureChen Ruijie and Kaoru Horiguchi10641067Preliminary Tests of Heave and Settlement of Soils undergoingOne Cycle of Freezing-Thaw in Closed System on A Small Centrifuge 1070Chen Xiangsheng. A.N. Schofield and C.C. SmithFrost Susceptibility of Powdered Calcium CarbonateChen Xiaobai. Corte A.E., Wang Yaqing and ShengYuComparison of Two Ground Temperature Measurement Techniques atan Interior Alaskan Permafrost Site 1076Collins Charles M.. Richard K. Haugen and Timothy 0. HorriganPreliminary Study on the Freezing Point in SoilCui Guangxin and Li YiIO7 31079Calculation of Maximum Thawed Depth of Permafrost under theBlack-Colour Pavement Based on Geothermal Gradient 1082Cui Jianheng and Yeo Cuiqin)Observation on Periglacial Mess Movement in the Head Area ofUrumqi River and Laerdong Pass, Tianshan Mountains 1086, Cui Zhijiu, Xiong Heigang and Liu GengnianLong-Term Shear Strength of Frost-Thaw Transit ZoneDing Jingkang, Xu Xueyan and Lou Anjing1092The Compressing Properties and Salt Heaving Mechanism Study ofSulphate Salty SoilFei Xuelipng and Li Bin1096A Computationally Feasible Reduction of the O'Neill-MillerModel of Secondary Frost HeaveFowler A.C., G.G. Noon and W.B. KrantzGeocryology in Mt. TianshanGorbunov A.P.The Freezing and Frost Heave Regularities of Base Soilfor Arbitrary SlopeDirection and Gradient 1108Guo Dianxiang. Vei Zhengfeng and Ma YijunPeriglacial Period and Pleistocene Natural Environment ofWestern Mountains of BeijingGuo XudongThe Physics of Liquid Water in Frozen Powders and SoilHaiying FU J.G. Dash. L. Wilen and B. HalletAStudy of the Thermal State in the Permafrost at theSejong Station, AntarcticaMan UK and H.C. Sung11001105111311171119Two-Dimensional Stefan Problem around A Cooled Buried Cylinder1124Haoulani H.. A.M. Cames-Pintaux and J. Aguirre-PuentePermafrost Mapping Using Grass 1128Haugen Richard R.. Nancy H. Greeley and Charles M. CollinsCircumarctic Map of Permafrost and Ground Ice Conditions 1132Heginbottom J.A., J. Brown, E.S. Melnikov and O.J. Ferrians Jr.Permafrostin Greenland Studies1137Henrik Mai and Thorkild ThomsenSnow and Permafrost in the Tian Shan Mountains1144Hu Ruji and Ma Hang

Frost-Action Design and Applications of Enlarged TypePile Foundation Bridge in Waterlogged Area of Song Tong, Huang Junheng, Xu Zhenghai, Ge Huanyou and ZUO LiThe Shallow Cover Design and Construction Technology ofBuilding Foundations in Daqing RegionJiang Hongju and Cheng EnyuanThe Prevention 'and Treatment of Frost Damage .on Buildingsand Canals in Permafrost RegionsJiao TianbaoDetermination Method for the Coefficient of the Degreeof Sunshine and Sunshade on CanalsLi Anguo and Chen QinghuaSimilarity Analysis of Modeling Test of Frozen Soil under LoadLi Dongqing and Zhu LinnanAComposite Model of Multiple Actions for Forming Patterned GroundLi Guangpan and Gao MinConsolidation of Deep Layer Frozen Soila in Triaxial TestsLi Kun, Wang Changsheng and Chen XiangshengPermafrost and Periglacial Landforms in Kekexili Area ofQinghai ProvinceLi Shude and Li ShijieRegional Features of Permafrost in Mahan Mountain and theirRelationship to the EnvironmentLi Zuofu, Li Shude end Wang YinxueA Solution for the Icing Heave of Foundations in PermafrostRegions by Lowering the Ground Water TableLiu Shifeng and Zou XinqingThe Geographic Southern Boundary of Permafrost in theNortheast of ChinaLu Guowei, Wang Binlin and Guo DongxlngRoad Design and Senovations of the North Slope in Da Hinggan LingLuo WeiquanPracticeof Reinforced Concrete Strip Foundation in Permafrost RegionsMen ZhaoheA Microstructure Damage Theory of Creep in Frozen SoilMiao Tiande, Wei Xuexia and Zhang ChangqingThe Paleoclimate Charactersitics in Xinjiang since the Late PleistocenePan AndingThe Introduction of Application Methods for CT in Frozen SoilExperimental ResearchPu YibinPreliminary Data for Permafrost Thermal Regime and i ts Correlation' Meteorological Parameters near the Spanish Antarctic StationRamos M.An Experimental Study of Canal Lining Prevented from Frost Damageand SeepageRen ZhizhongThe Impact of 'Salt Type on deformation of Frozen Saline Soils 'Roman L.T.. Alifanova A.A. end Zhang ChangqingGeophysical Methods of Cryology Ecological ,MonitoringSedov B.M. and Yu, Ya. VaahchilavPermafrost in the Selenge River Basin (on the Mongolian Territory). Sharkhuu A.114811521155115911641167117111741178118311861190119311971202120812111215121912221223\

Deformation of Thawing Dispersed Large Detrital Rocks of Cryolite ZoneShesternyov D.M.Palsa Formation in the Daisetsu Mountains, JapanSone Toahio and Nobuyuki TakahashiA Study on Characteristics of Ice-Damage and Prevention ofHydraulic Projects in North ChinaSu Shengkui and Zhang TiehuaThe Lates Pleistocene Cryomere in the Region of Kopjes and theBig Mesetas, Patagonia. ArgentinaTromobotto Dario and Bernd SteinSeasonal Freezing and Thawing Grounds of MongoliaTumurbaatar D.Highwall Stability in Strip Mines in PermafrostVakili JalalDistruction and Rehabilitation of Shaft Lining Used in Frozen ShaftWang Changsheng and Liu RihuiPressure Influence on Pore Charactersitic of Frozen SoilsWang Jiacheng, Xu Xiaozu, Deng Yousheng, Zhang,Lixing,IU.P. Lebedenko and E.M. ChuvilinPermafrost Change for 'Asphalt Pavement along Qinghai-Xizang HighwayWang Shaoling and Mi HaizhenA Study on Preventing Frost Heave of the Shaft-Type Energy DissipatorWang ShirongField Experiment Research of Water and'Heat Transfer within Freezingand Thawing Silt Loam under Fixed Groundwater LevelsWang Yi, Gao Weiyue and Zhang LianghuiThe Effects of Gold Mining on the Permafrost Environment, Wuma MiningArea, Inner Mongolia of ChinaWaag Yingxue and Tong BoliangRecent Discovery of Periglacial Phenomena on Tu Wei Ba Shan(Broken Tail Hill) in Zhalainoer, Inner MongoliaWang Zhenyi and Lin YipuUniaxial Stress Relation of Frezen LoessWu Ziwang. Ma Wei, Chang Xiaoxiao and Sheng ZhongyanApplications of Data Base Technology in Frozen Soil ResearchXia ZhiyingFrost Heave Properties of Nonsaturated Compacted Cohesive Soiland its Application in Winter Construction of Core DamsXie Yinqi and Wang JianguoObservation and Research of Sorted Circles in Empty Cirque atthe Head of Urumqi River Tian Shan, ChinaXiong Heigang, Liu Gengnian and Cui ZhljiuStudy and Development of the Techniques against Frost Damage ofHydraulic StructuresXu Bomeng. Li Anguo and Shao LijunUnfrozen Water Content in Multi-Crystal IceXu Xiaozu. Zhang Lixin. Deng Youbheng, Wang Jiacheng,IU.P. Leb'edehko and E.M. ChuvilinThe Thaw Settlement of Railway Foundations in Permafrost Swamp RegionsYang Hairong. Liu Tieliang and Guan ZhifuThaw-Consolidation of Unsaturated Frozen SoilYang Lifeng, Xu Bomeng and Lu Xingliang1227123112351238124212471251125512591262,12651269127212741278128212871292129512981301

Notched Charpy Bar Impact Test on Frozen SoilYu Qihao and Zhu YuanlinPermafrost Characteristics and the Exploitation and Utilizationof Ground Water in Hanjiayuan, Da Hinggan Ling, ChinaYuan Haiyi and Liu XuekuiSeasonally Frozen Ground and its Behavior on Frozen Heave inthe Yamenzhen Region, Gansu Province, China ,Yue Hansen and Qiu Guoqing130413081312Culvert Engineering in the Permafrost Region on Qinghai-Xizang Plateau 1317Zhang Jinzhao and Yao CuiqinNumerical Analysis of Temperature and Stress on the Canal Subsoilduring FreezingZhang Zhao and Wu ZiwangThe Relationship between the Railway Project Construction andEnvironment Protection in Permafrost AreaZheng QipuRegularity of Frost Heave of the'seasonally Frozen Soil i nHetao Irrigation Area, Inner MongoliaZhou DeyuanFossil Periglacial Landforms in the Shennongjia Mountains. ChinaZhou ZhongminThe Research of Porous Slab Structures for Preventing FrostDamage of RoadsZhu Yunbing and Guo ZuxinDrilling Characteristics of Engineering Geology of Permafroston Da Hinggan Ling RegionZou XinqingQuaternary Geology and Geocryology in Northern Quebec, CanadaAllard Michel. Jean A. PilonRational Utilization of Water Resources in Permafrost Regions,Artificial Recharge of Groundwater StorageBurchak T.V. and L.M. DemidyukIce Wedge Development Slopes, Fosheim Peninsula, EllesmereIsland, Eastern Canadian ArcticLewkowicz Antoni G ,Lakes and Permafrost in the Colville River Delta, AlaskaWalker H. JesseSorted Circle Dynamics: 10 Years of Field Observations fromCentral AlaskaWalters James C.Author IndexGeneral Subject - Senior Author IndexList ofparticipants in VI ICOP13211326133013341338134213441344134513451346134713511354

OPENING PLENARY SESSIONTuesday, July 6, J 993ClIRNG GUODONG - Respec-ted Dr. Troy P&w&, l’residentof the <strong>International</strong> Permafrost Association;rcspected Medam Deng Nan, Vic.e President of theState Science R I I ~ Technology Commission of China:respe.:red Yadam Hu Qiheng, Vice President of thcChinese Academy of Sciences; distinguished delegates,distinguished guests, and ladie’s andgentlemen: Welcome to the formal sessions of theSixth TnLernational <strong>Conference</strong> on Permafrost. Iam Cheng Guodmng, Chairman of the Chinvst: OrganizingCommittee for the Confcrcncc. It gives megreat pleasure tu announce the formal opening ofthe Sixth lnternational <strong>Conference</strong> on Pernafrost.This <strong>Conference</strong> is Organized by the LanzhouInstitutecof Glaciology and Ceocryology w i . ~ h t.hec0llaboration of the State Key Laboratory ofFrozen Soil Fngineering of the 1,anzhou Institute,and the Center for Internatimnal Scieuti I ic.Sxchange of the Chinese Academy of Sciences,under thc auspices of the Chinese Society otGlaciology and Geocryolugy. The society is theAdhering Body for the lnterr~ational PermafrostAssociation. On 5ehalC of t.ht!se institutions,1 have the honour of expressing the warmestwclcome to all parlicipants of this <strong>Conference</strong>.‘Today, T will first introduce the digniLacius atthe front tahle :+nd then say a few words ahoutthe <strong>Conference</strong>. This will bc followed by anopening address fl-om the presidellt of thclntcarnational Perma1 rust Association.It is my pleasure to introduce t-he persons atthe front tahle. A t my far right is t’rofessrlr7,hu Yuanlin, Secretary general of the <strong>Conference</strong>,and Deputy Directur’ of the State Key I.~bvr-atoryof Frozen Soil Engineering, N e v ~ LO Profcssor%hu is Vr. 7hang Hcngxuan, Vi(’+ Chairman of theCold Regions Tjevclvpmcnt and I,esearc.h Suciety ofChina. hext. is DI. .Jerry Brown, an oversraymember of (.he 0r.yanizing Committee for thisConfcrcnco and who contrihut.e(l n grc.at amountoi hclp during the preparat.11jn of this Conferencr.Next to Pr. Drown is Academician Shi Yafcng,Chairman of the Consultative Committee for the<strong>Conference</strong>, Honorary President of the ChineseSOC iet.y of Glaciology and Geocryology, aIId!lorlorary Direct-or of the Lanzhou Institute ofGlaciology and Geocryology. Next is Dr. HughFrench, the leader of the Canadian Delegation,~11d Dcnn of Science at the University of Ottawa.On my immediate right is Madam Hu Qiheng, Vicc-President of the Chinese Ac-ademy of Sciences.A n d on my i.mmediatc left is nr. Troy t’Cw6,President of the Tnternational Permafrost Association.Next tn Dr. PCwC is Madam Dcng Nan,Vicc President. of the State Sciencc and TechriologyCommission of Chirla. Her attendance is a veryexc~ting event. Next to Madam Deng is Dr. RossYackay, Secretary General of the TPA, a leadingpermafrost researcher and holder of numeroushonors. Next to Dr. 41eckay is Academici~n TuGunngzhi. Member of the Presidiumof the Chinese Academy of Sciences, Chicfof t.he Division of Geoscicnces. Next to AcademicianTu is Ilr. Love1 1, leader of t.he US Delegation,and. the Chairmrrn of the Advis(.)ry Committeeon Working Croups of the TPA. Next is Dr.Kamensky,leader of t h c Russian Delegation, and Directorof the Permafrost 1nst.i tute ot the RussianAcademy 1)f Sciences.Let me turI1 rlow to the Conferencc. Yore than277 research papers and poster papers were submittcd,among t.hem about 159 rrom overseasscientists and en’gineers irlvolvcd in geocryology,and 124 from Chinese participallts. About 300frozen ground researchers from 22 countries areattending !.his <strong>Conference</strong>. This gives us anopporLurlity to review thc state of our knowledgeon the subjcct. Me w i l 1 bc able to henelit fromthis opportunity to develop a vision of t.hcfuture for permafrost research and develupmentin the world. I would also I ikc to point outthat many papers have been written primarily hhyoungresearchers; thus it seems that the Conferericrwill achieve one of its primnr) ubjcctives965

'iich is to promote a ncw generation ljf scieqtistsand enginccrs intercstcrl iu geocryology.In this time de havc a many problcms wiLh~puyulaLinn growth, resource dcvcloyments and theproI.ecLiun vf the euv ironment. Many of our per-mcrrrostregions are rich in rcnew:+hle and nonrenewableresources which are rf+guired for thebenefit of t.he human race. To develop thesercsouic.es, we have to continually scek L o improvcour engineering dcsign and construct ion techniquesand help promole a clean environment. Linkedwit.h this is the realization that permafrostscience, as the hasic rcscorc.h, involving theorigin, dist.rihution and nature of permafrost andrelated ficlds such as hydrology are cssential tosound engineering design and clean safe tnvironment.A t present, considcrable attention basbeen direcLed to concerns regarding yotent.ia1c.honges in permafrost. due LC human ac.tivi ties andcarbon dioxide-induced c1imat.e warming. We mustremember that the ground is the very foundationon which thc ocusystems evolve rrnd huwan infrastructuresare built. Climatic warming wi 11change the permafrost terrain, thus changing thevery foundation UT the ecosyst-ems and infrastructures.It is, therefore, clear' Lhat wc are facedwith numerous challenges. To meet these challenges,we need not only to work hard, hut also to learr.through the experiences of others. For thesereasons, international c-onferences and internatiorlalstudies are essential. The lntcrnationalPermafrost Association is thus 3rganizetl toserve this purpose. Let's now call on Dr. TroyP&w&, President of the <strong>International</strong> PcrmafrostAssociation.TROY L. PkWk - Thank you Professor Cheng, MadamDcng Nan, Madam Hu Qiheng, distinguished guests,and ladies and gen~lemen: The <strong>International</strong>Permafrost Association is most pleased andhonored that the People's Republic of China ishosting the Sixth <strong>International</strong> <strong>Conference</strong> onPermafrost. This marks a turning point in thehistory of IPA. Now each of the founding countrieswill have hosted an Tnternational Perma-Frost <strong>Conference</strong>.Madam Deng Nan, Madam Hu Qiheng, it is mypleasure to thank you for- the continuing interestand su.pyort that your respective organizations,and your country in general, havc given toresearch and applic.ation of permafrost s~udies,not only in China, but to international cooperationin this field.On behalf of the IPA, I wish to extend ourgreat appreciation to the Chinese Society ofGlaciology and Geocryology and to the LanzhouInst i Lute of Clac iology and Geocryology underwhose auspices and organization this conferenceis being held.It has heen ten years since the officialfound irlg of the Tnt-ernational Permafrost. Associationin 1983 at the Fourth <strong>International</strong> Confert.nceon l'ermaf rust held at Fairbanks, hleska.Rut it has been 20 years since the idea for aninternationti1 organization was advanced by P.I.Melnikov at the Second <strong>International</strong> <strong>Conference</strong>on Permatrost held at Yakutsk, USSR, in 1973.Our international activities started with theFirst Intarnatjonal <strong>Conference</strong> on Pcrmrrfrostheld at I'urdue University, Lafayettc, Indiana,USA, in 1963. IL has been my privilege to participatei n all of Lhe conferences to date andserve as Vice President of IPA, 1983-1988, andPresident, 1988-1993. Ross Mackay has been ourfoun+ing, hard-working, and guiding SecretaryGeneral for the pioneering 10 years. Ross andI cxtend our heartfelt thanks to the other membersof the Executive CommiLLees who guided theAssociation through thesc early years: P.T.Meln i kov, President, 1983-1988; Kaare Float.e,Vice President, 1983-1988; and Cheng Guodong andV.P. Melnikov, Vice Presidents, 1988-1993. Forthe last fivc years, pcrsonnel of the StandingCommittees, Working Groups and National AdheringBodies have greatly aided in the development ofour active organization.Perhaps it would be well to review progressof the IPA, especially of the last five years.Four of the 20 Adhering Member Countr1.e~ (Canada,China, USA, USSR) were charter members, and theothers joined in the last ten years. Our constitutionand bylaws were formalized in 1988 andrevised in 1993.With a firm early foundation, IPA has undergonean extensive maturing and expansion. TheExecutive Committee and members of the Councilhave met annually: 1988 (Oslo), 1Y89 (Yamburg,Siberia, USSR), 1990 (Quebec City), 1991 (Beijine),1Y42 (Washington, D.C.) and 1993 (Beijing).The council met in 1990 and 1992. In 1989 TPA wasapproved as an Affiliated Organization of theTnternational Union of Geologioal Sciences (Illcs),one of the largest and most active non-governmentalscientific organizations in thc world.The establishment and functioning of StandingCommittees and Wor'king Groups during the lastfive ycars.probably has been the most importantaction of the TPA to date.The 24-page IPA news bulletin, Frozen Ground,has been well-received worldwide since itsinitiation in 1990 when it grcw from the ori.gina1IPA newsletter. Puhlication is thc courtesy of

the Cold Pegions Research and Engineering 1,abora-Lory, 'lanover, N.Y. USA, and 1600 copies arepr-intell and dist.ributed throughout thc world bythe.~+~Iher ing mcmbers. A pcrmafrost and ground iccmay of thc Northern Ycmispherc is being preparedat :d scalc of 1:10,03O,OOU and will be publishedby thc U.S. Gcological Survey.The Norking (;roups consI.iI.ute the he;3rL of t.he"act.inn cen~er". Six groups wcre organized in.July 1088 i3nd une j n 1992, and onc is in 1993. Itis.a plr;>sur,e trr reyurt tha,t thc Working Groupshuvt. bccn cxccedingly active since their inceptionw1t.h flcld symposia on various subjects in Switzerland,Rurui;j, United States, Canada, Sweden,Net.hrrlonllsr FI ancc, and others; and publicationof sympi~sium rvsults as wcll as bibliographiesand preparation o f ;3 mu1 t.i1 ingucrl glossary.Thc importance of permafrost research irl sciencea r i d engineering is being Inore widely apyrt"Clatvd because of 1nt.ernat.ional Confcrcnces onI'ermafrost. sut..h :js this one, as well as activitiesof our liarking Croups, and TPA in general.Past climat PK ~ . w r bc ~ bcttcr understood hy aknowledge o f [tic history of frozen ground, it.s'forrnatio:~ ;311rl degradation over geologic time; abetter undcrstonding of permafrost will permitmore suc.cessfu1 construction proceduces in polarand mount.ain areas; even a better understandingof t.he environmcnt of some of the or-her planetswill he pos.;ible btAcausc of our current andfutllre wurk with frozen ground.TII (;I osing, I believe the ful.ure Tor. pcrmafrostresetjrch looks bright. Rest wishes and success,espccially tn he young scientists and engirleersof the world, for- ~:ontinued work in the st.urly offrozen grou~~rl.Ill1 QIHENC - Mr. Chairman, larliea and gentlcment,L t is t~ grvat honour for tile t.o be w i t h you atthe Openiilg St.ssLon of this conferer1c.e. On behalf1)l t l l c Chinese Academy o f Scicncc, 1 would like'to wijrmly welcome all of you t.o thc Sixth Lntcrnat11rnaI Confcrcncc on PermafrosL and tu exprcssour thanks tu the <strong>International</strong> PermdFrosLAssociatiorl lor guirl~rncc and support of thisCunfcrcnce.it is the first time that this <strong>Conference</strong> ishcld in China sincf its first mceting in 1963.T helieve I~;+I. !.his Cor~leten~e 11111st he veryimportant for Chinvsc gcocryologists in order 110stimulotc intcrnational cxchongcs irnrl couperalion,to ririse thc yosition of Chin;+ i n !.he world andto tuthsr develop the permafrost rcscarch inC h i [I n .Since lOClO's, the Chincsc scicntists hovf: bewdoing deep RII~ wide investigations on scasonally;+nd pernlanently frozen ground, especially on thewidely-distributed, high altitude permafrost.Meanwhile, they have made great achievements andcontrihutions to the research on frozen soilphysics, mechanics and engineering propertiesand to the design and construction of engineeringworks in cold regions. With further developmentof t.he Western and Northeast China, the study onfronen soil must play more and mure importantrulc in the economic development and relatedconst ruc.tion.Although we have gained great progress in theresearch on permafrost, there are still somedisparities compared with developed countries.A t present, the study on permafrost has beenclosely related to the glohal climate change andecological and environmental conditions. We arehappy to enhance scientific exchange and c-ooperationwith the scientists and engineers fromvariuus countries and to accelerate the developmentof permafrost science.Finally, please accept my best wishes for thesuccess of this grand conference. Thank you foryour attention.TU GUANGZHl - Ladies and gentlgment, friends andcomradcs, It is my great pleasure to extend toyou a warm wclcume on behalf of the Earth Scier1c.e~Division of the Chinese Academy uf Sciences.You as scifntists and engineers have gatheredhere from all over the world to attend thc Sixth<strong>International</strong> <strong>Conference</strong> on Permafrost.Among the various disciplines of earfh sciences,geocryology on the study of permafrcst isa relatively new one, but it has made rapidprogress in recent years. This could be judgedby thc number of papers presented at this conferencewhich cover a wide area of research. Acasual gl.ance at .these papers would reveal thestrong interest. fowards geocryology from otherbranches of earth sciences.'As an economic geolog.ist who is engaged in thestudies of gold rleposits at the present 'time, 1have found that., as is t'te case in Siberia, Canadaand Alaska, the placer gold deposits in China arepreferentially located in regions of permafrost.T 1 seems probable that gc'ld accumulates Lo formplacer mainly by chemical or biochemical means.This serves a good example of c.luse ties betwccngeocryology on the one hand and ore explorationand geochemistry on the other.China is a t.hird world country. We need athorough development in science and technology.China could uffer a good oppurtunity and a soundbackground for almost all branches of carthsciences. W e welcome earth scientists from a13

over thc world to IIOIIIC 1.c Chinv to exchungeopinions, scientific achievements an? ideas wi.thus. I do hope our conference would prove to br~ Bsuccessful one and you will enjoy your stay HLthis conference and in Chinu as wel.1. Thank you.SHI YAPENG - Ladies and gentlemen, anrl honuureriguests, On behalf of the Chinese Society ofGlaciology anrl Geocryology and the Lanzhou lnstituteof Glaciology and Geocryology (LIGG), ChienseAcademy of Sciences, 1 would like to welcome youtu Rcijing and the Sixth <strong>International</strong> Confererl(eon Permafrost.1 am especially delighted Lhat this confcrenc.cis being held for the first time in a developingcountry in Asia.Since 1978 end t.he Third <strong>International</strong> Conferenccon Permafrost in Edmonton, Canada, theChinese permafrost community h8.u I-~een activclyinvalved with the various Tnternatiunal Permafrost<strong>Conference</strong>s and very interested in the adva~lcesand developments in geocryology all over thcworld.As you all know, ladics and gentlement, Chinahas q long hisLory, but its permafrost researchis relatively recent. However China is the thirdlargest permafrost country in the world, with 2millian km' of the teritory underlain by permafrost.The inceptjon and developments of permafrostscience in China was clusely associatedwith exchanges of scholars with forcign countries.For example, some of the pioneer permafrost scientistsin China were trained S n the formerSoviet Union. Nowdays, with the increase ofinternational colleboration, Chinese graduatcs~udents anrl visiting scholars arc sent abroadto learn more about advanced sciencc and technology.These exchange programmes have, in one wayor another, assisted China in keeping pace withthe state-of-the-art of the international permafrostand its related sciences. Here, it is,therefore, mort? a.ppropriate them ever to expressmy sincere thanks and appreciation to the hostingcountries and organizations (institut.es anduniversities), and supervisors of these Chinesescientists.Permafrost provides opportunities and challengesfor both engineers and scientists specializingin cold reRions to cope with variousconstruction and geotechnical problems and toreveal the mystery of many periglacia% landformsand processes. We hope, that with the increaseof international collaboration and exchanges ofideas and data, more and more features andmysteries in geocryology will be uncovered.I 'd also like to say hnw proud T ;im of thenumber of yartlcipants from thc LIGG. Those ofyou who w'ill be part.icipant.ing i n vne of thetwo field trips will have the occassion,to visit1,atizhou and see € i rst h;1nd the research t-hat. i songoing there.T am also very p1east.d to see so many familiarraces avd have so many expert.s from a1 1 ovvr tha22 countries of the world gathering here in Beijingto exchange ideas and discuss researchresults. 1.f the number of participants is anyindicator, we should expect a very rewarding andsuccessful conference. Ladies and gentlemen,onceagain, welcome to China, welcome to Reijing, andwt'.Icome to> the Sixth <strong>International</strong> Confcrencc onPermRTrost.C. WILLlAM LOVELL - Thank you, Mr. Chairman. Onbehalf or t.he 11.S. Committee tor the IPA, the1I.S. National Academy of Scierrc-es, and the manyscientific and engineering specialists for permafrostirr the United States, T bring greetings tothe at-tendances of t.he Six1.h TnLerrlational Cor]-fercnce on Permafrost.I had the pleasure of serving on the OrkanizingCommittee of the vcry fast ICOP he1.d in 1963at Purdue University in Lafayette, Tndiana, USA.Therefore T can appreciate, in at least a smallway, the enormous efforts of the OrganizingCommittee for the Sixth ICOP, led so ably byCheng Guodong. The contributions of the HonoraryCommittee, the Consultative Committ'ee and thegovernmental and nongovfrnmental sponsors werealso essential.Some 52 U.S. scientists and engineers havetravelled to Reijing tu attend the 6th TCOP andto delivcr 29 papers. I wuuld like to givespecial recognitive to may 10 c-odelegates on theU.S. National Academy of Sciences Delegation.They are: Dr. Rernard Hallet, Deputy Chiefr)degat.e, Dr. Roger G. Barry, Mr. George Gryc,Mr. Rupert (:. Tart, Jr., Dr. Chien-Lu Ping,Dr. John P. Zatling, Dr. A.H. Lachenbruch, Dr.Jerry Brown, Dr. Troy L. P&w&, Ms. Sally A.Shoop.We all. anticipate with great pleasure thetechnical, cultural and social associations thatw i l l result from this conference. And we greatlyappreciate the many fjne efforts of our Chlnesehosts whi.ch make all of this possible. Thank you.HUGH FRENCH - Mr. Chairman, Mr. President, HuQihene, Tu GuanRzhi, ladies and ~entlemen, Onbehalf of the National Research Eouncil ofCanada, and the Canadian permafrost community,968

1 bring greetings from Canada, I bring greetingsnot only from my colleagues here with me in Rel-,jing, but also from numerous colleagues who wereunable to attend. Our permafrost community coversa widc range of permafrost science and engineeringinterests. They exist at the academic,governmentaland geotechnical (commercial) levels.Canada, like China, is a vast country. UnlikeChina, it only has a population of less than 30millian. But, like China, it possess significantand extensive areas underlain by permafrost. Assuch, we have much in commun and much to learnfrom each other as we continue to develop ourrespective countries, both materially and c.1-turally.Canada was the bust of the 3rd Tnternational<strong>Conference</strong> on Permafrost, held in Edmonton in1978. There was a small delegation from China tothat conference - approximately 10 persons.Today, as we can see, Chinese permafrost sciencehas grown and prospered, and it is impressive.Equally impressive is the amount of organizationthat must go into running a conference suchas this, planning the technical program, arrangingthe field excursion and local programs, theaccommodation end reglstratlon, and the banquetsand reception is a daunting task. I recognizethis arld thank you, Mr. Cheng, and your collea-'gues for the opportunity t3 benefit from thisconference.Therefore, on behalf of all Canadians present,and numerous others who are unable to be here,I extend my greetings and offer you my bestwishes for a most successful and productiveconference in an atmosphere of cordlalily andfriendship.R.M. KAMENSKY - Mr. Chairman, ladies and gentlemen, dearfriends and c.olleagues, On behalf of the Russian delegationI greet you cordially at the VI <strong>International</strong><strong>Conference</strong> on Permafrost. We are gathered here in Reijingto extend our knowledge of permafrost gained during thepast five years and discuss the strategy of furtherresearch among international community of permafrostresearchers. I am happy to say that during the lastyears international cooperation has strengthened instudyi.ng permafrost. For example, the Permafrost Instituteof the Russia11 Academy Sciences of is conductingJoint work with the Japanese scientists. We work inclose cooperation with Chinese scientists. Thus, a jointRussian-Chinese expedition worked on the Tien-Shan tostudy alpine permafrost. It resulted in joint publicationprinted at the Permafrost Institute. We have brought thecopies to Beijing and they are available for purchase.We are conducting research in cooperation with twoinstitutes in Kharbin, China, in the field of permafrostengineering. In Sept. 1993 an international workshop on!'rotection of Engineering Structures from Frost Heavewill be organized jointly with them. We hope that the<strong>Conference</strong> in Beijing will be successful. We are gratefulto the Organizing Committee, our colleagues in the LanzhouInstitute of Glaciology and Geocryology for inviting11s to the <strong>Conference</strong> and for warm hospitality. Thankyou.JERRY BROWN - It is an honor and a pleasure tomeet you here today and to have served as anoverseas member of the Organizing Committee. Onbehalf of the many countries and individualsparticipating in the conference, I extend ourcollective appreciation to the Chinese OrganlzingCommittee and sponsors for hosting and organizingthe VICOP. In addition to those attending theconference in Beijing, there are many Individualswho assisted us, but were unable to attend.%These include authors and reviewers of papers.In my role as Chairman of the IPA Editoral Committee,1 wish to extend special appreciation tothe conference General-Secretary, Professor ZhuYuanlin and his colleagues for organizing theconference publications. The preconference proceedingsvolume is a major accomplishment and itwill convey the results of the conference to thescientific and engineering commanities throughoutthe world.Finally, Secretary General Robin Brett of the<strong>International</strong> Union of geological Sciences, theparent organization of the IPA, has extended hisbest wishes to Professor Cheng Guodong for asuccessful conference. I also wish the conferencegreat success and thank the organizirs andsponsors in their contributions to the advancementof permafrost science and engineering.Thank you.ZBANG HENGXUAN - Ladies and Gentlemen, It is agreat event in the sci.entific world of the coldregion countries that the remarkable 6th <strong>International</strong><strong>Conference</strong> on-Permafrost is opening inBeijing, the capital city of China. On this,occasion and on behalf of the Cold Region Developmentinstitute of China, I would like to givesincere congratulations to the conference, andwarm welcome to the scientists from differentparts of the world.With the advance of cold region development,the study on frozen earth has become more andmore important. The <strong>International</strong> <strong>Conference</strong> on964

~ of'Permafrost,which is held every five years, hasprovided the scientists and engineers with anopportunity for regular contacts and exchangesof experience, to discuss the curr'ent status andthe future of research on frozen ground. Thisindicates that the research and development offrozen ground has opened a new road for themankind to make use of frozen ground and overcomethe damage caused by the frozen earth. Ibelieve that the 6th <strong>International</strong> <strong>Conference</strong> onPermafrost will record a new chapter in theworld history of frozen ground research..The Cold Region Development Institute ofChina is an academic organization which consistsexperts, scholars and leading officials. Itaims at the research work of cold region developmentand the promotion of economic and technicalcooperation will greatly promote the cooperationwill greatly promote the cooperative developmentof the international r-ld regions. Amajor event will he held in ,anuary, 1994 inHsrbin to celebrate the 10th anniversary of theIce and Snow Festival and the 20th anniversaryof Ice-Lzntern Snow. At the same time, we w i l lhold the <strong>International</strong> Research <strong>Conference</strong> onthe Development of Tce and Snow Culture. Theconference will not only activate and enrich theartistic and cultural life of the people in coldregions, but also promote the regional developmentas well as tourism, trade, economic andtechnical exchange and coo?eration and friendlyrelations between different parts of the world.Experts and scholars are welcome to come to themeeting.During the course when the world economy isbecoming more and more internationalized, modernizedand divided into groups, the developmentand cooperation between international regionsare given more anti more attention by many countriesand regions. The development of the countriesand regions in the cold area is constrainedby adverse natural environment, which set manynegative conditions for their economic development.We must combine different forces, toconvert the adverse conditions into favorableones. We will do our best to create a comfortableenvironment and land for the people to live andwork. We wish the conference a success. We hopeyou enjoy your stay in Beijing. Thank you.ZHU YIJANLIN - Dear Chairmen, ladies and gentlemeniT am very glad to be with you here attendingthe Sixth <strong>International</strong> <strong>Conference</strong> on Permafrost.On behalf of the State Key Laboratory of FrozenSoil Engineering, the Lanzhou Institute ofGlaciology and Geocryology, Chinese Academy ofSciences, I would like to express a warm welcometo all of you for attending this conference.Our State Key Lahoratory of Frozen Soil Engineeringis a new and modern laboratory equippedwith various kinds of test equipment and devicesand living accommodations. It is one of the bestcold regions science laboratories in the world,and is open to the world.Scientists and engineers, especially youngpromising rcsearchera are encouraged to apply toour Laboratory Science Foundation and do researchwork at our Laboratory. We would like to inviteYOU to visit our Aaboratory after this conference,I hope this conference will turn out to be asuccessful conference. And I wish all of youhave a good time during the conference, goodtime as Reijing and gopd time in China. Thankyou for your attention.970

CLOSING PLENARY SESSIONFriday, July 9, '1993CHENG (:IJODO!JC, - T,adies and gentlemen. Wclcomf tothe final formal scssion of the Sixth <strong>International</strong><strong>Conference</strong> on Permafrost. We have bcenmecting here in Beijing for the past five days.to freely exchange ideas and knowledge, discussjoint studies, develop friendship. Now wegathered for a few ilems of business. T wouldlike to thank all of you, on hehalf of theChinese Organizing Committee for attending the<strong>Conference</strong>. I am sure you all agree that theconference has been extremely beneficial 6o allof you.Now 7 would first like to call on ProfcssorAlbert Pissart to read the necrology.!de will hear now from Professor 'Troy P&w&.Under his very capable leadship and yuirlarlce,the TPA and continuing development of permafrostscience and enginecring have heen internationallyassured. Professor PbwB was giver) special recognitionby the University of Alaska in lY91, whenhe receive an honorary degrec of Doctor ofScience. Such an award is the highest tribute aUniversity can convcy to a distinguished individual,and i n his case recognized his major contributionto the understanding of Alaska, geologyand his internatioual lcadership in permafrostresearch. Jn addition to the honors bestuwcd onhim by the I:nivers-ity, the State of Alaska'slcgislature gave him an official cert-ifjcaterccognizing his year's of dedication to theDniversity of Alaska and its st.udents, as wellas his nllmrrous accomplishmer1ts associated withpermafrust rcscarch in Alaska and world wide.It's my pleasure, on hehalf of the IPA and t.heChinese Organizing Committee to present him agift in honor of Profcssor Pew& for tlis tremcndouscontributions to the IPA and pcrmafrostresearch.Next, 1 would like to call on Professor RossMackay. As a Secretary General he has made tremendousr-ontribuLlorrs to the TPA. Besides, ht.remains our mos't productive permafrost scientist,and spent more timc! in the field since hisretiryment than many w i l l in their entire lives.The award of Lhe 1991 Logan Medal signi firs thesubstance and leadership of his research. !lismost recent work on thermal contraction c.rac.kingprovides a remarkable synthesis of detailed' field obscrvations and the theory of crack propagationin solids. It is my honor on behalf ofthe Chinese Organizing Committee to presentProfessor Mackay a gift in honor of his lifetime dedication to permafrost research and tothe IPA and with our sincere good wishcs.Let me try to explain the meaning of thc gift.,This is a Chinese character which means longevity.Tn China, crane also symbolizes longevi t.y,and the old man is R long-1 i ved man. So we wishPruf. Mackay a long, long life.We w i l l now have a fcw closing remarks fromthe leaders of the delegations, we will hearfirst from Dr. Lovell, representing the IISDelegation.Next., lct.'s call on l'rofessor French, representingthe Canadian Dclcgation.Out ncxt speaker is Dr. Kamensky, representingthe Russ i.an Del cgatiun.I would like now call on Acadcmician Shi Ya-Feng, representing Chinese Delegation.Now, we w i l l hear from members of the new Il'AYxecutive Committee. Firstly, I would like tr)take this opportunity to express my he:

turn sc.r ve the r~eeds of ail pcoplc. 1,lr s~ientlsts:and er~y,ineers must join our mirlds and hands toexl~lr~re the sccxcta o f scie~ce and opcn newrvalms oi I-ivili~ation tor thc common progressof rnarlk inil. 'The IPA shuulrl 1)romot.e this kinrl orunion in the fteld of Gcocrynlugy. To reach thisgoal we sti II h.ive a long way tu xu, and arcfaceil with numer~o~~s challenges, which aretcchnulugic, scientific nrtd social in nature.Iiut wt! w ~ l do l our br.st 1.0 meet the challenges,to turthcr ;+I 1 research in yervlnrrost.Now T would like tu !.all. on I'rofcssor HughT:r.er~

1encouraged our Russian colleagues to considerthe establishment of R data subccnter' forpermafrost and ground ice. Wc heard plans forthe 1998 conference in Carrada and encouragedour European collecrgues to begin planniug forthe eighth conference in 2003 i n Europe. Overthe next five years I will be in cont,act withmany of you and hope to learn of you activitiesfor reporting in Frozen Ground. For t.hose ofyou going on the post-conference field tripsI wish you success and for those of you retu'rninghome, have a safe journey. Thank you and Ihope to see you all in 1998.' 973

FIELD TRIP A-1 FROM LZUJBHOU TO LHASA(JULY 12 - 22, 1993)Following the VIth <strong>International</strong> <strong>Conference</strong>Permafrost in Beijing, two technical field tripsdeparted for Lanzhou and Urumqi. About 20 membersof the confeFence and ten accompanying Chineseguides and support personnel took part in thepost-conference field trip to the Qinghai-Tibet-Plateau, the "Roof of the World". During the sixdaybus trip and over 2000 km from Lanzhou toLhasa (cp. Figure 1), the field trip permittedthe participants to observe selected perigladial,glacial and permafrost sites on the plateau.During this short period we gained an impressionof the characteristic permafrbst and permafrostrelated phenomena, of the distinctivegeoecologiCal and biological zonation of thisregion as well as of, the local people, theirliving conditions in and adaptions to this uniqueenvironment.JULY LO AND 11; LANZHOULocal sightseeing and city excursionVisit to the State Key Laboratory of FrozenSoil Engineering, Lanzhou Institute of Glaciologyand Geocryology (Figure 2).An evening banquet prepared by the staff ofLIGG (Figure 3) highlightened the visit.JULY 12: LANZHOU - OXNGHAI LAKE VIA XININGFigure 1: Route mapThe route led through an impressive loesslandscapewhich is characterized by linear downcuttings.Xiqing City (2275 m a.s.1.) with about570.000 people is located in the northeasternpart of the Qinghai-Tibet-Plateau. The Qinghai-Tibet-Highway begins at Xining and ends after adistance of 2000 km in Lhasa.There was a visit and photo stop at the TaerTemple about 30 km southwest of Xining. TheGelupa monastary is one of the most importantLama monastaries in modern China.After having passed an agriculturaly dominatedlandscape of extended fields of rape and corn andpastures we arrived at the hotel at the QinqhaiLake (3194 m), China's largest salt water lake.974

Sand dunes at Xidatan on the.valley bottom forma large chain of active crescent sand dunes of 'local Quaternary deposits. In the 1970s threedunes moved over the old highway at an elevationof 4350 m a.s.1..Overnight in Nachitai Military Depot at3550 m a,s.l..JULY 17: NACHITAI TO WD%O~,IANGAfter crossing*the Kunlun Mountains Pass(4776 m) in the mid-section of the Eastern KunlunMountains the open plateau was reached.Near by the pass a large open-system pingo whichdeveloped during the Pleistocene wag visited(Figure 7). Recent climatic conditions in thearea: mean annual air temperature is below - 5'C,yearly precipitation is at least 280 mm and thewinterfrost lasts about 7-8 months.The descGnding to 4500 m a,s.l., which is thealtitude of most parts of the plateau, the regionis characterized by smooth depressions. Damage tothe road is mainly caused by the melting ofground ice.Along the Qingshui River permafrost thicknessis 10-50 m. Numerous thermokarst lakes of varyingsize and form, the largest being several squarekilometers, and pingos are present.The Wudaoliang Basin has permafrost thicknessof 30-60 m.The Hon Xi1 Permafrost Observation Site ismainly used to study the effects of vegetation onthe ground temperature regime and the permafrostformation.Overnight at the Wudaoliang Military Depot(4700 m a.s.1.).JULY 16: GOLMUD TO NACHITAXThe Qinghai-Tibet Highway crosses the KunlunMountains, the high plains around the YangtseRiver, the Tanggula Range and the plateau basinregion of northern Tibet. Permafrost underlies800 km along the excursion area. On the firststage of the travel, the wide valleys of GolmudRiver and its tributaries, large dilluvial fansand prominent moraines in the river valleys areobserved.Between 4150 and 4200 m a.s.l., in the uppecpart of Golmud River Valley and Xidatan Valley, afault block valley 150 km from Golmud, permafrostis discontinuous with a mean annual groundtemperatures of 0.1 to 1.o'c.The northern lower limit of continuouspermafrost was reached at about 4350 m a.s.1..Permafrost is continuous in a distance of 550 kmfrom 4350 m on the north slope of the KunlunMountains southwards to Amdo on the southernslopes of the Tanggula Mountains.From Xidatan Valley two moraines of the lastglaciation are seen 4400 at m and 4300 m on thenorth-facing slopes of East Kunlun Nountains. Anouter end moraine with an extension of about10 km is developed along the south side of thevalley. From the highway can be seen various18: WUDAOLIANG m,,TANGGULA MITaTTARY DEPOTperiglacial phenomena such as rock glaciers above4900 m a.s.1. to the northwest of the KunlunThe mean altitude from Wudaoliang towards thePass, detritus slopes, solifluction slopes at ,Tanggula Mountains is slightly above 4700 m.4520 m a.s.1, and pingos.Photo stop at sand erosion slumpings, a specialA ground temperature observation field of the kind of thermoerosion in the wind blown sandLanzhou Institute of Glaciology and Geocryology deposition area of the plateau. The slumpingwas visited.976

consists of loose paleo-windblown deposits suchas fine-grained sand, silt and clay. The sparse -vegetation consists of traces of pastures. Theice content of the permafrost in these layers wasestimated to be 30%, a value questioned by someof the participants as being too low.Visit to the Fenghuoshan Observation Station(Figure E ) , where the most intensive permafroststudies of the Qinghai-Tibet-Plateau are carriedout. It is a major test facility and was used bythe Northwest Branch of the China Academy ofRailway Science for testing embankments and pilesin anticipation of a railroad across the Plateau.The station also has ongoing long termtemperature records.After crossing Tanggula Pass, which representsthe border between Qinghai Province and Tibet,the North Tibetan Plateau was entered. It is opento the south-west monsoon which causes anincrease in the annual temperature andprecipitation compared with the northern Qinghai-Tibet Plateau. Due to desqending altitude thecontinuous permafrost is gradually replaced bydiscontinuoup, sporadic and island permafrost.The limit between contipuous permafrost andislano permafrost runs near highway milestone116, north it is continuous and south it occursas island permafrost.overnight in Nagqu (4507 m a.s,l.) an importantcity in the eastern part of the plateau basinregion of Northern Tibet.JULY 20 : NAGQU TO LHASAVisit to an explosive seasonal frost mound inthe intersection zone between the NW strikingcompression torsion fault and the NE strikingtension torsion fault in front of Wuli Mountains.Photostop at the hot springs in theintersection of two faults in the Buqu RiverBasin at the north slope of Tanggula Mountains.Here in the area of discontinuous permafrostgeothermal phenomena like hot springs as well asthe presence of river and lake beds cause largethaw areas. Buqu River Valley is a prominentexample for such a thaw area.Overnight at the Tanggula Military Depot(5060 m a*s*l.).&&.X19: TANGGULA TO NAGQUCrossing of the Tanggula Mountains Pass. Withan altitude of 5231 m a.s.1. the highest point ofthe Qinghai-Tibet highway was reached (Figure 9).Here the end moraines of the last glaciationcould be seen near Basico Lake at the east of thepass.The route led along g4aciated mountain ridges,through a shrub and grassland vegetation, foreststeppes and a densely populated, rich cultivatedlandscape whiCh appgargd with descendingaltitude.Photostap at hot springs of the Gulu geyserarea near by a Tibetan settlement in front of theNianqing Tanggula Mountains. The group received awarm welcome from the local people (Figure 10).Visit to the Ygngbajing hot geothermal area andthe terrestrial steam generating electric stationabout 90 km west of Lhasa.Descent throuqh the V-shaped Londui Valley toLhasa. At 4100 m' a.s.1. the valley widened andwas occupied by cultivated fields of rape, andfrom 3950 m a.s.1. on with barley fieldsdominating the landscape. Closer to Lhasa from3700 m on downwards, they were replaced by cornf ields.Arrived in Lhasa (3658 m a.s.l.), the capitaland religious and cultural centre of Tibet, at 6o'clock p.m.' 977 '

~ areaFigure 10: Tibetans welcome th excursion groupwhile visiting hot springs the in Gulu geyser(by E. King)JULY 21 AND 22 LHASA:City excursions to temples, monastaries, localmarkets etc..JULY 23: FLIGHT FROM LHASA TO CHENGDUMost of the group returned to Beijing on July24. The excursion and the special circumstancesunder which it took place will certainly beremembered far a long time by all theparticipants.WFERENCEs :Liu Tungsheng & Yuan Baoyin (1991): QuaternaryGlaciations and Periglaciations in the Qinghai-Xlzhang (Tibetan) Plateau. Excursion guidebook. <strong>International</strong> Union' for QuaternaryResearch, XI11 <strong>International</strong> Congress, Beijinq,China, 1991.Guo Oongxin h Zhao Xiufeng, eds (1993): A Guideto the Permafrost and Environment of theQinghai-Xizang Plateau. Excursion guide book,VIth <strong>International</strong> <strong>Conference</strong> on Permafrost,Beijirrg, China, 1993.ACKNOWLEDGEMENTS:Thanks are given to Lotens Xing for contributionsand Jerry Brown for the improvement of theenglish manuscript.hElisabeth SchmittIPA, Chair Editor ial comm ittee

Tian Shan Field Trip, A-2(July 11-18, 1993)JULY -EXCURSIONIN URUMQIUrumqi (1.2 million people of severalnationalities) is the capital of Xinjiang Uygur ,TUInY. 14: URUMQI- TIANCHI LAKE - U mAutonomous Region. It is located on an alluvialfan at the northern foot of Tian Mountains.Left Urumqi heading to Tianchi (Heavenly) LakeAccording to its latitudinal and longitudinalabout 60 km east of Urumqi in the San Gong Riverlocation (about 43" N and 86" E) it has a dry, Valley on the north slope of Bogda Shan. Thecontinental climate. Trees and irrigation furrows route passed steppe and pasture in the foothillfollow all roads. The irrigation water source is zone northeast of Urumqi and irrigation systemsthe Tian Shan. The city is characterized by dense on arable land and forests in the Fukong area.traffic and large industry causes a remarkable The transition from desert to steppe and forestair pollution.belt (conifers) took place in a horizontalVisit to the Xinjiang Museum of Geology and distance of about 30 km.Mineral Products.The visit to Tianchi Lake (1910 m a.s.l) in amountainous setting provided an opportunity for*gQxExcursion to Turpan in the middle part of'theXinjiang Uygur Autonomous Region. As anintermontane basin in the Tian Shan it issurrounded by Mt. Bogdas in the north, Mt.Karawuquntag in the west, Mt. Jueluotag in thesouth and Kumtar desert in the southeast. Itslowest point is 155 m m.s.l..The route was characterized by the transitionfrom steppe to desert and allowed a view to thesnowy Bogda Shan (5445 m a.s.1,) during theentire day.Photostop at the Lake of Chaiwo Pu where saltexploitation and irrigated fields could beobserved. In the upper Turpan basin theconsequences of recent flooding (roaddestruction) across the extensive arid plain eastof Urumqi (appaerently due to the lack ofvegetation) after night rainfall in thesurrounding mountains could be observed. On theride through the stony desert (hammada) greenoasis Turpan and an oil exploitation field werepassed.Visit to ancient Bizalik Thousand Buddha Caveson the cliff of Mutuo Valley, 48 km notheast ofTurpan City. The sandy soil is covered withventifacts.Visit to the ancient city of Gaochang nearbythe caves. It was built in the 1st century,destroyed in the 14th century and has been key apoint along the ancient silk road.Visit to the grape Valley, which is an 8-kmlong and half-kilometer wide oasis rich in water.The 400 hectares of cultivated land include 220hectares of wine-growing.Visit to the ancient, but still active Korezirrigation system and the Wudaolin forest beltwhich is fed by this irrigation system of wells,underground-and surface-channels.speculation on alternative mechanisms offormation of the lake (ancient bedrock landslideor moraines). The bluff at the outlet showedunstratified coarse and fine deposits and layeredfluvial sediments. The sunny morning allowed abeautiful view to the snowy summit of Bogda Shanwhile in the afternoon it became cloudy.JULY. 15: URUMOI - URUMQI RIVE-CISTATION - CANGF ANGGOU SECTIONOLOGICALDrive to Tian Shan Glaciological Station withseveral stops en route to inspect and discussperiglacial features. Late Pleistocene alluvialfan sediments of Urumqi River with an alternatebedding of coarse, dark fluvial and Pine, lightsediments without clear signs of stratificationwere passed. The question of the origin of thefine sediments (fluvial or eolian) and thegenesis of the fan were up for discussion. Nosigns of periglacial features could be observed.About 10 km in front of the mountain rangeUrumqi River terraces could be seen. Two mainterraces were clearly distinguished in thedistance. The height of these terraces wereestimated to be about 50 m. The edges of theloess covered terraces (Figure 1) are maturelydissected due to water erosion.In the gorge of the upper reach of the UrumqiRiver (Daxigou River) the Yingxiongqiaohydrolgical and climatological station is locatedat 1830 m a.s.1.. The observations include waterlevel, temperature, discharge and suspendedsediments of the water, precipitation, humidity,air temperature as well as river ice conditions.At the upper end of the fault basin of Houxia(Rear Gorge) the base camp of Tian ShanGlaciological Station is located (2130 m a.s.1.).' 979 '

JULY 16: GLAUOWGI CAL STATION - SHENGLI WABAN- GTICIOMGICAL STATIONThe highly interesting day was devoted toalpine settings and glaciers in the Urumqi River Rock Glacier No. 4 (Figure 3)is located on ahead waters. The glaciological station summer north facing slope In front of Glacier No. 3 andcamp (Figure 2). Glacier No. 1, the precipitous cut by the old highway. Since 1960 mean advanceroad past Wangfeng Highway Maintenance and of RE4 is 18 cm/year, The rock glacier is notubiquitous glacial and periglacial landforms tongue shaped. Patterned ground between RG4 andprovoked lively discussions among members of the the moraine of Glacier No. 3 could be observed.group.The gravel road along the upper Urumqi River(Daxigou River) climbed a second gorge and a V-shaped valley and passed Uigur camps. No signs oflate-Pleistocene glaciation like U-shaped valleysand Nunataks could be observed. Wood and treesdisappeared and an alpine setting with debrisslopes characterized the landscape. On northfacing slopes above 2900 m high mountainpermafrost is present.Permafrost areas were reached at about3000 m a.s.1. near Wangfeng Highway MaintenanceSquad. Here, clear forms of glaciation as hangingvalleys and f?d moraines of +he Wangfengformation (C age of 14920 ?: 750 years B.P.)were developped. Under periglacial conditions aseries of gelifluction steps and tongues wereformed .The road to Shengli Daban pass (4020 m a.s.1.)was built in 1960 on a steep, dangerous debrisslope covered with large blocks. Rockfalls areobeserved at many places. Ongoing repairs of theroad are necessary. The elevation o'f thesurrounding mountains is about 4300 m a.s.1..View to retreating Glacier No. 3 on thenorthern slope of Mt. Karawuquauntag. The lateralmoraine contains buried glacier ice. A snowavalanche could be observed on the steep hangingcirqueglacier.The 20 m depth borehole No. 5, drilled ingneiss bedrock in a sharp turn of the road(3900 m a.s.1.) is the highest borehole in theChinese Tian Shan. The mean annual groundtemperature is -4.9-C and permafrost thickness isestimated to be 230 m. The active layer thicknessis 1.75 m in September.Figure 3: Rock Glacier No.4 (3550 m a.s.1.) cutby the old highway980

The glaciological station summer camp (3540 ma.s.1.) is a climatological station with 35 yearsrecord starting in 1958: Borehole No. 4, 18 rn indepth, was drilled in 1990. The depth of zeroannual amplitude is 15 m. Permafrost thickness isestimated to be loo m.The frost mound next tu a small creek(Figure 4 ) with an ice core is probably caused byan injection at about 3500 m a.s.1.. Its heightis about 1.2 m and it is partly eroded due tothermokarst. Several smaller frost mounds arelocated in the area.Walk along the moraines of the Wangfengformation. Road cuts show an unstratified tilland glaciotectonic thrust sheets. The end moraineof the lower limit is locatedabout 2900 ma.s.1..A profile at the Red May Bridge on a terraceabout 20 m above Urumqi River shows glaciofluvialsediments as well as talus and lakedeposits.JULY 17, 1993: GLACIOLOGICAL STATION - GLACIERNO. 1 - URUMQIThe 3.1 km long gravel road from the summercamp to the glacier was built in 1981 mainly byhand and in view of permafrost characteristics.Stone stripes are formed beside the road. Walkdown the moraine onto the south branch of theglacier.Glacier No. 1 (Figure 6) is a small glacierwith two branches in north facing cirques. It isretreating, south and north branch are stillconnected by a very short section but willprobably separate in a few years. Mass balancemeasurements are carried out every four to sixweeks since 1959. High ablation and accumulationduring summer. During the visit a fresh and wetsnow cover on the ice could be observed.Figure 6 : South and north branch of GlacierNo. 1. in the background and the active RockGlacier No. 2 in the foreground.981

An empty cirque, sorted circles, stripes andprotalus ramparts could be observed in thesurroundings of the glacier. The thickness of theactive layer is about 1.5 m. No solifluctionmeasurements take place. High mountain vegetationcharacterizes the landscape.The active Rock Glacier No. 2 is located nextto the glaciological station. The rock glacier isnot tongue shaped but. very similar to the form ofprotalus ramparts. It is situated below a steepnorth facing debri slope, Its active zone is inthe middle part of the front slope.Return to Urumqi.JULY,10 : FLIGHT FROM URUMOI TO BE,IJING.WERENCESChaohai, L. et al. (1991) : Handbook of the TianShan Glaciological Station, Lanzhou.Euoqing, Q., Shije, L., Huijun, J. & Lin, 2.(1993): Guide book of the field trip A-2 of the<strong>International</strong> <strong>Conference</strong> on Permafrost.Lanzhou.ACKNOWLEDGEMENTS: Thanks are given to Jerry Brownfor the improvement of the english manuscript.The report based on field notes and reports byRainer Lehmann, Germany, and Chuck W. Slauqhter,USA. All photographs are taken by R. Lehmann.edited by Elisabeth Schmitt,IPA, Chair Editorial Committee982

INTBRNATIONAL PERMAFROST ASSOCIATIONSTANDING COMMITTEES AND WORKING GROUPSMEMBERSHIP AND PURPOSEThe following Committees and Working Groups wereapproved for five years at the IPA Councilmeeting, July 8, 1993, Beijing. China. WorkingGroups and Committees are expected to report onprogress at Council meetings and regularly inthe Frozen Ground News Bulletin. Working Groupmembership is limited to Chair, Secretary andsix full members; exceptions to this limit aremade by the Council. Ex Offic'io members fromthe IPA Executive Committee and Working GroupChairs may be represented on other WorkingGroups. There may be an unlimited number ofcorresponding members on Working Groups andinterested individuals should apply to theWorking Group Chair.IPA EXECUTIVE COMYITTEECheng Guodong, President (China)H.M. French, Vice President (Canada)N.N. Romanovsky, Vice President (Russia)Jerry Brown, Secretary General (USA)ADVISORY COMMITTEE ON WORKING GROUPSC.W. Lovell, Chair (USA)W. Haeberli (Switzerland)S.E. Grechishchev (Russia)Ex Officio: H.M. French, IPA Executive CommitteeFINANCE COMMITTEEO.J. Ferrians, Jr., Chair (USA)A. Pissart (Belgium)Zhu Yuanlin (China)Ex Officio: J.Brown, IPA Executive CommitteeEDITORIAL COMMITTEEE. Schmitt, Chair (Germany)M.A. Grave (Russia)J.A. Heginbottom (Canada)K. Hall (Southern Africa)Xu Xiaozu (China)Ex Officio: J. Brown, IPA Executive CommitteeWORKING GROUPSMountain PermafrostTerminologyGlobal Change and PermafrostData and InformationPeriglacial Procekses and EnvironmentsCryosolsFoundationsSeasonal Freezing and Thawing of PermafrostAreasWORKING GROUPSMountain PermafrostPurpose: To improve the exchange of informationon, describe the state of knowledge about, andstimulate research activities concerning permafrostat high altitudes and in rugged topography,especially at low latitudes. The objectives ofthe WG over the next five years are to (1)promote application of computer models to predictpermafrost occurrence, (.2)' organize inter -comparisons of results of aodelling and fieldmapping, (3) coordinate long-term monitoringwith regard to warming trends, (4) investigateenergy exchanges in the active layer and (5)improve understanding of permafrost creeplrockglacier formation.W. Haeberli, Chair (Switzerland)F. .Dramis, Secretary (It. ly)S.' Harris (Canada)A. Gorbunov (Kazakhqtan)M.M. Koreisha (Russia)Ex Officio: Cheng Guodong,IPA Executive CommitteeTerminologyPurpose: To develop a set of internationallyaccepted permafrost terms for engineering andscientific use, with language equivalents. Overthe next five years the HG plans to completedevelopment of the multi-language index withthe addition of definitions, complete theEnglish-Russian dictionary of over 2000 termsand incorporate the new Chinese-Russian-Englishglossary into a common index. The ldorking Groupwill disserminate and encourage use of suchterminology.R.O. van Everdingen, Chair (Canada)V. Konishchev, Secretary (Russia)J. Akerman (Sweden)A. Corte (Argentina)F. dramis (Italy)O.J. Ferrians, Jr. (USA)J, Karte (Germany)0. Gregersen (Norway)J,P, Lautridou (France)Qiu Guoqing (China)Ex Officio: N.A. Romanovsky,IPA Executive CommitteeGlobal Change and PermafrostPurpose: To identify the effects and consequencesof global changes in temperature and relatedphenomena upon the nature of permafrost and itsdistribution. An annotated bibliography ofpermafrost and climate change was completed andthe special issue of Permafrost and PeriglacialProcesses on the same topic was published anddistributed at the VICOP. Activities over thenext five years include preparation of aninventory of existing ground temperature sites,standardization of data collection and archivingin cooperation with the WG on data and information,more direct involvement with global modellingactivities, and convening of wr;-kshops onrelated problems. The Working Group is encouragedto interact with other national and international~~projects groups concerned with global change(e.g. IGBP, IPCC).

F.R. Nelson, Chair (USA)A. Taylor, Secretary (Canada)O.A. Anisimov (Russia)M.K. Gavrilova (Russia)T.E. Osterkamp (USA)Ex Officio: Cheng Guodong, IPA ExecutiveCommitteeR.G. Barry, WG Data and InformationW. Saeberli, WG Mountain PermafrostData and InformationPurpose: To improve and standardize and collection,archiving, documentation and disseminationof permafrost data. Activities over the nextfive years include a proposed workshop on DataPriorization 1994, continued efforts to assembleand publish permafrost data under the auspicesof WDC - B (Boulder, develop a carto-bibliographicdata base, and update rhe permafrostbibliographic compilation. The Working groupwill collaborate with the WG on Terminology,Global Change and other national and internationalcommittees and agencies concerned withrelevant data.R.G. Sarry, Chair (USA)J.A. Yeginbottom, Secretary (Canada)J. Akerman (Sweden)M.J. Clark (United Kingdom)Chen Xianzhang (China)E.S. Melnikov (Russia)Ex Officio: F.E.Nelson, WG Global ChangeR.O.van Everdingen, WG TerminologyPeriglacial Processes and EnvironmentsPurpose: (1) To investigate the frequency andmagnitude of periglacial processes, especiallythose occurring within the active layer, (2) toevaluate different methodologies and techniquesfor process measurements, and (3) to predictthe effects of potential climate change onperiglacial environments using contemporarydata and the stratigraphic record. Over thenext five years the WG will conduct severalsymposium, field trips and prepare a handbookon recommended techniques for investigatingperiglacial processes. The WG operates jointlywith the IGU Commission on Frost Action Environmentsand its President J.-P. Lautridou,(France) and Secretary C. Harris, (UnitedKingdom).A,G. Lewkowize, Chair (Canada)C. Harris, Secretary (United Kingdom)J. Akerman (Sweden)Cui Zhijiu (China)R. Hallet (USA)A. Pissart (Relgium)V. Solomatin (Russia)V. Vandenberghe (The Netherlands)Ex Officio: J.P. Lautridou (France), IGUCommission on Frost Action EnvironmentsCryosolst’urilvst.: To dcvt.lup and n~djnt~in close workingrelations between soi 1 and permafrost scientist.&throughout the bipolar regions, and to develol,projects to correlate and/or consolidate thevast amounts of informetion, maps, and data onsoils that are of interest to IPA. Activitiesover the next five include charactcrization ofsoil climates, map the relationship of permafrostand cryosol distribution and participate inseveral field trips and conferences: soilcorrelation in Northwestern North America inlate July 1993; field trip to Yagadan-Kolymaregion in summer 1994, attendance at the <strong>International</strong>Soil Science Society (ISSS) Subcommissionon Frozen Soils in 1994, and organizationof the Second <strong>International</strong> <strong>Conference</strong> onCryopedology in Syktyvkar in 1996. Coordinationand liaison will be proyided with the ISSS andother organizations having common interests.D. Gilichinsky, Chair (Russia)C.L. Ping, Secretary (USA)J. Bockheim (USA)G. Broll (Germany)Wang Ilaoqing (China)9. Jakobsen (Denmark)G. Mazhitova (Russia)C,, Tamocai (Canada)FoundationsPurpose: To collect information on the practiceof foundation engineering in various permafrostregions of the world and to synthesize guidelinesfor effective engineering practice. Activitiesover the next five years include preparation ofconcise state-of-the-art-reports on such topicsas pile foundations and foundation research andconduct seminars and workshops on related topicsincluding a special seminar on foundationfailures. The WG encourages monitoring andreporting of the performance of foundations inpermafrost and works closely with the WG onSeasonal Freezing and Thawing and national andinternational engineering and geotechnicalorganizations and societies including theCanadian Geotechnical Association and theAmerican Society of Civil Engineers.J.W. Rooney, Chair (USA)K, Flaate, Secretary (Norway)R.M. Kamensky (Russia)L. Krustaley (Russia)P.J. Kurfurst (Canada)R.C. Tart, Jr. (USA)Zhu Y uanlin (China)ex 0 f ficio: A. Phukan (USA), WG SeasonalFreezing and ThawingSeasona 1 Freezing and Thawing of Permafrost AreasPurpose To improve the exchanRe of informationon,-describe the state of knowiedge about. andstimulat,e research activities concerning frostaction in soils and measures to protect againstits harmful effects in permafrost areas. The WGwill assist in organizing the next :;ymposium onFrost in Geotechnical EngineeEing ir Lulea,Sweden, in March 1997, and work closely withthe WG foundations and other WG on problems ofcommon interest. The WG interacts with othernational and international groups and crganizations(e.g. <strong>International</strong> Society of SoilMechanics and Foundation Engineering (ISSMFE)and its Technical Committee on Frost (TC 8); theAmerican Society of Civil Engineers (ASCF) andits Technical Council on Cold Regions.Engineering(TCCBE); the <strong>International</strong> Symposi,um for GroundFreezing; the <strong>International</strong> road Federation).. 981

A. Phukan, Chair (USA)E. Ladanyi, Secretary (Canada)M. Fukada (Japan)H.L. Jessberger (Germany)S. Knutsson (Sweden)G.Z. Perlstein (Russia)K. Senneset INorway)E. Slunga (Finland)EX Officio: J.W. Rooney fUSA), WG Foundatior IS

RESOLUTION: INTERNATIONAL PERMAFROST ASSOCIATIONApproved July 8, 1993 at the IPA Council Meeting, Beijing, ChinaWHEREAS the importance of permafrost is reflectedin both international governmental and nongovernmentalreports and science plans (IntergovernmentalPanel on Climate Change (IPCC);IGBP core projects: <strong>International</strong> Global AtmosphericChemistry Project (IGAC); Land.-OceanInteractions in the Coastal Zone Project (LOICZ);Biospheric Aspects of the Hydrological CycleProject (BAHC); and Global Change and TerrestrialEcosystems Project (GCTE);WHEREAS the distribution and properties of permafrostare of increasing interest to thoseconcerned with assessing the influence o fglobal climate change on high latitudes andhigh altitudes; .WHEREAS permafrost I s sensitive to climate andcontains a memory of past climate changes; 'WHEREAS the IPA is concerned with the advancernent,ofknowledge on the formation and degradationof permafrost at regional and global scales:Be it RESOLVED that the IPA, consisting oE 20adhering national bodies, representing manyearth science and engineering disciplines, seeka more active role in the IGBP-cor? programs bycommunicating IPA interests and acLivities torelevant IGBP programs, IPCC assessments, andother programs;FURTHERMORE the IPA notify other national andinternational scientific and engineeringorganizations of its present working groups'plans and activities including the availabilityin early 1994 of the IPA 1:10,000,000 map ofpermafrost and ground ice of the NorthernHemisphere;Finally, be it RXSOLVED that relevant IPAworking groups give particular attention toglobal climate change and prepare status andtrend reports for the Seventh <strong>International</strong><strong>Conference</strong> on Permafrost, to be held in Canadain August 1998.i

Permafrost and Changing ClimbleCF.E. Nelson', A,E. Larhenbrueh2, M.-k. Woo3, E.A. Kostec',T.E. Osterkarnp', M.K. Gavrllora6, Cheng Guodonp'.t'Dcpanment of Geological Sciences, Cornell University, Ithaca, ,NY. US4 14853Chairman, IPA Working Group on Permafrqst and Global Change'U.S. Geological Survey, Menlo Park, CA, USA 940253Depanmetlt of Geography, McMaster University, Hamilton, Ontario, Canada,L8S 4K1'Geographical Institute, University of UtrcchtHeidelbcrglaan 2, 3584 CS, Utrecht, "'he Netherlands'Geophysical Institute, University of Alaska, Fairbanks, Alaska, USA 99775'%ermafrost Institute, Yakutsk, Russia 677018'LanEhou Institute of Glaciology and Geocryology, Lanzhou, P.R. China 730GOOINTRODUCTIONThe focus on climate change and its effects on ecosystems andhuman activities has so intensified in recent years thrt it isnow a central issue in many of the natural and social sciences.Governmental interest and public support for research on thistopic has provided the impetus for highly visible and coordinatednational research initiatives, and intens debates centered onapproptiate responses to thc issue permeate public-policydiscussions (e.g., Levine rt al. 1992, Messner ef al. 1992,aothman and Chapman 1993, Winh 1993). Climate change has evenprovided a tentative, if controversial, basis forintergovernmental mpcration (Houghton el al. 1990, von Mnltke1989, Agarwd and Nasain 1990). Although permafrost has thepatentis1 to be a major element in the fabric of climate-changeresearch, it has to date occupied only a tangential position withrespect to both policy discussions and construction of*This contribution is bwd on an invited lecture by A.H.Lachenbruch and proscntations by the other authors in a plenarysession on "Qlobal Climate Change and Pcrmafrost" at the Sixth<strong>International</strong> <strong>Conference</strong> on Permafrost, Beijing, July 6, 1993.'Permanent addrew Department of Geography, Rutgers University,New Bnmswick, NJ, USA 08903integrative climate-change scenarios. A new resolution, whichappears elsewhere in this volume, puts the <strong>International</strong>Permafrost Association on record as giving permafrostlclimatechangeinteractions high priority (IPA 1993).Permafrost is a temperature condition of the solid Earth; itsexistence depends on the local heat balance in cold regions. Thewidely discussed models for contemporary greenhouse warminggenerally predict that effects will be greatest in these regions(e.g., Budyko and Izrael 1987, Maxwell and Barrie 1989, Roots1989: but see Kahl er 01. 1993, Walsh 1993) and, in general, thatthey will alter the surface heat balance and the temperaturc anddistribution of pcrmafrost (Nelson and Anisimov 1993, Riseboroughand Smith 1993). Relatively rapid changes may occur in theposition of the top of permafrost (depth of summer thaw) and inthe distribution of warm permafrost near its ,southern limit;these changes can impact the dynamics of a broad rangc of surfaceprocesses and may relcasc greenhouse gascs, currently stored in, permafrost, to the atmosphere. Changes in the position of thelower boundary of permafrost will, however, gencrally beunimportant for hundreds or thousands of years, during which timcthe downward propagating thermal wave will preserve a lingeringrecord of the climatic event at depth. In the presence of achanging climate, therefore, permafrost can play at least threeimportant roles: 1) as a rrcorder of climatt change, 2) as anrrgent of environmental changes that affect ecological and humancommunities, and 3) as a facilitator of further climate changc.987 '

IPERMAFROST AS A RECORD KEEPERIn cold continuous permafrost, as in other impermeable carthImatcrials, there is little movement of groundwater and heattransfer is almost exclusively by conduction, a process thatIIIIfollows relatively simple mathematical rulcs.climatic change in theThus, if apast altered mcan annual temperature atthe base of the active layer, thc changc would propagate slowlydownward into the permafrostat a rate that can be calculated.In effect, the ground "remembers" the major evcnts in its surfacetemperature history,today to depthsand careful tcmpwature measurements madeof a few hundred meters can provide informationon the history of local surface temperature during pastcenturies.In high latitudes, bodies of water that du nntfreeze to thebottom in winter generally have mean bottom temperatures near orabove O"C, whereas the mean tempcrature of theadjacent landsurface may be much lower (e.g., -10' to -15'C near the shorcs 11fthc Arctic Ocean today).concerned, thc migrationleavea a dramatic "climatic change"shore is advancing on therctrcating from the land.temperature meawrementsInsofar as the solid earth ISof a lake or ocean shoreline thereforein its wake a warming if theland or coohng if the shoreline isGcothermal methods can be applictl tuin wells nn the submerged Arcticcontinental shelf to interpret the chrnnolngy rlf rapid shorelinetransgression in progress onand of the inundation of the Arctic continentalthe last glaciation. Effectson low-lying permafrost canmuch of the Arctic shnrcline today,shclf followingof predicted global sea-level risebe estimated by thc same methods(Lachenbruch 1957, Lachenbruch et nl. 1982, 1988a, Wang 1993).Additional information about the history of the recentclimatic system can be obtained by measuring the tempcrature,depth, and ice content of existing permafrost and modeling thesurface conditions thatmust have existcd to generate thepermafrost observed. Where ice content is high and permafrost isdeep, as in Prudhoc Bay, Alaska or eastern Siberia, suchI calculations provide climatic information on timescales* 5approaching 10 ycars, the period of glacial cycles (Balohacv etal. 1978, Harrison 1991, Osterkamp and Gosink 1991).In addition to the information on climatethe present thcrmal statehistory contained inof permafrost, much can be learnedabout past surface conditions from the cold-climatefeatures and organic materials preserved in(Carter er 41. 1987, Mackay 1988).ice-wedge polygons, which form from thegeomorphicburied permafrostOf particular intcrest arepercolation of summermeltwater into a network of dcep thermal contraction cracks thatform during wintcr in cold brittle permafrost. Evidence on theseasonal growth and deteriorationof past generations of icewedges reveals changes in past surface conditions whcre thesenetworks of massive ice arcpreserved in the stratigraphic record(Lachenbruch 1966, Mackay and Matthews 1983).Climate Signals in Permafrost Temperatures:From Climate to PermafrostMany important environmental changeschange in permafrost regions dorather in the active layer above it. The permafrostassociated with climatenot occur in permafrost, butgenerallyshares its upper boundary with the base of the active layer(Figurc I), and acts as a "listeningchanges that occur there.and above the activepast" for temperatureHowever complex the thermal changes inlayer may be, the only parameter rccordcd bypermafrost is the temperature changc that makes its way to thebase of the active layer, for practical purpascs,the mcun annuoltemperature change. It is this quantity, the mean annualtemperature at the topof permafrost (Tpo, whose time history iscstimated in "climatic" reconstructions from borehole temperaturemeasurements.Such rcconstructions are bascd on the assumption that, exceptfor the effects of steady heat flow from the earth's interior, achange in mean annual temperatureby a past change in surface tempcrature. Thisheat transfer iswith depth can bt caused onlywill be true whercexclusively by conduction with non-changingproperties, and where effects of sources and sinks, if theyexist, cancel each other over the yearly cycle.conditions are generallyAlthough thcscsatisfied in most cold permafrost, theyclearly do not apply in the overlying active layer, snow, andnir In general, therefore, Tmr (mean annual air tcmperature,Figure 1) will diffcr from 'i',~(mean annual tempcrature of the"solid surface": snow in winter, ground in summer). and Tss willdiffer from Tpf (mean annual tempcrature at the top ofpermafrost, Figure l), whether 01 not the climate is changing.Effects above the solid surface are addressedby the radiationbalancc of climatology, considered briefly below. Between thesolid surface and permafrost, the temperature offset (Tsb- Tpf,Figure 1) is determined by the complex dynamics of thesnowlactive-laycr system. Clearly, thcrc aremany differentFigure 1. Measurement sites for the differently defined surfacetemperatures: Tpf at upper surface of permafrost, Tgs at groundsurface, Tss at the solid surface of the snow pack when it ispresent and ground surface when it is not, and T-ir in a standardabservatory thermometer shelter.

types of surface change, climatic andthe same change inotherwise, that can produceperdrost temperature, and someenvironmentally ~mpartant thermal changes in the active layermight have little effect on 4f. It is imponant, therefore, tomaintain a distinction between changing permafrost surfacetemperature and changing climate.Simple references tocharacterid or defined by an increase in"climatic warming' imply that it can bemean annual airtemperature near the Earth's surface. Figure 2 is a reminder ofthe inadequacy of such a definitionfor predicting environmentaleffects controlled by the active layer, which is mast sensitiveto summer temperature. The three cases illustrated in Figure 2a11 represent a 4 " mean ~ annual warming (from -9' to -9'~).achieved respectively, with warmer summers,much warmer winters with somewhat cooler summers.might have dramatic effectsmight have little or none.warmer wintcrs, orThe firston surface environments, the thirdThe second is more consistent withpredictions from most general circulation models (IPCC WorkingOroup 1990, Mnxwell 1992).These relationships can be addressedin an approximate fashion by the ratio of freezing and thawingdegrce-day sums (Nelson and Outcalt 1987).the warming will generally propagate to permafrast,be detectable for centuries,In all three cases,where it maybut a distinction among the threeoriginal surface conditions will not be possible.4°C Warming Scenarios, Alaskan ArcticNext to the shift in mean air temperature, the mog obviousagent to shift Tpf is winter snow cover, which insulates theground in winter, causing Tpr > h s (the *mow offsct", e.&,Goadrich 1982, Zhang and Ostertramp 1993); a secular increaw insnow cover (8 born fide climatic change) can cause a coIIspicuoUSwarming signal in pcrmafrost. A more subtle dTpf signal from thesnow can occur with no change in snow cover or mean airtemperature if the amplitude of seasonal temperature variationincreases; the same snow cover causes more warming because ofgreater damping in colder winters as the climate becomes morecontinental (Lachcnbmch 1959). This may be the principal causeof the difference in Tpf from -9°C at Barrow on the AlaskanArctic coast to 4°C at Wmiat, 80 km inland. Both have similarmean air temperatures and snow cover, but Umiat has a morecontinental climate.In general, rhe mean annual temperature will not be the sameat different levels between the air/solid interface and the topof permafrost, as explained schematically in Figure 3. Forseasonal effects that make it easier for downward heat transferthrough the snow and active layer than upward transfer out ofthem, Tpf will be greater than 'ha when the two have reachedcquilibrium. Examples are infiltration (with possible lateralflow) of summer meltwater and rain, a unidirectional process thatadvects heat downward, and the "snow offset," which insulates theground tclcctively in winter. These are designated Type I1processes in Figure 3. Examples of Type I processes, which cancause a steady negative offset of permafrost temperatures, arethe change in thc active layer from a good thermal conductor whenit is frozen in winter to a poorer one that inhibits downwardheat flow when thawed in summer. An effect of the same sign isthe upward transport of moisture (and dehydration, reducingconductivity) in the active layer to supply surface evaporationTYPE I TYPE 1Summw WlPrMlFigure 2. Three scenarios for the same warming (from solid todashed curves) of meun ground surface temperature (Tag). Summerconditions that control the environmentally sensitive activeLayer ate quite different for each, us indicated by the ratio ofannual freezing degree days to thawing degree days (see Nelsonnnd Outcalt 1987).Figure 3. Snow and active-layer processes 0.pC I) that causethe periodic seasonal influx of hwt in summer (Qio) to be lessthan the seasonal outflow in winter (Qont) generally contributeto a decrease in mean temperalure with depth. Those that causethe periodic summer inflow to exceed the periodic winter outflow(Type YI) generally contribute to an increase in temperature withdepth. Environmentdy induced changes in any such snow atactive-layer processes are remembered in the permafrosttemperature record as changing "climate."989

and transpiration; it tends to counteract heat gained by downwardconduction in summer. A number of other seasonal processesinvolving non-conductive transport of scnsible and latent heat bymigrating moisture, sometimes involving vaporization,condensation, and ice segtegalion in the active layer have beendescribed (e.g., Nakano and Brown 1972, McGaw rt ~ 1 1978, . Wallet1978, Mackay 1983, Nelson et al. 1%S, Outcalt et ul. 1992).Subtle changes and inreractions among the factors that controlsome of these processes, for example, prolongation of the zerocurtain by non-linear interaction of latent heat of soilfreeze/thaw and snow cover (Goodrich 1982), could cause n changein heat balance within the active layer that would be rememberedby permafrost temperatures as a change in Tpf.,d F O I The permafrost surface temperature,important single parameter for the, integratedenvironmental conditions. Understandingimplications, however, requires a knowledgethermal processes that control it as wellinformation whenever possible. Whetherdirect response to changing climatic parameters,precipitation and snow cover, or to changingconditions, drainage patterns, or theevolving plant communities, they areinformation that we should attempt tofactors, including climate change, landhelp us to understand the dynamics ofTpf, may be the mosthistory ofits full range ofof the physir- ofus supplementarypast changes in Tpf wereincludinggeomorphicheat and water balance ofunique source ofrelate" to environmentaluse, and others that willchanging ecosystems.aa-O T H ~ ,LFigure 4. Borehole temperatures In part b are from sites denotedby solid circles in part a. Shaded region for each curve is therecenk warming anomaly; numbers by CU~VCS denote time in yearsbsforc 1984 for start-up of best-fitting model of lincar warming.Insert in part b summarizes statis!ics for starting date andtotal warming for best-fitting step and linear models (afterLachenbruch and Marshall 1986).

The,Tcmpcmture Si-from Recent Warming EventsFipra 4b &owa a group of typical temperature profiles fromoil wells on the Alaskan Arctic coastal plain. With fewexcepions (see map of Fipro 4a), they generally have a linearportion at de@ and a curved portion near the surface.simpleat intcrprcution of the profiles is that the linearTheportion at depth represents steady heat flux from the Earth's2interior (which we mawre to be 0.05 to 0.10 Wlm ), and thecurved pan represents a rent warming event propugatingdownward (Wed nrca). ?emount of surface warming .(dTpf) isthe diffsrencs bawasn the tempcratures obtained by extrapolatingthe straighr and curved pans to the surface, a few 'C ia thesccases (see CUIVE CTD, Figure 4b). The duration of the warmingwont, severaldecades to a century according to simplehcat-condudon models. is estimated from thepcneVation of thedepth ofmomaious curvature, referenced in theright-hand margin of Figure 4b. Arrowsrho anomaly should be 10 or 100 years afteror linw change in aurfacc tcmpcratnre.Figure 5 cotuparea three formal reconamnionstemperature history fipf) from the(Figure 4b).low where the bottom oftho start of a stepof surfaceanomaly (shaded region) at AWUThe timing and magnitude of the step and linearfunctions were determined by the least squares method describedby Lachenbruch and Marshall (1986), and tbe smooth curve (Clow etnl. 1991) by tho msthod of Speard expansion (Parker 1977, Clow1992), in which the unknown time history is represented by aseries of orthonormal functions.the= dataAll of the reconstructions fitwithin observational error (f 0.05"C) and there islittle baiis for prefemng one to another. p c resolution canbc greatly improved,however, by increasing the measurementprecision to the prewntly feasible f O.00IoC.) The results showthat although theae mc.asurernems do not resolve details of thesurface temperature history, there can be little doubt about thebig picture; at AWU and most of the other holes in this 2 x la32km region, there WM a marked, but Laterally variable increasein temperature athlsius during thethe top of permafroat ("0 of a ftw degreestwentieth century (Lachenbmch and Marshall1986, Lachcnbmcb et al. 1988a. Clow et d. 1991). Some sitesshow a more recent -ling, believed to bc related to enginwringdisturbance (Laehenbrucb and Matsbnhall 1986). A reason for thisdramatic and locally variable change has not bccn found in thehitcd rccords of surface temperature or Snowfall from theregion, and its causa remains unknown.From Figure 6, it canbc w n that heat accumulation in thesolid Earth from the unbalanced downward "climatic" flux (c) mustbe of the me order as the upward geothermal flux (g), which itnullifies to yield ncar-zero gradients near the surface (Figure6b). "he catimatos for the AWU of g N 0.06 W/m2 and c, .* 0.16Wlm' (Lachenbnrcb ut al. 1988r) arc illugtratod in Figure 6. itb of interest to campwe these mdost solid&fluxes to the4.03.02.01 .o0.01 .0125 rnn 75 50 25 0The (YBP)Figure 5. Best-fitting step, ramp, and spectral expansion modelsfor rcccnt warming event at AWU, Figure 4 (after Lachenbruch etal. 1988a, Claw et al. 1991).more vigorous activity on the other side of the Fanh's surface.The larger "top-side" fluxes drive the ecosystem, and theirbalance point determines the Earth's surface temperature (Wcllcrand Holmgren 1974). They consist of the incoming and outgoingradiation components and their difference (the net radiation,Figure 68), which, pmony othcr things, is responsible for meltingsnow and evaporating water, allowing them to return to the sea orthe atmosphere to balance both the thermal and hydrologic budgctsin preparation for the next annual cycle. Thc numbers on thearrows (Figure 6a) indicate that there top-side climatic fluxessum to zero as they should if thc climate is not changing.However, from Figure 4 we know that climatc is changing; at Awunaabout 0.16 Wlm' more has bun going into the Earth than out for ahalf century or so. As this is just lllOhh of the net radiation(itself a difference of large numbers), we cannot detect it bytrying to keep a balance sheet of fluxes at the Earth's surface.Thus, while the unbalanced climate flux (C in Figure 6) is aninconspicuous .wcond-order effect in the climatic system, it is aconspicuous firscnrdcr effect that dominates the thermal regimeof the upper IMJ m in the solid-earth system (Figures 4 and 6b).The solid Zurth is a good monitor for changes in thc surfqccenergy balance. The downward flowing heat cannot go far in accntury because the Earth is a poor conductor; it is allcontained in the stippltd region (Figure 6b), which representsthe copplcte climatic cvcnt from statt to finish.PERMAFROST AS AN AGENT OF ENVIRONMENTAL CHANGEIt is paradoxical that in permafrost regions the portion of. the ground that has the greatest influence on surface dynamics isthe very portion that is not permafMSt: the active layer and itsvcgctation. The permafrost, of course, imparts to the activelayer its important characteristics: a base generally at- 991

Figure 6. Typical thermal conditions in permafrost regions ffthc Alaskan Arctic. a. Average annual energy fluxes (Wlm ) aboveand bclow the Earth's surface. b. Geothermal regime: g is stepdygwrbermal heat loss, C is heat gain from warming climate,stippled region represents total heat accumulation by the solidearth from warming event.subfreezing temperature and impermeable to moisture, withconditions generally prohibitive for penetration by roots.typical conditions, the active layer is the growth medium forbiotic systems and the reservoir for theirwater and nutrientsupply (Oeruper et al. 1980), the Locus of most terrestrialUnderhydrologic activity (Hinzman et al. 1991, Kane et ul. 1992). anda boundary layer across which heat, moisture, and gases areexchanged between rbe solid earth and atmospheric systems.With climate warming, (increased mean annual andlor summertempcralure), the depth of thaw and settlement of the surfacewill generally increase (e.g.,Kane et al. 1991, Nakayama ef nl.1993, Waelbroeck 1993). but not uniformly; for example, deepeningtroughs can form over a network of large ice wedges, therebyaltering drainage patterns and the distribution of wet and dryhabitats.The changed distribution and motion of the soil watercan have a dominant effecton the thermal and chemical balance,including the rates of biogeochemical reactions, the productivityof living systems, the decomposition of organic matter,generation or uptake of COZ and CHI, and other charactcristicr ofthe activelayer that influence (and are influenced by) thedistribution of plant communities (Oechel and Billings 1992,Shaver er al. 1992).Changes in thethemechanical and thermal condition of permafrostas its moisture changss state during climateinteracting surface effectschange can cause(mechanical, thermal, hydrological,and biological), whose major impacts are treated in thesubsections.followingTerrain Susceptibility and Implications for DevelopmentsThe environmental impacts of permafrost and ita growth anddeterioration with changing climate depend primarily on theamount and form of the ice itmassive ice bodies are common.manifestations of the dramatic change incontains; both interstitial ice andThe impacts are almost alltransfer properties associated with phase change.interstitial ice, forstrength and heat-Melting Ofexample, decreases strength and increasespermeability, permitting increased water flow.leads to increasedmelting in an'unstableThis, in turn,advcctive heat transfer, and acceleratedprogression that can cause collapse ofmassive ice, soil flowage, and disruption of the landscape.In warmer permafrost,near the margins of its extent,increasing surface temperature can cause summertoo great to refreeze in winter.conduits fortransfer (e.&,thawing to depthsThe resulting taliks can begroundwater flow and associated convective heatin warm permafrost may beAnisimov 1989). The presence of unfrozen watera common occurrence, owing to smallamounts of soluble impurities commonly present in groundwatct(Osterkamp 1989). This unfrozen water changes the thermalproperties of the permafrost, makesand creates a distributedpermafrost.them temperature dependent,latent heat sink throughout the body ofThese factors can result in very complex thermal andhydrologic regimes in marginalwill cause them to deteriorate rapidly.permafrost areas and in many areasThe latitudinal gradientof mean annual surface temperature near the warm margins ofcontinental permafrostgenerally ranges from o.~~-~.o~c/~oo km(&molotchiLova 1988, Mackay 1975, Anisimov 1989). Thus, in someregions like central Canada and Siberia, climatic warming of afew degrees can subject vast areas to marginal decay.Thisprocess. together with the effects of a thickening active layer,will bc responsible for prompt (10°-102 years), major impacts onthe dynamics of permafrost regions subjected to a warming climate(Stuart 1986. Nelson and Anisimov 1993).Unlike the slow Loss of permafrost by heat conduction frombelow in continuouswarm discontinuous permafrost is apermafrost areas, however, thc rapid loss ofprocess that is not wcllknown. There are few published data for such aspects as thawingratts at the permafrost tableand base, processes of talikdcvelopmenr, lime scales for thawing, or the importance oflateral heat flow.In the absence of such empirical data,simulation studies assume great importance. Riseborough (1990)carried out a numerical simulation on the effectsof latent heaton the thermal response of rl.~ upper 20 m of permaltost. Whereunfrozen water is present, henear-surface ground temperature trends aresuggested that short-term,not necessarily an992

absolute measure of the magnitude and rate of change to thesurface thermal regime. Risehrougb and Smith (1593) linked thethermal response of permafrost in the Fort Simpson area ofCanada's NorthweJt Territories to a 2x cO2 warming scenario fromthe Canadian Climate Centre general circulation model.Randomized weather mords, superimposcd on the trend provided bythe GCM, were used to drive a numerical heat-transfer model thatevaluated the thermal responss of permafrost over a 140-yearperiod. Results from 121 replications wete highly dependent on ,particular details in the simulated surface climate, and gave awide range of times for diwppearance of the permafrost profile.Interannual climatic variability appears to be an importantfactor in the evolution of permafrost under an overall warmingtrend.Despite our lack of data on thawing permafrost. there isabundant historical evidence for thermokarst development overextensive regions. Evidence for the existence of thermokarstrelief at subcontinental scales, in the form of alas complexes,ice-wedge pseudomorphs, truncated wedges, and interpermafrosttaliks, has been found across Siberia, and has been related tospecific intervals characterized by distinct climatic warming(Baulin et al. 1984, Kondratjeva et al. 1993). These ihclude theclose of the Pleistocene (Baulin and Danilova 1984), the"Holocene Optimum' in Yakutia (Soloviev 1959) and a late Holomnewarming 300-25oI1 years BP in western Siberia (Solomatin 1992).Some appreciation for the potential impact of climate-inducedpermafrost degradation in areas of intensive development can begained through examination of currcnt problems involving ice-richterrain in Siberia. 'Although permafrost degradation induccd bydirect anthropogenic activities such as clear-cutting of forestsare not optimal analogs for climate-induced effects, their scalecan in some instances allow us to visualize the regional impactsthat worst-case climatenhange scenarios could eventually have inareas of ice-rich permafrost. Moreover, local- and regionalscalehuman activities are likely to be superimposed on broaderclimatic influences, and their effects may therefore be additive.Regions that contain cxtensive areas of warm, ice-richpermafrost, as in western Sibcria and the far East, areparticularly sensitive to landscape disturbance and increasedthaw depths (Vyalov et d. 1993). At exploratory drill sites innorthern Chukotka, active-layer thickness increased by a factorof two to three in terrain subjected to removal of vegetationcover, compared with adjacent undisturbed areas. The Bykovskyfish-processing plant in Yakutie, which supplied fish to theentire USSR during World War 11, is an example of problems thatmay be encountered in many settlements under a warming climate.The ice content of the ground beneath Bykovsky is as high as 90%.with only 0.6-1.0 m of mineral-soil overburden. DiMurbance ofthe upgcr 5-15 cm of moss and turf cover bas initiatedthermokarst processes that will necessitate abandonment of thecommunity within the next ten years (Origoriev et al. 1990).Climatic warming alsahas serious implications for thestability of engineered works in mountainous regions (Haeberli 1w,heberli er al. 1593). SpiSc examples include theQinghai-Ximng Highway in Tibet (An et nl, 1993, Cheng, thisvolume) and projects described by UIrich and King (1993) in theAlps.Hydrology of Permafrost AreasSeveral points we notewortby when considering thehydrological consequences of permafrost thaw.e Frozen ground has very low hydraulic conductivity (Burt andWilliams 1976) and most permafrost materials can bc regarded BSimpervious to storage and transmission of significant quantitiesof water.Thus, infiltration and percolation is limited (WOO1986). and groundwater flow is restricted to taliks andseasonally thawed zones (van Everdingen 1987).Thickening of theactive layer and formation of new taliks accompanying climaticwarming will expand subterranean passageways for storage andcirculation of groundwater.a Permafrost retains varying amounts nf water in the form ofground ice, much of which remains in storagc for centuries or ~millennia.Some of this ice will be released to the hydrologiccycle when permafrost degrades, but once melted and drained, thissource of water will be depleted.3 Heat and moisture fluxes strongly affect each other. Latentheat is involved in the freezingand thawing processes while thethermal conductivity and capacity of thawed soils differ greatlyfrom those of frozen soils (Farouki 1981, Lunardini 1981).Increased thawing offeedbacks.the permafrost will have hydrologic0 Climatic warming in permafrost regions will likely bcaccompanied by changes in the precipitation regime.Alterationsin the magnitude and timing of precipitation, as well as the formof precipitation (e.g., rain YS. snow), will influence thehydrologic rhythm of a region under a changed climate. Climaticchange, with its attendant effects on permafrost and vegctationdevelopment, will affect all the water balance components, whilemodifications of several hydrologic elements will havesignificant feedback on the permafrost,basins.Precipitation is themajor source of water input to permafrostWarmer conditions will shorten the freezing period. Atlow elevations, snowfall may occur later and snowmelt may beearlier than at prcscnt, so that peak melt events may not beconcentrated or as intense. At high elevations, the effect oftemperature lapse rate onof the effects of temperature increases.asthe freezing altitude will countcr someTogether with incrcascdstorminess and orographic influences, conditions may even favormore snowfall at high elevations while rainfall increases in thelower mnes.For areas where the ground snow cover does notdecrease significantly, the insulating effects on the permafrostmay not diminish in the future.993 '

Evaporation in permafrost regions ia limited by the longduration of mow md ice cover on the ground and by the low levelof energy available during the thawed season. A warming scenariowill shorten the snow- and iw-covcred period (except perhaps forhigh altitudes), and provide more heat in summer. Pan of thiswill be due to reduced ground heat flux as the frnzcn substratedccpens to lessen the ground temperature gradients, and part isdue to the higher sensible heat available. Computer simulation(e+, Kana et al. 1991) suggests that higher evaporation willaccompany higher summer temperatures. Should climatic changecontinue, however, thc vegetarion will likely be different fromtoday, probably ar all scales, from the species to the biomelevel. If the lichens and mosses that tend to be suppressors ofevapotransiration (Rouse et al. 1977) are replaced by transpiringplants, evaporative losses will Increase. Large-scale hydrologicchanges may omr io the extensive wetlands that occupy largetracts of the discontinuous permafrost regions of Eurasia andNorth America. Enhanced evaporation asaociated with theprojected warming will lower the water table, followed by changesin the peat characteristics as the wetlaud surfaces kome drier(woo 1992).Permafrost degradation will thicken the active layer, Thisshould allow greater infiltration and water storage, especiallytor rain that falls during the thawed period. Thus, there willbe more groundwater available to suaain baseflow in the fall,resulting in an extended sueamflow season. In terms of runoffresponses lo basin water input, Slaughter er d ' s (1983) studyin central Alaska provides a modern analog to what the futurepatrern may be: small basins with low percentages of permafrosttend to produce smaller ranges of flow conditions, yielding fewerhigh flows of large magnitude and more discharge during low flowseasons.Srreamflow of most rivers in permafrost areas exhibit a nivalregime, which is characterized by high snowmelt runoff in spring,followed by low summer flows that are occasionally raised byrainfall events (Church 1974). In the temperate zone, raingeneratedpeaks dominare and fhc seasonal runoff pattern followsa pluvial regime (Woo 1990). As climatic warming occurs,particularly if winter temperature increases are pronounced, thesnowmelt pak flow may be similar to or lower than the present,depending on whether snowfall increases or otherwise. Raininducedflows may be more frcquent if rainstorms becomeprevalent, although higher evaporation and more deeply thawedactive layers available for moisture storage may reduce therunoff pks. The baseflow period will be extended, depending onhow much later the winter arrives. In general, the nival regimerunoff pattern will weaken for many rivers in the permafrostregion, and thc pluvial influence upon runoff will intcnsiEy forrivers abng the margins of the discontinuous permafrost zones.The impacts of global warming on the hydrologic system arefull of time lags; examples include such considerations as unevenstorage changes causing variable response time, and whether thechanges are steady or involve thresholds (see Woo et 01. 1992).Other considerations include fctdbacks between the hydrologicaland external systems or among different hydrological cltments:"random" signals emanating from climatic variability, chanceoccurrences of events and phenomena; and human factors, includingdevelopment in permafrost areas, activities involved in resourceextraction, and changes in land use. Our task is to deduce thehydrologic, tendtncies under climaticby uncertainties in the climate model predictions.change forcing, constrainedGiven thecurrent limited level of knowledge on permafrost hydrologic andclimatic processts, only pwralized statements on the impacts,not detailed quantification, should warrant our confidence.Ecological and Geomorphic EffectsComprchensivt overviews of thc very large litcraturc onecological responses to climatic change are provided by Gates(1993) and Solomon and Shugan (1993). A rich literature alsoexists on permafrost as an ecological factor, much of itpotentially useful in the context of climatic change (e.g., Gill1975, Llrltin and Billings 1983, Morrissey ef 01. 1986, Burn andFriele 1989, Evans et 41. 1989, Gross et a!. 1990, Zdtai andVitt 1990, Laflcur et al. 1992, Lavnic et 01. 1992).Considerable interest has focused on geographical shifts ofthe forest-tundra transition as a respnnse to climatic change: arecent publication by Timoney ef al. (1992) provides a firm basislor monitoring northward migration of the forest tundra innorthwestern Canada, as well as for simulatinn studies. Mucheffort has already bten expended on predicting latitudinal shiftsof the boreal and tundra zones (e.&, Solomon 1986. Kauppi andPosch 1988, Monserud ef 01. 1993; also see discussion in Gates1993 and Solomon and Shugart 1993). Experiments linkingvegetational responses to GCM-derived climate scenarios generallypredict poleward migration of vegetation classes, and shrinkageof the area occupied by tundra. Simulation experiments by Bonan@r al. (1992) suggest, however, that important climate feedbackmechauisms could be triggered by shifts in the forest-tundratransition, includinp tbose accomplished by such anthrnpogcsicactivities as extensive deforestation in the boreal zone.Permafrost is an important ecological factor in the arctic andsubarctic aqd should be considered explicitly in future modelingexercises*Ueomorphic responses to climate warming were reviewed indetail by Wuo er al. (1992). Development of thermokarstphenomena at regional scales is likely to occur, and hasimportant implications for resource management, as well as thcother impacts on human activities outlined above.Potential also exists for widespread slope failure in responseto a general climatic warming trend, although relationships arelikely to bc complex and spatially inhomogeneous. Lewkowicz(1992), for example, presented evidence that rapid rates of thaw994 *

I .penetration into ice-rich subsurface layers may bc a moreimportant factor in triggeringactive-layer detachment slidesthan is the simple maximum annual depth of thaw. This conclusionagain suggests the importanccof interannual climatic variabilityfor realistic awssments of tbe response of permafrost toclimatic warming (cf. Riseborough and Smith 1993).Slaughter and Hamann (1993) suggested that degrading pingosin the discontinuous permafrost wne could be uscd to monitor thconset or progression of climatic warming; palsas and other near-surface, ice-rich Landforms in subarcticuseful in this capacity.Concerns havc bcenpearlands may also proveexpressed about the effccts on coastalpcrrnafrost of a greenhouse-induced rise in sea lcvcl (e-p., Roots1989, Barnes 1990, Woo et af. 19992). Recent observational andmodeling work focused on the response of the Antarctic ice shcctto climatic warming suggests, however, that the ice sheet mayreceive increased nourishment, and that a large increase in sealevel could bc delayed substantially (Bentley and Geovinctto1991, Sugden 1992).PERMAFROST AS A FACILITATOR OF CLIMATE CHANGEA very large amountof carbon has been stored in pcatlands ofthe boreal and tundra regions during the Holocene. Much of thiscarbon is sequestered in permafrost;warming, gencral retreatunder conditions nf climateand thinning of permafrost may convertthese regions into net sources of carbon (Billings 1987, Oechelet al. 1993, but also see Yarie and Van Cleve 1991, Bonan PI a!.1992, Kolchugina and Vinson 1993). initiating fecdbsct cffcctsand enhanced warming.Methane in shallow permafrostA substantial body of literature has accumulatcd in rccentyears on methane emissions innorthern regions (e.g., Uarriss etdl. 1985, Moore and Knowles 1987. 1990. Whalen and Reeburph 198s.Moore 1990a. IWb, Morrissey and Livingston 1992, Torn andChapin 1993).The intense interest in this subject is motivatedby recognition of methane's effectivencss as a grecnhnusc gas nndby its rapidly increasing conccntratinn in thc atmc~aphcrc(Cicerone and Oremland 1988). Although fcw data arc available,large amounts of CH4 may be stored in permafrost rcgions; awidespread increase of thaw depth in response to climatic warmingcould, in principle, release asubstantial proportion of thismethane, triggering feedback effects in the atmospheric system.Although there has ken a tremendous upsurge of research onmethane in northern regions, detailsvery poorly known.sampled emission rates (e.g.,of the mcthanc budgct arcOreat variability has been observed inMorrissey and Livipgston 1992) nndthe few data available on tho methane content of permafrost alsosuggest high variability (Kvenvolden and Lorenson 1993).Moreover, the relationahips between methane flux, soil moisture,nnd soil temperature are poorly known (Vourlitis et nl. 1993).Two recent papers attempted to model the release of methanefrom permafrost,using one-dimensional numerical heat-transfermodels driven by GCM-generated climate-change scenarios.their estimates of the methane content of permafrost on dataBasingobtained from areas near Fairbanks (Kvenvolden and Lorenson 1993)and Pmdhoe Bay (Moraes and Khalil 1993, Rasmussen et al. 1993)in Alaska, estimates of the maximum global rntes nf methanerelease were, respectively, 25-30 and 5-8 Tglyr. These resultsunderscore he uncertainty surrounding this topic.Far morc dataon the methane content of permafrost are necessary before trulyreliable estimates can be made.Furure modeling effortsinvolving circumarctic extrapolation should also be based on thelarge-scale, high-quality map of permafrost distributionschcduled for publication in early 1994 (Brown 1992).rather thanone of the many published small-scale depictinns, the majnrity ofwhich are poorly documented (see Nelson 1989, Nelson and Anisinw1993).Gas hydrate: a recordkeepcr and agent of climate changeSediments in cold regions may trap large quantitiesgas, largely mcthane, in icc-like crystalline structuresof naturalcontaining water molecules and called gas hydrates or rlathrutes(Katz ez al. 1959). They store natural gas efficicntly (up tow 150 times as much methane as anequal volume of free pns untlcrstandard conditions of temperature and pressure. Gas hydratesare of interest inconnection with global climate change as apotential source of atmospheric methane, one of the mostimportant greenhouse gases.potential commercial source of energy.Gas hydrates aretemperature andlorsuperficial part of the solid earth. Theshown in Figurewherein fluid pressure is equalThey are also of interest as astable under special conditions of lnwhigh pressurc that obtain in only a small,stability ficld is7. assuming the common ("hydrostatic") condition.to the weight of a column ofwater extending to the surface. For thc typical Arctic coastalconditions illustrated for Barrow,methane hydrate isAlaska ("0 yrs," Figure 7a),stable between about 200 and 700 m, andpermafrost extcnds downward below a thin active laycr to a dcpthnf about 400 m. At Prudhoe Bay, Alaska, the higher conductivlty(lower thermal gradient) ice-rich permafrost extcnds to a depthof 600 m and the hydrate stability ficld to more than 1 km(Figure 7b). Understeady conditions, hydrates will be stable onthe continent only in colder permafrost regions.steady-state temperature gradientsutfacc temperaturegcotherm to interscctThe principal globakhangeFor the typicalillustrated in Figure 7a, \hewould have to be below about -5'C lor thethe hydrate field.question about gas hydratc is:will a warming climate destabilize gas hydrates, release mcthaneto the atmosphere, and thercby enhance thecontribution eo the greenhouse effect? Thewarming by itsgeothermal effect of3 sudden increase of surface temperature by 10°C is illus!rat~d' 995

Significant amounts of methane will not be released fromdestabilized methane hydrate in permafrost areas by present-dayclimate change for a millennium or more.e Presentday contributions to atmospheric methane fromdestabilized hydrate should be sought on the Arctic continentalshelf, where warming started with inundation thousands of yaarsago (Lachenbruch et al. 1982, 1988b. Kvenvolden 1988, MacDonald1990, Judge and Majorowicz 1992, Osterkamp and Fei 1993).Figure 7. Thermal. decay of methane hydrate stability field(shaded) beneath a transgressing Arctic shoreline. Land surfacewith geotherm typical of Barrow, Alaska (a), or Prudhoe Bay.Alaska (b), is inundated by sea (mean annual bottom temperature,-1°C) at time t= 0 years. Destabilization takes ten times longerat Prudhoe Bay owing to latent heat absorbed by its highice-content permafrost (Lachenbruch et al. 1982, 1988b, 1988c.Lachenbruch and Saltus, unpublished calculations).in Figure la. This is much more severe than predictions ofclimate models, but it is an accuratereprcscntation of the"climate change" that occurs when Arctic seas ovcrridc the land.a process that must have occurred nvet millions of squarckilometers on the Arctic cnntincntal shelf as it was inundated byrising sea lcvcl following the last glaciation.continues today at rates exceeding aTransgressionmeter per year along much ofthe Arctic coastline (e.&, Caner et d. 1987, Mackay 1986).Figure la suggests that the effects of a thermal disturbancetoday would not reach the top of the gas hydrate field at -200 mfor centuries, and for initialconditions typical of Barrow,Alaska today (Lachcnbruch et 41. 1988~). significant methanerdoase could not occur for thousands of years. If,"Barrow model" in Figureunlike thela, the permafrost were ice rich as atPrudhoe Bay (Lachenbruch et al. 1982, Lachenbruch et al. 1988b),the destabilivrtion would take tens of thousands of years (Figure7b. Although the phase boundaries of gas hydrates vary withComposition, fluid pressure, and salinity, these examples serveto illustrate two rather generalpoints:INFORMATION ISSUESResearch into the effectsgenerated a substantial increasc inthis topic sinceof climatic change on permafrost hasthe number of publicatians onthe last permafrost conference (Koster and Judge1994). Although warming trends have consistently been predictedto be most pronounced in the high latitudes (e.&, Roots 1989),the permafrost community has, with only a few exceptions (e.g.,Lachenbruch and hhrshall 1986), not communicated the importanceof its subject as a recorder, facilitator,climate-induccd changes to the largercspecially in high-impact, integrative medla.Interest inand agent ofscientific community;the potential effects of climatic changc onpermafrost is widespread and increasing; the situation isrcflected in the growth of literature on this general topic overthe past five years. Rcvitwshave been published recently byBarry (1985). Williams and Smith (1989). Pissan (1990). Streetand Melnikov (1990), Vinson and Hayley (1990), Haeberli (1992),Koster (1991, 1993a), Koster and Nieuwcnhudzen (1992). Melnikovand Street (1992). C)avrilova (1993), and Haeberli er d(1993).Several of these publications built upon earlier effortsMcBeath (1984) and French (1986).Despite worldwide interest in climatebychange issues and theassumption that nonhtrn wetlands and permafrost regions play avital role in theglobal climate system, relatively few studieshave appeared specifically dealing with these subjects andpresenting new data. The IPCC Scientific Assessment of GlobalChange (Aoughton et al. 1990) reflects this minor attention toclimate-permafrost interactions, with only two short paragraphson high-latitude methane sources (p. 22) and temwrature profilesin permafrost (p. 225).Similarly, permafrost bibliographies produced by Hcginhttomand Sinclair (1985) and Brennan (1988, 1993) rcvcal that, despitea large number of cirations, only a limited number of .publications explicitly present data on the response ofpermafrost to former, present, or future climatic changes.similar conclusion can be drawnAfrom the overview presented inTable 1, which reflects items published throagh 1992. Studies on*frozen ground properties" (125) and "permafrost engineering"(203), predominantly of North American origin, provide thenecessary background for climatenot address rhe subject in depth.change impact analysis, but doStudies on "palCWnVi~O~~ental, .,9 996

tI 363 72 11 37 1839 4 6 306 17 28 13 3 4. t!4047 lo 1507 433 I92I320 3114 31t 13ti 6 Olfrhilrl. pelnlalro6T ~rpllllrlwn wd EUllbtrUStIUn6 7 2 LI7 *r,,,,,

CONCLUSKONS: AN AGENDA FOR RESEARCHClimatic warming raises thespcncr of a very severe andspific sat of problems in high-latitude regions, To addressthese chnllengcs in an effective mnncr will require coordinatedand sustained effort by the permafrost research community.of the more prominent'below.ModelingGeneral circulation modelscurrently used to investigatedistribution of climatic change.isnues in need of resolution are outlined(GCMs) are the primary meansand predict the nature, extent, andGCMs are complex,threedimensional models of .the earth's atmosphere and oceans;because they are cxtrcrnely consumptive of computing resourcesSomethey are implemented over very coarse geographical Lattices. Theinternode distances of these models, often several degrees oflatitude andlongitude, is of profound importance for applicationof their results to a variety of scientific problems. Issues ofscale dependence and theinvolving regional climate are thereforeefforts at present (e.g., Karl et 01.,and Mearnsapplicability of GCMs to problemsthe focus of intensive1990, Boville 1991, Giorgi1991, Grotch and MacCracken 1991, Gutowski er al.1991, Hewitson and Crane 1992).With only the few exceptions noted in earliercontrast to the efforts of botanists and ecologists (e.g.,sections, snd inSolomon 1986, Kauppi and Posch 1988, Thomas and Rowntree 1992),permafrost gcientists have not made extensive use of GCM resultsin their discussions about potential impacts of climate change.Although scale-related problems areinvolving the formulationsubstantial, as are othersof explicit linkages between GCMresults and the transient response of permafrost to climati'cchange (see discussion in Burn and Smith 1993), generalcirculation models arc the scientificcommunity's primary toolsfor forecasting climate- change scenarios. As part of the<strong>International</strong> PermafrostAssociation's focus on climate-changeissues (IPA 1993). high priority should be given to developinglinkages between permafrost research and models of climatechange, over a hierarchy af geographical scales.Technological isauesTechnology introduced in the wlrnmercial market during thc lastdecade has made pos~ble substantial improvements in thefeasibility, cost, and scale of field measurement programs inremote cold-cllmste regions. As in many other areas of science,technological innovations will allow us to expand the scope ofour investigations to such an extentquestions wc ask is fundamentally altered.that the nature of theIntroduction ofrelatively inexpensive battery-powered, computer-controlledmeasuring and recording systems, for example, has made itpossible to oMniD bigh-frcqucncy time saries at remote sites,allowing extensive evaluation of local conditions and theirvariabitity. Accuracy requirements CUP be stringent, howevcr,particularly where permafrost temperatures are near OOC, or inspecial applications. Some of the less costly instrumentation isnot optimally cofllpred for the temporal intervals, sensingresolution, or tcmpcratwe canditions required for permafrostresearch. Recent unpublished calculations by Zhang andOstcrkamp, for example, suggest that highly accuratedetermination of in situ thermal diffusivity may require accuracyon the order of *O.OIoC in the mcasurcments. Permafrostresearchers should communicate their needs and standards tomanufacturers clearly and in detail, particularly in the sphereof low-priced instrumentation, which may require only inexpensivemodifications to achieve a configuration that can k usedeffectively and widely.In the even more demanding application of using theinformation stored in permafrost as a paleoclimatic signal, toachieve a reiiable means of identification, reconstruction, andfuture monitoring of recent changes in the temperature at thepermafrost surface requires that temperatures be monitored with aPrecision of * milliKelvin (mK). If borehole temperatures Canbe remeasured over a period of years with millidegree precisinn,we Can overcome major obstacles to the reconstruction of surfacetemperature posed by steady-statc effects of inhomogeneity andthreedimensionality. Commercially available thermistortranduccrs, metering bridges, and tmreholc logging cables areadequate to Construct systems with a sensitivity of a fraction ofa mK. The two principal limitations to attaining thc desirddprecision in the field are: 1) returning to the same measurementdepths with adequate precision on relogging, and 2) controllingthe random convection of the generally unstable fluid in the, borehole. These problems are currently being addressed by a teamat the U.S. OcologiCal Survey in Menio Park.Remote sensing has become an important tool in many aspects ofPdar research (kg., Hall and Martincc 1985, Epp and Matthews1991); permaftOSt is no exception. Remote sensing has proveneffective for relating vegetation and other terrain conditions toPermafrost (e.&, Morrissey et al. 1986. Morrissey 1988, Belangeref 01. 1990, Liang and Ou 1993, Peddle and Franklin 1993).Ground-penetrating radar is extremely useful for delineatingpermafrost distribution at local scales (Arcone et ale 1982,Dmlittfe et 41. IW), and should prove equally effective formonitoring areas with the potential for stability problems underconditions of general warming (Judge et 01. 1991). Asremote-sensing technology evolves, it is likely to increase inimportance as a gmcryological tool. Enhanced rcsolution fromspace-borne sensors, for example, may provide a means forobserving large-%ale mass movements or the degradation ofnear-surface permafrost iu subarctic peatlands on a circumpolar''998 -

asis, rather thanat the local or regional scales necessitatedby the us0 of sequential aerial photography (Thie 1974).Measurement Programs and Data ArchivesAs important as the contributions of individuals or smallgroups of investigators may be, maximum advantage can be made oftechnological advances only if the data derivedmade available to the larger scientific community.from them areRecentstudies suggest strongly that to view climate change as operatingprimarily at low frequcncies is ovcrly simplistic (Eradley andJones 1y92, Stcvens 1993; instead, integrated paleoclimaticevidence indicates that climate change operates at avariety oftemporal and spatial scales. and that pronounced, regionaloscillations take place at the d-.dl scale. The implicationsof these findings are profound for studies of permafrost, andunderscore the criticalbasis.Despite its importance toand applied science, there hasneed for integration of data, on a globala large numbcr d fields in basicbeen little consensus about orattempts to achieve a coordinated, international implementationof field measurements of ground temperature. Records areextremely few in many pans of the circumpolarstandardization prescnts a formidablc barrier tointegrative studies of existing records.regions andcomparative orGiven the very highlevels of interest and concern about the amplification of globalwarming in the polar regions, it is imperativcthat aninternational measurement program be implemented on a globalbasis.A program of ground-temperature measurements is neededurgently, similar inExperiment) effort to monitorwarming on arctic biota (Molau 1992).ground-temperature stations, ideally at orwill establish critical empiricaltemperature.mpe to the ITEX (Intcrnarional Tundraand evaluate the cffects of climatcA circumpolar network ofnear ITEX locations,Linkages between air and groundMoreover, sukh a network can provide informationabout the nature, extent, and rateof thaw that will be necessaryfor realistic regional and global planning activities.Much of the foregoing discussion centers on climatechange,its detection, and the derivation of recent climate history fromrelatively deep boreholes in thick permafrost.There is an acuteneed to inventory borehole locations and archive the data fromthcrn. This review also draws attentionobaervutional data describingto the shortage ofpermafrost thawing ordisappearance, and emphasizes the impanance of the active layerfor permafrost evolution under a wurming climate.Detection anddocumentation of warming in thc upper layers of permafrost, aswell as changes in the active laycr, are critical to theadvancement of our knowledge about climate change itself, as wellas the response of prdrost to it. Brennan and Barry's (1989,also see Barry and Brennan 1993) suggestiocs for programs tocollect, standardize, and archivepermafrost-related data, on aninternational basis, must k given high priority if we hope toobtain the backgroundnecewry to monitor cbnn&es in permafrostareas sensitive to climatic or anthropogenic disturbances. Onlywith an integrated, global database can adequate understandingattained about the degradation of permafrost, its effects ongeomorphic proccssea, and its potential to threaten the integrityof man-made structures andAcknowledgmentsThe authors thank Drs. Jerryreviews of the manuscrip.REFERENCESfacilitics.beBrawn and Roger Barry for helpfulAgarwal, A. and S. Nsrain (1990). Global Warming in an UnequalWorld: A Case of Environmental Colonialism. Centre for Scienceand Environment, New Delhi.An, W., 2. Wu, Y. Zhu, and A. Judge (1993). Influence of climatechange on highway embankment stability and permafrost in thepermafrost region of thc Qinghai-Xizang Plateau. InProceedings of the Sixth <strong>International</strong> <strong>Conference</strong> onPermafrost, Volume I, 11-16. South China University ofTcchnology Press, Wushan, Guangzhou, China.Anisimov, O.A. (1989). Changing climate and permafrostdistribution in the Soviet Arctic. Physical Geography 10(3),285-293.Arcane, S.A., P.V. Sellmann, and A.J. Delaney (1982). Radardetection of ice wedges in Alaska. CRREL Report 82-43, 15 pp.Balobaev, V.T., V.N. Devyatkin, and I.M. Kutasov (1978).Contemporary geothermal conditions of the existence avddevelopment of permafrost. In Sauger, F.J., (ed.), Proceedingsor the Second Internationul <strong>Conference</strong> on Permafrost, USSRContribution, 8-12. National Academy of Sciences, Washington,DC .Barnes, P.W. (1990). Effects of elevated temperatures and risingsea level on arctic coasts. Journal of Col&RegionsEngineering 4(1). 21-28.Barry, R.0. (1985). Cryosphere and Climate Change. U.S.Department of Energy, Office of Energy Research, Office ofBasic Energy Research, Washington, D.C., WE\ER4235.Barry, R.G. and A.M. Brennan (1993). Sixth <strong>International</strong><strong>Conference</strong> OQ Permafrost: Towards a permafrost Informationsystem. In Promedings of the Sixth <strong>International</strong> <strong>Conference</strong>on Permafrost, Volume 1, 23-26. South China University ofTechnology Press, Wushan, Guangzhou, China.Baulin, V.V. and N.S. Danilova (1984). Dynamics of LateQuaternary permafrost in Siberia. In Velichko, A.A., (ed.),Late Quaternary Environments of the Soviet Union, 69-77.University of Minnesota Press, Minneapolis.Baulin, V.V., B. Belopukhova, and N.S. Danilova (1984). Holoccnepermafrost in the USSR. In Velichko, A.A., (ed.), LateQuaternary Environments of the Soviet Union, 87-91. Universityof Minnesota PresJ, Minneapolis.

Belanger, J.R.. S.R. Dallimore, and P.A. Egginton (1990).Applications of thematic mapper thermal infrared imagery forPermaprost pnd terrain studies, Richards Island, N.W.T. InPrdings of the Fifth Canadian Permafrost <strong>Conference</strong>,231-238. Centre d'etudes nordiques, Universith LavaUNationalRcmrch Council of Canada, Quhbcc.Bentley, C.R. and M.B. Oeiovinctto (1991). Mass balance ofAntarctica and sen level change. In Wcller. G., C.L. Wilson,and B.A.B. Severin, (cdn.), Proceedings. <strong>International</strong><strong>Conference</strong> on the Role of the Polar Regions in Global Change.Volume II. 481-488. Qoophydcal Institute, University ofAlaska, Fairbanks.IBillings, W.D. (1987). Carbon balance of Alaskan tundra and taiga' ccosystems: past, present und future. Quaternary ScienceReviews 6(2), 165-177.Bonan, G.B., D. Pollard, and S.L. Thompson (1992). Effects ofboreal forest vegemtian on global climate. Nature 359,716-718.Boville, B.A. (1991). Sensitivity of simulated climate to modelresolution. Journal of Climate 4(5), 469-485.Bradley, R.D. and P.D. Jones (1992). Climate Since A.D. 1500. 679pp. Routledge, Chapman & Hall, London.Brennan, A.M. (1988). Permafrost Bibliography Update 1983-1987.Glaciological Data Report GD21. 225 pp, World Data Center Afor CJlaeiology, Boulder, CO.Brennan, A.M. (1993). Permafrost Bibliography Update 1988-92.Qlaciological Data Report 0J3-26.World Data Center A forGlaciology, Boulder, CO.Brennan, A.M. and R.G. Barry (1989). Ice Core Update 1980-1989.Permafrost Data Worksbop. 127 pp. Glaciological Data ReportGD-23, World Data Center*for Glaciology, Boulder, CO.Brown, J. (1992). A chamarctic map of permafrost and ground iceconditions. In clilichinsb, D.A., (ed.), Proceedings, 1st<strong>International</strong> <strong>Conference</strong> on Cryopedology, 15-21. PushinoResearch Centre, Pushino. Russia.Budyko, M.I. and Y.A. Iuael (1987). Anthropogenic ClimaticChange. 405 pp. Hydrometeoizdat, Leningrad, in Russian;English edition published by The University of Arizona Press,1992.Burn, C.R. and P.A. Friele (1989). Geomorphology, vegetationsuccession, soil characteristics and permafrost inretrogressive thnw dumps near Mayo, Yukon Territory. Arctic42(1), 314.Burn, C.R. snd M.W. Smith (1993). Issues in Canadian permafrostresearch. Progress in Physical Geography 17, 156-172.Burt, T.P. and P.J. Williams (1976). Hydraulic conductivity infrozen soils. Ea& Surface Processes 1, 349-360.Carter, L.D., T.A. Heginbbttom, and M.-k. Woo (1987). Arcticlowlandu. In ad, W.L., (ed.), Geomorphic Systems of NnrthAmerica, Centemid SpGcial Volume 2, 583-627. GeologicalSociety of America, Boulder, CO.'Christensen, T.R. (1993). Methane emission from Arctic tundraBiogeochemistry 21(2), 117.139.Church, M. (1974). Hydrology and permafrost with reference tonorthern North America. In Proceedings, Workshop Seminar onPermafrost Hydrology, 7-20. Canadian National Committee,<strong>International</strong> Hydrological Decade, Ottawa.Cicerone, R.J. pnd R.S. Oremland (1988). Biogeochemical aspectsof atmospheric methane. Global Biogeochemical Cycles 2(4),299-327.Claw, 0.D. (1992). The extent of,temporal smeriring insurface-temperature histories derived from boreholetemperature measurements. Global and Planetary Change 6,81-86.Claw, G.D., A.H. Lachenbruch, and C.P. McKay (1991). Inversion ofborehole temperaNrG data for recent climatic changes:examples from the Alaskan arctic and antarctica. In Weller,G., C.L. Wilson, and B.A.B. Severin, (eds.), Proceedings,<strong>International</strong> <strong>Conference</strong> on the Role of the Polar Regions inOlobal Change, Volume 11, 533. Geophysical Institute,University of Alaska, Fairbanks, AK.Dingman, S.L., R.Q. Barry, a. Weller, C. Benson, E.F. LeDrew, andC.W. Owdwin (1980). Climate, snow cover, microclimate, andhydrology. In Brown, J., P.C. Miller, L.L. Tieszen, and F.L.Bunnell, (eds.), An Arctic Ecosystem: the Coastal Tundra atBarrow, Alaska, 30-65. Hutchinson & Ross, Stroudsburg, PA.Doolittle, J.A., M.A. Hnrdisky, and S. Black (1992). Aground-penetrating radar study of Goodream Palsas,Newfoundland, Canada. Arctic and Alpine Research 24(2),173-178.Epp, H. and S.B. Matthews (eds. 1991). Remote Sensing of ArcticEhvironments. Arctic 44(Supplement I), 171 pp.Evans, B.M., D.A. Walker, C.S. Bensnn, E.A. Nordstrand, and G. W.Peterson (1989). Spatial interrelationships between terrain,snow distribution and kcgetation patterns at an Arcticfoothills site in Alaska. Holarctic Ecology 12(3), 270-278.van Everdingen, R.O. (1987). The importance of permafrost in thehydrological regime. Canadian Bulletin of Fisheries andAquatic Scienccs 215, 243-276.Farouki, O.T. (1981). Thermal Properties of Soils. CRRELMonograph 81-1, 151 pp. U.S. Army Cold Regions Research and. Engineering Laboratory, Hanover, NH.French, H.M. (1986). Climate Change Impacts in the CanadianArctic. Pr6ceedings of a Canadian Climate Program Workshop,3-5 March 1986. 171 pp. Environment Canada, Orillia, Ontario.Gates, D.M. (1993). Climate Change and its BiologicalConsequences. 280 pp. Sinauer Associates, Sunderland, MA.Gavrilova, M.K. (1993). Climate and permafrost. Permafrost andPcriglacial Processes 4(2), 99-1 11.Gersper, P.L., V. Alexander, S.A. Barkley, R.J. Barsdate, andP.S. Flint (1980). The soils and their nutrients. In Brown,J., P.C. Miller, L.L. Tieszcn, and F.L. Bunnell, (eds.), AnArctic Ecosystem: The Coastal Tundra at Barrow, Alaska,421432. Dawden, Hutchinson & Ross, Stroudsburg, PA.' 1000 '

Gill, D. (197s). Influence of white spruce on permafrost-tablemicrotopography, Mackenzie River Delp. Canadian Journal ofEarth Sciences 12, 263-25?Giorgi, F. and L.O. Mearns (1991). Approaches to the simulationof regional climate change: a review. Reviews of Geophysics29(2), 191-216.Goodrich, L.E. (1982). The influence of snow cover on the groundthermal regime. Canadian Cleotechnical Journal 19, 421-432.Grigoriev, M.N., I.V. Pozdniakov, and V.P. Romanov (1990).Cryogenic processes in ice complex ground causcd byman-induced disturbance of the surface. In Ratsionalnoieprirodapohvanie v kriolitozone, IMZ SO AN SSSR, YakutskGross, M.F., MA. Hardisky, J.A. Doolittle, and V. Klemas (1990)Relationships among depth to frozen soil, soil wetness, andvegetation type and biomass in tundra near Bethal, Alaska,U.S.A. Arctic and Alpine Research 22(3), 275-282.Grotch, S.L. and M.C. MacCtacken (1991). The use of generalcirculation models to predict regional climatic change.Journal of Climate 4(3), 286-303.Gutowski, W.J.J., D.S. Gutzltr, and W.-C. Wang (1W1). Surhccenergy balances of three general circulation models:implications for simulating regional climate change. Journalof Climate 4(2), 121-134.Haeberli. W. (1992). Possible effects of climatic change on theevolution of Alpine permafrost. In Boer, M.M. and E.A. Koster.(eds.), Cfrwnbouse-Impact on Cold-Climate Ecosystems andLandscapes. Catena Supplement 22, 23-35, Catena Verlag,Cremlingen-Destedt, Germany.Haeberli, W., Cheng Guodong, A.P. Gorbunov, and S.A. Harris(1993). Mountain permafrost and climatic change. Permafrostand Periglacial Processes 4(2), 165-174,Hall, D.K. (1988). Aswssmcnt of polar change using satellitetechnnlogy. Rtviews of Oeophysics 26(1), 26-39.Hall, D.K. and J. Martincc (1985). Remote Sensing of Ice andSnow. 189 pp, Chapman and Hall, New York.Hallct, B. (1978). Solute redistribution in freezing ground. InProceedings of the Third <strong>International</strong> <strong>Conference</strong> onPermafrost, Volume 1, 85-91. National Research Council ofCanada. Ottawa.Harrison. W.D. (1991). Permafrost response to surface temperaturechange and its implications for the 40,00(1-year surfacehistory at Prudhoe Bay. Journal of Qeophysical Research%B(2), 683695.Harriss, R.C.. E. Gorhum. D.I. Sebacher. K.B. Bartlett. and P.APlebbc (1985). Methane. flux from northern peatlands: Nature315, 652-654.Hewitson, B.C. and R.G. Crane (1992). Regional-scalc climateprediction from the GISS GCM. Global and Planetary Change 97,249-267.Hineman, L.D., D.L. Kane, R.E. Gieck, and K.R. Everett (1991).Hydrologic and thermal properties of the active layer in theAlaskan Arctic. Cold Regions Science and TechnoloSy 19(2),95-110.Houghton, J.T., G.J. Jenkins, and J.J. Ephraums (1990). ClimateChange. The IPCC Scientific Assessment. 364 pp, CambridgeUniversity Press, Cambridge.IPA (1993). Resolution on permafrost and global climate change,approved at 1PA Council Meeting, July 8, 1993, Beijing, China.Frozen Ground 14, in press.Judge, A. and J.A. Majorowicz (1992). Geothermal conditions forgas hydrate stability in the Beaufort-Mackenzic ma: theglobal change aspcct. Global and Planetary Change 6(2/4),25 1-263Judge, A.S., C.M. Tucker, J.A. Pilon, and B.J. Moormqn (1991).Remote sensing of permafrost by ground-penetrating radar attwo airports in Arctic Canada. Arctic 44(Supplement l), 40-48.Kahl, J.D., D.J. Cbarlevoix, N.A. Zaltwva, R.C. Schnell, andM.C. Serreze (1993). Absence of evidence for greenhousewarming over the Arctic Ocean in the past 40 years. Nature361(6410), 335-337.Kane. D.L.. L.D. Hinunan, and J.P. Zarling (1991). Thermalresponse of the active Lyer to climatic warming in apermafrw environment. Cold Regions Science and Technology19(2), 11 1-122.Kane, D.L., L.D. Himman, M.-k. Woo, and K.R. Everett (1992).Arctic hydrology and climate change. In Chapin, F.S. fly,R.L. Jefferies, J.F. Reynolds, G.R. Shaver, and J. Svoboda,(eds.), Arctic Ecosyscms in a Changing Climate: anEcophysiological Perspective, 35-57. Academic Press, SanDiego.Karl, T.R., W.-C. Wang, M.E. Schlesinger, R.W. Knight, and D.Portman (1990). A method of relating general circulation modelsimulnted climate to the observed local climate. Part I:Seasonal statistics. Journal of Climate 3(10), 1053-1079.Katz, D.L., D. Cornell, R. Kobnyashi, F.H. Poettmann, J.A. Vary,J.R. Elenbass, and C.F. Weinaug (1959). Handbook of NaturalGas Engineering. 802 pp. McGraw-Hill, New York.Kauppi, P. and M. Posch (1988). A caw study of the effects ofC02-induced climatic warming on forest growth and teh forest' sector. A. Productivity reactions of northern boreal forests.In Parry, M.L.,. T.R. Carter, and N.T. Konijn, (eds.), TheImpact of Climatic Variations on Agriculture. Volume 1.Assessments in Cool Temperate and Cold Regions, 183-195.Kluwer Acudernic, Dordrecht, The Netherlands.Hcginbottom, J.A. and M.P. Sinclair (1985). Cumulative Index toPermafrost <strong>Conference</strong> Proceedings (1958-1983). 215 pp.Geological Survey of Canada Open File Report 1135, Ottawa.1001 -

Kolchugina, T.P. and T.S. Vinson (1993). Climate warming and thecarbon cycle in the permafrost zone of the former SovietUnion. Permafrost and Periglacial Processes 4(3), 149-163.Kondratjeva, K.A., S.F. Khrutzky, and N.N. Romanovsky (1993).Changes in the extent of permafrost in the late Quaternaryperiod in the territory of the former Soviet Union. Permafrostand Pcriglacial Processes 4(2), 112-':9.Koster, E.A. (1991). Assewment of climate change impact inhigh-latitude regions. Terra 103, 3-13,Lachenbruch, A.H., S.P.J. Galanis, and T.I-&J. Moses (1988b). Athermal cross section for the permafrost bd hydrate stabilityzones in tbc Kupruk and Prudhoc Bay oil fields. U.S.Geological Survey Circular 1016, 48-51.Lachenbnrch. A.P., J.H. Sass, LA. Lawver, M.C. Brewer, B.V.Marshall, R.J. Munroe, J.P.*lly,S.P.J. Galanis, andT.H.J. Moses (1988c), Temperature and depth nf pcrmafrost onthe Arctic shpc of Alaska. In Grye, G., (ed.), Geology andExploration of the National Petroleum Reserve in Alaska. U.S.Qeologid Survey Professional Paper 1399, 645456.Koster, E.A. (1993a). Qlobal warming and periglacial landscapes.In Roberts, N., (ed.), The Changing Global Environments,127-149. Boston, Basil Blackwell.Koster, E.A. (1993b). Introduction-present glcbal change andpermafrost, within the framework of the <strong>International</strong>Geosphere-Biosphere Programme. Permafrost and PeriglacialProcesses 4(2), 95-98.Koster, E. and A. Judge (1994). Permafrost and Chmatic Changc:an Annotated Bibliography. in press pp. Woi-Id Data Center Afor Glaciology, Boulder, CO.Koster, .E.A. and M.E. Nieuwenhuijzen (1992). Permafrost responseto climatic change. In Boer, M.M. and E.A. Koster. (cds.),Ereenhouae-Impact on Cold-Climate Ecosystems and Landscapcs.Catena Supplement 22,- 37-58. Catena Verlag,Cremlingen-Destedt, Germany.Kvenvolden, K.A. (1988). Methane hydrates and glnbal climate.Olobal Biogeochemical Cycles 2(3), 221-229,Kvenvolden, K.A. and T.D. Lorenson (1993). Methane inpermafrost-preliminary results from coring at Fairbanks,Alaska. Chemosphere 26(1-4), 609-616.Lachenbruch, A.H. (1957). Thermal cffccts of the ocean onpermafrost. Geological Society of America Bulletin 68,1515-1530.Lachenbruch, A.H. (1959): Periodic heat flow in a stratifiedmedium with application to permafrost problems. U.S.Geological Survey Bulletin 10834, 36 pp.Lachenbruch, A.H. (1966). Contraction theory of ice-wedgepolygons: a qualitative dimdon. In Permafrost<strong>International</strong> <strong>Conference</strong>, 63-71. US. National Acadcrny ofSciences-National Research Council 1287, Washington, D.CLachenbruch, A.H. and B.V. Marshall (1986). Changing climate:geothermal evidence from permafrost in the Alaskan arctic.Science 234, 689-696.Lachenbruch, AH., J.H. Sass, B.V. Marshall, and T.H.J. Moses(1982). Permafrost, heat flow, and the geothermal regime atPrudhoe Bay, Alash. Journal of Geophysical Research 87(11B),9301-9316.Lachenbruch, A.H., T.T. Cladouhos, and R.W. Saltus (1988a).Permafrost temperature nnd the changing climate. In Senneset,K., (ed.). Promedings of the FX&h <strong>International</strong> <strong>Conference</strong>on Permafrost, Volume 3, 9-17. Tapir Publishers, Trondheim,Norway.Lafleur, P.M., W.R. Rouse, and D.W. Carlson (1992). Energybalance differences and hydrologic impacts across the northerntreeline. <strong>International</strong> Journal of Climatology 12(2), 193-203.Lavoie, C. and S. Payettc (1992). Black Spruce growth forms as arecord of a changing winter environment at treeline, QuBbec,Canada. Arctic and Alpine Research 24(1), 4049.Levine, M.D., J.A. Sathaye, and F.P. Craig (1992). Strategies foraddressing climate change: policy pcrspcctives from around thcworld. Energy 17(12), 1121-1136.Lewkowica, A.0. (1992). Factors influencing the distribution andinitiation of activc-layer detachment stidcs on EllesmereIsland, arctic Canada. In Dixon, J.C. and A.D. Abrahams,(eds.), Periglacial Oeomorphology, 223-250. Wiley, New YorkLiang, F. and Z. OU (1993). Application of remote sensing imagesto the investi~ations of the changes of permafrost environmentin burned forest regions, Da Hinggan Ling, China. InProceedings of the Sixlh <strong>International</strong> conference onPermafrost, Volume 1, 388-392. South China University ofTechnology Press, Wushan, Uuangzhou, China.Luken, J.O. and W.D. Billings (1983). Changes in bryophyteproduction associarep with a thermokarst erosion cycle in asubarctic bog. Lindbergia 9, 163-168.Lunardini, V.Y. (1981). Heat Transfer in Cold Climates. 731 ppVan Nostrand Reinhold, New York.MacDonald, G.J. (1990). Role of methane clathrates in past andfuture climaees. Climatic Change 16, 247-291.Maclray, J.R. (1975). The stability of permafrost and recentclimatic chnnge in the Mackenzia Valley, N.W.T. GeologicalSurvey of Canada Papcr 75-1, Part B, 1731176.Mackay, J.R. (1983). Downward water movement into frozen ground,western arctic mast, Canada, Canadian Journal of EarthSciences 20, 120-134.Mackay, J.R. (1986). Fifty years (1935 to 1985) of coastalretreat west of Tuktoyaktuk, District of Mackenzie. GeologicalSurvey of Canada Pap 86-1A, 727-735.Mackay, J.R. (1988). Pingo collapse and paleoclimaticreconstruction. Canadian Journal of Earth Sciences 25(4),495-51 1.Macllray, J.R. and I.V.J. Matthews (1983). Pleistocene ice and sandwedges, Hoopcr Island, Northwest Territories. Canadian Journalof EIanh Sciences 20, 1087-1097.1002 '

Maxwell, J.B. (1992). Arctic climate: potential for Change undcrglobal warming. In Chapin, F.S. 111, P.L. Jeffcries, J.F.Reynolds, G.R. Shaver, and J. Svoboda, (eds.), ArcticEcosystems in a Changing Climate: an EcophysiologicalPerspective, 11-34. Academic Press, San Diego.Maxwell, J.B. and L.A. Barrie (1989). Atmospheric and climaticchange in the Arctic and Antarctic. Ambio 1, 42-49.McBeath, J.H. (1984). The Potcntinl Effects of CarbonDioxide-Induced Climatic Changes in Alaska, Proccedings of a<strong>Conference</strong>. 208 pp. School of Agriculture and Land ResourcesManagement, University of Alaska, Miscellaneous Publication83-1, Fairbanks.McGaw, R.W., S.I. Outcalt, and E. Ng (1978). Thermal propertiesand regime of wet tundra soils at Barrow, Alaska. InProceedings of the Third <strong>International</strong> <strong>Conference</strong> onPermafrost, Volume 1, 47-53. National Rescarch Council ofCanada, Ottawa.Melnikov, P.1. and R.B. Street (1992). Terrestrial component nlthe cryosphere. In Tegart, W.J.McG. and G.W. Sheldon, (eds.),Climate Change 1992: thc Supplzlncntary Report to the IPCCImpacts Assessment, 94-102. Australian Government PublishingService, Canberra.Mcssner, W., D. Bray, G. Germain, and N. Stcht (1992). Climatechange and social order; knowledge for action? Knowlcdge andPolicy 5(4), 82-10.Molau, U. (1992). ITEX Manual. 9 pp. Man and the BiosphereProgram, Goetborg, Swcdcn.von Moltke, K. (1989). <strong>International</strong> agreement to stabilizeclimate: lessons from the Mantreal Protocol. Climatic Change14. 211-212.Monserud, R.A., N.M. Tcbebakova, and R. Lccmans (1993). Globs1vegetation change predicted by the modified Budyko model.Climatic Change 25(1), 59-83.Moore, T.R. (199Oa). Gas exchanges between peatlands and theatmosphere. Canadian Geographer 34, 86-87.Moore, T. (1Y90b). Spatial and temparal variations of methaneflux from subarctichorthern boreal fens. OlobalBiogeochemical Cycles 4(1), 29-46.Moore. T.R. and R. Knowles (1987). Methanc and carbon dioxidcevolutinn from subarctic fens. Canadian Journal nf SoilScience 67, 77-81.Moore, T.R. and R. KnoGIes (199@). Methane emissions from fen,bog and swamp peatlands, Quebec. Biogeochemistry 11, 45-61.EZoraes. F. and M.A.K. Khalil (1993). Permafrost methane content:2. Modeling theory and results. Chemosphere 26(1-4), 595-607.Morrissey, L.A. (1988). Predicting the occurrence of permafrostin the Alaskan discontinuous zone with satellite darn. InSeoneset, K., (ed.), Proceedings of the Fifth <strong>International</strong><strong>Conference</strong> on Permafrost, Volume 1, 213-217. Tapir Publishcrs,Trondhcim, Norway.Morrisscy, LA., L.L. Strong, and D.H. Card (1986). Mappingpermafrost in the boreal forest with thcdatic mapper satellitedata. Photogrammetric Engineering and Remote Sensing 52(Y)1513-1520.Nakano, Y. and J. Brown (1972). Matbematical modeling andvalidation of rbc thermal regimes in tundra soils, Barrow,AlasL8:Arctic and Alpine Research 4(1), 19-38.Nakayama, T., T. Sone, and M. Fukuda (1993). Effects of climaticwarming on the active layer. In Proceedings of thc Sixth<strong>International</strong> Confercnce on Permafrost, Volume 1, 488-493.South China University of Technology Press, Wushan, Guangzhou, ,China.Nelson, F.E. (1989). Permafrost in eastern Canada: a review ofpublished maps. Physical Geography 10(3), 233-248.Nelson, F.E. and O.A. Anisimov (1993): Permafrost zonation inRussia under antbropogenic climatic change. Permafrost a;ldPeriglacial Processcs 4(3), 137-148.Nelson, F.E. and S.I. Outcalt (1987). A computational method forprediction and regionalization of permafrost. Arctic andAlpine Research 19(3), 279-288.Nelson, F.E., S.1. Outcalt. C.W. Goodwin, and K.M. Hmkel (1985).Diurnal thermal regime in a pat-covered palm, Toolik Lake,Alaska. Arctic 38(4), 310-315.Oechel, W.C. and W.D. Billings (1992). Effects of global changeon the carbon balance o€ Arctic plants and ecosystems, InChapin, F.S. 111, R.L. Jcfferies, J.F. Reynolds, G.R.Shaver, J. Svoboda, and E.W. Chu, (cds.), Arctic Ecosystems ina Changing Climate: an Ecophysiological Perspective, 139-168.Academic Press, San Diego.Oechel, W.C., S.J. Hasrings, G. Vourlitis, M. Jenkins, G.Riechers, and N. GruIke (1993). Recent change of Arctic tundraecosystems from a net carbon sink to a source. Nature 361,520-523.Qstcrkamp, T.E. (1989). Occurrence and potential importance ofsaline permafrost in Alaska. In Proceedings ul the Workshop onSaline Permafrost,, Winnepeg, Manitoba.Osterkamp, T.E. and T. Fei (1993). Potential occurrence ofpermafrost and gas hydrates in the continental shelf nearLonely. Alaska. In Proceedings of the Sixth <strong>International</strong><strong>Conference</strong> on Permafrost, Volume 1, 500-505. South ChinaUniversity of Technology Press, Wushan, Guangzhou, China.Qsterkamp, T.E. and J.P. Oosink (1991). Variations in permafrostthickness in responsc to changes in paieoclimatc. Journal ofGeophysical Research 96B(3), 4423-4434.Outcalt, S.I., K.M. Hinkel, and F.E. Nelson (1992). Spectralsignature of coupled flaw in the refreezing active layer,northern Alaska. Physical G-grllphy 13(3), 273-284.Parker, R.L. (1977). Understanding inverse theory. Annual Reviewof Earth and Planetary Sciences 5, 35-64.'Morrissey, L.A. and G.P. Livingston (1992). Methane emissionsfrom Alaska Arctic tundra: an assessment of local spatialvariability. Journal of Geophysical Research 97(DIS),16,661-16,670.Peddle, D.R. and S.E. Fraultlin (1993). Clwification ofpermafrost active layer depth from remotely sensed andtopographic evidence. Remote Sensing of Environment 44(1),67-80.

Pissan, A. (1990). Advances in periglacial geomorphology.Zeitschrift fur Gcomorphologie Supplcmcntband 79, 119-1 3 1Rasmussen, R.A., M.A.K. Khalil, and F. Moraes (1993). Permafrostmethane content: 1. Experimental data from sites in northernAlaska. Chemosphere 26(1-4), 591-594.Riscborough, D.W. (1990). Soil latent heat as a filter of theclimate signal in permafrost. In Proceedings of the FifthCanadian Permafrost Confercncc, 199-205. Centre d'etudesnordiques, Universite LavallNational Research Council ofCanada, QuBbec.Riseborough, D.W. and M.W. Smith C1993). Modelling permafrostresponse to climate change and climatc variability. InLunardinai, V.J. and S.L. Bowen, (cds.), Proceedings, Fw~rth<strong>International</strong> Symposium on Thermal Engineering & Sciencc forCold Regions, 179-187. US Army Cold Regions Research andEngineering Laboratory, SFcial Report 93-21, Hannvcr. NH.Solomon, A.M. and H.H. Shugart (eds. 1993). Vegetation Dynamics &Global Change. 338 pp. Chapman & Hall, New York.Solovicv, P.A. (1959). Kriolitozana severnoi chastitcno-Amginskogo mezhdurechia. 144 pp, Izdatelslvo AN SSSR,Moscow.Stevens, W.K. (1993). In new data on climate changes, decades,not centuries, count. The New York Times December 7, C4.Street, R.B. and P.1, Melnikov (1990). Seasonal snow cnvcr, iceand permqfmst. In Tega'rt, W.J.McG., G.W. Sheldon, and D.CGriffiths, (eds.), Climate Change. The IPCC ImpactsAsscssment, 7.1-7.33. Australian Government PublishingService.Stuart, A. (1986). Thc development of a spatial permafrost modelfor northern Canada and its application to scenarios ofclimate change. KelResearch Corporation, Downsview, Ontario.Roots, E.F. (1989). Cliplate changc. high latitude regionsClimatic Change 15(1/2), 223-253.Rothman, D.S. and D. Chapman (1993). A critical analysis ofclimate changc policy research. Contempnrary Policy Issucs11(1), 88-98,Rohse, W.R., P.F. Mills, and R.B. Stcwar! (1977). Evapormon inhigh latitudes. Water Resources Research 13. 909-914.Schmitt, E. (1y93). Global cllmatic change and some possiblegeomorphological and ecological effects in arctic permnfrnstenvironmcnts, Isfjorden and Liefcdcfjorden, northernSpitsbcrgcn. In Proceedings of thc Sixtb <strong>International</strong>Cunferencc on Permafrost, Volume 1, 544-549. South ChinaUniversity of Technology Press, Wushan, Guangzhou, China,Shaver, G.R., W.D. Billings, F.S. Chapin 111, A.E. Gihlln. K.J.Nadclhoffer, W.C. Oechel, and E.B. Rastetter (1992: Globalchange and the carbon balance of Arctic ecosystcms. Biosclencc42, 433-441.Slaughter, C.W. and R.I. Hartzrnann (1993). Hydrologic and watcrquality charactstistics of a degrading open-system pingo. InProceedings of the Sixth <strong>International</strong> <strong>Conference</strong> onPermafrost, Volume 1, 574-579. South China University ofTechnology Press, Wushan, Guangzhou. China,Slaughter, C.W., J.W. Hilgcrt, and E.H. Culp (1983). Summerstreamflow and scdimcnt yield from discontinunus permafrostheadwater catchments. In Proceedings of the Fourth<strong>International</strong> <strong>Conference</strong> on Permafrost, Volume 1, 1172- 1 177.National Academy Press, Washington. DC.Smith, M.W. and D.W. Riseborough (1983). Pcrrntlfrost sensitivityto climatic change. In Proceedings of thc Fourth <strong>International</strong><strong>Conference</strong> on Permafrost, Volume I, 1178-1183. NationalAcademy Press, Washington, D.C.Solomatin, V.I. (1992). Deoecologia Severa, 270 pp. IzdatelstvoMGU, Moscow.Solomon, A.M. (1986). Transient response of forests toC02-induced climate chhnge: simulation modeling experiments ineastern North America. Oecologia 68, 567-519.Sugden, D. (1992). Global warming: Antarctic ice sheets at risk?Nature 359(6398), 775-776.Thic, 1. (1974). Distribution and thawing of permafrost in thesouthern part of the discontinuous zonc in Manitoba. Arctic27, 1S9-2M).Thomas, G. and P.R. Rowntree (1992). The boreal forests endclimate. Quarterly Journal of the Royal Mctcorolopy Socicty11SB(505), 469-497.Timoney, K.P., G.H. La Roi, S.C. Zoltai, and A.L. Robinson(1992). The high subarctic forest-tundra of northwcsternCanada: position, width, and vegetation gradients in relationto climate. Arctic 45(I). 1-9.Torn, M.S. and E.S. Chapin I11 (1993). Environmental and bioticcontrols over methane flux from Arctic tundra. Chemosphere2q1-4). 357-368.Ulrich, R. and L. King (1993). Influence of mountain permafroston construction in the Zugspitze Mountains, Bavarian Alps,Germany. In Proceedings oi the Sixth <strong>International</strong> <strong>Conference</strong>on Permafrost, Volume 1, 625430. Proceedings of the Sixth<strong>International</strong> Cunference on Permafrost, Wushan, Guangzhou,China.VanC-nberghe, J. and A. Pissart (1993). Permafrost changcs inEuropc during the last glacial. Permafrost and PeriglacialProcesses 4(3), 121-135.Vinson, T.S. -.n? D.W. Hayley (1990). Global Warming and Clin-xt!c'Change. Journal of Cold Regions Engineering 4, 73 pp.Vourlitis, O.L., W.C. Oechel, S.J. Hstings, and h1.A. Jenkins(1993). The effect of soil moisture and thaw depth on CH4 fluxfrom wet coastal tundra ecosystems on thc North Slope ofAlaska. Chemosphsrc 26(1-4), 329-337.Vyalov, S.S., A.S. Gcrasimov, A,J. Zolotar', and S.M. Fotiev(1993). Ensuring structural stability and durability inpermafrost ground areas at global warming of the Earth'sclimate. In Proceedings of the Sixth <strong>International</strong> <strong>Conference</strong>on Permafrost, Volume 1, 955-960. South China University ofTechnology Press, Wushan, Guanglhou, China.1004

Waelbrocck, C. (1993). Climate-soil processes in the presence ofpermafrost: a systems modelling approach. Ecological Modelling69(3,4), 185-225.Wal& J.E. (1993). The elusive Arctic warming. Nature 36U6410).300-301. 'Wang, B. (1993). A numerical simulation of coastal retreat andpermafrost conditions, Mackenzie Delta Region, Canada. InPr-ings of the Sixth <strong>International</strong> <strong>Conference</strong> onPermafrost, Volume 1, 664-669. South China University ofl'echnology Press, Wusban, Changzhou. China.Weller, a. and B. Holmgren (1974). The microclimates of theAr& tundra. Journal of Applied Meteorology 13, 854-862.'Whalen, S.C. and W.S. Reeburgh (1988). A methane flux time seriesfor tundra environments. cilobal Biogeochemical Cycles 2,399410.Williams, P.J. and M.W. Smith (1989). The Frozen Earth:Fundamentals of Qeocryology. 306 pp. Cambridge UniversityPress, New York.Winh, T.E. (1993). Developments in US Policy toward globalclimate change. US Department of State dispatch 4(22), 399.Woo, M.-k. (1986). Permafrost hydrology in North America.Atmosphere4cean 24, 201-234.Woo, M.-k. (1990). Consequences of climatic change for hydrologyin permafrost zones. Journal of Cold Regions Engineering 4(1),15-20.Woo, M.4 (1992). Impacts of climatic variability and change onCanadian wetlands. Canadian Water Resources Journal 17, 63-69.Woo, M.-k, A.G. Lewkowicz, and W.R. Rouse (1992). ResponseCanadian permafrost environment to climatic change. PhysicalGeography 13(4), 287-317.of theYarie, J. and K. Van Cleve (1991). Changes in the sourcelsinkrelationships of the Alaskan boreal forest as a result ofclimatic warming. fn Weller, a., C.L. Wilsan, and R.A.B.Sevcrin, (eds.), Proceedings, <strong>International</strong> <strong>Conference</strong> ttwRole of the Polar Regions in Global Change, Volume 1143&439. Geophysical Institute, Univcrsity of Alaska.Fairbanks, AK.amolotchikova, S.A. (1988). Mean annual temperature of groundsin East sibcrh. h Senneset, K., (ea.), Proceedings of theFifth <strong>International</strong> <strong>Conference</strong> on Permafrost, 237-240. TapirPublishers, Trondheim, Norway.Zhang, T. and T.E. Osterkamp (1993). Changing climate andPrmafrOst temperatUrCS in the Alaskan Arctic. In Proceedingsof the Sixth <strong>International</strong> <strong>Conference</strong> on Permafrost, Volume 1,783-788. south China University of Technology Press, Wushan,Guangzhou, China.ZOlbi, S.C. and D.H. Vitt (1990). Holocene climatic change andthe distribution of peatlands in western interior Canada.Quaternary Research 33(2), 231-240.1005

PRESENT HUMAN INDUCED CLIMATIC CHANGE AND CRYOECOI.,OGYMaria K. GavrilovaPermafrost Institute, Siberian Rranch, Russian Academy of Scicnccs, Yakutsk, 677018, RussiaPermafrost is a conscqucncc of cold climate. But the climate is not constant, It changes continuously.Until recently cryogenic processes were altering under effect of natural climate changahility. M:ln's af"fect changing microclimate of locality has incrcascd rcccntly. Under large-scale activity it leads to CBI:IStrophes.A great danger may be caused by climatc warming under atmosphere pollution. It will cause reinforcementof thermokarast phenomena.INTRODUCTION-- - - .Climate of any place undergoes continuous changc bothshort-term and long-term (Gavrilova, 1992 a). At prcscnt, mancan not affect large-scale climate forming factors such as cosmic,astronomic, geological and others, but can only influence to someextent the geographic factors as the atmosphere transparency andterrain.Predicted macroclimatic Change resulting from hurnan-induccdatmosphcric pollution and an increase in carbon dioxidecontent in the atmosphere in the future. Meanwhilc, permafrost rcscarchsare concerned about human-induced changes of a lowerscale, i.e. mesa-, micro- and nanoclimatic changcs causcd by tcrrainand surface disturbances.CI,TMATE AND NATURAL CRYOGENIC PHENOMENA. . -.. .. .- . . ... . "- ..... . - . ". . ...." . .-. .Pcrmatrost is characterized not only by negative temperatures,but thc presence of cementing L'rozen water ranging fromnllcroparliclcr to huge Ice bodies as well. sharp temperature fluctuallonsand phase transitions of watcr givc risc to cryogcnic processessuch ac trost fracturing, frost hcaving, ground icc melting,ground subqidence, thermokarst, thermal abrasion, thermaldenudation, thermal erosiori, solifluction, formation of open andclosed taliks and others.These processes are active under natural conditions, and theyarc associated with gcncral climatic and othcr physico-gcographicalconditions. In recent ycars, it has hccn revcaled that they followa cyclic pattern coinciding with short-term (tens to hundrcds ofycnrs) clirnatc fluctuations. For cxamplc, during the years of heavyprecipitation the fate of chel-mal erosion incrcascs, and many newthcrmokarst lakes form, while during dry years the lakcs degrade,tilliks freeze below lakes, pingos form actively (Bosikov, IYYO).I)urine \v:lrnmcr periods ground water sccpagc and icing fbrm;~~lonoccur intcnscly. Cooling is accompanied hv upward devc!opmcnt ofrcliel., while warming by 1iounw;ird dcwloprncnt (Shpolinnsk:li,l.1992).Cyclic:ll chnnpcc 111 cryopcnic procc\ws occur on ;I ~cawnc~llvhasi.; nc well. For cs;lmple. 111 :II-CRS 01 pcI-cnn1;11 Ircczlng ol groun,l.landslides are more active in fall (when frozcn material havcthawed), and in areas of seasonal freezing In spring after snow rnciting(Rosenbaum. Mudrov and Tumel, 1990). Intensified fracturingtakes place from November to March, i.e. when air and surfacetemperatures arc the lowest (Sukhodrovsky, 1969).MICROCI.IMATE AND ANTHROPOGENIC CRYOGENKC.... ". . .." . . .PHENOMENA. ..Under natural conditions the cryogenic pl-ocesses take placeslowly and in a smallcr cxtcnt, and the ratc of rccovcry is hichcr.Human activities cause acceleration of cryogenic processes, an increasein tile Rrea of adverse impacts, and the pcrmatrost tcrrainvery oftcn lose? its ability to recovery. A new gcosystcm forms. andpcrmafrost may disappcar in somc areas.Dernrestation accompanying development of the region, rcrnovalof vcgctation and snow cover, irrigation and drainage improvemen:$cause. significant changes in the heat and moisture balancesboth at the ground surface and in the upper layer 01- thcground. This can lead to local permafrosl degradntlon oraggradation, cspccially in thc island pcrmafrost ?one, Tor example,in the Baiktll-Amur Railway Region.Invesilgations on heat balance and microclimate show thatdclbrcstation :csults in intcnsification of seasonal thawing: in theNorth it is by a tactor of I .5, in Ccntral Yakutia by a factor of 2,and in southern Siberia soils frozcn in winter thaw cntircly. In theNorth removal of the soil cover (as the iorest litter) causes an addi-

tional increase under a disturbed surface freeze in winter more intensely.Observations in southern Yakutia show that if the mossand snow covers are removed, the upper soil temperatures are twiceas small in winter (Gavrilova, 1977,1973, 1978).The effects of human activities on permafrost depend on thescale and duration of the impact, as well as the climatic zone, terrainconditions, the nature ofthe warm and cold seasons, trends innaturally occuring processes. If human-induced and natural processescoincide in time, cryogenic processes intensify, if not, theymay be wcakened or even damped. On the whole, in cold Yakutiaclearing oi'trecs results in a lowering of the upper ground tcmperaturcs,while in warmer southern Siberia it causes permafrost to disappearin clear-cut areas.CATASTROPHIC.- CONSEQUENCES - OF MAN ACTIVITY 'INTHE CKYOLITHOZONE"". -"_ -In the permafrost areas, detrimental impacts of human activitiesbecome evident immediately. Deep impassable ruts and bogsformed along the paved roads in rural areas are the first evidences.If the early settlers of Siberia could simply displace the heaved gatesand fences, nowadays, this can not be applied to large blocks ofmulti-storey buildings in cites and large settlements, requiring urgentmaintenance measure;.Well-known Rykovsky fish-processing plant located in ArcticYakutia, which supplied tish to the entire country during the secondworld war, will have to be abandoned within IO years. The icecontcnt of the ground under the Bykovsky settlement is 90%, theice bodies are up to IO rn wide at the top, the thickness of a coveringground layer is only 0.6 to 1 m, average natural thaw depth is0.6 m (Grigoriev, et al., 1990). Disturbance of the upper 5 to 15 cmmoss and turf cover has initiated cryogenic processcs. The rate Ofshore erosion is up to 10 m / yr.Warm permafrost is particularly sensitive to landscape disturbance,as in wcstern Siberia and the Far East. At exploratorydrill sites in northern Chukotka active layer thickness increases bya factor of 2 to 3 as compared to adjacent undisturbed areas, whenvcgetational and soil cover are removed, and the thawed zone enlargesaround (Kotov and Maslov, 1992). Funnel-shaped hollowsform in (he mouth of deep oil and gas exploratory boreholes causedby melting of blocks of ground ice,Big towns generally favor the preservation of permafrost andeven strcngthcn it duc to the existence of an insulating "culturallayer" and shadowing cflect of buildings. In the city of Yakutsk, ata depth of 10 m the ground temperatures lower to 4 to SoC belowzero during 4 to 5 years, if proper construction and maintenanceapproaches are used (Popenko et al., 1990).Problems arise if construction is performed carelessly neglectingstandards and codes for placing of foundations, constructionmaterials arc of poor quality, maintenance is made incorrectly,foundations are flooded by industrial and domestic waste. InYakutsk, 40% of large-panel residential buildings require urgentrepair measures, and 20% may collapse soon.Settlements located in the permafrost zone are sources ofchemical contamination of the environment makarov, 1992). Poorpermeability of permafrost, small depth of thaw, low temperaturesand other factors facilitate the decay of chemical substances. con-centrated solutions arq transported by surface water contaminatehuge areas. In some locations weak platie-frozen pound and taliksmay develop. Salinized frozen ground erodes the concrete foundations.Contaminated frozen ground becomes an awressive factarIimpeding man's activity,There is increasing concern about construction of linear structurea (railwayes, highways, pipelines, power transmission lines, irrigationchannels, etc.). The reason is that during construction andmaintenance not only the construction sites are disturbed, but adjacentareas as well. In many casm access roads have not been builtin advance, and drainage facilities have been made poorly. Jt hasresulted in deepcr thawing, ground subsidence, thermal erosion,ground water seepage, icing, paludification and landslides.for the Baikal-Amur Railway the maintenance casts have alreadycxcceded the construction costs. Zones of thawed ground develop along oil and gas pipelines. In central Yakutia these zones are0.75 to 1.5 m wide (Popov, 1992). Reclamation channels are highlysusceptible, and when they are destroyed, it results in that agriculturallands become unusable because of water erosion of' soils. Thewater in reclamation channels causes melting of ice resulting in depressionsand in cave-ins, and water is discharged to the formeddepressions. The bed of a channel made for basin irrigation. in centralYakutia was designed to be 2 m wide 20 years later it is 10 to 20m wide (Ugarov, 1992; Chang and Melkozerav, 1992).Rccently, drainage and irrigation imprpvements have been in+troduced in agricultural areas of Traisbaikal, Yakutia andChukotka because of dry climates there. These improvements hawbeen applied to 8.5% of agricultural lands in central Yakutia. Theyhave already resulted in grw production of fodder grass 15-18%,potatoes 80%, vegetables 100%. However, improper use of irrigationmethods (excessive irrigation in particular ) results inpaludification,'formation of depressions and cave-ins, heaving asthe permafrost is impervious to water. One quater of reclaimedlands have already had to be abandonad. Restoration is requiredthrough tilling of depressions, cutting pf mounds and grading.More frequent cycles of irrigation with smaller amount Of Waterhave been recommended by the researchgrs of the Permafrost Instituteto prevent water from reaching the pound icetable(1ntensifikatsia ..., 1988; Gavriliev. 1991). ConStruCtion ofhydroelectric power plants exerts harmful infhencts 4" environment(Bianov et ai., 1992). Water reservoirs must impoundedfor power plants because the riwr discharge is ,not uniformthroughout the year. However, forests and meadows are flooded,water quality deteriorates, hydrological and hydrogtalogical regimeschange, earth materials are redeposited. Impnunded Waterwarms up the reservoir bed and shores.The Vilui hydroelectric power plant was the first ,proict Onpermafrost. The water impoundment lasted 7 years (1967-1973). Inthe first year the depth of thaw in the water reservoir bed increasedto 5 m, in the second year 7 m, in :he third year 8.5 m (ICamenskY Ctal., 1973). At present, thickness of the thawed layer is 20 m(Kamensky and Olovin, 1990). Fifteen-year observations(1977-1989) show that the retreat rate of low ice content shores is 1to 2 m / yr, that of ice-rich shores is 4 to 5 m / yr (Konstantinov,1992).The modification of shorelines in the Amguema hydroelectricpower plant which is under construction in Chukotka has been1007

~ ofpredicted to owur a8 following : during the tiit year of operationthe upper 1.5 to 2 m layer of unconsolidated rocks will be subjectedto thermal erosion, in 2-3 years ground ice will extensively melt,later the water reservoir will enlarge 5% due to thawing ofpermafrost; finally vat and moss cover will come to the surfacethat can causc,damage to the power plant units(Tishin, 1990).PREDICTED CLIMATE WARMING AND THERMOKARSTNowadays mankind is concerned about predicted man+ducedglobal climatic change (warming) related to atmospheric pollutionresulting from industrial development. It is predicted that inthe middle of the 21st century global air temperatures may rise by 3k 1.5"c (Antropogenie izmencnia klimata, 1987). Climatic warmingis forecasted to be greater at high latitudes, i.e. in thepermafrost zone (5 to 10°C per year) (Gavrilova, 1992 b). It is ob.vious that the earth surface and uppcr soil-ground will immediately response to thcsc sharp changes. Cryogenic procesases will develop extensively as they arc dependent upon both climatic conditionsand 1ocal.buman-inductd impacts (Kachurin, 1961; Shur,1977; Romanovrky, 1977).Thennokarst development has the following stages: melting ofground ice, fornation of a depression, filling with water, thermaland physical impact of water. drainage of a jake, diminishing of athermokarst feature.Melting of ice may commence due either climatic warming orlocal disturbance of an insulating layer (vegetational and soil cov-WS).Evidences of relict thermokarst relief are found in western andcasteh Siberia which developed 10 thousand years ago. It coincidedin time with the period of Late Pleistocene temperature maximum.The second phase of intensive thetmokarst development (ofa smaller extent) took place 5 to 6 thousand years ago in centralYakutia (Yakutian temperature maximum) (Soloviev.1959) and 2.5to 3 thowand yea- ago in western Siberia (Late Holocene warming)Gcocccdogia Sever@, 1992). The thermokarat features that canbe observed at present in antral Yakutia are 200 to 300 years old,and those in western Sibcrip wd Chukotka are 500 to 700 years oldvontirdiaro, 1972). Although it is believed that during the past 300to 400 years the climate (air temperatures) has not been changing,secular variations in humidity could have effect.Thennokarst development depends on all climatic factors:cloud patterns (radiative heating), temperature (general heat background),precipitation (insulating effect of snowcover, water balanceconstituents), air humidity and winds (evaporation rate ) andothers (Efimov. 1950; Nemvhinov, 1958; Sploviev, 1961).Precipitation is 'a very important factor in development ofthermokarst features and associated lakes. Precipitation changes ona alasses had less watmduring the periods of 3890-3900 and 1914to 1947, while between these two periods they had relatively greateramount of water (Soloviev, 1961). Direct relationship with precipitationis found through meteorological research of precipitation variations(Gavrilova, 1987k Bosikov_(l991) ghes aore &tailedchronolcgy of high and low water prccipitation data as well.Chigir (Geologia Scvera, 1992) bclievs that forecasts of changein thcmyogeasystem should take into account the comb'ination ofsecular climatic sequences: warm and humid, warm and dry. coldand humid, cold and dry. The cryogenic processes will developdifferently in different periods.It is predicted that by the middle of the 21st century theamount of precipitation will increase at high ' latitudes(Antropogenie izmenenia Klimata, 1987).CONCLUSIONThus, if global climate does change significantly in future, thecryogenic processes will develop intensely, thermokarst in particular.Thermokarst development will be caused both bymacroclimatic and microclimtic changes, i.e. global warming, increasein precipitation amount, and extensive human activities.REFERENCESBudyko, M. and Y.Izrael (1987).,Antropogenie izmenenia klimata.407 p. Gidromoteoizdat, Leningrad.Bianov, G., M.Malyshev and I. Sergeev (1992). Ecological aspectsuse of hydroencrgy in the permafrost zone: Ratsionalnoieprirodopolzovanie v kriolitozone, 46-50 pp. Nauka, Moscow.Bosikov, N.(1990). Changability of general humidity in the areaand dynamics of cryogenic processes: Ratsioalnoieprirodopolzovanie v kriolitozone, 58 p. IMZ SO AN SSSR,Yakutsk.Chang, R, and G. Melkozerov (1992). Peculiarities of structure andperformance of reclamation channels in Yakutia:Ratsionalnoie prirodopolzovanie v kriolitozone, 86-92 pp.Nauka, Moscow.Efimov, A.(1950). Development of thermokarst lakes in centralYakutia: Isledovanie Vechnoi Merzloty v Takutskoi Respublike(V01.2), 98-1 14 pp. Izdatelstvo AN SSSR, Moscow.Gavrilova, M.(1973). Climate of central Yakutia (Klimat' Tsentralnoi Iakutii). 2nd edition. 120 p. Knizhnoe izdatelstvo,Yakutsk.Gavribva, M. (1978). Climate and Perennial freezing of ground(Klimat i mnogoletnee promerzanie gornykh porod 1. 214 p.Nauka, Novosibirsk.Gavrilova, M.(1987). Analysis of changes in natural climatic conditionsin Yakutia to the beginning of the next century:Prirodnye uslovia osvaevaemykh regionov Sibiri. 146-1 59 pp.IMZ SO SSSR, Yakutsk.Gavrilova, M.(1992 a). Impact of global climatic change on thepermafrost zone: Ratsionalnoie prirodopolzovanie vkriolitozone, 4-8 pp. Nauka, Moscow.Gavrilyev, p. (1991). Reclamation of permafrost areas in Yakutia(Melioratsia merzlotnykh zcmel v Iakutii). 182 p. Nauka,Novosibirsk.Solomatin, V.(1992). Geoecologia Scvera (vvcdenie vgeokrioecologiu) (Northern geoecology. Introduction intogeocryecology). 270 p. Tzdatelstvo MGU, Moscow. .~Crigoriev, M., 1.Pozdniakov and V.Romanov (1SIM). Cryogenicprocpx,.in im. complex. gmcf-iiausid by man-induced disturbanceof the surface: Ratsionalnoie prirodopolzovanic vkriolitozone, 60 p. ZMZ SO RAN, Yakutsk.,IntensiMratsia koiioproizvodstva v zone reki Amgi.Recomendatsii (Intensification of production of fodder in theAmga River area) (1988). 40 p. Knizhnoie izdatclstvo, Yakutsk.Kachurin, S. (1961). Thermokarst development in the USSR terri-

tory (Termokarst na territorii SSSR). 292 p. Izdatelstvo ANSSSR, Moscow..Kamensky et al. (1973). Guide. North-western Yakutia (Putevoditel. Severo-Zapadnaia Iakutia). 45 p. IMZ SO ANSSSR , Yakutsk.Kanmensky, R. and B. Olovin (1990). Changes in climatic andpermafrost conditions in the zone of thermal impact from thelarge water reservoirs in the Far North: Ratsionalnoieprirodopolzovanie v kriolitozone, 20-21 pp. IMZ SO ANSSSR. Yakutsk.Konstantinov, L(1992). Shore erosion of the water rcscrvoir of theVilui hydroelectric power plant: Ratsionalnoieprirodopolzovanie v kriolitozone, 57-63 pp. Nauka, Moscow.Kotov, A. and V. Maslov (1992). Man-induced changes in thepermafrost zone of the Nizhneanadyr depression:Ratsionalnoie prirodopolzovanie v kriolitozone, 159-165 pp.Nauka, Moscow.Makarov, V.(1992). Ecological-geochemical problcms in thepermafrost zone in Yakutia: Ratsionalnoie prirodopolzovaniev kriolitozone, 133-139 pp. Nauka, Moscow.Nemchinov, A.(1958). Cyclical variations of water levels in lakes incentral Yakutia: Nauchnie soobshenia Yakutskogo filiala ANSSSR (Vol.l), 30-37 pp. Knizhnoo izdatelstvo, Yakutsk.Popenko, H., A. Petrov, and S. Petrova (1990). Man-induced impactson frozen Quaternary deposits in Yakutsk:Ratsionalnoieprirodopolzovanie v kriolitozone, 96 p. IMZ SO AN SSSR,Yakutsk.Popov, V. (1992). Examination of a gas pipeline laid in the area ofice wedges: Ratsionalnoie prirodopolzovanie v kriolitozone,102-107 gp. Nauka, Moscow.Rornanovsky, N. (1977).Theory of thermokarst: VestnikMoskovskogo Universiteta, seria geologia, No.1, 65-71 PP.Rosenbaum, G, Y. mudrov and N.Tume1 (1990). Effect of climaticchange on solifluction: Ratsionalnoie prirodopolzovanic vkriolitozonc, 12 p. IMZ SO AN SSSR, Yakutsk.Shpolianskaia, N. (1992). The permafrost zone as an indicator ofpresent clirnatit-. change (Western Siberia): Ralsionalnniepriodopolzovanie v kriolitozone, 30-35 pp. Nauka, Moscow.Shur, Y. (1977). Thermokarst (Thermokarst). 81 p. Nedra,Moscow.Solovicv, P. (1959). Permafrost in nothern part of the Lene-Amgaintcrfluvial (Kriolitozona severnoi chasti Leno-Amginskogomezhdurechia). 144 p. Izdatelstvo AN SSSR, Moscow.Solviev, P. (1961). Cyclical variations of watcr content in thc alasslakes in central Yakutia related to climatic fluctuations: ,Voprosy geografii Yakutii, 48-54 pp. Knizhnoe izdatelstvo,Yakutsk.Sukhodrovsky, V.( 1969). Present relief formation in ccntralYakutia: Voprosy geografii Yakutii (V01.5), 148-159 pp.Knizhnoe izdatelstvo, Yakutsk.Tishin, M. (1990). Prediction of shoreline modifications in the waterreservoir of the Amguema hydroelectric power plant:Ratsionalnoie prirodopolzovanie v kriolitozonc. 26 p. 1MZ SOAN SSSR, Yakutsk.Tomirdiaro, S. (1992). Permafros; and development of mountain-' ous and lowland arcas (Vechnaia merzlota i osvoenie gornykhstran i nizmennostei ). 174 p. Magadan.Ugarov, 1.(1992). Methods of construction of irrigation. channels in,areas with ground ice occuring close to the surface:Ratsionalnoie prirodopolzovanie v Kriolitozone, 92-94 pp.Nauka, Moscow.1009 '

RECENT PERMAFROST DEGRADATION ALONE THE QINGHAI-TIBET HIGHWAYCheng Guodong", Huang Xiaoming** and Kang Xingcheng*"Lanzhou Institute of Glaciology and Geocryology,Chinese Academy of Sciences, Lanzhou, China**Northwest Institute of Chinese Railway Academyof Sciences, Lanzhou, ChinaThis paper presents some observation results of climatic change and the responseof permafrost along the Qinghai-Tibet Highway during the last decades, anddiscusses a problem about the Qinghai-Tibet Plateau and initiation of climaticchange.Connecting Lhasa and Xining, the respectivecapitals of the Tibet Autonomous Region andQinghai Province, the Qinghai-Tibet Highwaytraverses more than 560 km of high-altitudepermafrost. Six meteorological stations arelocated along the highway; of these, Wudaoliang,Fenghuo Shan, and Tuotuohe are in the area ofthe plateau underlain by permafrost.Instrumental records show a coincidence oftrends in climatic changes between Xining inwestern China and those l.n eastern China andthe rest of the globe (Fig.2). The worldwidewarming that began in the 1880s reached amaximum in the 1940s (Zhang, 1976; Cheng, 1984).The instrumental records of five meteorologicalstations along the Qinghai-Tibet Nighway showdecreasing air temperatures during the 19609,and a pronounced reversal of this trend in the1970s. The only exception to this trend wasthe Wudaoliang station during the early 1980s(Fiy.3), although the reasons for this situationare not understood at present.PERMAFROST DEGRADATION ALONG THE QINGHAI-TIBETHIGHWAYEvidence of permafrost degradation has beenfound in areas of both discontinuous andcontinuous permafrost in the vicinity of thehighway, Temperature measurements and drillholesalong the highway have demonstrated in manylocalities that near its lower altitudinal limitp~rmafrost is separated from the active layerby a thawed layer (Wang 1993).Another line of evidence in the region is arise in the base of the permafrost. Groundtemperature measurements from 1974 to 1989 ina borehole in the northern Jingxian valley atb530 m shows the permafrost base rose 5 m in15 years. The mean annual ground temperatureat A depth of 20 m increased by 0.2-0.3"Cduring the same period (Table 2). Moreover,some permafrost bodies have disappeared entirely,zic5s-Fig..l Locations of meteorological stationsalong the Qinghai-Tibet Highway

Table 1. Elevation of meteorological station.8 along the Qinghai-Tibet HighwayStation Golmud Wudaoliang Fenghuo Shan Tuotuohe NagquElevationm, a.s.1.2806 4800 4700 4507 36580.40.30.20.10.0Fig.2 Comparison of average temperature changes since 1870 in the world and in Xining, ChinaSuccessive 5-year means expressed as departures from the means for 1880-84 in the wholeworld and successive 5-year means expressed as temperature grades in China(Modified from Zhang, 1976)10-2-6It2""3"I="- 101950 1955 1960 1965 1970 1975 1980 1985 19901. Lhasa 2. Nagqu 3. Wudaoliang 4. Tuotuohe 5. GolmudFig.3 5-yaar running mean curves of air temperature on the Qinghai-Tibet PlateauTable 2. Rising of permafrost base at thenorth of Jingxian valleyarea attains thicknesses- of about 100 m. RecordsDepth of permafrost 1514 12 101992, mean annual temperature at a depth of 15 mbase (m)increased by 0.2'C, by 0.4'C at 10 m, by 0.8'Cat 5 m, and by 0,4"C at a depth of 3 m (Fig.4).Year 1974 1979 1985 1989 from Fenghuo Shan show a warming trend in bothair and ground temperatures, Between 1980 and(after Wang)Ground temperature measurements in two otherboreholes near Fenghuo Shan also illustrate theGround temperature measurements from 1975 towarming trend in permafrost in this area. Mean1989 in a borehole at 4670 m a.s.1. near HighwayMaintenance Squad No.124, for example, illustrateannual temperature at the 15 m level in a boreholeon a south-facing slope increased by 0.1'C'the process of thinning and disappearance ofpermafrost near its former lower elevational 'limit (Table 3).from 1982 to 1992 (Fig.5). At the correspondingAreas of the Plateau underlain continuouslyby permafrost have also been affected by degra-dation. Fenghuo Shan station was established in1961 by the Northwest Institute of Railways ofthe Academy of Sciences. Permafrost in thisdepth is a borehole on a north-facing slope,mean annual temperature increased by a similaramount over a 28 year period (Fig.6).' 1011

.Table 3. Ground temperature in borehole 124-4("c)1 2 3 4 5 6 7 8 11 15 190 0.2 3.5 0.5 0.879-08-05 0.6 -0.1 -0.3 -0.1 -0.2 -0.20.184-08-07 -0.2 -0.3 -0.3 -0.10.1 o 0.2 0.7 0.8 0.989-07-26 0.7 0.1 0.10 0.1, 0.1 0.2 0.2 0.6 0.0 0.9(after Wang) parr1980 81 82 83 '84 85 86 87 88 89 I990 91 92-2.0 -- 2.5 -I 4II I II I I I I I l l I-3.0-Fig.4 Mean annual air and groundTsmpmmtvm ('01 -2 3temperatures at Fenghuo Shan StationTomparature("C)0,2 -3 4I I I5-IO -15 -20 -25 -30 -Fig.5 Mean annual ground temperature. deep borehole at Fenghuo Shanin 20 rn35-Fig.6 Mean annual ground temperaturedeep borehole at Fenghuo Shanin 35 m

REFERENCESThe warm and colkl periods in eastern China,reconstructed for the last 500 years by meansof ;henology, are comparable to those obtainedfor the Qilian Mountains through dendroclimatologicalmethods. Based dn comparisons betweeneastern China and the Qilian Mountains (Table4),the onset of both warm and cold pericds averages15 years earlier in the latter (Zhang et al.,1976, 1981; Cheng, 1YR4; Tang and Li, 1992).The air-temperature grade of the Qinghai-Tibet Plateau and the warm-cold periods in theQilitln Mountains have been reconstructed, basedon tree-ring data. Comparison between these datasets revealed that the relatively cold and warmperiods in Tibet were about 10-40 years, onaverage about 30 years earlier than those inthe Qilian Mountains (F'ig.7), Accordingly, ithas been suggested that the Qinghai-TibetPlateau may be an important source region forclimatic variations lasting 30-100 years (Tangand Li, 1992).Cheng Guodong, (1984) Study of climate change inthe permafrost regions of China - A reviewin Final Proceedings of the 4th <strong>International</strong><strong>Conference</strong> on Permafrost, National AcademyPress, Washington, D.C., pp.139-144.Tang Maocang and Li Cunqiang, (1992) An analysison the Qinghai-Xizang Plateau as "a distrubingsource region" for climatic changes, inProceedings of the First Symposium on theQinghai-Xizang Plateau, Science Press,pp. 42-48.Wang Shaoling, (1993) Permafrost changes alongthe Qinghai-Xizang Highway during the lastdecades, Arid Land Geography, Vo1.16, No.1,pp. 1-8.Zhang Jiacheng et al., (1976) Climatic changesand causes, Science Press, Beijing, pp.288.Zhang Xiangong, Zhao Qin and Xu Ruizhen, (1981)Tree rings in Qilian Shan and the trend ofclimatic change in China, in Proceedings ofNational <strong>Conference</strong> on Climatic Change,Science Press, Beijing, pp.26-35.Fig.7 Comparison between the air temperaturegrade of the Qinghai-Tibet Plateau andthe warm-cold periods in QilianMountains during the last thousandyears (after Tang and Li)black - warm period:empty - cold period.Table 4. Comparison of cold and warm periods between Qilian Mountains and east ChinaPeriodColdEast China(from phenology)1470-1520 A,D.1620-17201840-1890After 1945Qilian Mountains(from dendroclimatology)1428-1537 A,D.1622-17401797-1870after 1924Difference ofmedian year.12-1 231Warm1550-1600 A.D.1720-18301916-19451538-1621 A.D.1741-17961871-1923-7633

RESEARCH ON P EWROST AND PERICLACIAL PROCESSESIN MOUNTAIN AREAS - STATUS AND FERSPECTFJESWilfried HaeberliLaboratory of Hydraulics, Hydrology and Glaciology (VAW)Swiss Federal Institute of Technology (ETH), Zurich-SwitzerlandIWRODUCTIONDuring the past decades, intensified human activitytogether with the sensitive reaction of snow and iceto ongoing and potential future warming trends led toremarkably increased awareness not only concerningthe scientific interest, beauty and vulnerability of coldmountain rages but also with regard to environmentalaspects, technical problems and natural hazardsencountered in such remote areas. In view of this development.the <strong>International</strong> Permafrost Association(lPA] established a working group on mouniain pma-@st at the Fifth <strong>International</strong> <strong>Conference</strong> on Permali-ost.held at Trondheim, Norway. Within the frameworkof this activity and in cooperation with the PAworking group on periglacial enuironments, internationalworkshops were organized at Interlaken.Switzerland, in 1991 and at Calgary, Canada, in 1992.These workshops not only enabled exchange of experienceamong specialists from various parts of the worldbut also aimed at assessing the state of knowledge inthe field and at formulating research recornmendationsas well as internationally co-ordinated projectsfor the future. The following text attempts at drawingthe main conclusions from these reflections.STAlVSA series of situation reports were compiled for theInterlaken workshop. In addition, a review on mountainpermafrost and climatic change was written forthe IPA working group on present global change andpermafrost and discussed at the Interlaken workshop.These reports reflect a broad consensus among expertsand actively involved scientists about especiallyimportant topics concerned. They can be summarizedas follows:Prosvectina for mountain mvmafrost clnd rnavvinqof associated Dhenomena (King et al. 1992) requires acombination of techniques to be applied. Direct investigations(drilling. digging). indirect (geophysical)soundings and measurements (seismic refraction, geoelectricresistivity, bottom temperatures of wintersnow, or radar), natural indicator phenomena (perennialsnow patches, cold springs emerging from snow-free areas, intact rock glaciers, push moraines s.~.,specific vegetation patterns etc.) and computer sirnulationsbased on empirically calibrated algorithms anddigital terrain information using Geographical InformationSystems (Keller 1992) are most promising. Thetime of pure guessing with regard to the preserlce orabsence of mountain permafrost and to its main characteristics(temperature, thickness, ice content)should now have definitely passed. Evidently, a minimumlevel of sophistication and funding is necessaryfor adequate research on permafrost and its relationto periglacial processes and phenomena in highmountain areas. The most urgent need concerns detailedquantitative information from core drilling andborehole measurements (cf., for instance, VonderMflhll and Holub 1992).The study of the distribution of mountain Bemafrostand climate (Cheng and Dramis 1992) first of allencounters dimcult problems with definitions (whereis the limit between "mountain" and "lowland' permafrost?]as well as with concepts ofizonation or "beltation"(continuous - discontinuous - island sporadicpermafrost). Development of a common terminologywould facilitate the discussion about global patternsof mountain permafrost distribution. Consideration ofclimatic parameters such as mean annual air,temperature,freezing and thawing indices or solar radiation(cf.. for instance, Hoelzle 1992) has been successful inat least half-quantitatively determining the influenceof altitude, latitude and continentality on large-scaleoccurrence patterns and characteristics of mountainpermafrost. However, continued research is needed tounderstand regional variability in more detail. In particular,the surface energy balance in mountain permafrostareas and active layer processes (especiallyadvective energy fluxes involved with steeply inclinedblocky surface layers) up to now essentially remainblack boxes in our understanding and striking obstaclesto scientific progress;A considerable variety of field and laboratory investigationsis being devoted to processesin the wrialacial mountam belt as related to seasonallyand wrmy&llu frozen uround (Lautridou et al. 1992).One special focus of interest concerns the chain ofprocesses and forms linking mechanical weathering ofrock walls, cliff recession, scree formation, and debris1014

displacement by solifluction, avalanches, debris flowsand permafrost creep (cf. also Olyphant 1983). Aspecial meeting held at Caen. France in 1991, devotedto problems of cryogenic weathering and organizedunder the auspices of the ICU commission on frostaction environments and the IPA working group onperiglacid environments illustrated the advantages ofcombining laboratory experiments with extensive fieldmeasurements and theoretical mbdelling (cf.. for instance,Matsuoka 1991). The model of crack propagationdue to segregation ice growth in water-saturatedrocks with interconnected cracks as proposed by Halletet al. (1991) deserves special attention, especiallywith regard to the influence of permafrost conditionsin bedrock walls. The corresponding effects relating tothe continuous presence of ice in cracks at greaterdepth, the influence of long-term warming trends andthe potential destabilization of heavily fissured rockmasses in greater volumes remain to be investigated.Thanks to a growing number of precise field measurements(drilling. geophysics. photogrammetry etc.),permafrost creev on doves and rock olackr fimsti~nare now much better understood than a few decadesago (Barsch 1992). Perennial freezing and supersalurationin ice of non-consolidated materials such astalus, till or even anthropogenic deposits induces fundamentalchanges in their geotechnical properties byreducing internal friction between rock particles andby enabling large-scale stress transmission and cohesiveflow behaviour according to the rheology of icerather than to that of non-consolidated debris. Rockglaciers and push moraines S.S. as the landforms resultingfrom cumulative straining of perennially.frozensediment bodies are composite products of various interactingand competing processes involved with thebuild-up and transformation of talus or moraines byvarious slow and rapid processes. Inherent to thecomplex origin of such landforms with their manifoldtransitions and analogues is the fact that definitionsand terminologies necessarily remain problematic.Emphasis should therefore rather be on the investigationof the thermal and rheological conditions involvedwith long-term preservation of ice underneath the surfaceand the creep mode of ice/rock mfxtures (cf., forinstance, Vonder Milhll and Haeberli 1990, Wagner1992). Such an approach also has the potential of improvingthe paleoclimatic interpretation of relict featuresof permafrost creep (cf., for instance. Dramisand Kotarba 1992, Sollid and Ssrbel 1992).A great amount of work remains to be done in orderto learn more about the complex relaflons a n ateractions between mountain aennafrost. olaciesnow and wata (Harris and Carte 1992). The radic%change of permeability at the permafrost table is a vitalsource of soil moisture in arid regions and greatlyspeeds up suprapermafrost runoff durlng snowmelt orliquid precipitation. Soil and active layer texturethereby plays ari important role and also heavily influengsconductive and advective fluxes of sensible aswell as latent heat. The behaviour of intra- and subpermafrostwater is best studied using tracer experi-ments, borehole observations (cf., for instance, Tenthorey1992, Vonder MCihll 1992) and physico-chemicalanalyses of water-cycle components. The snowcover as one of these components is linked to mountainpermafrost through especially close interactionsinvolving thermal ground insulation during cold periods,reduced snow metamorphosis within and retardedspring runoff from the winter snowpack in permafrostareas as well as intense ground cooling by latentheat, snow emissivity and albedo effects duringsummer to fall (Keller and Gubler 1993). The relationbetween permafrost and glaciers is quite well understoodin a general way and for static conditions (cryosphereschemes: maritime and continental climaticconditions) but detailed understanding of individualcases requires modelling of complex time-dependentflow and freeze/ thaw dynamics with fullglacier/pennafrost coupling - a challenge for the 21stcentury ?Experience with constructipn, environmental aroblemsand natural hazards in Jwrinlacial mountain beltsgrows but is still far from being adequate (Haeberli1992). Systematic site investigation and design recommendationsare needed for construction on perenniallyfrozen ground of buildings, hydropower andtransportation installations or protection work againstfloods, rockfalls and snow avalanches at high altitudes.Environmental aspects related to permafrostshould be more thoroughly studied in connection withthe preparation and maintenance of ski runs and thehazards from steep periglacial slopes must be morecarefully considered. During the extraordinary floodsin the Alps in summer 1987, the periglacial belt producedthe largest sediment volumes eroded and displacedby debris flows. It is also noteworthy that thelast three major rock fall events in the Alps all seem tohave had some - still unexplained - connections withperennial ice in bedrock cracks Wal Pola/Valtellina.Italia 1987; Tschierva/Engadin. Switzerland 1988) orincreased hydraulic conductivity in degrading permafrost(Randa/Matter Valley, S&itzerland 1991). Thethinking process concerning the effects of permafrostgrowth and degradation on steep slopes and rockwalls needs to be intensified. On the Qinghai-XizangPlateau, highway construction on relatively warm permafrostand the danger of increasing desertificationwith potential future warming and permafrost degradationare of greatest concern (cf. Guo and Zhao1993).The sensitivity of surface and subsurface ice withrespect to atmospheric conditions together with theintensity of morphodynamic processes In rugged topographyleads to strong reactions of mountain PWmafrostto clfmatic chanoe (Haeberli et al. 1993a). Thefocus of interest is with the marked warming periodsof the late Pleistocene and the 20th century as a basisfor better understanding and realistically anticipatingeffects of potentially continued if not accelerated futurewarming due to anthropogenic greenhouse forcing.Combined glacier/permafrost studies have beenquite successful in reconstructing regional patterns of1015

paleotemperature and paleaprecipitation during thelate glacial transition from predominantly cold anddry to generally warmer and more humid conditionssuch as it appears to have been characteristic formany mountain ranges. Backward extrapolation ofstatistical relations between borehole temperatures inpermafrost and meteorological parameters (air temperature,thickness of winter snow etc.) may help withassessing permafrost evolution in relation to 20thcentury warming and - for cases with supersaturated(low-permeability) permafrost ~ can be compared withanalyses of borehole temperature profiles using heatconduction theory. It should thereby be kept in mindthat both, the signals from and the effects on mountainpermafrost, are highly relevant to studies of globalwarming trends, because they reflect 3-dimensionalaspects (latitude/longitude/altitude) of cryospherechanges and concern one of the most heavily affectedecosystems on earth.PERSPECTIVESDecades of worldwide research in periglacialmountain belts have enabled the present state ofknowledge to be developed. This tradition of highmountain research ha5 long been predominantly directedat the most fascinating phenomena and atproblems which could be addressed and treated byindividuals, small research groups and modestinstrumentation. The future potential for fast changesand extreme disequilibria in cold mountain areas,however, is a challenge which more and more requiresdetailed process studies concerning aspects andquestions which have remained nearly untoucheduntil recently. Scientific progress is now especiallynecessary in the fields of non-steady evolutions of andinteractions between various components of themountain environment like permafrost and snow,glaciers, groundwater or bedrock under conditions ofchanging surface energy balance. In addition to suchprocess studies, international coordination andcooperation could be envisaged with regard topina. modellina and monitoring of mountain permafrostin order to reach a more complete view in spaceand time of present conditions and potentfal futuredevelopments. Correspondlng.research strategies firstof all depend on the scale considered.An overview at the scale of hmlispheres and conttngntScan be gained by compiling the existing literatureand by applying climatic indices of large-scalepermafrost occurrence such as mean annual air temperature~n combination with altitude information anda simplified classification of vertical belts. A first stepinto this direction was made with the extensive discussionand intercomparison of mountain permafrostconditions during the field trips in the Alps and in-theAlberta Rocky Mountains as connected to the IPAworkshops at Interlaken and Calgary. Previouslymentioned differences in relative extent of altitudinalpermafrost belts (continuous, discontinuous, sporadic)were recognized to be first of all due to differentdefinitions and terrninolo&. In the Alpine sense of theterm, a belt of discontinuous permafrost indeed extendsover at least 600 meters and probably evenmore in the Alberta Rockies. Striking differences innature, however. Tor the Alberta Rockies as comparedwith the Alps concern structural-geologic influences(predominance of bedded sedimentary rocks). therarity of the meadow belt, and the marked overlap betweenpermafrost and forest areas. Attempts arepresently being made at compiling a map of circumpolarpermafrost distribution for the northern hemisphere,which also contains general information concerningthe occurrence of low-latitude mountaln permafrost.The three reviews on permafrost in mountainranges of North +erica (Harris and Ciardino 1993).Europe (King and Akermann 1993) and Central Asia(Qiu 1993) prepared in connection with the presentcontribution for the same special session on mountainpermafrost and periglacial processes within theframework of the VI <strong>International</strong> <strong>Conference</strong> on Permafrostare examples of such cornpilations. Based onearlier reviews (cf. for instance Cheng 1983. Corte1988, Fuji and Higuchi 1978. Gorbunov 1978, Harris1988) and newly available information. it should bepossible to reach global coverage at a comparable levelwithin the corning years.The combination of mapping and modelling techniquesshould enable considerable progress to bemade in the corning years at the scale of indiuidualmouniain mnaes and arwas of mountains. Spatialsimulation of permafrost conditions by computer (GIs)models making use of digital terrain informationshould thereby be compared with maps and inventoriesof indicator phenomena such as intact rock glaciers.long-lasting snow, or suitable vegetation patterns.Concerning rock glaciers as permafrost indicators,not the origin of the ice but the thermalcondition for preserving ice in the ground over longperiods, Le. perennially negative ground temperatures,constitute the essential point. Other potentialindicator phenomena in permafrost-underlain peatlandas well as the Kurums of Asian authors needmore detailed investigation. The presently availableempirical rules from the Alps for predicting permafrostdistribution as a function of topography can only beapplied to other mountain areas after transformationof the involved functions (aspect. altttude. snow/avalanche' effects) according to large-scale variationsin mean air temperature. solar radiation and precipitationregimes (effects of latltude and continentality).Systematic application and field testing of modifiedmodel versions for selected mountain ranges couldhelp with adapting and further developing, i.e.generalizing the algorithms. which would then alsosewe as a key for simulating past and potential futurechanges as a response to climatic forcing. At the sametime. models are to be developed and tested whichmore closely relate to physical processes. especially tot.he surface energy balance with its prinripal influencingparameters: solar radiation, temperature and snow(cf. Hoeble et al. 1993).

~ <strong>Conference</strong>The basis for such improved models is formed bydetailed process studies, modelling of transient reactionsand long-term monitoring at the scale of &lcatchments and selected research sites. At present,neither the climatic scenarios provided by GeneralCirculation Models (CCMs) nor the present knowledgeabout mountain permafrost as a function of topographyand microclimatic conditions enable any exactregional/local predictions to be made. Long-term measurementsof the most important parameters involved(temperature, rheological properties, ice content, hydraullcpermeability. frost heave and thaw settlementetc.) are therefore of high priority. The techniques(drilling, borehole observation, photogrammetry, semiquantitativeanalysis of infrared aerial photography,runoff measurements etc.) for such long-term monitoringpurposes are available loday. Correspondingprogrammes have recently been initiated in someplaces (Haeberli et al. 1993b) but need internationallyco-ordinated support and completion. Results fromlong-term monitoring - as they become available - canbe used to calibrate models of transient response atindividual points (for instance, heat conduction,melting and thaw settlement in material with varlableice content, etc.). Calculations for Individual pointscan later be combined with spatial simulations ofsurface permafrost conditions in order to simulatetypical t;ansient effects at depth for extended areas.In a further step, such model simulations must betested and further developed by applying apprapriatesounding methods at characteristic sites indicated bymodel simulations. They could then hint at especiallysensitive areas and help assessing the representativityof monitoring at a restricted number of sites. In fact,our state of knowledge and preparedness with regardto assessing and hopefully mitigating potential effectsof realistic warming scenarios essentially depends onthe: establishment of adequate long-term monitoringprogrammes.RECOMMEN1)ATIONS AND ACKNOWLEDGEMENTSSystematic international coordination and cooperationconcerning permarost at high altitudes and inrugged topography should continue with emphasis bnmodelling, mapping and monitoring, especially in viewof potential future warming effects. A concentratedknowledge transfer and inter-comparison projectcould be carried out in various mountain areas of theworld with regard to the application of CIS-modelsusing digital terrain information in combination withfield mapping (rock glacier distribution, soil andspring temperatures, BE soundings etc.) for simulatingpresent-day. past and potential future permarostdistribution patterns. In addition, co-ordinatedplanhing of monitoring programmes (photogrammetry,borehole observations, surface energy balance etc.)should be stimulated and the short/long-term stabilityof supersaturated sediments on slopes (creep/ rockglaciers. thaw/ debris flows) be examined/modelled byapplying modern sounding/computer techniques.Thanks are due to the members of the IPA workinggroup on mountatn permafrost and to J.P. Lautridou,chairman of the IPA working group on periglacial enutronmentsas well as to the involved IPA council membersfor their most valuable and encouraglng help andassistance during the first 5 years of working groupactivity. The members of the PA working group onmountafn permafrost, J.P. Lautridou and a number offriends and colleagues at VAW/ETH Zurich reviewdan early draft of the present report.REFERENCESBarsch. D. (1992): Permafrost creep and rockglaciers.Permafrost and Periglacial Processes 3. 3, p.175 - 188.Cheng, G. (1983): Vertical and horizontal zonation ofhigh-altitude permafrost. Fourth <strong>International</strong><strong>Conference</strong> on Permafrost, Fairbanks,Alaska/USA, Proceedings. p. 136 - 141.Cheng. G. and Dramis. F. (1992): Distribution ofmountain permafrost and climate. Permafrostand Periglacial Processes 3, 2. p. 83 -91.Corte, AE. (1988): Geocryology of the central Andesand rock glaciers. Fifth <strong>International</strong> <strong>Conference</strong>on Permafrost, Trondheim. Norway, koceedings1, p. 718 - 723.Dramis. F. and Kotarba, A (1992): Southern limit ofrelict rock glaciers, Central Apennines. Italy.Permafrost and Periglacial Processes 3, 3, p.257 - 260.Fuji, Y. and Higuchi, K. (1978): Distribution of alpinepermafrost in the northern hemisphere andits relation to air temperature. Third <strong>International</strong><strong>Conference</strong> on Permafrost, Edmonton,Canada, Proceedings 1, p. 367 - 371.Gorbunov. A.P. (1978): Permafrost Investigation inhigh mountain regions. Arctic and Alpine Research10, 2, p. 283 - 294.Guo Dongxin and Huang Yizhi (1993): A guide to thepermafrost and environment of the Qinghai-Xizang Plateau (field trip Lanzhou - Lhasa:July 11 - 22, 1993). Published by the OrganizingCommittee of the Sixth <strong>International</strong>on Permafrost. 56p.Haeberli. W. (1992): Construction, environmentalproblems and natural hazards in periglacialmountain belts. Permafrost and PerlglacialProcesses 3, 2, p. 111 - 124,. 1017 9

173,Haeberli. W.. Cheng, G., Gorbunov, AP. and Harris,S.A. (1993, a): Mountain permafrost and climaticchange. Permafrost and PeriglacialProcesses 4. 2. p. 165 - 174.Haeberli, W., Hoelzle. M., Keller. F., Schmid, W., VonderMohll, D. and Wagner, S. (1993,b): Monitoringthe long-term evolu'tion of mountainpermafrost in the Swiss Alps. Sixth <strong>International</strong><strong>Conference</strong> on Permafrost, Beijing,China, Proceedings 1, p. 214 - 219.Hallet, B, Walder. J.S. and Stubbs, C.W. (1991):Weathering by segregation ice growth in microcracks at sustained sub-zero ternpera-tures: verification from an experimentalstudy using acoustic emissions. Permafrostand Periglacial Processes 2, p. 4, 283 - 300.Harris, S.A. (1988): The alpine periglacial zone, In:Clark, M.J. (ed.): Advances in PeriglacialGeomorphology. Wiley, Chichester, p. 369 -413.Harris, S.A. and Corte, A. E. (1992): Interactions andrelations between mountain permafrost.glaciers, snow and water. Permafrost andPeriglacial Processes 3, p. 2, 103 - 110.Harris, S.A. and Giardino, R. (1993): Permafrost inmountain ranges of North America. Sixth <strong>International</strong><strong>Conference</strong> on Permafrost, Beijing,China, Proceedings 2 (present volume).Hoelzle. M. (1992): Permafrost occurrence from BTSmeasurements and climatic parameters inthe eastern Swiss Alps. Permafrost andPeriglacial Processes 3, 2, p. 143 - 147.Hoekle, M.. Haeberli, W. and Kefler. F. (1993): Applicationof BTS measurements for modellingmountain permafrost distribution. Sixth <strong>International</strong><strong>Conference</strong> on Permafrost, Beijing,China, Proceedings 1, p. 272 - 277.Keller. F. (1992): Automated mapping of mountainpermafrost using the program PERMAKARTwithin the geographical information systemARC/INFO. Permafrost and Periglacial Processes3, p. 2, 133 - 138.Keller, F. and Gubler, H.U. (1993): Interaction betweensnow cover and high mountain permafrostat Murttl/Corvats'ch, Swiss Alps. Sixth<strong>International</strong> <strong>Conference</strong> on Permafrost, Beiling,China, Proceedings 1, p. 332 - 337.King, L. and Akermann, H.J. (1993): Mountainpermafrost in Europe. Sixth <strong>International</strong><strong>Conference</strong> on Permafrost, Beijing. China,Proceedings 2 (present volume).Lautridou, J.P., Francou. B. and Hall, K. (1992): Present-dayperiglacial processes and landformsin mountain areas. Permafrost and PeriglacialProcesses 3, 2. p. 93 - 101.Matsuoka, N. (1991): A model of the rate of frost shatering:application to field data from Japan.Svalbard and Antarctica. Permafrost andPeriglacial Processes 2, 4, p 271 - 281.Olyphant. G.A. (1983): Computer simulation of rockglacier developnient under viscous and pseudoplasticflow. Geological Society of AmericaBulletin 94, p. 499 - 505.Qiu, G. (1993): Mountain permafrost in central Asia.Sixth <strong>International</strong> <strong>Conference</strong> on Permafrost,Beijing. China, Proceedlngs 2 (presentvolume).Sollid, J.L. and Sorbel. L. (1992): Rock glaciers inSvalbard and Norway. Permafrost and PerilacialProcesses 3, p. 3. 215 - 220.Tenthor+. G. (1992): Perennial ntvCs and the hydrologyof rock glaciers. Permafrost and PeriglacialProcesses 3, p. 3, 247 - 252.Vonder Mtihll. D. (1992): Evidence of intrapermafrostgroundwater flow beneath an active rock glacierin the Swiss Alps. Permafrost and PeriglacialProcesses 3, p. 2, 169 ~Vonder Mtihll, D. and Haeberli, W. (1990): Thermalcharacteristics of the permafrost within anactive rock glacier (Murttl/ Corvatsch, Grisons,Swiss Alps). Journal of Glaciology 36,, -123, p. 151 - 158.Vonder MCllrll, D. and Holub, P. (199d): Borehole loggingin Alpine permafrost, Upper Engadin,Swiss Alps. Permafrost and Perfglacial Processes3, 2, p. 125 - 132.Wagner, S. (1992): Creep of Alpine permafrost, investigatedon the Murttl rock glacier. Permafrostand Periglacial Processes 3, p. 2, 157 - 162.King. L.. Gorbunov, A. P. and Evin. M. (1992): Prospectingand mapping of mountain perrnafrostand associated phenomena. Permafrostand Periglacial Processes 3, p. 2, 73 - 81.' 1018

PERMAFROST It4 THE MOUNTAIN RRNCIES OF NORTH AMERICAStuart A. Harriabepartment of Geography, univeralty of CalgaryCalgary, Alberta, Canada T2N 1N4John R. GiardinoDepartments of Geography and GeologyTexas A fi M University, College Station, TX 71843, USAMountain permafrost extends along the Western Cordillera of North America fromthe Arctic Ocean south to Mexico. It also occurs along the lowar coastalmountains in eastern North America as far south au the Gasp6 Peninsula, wherewarm tropical air and the effect of tha Gulf stream combine to keep the groundwarm. Coldest ground temperatures (-10' to -2OOC) and thickest permafrout occurin the Queen Elizabeth Islands. Most is post-glacial and the greater permafrostdistribution in the western mountainu is due to their greater altitude, theproximity to the cold Arctic air masses in winter, cold air drainage, rain shadoweffebts and chinook winde causing thin winter snow on covers the eaatern side ofthe Cordillera. Climatic warming may be occurring on the Prairies, but the uppermountain slopes are generally not showing warmer air temperatures in thidContinental and Subcontinental climate. Instead, it neems likely that majorshifts in air mass distribution may be necessary to alter these mountainclimates. In the mountaina of eastern Canada, is there evidence for degradationof permafrost.-Pioneer work was carried out in Canada by Brown(1967, 1978). in Alaska by Ferrians (1965). and in thecontiguous United Statea by Pew6 (1983). The altitudinalsediments, but may be relict in olner maLerials, includingmassive icy beds, some of which may represent buriedglacial ice (e.g., Lorrain ti Demeur, 1985; French ti Harry,1990) or other surface ice (e.q., snowbanks). Permafrostexiatence beneath the former ice sheets and basal icedistribution in the Eastern Cordillera was described by temperatures of -17OC are common today.Harris (1986), and more recently Heginbottom and Dubrauil(1993) have assembled the currently available for data theIn the Eastern Arctic Islands, nlountairlv the containmoat recent permafrost and ground ice map for the National permafroat with mean annual ground temperatures rdngingAtlas of Canada and for the Circumarctic map of permafrost from -2°C to -12°C beneath the alpine tundra vegetation.and ground ice conditions (Brown et al., 1993). This Ground ice appears to be of limited extent except inreport is baaed largely on these sources.peatlands.DISTRIBUTIONIn the Cordilleran of norrhwes'terr Nor-tn iunerica,condition% vary with topography, latitude, proximityto theThe mountain permafrost areas extend along theArctic air maases and to the Pacific air masses. Al.ong theeastern and western margins of North America, from the Arctic coast, mean annual ground temperatures are -4OCQueen Elizabeth Islandm in the north, south to Mexico, to -1O'C. but these increaae to -1'C to -4'~ southwardn inwhere pnrmafrost occurs above about 4500 m (PlwB, 1983: central Xukon and in the central Alaska Range. PermafroetHeine, personal communication, 1993) onthetops of occurs only on ths eastern side and on the tope of thevolcanoes. In the north, these areas grade into polarcoastal mountains around north the Pacific Ocean. At. Mountpermafrost at low @levatione, but in the western Sub-Arctic Garibaldi (near Vancouvor) it ia limited to the mountainsouth of the Brooks Range and south-Central Yukon, they topa. Ice contents can be high near the Arcti.c coast andbecome discontinuous to sporadic, and are limited to theHarris higher mountain fi Brown, ranges when mapped at a small southwards along in the eastern Cordillera through thescale (e.g., Yukon Territory, although they are highly variable and1982; Heginbottom & Dubreuil, 1993; Brown generally decrease southwards. Sufficient interatltiaL iceet al., 1993). The tectonically active, young mountains occuru in sediments on block elopes ro generatrp active rockform a broad belt. The lower limits fpr continuous (>EO% glaciers as far south ae southwestern Alberta, and theseof the landscape underlain by permafrost), discontinuous can also be found locally on high mountain peake as far(30-80% underlain by permafroat) and sporadic permafrost south aa Mexico. Apart from these cases, ice contents tend(do%) generally rise southwards along the Easternto be lesa than 10% by volume south of the Yukon Territory.Cordillera of Canada and the contiguous United States(Harris, 1986). Because the mountain tops become progreus-Mean annual ground temperatures are appreclhblyively higher in elevation southward8 to the Colorado-New warmer along the eastern ri.m of the Arctic Islands and downMexico border, mountain permafrost can be found through a to the Gasp&, am a result of the higher snowfalls east ofverywide range of latitude and on into Mexico.Hudaon Bay. Ground ice content is variable, generally butlow.Permafrost aleo occurs in the mountains along theeastern rim of the Arctic felands and down along theseWhere the mean annual ground temperatures ahv- - eremountains to the GespB, where it occurs sporadically on 3'C. there tends to be a marked effect of aeyect onmountain tops under alpine tundra vegetation f, Brown, (Gray permafroat distributiun which lnczeaaea lnverflely with1982). It disappears in New England as a result of the latitude in Canada. Most of ths permafrost is postinfluenceof the Gulf Stream and tropical air maasea froin glacial because moet of the area of present-day mounta,nthe Caribbean. These mountains are lower and geologically permafrost was covered by Late Wiscticein ice. Relictolder, but are still quite rugged and tectonically active permafrost from colder glacial climates has only beenin various sections.demonstrated on Plateau Mountain, eouthwesl.Aqberta (Marrigfi Brown, 1982), which was a nunatak.CHAPkTERISTICSCLIMATIC PROCESSESIn the Queen Elizabeth Islands, near-surface permafrosttemperatures range between -10" and -2O'C, and theMountain permafrost areas oxist becauQe the micropermafrostis thick. Gtound ice contents are highlyclimate pecmita sufficient heat removal from the groundvariable and closely related to the nature of the bedrock. surface for part of the ground remain to below uoc for moreThe permafrost is post-glacial in the surficial glacial than two consecutive years.Basic climatic factor5

favouring its formation include lowinter snow cover, highIn eastern Canada, rock glacier-dominated mountainswinds removing the cover snow above tree line, proximity to are normal, although the shorter time mince deglaciationcold Arctic air masses for long periods of time in winter, and massive bedrock limits their development in easternmaritime temperate air masses in summer (not tropical air Quebec and Labrador. They are associated with occasionalmasses), and cold air drainage. The Canadian Cordillera is palsas, patterned ground and ice wedges.characterized by rain shadow areas in the lee of thecoastal mountain ranges with chinook winds that remove EFFECTS much OF CLIMATIC CHANGESof the winter snow cover by ablation. The further east andnorth the area is, the greater the frequency of cold Arctic From the discussion on climatic processes, it isair masses in winter. Permafrost occurs in two situations, obvious that any changes in climate may greatly alter theviz., altitudinal permafrost on the upper slopes of the permafrost distribution in the mountains of North America.mountains resulting from the cooler temperatures at higher There ia good evidence for permafrost being formed on theelevation, and at the foot of mountain ranges because of higher parts of the Appalachian MOuntainS since the lastcold air drainage (e.g., Harris, 1982). The latter process glaciation' (Clark & Schmidlin, 1992), whereas similarextends southwards at least as far as southwest Alberta, evidence can be found in the Cordillera from Mexico northwhere it has become more prevalent in the last eight through years. New Mexico, Arizona (e-g., Blagbrough & Farkas,Heavy snow accumulation$ on tha slopes of mountain ranges 3968; Barsch & Opdike, 1971) and Colorado (Giardino, 1983)facing Pacific coast winds tend to result in a high to Alberta and British Columbia. Thus, in Banff Nationalelevation €or the lowest level of permafrost westwardePark, inactive near-slope rock glaciers can be fnund alongacross the cordillera in British Columbia as the climatic the Bow River ?alley .at an elevation 600 m below theregime changes from subcontinental to maritime.present-day regional permafrost limit. Similar occurrencescan be found in Jasper National Park, so there must haveFurther north, cold air drainage and prevalence of been some very profound changes in climate producing acold Arctic air in winter result in far lowor groundmarkedly different distribution of mountain permafrost intemperatures at a given elevation, culminating in the poet-glacial times both in eastern and western Northconditions on the Queen Elizabeth and Ellesmere Islands, America.where mean annual air temperature can be as low as -20°C.Summers are short and cool. This is a continental perma-Further north around Fairbanks, Westgate et al.frost climate.(1990) have shown that permafrost landforms have developedperiodically for over 3 Ma B.P. The first dated evidenceSouthwards in the Cordillera, permafrost occurs most for cold conditions in British Columbia is 3.5 Ma fromfrequently on the east and north-facing slopes of the Wells-Gray National Park (Hickson & Souther, 1984). whichhighest mountains where snowfall is reducea or the snow agrees with the evidence from Patagonia (see 1991). Corte,cover is removed by high winds. The main occurrences are In Central Yukon, periglacial features to date least 2.5along the eastern Cordillera, although ice caves can be Ma in non-glacial sediments at Old crow. Cold conditionsfound in the coastal ranges as far south as northernfirst developed on the rising high mountains of the St.California.Eliae Range as early as Miocene times (Armentrout, 1984).but appear to have mainly caused alpine glaciation,Along the east coast of North America, the higher although some permafrost conditiona may have also beenprecipitation tends to reduce the frequency of permafrost developed in nunataks at high altitudes.at a given elevation and latitude in of apita the proximityof the mountains to a cold sea current. The of occurrenceepermafrost end abruptly where the coastal mountains areadjacent to the warm Gulf Stream drift, and where themaritime Tropical air masses penetrate from the Caribbean.LANDFORMSIn areas not dominated by glaciers, e.g., EllesmereIsland, mountain permafrost regions can be divided intothree groups, viz.: those dominated by active rock glaciers,by active block streams and by solifluction/gelifluction landforms. The areas with active blockstreams are mainly limited tu the northeast Yukon Territory,which has a suitable combination of low precipitationand intense winter cold. These landforms are associatedwith active ice wedges and open system pingotl, as ae wellmassive ground ice in the valley floors. Spectacularicings (Lauriol et al., 1991) and seasonal frost mounds(Leffingwell, 1919: van Everdingen, 1978, 1982) may occurat the base of slopes.From east-central Yukon Territory and the BrooksFig. 1. Variation in mean annual air temperature atRange southwards to southwest Alberta and southeast BritishWatson Lake between 1939 and 1990 (bssld oncolumbFa, the Cordillera exhibits abundant active rockRES, Monthly).glaciers, The climate in moist and cold enough to providethe interstitial ice necessary for the formation of nearsloperock glaciers. In the southwest Yukon Territory andCentral Alaska, these are associated with open systemFigure 1 shows that the so-called "climatic change"pingos at the foot of slopes. Eastward3 at this latitude, of the last decade which is regarded as being the of resultextenaivs thick peat deposits have formed on the valley an increase in carbon dioxide, has not yet produced meanfloors and exhibit paleaa and peat plateaus. Although the annual air temperatures outaide the extremes measuredpingoa are confined to tne area north of the 60th parallel, during the last 50 years in central and southern Yukonpalsas and peat plateaus also can be found in the northern Territory (Wahl et al., 1987; Harris & Schmidt, 1993).parts of British Columbia and Alberta, Degrading iceSimilarly in southwest Alberta, the only mountain weatherwedges occur in south-central Alaska Yukon, and but active station at a permafrost site showing an appreciable rise inice wedgee have been reported from Mayo (Burn, 1990) and mean annual air temperature is Marmot X1 near Basin Jasper.certainly occur to the north. Ice caves are present from The others ohow greater variability and frequency of majorOld Crow southwards.cold air drainage events, no but major overall change.South of Kananaskis Lakes in Alberta, the landscapeIn cuntrast, there is evidence for warming at the fewof active rock glaciers is replaced by one with gelifluct- lowland weather stations along the Mackenzie River valleyion and solifluction forma, together with felsenmeer and in the Arctic, but data from mountain sites in this areablock slopes, on the Lower part of the mountains. Active are lacking. This warming trend i5 also seen in datarock glaciera occur in the moister, cold climates nearrecorded since 1970 from the Plains region of Alberta. Itmountain tops. Patterned ground is common in the block is quite obvious that if a major change in mean annual airfielda, but peaty permafrost landforms are rare. Ice caves temperature occurred in the mountains, there would beare common in areaa of thin surficial deposits overadjustments in the distribution of mountain permafrost, aslimestones, dolomites, or volcanic rocks.ia occurring presently at Marmot Basin, but it would eeem

that the mountain climates of western North America not are Giardino, J. R. (1983) Movement of ice-cemented rockreacting to the perceived climatic change inferred fromglaciers by hydrostatic pressure: An example fromnearby lowland stations. This may mean that whereas majorMount Mestaa, Colorado. Zeitschrift fur Geomorpholoqiechanges could occur in permafrost distribution at lowland27, 297-310.sites, the distribution of mountain permafrost may remainlargely the meme a6 at present. It seems likely that major Gray, J. T. and R. J. E. Brown (1902) The influence ofshifts in air mass dimtribution may be needed to'alter theterrain factors on the distribution of permafroatsubcontinental and continental mountain permafrost cli-bodies in the Chic-Choc Mountains, Gaspbie, Quebec.mates. The different air masses would generate a newProceedings, 4th Canadian Permafrost <strong>Conference</strong>,pattern of winds which would produce a different pattern ofCalgary, Alberta. National Research Council of Canada,ocean currents, and together these would have a profoundOttawa. 23-35.effect on the climate of the coastal mountains.Plong the east coast, the evidence for climaticchangc in the mountains is lacking, but there is strongevidence for a climatic change that has been occurring inthe last few decades in western and central Quebec, wherepalsas are degruding near the southern limit of permafrost.Features interpreted us ice-wedge casts in post-glacialsediments have been reported from the St. Lawrence Lowlands,so there is evidence for substantial Holoceneclimatic change in this area as well.Thus, the occurrence of permafrost appears to becomplex, and numerous environmental variables are responsiblefor controlling ite location. whereas the numerousvariables controlling permafrost development are known andunderstood, the relations between the various variables areproblematic at best.WERENCESAES, Monthly. Monthly Record. Environment Canada,Atmospheric Environment Service, Toronto.ArmPntfOut, J. M. (1984) Recalibration of Miocenemolluscan stagaa and consequent recorralation ofAlaskan formations. Geological Society of AmericaRbatracts with Programs 15(5), 267.Barsch, D. and R. G. Updike (1971) Periglaziale Formung amKendick Peak in Nord-Arizona wahrend der letztenKaltzeit. Gaographische Helvetica 3, 99-114.Blagbrough, Y. W. and S. E. Farkaa (1968) Rock glaciers inthe San Mate0 Mountaina, south-central New Mexico.American Journal of Science 266, 812-823.Brown, J., 0. J. Ferrians, Jr., J. A. Heginbottom and E, S.Melnikov (1993) Circumarctic map of permafrost andground ice conditions. Proceedinge, 6th <strong>International</strong>Permafrost <strong>Conference</strong>, Beijing.Brown, R. J. E. (1967) Permafrost Map of Canada. NationalResearch Council, Building Research Division, NRCX9761, ahd Geological survey of Canada, Map 1246A.Brown, R. J. E. (1978) Permafrost. Plate 32 in:Hydrological Atlas of Canada, Ottawa. Fisheries andEnvironment, Canada. 34 Plates.Burn, C. R. (1990,) Implicationu for palaooenvironmentelreconsteuctiop of Recent ice-wedge development at Nayo,Yukon Territory. Permafrost & Periglacial Processes 1,3-14.Clark, G. M. and T. W. Schmidlin (1992) Alpine periglaciallandforms of eastern North America: A review. Permafrost& Periglacial Processee 3, 225-230.French, H. M. and D. G. Harry (1990) Observations onbutied glacier ice and massive segregated ice, WesternArctic Coast, Canada. Permafrost L PeriglacialProcesses 1, 31-43.Harris, s. A. (1982) Cold air drainage west of FortNelson, British Columbia. Arctic 35, 537-541.Harris, S. A. (1986) Permafrost distribution, zonation,and rrtability along the eastern ranges of thecordillera of North America. Arctic 39, 29-38.Harris, 5. A. and R. J. E. Brown (1982) Permafrostdistribution along the Rocky Mountains in Alberta.[Roger J. E. Brown Memorial Volume, H. M. French, Ed.]Proceedings of the Fourth Canadian Permafrost confer-ence, Calgary, Alberta.Canada, Ottawa, 59-67.National research Council ofHarris, S. A. and I. H. Schmidt (1993) Peat plateaus, h160-167, Robert Campbell Highway, Yukon Territory.Journal of Paleolimnology. In the Press.Heginbottom, J. A. and M. A. Dubrouil (3993) A newpermafrost and ground ice map for the National Atlas ofCanada. Proceedings, 6th fnternational Permafrostconference, Beijing.Hickson, C. J. and J. G. Souther (1984) Late Cenozoicvolcanic rocks of the Clearwater-Wells Gray area,British Columbia. Canadian Journal of Earth Sciences21, 267-277.Lauriol, B., J. Cinq Mars and I. D. Clark (1991)Localisation, genbe et fonte de quelgues naleds duNOrd du Yukon (Canada). Permafrost & Periglacial Processes2, 225-236.Leffingwell, E. de K. (1919) The canning River region,northern Alaska. United States Geological Survey, Professionalpaper Xl09. 251 pp.Lorrain, R. D. and P. Demeur (1985) Isotopic evidence forrelic Pleistocene glacier ice on Victoria Island,Canadian Arctic archipelago. Arctic 6. Alpine Research17, 89-98.PBwB, T. L. (1983) Alpine permafrost in the contiguousUnited states: A review. Arctic and Alpine Research15, 145-356.van Everdingen, R. 0. (1978) FrOSt mounds at: Bear Rock,near Fort Norman, Northwest Territories, 1975-1976.Canadian Journal of Earth Sciences 15, 263-276.van Everdingen, R. 0. (1982) Management of groundwaterdischarge for the solution of icing problems in theYukon. In: Proceedings of the 4th Canadian PermafroatConforeme. H. N. fronch, Ed., Associate Committee onGeotechnical Res@arch, National Research Council,Ottawa, 212-226. .Wahl, H. E., D. B. Fraa@r, R. C. Harvey and J. B. Maxwell(1987) Climate of Yukon. Atmospheric EnvironmentService, Environment Canada, Ottawa. climatologicalseries 140. 321 pp.Corte, A. E. (1991) Chronostratigraphic correlations ofwaatgate, J. A., B. A. Stemper and T. L. Pew6 (1980). Acryogenic and glacigenic episodes in Central Andes with3m.y. record of Pliocene-Pleistocene loess in interiorPatagonia. Permafrost & Periglacial Procewdes 2, 67-Alaska. Geology 18, 858-861.70. .Ferrians, 0. J., Jr. (1965) Permafrost Map of Alaska.United states Geological Survey Miscellaneous GeologicalInventory Map 1-445. Scale 1:2,500,000.

MOUNTAIN PERMAFROST IN EUROPELorenz King' and Jonas kerman2' Geographical Institute, Justus Liebig University,D-35390 Giessen, GermanyDepartment of Physical Geo raphyUniverslty of Lund, S-233 62 Luni, SwedenAbstract:In several Euro ean countries a good knowledge about the properties and the distribution of mountainermafrost has gee, a vital factor for the development of these areas. This concerns particular1gvalbard (mining and construction activities), the Fennoscandian mountains and the Alps (traflc andprotection measures). The distribution of mountain permafrost in Europe is displayed and the mostImportant permafrost features for these mountain areas are mentioned. Whereas island permafrost isvery common in many mountain areas, and sporadic permafrost, too, continuous or discontinuouspermafrost is restricted to Svalhard, the Fennoscandian mountains and the Alps.INTRODII(JTION One of the was sessions also devoted to definition andclassification of mountain permafrost. It was agreed there, thatA "Circumarctic Map of Permafrost and Ground Ice mountain permafrostis perennially frozen ground in mountainConditions" at a scale 1:lO million bas been recently prepared by areas. This includes mountains in tropical, moderate and polardifferent member countries of the <strong>International</strong> Permafrost areas, and their common featureis a considerable altitudmalAssociation (IPA) and a first draft map has been presented and difference that produces special morpholoi d forms. It isdiscussed at the Sixth <strong>International</strong> Permafrost <strong>Conference</strong> ingenerally accepted today, that active rockgfaciers are the mostPeking 1993. In addition, the IPA Working Group on Mountain visible expresslon of mountain permafrost as creep of thick talus orPermafrost prepared a special session with contrlbutions frommorainic material. A great number of other eomorpholocicalAmerica, Europe and Asia (IIaeberli 1993). This paper is part of expressions for permafrost exist, too, and wilf be treated in thisthese presentations and summarizes the knowledge concerning paper.mountain permafrost in Europe.According to the considerable altitudinal differences, theAvailable information shows, that permafrost in Europe exists distribution pattern of mountain permafrost is rather a verticalin: than a horizontal with one, difficulties many involved for itsa) Northern Europe (incl. Iceland, the Svalbard Archipelago and presentation on maps. In a short horizontal distance, ermafrostthe Fennoscandian Mountains) may existas continuous (90 - loo%), discontinuous (& - 90 %),b) the Alps and sporadic (10 - 50 %) or island (0 - 10%) permafrost.c) the Pyrhkes.Although these definitions have been used for the PermafrostTo a smaller extent permafrost is also [Tesent in:Map of the Northern Hemisphere presented at Sixth thed the Tatry Mountains (Poland, Slova la),<strong>International</strong> Permafrost <strong>Conference</strong> in Pekin , and have alsoe{ the Carpathian Mountains (Romania), been adopted by the authors this for paper, Sould it be realized,f) the Abruzzi Mduntains (Italy) and that these mentloned termsare also used in a different manner n9) Scotland. only in Europe, but also in earlier studies in Norte.g. the term "sporadic" is often used for areas with less than 10%Permafrost areas are small in Europe, when compared to the occurrence, "patchy discontinuous" for10 to 50% and so on. TheAmerican Cordilleras or to Asia with their large areas of mountain addition of percentage values to these terms certainly is a goodand plateau permafrost. However, the European mountains are approach to definition problems and helps to avoidrelatively densely PO ulared or develo ed, and a good permafrost misunderstandings.knowledge is a vital Factor in and for tl!ese areas and theirdevelopment (cp. Haeberli, 1992). Therefore, there is a goodknowledge about mountain permafrost distribution and properties SVALBhRnin quite a number of European countries. During the IPA5 onsored international workshop on "Permafrost and Periglacial The Svalhard Archipelago is mainly located between about 74"8. nvlronments in Mountain Areas" in Interlaken, Switzerland, 16- and SO" northern latitude and belongs, to the area of continuous20 September, 19Y 1, much of this knowledge has been presented permafrost with a MAAT of about -4°C and lower (herman 1980,and dlscussed together with American and Asian permafrost1987, King et al., in prep.). Permafrost thicknesses of about 100 mspecialists, and research sites in the Alps have been visited befsre (along the west coast and the larger fiords) and 250 to 450 mand after the workshop. Meanwhile, most presentations have heen (further inland) have been measured, especially in the existin coalpublished in three issues of Permafrost and Periglucial Processes mines (Liest011980, 1986, Landvik et al. 1988). Large parts of(volume 3, 1-3, 1992). The references to these papers form anSpitsbergen, the main island of the archipelago, are an area ofexcellent bibliography on mountain permafrost.rugged mountainous terrain with only narrow coastal plains.. 1022

The central parts of Spitsbetgen and the large islands in the eastare built by youn sedimentary rocks and the mountains areplateaulike and Jvided by wide glacial valleys. 60% of thearchipelago are covered by glaciers.Ice-wedge olygons are frequent in the valley bottoms of thelarge wide vafeys of central Nordenskjold Land (Svensson 1976).However their distribution is not clear as they are often mixed upwith the even more common and more widely distributed soilwedges. These soil wedges are mistakenly often classified as icewedgesin ma s a d inventories, because their surficial appearancemight be sirnirar (herman 1980,1987). Pingos are common andfound in the large wide valle s of central Nordenskjbld andAndrte Land and more rare& on Ed 'eoya and Barents~ya. Amajori of the pingos are interpretef as open system pingos(Liestey1976) but a few have been classified as closed systempingos (Svensson 1973). Palsa-like frost mounds have he nreported from Nordenskjold Land and Nordaustlandet (herman1982, Salvigsen 1 77). Icings of different origin are common allover the re ion ( w kerman 1980) and most commonly produced bythe winter &charge from subpolar glaciers. An interesting formare icings produced in association wlth some of the surgingglaciers.Steep rock slopes often produce vast amounts of debris, thatform rockglaciers in facoured places. These henomena ofmountain permafrost have been studied by 8umlum (1982) and bySollid and Sorbel (1992), also in cooperatlons with colleagues fromETH-Zurich (Hoelzle, In prep.). Push moraine formation is oftenfavoured by the existence of permafrost in glaciofluvial or marinesediments in the valley floors. The mechanisms of push moraineformation has been studied by Van der Wateren (1Y92), Lehmann(1993; cp. also Gripp 1926, Sollid & Sorbel 1988).Although Greenland belongs to the North American continent,it is also regarded as part of the "Nordic Countries" of Europe forhistorical reasons, and a large number of permafrost studies havebeen done by European scieyp, especially from Denmark.References are compiled in erman (In prep.).ICELANDIceland experiences a mild and humid oceanic climate.Permafrost is limited to the central hi hlands above450 m ad.(Thorarinsson 1951, Schunke 1975). l%e majority of thepermafrost observations in Iceland arc connected with the flisurfaces(Islandic bogs), that are characterised by numerous smallponds alternating with level surfaces of wet ground and by amultitude of large hummocks, called rist (or dys). These rhsts arethe equivalent of" alsar" in the Scandinavian terminology andfound between 40tm and 800 rn ad. Their lower limitcorres onds fairly well with the 0°C isotherm. At higher altitudesthe parsas (rdsts) are larger, higher and more stable. Hcights mayreach 3 m and diameters up to 30-40 m are common.Observations of permafrost in terminal moraines, rockglaciersand push moraines have been mentioned by Eyles (1978),Humlurn (1985) and Rutten (1951. Permafrost is certaini alsopresent in the glacier-free rockwal I s and summits above 8&l ma.s.l., but no systematic studies have been done.FENNOSCANDKAIn Fennoscandia, the traditional opinion until recently was, thaterrqafrost in Fennoscandia, even in the northernmost parts,!asically was restricted to the palsa bogs (Fries& Bergstrom 1910,Hamberg 1905, Rapp & Rudberg 1960, Wramner 1973). However,recent studies made'clear, that the extent of permafrost is muchmore widespread and that the majority of the permafrost areasbelong to "mountain ermafrost" (0strern 1964, Svensson 1962b,Ra p & Clark 19 1 Ring 1976, 1982, 1983, 1986, Rapp 1982, King& %p ala 1987, kerman & Malrnstrfim 19 6, Jeckel 1988, Sollid& S@r&l 1992,0degard et al. 1992, cp. also L erman, manus.).The distribution of permafrost in Fennoscandia is basically avertical zonation as follows:Island aermaft'ost with a high or medium high ice content iscommon in and around the Fennoscandian mountains and is found *from liardangervidda and Dovre in southern Norway (Sollid RrSclrbel 1974) up to the Varanger Peninsula in the north. Withinthis class permafrost is more or less restricted to organic soils. toogs and to als& or palsa-like features (Svensson 1962a, 1986,khman 197f Meier :. ' 198.5, Seppala 1988). Annual and short-livedfrost blisters are also common, here. Island permafrost is quiteoften observed at altitudes above 1000 m a.s.1. in the south, andreaches down IO sea level altitude in the north. In thesouthernmost mountains of Hardangewidda (Norway) andJamtland and Hiirjedalen (Sweden) it my even be foundconsiderably lower than 1000 m ad. at very selected places,mainly bogs. In Jiimtland several sites between 650 and 750 m a.s.1.with palsa-like forms and small permafrost lenses have beenreported by Smith (191 I) or Lundquist (1962).Saoradic Dermafrost occupies the mountain areas above 1.200m ad. in thk south (Jotunheimen) and above 750 m ad. innorthern Sweden (Kehnekaise). Further north and inland, and dueto increased continentality (cp. King & Seppsla 1987, MalmstrBm1988), this belt can also be found at lower altitudes of 300 m to 400rn. The stron gradient in continentality from west to east isdisplayed in bigure I (cp. also valuefi in table 1). Icings, often inassociation with karst drainage in the mountains and with smallgroundwater springs are quite common in the sporadic ermafrostbelt. Pingo-like features regarded as transitional forms getweenalsas and pingos are found in the Abisko, Finnmarksvidda andkastosjaure areas (Svensson 1969, La erbiick & Rohde 1985,Akerman & Malmstrom 1986): in thefower levels these forms areoften relict.+s at altitudes of about 1600 mWith increasing altitudes, the sporadic permafrost belt graduallychanges into discontinad. In Jotunheimen, and 1 00 m a s.1. in the Kebnekaisemountains. This permafrost belt shows rock8laciers and ice-coredmoraines in the steeper, alpine type mountam (Barsch 1971,Bstrem 1964) and pronounced large-scale pol gon patterns in thesmoother Scandinavian fell-type mountains (iap & Clark 1971,Rapp & Annersten 19691. Rock glaciers may reacK down into thesporadic zone, especially in the more continental, easternmountain areas (Barsch & Treter 1976), where the periglacialareas have. quite a considerable vertical extent due to a highglaciation limit and relatively low mean annual air temperatures.Wmoremoremaritime CLIMATE continentalFigure 1: Schematic section across the Scandinavian mountainranges. From west to east the lower permafrost limits dropconsiderably and the altitude of the glaciation limit increases. Thuspermafrost is mainly found in the eastern (and northern) partswith a more continental climate; the western mountain rangesexperience much precipitation and are often glacierized.ElI1023

The &~IOUS aermafrosy belt is found above 2060 m inJotunhcimen (NoGay) and above 1600 m ad. in northernSweden (King 1984). It is an area where patterned ground is oftena5 well developed as in many hi h arctic areas, if slope andsediment cover are favourable. fermafrost thicknesses of about 70m have been measured even at the relatively low altitude of 1200m a.s.1, (Ekman 1957) and the thickness of permafrost is estimatedto be more than 100 rn at about 1500 m ad. and probably even200 to 300 m in the highest elevations of the Kebnekaise mountain(2400 m ad.) in northern Sweden (King 1984, 1986).U S AND JURAResearch in mountain permafrost has a long tradition inEurope and started mainly in the Alps with the investi ation ofrockglaciers (Barsch 1Y78, Haeberli 1985). It is generafly agreedtoday, that active rockglaciers are the most typical phenomena forwidespread s oradic or discontinuous mountain permafrost(Barsch 19927. Rockglacier research is still continued in the Swiss,Italian, French and Austrain Alps, and there are a large numher ofthorougly investigated rock laclers in all these mentionedcountries (cp. Haeberli et 3. 1992, Belloni et al. 1993, Evin et al.1993). Research includes studies of rockglacier movement over afew and up to 25 years, rockglacier drilling and detailedgeophysical borehole logging, as well as man eophysicalsounding techniques (eg Francou et al. 1992 haeberli et ai. 1979,Von der Muhll et al. 1990, 1992). In more recent ears, similarstudies have also heen conducted on Iceland, in tieFennoscandian mountains, as well as on S itsbergen (cp.references in previous chapters). Freeze/tfaw cles in ermafrostareas (e. . King 1990) and creep of permafrost~~laeber~ 1985,Wa ner f992) are other important research topics. Importantmet%odolopical advances in mountain permafrost research alsooriginated In the Alps: the BTS-method has become an acceptedworkin tool for ma in permafrost very efficiently (e.g.Gugliefmin et ai. 19$!, fioelzle et ai. 1993). Modelling ofmountain permafrost distribution and its automated mapping willbring new knowled e and new ideas about permafrost distributionof lar e areas that kave not been investigated detailedly until now(cp. f!eler 1992). A methodological review concerning mappingand prospecting of mountain permafrost is given by King et ai.(1992 . It shows that the Alps represent an Important researcharea or comparative studies in other mountain areas of the world(Cheng et al. 1992, Gorbunov 1978).In addition to scientific projects (Haeberli 1993), research inmountain permafrost has also been greatly promoted byconstruction measures. In the Alps, construction includes railways,cable cars, skilifts, restaurants and hotels, communication towers,hydropower installations, high power transmission lines, androtection measures a ainst natural risks. Construction sites can&e found in all permafost belts, from the s oradic to thecontinuous one (Haeberli 1992). Many bezock exposures havethus been created and allowed the study of ice-filled bedrock-jointsor temperature gradients in tunnels. The scientific permafrostcommunity is ve grateful for these engineering activities, andshould carefully %low these, wherever ossible. Cum arabledevelopment activities have only startexin Northern gurope,Eastern Europe and the Pyrinies, and we will certainly learnmuch more about mountam permafrost distribution in Europe inthe years to come.In the Alps, permafrost occurs from the southern French andItalian Alps over the central Swiss Alps to the Austrian andGerman Alps. Predominantely sooradic oermafrad with low icecontent in bedrock and high ice-content in nonconsolidatedsediments exists at altitudes between 2000 m and 2500 m ad.Patchy vepetation and alpine meadows cover these areas and relictand inactwe rock laciers are ty i d and more numerous thanactive ones here. hand ~ermat%[ extends much further down, toaltitudes considerably lower than the treeline in general. Theoccurrences are limited to special places, as e.g. snow-freerockwalls and slopes exposed to ttie north, long lastin avalanchedeposits and ice-caves in limestone areas. There, the hAAT maybe markedly higher than 0°C.Discontinuous Dermafrnst is common at altitudes above about2500 m a.s.1. and gradual1 changes into n In s ermafrost ataltitudes above 3000 to 3&0 m a.sll. Thc?%~~~r~afrosttemperature is -5" to -6°C at Jungfraujoch (3500 m ad., northernSwiss AI s) and about -15°C with maximum ermafrost thickncsscsexpectrfto exceed 1000 m on M,onte Rosa fi.500 m ad.,southern Swiss Alps; cp. Haeberli & Funk 1991).In contrast to the Fennoscandian mountains the AI s have notbeen shaped by a continental ice-sheet, and flat or rolEnmountain landscapes are missing and steep slopes preva!. Due tothis geomorphological characteristics of the Alps, the permafrostfeatures so typical for the "Scandinavian fjell" (palsas, pingo-likefeatures, ice-cored moraines, vast large scale polygon patterns) aremissing in the Alps or are restricted to a few selected places.Active phenomena of creeping ice-supersaturated sedlments frommoraines and talus (rockglaciers) dominate instead, and thenumber of rockglaciers matches the number of glaciers. All formsof solifluction are quite widespread, too.The Jura mountains are located in France and Switzerland,northwest of the Swiss Alps. For large areas, they consist ofparallel folds of mesozoic sediments, often limestone. In thenortheast, their continuation reaches to the cuesta-like Alb Imountains in Germany. Alltogether, this mountain area is severalhundred kilometers long and reaches above 1500 m a.s.1. at manypoints. Althou h the MA.+T is abave freezing point, ice-caves(island permatost) may be found at many places even at altitudesbelow 1200 m a.s.1. (Pancza 1992).OTHER EUROPEAN MOUNTAINSBesides the main mountain ranges of continental Europe, the Alpsand the Scandes, there are a great number of larger and smallermountain ranges, where ermafrost has been roven or where itcan be expected, e. thehrenkes, the Ca atiians and theAppennine. Table Hgives the approximateTower limits for island,sporadic and discontmuous mountain permafrost in the mountainsof continental Europe.The lowermost limits for island permafrost include areas of icecaves in karst areas, and the MAAT may therefore reach valuesconsiderably above the freezin point there. At the lower limit ofsporadic permafrost the MAAf 1s in the order of -1°C to -1.5'C,and above the altitude of the -3.X MAAT discontinuouspermafrost (> 50%) can be expected. The maximum 'altitude of therespective mountain range gives a first idea of the area affected byperennially frozen ground.In -[knees, geomorphological research has a longtradition, on the French side as well as on the Spanish one and theinterest in mountain permafrost research is present, too (cp.Gutierrez et al. 1981). There are more than a dozen activerockglaciers in the Spanish PyrknCes in areas reaching above 2800m ad. (Agudo et al. 1992, San Jost et al. 1992, Serrano et al.1991). According to David Palacios (written communication) thearea abovc 2800 m a.s.1. is regarded as the altitudinal belt ofsporadic permafrost (10 - 50 %). In addition to the studies ofactive permafrost, Chueca (1992) has mapped 170 rockglaciers asrelict permafrost forms. The joining of the Spanish national bodyto the <strong>International</strong> Mountain Permafrost Association willcertainly have a positive effect on more detailed mountainpermafrost research in the Pyrenees, where more areas areexpected to he underlaid by permafrost according to the existingMAAT of -1.5"C and lower.In the central part of the Italian-Appennine, the AbruzziFountains, there are three small mountain areas where permafrostcan be expected: Gran Sasso, Maiella and Mente Vellino (Dramis& Kotarba 1992).In the Sauthern Camw,-Romania, inactive and activerockglaciers have been reported in altitudes above 2000 m a.s.1. byUrdea (1992). In the Fagaras, the Paring and the Retezatmountain massifs many mountain tops and crests reach altitudesbetween 2300 and 2500 m ad. and the MAAT at Omu- 1024

Table 1: Approximate lower limits for island (0 - lo%), sporadic (10 50%) anddiscontinuous (> 50%) permafrost in continental Europe. Continuous permafrostexists only in the Alps and in the highest parts of the Scandinavian mountains.Y ugoslaviaAlbaniaBulgaria(2505 rn a.s.1.) is -2.6"C. Here a ain, a large number of relictrockglaciers exist below about f000 m a d. and proove formerpermafrost conditions (Ichim 1978, Urdea 1993).tlLtl!m,The mean annual air temperature in the Northern Cam 'es ecially in the Tatry mountains (Poland, Slovakia) is lower than -I.!&, too, and these climatic conditions undoubtedly favour thedevelopment of sporadic permafrost. Whereas relict permafrostfeatures have been studied, research on the distributlon of activepermafrost is sti!l urgently needed (cp. Czudek, 1993).In Scotland the existence of small permafrost island in thehighlands cannot be excluded accordlng to thexistingtemperatures.ACKNOWLEDGEMENTS:Additional contributions for this report came from the followincolleagues: Matti Seppala (northern Finland), Wilfried Haeberfi(Switzerland), Michele Evin (France), Francesco Dramis (Italy),David Palacios (Spain), Adam Kotarba (Poland) and Petru Urdea(Romania).R,EFERENCES,SVALBARDAkerman, H.J. (1980): Studies on Perflaciql Ceompgholoq inWest Spitsbergen. PhD thesis. Lun s Unwersltets eogra lskaInst. Scr. Avh. LXXXIX: 297 pp.herman, H.J. (1982): Observations of palsas within thecontinuous errnafrost zone in eastern Siberia and in Svalbard.Geografisk Pidsskrift 82: 45-51.herman, H.J. (1987): Periglacial forms of Svalbard - A review. In:Boardman, J. (ed.): Periglacial Processes and Landforms inBritain and Ireland, Cambridge llniversity Press: 9-25.Akerman, H.J. (in prep.): Mountain permafrost in the Nordiccountries, a review. - In preparation for Lunds Univ.Naturgeogr. Inst., rapporter och Notiser (1994).Gri p, K. (1926): Uber Frost und Strukturboden auf Spitzbergen.leitschr. Gesellsch. f. Erdkunde: 351-354.Humlum, 0. (1982): Rock Glaciers in Northern Spitsbergen: Adiscussion. Journ. Geol. 90: 214-218.King L. & M. Volk (in trep.): Glaziolo ie und Glazialmorphologiedes 1-iefde- und Boc f~ordgebietes, Pfordspitzbergen. - Inpre aration for Zeitschrift fur Geomorphologie N.F., Suppl.-bd.(1984)Landvik, J.Y..et al. (1988): Glacial history and permafrost in theSvalbard area. Permafrost, Fifth <strong>International</strong> Permafrost Conf.Proceedings 1: 194-198, Trondheim.khmann, R. (1993): The Significance of Permafrost in theFormation and Appearance of Push Moraines, Examples of theCanadian Arctic and S itsbergen. Sixth <strong>International</strong><strong>Conference</strong> on Permafost, Beijing 1993, Proceedinks vel. 1:374-379.Liest~l, 0. (1Y76): Ping 5 springs, and permafrost in Spitsbergen,Norsk Polarinstitutt f;bok 1975: 7-29.Liestal, 0. (1980): Permafrost conditions in Spitsbergen. Frost iJord 21: 23-28.Liestal, 0. (1986): Permafrost pP Svalbard og pb fastlandet.Klimatiske forutsetnin er, uthredelse og tykkelse.Fjellsprengningsteknik~, bergmekanikk, geoteknikk. Tapir.Salvigsen, 0. (1Y77): An observation of palsa-11 e forms inNordaustlandet, Svalbard. Norsk Polarinst. ftr bok: 364-367.Sollid, J.L. & L. Sorbel (1988): lnfluence of temperaturecondititinsh formation of end moraines in Fennoscdndia andSvalbard. Boreas 17: 553-558.

Sve won 11 1976): Iskilar som klimatindikator. Svensk Geogr.Ar;hok 52L5.7.Van der Watcrw, 1.). (1992): Structural Geology andScdirnentology of Push Moraines. Diss., Faculteit derRllillrrr.lijke'Wt.tenschappen, Univ. van Amsterdam: 230 pp.ICEI.ANX)Eyles, N. (1978): Rock glaciers in Esjufjbll nunatak area, southeastIceland, Jijkull 28: 53-56.Humlum, 0. (1YXS): Genesis of an imbricate push moraine,Hofdahrekkujull, Iceland. Journal of Geolog 93: 185-195.Rutten, M.G. (1951): Polygon soils in Iceland. Geologie enMijnhouw, Mai 195 1Schunkc, E, (1975): Die Periglazialerscheinungen Islands inAhhiingigkcit von Klirna und Substrat. Ahh. Akad. Wiss.Gijttingen 30: 273 pp.Ihorarinsson, S. (1951): Notes on patterned, round in Iceland,with particular reference to the Icelandic kia's". Ceogr. Ann.33: 144"156.FE.NNOSCANDIAAhman, R. (1977): Palsar i Nordnorge. Medd. Lunds Univ. Geogr.Inst. Scr. Avh. 78: 165 pp.Akcrrnan. H.J. 6r l3. Malmstriirn (1986): Permafrost mounds in theAhisko area, Northern Swcden. Geogr. Ann. 68 A, 3: 155-165.Akerrnan, H.J. (in prep.): cp. section on SvalbardBarsch, El. (1971): Rock glaciers and icc-cored moraines. Geogr.Ann. 53 A, 3-4: 203-206.Barsch, D. Rr IJ. Treter (1976): Zur Verhreitung von Periglazialphiinomenenin Honrlane/Norwegen. Geogr. Ann. 58 A: 83-93.Ekman, 5. (1YS7): Die Gewgsser des Ahisko-Gehietes. Kungl.Vetenskansakademiens Hmdlingar 4, e ser. 6,6: 172 pp.Fries, T. Rr E. Bergstriim (1910): Nigra iaktagelser iifver palsaroch deras fiirckomst i nordli aste Sveriga. Geol. Foren.Fiirhandl. Stockholm 32: 19f-205.Hamberg, A. (190s): Till Mean om alltid frustn mark i Sverige.Ymer 24: 87-93.Jeckcl, P.P. (1YXK): Pcrmafrost and its altitudial zonation in N.Lapland. In: Permafrost, Fifth <strong>International</strong> Permafrost<strong>Conference</strong>, Prtrcecdings Vol. 1: 170-175.King, L. (1976): Perrnafrostuntersuchungen in Tarfala, SchwedischLappland, mit Hilfe der llammerschlagseismik. 2. Glctscherk.11. C;\~ia!geol. 12,21 187-204.Kin L (19x2): Qualitative und uantitative Erfassun vonFkrmafrost in Tarfala (Schwelsch Lap land) und fotunheimen(Norwegen) mit Hilfe geoelektrischer gndierungen. Z.Geornorph.. Suppl.Bd. 43: 139-160.King, L. (1983): High mountain permafrost in Scandinavia. In:Permafrost, Fourth <strong>International</strong> <strong>Conference</strong>, Proceedings Vol.1: 612-617, Nar. Acad. Press, Washinfiton, D.C.King, L. (1984): Permafrost in Skandinavien,Untersuchungsergehnisse aus bppland, Jotunheimen undDovre/Rondane. Heidelberger Geogr. Arb.76: 174 pp.King, L. (1986): Zonation and ecology of high mountainpermafrost in Scandinavia. Geogr. Ann. 68 A: 131-139.Kine, L. & M. Seppiiki (1987): Permafrost thickness anddistribution in Finnish Lapland - results of geoelectricalsoundings. Polarforschung 57,3: 127-147.LagerbHck, R. & L. Rodhe (19851: Pingos in northernmostSweden. Geogr. Ann. 67 A, 3-4: 239-245.Lundquist, J. (1962): Patterned ground and related frostphenomena in Sweden. Sveriges Geologiska Undersijkning, Ser.c 55: 101 pp.Malmstriim, B. (1988): Occurence of permafrost on the VarangerPeninsula, northernmost Norway. Norsk Geogr. Tidsskr. 42:239-244,Meier, K.-D. (1985): Stud-ien zur Verhreitung. Morphologie,Morphodynarnik und Okolo ie von Palsas auf der Varanger-Halbinsel, Nord-Nonvegen. kssener Geographische Arbeiten10: t14-204.@degSrd,R. et ai. ( 1992): Ground temperature measurements inmountaln permafrost, Jotunheimen, Southern Norway.Permafrost and Periglacial Processes 3: 231-234.Ostrem, G. (1964): Ice-cored moraines in Scandinavia. Geogr.Ann. 46: 282-337.Ra p. A (1982): Zonation of permafrost indicators in Swedishkpland. Geogr. 'I'idsskrift 82: 37-58.Rapp, A. & L. Annersten (1969): Permafrost and tundra olygonsIn northern Sweden. In: Ptwt, T. (ed.): The periglaciafenvironment. Montrtal, Canada.Ragp, A. &.G.M. Clark (1971): Large nonsorted polygons inadjelaota National Park, Swedish Lapland. Geogr. Ann. 53 A:71-85,Rapp, A. Rr S. Kudher (1960). Recent periglacial phenomena inSweden. Bid. Pcrigf 8:.143:154.Seppala, M. (1988): Palsas and related forms. In: M.Y. Clark (ed.):Advances in Periglacial Geomorphology: 247-278.Smith, H. (191 I): Postglaciala regionforskjutnin ar i norraHarjedalen och siidra JBrntlands fjalltrakter. keel. F8ren.Stockholm FBrh. 33: 503-530.Sollid, J.L. & L,. Sarhel (1974): Palsa bogs at Hau~tj0rn/o,Dovrefjell, South Norway, Norsk Geografisk TldsskrIft 28,l: 53-60.Sollid, J.L. & L. Smbel(1988): Influence of temperatureconditions in formation of end moraines in Fennoscandia andSvalbard. Boreas 17: 553-558.Sollid, J.L, & L. Sorbel (1992): Rock laciers in Svalbard andNorway. Permafrost and Perlglaaa! Processes 3,3: 215-220.Svensson, H. (19624: Nigra iakta elser frh pals ornrfidon. NorskGeogr. Tidsskrift 18, 5-6: 212-2h.Svensson, H. (1962b): Tundrarlygons. Photographicinterpretation and field stu les In North Norwegian polygonareas. Arb. Norges Geol. Unders 223: 298-327.Svensson, H. (1969): A type of circular lakes in northernmostNorway. Geqgr. Ann. S1 A: 1.12.

statusSvensson, H. (1986): Some morphoclimatic aspects of periglacialfeatures of northern Scandinavia. Geogr. Ann. 68 A:123-130.Wramner, P. (1973): Palsmyrar i Taavavuoma, La land.Giiteborgs Univ. Naturgeogr. Inst. Rapp. 3: 14d)pp.ALPS AND JURABarsch, (1978): Active Rock Glaciers as Indicators forDiscontinous Alpine Permafrost, an Example from the SwissAlps. Proceedings of the Third <strong>International</strong> <strong>Conference</strong> ofPermafrost, July 10-13, 1978, Edmonton, Alberta, Canada. Vol.1: 349-352, Ottawa.Barxh, D. (1992): Permafrost Cree and rockglaciers. Permafrostand Periglacial Processes 3,3: lkI88.Belloni, S., A. Carton, F. Dramis & C. Smiraglia (1993):Distribution of permafrost, laciers and rock glaciers in theItalian mountains and correfations with climate: an atteqt tosythesize. Sixth <strong>International</strong> <strong>Conference</strong> on Permafrost, eyng19Y3, Proceedings Vol. 1: 272-277.Cheng, G. & F. Dramis (1992): Distribution of mountainermafrost and climate. Permafrost and Periglacial Processes 3,8: 83-91.Evin, M., D. Fabre, A. Assier & Ch. Guillain (1993): Les glaciersrocheux du Roure. Socittt ilydrotechnique de France, sectionglaciologie, Grenohle: 6 pp.Francou, B. & L. Reynaud (1992): 10 ears surficial velocities on arock lacier (Laurichard, French dps). Permafrost andPerigyacial Processes 3,3: 209-214.Gorbunov, A. P. (1978): Permafrost investigation in high mountainregions. Arctic and Alpine Research 10,2: 283-294.Guglielrnin, M. & C. Tellini (1993): First example of permafrostmapping with BTS in the Italian Alps (Livi no, Sondrio, Italy)Acta-Naturalia de I’Ateneo Parmenese 29,gN. 1/2 39-46. ’Haeberli, W. (1985): Creep of mountain permafrost: internalstructure and flow of alpine rock laciers. Mitt. derVersuchsanstalt fiir Wasserbau, kfydrologie und Glaziologie 77,ETH Zurich: 142 pp.Haeberli, W. (1992): Construction, environmental problems andnatural-hazards In periglacial mountain belts. Permafrost andPeriglacial Processes 3.2: 11 1-124.Haeberli, W. (1993): Research on permafrost and periglacidprocesses in mountain + areas and perspectives. (Thisvolume)Haeberli, W., M. Evin, G. Tenthorey, H.R. Keusen, M. Hoelzle, F.Kellef, D. Vonder Muhll, S. Wagner, M. Pelfini & C. Smiraglia(1992): Permafrost research sites in the Alps: excursions of the<strong>International</strong> Workshop on Permafrost and PeriglacialEnvironments in Mountain Areas. Periglacial Processes 3,3:189-202.Haeberli, W. & M. Funk (1991): Borehole temperatures at theColle Gnifetti core-drillin site (Monte Rosa, Swiss Alps):Journal of Glaciology 37,725: 37-46.Haeberli, W., L. King & A. Flotron (1979): Surface movementsand lichen-cover studies at the active rock glacier near theGrubengletscher, Wallis, Swiss Alps. Arctic and Alpine Res. 11,4: 421-441.Hoelzle, M., W. Haeberli Br F. Keller (1993): Application of BTSmeasurementsfor modelling mountain perma rost dlstrlbutlon.Sixth <strong>International</strong> <strong>Conference</strong> on Permafrost, Beijing 1993,Proceedings Vol. I: 272-217.Keller, F. (1992): Automated map ing of mountain permafrostusing the program PERMAdT within the geo raphicalinformation system ARC/INFO. Permafrost and B eriglacialProcesses 3,2: 133-138.King, L. (1990): Soil and rock temperatures in discontinuousrrmafrost: Gornergrat and Unterrothorn, Wallis, Swiss Alps.ermafrost and Per~glacial Processes 1,2: 177-188.King, L., A. P. Gorbunov & M. Evin (1992): Prospecting andmapping of mountain permafrost and associated phenomena.Permafrost and Periglacial Processes 3,2: 73-81.Pancza, A. (1992): gtlivation des parois rocheuses dans unefacikre du Jura NeuchBtelois. Permafrost and Periglacialrocesses 3,2: 49-54.Vonder Muhll, D. & W. Haeberli (1990): Thermal characteristicsof the permafrost within an active rock lacier(Murttl/Corvatsch, Grisons, Swiss Alps!. Journal of Glaciology36, 123: 151-158.Vonder Muhll, D. & P. Holub (1992): Borehole lo ing in AlpineFmafrost, Upper Engadin, swiss ~ ps. Permakt anderiglacial Processes 3,2: 125-132.Wa ner, S. (1992): Creep of Alpine permafrost, investi ated ontte Murttl rock glacier. Permafrost and Periglacial jrocesses 3,2: 157-162.OTHER EUROPEAN AREASAgudo, C., E. Serrano & E. Martinez de Pison (1989): El laciarrocoso activo de Los Gemelos en el Macizo de Posets birinboAragones): Cuaternario y geomorfologla. Vol. 3: 83-91.Chueca, J. (1992): A statistical analysis of the spatial distributionof rock glaciers, Spanish Central Pyrenees. Permafrost andPeriglacial Processes 3,3: 261-265.Czudek, T. (1993): Pleistocene Periglacial Structures andLandforms in Western Czechoslovakia. Permafrost andPeriglacial Processes 4, 1: 65-76.Dramis, F. & A. Kotarba (1992): Southern limit of relict rockfaciers, Central Appennines, Italy. Permafrost and Periglacialrocesses 3,3: 257-260.Gutierrez, M. Kr J.L. Pena (1981): Los glaclares rocososy elmodelado acompahante en el Brea de la Bonaigua (PlrinCo deUrida). Bol. Ceol. Mintro 92: 11-20.Ichim, 1. (1978): Preliminary observations on the rock lacierphenomenon in the Romanian Carpathians. Revue kournaineGtol. Gtof. Gtogr., Gtographie 23,2: 2Y5-299.San Josk, J.J. De, C. Agudo, E. Serrano & F. Silio (1992):Auscultaci6n topografica y fotogrametrica del 8laciar rocoso deLas Ar ualas (Pirinto Aragones): datos prelimlnares. In: F.Lo ez bermudez. C. Conesa Garcia & M.A. Rornero Dim(e&): Estudios de geomorfologia en Es aha, actas de lareunmn de la I1 reunion nacional de GI&: 423-431, Murcia.Serrano, E., E. Martinez de Pison et al. (1991): El glaciarnoroccidental del Besiberri (Pirinto de Urida). Pirintos 137:95-109Urdea, P. (1992): Rock Glaciers and Periglacial Phenomena in theSouthern Carpathians. Permafrost and Periglacial Processes 3,3: 267-273.Urdea, P. (1993): Permafrost and Periglacial forms in theRomanian Carpathians. Sixth <strong>International</strong> <strong>Conference</strong> onPermafrost, Beljing 1993, Proceedings Vol. 1: 631-637.1027

STUDIES ON MOUNTAIN PERMAFROST IN ASIAQiu GuoqingLanzhou Institute of Glaciology and GeocryologyChinese Academv of Sciences,T..anzhou 73000,ChinaAsia has the largest permafrost area in the world. Thc existence of frozen ground and frost action wasknown long ago, while the systematic investigations on mountain permafrost were carried out since the50's of this century. The mountain permafrost can be divided into threeAbroad categories, ;.e., thepermafrost dcveloping in mountains in the continuous belt of the Eurasian continental permafrost zoneon a frigid background, the permafrost developing in mountains out side the Eurasian continentalpermafrost zone on a warm to temperate background and thc permafrost in mountains in thediscontinuous belt of the Eurasian continental permafrost zone. The distribution of mountainpermafrost has a closc relation to the climatic variation in three dimensions. Much more work is neededfor searching a perfect classification of mountain permafrost.HISTORY. . " "Asia has the largest permafrost area in the world. The existenceof frozen ground and frost action was known long ago. InChina, the Rook of Rites: Seasons, that was written 2000 years ago,described cxactly the seasonal fluctuation of climate and the processesof ground freezing and thawing (Shi et al.,1964); the outstandingChincse geographer Xu Xiake had reported the existence.Of block field and the difference in frost action bctween the northandsouth-facing slopes in the Mt. Wutaishan of Northern Chinain 1633 (Xu, 1980). In West Tianshan, frost fissures and frostmounds were first reported by V.I. Roborovsky in 1893; the firstinformation about permafrost in AK-Say basin of Inner Tianshanwas given by A.l. Reasonov in 1913 (Gorbunov,l993). With the exploitationof Siberia since the 19th century, many data and knowledgewas obtained and the geocryology became an independentdicipline lirstly in Russia in this century.Since 1950'S, permafrost investigations werc undertaken in theNortheast China in the interest of forest cxploitation and railroadconstruction in the Mt. Da-Hinggan Ling region. During the periodfrom 1950's to 1970'S, comprehensive invcstigations on naturalcondition and resource were carried out in West China, many publications,as the rcsults of those expedilions, emphasized the importanceof altitudinal zonation and the frost actlon in mountains.Special permafrost researches were carried out in the Mts. AltaiShan, Qilian Shan, Qinghai-Xizang Plateau of China, thc Mts.Zayilisky Alatau, West Tianshan, Inner Tianshan and Pamir of theUSSR / CIS, the Mts. Rhentei and Khangai of Mongolia sincc1960's. In the past 3 years, a joint expedition in Northern Tianshanwas undertaken by the Lanzhou Institute of Glaciology andGcocryology, Academia Sinica and the Pcrnmaf'rost Instilute ofSibcria Branch of the Russian Academy of Sciences.On the basis of the previous work, it is possible to review anddiscuss some aspects of the permafrost studies in mountain areas ofAsia.DISTRIRUTTON OF MOUNTAIN PERMAFROST-According to Gorbunov (1988), the mountain permafrost"should include only those percnnially frozen soils which occurover 500 m above sea level, and is absent below this level". Chengand Dramis (1992) suggested that the term "mountain permafrost"is used to include both alpine permafrost and polar mountainpermafrost, Which is not conventionally considered as alpinepermafrost. In this revicw, Cheng's concept is used.Generally, the mountain permafrost can be dividzd into threebroad categories, Le.,1. the permafrost developing in mountains in the continuousbelt of the Eurasian continental permafrost zone, on a frigid background;2, the permafrost developing in mountains outside theEurasian continental permafrost Lune, or on a warm to tempcratebackground;3. the permafrost developing in mountains in thediscontinuous belt of the Eurasian continental permafrost zone,this is the transitional type of the first two.The pcrmafrost in mountains of the northern part of EastSibcria is the typical one of the first type. There, the mean annualair temperature ranges from -.7.5 to -15 OC or lower, permafrost ispresent everywhere with a mean annual ground tcmperarure of -3to -1 1 "C in plains, depressions and valleys, -7 t,' -15 'C at mountains(Ershov et al, 1989 a). Such a mountain permafrost developingon an extremely frigid background without a lower limit wasnamed the "Verkhoyangsk type alpine permafrost by Gorbunov(1978). Such kind of mountain permafrost is also seen in the Mt.

RYrranga and Mt. Putorana of the northern part of Central SiberiaPlateau,Outside the Eurasian continental permafrost zone, thc &vel-OPment of mountain permafrost depends mainly on the variationof climate in three dimensions, For examplc, in the ChineseTianshan, according to the statistics using the data from 17stations, the dependence of the mean annual air temperature T onthe latitude x], longitude X2 and altitude X3 is known as follows(Qiu, 1993):T= 109.28-2.19X1-0.024X2-0.005IX3 (1 1The'correlation coefftcient R of (I) is 0.9596. The correspondingstandard rcgrcssion coefficients are:R1 "0.48 forX1;B2 = 4.025 for X2;B3 =-0.93 for X3.For the vast area from Mt. Qilian Shan to the Mt.Himalaya,according to the statistics by using the data from 78 stations,T=66.25-0.92X1-0.14X24.0056X3 (2)The corrclation coefficient R of (2) is 0.9530. The correspondingstandard regression coefficients are:R1 =-0.77 forX1;B2 = -0.20 for X2;B3=-1.12forX3.Thus, it could be concluded that in the mountains of WestChina the mean annual air temperature decreases very obviouslyfrom lowland to high mountain, decreases obyiously from south tonorth and slightly decreascs from west to east. Such a variation ofmean annual air temperature in three dimensions influences in thedevelopment of permafrost in high mountains, the cold islandsstanding on the tcmperate-warm background in the middle-lowlatitudes while the lower limit of the discontinuous mountainpermafrost ascends southwards (Zhou et al, 1991):2700-3100 m a.s.1. in Mt.Tianshan;3500-3900 m a.s.1. in Mt.Qilian Shan;4150-4200 m a.s.1. at the Xidatang of the Mt. Kunlun Shan;4600 m a.s.1. on the south sidc ofMt.Tangula, and5100-5300 m a.s.1. in the Mt. Himalaya Shan.Such a latitudinal variation of the lower limit is also sccn in thecentral Asia part of CIS, e.g., it riscs from 3200-3700 m a.s.1. in theMt.Zayilisky Alatau up to 3600-4000 m in Pamir (Gorbunov,1978).The lower limit of the mountain permafrost also descendseastwards. For example, along the 43'N latitude, it lies at 3300ma.s.1. in the Ralshaya Almatinka Permafrast Station in Kazakhstan,2900-3250 m a.s.1. in the Chinese Tianshan Glaciological Station(Qiu, 1993),1900-2000 m a.s.1. in the Mt. Changbai Shan (Zhou etal, 1991) and the Northeast Korea (Viktor An, 1993), and down to1700 m a.s.1. in Hokkaido.In the Mt.Qilanshan at 37'-40"N. the permafrost lowcr hitdescends from 3900 to 3500 m a.s.1. eastwards; in the Mt.WutaiShan at 38"N it lies at 2300 m a.s.1. In the Mt. Kunlun Shan at36"-34'30'N, it descends from 4400-4500 m in the northwest,4150-4200 m a.s.1. in the middle and 3850-3900 m at the southeastend; in the Mt. Taibai Shan (3CN) in the Shannxi Province, it liesat 3000 m a.s.l.(Zhou et al, 1991). All of these document aneastward descent of the lower limit of mountain Permafrost, andthis would result from the decrease of air ternperaturc in the Samedirection.The permafrost regions in the Mongolian Republic and theNortheast China are in the south part of the Eurasian ContinentalPermafrost Zone. 63% of the territory of the Mongolian Republicis underlain by permafrost (Lombodchen, 1993), where the lowestand southernmost position of the permafrost islands was consideredas both the lower limit of the sporadic permafrost belt and thcsouth limit of the permafrost zone(Joint Soviet-Mongolia ScientificResearch Geological Expedition, 1974). Although the continuityof permafrost tends to increasc northwards, showing a latitudinalzonation, it is also obvious that the permafrost with higher continuity,lower temperature and a greater thickness occurs in the coldcenters or thc three mountain systems i.e., the Mt. MongolianAltay in the southwest, the Mt.Khangai in the west and theMt.Kentei in thc middle part of Northern Mongolia, standing onthe background of sporadic island permafrost belt, in other words,the latitudinal zonation and altitudinal zonation are of the sameimportance in the development of permafrost, while in thc otherkinds of mountain permafrost region there is not such a wide sporadicpermafrost belt. In Northeast China, the south limit ofpermafrost is also determined according to the southernmost positionof permafrost islands( Guo et al, 1981). The west secor of thesouth limit approximitely coincides with the 0 to -1 "C isotherm,the middle --O'C, and the east--0 to t 1'C. Thc W-shanped runningof the south limit might result from the special terrain with theMt.Acrshan (700-1200m a.s.1.) in the southwest and the Mt.XiaoHinggan Ling in the southeast and the lower plain in thc middle(Guo et al, 1981). The fact that the permafrost increases in continuityand thickness and decreascs in temperature shows a latitudinalmnation. The coldest and thickest permafrost occurring in thenorth is the result of the altitudinal zonation,and the terrain has alsoefTect on it, Although different in terminology, the distributionof permafrost in Northeast China and Mongolia can be compsrcdto each other. The sporadic scarce-island- and island-pcrmalrostbelts in Mongolia is corresponding to the island-permafrost bclt inNortheast China, the Predominantly continuous belt in NortheastChina might correspond to the discontinuous belt, according tothere continuity. The mountain permafrost in the Zabaykal rcgionand the southern part of East and Central Siberia is on the backgroundof the discontinuous permafrost belt of the Eurasian ContinentalPermafrost Zone.CLASSIFICATIONAn important problem in mountain permafrost studies is toprovide a perfect principle for the classification related to the distributionof permafrost.It was suggested that the alpine permafrost be divided into thecontinuous, discontinuous and sporadic zones like that in the polarand subpolar regions, but the definitions would be somewhat different.If above a certain altitude the permafrost distributes everywhere,then this altitude could be defined as the lower iimit of thecontinuous permafrost; if above a certain altitude only on somesides of the slopes can the permafrost occur, then this altitudc couldbe defined as the lower limit of the discontinuous pcrmafrost; belowthe discontinuous permafrost zone, only in scmc localities witha special cryogenic condition can the permafrost occur, then this

kind of permafrost could be defined as the sporadic one(Corbunov,l978). The permafrost that develops under the thickforest and moss cover and the permafrost that develops in thehigh-porosity block fields is the typical sporadic permafrost inMt.Tianshan. This classification is quite simple and useful inpcrmatiost investigation and mapping to distinguish the developmentcondition and distribution characteristics of thecryolithonones. It is questioed that the mounatin mass is cutseparately and is discontinuous, the individual summit of themountain mass rose up to a high absolute altitude with low a meanannual air temperature,say, as low as -10 to -1S0C, of course, thesummit itself is subject to perennial freezing to a depth of severalhundred meters, the temperature in the cryolithozone may be lowerthan -5 to -10 'C, and all sides of the slopes are undcrlain bypcrmafrost and could be defined as the continuous permafrostzone;however, on the background of the whole mountain system inmiddle-low latitudes, the cryolothozone is only a small island occupying.Only several percent of the whole mountain area, in thisway the cryolothozone is discontinuous. Thus, there was anothersuggestion that the permafrost be divided into the stable, lessstable, unstable and extremely unstable ones on the basis of theirtemperature and thickness (Cheng and Wang, 1982). This classificationmight be more perfect theoretically, however, it needs manydata in temperature and thickness obtained from boreholes andother observation sites, so it could only be used in some areas beingstudied in detail. A perfect classifiction should be available for theplanning of engineerings and for the prediction of possible changeof environment in the near future, in addition, it should also besimple and clear on the basis of parameters easier to obtain. Muchmore work is needed for searching a perfect classification of mountainpermafrost.REFERENCE". ., ."Cheng Guodong and Wang Shaoling,1982,0n the permafrostzonation of the high-altitude permafrost in China. Journal ofGlaciology and Geocryology, 4(2),pp.l-17.Cheng Guodong and PDramis, 1992, Distribution of mountainpermafrost and clirnte. Permafrost and Periglacial Processes.3.2,pp83-91.Ershov, E.D. et al, 1989 a, Geocryology in USSR (EasternSibcria),"NEDRA" Press, Moscow. p.515.Ershov,E.D. et al, 1989 b, Geocryology in USSR(Centra1 Siberia)"NEDRA" Press, Moscow. p.414.Ershov,E.D. et al, 1989c, Geocryology in USSR (Mountain Regions in Southern USSR)."NEDRA" Press, Moscow. p.359.Gorbunov, A.P., 1978, Permafrost investigations in high-mountainregions. Arctic and Alpine Research.VOl.lO,N0.2.pp283-294.Gorbunov, A.P., 1988, The alpine permafrost zone of the USSR.in:Proceedings of The Fifth <strong>International</strong> <strong>Conference</strong> onPermafrost. Vol.1. Tapir Publishers,Trondhcirn.pp.lS4-158.Gorbunov, A.P.,1993, Geocryology of the Tianshan. in:Proceedingsof The sixth <strong>International</strong> <strong>Conference</strong> on Permafrost.Vol.2.pp. 1105 -1107 .Guo Dongxin, Wang Shaoling, Lu Guowei and DaiJinbo,l981,Division of permafrost regions in the Da andXiao-Hinggan Ling of Northeast China. Jownai ofGlaciology and Geocryology, 3(3), pp.1-9.Joint Soviet-Mongolian Scientific Research GeologicalExpedition, 1974, Geocryological conditions of the MongolianPeople's Republic. "NAUKA" Press, Mogcow. p.200.Eomborichen, R.,1993,Cryogenic processes and phenomean inMongolia. in : Proceedings of the Sixth <strong>International</strong> <strong>Conference</strong>.Vol.1. South China Technology University Press.pp.411-415.Permafrost Institute of Siberian Branch, Academy of Sciences,USSR, 1974, General Geocryology (ObshayaMerzlotovedenye). Science Press, Beijing. p.318.Qiu Guoqing, Huang Yizhi and Li Zuofu, 1983, Alpine permafrost- in Tianshan, China. in: Procecdings of 4th <strong>International</strong> <strong>Conference</strong>on Permafrost. National Academy Press, WashingtonD.C. pp.1020-1023.Qiu Guoqing, 1993, Development condition of alpine permafrost inMt.Tianshan, China. Proceedings of Sixth <strong>International</strong> <strong>Conference</strong>on Permafrost, Vol. I. South China Technology UniversityPress. pp.533-538.Shi Yafeng et al, 1964, Basic characteristics of existing glaciers inChina. ACTA Geographica Sinica.30(3). pp.183-208.Shi Yafeng et al, 1988, Explanation to the map of snow, ice andfrozen ground in China (1:4000000). Cartographic PublishingHouse, Beijing. p.32.Viktor An, 1993, Permafrost in the north of Korean Penisula. in:Procecdings of the Sixth <strong>International</strong> <strong>Conference</strong> onPermafrost, Vol.1. South China Technology UniversityPress.pp.843-845.Xu Hongzhu, 1981, Xu Xiake's Travels. Acient Literature PublishingHouse, Shanghai. p.82.Zhou youwu, Qiu Guoqing and Guo Dongxin, 1991, Quaternarypermafrost in China. Quaternary Science Revicw.vo1.1o.pp.511-517.1030

LINEAR CONSTRUCTION IN COLD REGIONS - PAVED ROADS AND AIRFIELDSTed S. VinsonProfessor, Department of Civil Engineering, Oregon State University, Corvallis, OR 97331The performance of paved roads and airfields in cold regions may be categorized with rcspect to three modes ofdistress: (1) distortion and pavement faulting, (2) disintegration, and (3) cracking. By far the most prevalent problemfor an asphalt concrete pavement is cracking. Cracking may be traffic/load associated or non-traffic/load associated.TrafficAoad associated cracking may be related to fatigue failure and subgrade rutting in an asphalt concrete pavementstructure. Low temperature cracking may be caused by two distress mechanisms. First, trnnsverse cracks can extendthrough the entire pavement structure and into the subgrade. This type of transverse crack is associated primarily withthe thermal contraction of soil (in the base, subbase, and/or subgrade) rather than'the asphalt concrete surface layer.Second, transversc thermal cracks can occur wholly in the asphalt concrete surface layer. Methodologies presently existto allow the resistance of an asphalt concretej&ement to fatigue, subgrade rutting, and low temperature cracking tobe quantified and easily incorporated in pavement design for the subarctic and arctic.INTRODUCTIONRoads and airfields located in the arctic and subarctic may beclassifid iuto three distinct groups: (I) gravel surfaced, (2)"flexible"- bituminous surfaced, and (3) "rigid"- portland cement concrete(PCC) surfaced. For example, the Alaska Department of Transportationand Public Facilities (Alaska DOTPF) maintains a total centerlinenetwork of about 8000 km (5000 mi) of which about 3520 km (2200mi) are paved. Essentially all of the paved roads are bituminoussurfaced (asphalt concrete (AC), bituminous surface treatment, sealcoat, etc.). However, in Russiaessentially all paved roads in the arcticand subarctic are rigid portland cement concrete (Le., prefabricatedslabs).Virtually all roads and airfields in the subarckic and arctic wereinitially constructed of a gravel fill placed directly on the existingvegctation to take advantage of the insulation and latent heat capacityof the surface organic laycr (Crory, 1988). The bearing surface wascreated by adding fines to "bind" the gravel, which was then compactedand graded. The binder material wns necessary because theaggregatcs which were used were not crushed.Since the number of paved roads and airfields in the subarctic andarctic is not great, their performance and factors which cause distressto asphalt concrete may not be fully appreciated by professionals involvedin their design and construction. Thc purpose of this paper isto providaa review of the performance of paved roads and airfields inthe subarctic and arctic and to briefly discuss design mcthodologies torelate to three major distress modes in an asphalt concrete pavement,namely, fatigue, subgrade rutting, and low temperature cracking.PERFORMANCE OF PAVED ROADS ANDSUBARCTIC AND ARCTICAIRFIELDS IN THERoad and airfield pavement performance in cold regions may becategorized with respect to three modes of distress (Vinson et ai.,1986): (1) distortion and pavement faulting, (2) disintegration; and (3)cracking. In the discussion Khat follows, a background of these distressmodes together with a limited review of the current stateof-the-art ofresearch and/or practice is presented.Distortion and Pavement FaultingFrost heave and thaw degradation in the active layer are primarycauses of distortion (movement) and faulting of pavements in the arcticand subarctic. Distortion distress is also associated with buriedstructures remaining fixed while the surrounding mil heaves or"jacking" out of the ground under successive freeze-thaw cycles.Frost heave distortion is greatest toward the end of the winter.Evidence of differential movements for asphalt concrete pavementsafter spring thaw is generally suggested by scrape marks from a snowplow blade. Very often, howeverp differentid movements are nitobvious after spring thaw. Remnant evidence of differential movementmay be associated with loss of fines beneath a distorted section thatresults in "lipping" at a crack or the formation of a "birdbath"..The greatest differential movements observed over a one-yearperiod are generally associated with differences in soil frost heaveresponse, for example, a difference in soil response may occurbetween a backfill material in a culvert trench and the adjacent soilunderlying the pavement. If the backfill soil is identical to the adjaccntsoil underlying the pavement, the differential movements may bereduced, but generally cannot be eliminated if the soil is frostsusceptible.Berg and Johnson (1983) noted that drains, culverts, or utilityducts placed under pavements on frost susceptible subgrades oftenexperience differential heave and should be avoided. They provideguidelines for transition zones for culverts or utilities that must beplaced beneath pavements on frost susceptible soils 88 well 96longitudillal and transverse transitions accommodate to interruptions inpavement uniformity.Rice (1975) succinctly identified the causes of frost heave as thethree W's: winter, water.wick (is., cold temperatures, access to1031

watcr, n frost susceptible soil). It is univcrsally recognized thatmitigation of distress relatcd to frost heave involvcs the elimintltion ofone or more of these factors. Considering the evolutionary history ofmost roads and airficlds it may not be prnctical to rcmove frostsusccptihle soils since they often comprise must of thc emhankmcnt.Consequently, insulation is expericncing increased use to prevent theadvance of the freezing front into the embankment to mitigate frostaction (Kcstlcr and Berg, 1989).A considertltion of frost heavc suggests the related problem ofthnw weakening in the pavement structure. Pavement deteriorationunder repeutd loads is a process of cumulative damage. During springthaw, thc supporting capacity of a pavement surface layer provided bythe base, suhhase, or subgrade, can be reduced owing to excess porewater in the supporting layers. Under these conditions damageaccumulation for a given traffic volume and load is greatest and canlead to a substantial reduction in overall pavement life (Rutherford andMahoney, 1986). An example calculation which follows the mechanisticapproach to pavement design (Mahoney and Vinson, 1983) suggeststhe magnitude of the problem at the Nome Airfield (Vinson andRooney, 1991). Considering the potential for a fatigue failure in theasphalt concrete under three design aircraft'loadings at the NomeAirfield, it wns demonstrated that 85 56 of the annual damage occursin a one month "typical" spring thaw condition.Adequate drainage provisions can mitigate thaw weakening anddamage accumulation in a pavement structure. As thaw progressesfrom the surface downward the water released can, in general, onlydrain upwards (since the ground is frozen beneath and lateral redistributionis oftennot possible owing to slowerthawing and/ or lesspermeable soils in the vicinity of the shoulders). The situation pointsto the definite need for free draining base and subbase courses andlongitudinal drains to remove the thaw water. Impedance of subsurfacedrainage elements caused by frozen soils must be considered in thedesign process (Berg and Johnson, 1983).Distortion distress of pavements in the subarctic and arctic causedby thaw-consolidation of underlying ice-rich permafrost is related toconductive and/or convective heat transport processes. Representativecase histories which document this problem have been presented forNome (Rooney et al., 1988), Kozebue (Esch and Rhode, 1976), andBethel (McFadden and Seibe, 1986) airfields. It is extremely importantwhen seeking a solution to a permafrost degradation problem toidentify the contribution related to conductive versus convective heattransport processes. Permafrost degradation related to conduction maybe associated with an inadequate thickness of gravel fill and/or thechange in albedo of the surface when it is pavcd. Traditional solutionsfor a conductionproblem are to increase the thicknessof the fill(Hennion and Lobacz, 1973) or insulate the problem arm (Esch 1973,1986; Esch and Rhode, 1976; Rooney et ai., 1988). Non-traditionalsolutions include changing the surface albedo using paint (Fulweiderand Aitken, 1963; Berg and Aitken, 1973; Berg and Esch, 1983), the ,usc of thermoprohes (or therrnosyphons) (McEadden and Siebe, 1986).prevention of snow acting as an insulation blanket during the winter(Zarling et al., l988), use of air duct systems (Zarling et al., 1983),and prethawing followed by dynamic consolidation (Rooney et ai.,1988).Permafrost degradation related to convective heat transport isassociated with ground water flow beneath the airfield or through theembankment. The influence of convection is to cause a positive inflowof heat into the system, thereby slowing down or preventing long-termfreeze back or, alternatively, accelerating the depth of thaw. Thesolution to a convective heat transport problem generally involves themodification of an existing subdrain system and/or the enhwccmcnt ofsuhsurface drainage away from the pavcrncnt structure embankment,Rutting distortion can occur in the asphalt concrete surface laycr.The mechanistic approach of pavement design present4 in a latersection of this paper considcrs rutting related to the accumulation ofdeformations associated with vcrtical strain at thc top of the suhgrade.A discussion of rutting related to the use of low viscosity wphdtcements in cold regions is given by Janoo (1989). He notes that ruttingis related primarily to the aggregate in the mix and to a much lesserdcgree the grade of the asphalt cement. To minimize the potential forrutting, a well-graded angular aggregate (with two or more fracturefaces) is recommended. The occasional practice of adding one-halfpercent to the design asphalt cement content should be avoided as itmay contribute substantially to rutting.DisinteErationDisintegration is the breaking up of a pavement into small, looseparticles. Disintegration may be acceterntd by freeze-thaw cycles, ortraffic loading, especially adjacent to cracks.In an asphalt concrete pavement, disintegration isgenerally relatedto insufficient asphalt cement content in the mix, poor compaction ofthe mix (which may be related to cold weather construction (Eaton andBerg, 1978)), overheating of the mix, or asphalt stripping. The firstthree problems can be avoided by carefully following conventional mixdesign practice and conscientiously supervising the construction of thepavement. The potential for stripping may be overlooked!Stripping of an asphalt concrete pavement is the loss of adhesionbetween the asphalt cement and the aggregate. Stripping is due to theactionofwater or water vaporin the asphalt concrete pavement.Specifically, water gets betwccn the asphalt cement film and the aggregate surface. Since the aggregate surfacc generally has a greaterattraction for water than asphalt, the water is drawn between theasphalt cement and aggregate surface and strips the asphalt away fromthe aggregate. The rate at which stripping takes place depends on thetemperature, type of aggregate, and viscosity and composition of the__asphalt (Tyler, 1938). A summary and evaluation of laboratory testprocedures used to identify the potential for stripping has beenprepared by Terrcl and Shute (1989).Two characteristic types of pavement failures are associated withstripping. If water enters the asphalt cement pavement through theupper surface, raveling of the aggregate occurs. If stripping occursfrom the bouom of the pavement upwards, random cracking and 'potholing is generally not detected until it is too late to prevent.Concern for stripping suggests that the asphalt concrete should bedensely compacted to achieve maximum impermeability. If thepavement has a high voids content, water will enter at the surface andcreate the potential for stripping. Further, water can enkr thepavement through cracks.Wintcr snow removal practices can also create potential forstripping. Snow plowed to the sides of a roadway prevents the frozenshoulders from thawing during warmer periods. The frozen shouldersact as a barrier to drainage of free water provided by deicingsaltslsolutions or snowlice thaw associated with the heat absorbingblack asphalt pavement.CrackingBy far the most prevalent airport pavement performance problemis cracking. Cracking may be trafficlload associated or non-trafficlloadassociated. With respect to roads, both traffic/ load and non-trafficlloadassociated cracking may exist. With respect to airfield pave1032

ments, non-trafficlload associated problems related to changes intemperature in the pavement structure and underlying#ground arepredominant.Fatigue of asphalt concrete refers to cracking caused by repeatedbending due to traffic. Fatigue related to pavement cracking is oftenreferred to as "alligator cracking." The mechanistic approach topavement design presented in a later section of this paper considersfatigue cracking related to the accumulation of deformation associatedwith tensile strains at the bottom of the asphalt concrete surface layer.Reflection cracks are an expression of the crack pattcrn in anunderlying pavement. They are causcd by horizontal and/or verticalmovements in the pavement beneath an overlay. Reflection cracks maybe observed in both asphalt concrete overlays on old PCC pavementsand asphalt concrete pavements. Despite extensive work usingtechniques such as stress and strain relief interlayers, geotextiles, orreinforcing in the overlay, an monomic solution to prevent reflectioncracks does not exist. Reflection cracks have been observed within 6to 12 months after construction of an overlay on a 50 mm asphaltconcrcu surbe that was cold milled to an initial thickness of 12 mm(Vinson et al., 1986). The only way to completely eliminate reflectioncracking is to remove the old pavement.Reflection cracks have been associated with the use of a cementtreated base (CTB). Vita et al. (1988) did not find this to be the caseat the Bethel, Alaska, Airfield which consists of a CTB underlying dnasphalt concrete pavement. CTB reflection cracks did not appear to bea serious problem at khel. In fact, the construction and performanceof CTB in cold regions has been very successful and it should be consideredfor more routine use in the future (Vinson ef al., 1984).Two distress mechanisms are believed to cause thermal crackingin the subarctic and arctic (Fromm and phang, 1972; h h andFranklin, 1989; Tian and Dai, 1988). First, trnnsverse cracks may becaused by the overall contraction of the entire pavement structureand/or underlying subgrade. This mechanism may cause the crack toextend through the entire pavement structure and into the subgrade.The crack can extend across the pavement surface into the shoulderand be several inches wide. This type of transverse crack is associatdprimarily with the thermal contraction of soil (in the base, subbase,and/or subgrade) rather than the asphalt concrete surface layer. In fact,they can occur in both paved and unpaved roads and airfields atintervals of 12 to 90 m and depths extending to 2 m (Esch andFranklin, 1989).Second, low temperature cracking is attributed to tensile stressesinduced in the asphalt concrete pavement as the temperature drops toan extremely low temperature. If the pavement is cooled to a lowtemperature, tensile stresses develop &g ,a result of the pavement'stendency to contract. Friction between the pavement and the base Layerresists the contraction. If the tensile stress induced in the pavementequals the strength of the asphalt concrete mixture at that temperature,a microcrack develops at the edge and surface of the pavement. Underrepeated temperature cycles or the occurrence of colder temperatures,the crack penetrates the full depth and across the asphalt concretelayer. Tian and Dai (1988) note that it may be possible for a thermalcrack to reflect up through the asphalt concrete layer from anunderlying stabilized layer if the coefficient of contraction of thestabilized layer is greater than that of the asphalt concrete layer.The primary pattern of low temperature cracking is transverse tothe direction of traffic and is fairly regularly spaced at intervals of 30to 60 m for new pavements to less than 5 m for older pavements. Ifthe transverse crack spacing is less than the width of the pavement,longitudinal cracking may occur, and a block paacrn can develop. Thecomplex pattern of cracking that may be observed inmany olderpavements is a result of (1) the increase in stiffneas (i.e., hardening)of the asphalt cement with age, and (2) the change in the geometry ofthe,pavement slab. On the taxiway at Fairbanks Airfield a 3 m "blockpattern" developed on the taxiway associated with longitudinal andtransverse thermal cracks (Esch and Franklin, 1988).Fromm and Phang (1972) noted that the transverse cracks causedby the overall contraction of the pavement structure and subgrade arenot as serious as cracks occurring wholly in the asphalt concretesurface layer. Cracks restricted to the asphalt concrete surface layerallow ingress of water which in turn increases the rata of stripping and'allows pumping of a fins granular base course. Water entering thecrack during the winter may result in the formation of an ice lensbelow the crack which produces upward lipping at the crack edge.Also, de-icing solutions may also enter the crack and cauw localizedthawing of the base which, in turn, may result in a depression aroundthe crack. Cedergren and Godfrey (1974) noted that 70% of surfacerunoff can enter a crack 1 mm wide.CURRENT PAVEMENT DESIGN PRACTICE FOR TRAFPIClLOAD RELATED CRACKINGCorps of Engineers ProcedureSince World War 11, the Corps of Engineers has developed pavement design procedures (toad and airfield) thaL CM be used to developstructural design requirements. The available design procedures forpavements subject to freezing and thawing in the underlying soils arebased on either of two basic concepts:Control of surface deformation resulting from frost heave (orthaw).Provision of adequate bearing capacity during the most criticalclimatic period.Based on the above considerations, three separate design approachescun be used:Complete protection method: Sufficient thicknesses of pavementand non-frost susceptible base course are provided to prevent frostpenetration into the subgrade.Limited subgrade frost penexration method: Sufficient thicknessesof pavement and non-frost susceptibb base course are provided tolimit subgrade frost penetration to amount8 that restrict surfacedeformation to within acceptable limits.Reduced subgrade strength method: The amount of frost heave isneglected and the design is bascd primarily on the anticipatedreduced subgrade strength during the thaw. +Hennion and Lobacz (1973) recommend that seasonal thawing andfreezing should be ratricted ta the pavement (surfacing and non-frostsusceptiblebssecourse) in'continuouspermafrost regions. Theconceptis comparable to the complete protection method wherein the criticalfactor is the dcpth of thaw rather than the depth of frost penetration.Alaska DOTPF ProcedureThe Alaska DOTPF (1982) haa issued pavement design guidelinesbased on research begun in 1976. The primary objective was to studythe various relationships that controlled the performance of flexiblepavements. Approximately 120 pavement sections were selected on theexisting state maintained road network for the rcscarch program. Thercsults of that study (McHattie et al, 1980) significantly influenced thedesign procedure.The Alaska WTPF design guidelines focus on the fact thatincreased fines (No. 200 minus material in the unbound layers of a- 1033

pavement structure) lead to increased thickness of asphalt concretesurfacing. In those areas of Alaska where asphalt concrete is readilyavailable, increased fines in the underlying layers may be acceptable.However, in remote arctic regions the production of asphalt concreteis expensive and should be minimized. Thus, it is prudent to specifynon-frost susceptible (NFS) materials in the underlying layers to theextent possible to minimize the following:The requirements for asphalt concrete (if not eliminate its need,altogether);The potential for frost heave;The potential for other types ofpavement distress such as alligatorcracking and rutting,Mechanistic DesiEn ApproachA mechanistic approach to pavement design involves (1) predictingstresses, strains, and deflections in a pavement structure owing to aspecified geometry and magnitude ofwheel loading conditions, and (2)adjusting ihe properties and thicknesses of the elements in thepavement structure to insure the predicted stresses, strains, anddeflections are within allowable limits. The mechanistic approachprovides flexibility in modeling apavement structure owing to thc factthat measured material properties &

T,in which,Q(T) = accumulated thermal stress for n particular coolingrate, T,cy = coefficient of thcrmal contraction,To, T, = initial and final temperature, respectively,S(t,T) = asphalt mix stiffness (modulus), time- and temperature-dependent,AT = tcmpcrature increment over which S(t,T) isapplicahle.The approximate solution suggested by equation (2) may yieldreasonable results providing that two input paramctcrs are correctlymeasured or assumed: (1) the coefficient of thermal concentration, and(2) the asphalt concrete mix stiffness. The tensile strength of theasphalt concrete mixture may be estimated or mcasured in thelaboratory in cither direct or indirect tension. In thc mechanisticapproach the fracture temperature is established hy equating the tensilestress calculatcd from equation (2) with the tcnsilc strength at thattemperature.The dctcrminatinn of both the asphalt concrete mix stiffness andthe tensile strength requircs that the rate of cooling in the hcld (andthe associated development of tcnsile stresses and strength) be relatedto a rate of loading or deformation in the laharatory (or In the case ofa creep test, a time after initial loading). To date, a procedure toaccomplish this ta6k has not been conclusively demonstrated to thepavement engineering community. Further, in the calculation ofthermal stress the thermal contraction coefficient is generally assumedto bc: 2 to 2.5 X lO+C. Recent measurements of the thermalcontraction of mixes with high voids contents or mixes employingmodified asphalt ccmcnt suggest this assumption could be in error hya factor of two or three. Further, age conditioning of the specimensfor the determination of the mix stiffness or tensile strength has notbeen considered in the application of this approach. Finally, it hnsbeen noted by several rcswchers that my approach that is fundamentallyrelated to a measurement of the stiffness of the mix will not beacceptable for mixtures which employ modified asphalt cements.Simulation MeasurementMonismith et al. (1965) were the first to suggest that the thermallyinduced stress, st,rength, and temperature at failure could be measuredin a laboratory test which simulated conditions to which a pavementslab was subjected in the field. The basic rcquirement for the testsystem is that it maintains the test specimen at constant length duringcooling. Arand (1987) made a substantial improvement to the testsystem by inserting a displacement "feedback" loop which irtsured thatthe stresses in the specimen would not relax because thc spccimenlength is continuously correct during the test.A recent version of this "4, sys m developed under the StratcgicHighway Research Program (SHRP) is presented by Vinson et al,(1993). The thermal stress restrain4 spccirnen test (TSRST) systemconsists of a load frame, screw jack, computer data acquisition andcontrol system, low temperature cabinet, and temperature controller.A heam or cylindrical specimen is mounted in the load frame whichis enclosd by the cooling cabinet. The chamber and specimen arecooled with vaporized liquid nitrogen. As the specimen contracts,LVDTs sense the movement and a signal is Sent to the computer whichin turn causa the screw jack to stre:ch the specimen back to itsoriginal length. This closed-loop process continues as the specimen iscooled and ultimately fails. Measurements of elapsed time, temperature,deformation and tensile load are recorded with the data acquiritionsystem.Under the SHW effort over 400 TSRSTs were performed tocharactcrize the thermal cracking resistance of asphdt-aggregatemixtures. Bas4 on the test results. the following conclusions areappropriate:TSRST results provide an excellent indication of low temperaturecracking resistance of asphalt concrete mixtures. A ranking of lowtcmpcrature cracking resistance based on TSRST fracture temperatureis in excellent agreement with B ranking basd on the physicalproperties of asphalt cements.Based on a substantial number of test results, asphalt type,aggregate type, and air voids contcnts are factors which have Imajor effect on the low temperature characteristics of asphaltconcrete mixtures. Softer asphalt cements and aggregates with arough surface texture and angular shape provide greater resistanceto low temperature cracking ofasphalt concrete mixtures. Fracturestrength was grcatcr for mixtures with low air voids content.The dqgrec of aging has a significant effect on low temperaturecracking resistance of mixtures. As the degree of aging of amixture increases, fracture temperature becomes warmer andfracture strength decreases. The degree of the influence of agingdepends on t,he asphalt type.Increasing the amount of ttsphult cement in the mixture does notimprove the resistance of the mixture to low temperature cracking.The TSRST can be used in routine mix evaluation for lowtemperature cracking resistance of asphalt concrete mixtures.. Vinson et al, (1993) have presented a framework to evaluate thelow temperature cracking resistance of road and airfield pavementsbmed on TSRST results.SUMMARY AND CONCLUSIONSProfessionals involved with the design of pavements in thewbmctic and arctic must bc concerned with three modes of distress:(1) distortion and pavement faulting, (2) disintegration, and (3)cracking. By far the most prevalent prohlem for asphalt concretepavcmcnts is cracking, followed by distortion related to differentialfrost heave. Methodologies prcsently exist to allow the resistance ofan asphalt concrete pavement to fatigue, subgrade rutting, wd lowtemperature cracking to be quantified and easily incorporated inpavement dcsign for the subarctic and arctic.REFERENCESAlaska DOTPF (1982) Guide for flexible pavement design andevaluation, Anchorage, AK. and Alaska DOTPF (1993) PreconstructionManual, Chapter 11-Design.Arand, W. (1987) Influence of bitumen hardness on the fatiguebehavior of asphalt pavements of different thickness due to bearingcapacity of subbase, traffic loading, and temperature. Proc., 6th<strong>International</strong> <strong>Conference</strong> on Structural Behavior of Asphalt Pavemenu, University of Michigan.Berg, R. and T. Johnson (1983) Revised procedure for pavementdesign under seasonal frost conditions. Special Report 83-27,USACRREL, Hanover, NH.' 103s

kg, R.L. and 13.W. Ajtken (1973) Some passive methods ofcontrolling geocryological conditions in roadway construction.Proc., North American Contribution, 2nd <strong>International</strong> <strong>Conference</strong>on Permafrost, Yatutsk, USSR.Berg, R.L. and D.C. Esch (1983) Effect of color and texture on thesurface temperature of asphalt concrete pavements. Proc., 4th<strong>International</strong> <strong>Conference</strong> on Permafrost, Fairbanks, AK.Ccdcrgren, H.R. an3 K.A. Godfrey (1974) Water: key cause ofpavement failure. Civil Engineering, ASCE.Crory, F.E. (1988) Airfields in arctic Alaska. Proc., 5th <strong>International</strong><strong>Conference</strong> on Permafrost, Vol. 3, Trondheim, Norway.Eaton, R.A. and R.L. Berg (1978) Temperature effects in compactingan asphalt concrete overlay. Roc., Cold Regions Specialty <strong>Conference</strong>,ASCE, Vol. 1, Anchorage, AK.Esch, D.C. (1973) Control of permafrost degradation beneath aroadway by subgrade insulation. hoc., North American Contribution,Zndfnternational <strong>Conference</strong>onPermafrost, Yakutsk, USSR.kh, D.C. (1986) Insulation performance beneath roads and airfieldsin Alaska. Roc., 4th <strong>International</strong> <strong>Conference</strong> on Cold RegionsEngineering, ASCE, Anchorage, AK.Esch, D.C. and D. Franklin (1989) Asphalt pavement crackcontrol atFairbanks <strong>International</strong> Airfield. hoc., 5th <strong>International</strong> <strong>Conference</strong>on Cold Regions Engineering, St. Paul, MN.Esch, D.C. and J.J. Rhode (1976) Kotzebue Airfield, runway insulationover permafrost. Proc., 2nd <strong>International</strong> Symposium on ColdRegions Engineering, ASCE.Fromm, H.J. and W.A. Fhang (1972) A study of tranwerm crackingof bituminoua pavements. Proc., AAFT, Vol. 41.Fulwider, C.W. and GIW. Aitken (1963) EffeEt of surface color onthaw penctration beneath a pavement in the arctic. Proc., 1st <strong>International</strong><strong>Conference</strong> on the Structural Design of Asphalt Pavemanta.Has, R., I;. Meyer, G. Assnf, and H. Lee (1987) A comprehensivestudy of cold climate airfield pavement cracking. Proc., AAW,Vol. 56.Hennion, F.B. and E.F. LobacZ (1973) Corps of Engineers technologyrelated to design of pavements in areas of permafrost. Proc..North American Contribution,lnd <strong>International</strong> <strong>Conference</strong> onPermafrost, Yakuuk, USSR.Bills, J.F. and D. Brien (1966) The fracture of bitumens and asphaltmixes by temperature induced stresses. Proc., AAPT. Vol. 35.Janoo, V.C. (1989) Use of low viscosity asphalts in cold regions.Proc., 5th <strong>International</strong> <strong>Conference</strong>on Cold Regions Enginwring,AXE, St. Paul, MN.Kestler, M. and R. Berg (19d9) Enginwring design and constructionin permafrost regions: a review. Roc., North American Contribution,2ndlnternational ConferencconPermafrost, Yakutsk, USSR.Mahoney, J.P. and T.S. Vinson (1983) A pxhanistic approach topavement design in cold regions. Roc., 4th <strong>International</strong> <strong>Conference</strong>on Permafrost, Fairbanks, AK.McPadden, T. and C. Siebe (1986) Stabilization of permafrostsubsidence in the airfikld runway at Bethel, Alaska. Roc., 4th<strong>International</strong> Specialty <strong>Conference</strong> on Cold Regions Engineering,ASCE, Anchorage, AK.McHatxie, R., B. Connor, and D. Esch (1980) Pavement structuralevaluationofAlaskan highways. RcportNo. FHWA-AK-RD-SCLI,Alaska WTPF, Juneau, AK.McLeod, N.W. (1987) Pen-vis number (PVN) M a meamre of pavingasphalt temperature susceptibility and its application to pavementdesign. Proc., Paving in Cold kens Mini-Workshop, Vol. 1,147-240.Monismith, C.L., G.A. Secor, and K.E. Secor (1965) Temperatureinduced stresses and deformations in asphalt concrete. Proc.,AAPT, Vol. 34, 248-285.Rice, E. (1975) Building in the North. Monograph, The OcophysicalInstitute, University of Alaska, Pairbanks, AK.Rooney , J.W., J.F. Nixon, C.H. Riddle, and E.G. Johnson (1988)Airfield runway deformation at Nome, Alaska. Proc., 5th Intemationel<strong>Conference</strong> on Permafrost, Trondheim, Norway.Rutherford, M. and J.P. Mahoney (1986) A thermal analysis ofpavement thawing. Proc., 4th <strong>International</strong> <strong>Conference</strong> on ColdRegions Engineering, Anchorage, AR.Terrel, R.L. and J.W. Shute (1989) Summary report on watersensitivity. Stmgic Highway Research Program, + NationalResearch Council, SHRP-NIR-89-1003, Washington, DC.Tian, D. and H. Dai (1988) Cold cracking of asphalt pavement onhighway. Proc., 5th <strong>International</strong> <strong>Conference</strong> on Permafrost,Trondheim, Norway.Tyler, O.R. (1938) Adhesion of bituminous films to aggregates.Furdue University Research Bulletin, No. 62, Val. 22.Vinson, T.S. (1989) Fundamentals of resilient modulus testing. Proc.,Workshop on Resilient Modulus Testing: State of the Rwtice,Oregon Stare University, Corvallis, OR.Vingon, T.S., D.H. Jung, and J.W. Rooney (1993) Pavementperformanceand low temperature cracking of subarctic and arcticairfields. Proc., 6th <strong>International</strong> <strong>Conference</strong> on Permafrost,Beijing, China.Vinson, T.S., J.P. Mahoney, and M. Kaminski (1984) Cementstabilization for road construction in Cold Regions. Proc., 3rd<strong>International</strong> Cold Region Engintering Specialty <strong>Conference</strong>,AXE, Edmonton, Alberta, Canada.Vinson, T.S. and J.W. Rooney (1991) A mechanistic approach topavement design for Nome Airfield. Proc., 6th <strong>International</strong> ColdRegions Engineering Specialty <strong>Conference</strong>, ASCE, Eebanon, NH.Vinwn, T.S., J.W. Rooney, H. Zhou, and N. CoetzsG (1991) TudorRoad rehabilitation, Anchorago, Alaska. Proc.. Road‘and AirportPavement Response Monitoring Sy*ms, ASCE, USACRREL.NH .Vinson, T.S., H. Zhou, R. Alexander. and R.O. Hicks (1989)Backcalculation of layer moduli and runway overlay design: U.S.Coast Guard Air Station, Kodink, Alaska. hoc., Stntc of the ArtSymposium on Pavement ResponseMonitoring Systems for Roadsand Airfields, USACRREL, NH.Vinson, T.S., I. Zomerman, R. Berg, and H. Tomita (1984) Surveyof airfield pavement distress in cold regions. Proc., 4th Interna-KiOd <strong>Conference</strong> on Cold Regions Enginering, Anchorage AK.Vita, C.L., J.W. Rooney , and T.S. Vinson (1988) Bethel Airfield,CTB pavement performance analysis. Proc., 5th <strong>International</strong><strong>Conference</strong> on Permafrost.Zarling, J.P., W.A. Braley, and D.C. Esch (1988) Thaw stabilizationof roadway embankments. Proc., 5th <strong>International</strong> <strong>Conference</strong> onPermafrost, Vol. 2.Zarling, J.P.,. B. Connir, and D.J. Ooering (1983) Air duct systemsfor roadway stabilizntion over permafrost areas. Proc., 4th<strong>International</strong> <strong>Conference</strong> on Permafrost, Fairbanks, AK.

CURRENT DEVELOPMENT ON PREVENTION OF CANAL FROM FROST DAMAGE IN PRCChen Xiao-baiLanzhou Institute of Glaciology and Geocryology,Chinese Academy of Sciences, Lanzhou, ChinaThe frost damage condition of canal was very heavy and has be-en paid much moreattention by scientists and engineers in China in last 20 years. The mainresults conducted in situ and in door are summarized including frost heavemechanism, classification, water migration and heave distribution along depth,frost heave prediction models and frost heaving force etc. Some anti-heavecountermeasures of canal used effectively and widely, such as flexible lining,anti-heave structures, and insulation structures as well, are introducedbriefly also.INTRODUCTIONThe annual water discharge in China is theNumber h in the world and the average for perperson is about one fourth of that in the world.The annual amount of water for aRriculture usingin China is near 88 percentage of total waterexpended in which the irrLgation amount is about94X of the azricultural one which is almost 82.7%of total amount of water expended in China.However, the uti,lization coefficient of theirrigation water is only 40 to 50X which meansabout 42% of the total amount water expended inChina was lost during transporting in c.ana1.Comparing with the utilization coefficient of 70to BOX in the developed countries, about 30 to50X of water discharge in China was osmosizedout off canal because of aoor lining. Up tonow, the length of lining canal is 10 to 50% ofthe total in different proviences, for instance,in Xin.jiang district there were 23000 km canalswith different types of lining which was 109, ofthe total in 1986: in Xanxi province the lengthof lining canal was 1009 km in 1988 which was45% of the total one: In Hunan province therewere 8000 km canals with lining which was 19.85Wof the tutal: In Sicuan and Fuje provinces thelength of lining canal was 39.37X and 22% of thetotal one respectively (Jj.an, 1992).The seasonally frozen ground is widely distributedto the north oE Changjiang River withthe half areas of the total in China. Thefrost depth increases from South to North andfrom the place with a low elevation to that witha high elevation and with the maxlmum one ofaround 3 m. The amount of frost heave i.s 10 to30 cm and some over 40 cm. In gerenal, thefrust heave will make the canal suffered frostdamage as the frost depth is over 50 cm, It issure that the frost damage to the canal 1s themain reason causing seepage. According to thestatistical data, about half lcngth of canalwith concrete lining, 2800 km, was suffered byfrost damage in Xinjiang; for the main irrigationnetwork in Xanxi province, the frost crackphenomena were very popular in canal withconcrete lining which occurred on the surfaceof. south slope and north slope of the canalwith 75.3% and 277 of total area respectively,The frost damage to the bridge, culvert aswell as sluice Rate projects was very heavy.Almost all of the timber bridges in Heilongjiangprovince were suffered frost heave wi-th themaximum frost displacement of 1 to 2 m: InTnner Mongolia district, about 359, of totalbridges with different types were in frostdamaRe, and 63% of total culverts were notworked well in which 25% of the total onessuffered very serious frost damaRe, as about71% of total sluices were stood frost damage aswell.A large amount seepage of water from canalcaused by frost heave made the land at the bothsides of canal, 100 to 200 m in wideth, salinealkalized.For intance, during the thirtyyears from 50’s to ~ O ’ S , the irrigation areaincreased to three times in Inner Msngolia whilethe area saline-alkalized increased to 10 timesmore than that of before at the first 10 yearswhich was about 73,8% of the total farmmingarea, and there were about 30X of ttltal farmminxarea saline-alkalized in Ninxia district.Since realizing that the frost damage was themain reason causing seepage of canal in thewhole regions of North and North-West China witha seasonally frost condition, the water conservancyinstitutes and colle~es in the regions aswell as Chinese Academy of Sciences paid moreattentions for understanding the mechanism ofsoi1,Erost heave and finding the more effectiveways to prevent the canals and hydraulicstructures from frost damage both in door andin situ. Some large observation stations *established in Heilongjiang, Gansu, Liaunin,Xanxi, Xiniiang, Inner Mongolia, Shandong,Shanxi and Ninxia Provinces (Autonomous regions)1037

in which different suhjecLs have been obscrvedsuch as the frost processcs under various typesof soi 1 and water conditions, and thc requlnritieson thc profiles of water redistribution,frost heave as well as frost heaving forcesalong the depth. In enginecring practico,different anti-heave countfrmeasurements wereconducted wit.h various structures widely includingreplacing r-layey soils with gravel, sand andblown sand with different contentof fine grainedsoils; increasing t.he dcnsity of spils withdynamic consolidation method; st-rengthcningdrainage;impruving soils with chemical agcnts;insulat-ing foundation bascs by ~near~a of effectiveinsulations; anti-heave strucl.ur'es such asarch and curvc structures, anchored structures,flexible ones as well as some special. structureswith large overlrurden pressurc etc. Based on alarge amount of rlnta collected from ohaervationstations as mentioned above for years and th,ewealthy experiences accumulated from pri>ct.ice, aTlesip,n N1>rr,1 for can:II was estahl ishcd by TheWater Conservancy Ministry of PBC as a profess.ivnalstandard (SL23-91), the other' one with :1 t.itleof Anti-hcave Design Norm for Hydrauiic. structureswi 1 I bc estah1"ishcd in not long future bythe Ministry. Besides above, accordir~g t.o thelocal conditions some special detailed rules andregulatious for dppl ication i n engineering practicehave been made already by Local government.Up 1.0 now, it started to the new stage of thatthe results of frost heave study and experiencehave been ayplicd to t.he practical projects incold regions in China.NlTWRESULTS OF APPLLED STUDYIn the period from 1978 to 1988, bccause thescientists and engineers have been engaging inunderst.anding the hehavors of frost susceptibilityof soils during freezing in door and in situ,and in the practice for accumulating the experiencesin order to deal with the frost damageproblems to the structures in cold regions, somewealthy results could he suned heriefly asfollows:1. Frost Susc.cptibi1it.y Classificotion of BaseSoilsAs well known. !.hat the indcx for evaluatingthe criteria of frost susceptibility of soilswas and is thc ratio of frost heave or the rateof frost heave widely. As the variatior, offrost depth in China from t.he north to theChangjiang River to the far north is 40 cm t3near 300 cm, it is difficult to predict thedegree of frost. dama~c to the structure byusing the indcxes. After analysing large amoun~of statistical data collected from differentreEions in Chi,na, the rep,ressive results showthat the denrec of frost damage to the structuresdepends 011 the total amount nf frost heave.Consequently, the total amount of frost heavehas been accepted as an index For evala;3tjng thedegree of frost damage of base soils with whichthe base soils wcrc devided into five classeswith the name of Non FH, Weak FH, FH, lleavy FHand Very heavy FH as well which is rcsponsihleto the total amount of Crust heave of h

~~~ thatby Jiling Province1 Water Conservancy 1nst.itute(SL23-91) In which the annual variation coefficientof frost depth Cv will decrease with theincrease of frost depth. The standard frostdcpth could be expressed by freezing index Iowith a function of FIo=cc. To/' whilc.a=2.84 exp(1.921.10-" .To). )luring the calculation offrost depth in practice, some coefficients offrequency model ratio Kf and slope correctionKS as well as groundwater influence Kz areconsidered as follows:Hd = Ho.Kf.Ks.Kzwhere, Hd is Ilesign Frost depth and Ho isstandard one.The ohservational works 011 an effect ofsunshine on the frost depth have been conductedalready in Gansu, Xanxi, Ninxia and Shandongdistricts as well. The results have beenapplied in the Norm yet in which Ks equals 0.65to 1.0 for the slope facing sun, 1.0 to 1.2 forthe bottom of canal and 1.2 to 1.5 for the slopewithout sunshine. The Rroundwater influencecoefficient Kz depends on not only GWT, but alsosoil types and i.s listed in Table 2 (SLZ3-91).Table 2. The Value of Groundwater InfluenceCocfficient K,Table 3. Replacing Ratio of Anti-heave Cushionof CanalG.W.L.'(em)Soil typeHw>Hdt250 Clayey soil withhigh 11 R siltysoilHw>HdtlSOClayey soil withmiddle 11 &silty soilHeplacing ratiO""E(x)Up. p.s.Dow. p.s. & Rot.50-70 70-80lt>Hdt150 Clayey soil withlow 11 & silly bO -- 50soilClayey soil withhigh or middle I, 60-80 80-100R, silty sui1mentioned -aboveClayey soil withlow T1 fi silty 50-60 60-80sol I.H,=>2.0 1.0 -0.95 1.0 1 .o2.O>Hw=>1.5 0.95-0.90 1-0 -0.95 1.01.5>Hw=>3.0 0.90-0.84 0.95-0.90 1.0 -0.97l.OO~ 5 0.84-0.75 0.90-0.80 0.97-0.940.5>Hw=>0 0.75-0.62 0.80-0.70 0.94-0.85Each signal's mean is shown in Table 4.Roccntly years, a theorical frost depth indifferent conditions with vrrrjuus angles betweensunshine ray and ground surface was calculatedby means of encrge balance method (Gou et a1,1993; Li, 1Y93).4. Sufficient Conditions of an Anti-heaveCountermeasuti by Replacing Cxycy Soil withSandy or Gravel (Chcn et al, 197Y)A t first, limitina - the content of finegrained soil as mentioned above: Secondly,designing a sufficient replacing thickness: andthirdly, having a nccessary drainage cundition.A l l of the three is the sufficient condition^of rcplac.ing clayey soil with sand or gravel foranti-heave.Sased on the observational data collected insitu, a suitable replacing ratio has beensupposed as listed in Table 3 (Zhu, Q.et al,1989b).Resides coarse sand and gravel used forcushion meterials, blown sand was and is widelyutilized .in Gansu and Inner Mongolia districts(Chen et al, 1986; Cheng et al, 1991, Zhu,D. etal, 1986a; Zhau et al, 1988) while a serviceinstruction has been provided by the experimentalresults in door conducted by the authors(Chcn et al, 1986). A detailed service instructionon application of blown sand to canalengineering projects has been established byInner Mongolia Water Conservancy Institute(IMWCI, 1987).5. An Effect of Groundwater Level on Frost Heave+G.W.L.-Groundwater level, Hw: Hd-Design frostdepth; E*+ -1'erccntage of replac.ing depth tofrost dcpth; 1Jp.p.s.-Upper part of slope:Dow.p.s. & Rut.-Down part of slope & Rottom ofcanal; TI-Liquid limit of water content.A largc amount of results collectcd fromobscrvat,ioual stations in situ and engineeringpractices show that there is a critical groundwaterlevel Hwo. While Hw=>Hwo, the frost heavew i l l be indeper~dent on groundwater level. Afteranalysing the statistical data, the value ofcritical groundwater level IIwo was listed inTable 4 (ST,23-91).Table 4. The Value of Critical Groundwater LevelSoi.1 typeHwo(m)C.-Clayeof water-C. wjth 11 C. with M C. with L SandyI1 & s. I1 & s. TL & S. soil2.0 1.5 1.0 0.5Y soil; S.-Silty soil: 11-Liquid limitcontent; 11-High: M-Middlq; L-Low.Rased on the results collected in situ fromXinjiang Liaunj.n. Gansu, Xanxi, Neilonuiiang" -et.c, a regression functj.on for expressing therelation bctwren frost heave ratio and groundwaterlevel Hw is as follows:q = a RXP (-b Hw)The function has heen conducted in dooralready (Chcn et al, 1988~). In order to use therelation conveniently in engineering practices,an apparent frost heave ratio f was recommendedwhich is defined as follows:f=h/(Hfth)xlOa%P~/(100tq)=~~~XP(-~lHw)1039 '

After regres:jion of observational results insitu, the value of the constants of a1 and 81is listed in Table 5 (SL23-01).Table 5. The Value of 01 and 61Soil type '-1 a1 81-C. with €1 I, 8 S.O=

~~For claycy soils with water supply at given The listed value is only avai. table for the slebgroundwptfr level dur ir~g frcezing, a new modclfuundatimn wirh a short size of mure than 2 rn.wab conducted considering i~~itial water contentW(X), initial dry unil. weighl: Yd(t/m'), frost Table 12. The Value of Horizontal Frost Ileaving 'pcnctration ru~e Vf(cm/day), groundwater levelForceHw(m), plasLicity index lp, and ion content, S(rrrmol/l.00g soil) as well wt1ic.h could be expressedby :Classificoti~rrl ';: Weak FH FA Vcry PI1 VeryE'HAfter curve fitting with 76 groups uf expcrimentaldal.a, it cou1.d be described as follows:T,=bs41 l.5Xl~"5~3.60R2yi345Y vf1.1825EXP(-1 .4502Hw)EXl'(0~14781p)EXP(-0.1225S~While Frost heave rate R could be expressed as:H=4.h293X10-6W3.5177 4.2673y,i EXlJ(-l.6059Hw)F,XP(o.1637Tp)~XP(-n.124~~)7, k'rost Aeavino ForceAs well knownthat thc deterrninati.on of thevalue of frosC heaving force is very big workwhich should be observed in situ with largescale and taken many years. Rased on the datacollected mainly from Heilongjiang ProvincialWater Conservancy Institute, Low TemperatureInstitutc and Transportation 1nstitut.e. apractic.al value for frost heaving force has beenprovided. Based on the affecting directlor1 ofthe force, it. could be divided thrce ki-nds offrost heaving force: tangential one T affectirlgaldng the lateral surface of foundation; normalune 0 affecting f.he bottom of foundation vertically;horizontal one "h affecting the surfaceol retaining wall horizontally. Tablc 3.0, 11and 12 are 1ist.ed the values already which canbe applied in engineering pratice (SL23-91).Tablc 30. The Value of 'Tangential Prost He~vingF (I 1- c eNun Weak FH Very FH Very heavyClassificationI.'! t.'H FH"For vcry smooth surfar-e of precast concrete pile,the value should time a cocfficient of 0.8; andfor the r.oarsc surlace of stone pi.^, the valueshould tine :.t coefficient of 1.2.'Table 1.1.The Value cf Normal Frost Heaving ForceStmulard va I ueI00Classification Standard value U (KPa)Non FH X-100 30-60 20-50 10-30l;:eak P9 100-150 63-100 50-80 30-60VI1 150-7.10 100-150 83-130 60-100Very FI4 210-290 150-220 130-1C10 103-150Very heavy' FH 290-390 220-300 190-260 1.50-2101 1 16.3 23.6 42.7 77 .o2 1 11.0 7.5.9 41.3 59.3L.'Y.-I.,inin~ type; W.D.-Wat.er dr:pth;Conc.-Cc.)ncretc;I

scale and practical experience in situ, a tesonablestructure design was proposed by the uuthurs(Chen et 111, 1987) with wh-ich the slope uf soilprotecticn layer' upon plastic film will bestable." 2. Allr i--heave Structures1) II shaped canal--Sincc 1975 the first Ushayt~rl canal vas set up in Xanxi province (Zhu,Q- et al, 1991) with an optium hydraulic section;Ind saving land 1/2 to 3/4 compared with thetradional ladder shaped canal, and increasinganti-heave capacity as well. In 2983, the I1shaped canal built hy concrete was set up inXanxi province for 7600 km and up to now it wasfinished about 20000 km i n China with a biggestd.ia;charge of 25.8 m3/s in irrigation canal. NUWthe machines for constructing U shaped canalwith cast in place or precast concrete with therhameters of 40 cm, 60 cm, 80 cm, 100 cm and120 cm or more havc been used widely in practice2) Arch shaped botlom uf canal--As well knownthat, thc lining c.orner of canal bottom is easybroken (luring freering bf:cause of concentraledfrost heaving stress. It is shown in pract-icethat the amount of frost heave at the archshaped bottom is about 1/2 to 1/4 of that at thecorner of bottom with a ladder shaped. Consequently,the arch shaped bottom of canal withconcrete lining is set up rat.hcr popu1arl.y inseasonally frost regions in China. 'The structurewith an arch shaped bottom is also used inthe tunnel for' disappering water head verysuc.cessfu1 in nortll-cast China with strong frostsusceptihility hnse soil (\:RII~,, S.Y., et nl,1993) which is so called anti-heave structurewith an arch shaped.3 ) 11-11 shaped sluice--In norm situation, asluice w i l l be designed as in optium hydrauliccondition with much complex structure. Howcver,this complex structure would not work well incold region with heavy frost heave because theoverburden pressure is so small that could notlimit the heave in an allowable one. The "-"shaped sluice with a concentrated load and aburied depth over maximum frost depth vas setup firstly in Xijiang district and later inHeilongjiang province which is uscd widely innorth China.4) Concrete box and Reofahric sandwedgeretaining well--Some retaining walls were setup hy means of concrete box with a space insidewhich were successful in Heilongjiang provincebecausc they were permlted to have nn allowabledisplacemenl during freezing. The sandwedgestructures were built up irl Inner Mongolia inwhich local soil was covered by Reofabrics andbuj1.t up one upon another. Since they are veryflexibale for dealing with frost heave, so itw i l l be a new anti-heave structure for buildingrrtai.nj.ng wall in future.3. Insulation StructuresThc insulation materials were used beneathconcr'ctc lining or foundation, or closed theinside surface of retaining wall for preventingfrost penetration into base soil, as a result,the frost heave of insulated base soil will bedisappeared ur limited. A pulystyrene form slahwas firstly used beneath concrete lining ofcanal in Shanxi province hy Academy of WaterConservancy, and Shanxi Water ConservancyInstitute. It was successful and had a goodcomprehens.ive benefit (Jin et al, 1987). Lateron, the polystyrene form slab was'userl bencat,hthe concrete lining c3f canal :rlopc at the plt+cewith water level fluctuationsuffered heavy frost damagewhich used LO bein a water transportcanal in Shandong,provincc. The thickness ofpolystyrene form slah is abuut 1/10 ro 1./15 ofthe design frost depth Ild which depends on thematerial properties (SL23-'31). Besides above,insulation slab is also madc of pearlite andm-i.neral wool (Jian, 1992). The total capitalexpenditure of. last two will. reduce about. 1/3compared with that of insulaLed by polystyreneform.All of that mentioncd above is an outline ofthe current development. on thc suhject forpreventing canal frorr! frost damage in China.Howevcr, there are lots of pr'oblerns facing tohe sulved incl.uding tcchnique, capital,adiminislration and mat.erii3ls etc. as well.REFERENCESChen, X.E. ct al. (1979) On anti-heave measureoyrcplacing clayey soil with gravel. Bulletinof Sciences. No. 20, (33.5-939.Chcn, X.B. et al. (1986) An experimental studyon frost susceptibility of wind-blown sands.J. Glaciology & Geucryology, No.3, 233-238.Chen, X.H. et al. (3.987) SaturaLe(l-dchydratedconsol.idation of f i I J with low densi1.y undits application to canal engineering, SCIENTTASINICA, No.7: A, 779-784.Chen, X.R. et al. (1988a) Frost susceptibilityof sandy gravel during freezing, YAN'PUGONCCHENG XIJRRAO, No.3, 23-29.Chen, X.R. et al. (1.988b) A frost heave modelof sandy gravel in open system, Proc. VTCOP, 304-307.Chen, X.B. & Wang, Y.4. (1988~) Frost heaveprediction model of clayey soils, Cold RegionsSciences and Technology, 15, 233-238.Chen, X.E. & Wang Y.Q. (1991) A new model offrost heave prediction model for clayey soils,SCIENCE I N CHTNA (n), No.iO, 1225-1236.Cheng, M.J. et al. (1991) Some anti-heavecountermeasures of canal lining in Tnnel-Mongolia, Technique of seepage pevention,N0.2, 107-110.Dcsign Norm OF Anti-heave for Canal Engineering(SL23-91)Gansu Water Conservancy Inst., (1986) A reviewon the results of prevent-ion of canal. fromseepage and frost damage in Gansu province,Tec-hnique of seepage prevention of canal,Special lssue (4), 2-5.Guo, D.X. et al. (.1993) On freezing and frostheaving rule of base earth of lined canal ofArbiLry slope direcliun and gradient, Proc.ICOP *Tnner Mongolia,Water Conservancy Tnst. (1087)An experimentaL report on prevention of canalfrom Erost damage by using blown sand.Jian, G. (1992) An outline uf current deve.I.opment,problems and suggestions on antiseepagetechnique of canal i n China,Technique of secpage prevention, No.3, 1-8.Ji.n, Y.T. et a1. (1987) An experimental studyon new material and new structure forpreventing canal lining Srom frost damage.ti, A.G. et al. (1993) Delermination of sunshineand sunshade extent coefficient oncanal, Proc. TCOP.Ljaunin Water Conservancy Inst. (1986) Ondesign for prcvcntion of ant i-seepagestruc.tures from frost damage of canal,Technique of seepage prevcntion of canal,Special Tsvue (4), 6-10.1042

Wang, S.R. (1993) A sttdy on basic types ofvertical tunnel for divappering water headand their capacity for anti-heave, Proc. ICOP.Wang, X.Y. (1987) A study of the relation betweenfrost heave ratio (and apparent one) andgroundwater level.Zhau, B.X. et al. (1986) An experimental resultson the prevention of canal from frost heaveby using blown sand in open system inNincheng county, Technjque of seepageprevention of canal, Speci.al Issue (4),82-89.Zhu, D.F. et al. (1986a) A discussion on antiheaveeffect of wind-blown sands in irrigationcanal, J. Glaciology & Geocryology,No.3, 239-244.Zhu, D.F. et al, (1986b) An experiment of theeffect of soil and water conditions on fTostheave of canal with concrete lining,Technique of seepage prevention of canal,Special Issue (4), 123-134.Zhu, Q. et al. (1988a) On the distribution offrost heave along depth in seasonally frostregions, J. Glaciology & Geocryology, No.1,1-7.Zhu, Q. et al. (1988b) Subsoil replacement usingsand-gravel for preventing frost heave damageof canal lining, Y. Glaciology & Geocryology,No.4, 400-408,Zhu, 9. et al. (1991) A review on the develoymentof anti-seepage technique of canal withconcrete lining in China, Technique ofseepage prevention, No.1, 1-7.- 1043

A MODEL FOR THE INITIATION OF PATTERNED GROUNDOWING TO DIFFERENTIAL SECONDARY FROST HEAVEG. C. Lewis+, W, B. Kmntzt, and N. Cainet1 INTRODUCTION1044 -

1046

where vi is the tlitnensionless ice velocity and XG is ;tnempirical constmt. The frost-heave velocity defined byequation 7 is ;Lssuwxl to be equal to the ice vklocityevaIu:tteci at the z;+; that is,The parameter B defined by e$:ttion 10 isproportional to the IiLtent heat content of thepermeating water divided by the conductive heattransfer characteristics of the soil; it is ch:tr:tcteristic .ofthe soil type: for clay B > 1. Eqmlion 6 is the Stefan condition for thismoving boundilry problcm. Tho dorivzttion ofequations 6 and 7 erntmdies much of the complexphysics occurring in the frozen fringe. These equationswere obt:tined hy conhining tqu:Ltions for the unfrozenwater and ice mitss hdances, energy h ~l~n~e, Drtrcyperme:ttion flow, Clitpeyrotl relation for the freezingtcmpcratlire, and over:tlI force bnlmce at the hottom ofthe lowest ice lens using the fonndism for rrdl1cing thefrozen fringc to il m;lthc1n;lti(:;tl plttne ils t1csc:ribotl byFowler (1989). The C1;yjeyron etp1:ltion is motlifictl toinclude cryostatic suction effects : t~l is given bywave length of the patterned DSFH. A more completedevelopment of this analysis is given by Lewis (1993).The principal results of this linear stability analysisare sumtn:trized in Figure 2 for three characteristic soiltypes: clay (curve a); silt (curve b); and sand (curve c).This figure was prepared assuming that the loading isdue only to the weight of the overlying frozen soil andT; = 0 "C. The modulus of elasticity of the frozen soilwas assumed to be 5.0 x IO7 3 (Drewry, 1986).hlI I I T I E l0 0.1 0.2 0.3 0.4 0.5 0.6rFigure 2: Dinlensionless wave number a and growth c*ellicient as functions of dimensionless frost-penetrationdepth e -B for: (a) clay; (b) silt; (c) sand.[F] V4nt = Pwhere w is the z-tieflectiotl of tho conter of tho plnts, Eis Young's rnodult1s, antl h is the thickness of the frozenregion.In order to ~t;JAish the conditions for wllicll ilp:ttterned form of DSFH is possible, it is neceswry tosolve the above system of equations. We explored thisinstability mech;tnism by cwrying out a linear st:hilityanalysis to tleterrnine the oonditions recluircrl for tho 'initiation of DSFH in the form of grountl-surf:lcecorrugations hrtving a wwe length X. This malysis candetermine the effect of soil type, frtwing depth, climilticconditions, ctc., on the initiation and charltcteristicThe lcft ordinate in Figure 2 is the dimensionlessfrtuzing front clepth which is plotted versus thedimensicinis wave number a. This figure indicatesthHt for C ~ L1oc:rtion C ~ of the freezling front there is apreferred unstable wave number, independent of the -soiltype, which satisfies the linear stability analysis for thecoupled heat transfer antl permeation described byequations 2-1ci and the elasticity constraint.given byequation17.The right ordinate in Figure 2 is the dimensionlessgrowth coefficient r which also is plotted as a functionof a. Figure 2 implies that the preferred wave numberand its growth coefficient decrease with increasingpcnetriltion of the freezing frdnt. These predictionsinclicate that the onedimensional SFH process for thespecified conditions is unstable such that it will evolveinto DSFH characterized by a corrugated heavedground surface under1:tin by a corrugatedfrost-penetration front.This linear stability analysis predicts that DSFH isthcorctic:llly possible in any soil characterized by acryostrttic,suction function for which I b I< 1. However,1047

7 REFERENCES

Initiation of Segregation Freezing Observed in Porous Soft Rockduring Melting ProcessSatoshi Akagawa,Institute of Technology. Shimizu Corporation,Tokyo 135, JapanIce lens appearance in freezing soil and porous soft rock varies greally. This difference is assumed 10 be governedby the cohesion or tensile strength of the porous materials, such as soils and soft rocks.In this paper, the author demonstrates the empirical results which show ice lens initiation in a melting frozen poroussoft rock where no pore water flow and negative to zero net heat tlow at the segregating ice lens is maintained.The mechanism of the ice lens initiation and growth in melting porous soft rock is discussed in the light ofthermodynamics and mechanical proprties of the rock.The discussion supports the assumption with the data demonstrated in this paper : the tensile strength of freezingporous mawrial may play the role in determining of suction pressure at growing ice surface, and may control ice lensinitiation and continuous growing properties.Table-1 Physical and mechanical properties Of tuffIce lens appearance in freezing soil and porous soft rock variesgreatly. Comparing the properties in frost heave tests, conducted withconstant boundary temperature conditions and an open pore water linesystem, it may be said that (1) many fine ice lenses are seen in frozennormally-consolidated soils, (2) a reduced number of thin ice lenses areseen in frozen over-consolidated soils, (3) a more reduced number of thickice lenses are Seen in Tertiary soils which have slight measurable tensilestrength, and (4) a few dominant thick ice lenses are seen in frozenporous soft rock.This difference seems to be related to the magnitude of theinteraction between soil grains which is measured as the cohesion and/ortensile smngth of each materral. Low cohesion may cause frequent min icelenses and high tensile strength may cause restricted thick ice lenses.Experimental results mentioned above reveal an assumption thatcohesion and/or tensile strength of freezing porous materials affect theproperties of ice lensing. In other words, tensile strength may play thesame role as super-cooling does at ice lens initiation in freezing soil.In order to examine the assumption mentioned above, unusual frostheave tests were conducted. The procedure of the frost heave tests wasspecially arranged to eliminate the super-cooling of pore water which mayassist the initiation of ice lenses. For this purpose, frost heave tests in thethawing condition were ntroduced . Theoretically and experimentally itwas understood and confirmed that ice lenses may not initiate or grow inthawing conditions (ex Gilpin, 1982). A certain amount of net heat flow atthe rowing ice surface is required to meet the latent heat budget of icenucfreation (Akagawa, 1990). In macroscopic view, the required net heatflow will never be supplied in melting process.In this paper, details of the frost heave tests conducted under thethawing conditions aTe explained with test results, and the process of icelens initiation in the thawing condition is discussed.IMENT3almhEdWelded tuff with the properties listed in Table-I WBS used as thespecimen. It has already been confirmed that the tuff frost heaves againstits tensile strength of 1.4 MPa in the freezing process (Akagawa et al.,1988), and that its properties which affect frost susceptibility, such as porestructure and/or unfrozen water properties, are similar to those of frostsusceptible soils (Akagawk 1991).Apparent specific gravityPorosity(%)Strength(MN/&CompressiveTensileTangent modulus (MN/m2)Secant modulus (MN/d )Elastic wave velocity(km/s)Poisson's ratioUnfrozenDry Saturated(20°C)Frozen.Saturated(-5C)(-15"c)(-25"c)1.40 1.72 ______ .__I___37.914.7 6.6 12.0 18.5 26.21.6 1.4 1.4 4.4 5.74812 3548 2859 2068 37634596 3126 3097 3410 62812.45 2.72 3.20 3.67 3.890.17 0.28 0.46 0.38 0.42AooaratusA triaxial frost heave test cell, shown in Fig.1, which was placedin 2 "c chamber waq used for this experiment. With this apparatuslateral heat tlow and side friction were reduced as much as possible. Theboundary temperatures of the specimen were maintained tiy circulatingtemperature controlled liquid in pedestals which are seen at both side of thespecimen in the figure. The temperature profiles were acquired withtemperature readings from platinum temperature transducers placed everylcm over the specimen. A pore water line was open from the waterreservoir to the specimen and the amount of the flow was measuredautomatically. The expansion of the specimen was measured with a digitaldeformation transducer which generated one pulse for every 1 urndeformation.€!m&bESample preparationThe trimmed specimen, which had the dimensions of 60mm indiameter and abut 90mm in height, wa? submerged in water under avacuum condition for more than 3 months. After the installation of thespecimen in the uiaxiat frost heave cell, following almost the sameprocedure as conventional triaxial compression test, pore water wascirculated from the bottom pedestal to the upper pedestal for 24 hours toavoid unfavorable air bubbles. Uurin this period, both pedestals weremaintained at a same temperature, +2E, to equalized the initial specimentemperature.

,3rd.thgwCooling liquid --.circulationThin rubkrmembrane...........shown in Fig.2-A-l to Fig.2-C-l . Tenlperalure changes in the freezingspecimens are also shown in Fig.2-A-2 to Fig.2-C-2 . Temperatureprofiles at some important priuds are shown in Fig.2-A-3 to Fig.2-C-3.These temperature profiles wcw drawn with temperature data at "a", "b","c", "d", and "e' in Fig.2-A,B1:-2, and correspond to the following;initial temperature profile, stcady urnperatwe profile at first freezingphase, steady temperature profile at second freezing phase, ice lensinitiation in first thawing phase, a:ld continuous ice lens growth inthawing phase.During the first freezing phase, "a" to "b', specimens seemd toexpand, however this was not frost heaving but expansion due. to thefreezing of bulk water and is usually called in-situ freezmg. In this phase,pore water was expelled from the specimen which is a typical phenomenaWarmingliquidcirculationPore water supplyPressure gaugefor unfrozenwater pressure atsegregating icelensFig.- I Triaxial frost heave cellTest procedureAs for the first stcp of this experiment, the specimen was frozenf~-om top to bottom uniaxial1 . controlling the tetnpcratures of the upperand lower pedestals at -IS 8 and 0c respectively. After theconfirmation of negative remperatures over the entlre specimcn, furtherlowering of the specimen temperature was conducted by ceasing thetemperature control at bottom pedestal. The frozen specimen was cooleddown further by heat extraction from the upper pedestal. Afer it had beenhe confirmed that the specimm had not expanded, the tcmperaturt: of thebottom pedestal was increased to the scheduled positive temperature inorder to thaw the specimen from the bottom upward. Since this was thcvery hcginning of the experiment, all the tcst conditions were maintainedthroughout the experiment.In order to achieve the above mentioned te6t procedure, the testconditions, shown in Table - 2, were applied.Photo 1 Freezing specimenTable-2 Tcst conditionsTest name A - - ~ __ ~-Freeze up .I sf thaw hack 2nd thaw b& I b&......................................................(r I (kgf/cmZ) O.(l95 0.095 0.005 0.095*r 3 (kgf/cm') 0 0 0 0Pw (kgf/cm' ) 0 0 0 0Tc ("c) -15 -15 -12 -12Tw ('c) Oor-2 4.5 +3.5 +5.55"- ,."" _._"" __ ~'Teat name B............................. Fl-ee1-t: up~~'Test name C""hack thaw ........................................Frcc/n: UJ thaw backc I(kgf/cd) 0.095 0.095 n I(kgf/um2) 5.62 5.62n 3(kgf/cd ) 0 0 n 3(kgf/crn2) 5.43 3.43Pw(k Yficrr? ) 0 0 Pw(kgf/crn2) .S.O() 5.1) 8 closcdTc (k) -15 -15 Tc (c) -15 -14.4Tw ("(3) 0 & nocont. +5.3 Tw ('C) Odi. nocont. +5.1

-1Time (tu)Fig 2-A-I0 20 40 60 80 100-Time (tu)Fig 2-B-10 20 40 60 80 100 120140Time (hr)Fig 2-C-1-020v E 40b 6o800 50 '100 150 200 250Time (hr)Fig 2-A-20204060800 20 40 60 80Time (hr)Fig 24-2I..i-1020I B 6o0 20 40 60 80 100 120Time (hr)Fig 242-2100."-15 -10 -5 0 5-20 -15 -10 -5 0 5 -20 -15 -10 -5 0Temperature ('c) Temperature ( c ) Temperature ("c )5 10Fig 2-A-3 Fig 2-B-3 Fig 2-C-3Fig 2-A Test results of Test-A Fig 2-8 Test results of Test-B Fig 2-C Test results of Test-CI8oof in-situ freezing. At a time "b, the specimen temperature went helowzero entirely and then further freezing was accomplished by releasingconuol of the wming pedestal temperature to avoid super-cooling. Thisprocedute led to the warmest temperature at "c" in Fig.2-A, B, C-2,where the specimen is lower than -4 c, as shown in Fig. 2-A, 8, C;3.During this period, the expansion was ceased abruptly. Immediately afterthe period "c", the temperature of the warming pedestal was increased to 4'c in Test A and 5'c in Tests B and C . A couple of hours later, as thelower part of the specimen became positive in temperature, gradualexpansion was started at about the period "d' in Fig.2-A, B, C-2. As timeelapsed, the heave fate increased to a rate of about 0.06mm/hr with porewater in-take. These combinations, such as expansion of specimen andpore water in-flow reveals that the frost heaving with ice lens segregationwas taking place. Actually, ice lens growth was confirmed through asemi-transparent rubber membrane as shown in Photo-1.In Test A, funher lhawing was conducted twice-at the periods "e"and "f'. During hoth thawing phaes, reduction of heave amount was seenfirst and then recovery of heave amount was confirmed as lime elapsed.However, the recovery was not very clear in the period "e" to "P',because of the insufficient time length.After the frost heave tests, specimens wcrc cut in two peices. Atypical photo of the ice lenses seen in the 'Test C specimen is shown inPhoto. 2. Due to the high rigidity of the specimen, the distance betweenthe bottom end of the specimen and warmest side of the ice lens did notchange over the Est durations. Therfore the location of ice lens initiationwas measured as this thickness. With these tnicknesses and thetemperature profiles at "d","e", "E". and "h", the segregationtemperatures, Ts, of ice lenses were observed as shown in Fig.3. Theseobserved Ts values scattered around the formerly observed Ts value, -1.4VI which was observed in normally conducted frost heave tests.

dPi= n t+Po : seoaration omsurewl I , condition at theDISCUSSXQjFrost heave tests condwtcd in this work clearly show a possibilityof ice lens initiation and growth during thc thawing process, although thisis only seen with a very slow thaw back process. The tcst results promptquestions. such as (1)what mechanism may enable ice lens initiation and(2)how is the heat budget for ice lensing maintained.In the following, a possible understanding of the above questionsis discussed.Since the observed Ts values in this experiment are somcwhalsimilar to what was formerly measured, icc lensing propenies observed inthe experiment arc also similar IO usual ice lensing properties (Akagawa,1988). The observcd segregation process of the ice lcnses in thiscxperment is understood to be the usual process. Then thennodynamiccondition in frozen fringe at the initiation of ice lensing, the pcriod "d" and"g" in Fig.Z.A,B,C-2, is drawn as a triangular section pointed to by anarrow ,"A", III Fig.4. A quasi-equihbrium state was reached because therewas no pore water in-flow and availahle pores in the fringc were filledwith ice and unfrozen water. The right most, or highest point of trianglesection shows an ice pressure that IS almost scparating the pore despite itstensile strength, n 1. According 10 thc generalized Clausius-Clapcyronequation, ice pressure devcloped (dPi) due to the depression of the localequilibrium tcmperature (dT) is exprcsscd as;where dPw is the suction pressure In unfrozen water, 1, is volumetriclatent heat of fusion, To is the freezing temperature of water in Kelvin atatmospheric pressure, y is the specific gravity and subscripts w and irefer to water and ice. In this case Ihe temperaturr should be around -1.4'c. Therefore. dPi should he about I .SSMp;i. lmmcdiately after thepore is enlarged by the separation force exerted hy the ice pressure, theice pressure may drop to the overburdrn prcssure of Y.3kPa. Since thispressure is about 6/lWO of initial dPi, the thermodynamic condition maychange to the other sidc of the 3I) drawing, sec exrow "C". The triangleshaped section shown by the other arrow reprcsents a relationshipgoverned by generalized Clausius-Clapeyron a1 dPi=O as;dPidPw = y i ( ---__ +y iLdT)ToThe thennodynamic condition at the time ~rnrnediatcly after ice lensinitiation in frozen fringc is shown as a deformed triangle, labled "R" inFig.-4, bwause tht thermal ficld will not change spontancously. Theciefobnned trianglc shows the reduction in ice prcssure of ( ~ and t thegeneration oftLPw, i.r. suction force. The relation brtween n t and dPwin this case is shown as ;a condition at ice lens segregation withno overburden oressuret -".7"i i dPw =S: suction pressu;1, 8 (dPi=O)dPw=yw(9+L,, y I TOFig.4 Ice lens initiation scheme with tensile strengthThe suction force induced by the tensile strength may he a driving forcefor flowing pore water from the water reservoir to the segregating ice lensthrough frozen fringe. As pore water migrates to the segregating ice. thelatent heat of ice nucleation may raise the local temperature. If the rise inlocal temperature is less than the temperature change equivalznl tu a thepressure change, Pi, from I.SSMPa to 9.3 kPa, i.e. in this case 1 .SS/1.11- 0.0093/1.1 I = 1.39 (re, the suction may continue to function pullingpore water to the growing ice lens. !4owever, although the value of thesuction may diminish a cenain mount.In this manner tensile strength may play a role in determiningofsuction pressure and therefore control ice lens initiation and the continuousgrowing that is observed in the thawing condition.In addition, F4(3) clearly indicates that if a freezing material has ahigher tensile strength, a higher suction force may be expected when thematerial frost heaves. Therefore it may be said that strong and stable icelensing occurs more often in freezing porous soft rock compared to usualsoil. This tendency may also be expected lo the ice lensing properties ofTertiary soils and ovcrconsolidawd soils.CLUSIONS(1) Porous material, i.e. welded tuff, frost heaves while it is thawingunder the condition that no super-cooling is available.(2) The tensile strength of freezing porous material may play the role thatsuper-cooling does at ice lens initiation,(3) Thc new ice lens initiation process discussed in this work is favorableto explain the data obtained in this experiment, and the variation of icelensing properties with respect to change in tensile s~de strength.REFERENCESAkagawa, S., S. Goto, tmti A.Saito (1988) Segregation freezingobservcd in welded tuff by open system frost heave test. Proc. 5th, Int. Conf. Permafrost, pp1030-1075.Akagawa, S. (1990) A method for controlling stationary frost heaving.Shimizu Tech. Res. Bull. No.9, ppl-8.Akagawa, S. (1991) Studies on the process of frost damages to stoneremains under cold environments and its preservation method.doctoral thesis submitted to Hokkaido IJniv.Gilpin, R,R. ( 1982) A frost heave interface condition for use in numericalmodelling. Proc. 4th Can. Permafrost Conf., pp4S9-465.YWdPw =. ...___ , t (3,yi .1053

THE ENSURING AND ECOLCGICAL SAFETY OF THE GAS PIPELINES OPERATING IN THE PERMAFROST ZONEVitaly P. Antonnv-DtuzhininRRTTWDEINFO (INGEOTES) Antonov-Druzhinin Private Enterprise,Novy Urengoy,Tumen Region 626718, RussiaThis papr contains data on methodical backgrounds, large-scale experimentalresearch and fie1.d test of pipelines in a permafrost zone. It is shown that inthe course of the experiment and during the obse'tvation of the operating gaspipline behavior displacements of up to 100 m were determined. I : isdesirable that they should bt: of a seasonally cyclic nature. This can beachi-evsd through the creation of a special temperature condition of gas beingtransported which prevents the effect of cryogenic heaving on the gas pipeline.The lat.t;er effect results in the disturbance of a seasonally cyclic nature ofthe gars pipline displacements. It may sometimes lead to vertical pipelinedisplacements of 300 mm and more. Designed displacements may exceed 2 m perseason. The latter phenomena seem to ke very dangerous for gas pipelinesopration i.n the pnnafrost zone.Z-THRGPGOENIC STRUCTURES IN GEQSY STEMS(LANDSCAPES)~ OFPERMAFROST ZONEProblems created by oil and gas fieldsdevelopment, in Arctlc regions attract muchattention in the discussion of interactionof civil and industrial buildinqs andst>ruct:ures with permafrost.The investigations carried out mustpermit us to eva1uat.e changes in naturalenvironment, to single out theanthr-opoyenic component of these chanqes,to enslure accident-free operntion of oiland qa.5 transport units, safet-y of people,envirooment.al control in the mineralresources production regi.ons of the Arctic.Consider t,he spatial-temporal structureof the Tas t.ranspor-t geotechnical system asa natural anthropogenic physicngeographicalohjert-, faking as an example the "pipelineenvirorlment" systein.These areas and zones are shown (in crosssection) in Fig. 1.The paper Fuhrnitt.ed t.n your considerationdi:cur;ses ecologi.ca1 after-effects o€ thecclnstruction and operation of gas pipeliresin a nerm;jfr-or;t zone and nertains to zoneslarqely determined by natllral conditions.Even in case of uking ident,icalconstruction technolcgy these zones havedifferent configuration.The main task of our investigations is toen.o,ure accident-free operation of oil andgar; transport units, saf-ety of; people. Achilled ga,s ~ipeli.ning experiment has beenperformed on a test site of an operatinggas pipeline. In the course of theexperimsnt we have received data on thefreezing peculiarities of different soilsin the gas pipeline base. The freezing ratewas found to be 0.5 to 1.6 rnm/hour and thatof the pipeline displacement 45-65 mm.Computer simulation of the temperaturefields of the soil around the gas pipelinebeing chilled has been carried out. Thesimulation results are in qood,agreementwith field test experiments. Classificationof permafrost zone geosystems (Landscapes)reflecting ecological hazard of natural gaspipelining has been dsveloped.However, the ecoloqical safety of gastransport system in the permafrost zone is1argel.y determined by its thermodynamicssystem. Most of all the thermodynamics ofthe gas transport system depends on the gaspipeline interaction with the surroundingsoil, air and water in region 8, zones 9-16(Fig. 1). Just. these processes arediscussed in this paper. It is known thatthe formation of a soil zone exposed to allthe year round thermal effect with the"ice-water" phase transition in soil is aprocess the most typical of the northerngas pipelines in a permafrost zone. Thisprocess manifests i.tself in: 1) theformation of thaw areas beneath thepipelines when gas is transported at atemperature above zero (t"C>O)- "warm" gas;2) the fobmation of freezing areas duringthe transport-ation of gas chilled to asubzero temperature (t"C

I1 - Region of geosystem anthropogenic transformation,which is formed during gas pipelineconstruction (region exposed to effect of gaspipeline under construction) including: 2 - zoneof complete mechanical surface destruction; 3 -zone of partial mechanical surface destruction;4 - buffering zone: 5 - zones of indirectFigure 1. The structure (in cross section) of thethermal effect: wi t.h "ice-water" phase tram.-tion in soil (6), without "ice-water" phasetransiti-on in soil (7); 8 - reqion of gposystemanthropogenic transfornmtion, which is formedduring gas pipeline operation (region of oysr-atinggas pipeline effect); includinq: 9 - zone ofwind regime change; 10 - zone of surface waterflow change; 11 .- zone of d.i rect thermal effect,consisting of: 12 - air zone exposed to directthermal effect; 13 - zone of surfare waterexposed to dirert thermal effect; 14 - ruck zcxlpexposed to all the year round thermal effectwit.11 "ice-water" phase transit ion in soil; 15 - ,rock zone exposed to all the year rmmd thermaleffect without "ice-water" phase t,r;msition insoil; 16 - rock zone exposed to seasonal directthem1 effect with "ice-water" phase transition;17 - pipe of the gas pipeline; 18 - soilsurface; 19 - boundary of zone exposed to ef fcctof engineering structure resulting from dynamicequilibrium of interrelated sahsystems (herclstrichis directed to natural geosystfm).gas trarlsport geotechnical system.geosystems of a landscape stow type: pretundraforests (light forests) growing onunfrozen sands (natural type I) and onpermafrost clay loams and sandy.loams withmean annul soil temperatures (tBv,) of about0 to l.O°C; tundras on permafrost sandyloams, clay loam and clays with tav, of -3.0 to-3.5"C (natural type 11) ; peat bogs with tmv+of -3.0 to -4 .0"C and -4 .o to -5.0"c(natural type IS1: type 111-1, free fromvein ice; t.ype 111-2, containing vein ice):high grass bogs with tav, of -2.0 to tl.O"C(natural type IV) .The main natural peculiarities nf the gaspipeline operation in sollthern foresttundrawere represented in one of our

Natural type ..Conficpration of zone 14 (Fig. 1) for "warm"qas transnortationIteract ionA points of temperature fieldcontrol. in soils surroundingpipe of gas pipelineA points of pipe displacementcontrolThawing-freezing soils boundary (bergstrich isdirected to freezing soil)Soils: 1) seasonally thawed2) seasonally frozenFigure 2. The natural and geotechnical conditionsof the gas pip :lining test site.According to field testresults(2)06.02.86According to computersimulation results(b)After 31 days of "cold" gastransportation23.03.86After 85 days of "cold" gastransport at ion26.04.86After 123 days of "coldgas transportation"r- 1IFigure 3, The tempprature field in the soil surrounding the gas pipeline.1056

Table 1. Physical properties of soils in the base of the gas pipeline test site (Fig. 2Y:(accepted in computer simulation)NaturalThermalHeat capacity perTotalconductivity,unit volume+Lithologicmoisturewei:J:l of(kg-cal/m-hour .c) . (kg-cal/m3composition'Ontent'pondingskele,t.on,W . forforfor foxto Fig.2)kg /mfreezing thawinq freezing thawingI I 1* I sand I 1600 I 16 I 1.75 I 1.55 I 480 I 590 I111-2 peat, 2.5 m 200200 1.150.1570 900sand 1600381.75 1.55480 590sandy loam 1650? 301.7 1.6 590035peat, 1 m. 200400 1.150.7510 900sandy loam 1650 30 1.1 1.6 590835111-1I I 1 I I Ipeat, 1 m 200 400 1.15 0.7 510 900clav loam 1100 32 1.55 1.45 590 835the results are duplicated by means of analog computationexperiments which was carried out on a testsite of an operating gas pipeline route(Fig. 2). The length of this site is about2.5 m. According to the program of thisexperiment the transported gas temperaturewas changing from positive to negativevalues. This enabled to observe differenttypes of interaction between the gaspipeline and its soil base,It was found that during thetransportation of gas at low temperatures(t"

Natural. condi.-t: i on?3.. Fre-tundra forqstsand lightfor~st.s nn sandysoil.~ wit-h discont:inuausuerrnafrostcompositionSoil3stateConfi urat ion-ofzone 11 (Fi.g. 1) forthe case of "warm"gas transportation: . - X X x X k x"-4. High boys with peatydeposit more than 1 mthick1 - Fipe of the gas pipeline, 2 - thaw- phase transition in the soil: soils compoing-freezingsoils boundary (bergstrich is sition: 4 -' sand, 5 - sandy loam, 6 - claydirected tn freezing soil), 3 - zone of the loam; soils state: 7 - seasonally thawing,all year round thermal effect with water seasonally freezing, 9 permafrost.Figure 4. Heat disturbance of the permafrost surrounding the gas pipeline with "warm" gas.5b. The displacement resulting from the effect ofcryoqenic heavi.ng .175% l9B9Figure 5. Seasonal displacement of the pipeline

soil surrounding gas pipelines, it is veeysimple and oh~ious. Fig. 3 shows a goodarjreernent between the computed andmeasurement results.FIELD TESTRJEARCH 05' PIPELINES IN PERMRE'KOST ZONE,~Field test and large-scale experimentalresearch of pipelines in permafrost zoneunder conditions of planes (sim.il.ar tothose of the permafrost zone of WesternSiberia) must take into consideration atleast five types of gas pipelining naturalconditions (Fig. 4) ._"Field~ TestIn our approach to field tests ofpipelines the object of investiqation isnot only the gas pipeline as a technicalobject but the gas pipeline and its naturalenvironment as an integral. gastranspiration system. During theobservations we study all zop.es shown inFig, 1. Such an approach is calledengineering-geological monitoring. Underthis term we understand observation, dataaccumulation, prediction of thegeotechnical system development as anatural anthropogenic object. The main aimof engi.neering-gedlogical monitorina isoptimization, provisior! cf reliability andecological safety of the geotechnicalsystem.At present the service of engineeringgeologicalmonitoring is carrying outobservations on 75-100 km sites ofoperating gas pipelines. These are mainsites of gas transporting systems ofWestern Siberia, which ace operating underpermafrost conditions.The observation system was designed withaccount. of methodical backgrounds mentionedin the first section of the author's paper.In the process of observation wedetermined:1) the thermal field in the soilsurrounding gas pipeline; 2) the dynamicsof d.irect thermal effect with "ice-water"phase transition in soil (zone 14, fig. 1);3) the ground water level beneath the gaspipeline; 4) displacement of gas pipelineand other technical characteristics.OPTIMIZATION OF GAS TRANSPORTINGGEOTECHNICKSYSTEM IN PERMAFROST ZONEThe brief summary may be presented asfollows :1. The "warm gas" pipslining isunreasonable because it is connected withthe permafrost thawing in the pipeI.ine baseand the activization of cryogenic processesdangerous for gas pipelines - thermokarst,thermal erosion, ground subsidence etc.2. The cold gas pipelining is dangerousbecnl.l:;e of the acti.vization of frostheaving ahd fracturing of s0il.s and theformation of frozen earth materials beneaththe pipelines.All the above-mentioned testifies to thenecessity of developing such a temperaturecondition of the transported gas whichwould permit to control cryogenic processesin the direct thermal effect zones of gaspipelines (mainly in zoms 13, 14, 15, 16in the region of operating gas pipelheeffect 8, Fig. 1)-The temperature condition of thetransported gas developed by us permi.ts toprevent dangerous cryogenic processes inthese zones. Thereat seasonal qas pipelinedisplacements become comparable with thecxtsnt of seasonally heavi.ng of clay loamin a seasonally nature. As a result, thegas pipeline remains at the elevation mark(Fig. 5a). In case of violation of thiscondition the forces of cryogenic heavingmay affect the gas pipeline. This resultsin the deviation from a design position(Fig. 5b) and possible dangerousdisplacements (Fig. 5c).The data presented are obtained duringfield nesting of pipelines in the processof engineering-geologial monitoring of thegas transporting systems in the North ofWestern Siberia. The pipe displacementillustrates in Fig. 5a is determined on thegas pipeline fixed in soil with anchordevices and transporting gas undertemperature condition developed by us.We have presented the methodicalhackgrounds and the results of large scaleexperimental research of pipelines in thepermafrost zone being carried out by thedepartment of northern geotechnical systemsof the trust of Engineering-geologicalMonitoring and Research. It is worth notingthat the development'of temperaturecondition for gas pipelining, which permitsto control cryogenic processes, is based ona special permafrost prediction.Using the computational algorithm we havesimulated a problem of finding the monthly .average gas temperature which ensures a 0.5 mthaw radius around the pipe all year round.The simulation results are presented inFig. 6.t,'CIn '?- 1a 1L-L4I0 I , " ,I46IloIFI4; i c i i i j @ i j-- - gas temperaturethe temperatureI I -- mmh51of soil. surfaceFigure 6. Average monthly gas temperatureprovide the minimum thaw area (0.5 m) .around the gas pipeline.The first year of operating. The initialsoil temperature -1 C.

~ ComplexENGINEERING AND GEOCRYOLOGICAL STUDIES OF THE CENTRALPART OF YAMAL PENINSULA CAUSED BY ITS DEVELOPMENTV.V.Baulin, A.L.Chekhovsky, 1.1. ChamanovaPNIIIS, Okrujnol PR. 18, 105058 GSP; Moscow, RussiaThe album of different scale engineering geocryological and forecast maps has been drawn up by thePNIIIS to design the projects of extraction, treatment and gas piping. The special features of,geocryological conditions of the territory are reflected the occurrence of ice beds and saline frozen soils.slope processes ect. The forecast of geocryological conditions changes is given by means of differenttechnogenic impact upon the environment.In 1990, Russia began to develop Bovanenkovo gas field situ-’ated in the central part of Yamal peninsula (west Siberia).Engineering and geocryological conditions of the peninsula arewmplicated and specific. Permafrost that have the annual average#temperature from -7T to -2OOC is spread everywhere in the gas‘field area. Permafrost from the surface and down to the depth of3-4 m is characterized by high content of ice (0.4-0.6) which decreases to 0.2-0.4 with depth increase. All the sea deposits that occupy75-80% of the gas field area are salinized. Their salinizationchanges from 0.1 YO to 1 % and more. Due to this fact frozen clayishsoil remain plastic even under low annual temperatures. This makesit more difficult to determine strength and deformation characteristicsof salinized frozen soil and significantly diminishes.the carryingcapacity of pile foundations.Sloping processes are extremaly dangcrous. Depending on themeteorological conditions and the active layer soil composition soilof the slopes at the steepness of 1.5T can be mobile. When carryingout field studies we obscrved the whole scale of slope processesfrom the classic slides to viscous soil flow on the foot of theactive layer. Destruction and damage of the vegetative cover facilitatesslope processes.On the territory containing sea deposits massive ice is widelyspread. It reaches more than 100 m lengthwise and is 1Om and morethicker. Slide 1 shows the map of massive ice spreading inBivanenkovo gas field. The map is made on the basis of abundentdata (over 100 massive ice stripping holes, dozens of outcrops ) andcomputer interpreted aerophotographs, spacephotos andtopographic maps. It was found’ that massiveice properties areconnected with numerous diagnostic indicators of its disposition(over 40): age of the topography, its morphology. number of lakesetc.The reliability of the method applied is 75-80% with massiveice at the depth of 3-5 m and 60-70% with ice at the depth of 5-10rn.PNIIIS’ studies in Bovanenkovo gas field were carried aut inthree directions: regional engieering and geocryological studies;survey of cryogenic processes and geocryological forecast; studiesof salinized frozen soil physical and methanical properties. In thepresentation, we are going to dwell on the first two directions..Regional engcering and geocryological studies were aimed atproviding the designers with the survey data at every stap of design work. For this purpose we have worked out map legends andcarried out engieering and geocryological mapping in three scales.A 1:100,000 scale map of ecological and geocryologicalregionalization contains data on engeering and geocryological conditionsin Bovanenkovo gas field area, permafrost resistance totechnogenic exposure and on environment- protecting measures(Slide 2). In the process of making the map we have determined everyidentified engiecring and gaocryological site’s rcsistana totechnogenic damages. The map also provides data,on allowabletechnogenic damages and on recommended complexenvironment-protecting measures for every engineering andgcocryological site. Depending on the potentialities of emergence oractivization of cryogenic processes at disbalanced natural conditionsresistance of the geological environment to technogenic exposurehas been estimated. we have identified three categories of engineeringand geocryological sites-resistant, relative resistant, andirresistant ones.of environment-protecting measures that arc suggestedfor every engineering and geocryological site give informationon vegetation removal possibilities; recommended fillingheight; critical quantity of snow accumulation; filling slopes reinforcement;and surfaoe drainage.Engineering and geocryological regionalization maps by constructionconditions at the scale of 1:25,OOO make it possible to estimatethe complexity and approximate cost of construction worksin every engineering and geocryological site (Slide 3). In the processof regionalizing such factors as slope angles, prescncc of massive icd1060

and cryopegs, soil salinity, its temperature and ice content in it, aswell as percentage of cryogenic damages and formations in the unit,and flood zone size are taker. into consideration.hccordine to the estimation of various combinations of theabove-mentioned geocryological factors several complexity levelsfor construction sites were graded.Finally, we have marked 3 levels of construction sites complexity-the least complex, relative complex, and complex sites. To theleast compleax sites we attribute plain terrace surfaccs and hasureysformed by soils with low ice content; gentle slopes and back floodlands are related to the second group; the most stecp slopes formedby soils with high ice content and river bed-side flood zones are related to the third group.Maps on engineering and geocryological conditions andregionalization by allowable technogenic load at the scale of1:5,000 are made for the southern part of Rovanenkovo gas field.Primary construction sites are concentrated here (Slide 4). Scenari.os of geocryological condition’s damages are worked out for everyidentified engicering and geocryological district and for 7 mainconstruction type’s. For evcry typeof construction technology withinengieering and geocryological districts we have suggested designdecisions that provide engieering construction stability.To observe requirement of preserving thc naturalgeocryological conditions and constructions stability it is necessaryto take various measures aimed at regulating ground temperatureand depending on the construction type. First, these are varioustypes of refrigerating facilities (ventilated basements and foundations,cooling tubes, thermopiles, heat insulating fillings). Besides,it is very important to install efficient water-conducting systems,belt-formcd frozen fillings for preventing slidings etc.In addtion to the general engieering and geocryological mapsand the general ecological and geocryological maps there were developedlegends for special geocryological maps and a gas field areamapping by ccrtain permafrost characteristics and cryogenic processcsdevelopment were carried out.On the special map or salinized frozen soil and cryopegs at thescale of 1:25,000 there were distinguished three types of salinization- a chloride, a sulphate, and a hydrocarbonate type, with heavydominating of the first type (Slide 5). Soil salinization in certaincases exceeds 1 %. On flood lands and slope terraces, upper parts ofsectionsto thc depth of 2-10 rn are formed by unsalinized andlow-salinized soils that have famed under continent conditions.Below them there occur salinized sca deposits. On the surface,salinized permafrost soils are deposited only within the surfaces ofsea terraces that are not affected by denudation.Therc were determined mechanisms of cryopeg forming andwide-spread occurence in the salinized frozen and,cooled soil. Itwas found that cryopegs occur at absolute marks from +20 m to-200 m within all the geomorphological levels but the most lensesare deposited at the 5-40 m depths in flood land.Cryopegs have formed in the course of the Pleistoccnc and theHolocene, they keep on forming at present as well. the map showstwo types of cryopegs, their mineralization, temperature, and brinepressure in lenses. Cryopegs of the first type are deposited at depthsfrom 4 rn to 10 m and are dynamic, their lenses are thin and differentin horizontal dimension. The existence of the lenses and theirmineralization are closely connected with short-period temperaturevariations in the upper part of permafrost. Cryopcgs of thesecond type are deposited at the depth of 10-40 m and deapcr, theyare characterized by quasi-stationary condition and considerablepressure. Lenses are up to several meters thick, they reach 300m in.,extent, and cryopeg properties are less dependent on the surfaceconditions. IOn slide 6 ydu can see the map df permafrost thickness inBovanenkovo gas field at the 1:200,000 scale. The map presents thethickness of the negative temperature soil layer containing ice andthe depth of the zero isotherm. It was verified that the thickness ofpermafrost grows with the geomorphological level age increase.Within the third sea .terrace, the thickness if permafrost containingice makes up 200-225 m, the zero isotherm goes at thedepth of 325 m. For flood land the figures are 150-175 m and250-275 m respectively. ,Slide 7 presents the thormokarst forecast map. It gives informationon the process dynamics under different types of the surfacedamages (vegetative cover destruction and cutting away ground ofdfferent thickness) with today’s climate and the forecast globalclimate warming.Technogenic damages of different scale can lead to reversableand irreversable changes of the initial geocryological situation. Inthe gas field area one can identify three groups of sites according toa potential risk of thermokarst development. The most risky aregentle slopes with massive ice close to the surface- here one can scethe connection of thcrmokarst development with thermal erosionand creeping processes. the map is built on thc data of field observationsand computer model testing.To finish the presentation wc would like to show the map oftechnogenic damage in Bovancnkovo gas field area at the scale of1:25,000.A forecast of changes in the geocryologic?l situation in termsof the identified types of technogenic damages is also given at themap. The damages arc classified by three indicators. Ageocryological changes forecast for 10 types of damages is made forthe periods of 3-5 years and 20-25 years. 13-14% of the gas fieldarea under development is affected by technogenic damages. Theprinciple conclusion is that the today’s level Of tcchnogcnic exposurehas caused the development of separate ( not IIUmerOUS) localscats of the intensified cryogenic proctssta. Both field obamationsand forecast indicate that within the boundaries of the most of thedamaged sites there goes a gradual rccovcry of the initialgeocryological conditions. This process passcs ahead of theecological situation recovery because the primary vegetation communitiesare being replaced by the secondary vegetation.The studies in Bovanenkovo gas field were carried out by thelarge group of PNIIIS’ explorers as well as by specialists from otherresearch institutes (MGW, Poundamentproyect. VSEGTNGEO).The authors of the report are very thankful for them for the ftuitfdand highly professional work without which the presentationwould not be possible.1061 -

METHODS OF LARGE SCALE ECOLOGICAL AND GEOCRYOWICAL CUSSIFICATIONOF THE NQRTHRRN PART OF WESTERN SIBERIAA.L. Chehovaky, 1.1. ShamanovaIndustrial and Research Institute for Engineering Investigations of Construction,Moscow, RussiaOn the basis of many yearn of complex engineering 'geocryological and ecological investigationsin the northern region of Weatern Siberia new methods oi large ecaleecological and geocryological classification including general recommendations onpossible engineering disturbances to nature have been developed.Ecological and geocryological classificationhas been made on the basis of estimation of territorialdifference of landscape and geocryologicalconditions and forecast calculations on thechanging of geocryological and landlrcape conditionsunder engineering disturbances to nature.Ecological protection of developing territoriesreguirem preservation and minimum impactto natural landscape and engineering geocryalogicalconditions and also provides safe design forengineering structures. While undertaking ecologicaland geocryological territory classificationand napping, it is necelrsary to provide stabilityof buildings and structure8 with minimum engineeringinterference into nature. On the basis ofmany yearn of complex engineering geocrylogicaland ecological investigations in the northernregion of Western Siberia (Yamal Peninsula), ourInstitute has developed new methods of large 1scale ecological and geocryologicaL classificationincluding general recommendations on possibleengineering disturbances to nature.Ecological and geocryological claesificationhas been mads on the basis of estimation of territorialdifference of landscape and geocryologicalconditions and forecast calculatione on thechanging of geocryological and landscape conditionsunder engineering disturbances-to nature.The final stage was the development of recommendationson providing for atability of engineeringstructures and prevention of the development ofdangerous cryogenic processes.The diagram of natural division ia bailed on indicationsdetermining the main peculiarities ofengineering geocryological conditions. Thepeculiarities of the given region axe geornorphology,meso and micro relief, slope of the surface,drainage,. vegetation, surface rtediment composition.The stated types of natural micro-regionsare complex indicators of gaocryological conditions.The type designs of natural micro-regions .are made in accordance with construction developmentconditions. The determining parameters fortype designs of engineering and cryological rituationaare:1) average annual temperature of permafrost;2) ice content of upper layers of permafrost;3) presence and depth of menomineral ice beds;4) presence and depth of saline soil andcryopegs ;5) surface slope.Taking into consideration these parametersthree type8 of mglneer-geological regions (EGR)have been distinguished.EGR Ir Flat terrace surfaces with slopes lessthan 1.5 and permafrost characterized by thetemperature from -5'C to -8°C. The deposits ofupper permafrost zones are icy till 3-4 m. Icebed deposits are found as a rule below 10 m. Thesails m e saline along the entire geological section.EGR 11: Terrace slopes and lake depressionswith surface slopes more than 1.5 and characteristicpermafrost temperature from -1°C to -6'C.The deposits of upper pennafrort layerer are characterizedby high ice contents. The depth ofsheet ice occurs below 2 m. The soils are salinealong the entire geological section. The cryopegzones are found at various depths and i.n someplaces they are having hydrostatic pressure.EGR 111: River flood plains with lake depressionsand surface slope8 less than 1.5 and permafrostcharacterizing by the temperature from-1'C to -4°C. The deposits of upper ,zones are highice content. Sheet ice deposits are possiblebelow 10 metres. The soils possess salinity below4-5 m depth. Cryopegs with hydrostatic pressureoccur below 6-8 m.EGR I is most suitable for construction. Lowaverage annual temperature of rocks, the absenceof sheet ice below 10 m depth creates favourableconditions €or foundation of engineering strnctureseven with the presence of salinity soil#.EGR XI is le88 suitable for construction sinceconsiderable surface slope determines the developmentof dangerous slope processes, soil'ssalinity over the entire section, high Ice contentof soils with sheet' ice and a wide tempereture interval of permafrost. The complex of negativefactors presents considerable difficultiesfor construction and engineering exploitation.EGR I11 occupies an intermediate position betweenthe EGR I and EGR 11. The difficulty Oferectinq enqineerihg structures is connected withrelatively high permafrost temperature, presenceof cryopegs with hydrostatic pressure and at the

low plain flood lands with flooding of the surface.The absence of salinity soils to the'depth4-5 metres and sheet ice to the depth 10 metresare favourable factors for construction.Nature protecting measures include a number ofactivities providing structural stability of soilunder the foundationrs according to principle I(preserving soil in permafrost .-condition). Themost characteristic four types of engineeringstructure8 considered are:I) apartment huildinga and industrial structures;11) gaa pipeline (gas transportation with subzeroaverage annual temperature is being planned):111) embankments of linear structures;IY) above-ground pipelines on piles.For each type of ntructurea the maximum poasiblechangee of geocryological conditions (depthof seasonal thawing, average annual temperatureof pemafrost) and engineering loads (types of engineeringdisturbances) are given here which naycanse the change@ in tolerance limits of engineeringgeocryological conditions. The increase ofsoil foundation temperacure and active layerthickness even in area8 most suitable for buildingconstruction areas of EGR I, is not permittedfor apartment and industrial buildings. This canbe explained by the fact that mils are salineand the rise permafrost temperature to -4OC andmore will transform soils from solid into plasticfrozen state.The increase of active layer thickness maylead to the formation of impoundments under build-ings and structures, thawing a€ soils and loss ofstructural stability. At the sites with averageannual temperature of permafrost higher than -4'C(EGR XI,. III), it is necessary to diminich thefoundation soil temperature to its transformationinto solid frozen state. To prevent the slidingaf solids an the slopea (EGR If) it's neceraaryto decrease the thickness of the active layer.Under construction and maintenance of structuresfor types I1 and 111 the necessary temparatureconditions of foundation soils are providedby the structure of the building it..self (roadand railway embankments) or by accepted temperaturemaintenance behaviour (gas pipelines withsubzero temperature of gas).For the structures of type IY - above groundpipelines an pilea, the sites with saline soilswith temperature higher than -4'c (EER If) areunfavourable. Such sitee need to diminish thefoundation soils temperature for 3 - -4'c andkeep the temperature condition of soils duringthe operational period with the urse of seasonalcooling devicsr. At the sites of EGR IXI with annualaverage temperature of permafrost higherthan - 4OC, the additional decreasing of soiltemperature is not needed since 4-5 m depth thesoils are not saline and are in solid permafroststate.The above mentioned requirementa ofgeocryological character are reached by fulfillingnumber of engineering practices directed tothe regulating temperature behaviour of foundationsoils and which depends on the type of engineeringsites and geocryological situation atthe construction site. The main methods of thesoil temperature preserving are the providing ofvented space under a building and placing ofcoolingpipes and channels in foundations of structures.The choice of the method is baaed on thethermal engineering calculations for specific. sites.The construction of underground pipelines withsubzero annual gas temperature and embankment8 oflinear structures do no+ need special actions forthe regulation of temperature of foundation soils.From the point of view of dangerous engineeringgeological processes on the territoriee closeto engineering structures, the most complex activitiesare needed for EGR 11. Active layerslides, which are widely apread on alapes innatural conditions, may become more active underconstruction development of the area. To preventthe development of thia process it ia recommendedto make earth fills of 1.2-1.5 m of height at theconstruction siter and adjacent territory, toform a good water overflow, to construct a permafrostbarrier. Within the limits of EGR I1 andEGR 111, the main activity is directed to preventthe increasing or dacreasing of permafrost .temperature. The main problem here is to providefree drainage of ground water at the constructioncites, organization of drainage stmctures, maximumlimitation of transport across the tundra insummer. EGR I11 requires also the prevention offormation of snow driftrs.The set of activities providing stability ofengineering structures and preventing the developmentof engineering geological processes in differentengineering-geocryologipal regions aregiven in a special table in the mapla explanation.These are connected with the rrcheme ofnatural micro-regions by common figure indexes ofngtura protection activities.The map, made on the bwin of the givenmethodology contain the infomation necessary foroptimal location of epwcific engineering sites,development of nature protection activities andrational use of geological environment. Theyserve aa a cartographic basis af engineeringgeocryological monitoring.

EXPERIMENTAL STUDIES ON ICE SEGREGATION AND THE MODES OF FROST HEAVINGChen Ruijie' and Kaoru Hofiguchi''State Key Laboratory of Frozen Soil Engineering, Lanzhou Institute ofGlaciology and Geocryology, Chinese Academy of Sciences, China.21nstitute of Low Temperature Science, Hokkaido University,Sapporo, JapanFrost heave rate of freezing soils has been paid close attention t;, andnumerous theoretical and experimental studies have been done since the beginningof research on freezing soils in cold regions. The temperature gradient acrossthe warm side of the ice lens has long been considered as the driving force ofwater flow, and the existance of a frozen fringe has been thought of as adefin-iteness and used as a banis for further research. Some experiments werecarried out in which ice lens grew without temperature gradient in the soil. Thecomparison between ice intrusion temperature and equili,brium temperatureindicated that the frost heaving occurred under general conditions (loose soilin natural field) usually is primary heaving.INTRODUCTIONSince Miller's putting forward the concept offrozen fringe, it has been accepted and usedwidely, and most of ice segregation processeshave been dealed with on the basis of the theoryof secondary heaving. The existence of frozenfringe in frost heave process be seldom suspected,though it has hardly been observed up to now.And the temperature gradient in the frozen fringehas.been considered as the driving force ofwater flow toward the ice lens.Frozen fringe is a thin zone of frozen soilfrom the growth edge of-the ick lens to O°Cisotherm, according to Miller's theory (1978).Konrad and Morgenstern (1980) got a theoreticalconsideration for one - dimensional frost heavewithout externally applied load, which was thatthe water intake flux is proportional to thetemperature gradient in the frozen fringe. Asearly as 1974, however, Vignes carried out amodel experiment by using glass tube andconfirmed that frost hea e rate is a functionof the ice-water meniscua temperature and js *independent of the teeperature gradient of theunfrozen part along the capillary. Besides,Ishizaki,et al., (1985), performed frost heaveexperiments and obtained a linear relationshipbetween water intake rate and the warm sidetemperature of the ice lens, Ozawa, et al.,,(1980), used micropore filter in the experimentsand had the ice grow on it, and also obtainedthe same results with lshizaki (1985), althoughthe experiments arc different in form. Horiguchi(1992) analyzed the frost heaving processtheoretically, wrote out the water and heat flowequations during ice segregation, and obtainedthe linear dependence of water intake rate onsegregation temperature.So far, however, most of the researchersthought that water intake ,in the forming of icelens necessarily results from the temperaturegradient in the frozen fringe. Hence, we carriedout an experiment in which the temperature ofthe soil sample keeps homogeneous in ice segregationprocess. More over, it becomes probable tomake cleat' the occurrence of the modes of frost:heaving by comparing ice in-trusion temperatureand equilihrium temperature.EXPERIMENTS AND ANALYSISThe apparatus uscrl in the experjments isdiagrammed in Fig.1, the soil sample used isFujino Mori clay, its grain size distributionis showed in the paper ahout ice intrusiontemperature.GTFig.1 Apparatusthat of the ice intrusion experiment except thatThe experimental procedure bas similar tothe screw lid not be used here. Af!:er seeding atthe bottom tube, ice advanced to the bottom o€the soil sample, and ice lens formed here. Thewater lev,eI in tube GT was observed decreasinggradually.

'FiglZ and Fig.3 are the diagrammatic sketchof the experiments.c0-81sSKwx\\\\\\ \\\ \ \ \\\IbI.?bTf9, e,> 0,wndaty havingFig.2 Diagrammatic sketch of the ice segregationin the experimentPig.4 Analysis of the modes of frost heavingconfined pressure (see this proceedings, Chen,et el.).nAPIII .I \Fig.3 Temperature condition of the experimentTb - buth temperature:Tf - warm side temperature of the ice lens,that is, ice segregation temperature.Because of releasing of latent heat in formingof the ice lens, ice segregation temperatureTf probably be slightly higher than the surroundingtemperature Tb. It indicated that the temperaturegradient on the side of water supply isin reverae to that of the so called frozenfringe. Tf though the latent hiat released isnot enough to make temperature Tf higher thanTb, Tf is equal to Tb at least. Then no temperaturegradient exists in the side of water supply.Now, measuring ice segregation temperature Tfaccuratly is still a difficult problem. But in aword, temperature gradient and frozen fringe arenot always necessary in the forming of ice lens.Fig,& is a sketch map about the mod.es offrost heaving achieved by comparing ice intrusiontemperature with equilibrium temperature, accordingto the definition of ice intrusion temperaturein the paper (see this proceedings). Whenthe ice intrusion temperature Bi of a given soilis lower than the equilibrium temperature Bewhich comes from Clausius-Clapyrom equation, icecan not intrude forwards to form pore ice in thegrowth side of the ice lens. The frost heavingthen be primary heaving, and no frozen fringeexists. Otherwise, the ice can intrude forwardsin the growth side of the ice lens to form poreice, frozen fringe, if the ice intrusion tern- 'perature 8i of the soil i,s higher than theequilibrium temperature Be. And the frost heavingthen be secondary heaving.Tf and Tc is the ice segregation temperatureand cold side temperature of the soil respectively.Fig.5 is the relationship between ice intrusiontemperature and confined pressure, thepotted line drawn up comes from Clausius-Clapyromequatipn. The curve represents change of ice+ntrusign temperature of silt or clay withFig.5 Analysis of the modes of frost heavingFor a given soil, if the cgnfined pressure4P reaches a critical value Ap, the ice intrusiontemperature 8i of it will be eqyal to the equilibriumtemperature Be. If PP

segregation temperature:then, from equation(l), we get:Comparing constan~s h and B which representsthermal and hydraulic condition of the systemrcspcctively, we get three kinds of situationsas follows:1) B >> AT t indicates that water intnke rate is controlledby thermal condition of the system(Ozawa, 1989).2) H i< AB -"A+U - n

The Rclationship between Ice Intrusion Temperatureacd Confined prcssureChen Ruijie' and Kaoru Horiguchi'' State Key Laboratory of FTOXn Soil Engineering, Lan Zhou Institute of Glaciologyand Gcocryology ,Chinese Acadcrny of SciencesInstitute of Low Tempcrature Science,Hokkaido Univcrsity,Sapporo;JapanIce intrusion tcrnperature is a temperature near and below OC at which the freezing ofpore watcr can advance continuously with no water migration, and unfrozen water contentand hydraulic conductivity of thc soil decrease sharply at it. Thc cxperiments indicatcdthat ice intrusion ternperaturc decreases with incrcasc of the confined pressure,that is, there are diffcrcnt ice intrusion tcmpcrature for different pore size because ofdifferentdegrcc ofcompression,WTRODUCTIONIn cold regions, the downward freezing of soil surhcc,whcn winter corning, usually is not duc to ice nucleationwith occurrance of supercooling, but due to icc intrusion.Thercforc, ice intrusion temperture is a tctnperature whichdiffers from the stablc frcczing ternpcraturc aftcr supercoalingprocess and its dependcce on confir~cd prcssurc is significantfor study on thc proccss of icc scgrcgation, cspcciallyon the modes of frost hcaving. Thc ficczing of fttcc watcrcan advance at about OC as wcll known. 7'hc frccing ofpore water, however, advance only at a tcmpcrature below0% , according to thc pore size because of adsorption. For agiven soil, different pore size responds to difkrent confinedpressure. The relationship between ice intrusion temperatureand confined pressure wdS obtaincd through ice seedingat the bottom of the soil.A little watcr was put inside the cylinder holder firstand the rubber tube FT was stopped up by a stopper to assurethat no air bubblc exist in bore hole F. An air-free porousplate N1 was then put on thc bottom plate in the cylinderholdcr and adequate amount of slurry of the soilsample was put on thc porous platc. Afterwards ,the topplatc N2 was pushcd into the holder and thc excess waterrnovcd up into thc tubc GT. 0 ring was used for sealing upthe holder. Certain amount of load was used on the topplate C to comprcss thc samplc Tor cnough long time untilthe water lcvcl in tubc GT did not move. Then the lid Dwas screwed down to fix top plate C, and the laad wastaken away.APPARATUS AND MATERIALSThc apparatus uscd is diagrammcd in Fig.1, bras cylindcrA is connected with bottom alatc R in which thcrc is aborc holc F, 2 mrn in diamctcr, connoctcd with a rubbcrtuhc FT. There is also a bore holc in thc ccnfcr of thc topplatc C, which is conncctcd with a rubbcr tube GT. Lidcan bc screwcd down along thc cylinder ta prevent plate Cfrom moving upwsrd. It means that soii samplc S, can beconfined in the cylinder with certain pore size.The materials used in the expcrimcnts are CorundumA (3-7 pm), Corundum B (24 pm)? Fuiino Mori clay C,Alluvial clay 13 and Manaitabashi clay E. Fig.2 and Fig.3 isthe diagramm of particle size of Corundum A and Brespcctivcly. Fig.4 is the grain size distribution of all matcrialsuscd.s I IICE INTRUSION EXPERIMENTSA cold liquid buth was uscd and controlled at a temperaturcnear and bclow OC , and its temperaturc wasmcasurcd by standard thcrmomcter.Fig. 1 Apparatus

The apparatus prepared was immersed in the cold liquidbuth which was controlled at -0.1C. A small length oftube GT was left outside of the buth so that the water levelin it could bc observcd. Tubc FT was then seeded by puttingit in liquid nitrogen after about 24 hours. The ternpcratureof the cold liquid buth was lowered by -0:lC step bystcp, and it was observed that thc water levcl in tube GTmarched a grcat dcal abruptly at ccrtain tcmperaturc. It indicatcdthat thc sccding icc had intruded into thc soil, thatis, ice intrusion occurred. 'I'hc tcrnpcraturc showcd by thestandard thcrrnomctcr is thc icc intrusion tcmpcraturc ofthis soil undcr givcn confined prcssurc. ,8'is5.360504c," 3002 2c!3z 100 IlL5 10 15 20 25Diameter of particle(pm)0.001 0.01 0.1 1Grain sizeFig.4 Grain size distributionFig.2 Particle size distributioaAIce intrusion temperature ("C)Ii 00 1 2 3 4 5 6 7Diameter of particle (pm)Fig.3 Particle size distributionP8xPore radius (pm)Fig.5 Relationship between ice intrusionTemperature and confined pressure '

RESULTS AND DISCUSSIONThe Experimental results of thc rclationship betweenicc intrussion temperature and confined pressure of the materialswere showcd in Fig.5 Curve A, R, C, D, and E represcntsCorundum A ( 3-7 pm), Corundum R ( 3-7 py),FujinoMori clay, Alluvial clay and Manahbashi clay ,.,rcspcctivcly. It can bc found that for fine soil, such asFujinoMori clay and Alluvial clay, thc icc intrusion tcm- IPCrdtUrC dccreascs with incrcasc of thc confincd prcssurc.Thc icc intrusion tcmpcraturc of cnrsc particlcs, such asCorundum, almost kccps constant with incrcasc of confinedprcssurc, that of Manaitabashi clay, howevcr, showsrclativcly low valuc compared with other samples, becauseof its fineness, though there is only one point obtained.Assuming that the freezing frant at each pore shows ahemispherical meniscus, thcrc will be a depress of thc freezingpoint. This depress corresponds to ice intrusion temperature.The right hand axis in Fig.5 is pore radius calculatedfrom the following equation which corncs from theClausius-clapyron equation and Laplacc equation:in which r,, and 0,respresents pore radlus and ice intrusiontemperature, respectively, To, pi, and L respresents meltingpoint of ice, density of icc and latent heat of melting ice,rcspcctivcly.It can be found that thc pore radius dccrcascs with in+crease of confined pressurc for fine soil. That is, ice intrusiontemperature of fine soil decreases with decrease of poreradius. The two kinds of come particles, i.c, Corundum Aand B, keeps constant because it is almost incomprcssibleand its pore radius keeps almost unchangc with change ofconfined pressure.REFERENCESKaoru Horiguchi 1989 Short Rcport: Relationship betweenConfined Pressure of a Sample and Ice Intrusion Temperature.Low temperature Sciencc,Ser.A,48.1069

PRELIMINARY TESTS OF HEAVE AND SETTLEME.NT OF SOILS UNDEIiC~OTNOONE CYCLE OF FREEZE-THAW IN CLOSED SYSTEM ON A SMALL C1ENTRIFUC;E

thouswd dayFig.5. Modelling of tuodels for prototype sand (2m) at lOOg and Z ~ levels JFIW dmost the same for the same soil. These facts indicate thatcc.mtrifug;c modelling of frost hcavr and thaw induced settlcment ofsoils is fcnsible if careful attention is paid to controlling the conditionsa d motlcl prcp;lratioll. Om can also fi11d from the figuresthat frost heave of china c1a.y is much larger (ncarly 1.5 times) thrtmthat of smd wihh the same thickness (2m), which can be mainlynt,trihuted to the larger wrttcr contcnt (0.93) in the forrncr thanthat (0.32) in tho latter.It is very interesting to observe in these tests that thaw inducedst.ttlement of sand are different from that of china day. Settlementof s a ~ is d almost qual to its heave, while that of china clay (slurryhefore freezing) is ncnr to twice its frost heave, and ils might hccxpctcd I,csulted iu a significnnt, hndy of water on the surface oftl~cllino. clay Irlodcls on completion of the tcsts. A similar findingP:I.II 1-w S~CII ill kaolin (Fig.3). This fact indicatcs that freezingcscrts a clesiccating effect 011 clay to some extcut similar to thatI'r1~1::ci in frcczc/tllaw treatment of orgn.nic vwer sluclgc.On coIrlparing Fig.4 with Fig.5, it is seen that frost heavc ratio(frost heave amount to original soil thickness) of sand is less thanthat of clha chy for thc samr dimcnsion (thick 2m) significantly,t111tl tllat tho hcave rato for sand changes from a slow to a sharpintrra.so and talm ahorlt 700 days (prototype equivalent) to reachits hcwr pdc, 011 the otllcr l ~ud, the latter (1.83mm/day or so)incretrscs almost constantly except during the very initial sta c,;mrl ils tluration of hravu is little longcr-ahout 975 days. For &etllaw inclucecl settlement patterns, s11nc1 differs from china clay. Tl~emtc of t,ho forlner cl~angcs from slow to very fmt, wllile the lattertlccrcast:~ ppdm~lly. Thc duration nf scttling, however, is very closeIiloagll tllat of sand is slight,ly shorter (880 days for snnd and 910days for cllirlit clay).

, -011 comparing frost he;l.vc of the same soil at diffetent g-levels,it CXTI be seen that the henvt: ratio slightly decrcmes with increaseof g-lcvc~. Fur example, thc ~ w ~ ratios v e uf cllina c~ny are 9 andS.4 'X ut 100~-level and 200g-lcvc1, resprctivcly, (from F$4.). Thcstme trcnrl can be found for sand in Fig.5, Fig.6 and Ig.7, andfor silt in Fig.8 and 9. This fact indicates that g-level may hassome suppressina effccts OII frost l~cavc, that is, the greater the g-level, the slightly less t,he frost heave in closed systcrn. Howevcr itt n q be possible to attributc this to the radially varyiIlg g-field ofthe wnttifugr, which rimy hc significant yrt~icularly in this sn~allwntrifugc. Trsts urldcrtdxw i u largcr Inac lme may be less affected.% Fig.8. Heave and settlement UP Fig.9. Heave and settlement ofprolatype silt (2.5m)prototype silt (8.2m)Silt umlcrlfoes the 1;ugrst m;lglrit,tlde of frost heavc, as showrl.it1 Fig.8 and Fig.9. In thc fnrmcr the frost hcave ratio rcaches 6.8'X' ant1 ill t.1~ latter that is INWC tu 7 (K, drubst close to that ofchills clny ill which, l~owc~vc~r., watcr contcnt (0.03) is larger thant,hat (0.30) ill silt.To frccac a prototype rrp~ivdent. silt. layer of thickness 8.2m1let:ds more t,hn 9 years at a t,enlperatrlrc -15"C, to drfreeze it,howrver, last,s about more than 13 years if temper:tt,ure on the soilsurface itlcreitSes from -WC to +~P,~C' in ;lpproxinintely one yearand is tlwl kcpt constant. To freczr silt of thickness 2.5m needs1.5 ycars tit B temperature -16"C, on ther other hand, to defraezeit liLst,s almost 2 years, according to the centrifuge modelling ofprotoype soil. One can find the freezing and thawing time for prototypechinn clay and sand in Fig.4, Fig.5, Fig.6 and Fig.7.The resu1t.s indicate that g-level does not appear to have obviousinflurnct: on patterns of frost hcave and thaw induced settlcmentof the same soil, a.9 shown in Fig.8 and 9; in Fig.4; in Fig.5, 6 and7.It sllould be noted that the frost heave ratio for testcd soilsseems lnrger for a clused system whcn their water contents are takeninto account. The frost heave of soil in cluued system is mainlyatt,rilnted to its water content. If it is assulncd to be purely due't,o tlw cxpn.nsiou of watcr as it forms icr then tho expected heaver21,(,io is: WG x0 01 . 111 theso tests, however, the heave ratio of silttc:a&s 7% (expccted 4.4%) with water content 0.3% that of clhaclay is about 9% (expecttd 6.4%) with watcr content 0.93; that ofkaolin gets 6% (cxprcted S.S%) wit,h water crrritent 0.7; and tha.t ofsmd is proxinmtely 3.4% (expected 4.1%) with water corltent 0.32.'111~ rcsmn arc not char : d sho~~lcl I w furthrr investigatctl.CONCLUS1C)NSFroln tllc ;how! sotne 1)rclirninwy coldusions call LC presented.a). It appears possihlc to model frost hcavt: and thaw il~rlllcedsettlerneent of prototype soils accura.t,ely in small centrifuge, andthus by extrapolation in large centrifuge, with time scah as N'ttnd displacement (heave, setticmmt) as N. This certain$ warrantsfurt,her investigat,ion and if substansiated will demonstratetllc ccntrifuge to bc a useful tool for the invcstigation of prototypefrost 11eavc and thaw induced scttlctnrnt prohlcmu in shortenedtinc sca1r.a.b), An incrcwsing g-levo1 tqqx:ars to have sonw suppression cffccton frost hcavr of soils, i.c. frost ht:avc decreases very slightlywith increase of g-level. Howcver this may he due to the limitatinusof tlw g-fieltl in tlle snl;lll wntrifuge used md nccds furtlwrinvcstigation.c,). Thc g-level docs not appmrs to have an olwious infinenreon patterns of frost henvr and tlmw induced settlsment.(1). Edge friction hstween soil nlodel side and model coIlt,zinersi& does not. Rppears t,o have it11 significant cffcct on frost hca.vrwhctl container sides are well grcasrd.e). The frost heave rat,ios of tested soils in closed system seemlrlrger than would he expcctcd fwnl calculat.ions based on the waterc(.)lltcnt nlune.f). A slmll centrifuge of the typc used has boen shown to ha.vcLL vcry useful cnpaMity for perfurIning initid investigations to testtllc feasibility uf certain concepts relatively quickly and at a lowcost. The results show that further work on larger (more expensive)rnachine is wort,hwhile in this firld.ACKNOWLEDGEMENTSThe authors would like to thank the staff of the GmtcrhnicalCcnttifuge Centre of C~rnbridge University Engineering Dcpartmeritantl ANS 95 A for their kind help, especially Dr. R Phillips,N Baiter. Thc first author wislws to express his sincere gratitudrt,o his sn1)crvisor Professor A N Schofield for his very valuable adviceancl kimlIlrss, nncl to Dr. M D Bnlton for his much appreciatedhelp and nomination of the first nuther a Fcllow Cummorlcrshipin Churchill Collcgc Canbridge which offcred llim the postfnr 1992/93 with accommodation, facilities and etc. And finally,thr first author is very grateful to Sir Y IC Pm Foundation (HongICong), the Cllinese Government and thr Britidi Council for theirfinancial aids.REFER.ENCESAndcrson, D M, and Pcnner, E (1978) Physical nnd therrnalproperties of frozen ground (pp37-102); in: Geotechnical cngineerinfor cold regiorls.Iictcham, 8 A (1990) Application of centrifuge tcsting to coldregions geotecllnical studics (internal Report of USACRREL).L~Iillcr, R I) (1OSO) Frcczing phenolnetla in soils (pp254-299); in:In applications of soil physics.Nixon, .7 and Ladanyi, B (1978) Thaw consolidation (pp164-215);in: Geotrchnical engineering for cold regions.Palmer, A C, Srhofield, A N, Vinson, T S and Wadhams,P (1985)Centrifuge modclling of underwater permafrost and sea icc;Proc 4th Int. Offshore Mechanics and Arctic Enfirieoring Symposiium.Texas, Vol. 11, (pyGC9). ,Schofield, A N (1980) Cambric #e gcwtechnlcal ccl-ltrifuge opcrittiuns;Twentieth R.anltine Lvrture, Geotcchnique. (pp227-263).Sclltsficltl, A N ;Inrl Snlitll, C c' (1993/4) Cold Rrgions Eugineeringin: Gcotc:chnical centrifugc technology. cd. h'. TaylorSmith. C: C (1991) Thaw induced srttleInent of pipeline in centri-. fugc model t,ests; P11.D. disscrtation of Camlxid c. University.Vinsnn, T S antl Ptllrnrr, A C (1DSS) Physical mode? study ofartic pipline setfletnrnt; Proc 5,th Int. Conf. Permafrost.Trondhcim. Vo1.2, (pp1324-1329).1072

FROST SUSCEPTIBILITY OF POWDERED CALCIUM CARBONATEChen Xiao-bail, Corte A.E?, Wang Ya-qing' and Shen yu''Lanzhou fnst, of Glaciology & Geocryology, Academia Sinica, Lanzhou, China'Regional Investigation Centre of Sciences & Technology, Mendoza, ArgentinaFrost susceptibility tests of a powdered calcium carbonate (FCC), collected fromthe Andes Mts near Mendoza, Argentina, were conducted both in closed and.opensystems, Experimental results show that an ice segregation and consequently anice lens and strong frost heave could occur in FCC while rich in moisture. Thefreezing point Tf('C) of PCC mainly depends on the moisture W ( X ) with a functionof Tf=-Sl99.62 W-3*27'. The frost heave rate of PCC R (mm/day) in a closedsystem increases with its water content W(%) intensively while W>17%. In anopen system, the frost heave ratio O(Z) decreases sharply with the frostpenetration rate Vf(cm/day), and its regression equation is rl=4,495 VT'*'P9.AS a result, the frost susceptibility of PCC is very similar to that of clayeysoils. After moisture was controlled or mixed with a special agent, the PCCcould be very light or not frost susceptible and then could be used as amaterial in subgrades or base of buildings.- INTRODUCTIONEXPERIMENT RESULTS AND ANALYSESPowdered calcium carbonate is rich in coldand arid areas, such as in the Padagonia regionsand Andes Mountains in Argentina, as well as inTaishan Mountains area of China. Up to now, alot of ancient buildings or structures in Chinaremain unchanged because of stabilized basesmixed with powdered lime. At present, the limestabilizer is widely used as an anti-heavecushions in China.The purpose of this paper is to determine thefrost susceptibility of PCC collected from theAndes Mts near Mendoza, Argentina and to find away of using it as a construction material.EXPERIMENT COUDITIOBSFreezing PointThe experiment results showed that the freezingpoint only depends on the moisture. Fig.1illustrates the relation curves of the poweredcalcium carbonate. After regression, the curvefunction can be expressed by:Tf=-S199.62 W-3'274, rm.943 (1)T.(C 1The samples of FCC, collected from the AndesMts near Mendozs, Aryent.i.na were less than 1 mmin size. The freezing point was determined bythe supercooling method with various watercontent. - For the .f,rost susceptibility test, thessmples with a given density and moisture wereput in plexiglass cells, 11 cm in diameter and2 cm in height. A thin grease layer and amembrane were set up between the surface of thecells and samples in order to reduce the frictionresistance. The cells were surrounded byinsulation. The multi-stages of temperat.ure atthe surface and the bottom of the samples werecontrolled by cycle' refrigerator baths with asensitivity of ?O.O2"C and an accuracy of tO.l"C.The temperature and the displacement of sampleswere determined by thermocouples and displacementgauges and were collected by HP 305L-SAutomatic Data Acquisition and Contra1 System.For an open system, tht. distilled water wassupplied from the sample bottom towards thefrost front from a special reservoir with aconstant water level. After testing, thessmples were cut into pieces for observation of.ice segcegation and measurement of the moistureprofiles along the depth.Fig.1 Freezing point of PCC vs water contentin where, Tf - freezing point of PCC, 'C:V - water conFent of samples, X :r -,?orrelation coefficient.Frost SusceptibilityThe results of 5 groups 3f frost susceptibilitytests for the PCC with different water contentand almost the same density, as well assame boundary temperature, are listed in Tablel,In which, sample CC-I. is in an open system, The1073

Table 1. The results of frost susceptibility tests on powdered CaCO,under different water conditionscc-1 126 1.62 25.51 -.13 22.58 lWLh 4.35openCC-2. 78 1.62 25*76 -.13 10.11 *3.11 closedcc-3 98 1.63 19.80 -.30 7.031.72 closedcc-478 1.62 15.15 -.71 0.270.08 closedcc -5 94 1.59 17.97 -.41 1.06 0.27 closed*t - elapsed time; h - frost heave; Q - water intake,frost heave and water intake processes are shownin Fig.2. It is obvious that the increase ofheave is respondent to water intake.Q(4xxl"-4-""""" ""0 '"fl4 \h(m1- I0 20--------- - --I I I I40 60 m a 0Hqrrsd tlina(h.s) *Pig.4 The temperature isothermal line and frostheave process of Sample CC-3O ~ " " " ~ ~ 1 1 l l0 20 49 60 8 o w o mElrpJcd tbne(hs)cc-2Fig.2 The frost heave and water intake processesof Sample C,C-IPig.3 and 4 illustrate the temperatureisothermal lines and heave processes. It isshown thst the heave will reduce with thedecrease of moisture in a closed system."II I L0 20 40 60 80Hrplsd tb(WFig.3 The temperature isothermal line and frostheave process of Sample CC-2Fig.5 is the frost action profiles of SampleCC-I, CC-2, CC-3 and CC-4 after freezing,respectively. In which a large amount of icelenses occurred in Sample CC-1, 1 to 2 mm inthickness and 1 to 2 mm in interval, bec.ause ofwater supply during freezing (sev l'hoto), frostheave occurred as well. For Sample CC-2 withan'initial water con ten^ oi 25.76%, a s.naI1pingo developed in' the top of ice lenses, 0 .2to 0.5 mm in thickncss and 0.3 to 0.5 mm inFig.5 Frost action profiles of Sample CC-1,cc-2, CC-3, cc-4interval. For Sample CC-3 with an initialmoiscure content of 19.80%, a fine ice crystaldeveloped in the centre of the top and ii fewice 3.enses under the crystal. For Sample CC-4with $111 initial water contcnt,of 15.15%, therewas only a thin ice layer 0.2 to 0.5 mm inthickness, which covered the upper surface. Thewater redist-ribution profiles of Sample CC-1,CC-2, CC-3 and CC-4 are illustrated in Fig.6.1074

of the frost penetration rate as shown in Fig.8and the regression equation is as EoZlows:in where, 0 frost heave ratio, %;vf - frost penetration rate, cm/day;Eq.(2) is very similar to that of the authors'earlier works (Chen and Wang, 1983, 1987, 1988).Photo: Ice lensos occurred in Sample cc-I4 liHim)4 13-L------- Ec-2 10 20LP"40 60 80~1802803M]W%)Fig.6 Water redistribution profiles of samplesafter free;..ingFrom Table 1 we also know that the frostheave rate of PCC in a closed system increaseswith the'moisture intensively whi1.e W>lhZ asshown i n Fig.7 which will be useful for controllingfrost heave by using limited water content.Fig.8 Frost heave raiio vs penetration rate OfPCC in an open systemApplication of An Anti-heave AgentAs mentioned above, the powdered calciumcarbonate is frost susceptiblc while the moistureis large enough. In order LO use the PCCas a material for subgrades or bases, besldescontrolling t.he water content, a special antiheaveagent could be used for reducing the frostsusceptibility. As an example, if the PCC ismixed with a given content of a special agentwith a lower cost, the freezing point might dropto -11.37OC. Consequently, this kind of ant.iheaveagent is powerful for changing the powderedcalcium carbonate with strong frost susceptibilityinto a very light or non-frost susceptiblemalerial.PRELTMINARY CONCLUSIONS1. The powdered calcium carbonate collectedfrom the Andes Mts, Argentina is frost susceptihle,and 3 large amount of ice lenses will occurforming pingues when it is rich in moisture;2. After controlling the moisture or whenmixed with a special anti-heave agent, thepowdered calcium carbonate could be very 1.ighLor non-frost susceptible and then might be usedRS a material in subgrades or bases of buildings.REFERENCESIIQ -"i" I I Ix3 15 x) 25 30W(%)Fig.7'FrosL heave rate of PCC vs water contentin a closed systemChen, X.R., Wang, Y.Q. et af., (1983) .Influenceof penetration rate, surcharge stress andgroundwater tab.le or1 frost heave, Proceedingsof VT TCOP, 131-135.Chen, X.B., Wang, Y.Q. et al., (1987) IceseEregation and frost susceptibility 01 sandygravel, Bulletin of sciences, 32(23): 181%-181 5.Chen, X.R., Wang, Y.Q. et al., (1988) Frosthcave model of sandy gravel in open system,Proceedings of V ICOP, 304-307.For Sample CC-1 in an open system, the frostheave ratio decreases sharply with the increasc1075

Comparison of Two Ground Temperature Measurement Techniques at anInterior Alaskan Permafrost SiteCharles M. Collins. Richard K. Haugen, andTimothy 0. HorriganUS Army Cold Regions Research and Engineering LaboratoryHanover, Kew Hampshire 03755-1290Two temperature measurement systems, aremoved each time a series of readings was made.string of Chinese-fabricated mercury thermometersFor statistical analysis, we grruped theseand a thermistor assembly, were installed in depths into three zones: the upper 1.2 m, theadjacent bnreholes drilled in fine-grained, peren- middle range from 2.1 m dowa to 4.5 m, and thenially frozen silt. The Chinese thermometers were lower Erom 6.0 m down to 10.0 m. During the summerprovided by the Chinese Academy of Railway Sci- months of 1988 and 1989 (Fig. l), the thermometersences, LFzhou. The thermometers read to 0.1 "C, yielded higher readings (as much 1.5 as 'Cand were supplied with correction tables also to higher). Conversely, during the winter months, th0.1 'C. The thermometers were assembled into a thermometer readings were lower. The pattern oflinear string to replicate, as closely pos- as seasonal differences was not evident during thesible, the thermometer strings that are commonly final year of the test evaluation (1990). In theused in China for monitoring ground temperatures. middle zone (2.1 m through 4.5 m), the pattern wasThe sensors were installed at the Caribou-Poker similar, warmer during the warmer months, andCreeks Research Watershed in interior Alaska (65' slightly cooler during the cooler months. The10' latitude, 147" 30' longitude). The thermom- warmest period of the year at this zone occurseters were placed at the surface and -0.6, at during the late fall. The thermometer readings for-1.2, -2.1, -3.0, -4.5, -6.0, -9.0, and -10.0 m, this middle zone are skewed upwards only duringwith insulating spacers placed between each ther- the warmer period (during smer the and fall) andmometer. The cable was suspended in air 5-m- in a not during the winter cooling period. 6.0 Below mdiameter casing installed in a 10-m-deep borehole. (Fig. 3 ), the two time-series €or the lower zoneThe thermometers were pulled from the hole peri- were almost parallel to each other. The thermomodicallyto obtain the temperature readings. The eter readings were higher than the thermistorthermistor cable consisted of an assembly YSI of readings by an average of 0.17 'C, suggesting that4407 glass bead thermistors placed at the surface 'the calibration of the t,hemometers was higherand at -0.3, -0.6, -0.9,-1.2,-1.5, -2.1, -3.0, than that of the thermistors. (The average differ--4.5, -6.0, -9.0, and -10.0 m. The assembly was ential between thermometer and thermistor readinginserted in a similar 5-cm-diameter casing in- was 0.17 and 0.18 "C in the top and middle zones.)stalled in a 10-m borehole located 30 cm from the Figure 4 shows the time series of the differencesthermometer string. The pipe containing the ther- between the thermometer and thermistor readings.mistors was filled with silicon fluid and capped. The upper and middle zones show seasonal patternsThe cable was extended some distance from the pipe of difference for most of the record; the lowerto avoid surface disturbance. Simultaneous mea- zone is stable in this regard.surements were made on the thermometer string and We attribute the minor differences inthe thexhistor assembly over three years. The measured values between the two sensor types t.othemistor string remained undisturbed within the apparent difference of about 0.17"C in calibrationsilicon fluid environment. The thermometer string, between the two sensors, and a possible influencesuspended in air with the insulating spacers, was of ambient air temperature on the mercury thermom-

0.0mm7Om7-0.5-1 .o-1.5-2.0-2.5-3.0Figure 3 . Average temperatures (TI, below 6.0 m, thermwmet.er & thermistor methods. 1077

~~ .-L2 1.5.-leE 1.0alc+ 0.5Y50U50.0-0.5E -1.0OIc-1 5m" mu "--I -.- -DlFF top layer0 mF-DlFF med layer "-*- DlFF bottom layerI.F-m-Figure 4 . Difference between thermometer and thermistor readings.eters at the time of reading. Average annualthat only minor differences exist between the twoground temperatures calculated based an the ther- types of ground temperature measurements. Ourmometers and the thermistors were within O.0loC Chinese colleague, Ding Jingkang, has obtained afor all levels if the apparent calibration differ- similar set of data at the Fenghuoshan Permafrostence is taken into account. We concluae therefore Research Station on the Qunghai Xizang Plateau.- 1078

PRELIMlNAHY STUDY ON THE FREEZING POINT I N SOILCui Guangxin and LiY iMining and Technology University, ChinaIt is pr'e3erlted that the freezing point aifected the depth of the freezing wallof B freezing shaft. due to thp differences of depth, water content and load inminixg engineering. The study method a1111 some rcsults are introduced. ,conductivity; r--coodinate radius: 't--tirne;XI ,X2--heat conductivity of unfrozen and frozensoi 1, respcctivc1.y: E--the coordinate of thefrozen well; B--latent heat, 8"temperaturefunct.inn: Ro--coordinate of freezing tube:I. Y "temperature of cooling source: to--preliminary temperature of soil; tD--frcez,jngpoint.Fig.1 l'empcrature fieldali)r~g the freezing wallThe strength anrl s~.ability of the freezingwall arc very important for enginccring. Thethickness ol the frcczi.ng wall is determined onthe basis of strength and stability. So far indesign and construc~ion, thc zero degree line isdefined as the boundary uf Lhe freezing wall,that is tD=O"C. Both in tests and engineeringpractice i t was proved t.ir;rL when there isuver~burtien P, the frcczing point of surface soil Iis lowrr tha11 0°C and not a constant, tD=f(p,w,s).Uowever the function has not been obtained, thcthickness of the freezing wall in design is not1079

equal to that in field. For example, the designedthickness of the freezing wall according to O0Cline is 7 m, while the boundary is along -2°Cline, so that the thickness in field is only5.6 m. It's dangerous to engineering. Designersand .technologists had to make them by decreasingthe boundary temperature lower than 0°C. 11'surgent and a requirement to obtain ttre functionof to=i(l',w,s) in thc design and ~onstruction offreezing shafts.IIIPressuresystemINTKO~ucrIo~The study of freezjng point under ovrrburdcnbegan in the USSR in the 1960s, and it began i nthe 1970s in China.in recent years, the freezinp yuint of soilwas fiet.ermined by the temperature increasing dueto released latcnt heaL Crum the curve of te~peraturevs. time obtained by thermalmeterduring soil freezing. H!I~ Lhis mcthod is notadaptivs to wl~en thr temperature .is not increasing.So this methud has great locality and error.Liu Zongc.hao (1987) proposed a criterion ofdetermining thc freezing point using volt changeduring soil freezing, and experirnent,al researchwas studied for the range of 4 MPa in an opensystem and the range of 10 MPa i n a closedsystem. This method is simple, arid volt changeis obvious. ?owever, it needs the premise Lhatvolt change is unique. Jt was regretted chatthe volt change around the l-reeaing point wasnot unique, RO t.hrrt some mistaken determinationwas made.Vreezillg point can be determined accorriing tothe diftarence of resistivity between the frosennone and unfroz,en zone. Gu Zhongwci (1082) showedthat, resisti-vity of soil depended on water content,salt content and soil properties. Thcaathor and-assistants propose an idcntificati.onmethod to determine the freezing point underoverburdcn usi.ng the mutation of resistivityduring soil Ireezing, and the freezirlg poinlfinder' is provided. He Ping (1'790) tasted thefreezing point for different. soils and 1,rovF'dthat there existed a11 obvious mutation ofresist.ivity during soil freezing under lowoverburden.RESEARCH 0F"INGPOINT UNDER OVERBURDENR~!Z~USF it is required to dctcrmine thefreezing point under overburrlcn in productionfi.c. freezing shaft.), we proposed a decisionscheme of I he rreezing point i.11 the range uf8 KPa of uverlruriiel~ for clay, sandy clay, clay~ysand and sand usj.ng t.be changc of restivitychangc durj.ng freezixg.Fjguror 2 shows the test systcm. %t is composedof a test nell, ~emperaturc sensor, €Itlidpressurcrcll and freezing t.ubtr. Thc scnsorohserverl 1.h~ frost linc and temperat.ure. Tnsaturated soil, pressure t.ransfcr exists, sothat prcssurc is applied by iniecting highyressurc liquid through a pressure hulc. l.iquidpressure is 1:nntrollcd by a pressure-sensor andpressure-control system. Liquj.d pressure in thefluid pressure cell i:; transferred to the saahplethrough a soft tubor Illembranc so as not to mixliquid wlth thc sample.It is important to det.erm'i.nc thc frost lineduring thc ,test. Although it. has proved thatresistivity changed obviously during soil freesing,the relation het:ween thc spot of resistivity mutation and freezing point (~f soil. has notheen idctntifiod by the !.est..UUFi.g.3 Freezing polnt finder'The 05 jective of the project is:1.. Find t,he relation between resisti.vitychange and frost. line during soil freezing.2. CorPect the freezing point fin(1r.r wi.thunclear magnetic resonance.PRlMARY RESULTS OF THE LAW OF RESTSTIVITY CHANGEDURING FREEZINGClay and sand are saturated and luadcd in thetest. The sensor with the rreezing point finderis composed of thermalcouples and clcctrodes tomeasure resistance. Tl~ermalcouples are locatedbetween two electrodes. The frozcn liquid isNaCT solution (-2°C). .Fig.4 shoks the processes of tcmperature andresistance i n saturatod sand. Where, linc 1 isthe curve of ternperal.ure v::. time, and line 2is the curve of rc$:iotnnce YS. tine. The Lemperaturehecorr~es obviously steady from 10 to 70mi~~rrtes, :~nd thc freezing point can be easilyde~ermined. Bcfore freezing, the change ofresistance is steady, after fieeP,irlg, itincreases rapidly. lt is intcresting thnr. the.distinct increase of resihtancc occurs at 60minutes, which curresponds to Lhe later period(not.. the beginning) of the temperature. It isindicated \.hat. ttre micro-phasc change heginsJ!: 10 minures. The ir:e forI:ls on B 1.arge scale athO mijiuteu and t.his res1rlt.s from a rapid increnseof rcsistance. Fig.5 shows Lhe relation betweenLemperaturc and resistivity. There exists apoint. of inflection E that corresponds tu thefreezing pujnt (point A in Fig.4.). Fig.6 is theamplified irlflected point. from Fig.',. Hcorr~sponds to super-cooling tempcrature. Atthis moment, the resi.stance does ,lot have an' 1080 '

acute change, soil is unfrozon. The temperatureat. the inflected point is -O.Ob°C, that is thefreezing poinL.I., 16006r.. * I \/{ ,200 2-0 20 40 60 80 100 120Time (min.)Fj8.A Process o€ temperature and resistancein saturated sandY30 60 90 120 150Time (min.)Fig.7 The proccss of resistance and temperaturein saturated clayIOoo r-6 -4 -2 0 2 4 6Temperature ("C)Fi g.5 The relation Irr.twecn resistance aridtemperature in saturaLed sc

CACULATION OF MAXIMUM THAWED DEPTH OF PERMAFROST UNDERTHE BLACK-COLOUR PAVEMENT BASED ON GEOTHERMAL GRADIENTCui Jianheng and Yao CuiqinThe First Survey and Design Institute of Highway,The Ministry of Communications, Xian, China ,From the engineering state of roads, geological conditions, present research situation and accumulatedinformation in the permafrost area of Qinghai-Xizang, the paper, evading the difiicult problem of thethermophysical process of seasonally frozen-thawed layer, directly considers the thermophysical processof frozen-thawed boundary and uses the geothermal gradient caused by various factors(air temperature,terrain, height above sea level, geographical latitude, ground surface condition, soil type,thermophysical property and the effect of water), presents a method to calculate the maximum possiblethawed depth of premafrost in a certain period.INTRODUCTIONOn the Qinghai-Xizang plateau. thc amount of heat absorbedby black colour road surface is in larger proportion to that absorbed by ground surface. Asphalt pavement restrains the evaporationof heat, in addition, the erects of permeation and convectionof water have very large influences on the scasonally thawed depth.when precipitation permeates into soil, its heat and the radiationheat of soil will be taken to the thawed boundary. It is difficult todescribe the complex thermophysid proccss by using athermophysical equation. But the mthermal gradient effected byvarious factors can be found by the observation of ground temperature.If the thermal flow along the road direction is neglected, thetemperature accumulated under black road is expressed by a twodimensional Iield(Ding Dewen,1983):The horizontal and vertical heat flows caused by geothermalgradients (aT/ aX and aT / ah) make up the total heat flow of soil:the relation of vertical heat flow to total heat flow is defined as theside radiation coefficient,. aTI -dxdvum and the rclation of vertical heat flow to total heat flow will beconstant.The paper only discusses the relationship of maximumseasonally thawed depth and the vertical geothennal gradient.HEAT EXCHANGE ON THE FROZEN-THAWED BOUND-ARYThe heat exchange on the from-thawed boundary obeys thelaw of energy conservation and exchange. For the frozen-thawedboundaty, the heat equilibrium condition must be obeyed,, it can beexpressed by the Stafen equation:where, I, and 1, arc heat conductivity of uppcr and lower layersfor phase change boundaty, t+nd t,are soil temperature of uppcrand lower layers, C is the thawed depth.Q, = L W Tdr is the consumption heat amount per unit volumeduring freezing or thawing, L is phase change latent heat Qf ioc, wis natural water content, Tdis dry density of soil.The left portion of the equation is the differenm of heatamount getting in and out the boundary, the right portion of theequation means that the phase change boundary moves at the rateof dF / dz because the heat amount absorbed by the boundary iscansumed by soil during thawying or freezing. Equation (1) is alsoexpressed as:when the boundary condition of the temperature field is defined,the side radiation condition will certainly be in dynamic equilibri.1082

geothermal gradlent at upper and lower phase change boundary.< iq the moving rate of the boundary.If q, = qT, 5 = 0, the boundary docs not move. If' the heat flowof- cvery point is constant, the upper limit is in the steady state. Ifthe upper limit descends, it certainly absorbs heat from thefrozen- thawed boundary. The main condition of the upper limitdescending is that annual average heat flow decreases jumpingfrom the upper laycr of soil to the lower layer. The dynamic state ofthe upper limit can be judged by the annual average geothermalgradient of the uppcr and lower layer of soil at the upper limitboundary, because a differcr,t geothermal gradient is correspondingto different heat flow. It has:If gT> (Am / A,.)gm, it is shown that the upper limit will descend. Ifthe annual avcrage heat current of the upper layer soil is equal tothat of the lowcr, the upper limit is in the steady state. Thegeothermal gradient at p,=(dm/ AT)gm is regarded as the criticalgradient of upper limit movement. If the thennophysicalparameters of permafrost do not change, the moving rate of theboundary may be expressed as the function of geothermal gradient(ET):If 5' =O , g, is the critical gradient at which the table starts moving.If- thc gradient is constant, the moving rate is an invcrscmensurcmcnt of'thc watcr amount (w - Td) of frozen soil.The above discussion is for the uniform medium, but thc a~tual soil laycr is not uniform.VERAGE STEADY TRANSFER HEAT PROCESSThe idea of effective annual average temperature will be introducedhere. The purpose is that a uniform soil layre, in which theheat conductivity is equivalcnt to the actual one, is used to expressthc actual soil laycr.If the state is steady and cffcctive annual average temperaturedoes not change, bccausc of no inner heat source, the averagc heatflow is constant, an average steady tempcrature field is formed andthe annual avcrage transfer heat amount is constant. Because theheat exchange between the ground surface and atrnosphcric layerchanges periodically, the heat exchange between the annual temperaturechange layer and the bottom of thc layer is relativelysteady. Therefore heat exchange between the annual change layerand outside is steady certainly in an annual period. The algebraicalsum of heat amount QT absorbed by the ground surface in an annualperiod with evaporation heat Qm equal to the heat quantity(q,t,) absorbed into the layer under the annual change laycr. Theannual average heat flow in ground is:where,q, is the heat flow entering into the lowcr boundary of theannual change layer, A is equivalent heat conductivity, t, ispermafrost temperature, to is effective annual average temperatureof ground surface, h, is the annual average depth of ground temperature.Therefore, the soil layer with different heat conductivitycan be replaced by a. uniform soil layer with equivalent heat conductivity:where hiis the depth of the lower boundary and Ai is the heat con--ductivity of'the soil layer of number i, respectively.It is shown that the heat exchange between thc annual averagestcady active layer and the bottom of the layer is not certainly equalto zero. The annual average temperature of frozen soil is not equalto the one of ground surface and is also not equal to the annual averageair temperature.For a largc continuous permafrost area, black road surface isonly a limited boundary condition, the heat transmitlcd from thesystem to permafrost is not enough to change the hcat state ofpermafrost. The temperature field under black road .may cornc toan equilibrium state in thc boundary condition for a period. Theequilibrium is dynanic, IL stcady artilicial upper limit may beformed under the large area continuous black road surface ofpermafrost, the equlibriurn is corrcsponding to the climate and environmentconditions for a period. The formation of a stcady upperlimit under the black road surfacc is essentially the equilib:iumquestion of excessive absorbing hear of black surface and heat currentin ground. In an average steady state, the average heat currcntpassing through the permafrost upper limit in a annual period is:where, tT is effective annual avcrage temperature at the steadypermafrost table, h, is tne cffective depth of a steady Permafrosttable. Therefore the critical gcothermal gradient can be obtained inthe average steady equilibrium state:The observation data of ground temperature shows that theannual average tcmperature at 0.h under the black road surface isa plus. It is shown that seasonally thawed depth is more than thefrozen depth for black road surrace in the average steady equilibriumstate. Therefore the unfro7xn layer exists ccrtainly betweenpermafrost and the seasonally frozen soil layer. The annual avcragetemperature at the uppcr limit is zcro. The equation (9) may bewri ttcn as :The above discussion is only for a one dimetlsional heat conductionproblem. But, in ground there is heat convection from wat-

er as well as heat conduction, the temperature field is not onedimensional. Therefore the relationship of the average verticalground temperature gradient and maximum thawed depth dependson ground temperature. The critical average vertical geothermalgradicnt (g,) is determincd by ohscrvcd ground tcmpcrature.whcrc, A is the change rate of Ah,. with g, , At a constant healflow condition, the influencing factors on A are the frozen soil typeand heat quantity (0,) consumed during ice thawing. It is shown asequation (9) that the critical gradicnt of ground tempcrature is theratio of equivalent heat conductivity of the upper layer Soil to lowerlayer soil at the tablc and times of average gradient of grdund ternperaturein the annual change layer. when the heat conductivity isconstant, go is related to (t.,.-tc) / (hc-hT). Generally VeakinL thetemperature (1, ) of frozen soil and the annual change depth (h,) isrelated to the height above sea level and latitude, By Convertingnatural soil of every observation point into gravel (&=2.56- 01i. 1DATA ANALYSTS3 4.. ~-0.5 1 / -0.51 /The data was observed every ten days in the period of plurtcmperature. Obscrvation intcrval is 0.Sm in depth. Because roadsurfacc tcrnperature is intluenced very largelyby the periodic airtempcrature of daytime, the deviation of the data is large. But theground tcmperature at 0.5m depth is influenoed very little by airtcrnpcrature of daytime, and for artificial road, the boundary conditionsarc similar in each of the road sections. The thermophysicalconditions at the depth of 0.5m may be defined as the upper-0.5boundary condition in the research system. In addition, thc changeof ground temperature in a year is periodic, average ground tcm- l.O}AhlIxraturc in a plus temperature period is a constant proportion tothe one year period, and thawing happens in a plus temperature pkriod. The unfrozen layer exists in the average steady equilibriumstate. When a seasonally frozen Iaycr exist, the heat cannot bc11-ansimitted into the unfrozen laycr. After the seasonally frozenlaycr thaws, the permafrost layer can absorbed the hcat from theupper layer. Thcrcforc, thc train of thought to look for the relalionshipof average ground tcmpcrature at 0.5m in plus temperatureperiods and maximun lhawed depth of' permafrost does notchange the substance of thc problem.Table 1 shows thc obscrvation data of ground temperature and-0.5 -0.51Fig.1 h vs. gcaicula!inn results ol' 6 ohscrvation points in different heightsnhovc sca level and latitudc from 1YS5 to 1990. Where Te is fhe averageIcmpcrature Iiom June to October at 0.5m depth, gr is calledthc calculated avcragc gcothermlal gradient, is the 1-atio of thcaverarre temperature differepcc bctwccn at 0.5m dcpth and at max.imum thawed depth from June to October to thc seasonally thawedwhcrc, K. is converting coefficient of soil layer (K =A, / An), thercgrcssion coefficient is 0.94.From rhc definition of calculation gradient of ground tempcradcpth.T!lc thawed dcpth is the depth where thc tempcrature is ture. thc actual maximum thawed depth of calculation points at a-0. I'C. Figure I shows the relationship of thawed depth h ofcvcry constant condition can be obtained:observation point year by ycar and calculated average geothermalgradicnt (gT). The line, passing through the points whert: g-, arestn;~ileI and Ah, arc largcr, is in unfavourablc conditions. The intersectingpoint of the straight line with g, axis is the critical calculardgradicnt (go). Wndcr the unlhvourable condition, Ihc rclation- where, Te is the avctagc tcrnperature at O.Sm dcpth from June toship ol-g, and Ah,. is:October(T).Ah, = A(g, --x,) (1It is confirmed by observation infomation that the g also ex-1)ists under na:ural ground surface. The natural upper limit isw/ m "C), the rclaticnship (see table 2) of g,' and height (Hs)above ses level and latitude (L) is:thought as thc average steady equilibrium limit of permafrost (seeTable 3,.Fig.?, Fig.3). The relationship of the geothermal gradientundcr natural ground surface and seasonally thawed depth can bethe referencc of' black colour pavement.CONCLUSIONThe theory of average steady heat transfer processes are introducedto research'the rciationships of steady table and the gradientof average ground lempcrature by means of the uniform soil layrereplacing the natrual soil layer with different conductivity.The vertical portion quantity of geothermal gradient measured

4750fable 1. The observation information of ground tempcrahucTable 2. Thc deta of do , H, and LNO.Ha L 1 Bo 8'0(rn) ("1 (W /mV) cC/m) ("C/ m)1 4149 35.61 1.65 2.70 1.742 4615 35.28 1.80 2.05 I .443 4140 35.18 1.66 2.31 t .544 4579 35.36 1.78 2.13 1.485 I34.77 1.70 2.60 I .756 4940 34.12 1.92 2.15 2.06"5.00 1 2.23Table 3. The temperatun: information of the natural groundsurface in the wcst of Xiesui riveryear-TC Tr h, gr"Ah,("C) ("c) (m) I0Cc/m) (m)85 2.586 4.10 "0.S 2.5 2.30 087 4.40 -0.5 2.67 2.45 0.1188 4.40 -0.5 2.67 2.26 089 3.9 -0.4 2.38 1.98 -0.2990 5.0 -0.1 2.75 2.71 0.37Note: g, is calculated by using gT-Te-T,r / h,-0.5.T, andr,. are the mean ground temperatures at 0.5 rn and the permafrosttable from June to October, h,last year.is the depth of the permafrost table0.5 /I 21'-0.50 lTT-kFig.:! H vs. g (natural ground surface in the north of Xieshui river)may be considered (the actual gradient of ground lemperaturc is influtnoedby heat conduclion, heat convection, et a].) by means ofthe statistical method of pure theroy which is uscd to develop theresearch of heat conduction, it is in accord with the actualsituation.The method given in he papcr can be used to calculate themaximum thawed depth not only for the pavement in large areas ofpermafrost but also for the natural ground surfacc and engineering.The method may be as used as a train of thought for continuing research.ACKNOWLEDGEMENTS4 Eip.? Mean ground temperature8cw~pt's in Kekcxili from Juneto October in 1995,1987,1989The authors wou 11d lil ce to thank very sinccrcly Plofcssors YuWenxue and Wu Jingming for the valuable help given.REFERENCESDing Dewen, 1983, A New Theory on 'I'cmpcrature and IJcat Regimein the Annual Average De$h of Ground Temperature.Proceeding of Second National Conl'erence cn Permafrost.Ding Dewen, 1983, Determination of Characteristic Valw of TcmperatureField in Annual Average Depth Gf Ground Temperature.Proceeding of Sermd National <strong>Conference</strong> onPermafrost.

OBSERVATION ON PSRTGLACIAL YASS MOVEMENT IN THE HEAD AREA OFURUMQT RTVEK AND LAERDONG PASS, TIANSHAN MOUNTAINSCui Zhijiu', XiongHeigang' and Liu Gengnian''Department of Geography, Peking University'Department of Geography, Xingjiang UniversityObservation of the movement rare and forms of talus, rock glacier, active layerand gelifluction in the periglacial environment in Tianshan Mountains have beenconducted since 1986. The talus, which is strongly developed between 3300 and3900 meters above sea level, is supplied by snow-avalanche, rock fall andalluvial forms. The down-slope moving rate of surface gravel on talus is 0.82-10.4 cm/yr, the mean value is 3.15 cmlyr: and the mean rate of the front lineof the talus is 2.22 cmlyr. The rock glaciers evolved from talus and have theshape of lobate. and are distributed at 3300-3950 meters a.s.1. Front movementrate is 0.96-49 cmfyr, mean rate is 21 cmlyr. The active layer is at 3460-3540meters, with a slope dj.p angle of lO"-lf' and slope orientation of 9O-17Ov, andit creeps down slope with a mean surface rate of 1.13 cm/yr. A 1,arge amount ofgelifluction lobates are distributed at the north-ward slope near Ltierdong Passwith an altitude of 2800-2900 meters, mean surface rate of 13.9 cm/yr and frontrate of 2.16 cmlyr.C1,TMATIC CONDITIONSRased on the data from Daxigou MeteorologicalStation at 3545 meters in the head'aren ofUrumqi Qiver, where the annual mean temperatureis -5.34"C and annual mean precipitation is 430em, the times of temperature oscillation around0°C are 130. The Hydrographic Station at 3805meters in the empty cirque at the head of UrumqiKiver shows an annual mean temperature of -7.5'C,and an annual precipitation of 380-Lf10 mm (1987-1.989). Annual mean temperature is -2.h'C andprecipitation is 830-840 mm (deduced from dataOE Snow Avalanche Observation Stntion).TALUSTalus is one of the most common alpine features,and ranges widely from the piefirnont t,othe divide and is distributed mainly in analpine periglacial environment with greater size,density and activity. Tn periglacial areas,strong freeze and thaw actions favour the breakingof bedrock and the abrupt valley side byglaciation offers a favourable relief for talusformation. Talus appears greatly at. 3609-3950meters above sea level.Observation was made to measure surface gravelmovement and baseline movement of talus in 198h-1991 fXiong Aelgang, et al, 14921, for- theresults see Table 1. Talus-1, located beneathGlacier 5, consists of til? from Glacier 5 andweathering debris from the upper slope. Talus-2is located at the trough base to the right ofGlacier 5 and mainly consists of frost. weatheringdebris from the upper slope. Talus-3 and Talus-4are located at the back and left side of theempty cirque. The observation started in June,*+The research was supported by NSFC and TianshanGlscial Observation Station1986 and continued until August, 1991. Talus-5 ,and Talus-h are located at the western side ofthe empty cirque, Talus-7 is located at theriegel of the empty cirque, Talus-8 and Talus-9are located at the trough of Glacier 8. Thesurface gravel movement of the upper, middle andlower lines was taken to show the differentmovement ra~es at the talas. Iron poles wereused as the datum point to determine the baseline movement of the talus. The results areshown in Table 1. Mean movement rate of thesurface gravel is 0.82-10.4 cmfyr, total meanrate is 7.15 cm/yr; the mean rate of the upperline is 2.7 cmfyr, middle line is 3.6 cmfyr,lower line is 3.0 cm:yr. Velocity of talus baseline is 0.26-5.7G-cm/yr, total mean rate is2.22 cmfyr.BLOCK SLOPEZhu Cheng** arranged a block slope measuringsite at 3990 meters, with br?dr.,ock slope angleQ€ 27", slope aspect of llOo, slope length of275 meters, debris width of 5.5 meters, debrisslope angle of 34', and debris s1op.e length of69 met-ers. Lithology is quartz schist. Remeasurementwas made in 1990 and it shows a 30-80 cmdisplacement among 13 points, the mean displacementis 38.46 cm and mean rate of surface gravel/-on the block slope is 7.69 cm/yr.The processes of surface gravel on talus arevery complex (Washburn, l97Y), they can be roll--ing, sliding, creeping and subsiding, etc. Themovement is sporadic. The measurement may onlyrepresent the condi til.'. in >I limited period.For example, surface gravel mean rare of 2.5cm/yr and maximum rate of 5.8 cmfyr was measuredby Rampton (1974) at 25' slope.talus in Yukon**Zhu Cheng, 1990, comparison of periglacialprocess among Tianshan Mountains, WesternAntarctica and Andes, Ph.D. thesis, PekingUniversity, 171 pp.1086 '

Table 1. Altitude, form and movement rate of the talus (Talus 1-4 in 1986-1991, Talus 5-9 in 1990-1991)Altitude Slope Slope length Middle width Slope angle Height of the Slope angle Rate of Rate ofNumber a.s.1, aspect of the talus of the talus of the supply area of supply surfacegravel base line-1meter degree meter meter talusdegree meter area cm yr degree cm yr-lTalus-1 3700 35" 270 115 25"-30" 4 50 450-55O 1.8 0.261.26 . 0Talus-2 3700 40' 90 65 30"-35' 450 60"-75" 10.4 4.1.88.40.8Talus-3 3950 135O 175 110 25'-32" 375 45"-60' 0.9 1.520.821 .h8Talus-4 3900 230" 210 95 25"-32" 350 , 50"-70' 1.25 1.821.49Talus-5 3900 75" 45 40 28" 560 41' - 1.4Talus-6 3920 80' 65 60 30" 610 .40" - I .03Talus-7 3040 230' 70 35 27' 355 39" - 0.88Talus-E 3640 265' 130 1728" 520 32" - 3.1Talus-9 3680 265' 180 35 28" 720 34.5" - 5.78n rL.3Territory, Canada. Rate of 6.5-111.0 cmfyr on25' slope talus was measured by Gardner (1973)in Lake Louise, Canada. The rate is 1-450 cmfyrin Chambeyton, French Alps (Michard, 1950). Therate is 0-22 cm/yr on 14"-38.4' talus slope inLapland (Rapp, 1960). It is suggested that therate of surface gravel on talus is in the scaleof several centimeters, the extreme scale canbe several meters per year. The activity and thesize of the talus are related to topography,climate, lithology and geulogical structyre.ROCK GLACIERThe rock glaciers in the head of IlrumqiRiver, mostly lobate-shaped and a few torrgueshaped,are distributed at 3300-3950 metersa.s.1. on the northward slope (Schatter Seite).Most of the rock glaciers are evolved fromtalus, known as talus rock glacier: a few frommoraine, known as glacial rock glacier (Barsch,1969; Corte, 1987). Form, composition, velocityand structure type have been measured since 1986(Cui and Zhu, 1989: Zhu;19891, the results areshow in Table 2. RC-1 is tongue-shaped with aratio of length and width of 1.6; the othersare lobate-shaped with a ratio of length andwidth of 0.2-0.58. The direction of movement ismainly northward, RG-5 is westward and RG-6 issouthward, Moving rate of the rock glaciers is0.96-49 cmlyr, mean rate is 21 cmfyr. The rateof rock glaciers in the Alps is 7-500 cmfyr,mustly 140 cm/yr (Vietoris, 1972: Aarsh, 1969:1975; Chaix, 1943). The rock glaciers in AlhertaPark of Canada move down slope by 0.3-0.8 cm/yr(Osborn, 1975). The rate of rock glaciers inAlaska Range is 36-69 cm/yr (Wahrfaftig, 1959).The rate of rock glaciers in Front Range,Colorado and Galena Creek is 2-15 cm/yr and 32-217 (mean 133) cm/yr (White, 1971; Potter,1969).The rate of rock glaciers in a maritime environmentis up to 1 meter or more, whereas in acontinental environment the rate is severalcentimeters,ACTIVE LAYERThere are long distance telephone wires,which were built in 1979, that have inclineddown slope by a slowly act.i.ve layer creepingsince then, passing through the head of UrumqiRiver. The amount of displacement representsthe down slope creep of the activc layer, sothe surface creeping rate of the active laycrcan be deduced. The dip angle of 77 wire poleswere measured and the slope features wereconsidered at an altitude of 3460-35L0 metcrs.It is assured that the creep of the active layerand the incline of the wire poles are simultaneousand equally-quantitative, the surfac.ecreeps quickly and the pole end is nearlyinactive. The distance hetwcer~ two poles is 25meters and the underground part of the poles is1.5 meters. The surface creep distance:S=hxsina/sin(90'-a/2)where: S displacement distance s.itlce (1979 (cm). a 90' minus dip angleh 150 cmThe mean moving rate of t.he active laycr surfaceR=S/12R: cmiyrThe results are shown in Table 3.The first part A of 25 poles, striking 245"with slope direction of 170' and slope angle of10"-15", creeps down-slope by 0.22-24 cnliyr andthe mean rate is 1.0 cm/yr. The pole A-2 wasloosened until iL crept down-slope 28.8 rm,since 1979, with a mean creep rate of 2.4 c-nlfyr.The second part B uf 29 poles, striking 240'with slope direct:ion of 150' and slope angle of11"-17", creeps down-slope 0.22-2.4 cm/yr andthe mean rate is 1.11 cm/yr. The pole B-1 atthe block strew> crceps 2.4 cm/yr and is themost active. The third of 23 poles, striking215' wit.h siope direction of 90'-140" and slope1087

I rm~~ITob1.e 2. Form and uovement rate of the rock glaciersThicknessp[umber Length Widthrneter meters met e r sAltitude of rockFrontp,l.acier tcrmina?Movingslope angledmeter a.s.1.irecti.ondep,reecm yr3350KG- 2 70 120 3.600R$": 60 15083.9 350044" 35" 0.96352" 60" 11.23305" 41" 45.0' KG-4 27 101 27 3950 352" 35" 15.5tlG-5 55 130 100. s 3500 2800 43 O 49.0RG-6 20 100.5 41,5 3900 200 32" 6.2+,. ,.. lable 3. Creep rate of the active layer at the head of Urumqi Riverdeduced from inclined wire polesDip Total Dip Totalanglr mcveaent Mean angle movement MeanNo, of the at rats Slope feature No. of the at rate Slope feature" pol,@ surface cm yr- pole surface cm yr-I" -cm-A-J 88 5*24A-2. 79 20.80' A-3 86 10.47A-4 84 15.71A-5 a7 7.85A-6 86 IC,&-?A-7 na 5.24A-8 a2 20.95A-9 83 18.33A-10 85 13.09A-11 85 13.09A-12 84 15.71A-13 07 7.85A"14 85 13 .09A-15 YO 0!.-I6 89 2.62A-17 80 0A-18 05 13.09A-19 85 13.09A-20 84 15,71A-21 87 18.33A-22 a5 13.09A-23 '35 13.0YA-24 ' 82 20.95A-25 80 5.24H-1B '- 2E-3R- 4E-50-6B-78-88-9R-10E-11B-12E-13B-14B-15R-160-1779818687868987888587858887a084828228.ao23.5610.477.8510.472.627.855.24Y 3.097.8513.095.247.855.2415.7120.9520-950,442.400.87I .310.650.870.441.751.531.091.091.310.65I.. 0900*22(i1.091.091.311.531.091.091.750.442.401.960.070.650.870.220.650.41,1.090.651.090.440.650,441.311.751.75Part A, Strike245", slopeaspect 1709,A1-12 dry grassland, slopeangle loo,A8 at. blockstream.A13-19, Wet grassland, slope anglell', A15, 16, 17were supported bywire.A20-25, Dry grassland, slope angle1l0, A20, 21 atblock streAmslope 15"."-Part 8, Strike240", slopeaspect 150",81-12., Yet grassland, slope ll",B1, n2 at blockstream, B6, 8supported, &I1at thcrmalslump.R13-24, Dry,grass landslope 140,R16, 17 at stge am8-18 88R-19 86R-20 86R-21 86R-22 ahB-23 86B-24 86R-25 80B-2h 81R-27 79B-28 83E-29 85C-l 89c-2 88c-3 80C-4 87c-5 81C-6 86c-7 84C-8 89c-9 a4C-10 84c-11 83C-12 87c-13 83c-14 84C-15C-16 a6C-17 86c-laC-19c-20c-21c-22C--23".5.2410.L710.4710.4710.4710.4710.4726.1823.5628. 8018.3313.092.440.870.870.870.870.870.872.181 .Yh2.041.53I .09B18 supported"B2.5-29 Wet.grass land,slope 15'-17",B25 at stream,R29 at blockstream.2.625.2410.477 .BY23.5610.4715.712.6215.7115.7118.330.220.440.870.651,960.871 I 310,221.311.311.53Part C , St.rike%15", slopeaspect 9Oo-14O0(21-13 Wet grassland, Cl, 4, 8,12, 16 supported,C5 at. blockstream, C9 atthermal-karstdcpressi on,C.L3 at block7.85 0.65 stream, C14-2338.33 1.53 Dry grass larid,15.71 1.31 slope aspect83 18.33 1.53 YO', slope 14'.10.47 0.8710.47 0.8786 10.47 0.8786 10.47 0.8786 10.47 0.8786 10.47 0.0786 10.47 0.8786 10.47 0.87"

angle of 12', creeps down-slope 0.22-1.96 cm/yrand the mean rate is 0.99 cm/yr. The pole C--5at the block stream creeps 1.96 cm/yr and isthe most active. The total mean c-reep rate ofthe surface of the active layer is 1.13 cm/yr.The sorted stripe ground at 1900 meters inNorthwest B.C., Canada, creeps 15-33 cm/yr(Makey, 1974), which is greater than that inTianshan. The down slope creep rare by freezingand thawing and gelifluction is mostly around0.1-12 cmlyr and the mean is 2.5 cm/yr(Washburn, 1979. Table 6.2). The movement ofthe active layer is influenced by climate.slope direction and angle, ground-water content,landforms and composition.GELTPLUCTlONMeasurement and sampling were made to studythe gelifluction near Laerdong Pass (43"10'N, 'r'84'20'T; 2888 meters a.s.1.) i.n August 1990 andremeasurement was made in 4ugust 1991. grainsizeand water content of a pitting at onegclifluction lobate were measured according torhe stratification, the result are shown inTable 5. The form of the gelifluction i s shownin Table 4. The surface moving rate of lhegelifluction is in Table 6 and the front movingrate is shown in Table 7.The form of the gelifluction in this area islobate shaped. The ratio of length to width is0.33-0.95 for most. of the gelifluction, and themean ratio is 0.59, only TLG-4 is tongue shapedwith a length to width ratio of 1.51. The heightin the middle front of the gelifluctions is1.55-2.55 melers and mean height is 1.86 meters.The slope angle of the top surface is 10"-30."and mean slope angle is 20". The front slope is48"-70' and mean i.s 59". 'The ground surfaceslope in front of the gelifluction is 12e-330and the mean 1.s 20". Vein-ice 0.5-1.0 mm wasfound at the depth of 106 cm as pitting. grainsize analysis shows the mean content of sand,silt, and clay are 41%, 41X and 18s: respecL.ivelyWater con ten^ rangcs from 9.74% to 75.217: witha mean of 31X. The water cont~nt is an inverseratio to sand content and direct ratio to clayproportion. Hard plastic tubes-of 120 rm wercinserted into the gelifluction in August 1'390and their, angles were remeasured in August.1991 (for results see Tahle 6). Rate of thegelifluction surface is 6.28-16.8 cm/vr andmean rate is 13.9 cm/yr. Wood bars were stuckin the front of the gel ifluction to determinethe front movcmcnt rate in Augus~ 1.990 andremeasured in August 1991 (for results sccTable 7). The minus values of the cast-ern sideat TLG-3 show movement: of the ground in front.of the gelifluction is greater than that of thefront of the gelifluction. The mean rate of thegelifluction front relative to thc grourld is2.76 crn/yr wi.th a maximum of 12.5 c.m/yr. 'Thcqoving rate of geliflucti.on varies in dirferentenvironments with the range of 0.5-10.0 cm/yr'(Dyke, 1981; Raginton, 1985: Frcnch, 19711:Mackcy, 1981). In Laerdong Pass, t.he grcat.f.rmovement is suggrsted to result from the greaterslope gradicnt, precipitation and grounrl wat-el-.CONCLUSTONThe talus, which strongly developsat. 3f)OO"". . . . . . .Table 4. Forms of the gelifluction at Laerdong Pass (measured on August 8, 1990)GclifluctionNumber Length Width HeightaspectmeterSl.opeangleGelif 7 uct.ionFront Groundsurr~ceangleang I eAlti cude21.3.1.m c L e rFr e :; t. 19" 590 2 3 OTLG-1 18.3 25.7 2.5 15" Middle 1.0 52" 20"East 175' 540 24"2800West 70 4U0 3 1 OTLE-2 8.55 17.5 1.55 150 Middle 2Y" 54.50 J 2" 2890East 7.2" 58" 1hWest 25" 6 'i" 22"TLG-3 2.8 . 4 30.0 2 .O 13' Middle 2'7 6 1 " 17.5' 2895East 20" 70" 21"TLG-4 15.7 10.4 1.65 3 7 1 6

~~ ~~Table 6. Hate of top surface movement of the gelifluction at Laerdong Pass1'ri.mary dip The dip angle Length of theMoving rateNumber angle of the of the next tube undergroundtube (Aug. 1990) 1991) year cm (Aug.cm yr""" "TLE-1TLG-:!I'LG-3TLE-4'TLG-512012012012012016.814.76.2815.016.6Table 7. Front moving rate of the gelifluctionnt Laerdong Pass (in 8,8, 1990-24,6, 1991)NumberTLG-1'SLG-2TLG-3TLG "4TLG-5TLE-6Western side Middle Eastern sidecm cm cm1.10.52.0--7 .O 1.00.7 -0.5-2.0 01.75 1 .o -12.5 I3950 meters above sea level in the head ofUrumqi. River, shows a surface gravel moving rateof 0.82-10.4 cm/yr and a mean rate of 3.15 cm/yr,a base linc creeping rate of 0.26-5.78 cm/yr anda mean of 2.22 cm/yr.The'block slope, which develops on the slopeangle of 34" and slope orientation of 110" at3900 meters a.s.1. in the head of Urumqi River.shows surface gravel moving rate of 7,69 cm/yr.The lobate-shaped rock glaciers, whichdevelop at 3300-3950 meters a.s.1. in the headof Urumqi River, move down slope 0.96-49 cm/yrand the mean rate is 21 cm/yr.The active layer, which is measured on theslope gradient. of 10"-17" and slope orientationof 90"-170° at 3460-3540 meters a.s.1. in thehead of Ilrumqi River, creeps down slope with asurface mean rate of 1.13 cmlyr.The lobate-shaped gelifluctions, whichdevelop on the slope angle of 20' and slopeorientatioh nort,hward at 28811-2900 meters a.s.1.near Lacrdong Pass, snows the surface movementmean rate is 13.9 cm/yr and front rate is 2.75cm/yr."-REFERENCESRarsh, D., (1969) Studien und Messungen anElockgletschern in Macun, Unterengadin,Icitschrift fur Geomorphologie, Suppl. Bd.,8, pp 11-30.Baruh, D. and Hell, G., (1975) PhotogrammetrischeRewegungsmessungen am Blockglet.scher Murtel1, Oberengadln, Schweizer Alpen: IeitscherGlerscherkunde, U. Glazialgeologic 2(2),pp 111-142.Chaix, A., (1943) Les couless de blocs du ParcNational Suissei Le Globe, Memoires R2,pp 121-128.Corte, A.E., (1987) Rock glacier taxonomy, inROCK GLACIERS ed. Giardino; Shroder andVitck, Allen & Unwin, Boston, pp 21-39.Cui, Zhijiu and Zhu Cheng, (1969) The StructuralType of Temperature and Mechanism of Movementof Rock Glacier at the Ycad of Urumqi Fiver,Tianshan Mountains,, Chinese Science Bulletin,V01.34, N0.15, pp 1286-1291.Dyke, A.S., (1981) Late Holocene solifluctionrates and radio-carbon soil ages, central.Canadian Arctic, Geological Survey of Canada,Paper 81-1C, pp 17-22.Egginton, P.A. and French, H.M., (1985) Solifluctionand related processes, easternBanks Island, N.bI.'P., Canadian Journal ofKarth Science, 22, pp 1671-1676.French, H.M., (1974) Mass wasting at SachsHarbour, Yanks Island, N.W.T., Canada, Arcticand Alpine Research 6, pp 71-78.Gardner, J.S., (1973) The nature of talus shifton alpine talus slope, in Research in PolarGeomorphology, cd. Fahey and Thompson, 206pp. ,Mackey, J.R. and Mnthews, W.H., (1974) Movementof Sorted Stripes, the Cinder Cone, FaribaldiPark, B.C., Canada: Arctic and AlpineResearch 6, pp 347-359.Mackey, J.R., (1981) Active layer slope movementin a continuous permafrost environment,Garry Island, N.W.T., Canada Canadian JOU~IIAI.of Earth Sciences 18, pp 1666-1680.Michard, J.,(1950) Emploi de marques dans l'etudedes mouvements du sol: Rev. EeomorphuloguieLdynamique 1. pp 180-18Y.Osborn, G.D., (1975) Avancing rock glaciers inthe Lake L,ouise area, Banff National Fark,Alberin, Canadian Journal of Earth Science12, pp 1060-1062.Potter, N.J., (1969) Rock glaciers and masswastagein the Galena Creek area, northernAbsaroka Mountains, Univ. Minnesota Ph.D.thesis, 150 pp.Rampton, V.N. and J.S. nugal, 11974) Quatcrnarystratigraphy and geomorphic processes on theArctic Costal plain and adjancent areas,Demarcation Point, Yukon Terri'tory, toMalloch Hill, n.M., Canada, Geol. SurveyPaper 74-1, Part A, p 283.Rapp, A., (1960) Recent davclopment of mountains'slopes in Karkeragge and surroundings, N.Scandinavia: Geugrafiska Annaler 42(2-3),pp 65-200.Vietoris, I,., (1972) Ilden den Blockgletacherdes Ausseren Hochebev-kars: Zei tschr,Gletscherkunde und Glazialgeologie R(1-21,pp 169-168.Wahrhaftig, C. and A. Cox, (1.959) Rock glaciersin the Alaska Ranee: Geol. SOC. America Rul1.70, pp 383-636.Washburn, A.L., (1979) Geo~ryology, RdwardArnold, pp 230-231,White, S.E., (1971) Rock glacier studies i n LheColorado Range, 1967-1968: Arctic and AlpineResearch 3, pp 43-64.Xiong Hcigang and Liu Gengnian, (1992) Measure-ment and consideration on talus in TianshanMountains. Geomorphological yrucesscs andenvironment management conlerencc papercollection, in press.

. .Zhu Cheng. (1989) The characteristics of thesurface movement of the lobate shape rockglacier from the Debris fabric, Tfanshan,China, Journal of Glaciology &.Geocryology,V01.11, No.1, pp 82-88.1091

LONG-TERM SHEAR STRENGTH OF FROST-THAW TRANSIT ZONEDing Jingkang', Xu Xueyan2 and Lou' Anjing'State Key Laboratory of Frozen Soil Engineering, LIGG, AS, China'Northeast Institute Chinese Academy of Railway Sciences*Harbin Building Engineering InstituteThe frost-thaw transit zone is a complicated soil zone with a certain thickness,a beginning freezing temperature, frozen and thawed soil existing together,obvious rheological characteristics and a long term shear strength. The 10118term shear strength equation is obtained by Using several groups of relaxationcurves of the shear strength in the frost-thaw transit zoner The ability toresist sheqr stress is mainly due to the cohesive strength and the inner grindangle being quite small, The long term shear strength is about one-sixth, longterm cohesive strength is about one-fifteenth and the long term internal frictionangle is about one-twelfth when compared with their different instant values.INTRODUCTIONThe thaw of permafrost in foundations resultsin the strength of Eoundation soil decreasinggreatly. So the stability of foundations is aproblem that should be considered above all whenbuildings are designed according to the pri.ncipleof allowing the permafrost foundations tothaw, The strength of melting soil is small andthe temperature of the soil layer with a c'ertainthickness near the thawing front maintains thebeginning to freeze temperature. The soil zoneis high in water content and temperature and isthe transit zone of melted and frozen soil andis naturally a weak layer in the foundation offrozen soil. It is significant in the design ofslope foundations in permafrost regions and ininvestigating the mechanical. properties of thetransit zone (S. Goto and K. Minegisgi, 1988:B.J.A. Stukert and L.J. ?lahar, 1983).The frozen soil in the transit zone haveobviously rehological properties. This kind oftransit zone, with melted soil in low temperatureand frozen soil of high temperature commonlyexisting, is a more complex zone of soilcorresponding to the melted and frozen soil. Tounderstand the characteristics of the transitzone and to provide a basis for building designon slopes in permafrost regions, research wascarried out on the long term shear strength ofthe frost-thaw transit zone, and the results arepresented as follows.TESTING METHODSUsing different shear load with the samevertical load, a set of shear creep curves canbe obtained. Then, the long term strengthequation and long term strength can be obtainedand calculated with different vertical loads.Meanwhi1e;the long term strength curves andlong term value C, @ in the soil of the frostthawtransit zone can be obtained.Apparatus being used in the shear strengthtest is a large scale, sheet creep apparatus,Test specimens were subclay taken from the FenhuoMountain region of the Qingahang Plateau.The particle composition is that >0.05 mm takesup 37.7 percent, 0,05-0.005 mm takes up 47.7percent, CO.035 mm takes up 14.9 percent. Thephysical properties of the teat specimen areshown in Table 1.Table 1. Physical properties of the test specimenTemperature ofW(X) WL WP IP beginningfreeze ("C)to23.4 28.6 16.1 12.5 -0.1The specimen used in the test was remoldedsoil, with the density being controlled by thedesign dema:d of the earth filled foundation,r=19.9 KN/~ , rd-16 KN/~. The specimen was acylinder 9.5 cm in height and 11.28 cm indiameter. The finished specimen was placed intoa low temperature cabinet to freeze for 24 hours.For the purpose of controlling the positions ofthe frost-thaw transit zone, two pieces ofthermoelectric cell were put into the shearzone of the specimen,The test of long term shear sttength is ashear test with constant load, consolidation anddrainage. The rest procedures were as follows:1) ForminR and Controlling the Frost-thawTransit Zone2) Addin Vertical LoadWhen :he temperature of the shear zone is-1.5"C. half of-the vertical load is added andwhen temperature of the shear zone is -1.O"C.the whole vertical load is added.. 1092

3) Consolidation Time: 2.5 Hours4) ShearAfter the specimen is consolidated, the testcan begin. During the shear process the temperaturein the frost-thaw transit zone and shearload is kept stable.5) Failure CriterionWhen shear displacement reaches 30 mm, namely27 percent of the specimen diameter, we defineit as the limited failure.6) Grade of Added L3adFor the long term shear strength, the gradeof added load is shown in Table 2.HorizontalVerticalTable 2. Grade of added load(N)2800 2500 2300 2100 1900 1700 1500 13004000 J J J J J J J3000 J J J J J J2000 J J J J JThe grade of vertical load is 4000, 3300 and2000 N in the test of consolidation and quickshear for the frost-thaw transit zone and meltedsoil.TEST RESULTS1. The relationship between shear displacementand time. Under the action of applied load,the she,ar deformation obviously has the rheologiccharacteristics in the frost-thaw transit zone.When shear stress does not change, shear strainincreases with time. When the stress is smallerthan a certain value, shear strain appears as acharacteristic of attenuation. When the 'stressis over a certain value, shear strain shows 'thecharacteristic of non-attenuation as shown inFig.1, 2,, and 3. ,hf!et (mill.)Fig.2 Relationship curver between shear displacementand time when vertical load is P=3400N10or'1900, 170001 So0t (min.)Fig.3 Relationship curver between shear displacementand time when vertical load is P=4000NFig.4 shows that the strength of the frostthawtransit zone is located between the meltedand frozen soil, its cohesive strength is largerand internal friction angle is smaller whencompared with melted soil, and its cohesivestrength is smaller and internal friction angleis larger when compared with frozen soil, asshown in Table 3. This sufficiently indicatesthe transit properties of the soil in the zone.Testing temperature of the frozen soil below thetransit zone is about -2.5'C (the temperatureat the beginning of freezing is -0.1"C).19001300t (min.)Fig.1 Relationship curver between shear displacementand time when vertical load is P-2400NFigures 1, 2 and 3 show that the curves ofshear deformation in the frost-thaw transit zoneare teh typical creep curves of frozen soil.Under action of the vertical and horizontalload in each grade, the shear deformation offrost-thaw transit zone has the characteristicsof creep deformation. This indicates that thesoil in the frost-thaw transit zone mainly hasthe characteristics of frozen soil.Fig.4 Comparesion of shear strength2. Long term shear strength of soil in thefrost-thaw transit zone. As shown in the former,the soil in the frost-thaw transit zone chiefly1093

Table 3. Instantaneous value C, c$of tested soilType of specimen C(KPa) bFrozen soil under frostthawtransit zoneFrost-thaw transit zone 3007 ao 2629l/rI0.007 'u= 240 Kpa . ,m.Melted soil above frostthawtransit zone47320.005 -has the characteristics of frozen soil and thusexists as a long term strength.Fig.5 shows a set of relaxation curves ofshear strength in the soil frost-thaw transitzone obtained by Fig.1, 2, and 3.k 260O.OO30 1 2 3 'dFig.6 Lgt--l/-r curveTable 5. Long term shear strength offrost-thaw transit zoneIloo 200 300 400t (min.)Fig.5 Relaxation curves of shear strengthFrom Fig.5 it can be seen that the curveswith different vertical load are similar.Therefore the long term strength of the soilin the frost-thaw transit zone can be expressedwith the same equation with different parameters,that is:T =- RlgAtWhere A, B - parameters determined by thetest, which the unit of A is llminute, B is KPa,t - time, minute, T - shear strength, KPa.To obtain A and R, the curve of lgt-l/t isdrawn, as shown in Fig.6, where the slope oflines is 1/B. The calculated value is exhibitedin Table 4. The long term s:rength of !=50 and100 years are shown in Tablr 5.Table 4 , Parameters of the rquation for longterm strengthVertical400034002400(N) loadA R2955ah2216BOO800800Drawing curve of 0-T with the value inTable 5, long term value C, 4) of the frost-thawtransit zone can be obtained, as shown inTable 6.40034024080.978.574.3.78.576.372,3Table 6. Long term value of C, 9 in thefrost-thaw transit zoneLimit of C(KPa) year+50 642.410062 2.4From Table 6 , we can notice that the cohesivestrength decreases with the increase of time,but value of 4) does not vary basically and isvery small.3. Comparison of long term shear strengthwith instantaneous shear strength in the frostthawtransit zone.According to the boundary conditions of thetest for long term shear strength, an instantaneousshear test is carried out in the frost-.thaw transit zone. The speed of added load iscontrolled by time, that is the specimen issheared with the time of 40 to 60 seconds.Figure 7 shows the comparison of long termshear strength with instantaneous shear strengthin the frost-thaw transit zone. From Fig.7, itcan be understood that the long term shearstrength is much smaller than the instantaneousshear strength, with the former being one-sixthof the latter. For the values of C and $, thevalue of C of the former is approximately onefifthof the latter, the value of Q of theformeris about one-twelfth of the Latter. Thelong term and instantaneous value of C andin the frost-thaw'tfansit zone are.listed inTable 7.1094

~ work0 200 400 600u &Pa)Fig.7 Cornpartson. of shear strengthTable 7. Comparison of value C and 1 in thefrost-thaw transit zoneTime C(KPa) 1Long (50 years) 64 2.4Instantaneoustime (1 minute)300 29 .O4. Temperature variance of the shear zoneduring the shear process. It was found by thetest that the temperature of the .soil In theshear zone increases. This indicates that thedone by the exterior force, except forthe shear deformation, partly converts intoheat. The conversion of energy results in alarger affect on the mechanical characteristicsof the soil in the frost-thaw transit zone andfrozen soil in high temperature. ,In Table 8lists the variance of temperature of the centerand the border of the specimen in the fast sheartest.Table 8. Variance of soil temperature duringthe process of fast shearing (C)Type of,specimenFrozen soil undertransit zoneThe soil in frostthaw transit zonePositionCenter Roundarv0.345 0.2540.166 0Melted soil abovefrost thaw transit 0.110 0.102 .zone. .RemainThevaluesin thiscable ...all areaverageThis is because more heat in the border of theshear zone is lost. During the process of shearing,the environment temperature is controlledat -2°C and heat produced by shearing radiatedthrough the shear box. Naturally the loss ofheat at the border is more and the increase oftemperature is smaller. The rise of soil tempera-.ture in the shear zone has a certain relationwith the appearance of the peak strength of shearstress, in which the temperature of soil risesonly after a certain amount of displacement isproduced. In frozen soil, in general, when sheardisplacement reaches 9 percent of the specimensdiameter, thb temperature of the soil willincrease, but in melted soil and soil in thetransit zone this value is 6 percent. Followingclosely with the rise of soil temperature, thepeak value of shear stress appears. After thatthe temperature of soil increases continuously,but the shear stress decreases gradually untilthe specimen is sheared.CONCLUSIONS1. In clay, the soil in the frost-thaw transitzone i8 a mixture of frozen soil at O°C andmelted soil with obvious rheological characteristics,and its long term shear strength can becalculated with the equation of long termstrength presented in this thesis.2. In clay, the ability of resisting shearstress of the soil in the frost transit zone ismainly composed of cohesive strength since the 'internal friction angle is very small.3. In the frost-thaw transit zone itsinstantaneous shear strength is larger than theshear strengch of melted soil. When the stabilityis calculated for the slope in permafrostareas. the indexs C and @ should be determinedaccording to the length of time that load willbe added.REFERENCESB.Y.A. Stukert and L.J. Mahar, (1983) Proceedings,4th <strong>International</strong> <strong>Conference</strong> onPermafrost.S, Goto and K. Minegisii, (1988) Direct sheartest at a frozen/unfrozen interface,- Proceedings. 5th <strong>International</strong> Symposiumon Ground Freezing..a .From Table 8 It can be seen that during theprocess of fast shearing the variance of soiltemperature is the largest in frozen soil,second in the transit zone and smallest in themelted soil, It also can be understood fromTable 8 that the variance of soil temperaturein the center is larger than in the border.1095

THE COMPRESSING PROPERnBS AND SALT HEAVINGMECHANISM STUDY OR SULPHATE SALTY SOILFei Xucliang'fLi Bin''The State Key Laboratory of Frozen Soil EngineeringLIGG, Chinese Academy of Sciences, China'Xian Highway and Transportation Wniversity,Xian;ChinaIn the paper, a set of salt heaving laws of sulphate salty soil which is made by adding 3% Na,SO, inXian-loess in different dry density is obtained through simulated test. It demonstrates that when the.temperature is decresed from +25"C to -15T and the initial water content is cehain, the relation betweenthe salt heaving value of sulphate salty soil and initial dry density presents a type of parabolic. Italso exists a minimum value interval. In this test condition, this interval is within the range of1.80g/cm3 when water content is the best one of the heavier compaction standard test. It is quite importantto discover the interval for saline soil areas of highway engineering. In addition, the analysis oftest result reveals the mechanism of salt heaving in sulphate salty soil under different initial dry density.It shows why sulphate salty soil heaving ratio changes with the initial dry density.INTRODUCTIONThere is a large area of sulphate salty soil in China. In thewareas, there arc some peculiarity in highway engineering. First,when sulphate salty soil meets water, its strength is degraded fastand occurs deformation easily when vehicles go through. Secondly,sulphate salty soil can create salt heaving when temperature lowering.It leads to the dry density of the base and subgrade d&grading and strength reducing. The parts of salt heaving influencehighway's using quality greatly. These two main drawbacks dccreaseusable years of highway and cause much loss in finance.* The purpose of this study is, by simulated test in laboratory, toobtain the relation of salt heaving ratio to initial dry density(expreeed with rd) of sulphate salty' soil subgrade, then proposing arange of initial dry density for engineering. In the proposed range,thc salt heaving can be prevented from creating. And by analyzingtest result, the salt heaving mechanism has been concluded why saltheaving changes when initial dty density changes.TEST METHOD " AND MAIN PARAMETERS~ ~-~Test MethodSulphate salty soil samples used are made by adding 3%Na,SO, in Xian-loess, in order to obtain the salt heaving valueand salt heaving laws ofsulphate salty soil in quantity when free*ing. Formed temperature of the samples is +2OoC; formed methodis based on Highway Civil Engineering Test Rules (JTJO51-85 ofChina), Basic test plan is showed below.- Main Parameters ." of SoilMain parameters of soil are showed below.Test Conditions"There is a insulating material which is 15 cm in thick aroundthe model. Freezing of sample is controlled by Program Tempera.ture Controller and Cooling Bath, freezing rate is 0.5'C/ hour.Temperature range is from +25"C to -15°C.Size of sample is 8 cm in height and 10 cm in diameter. Temperatureand deformation of sample are detected by thermal-coupdensitywatercontentTable 1 Test design table1.2 1.4 1.6 1.7 1.8 1.89 1.9410% d J J13.5% J d J 4'15.5% d d d d d d J17% J d d dNote: J is indicated which sample is tested in freezing; unit ofdry density is g/ cm3.le8 and displacement sensors. All data is collected by Data Taker.TEST RESULT AND ANALYSISResultBy simulated test, S-T curves (salt heaving value changingwith temperature) which deformation ir- :hanged while cooling areobtained; as shown in Fig.] ,2,3,4.Tab.3 is the final salt heaving value taken from Fig.1,2,3,4,under every water content and initial dq density.

plain mil 2.647 36.1 18.2 18.5 1.89 14.24sulphateaalty soil33.86 15.65 18.21 1 .a9 15.51Q rcU I I/ 2Tempcrature ("C)Temperature ("C)Fig. 1 Salt heaving value vs. temperature (initial water content Fig.3 Salt heaving value w. temperature (initid water contentis 10%) (initialdrydcnsity(g/cm3)is: 1"1.3,2"1.65,3--1:5) is 15.5%) (initial dry density (g/*n') is: 1-1.89, 2-1.54,3-1.2, 4-1.6,5"1.7 6-1.8)15 > 5 -5 -15Temperature ("C)1097

~ .Table 3. Final salt heaving value (mm)watercontentdrydensity10% 13.5% 15.5% 17%1.2 2.982 7.088 5.109610H,O and separates from solution existed among soil grains. Saltheaving of samples increases faster in the S-T curve when temperatureis between +25T and -5OC. And the temperature range where1.3 4.3781.4 ' 2.156 5.6 4.33361.5 1.51 *1.6 1.81 4.53341.65 1.842 *1.7 4.362 3.5865 3.5881.8 2.73981.87 4.34741.89 8.933I .94 7.044* : initial dry density (g / cm3) of samples has been adjusted.According to S-T curves (Fig,l,2,3,4), one could know that:no matter which initial water cantent and initial dry density ischosen, the salt heaving increases constantly while cooling. Frominspecting of all the curves, one can discover that salt heaving is increasingquickly in range of from +lS°C to -5"C, increasing slow inrange of from +2SoC to +iST and -5'C to -15OC. Especially,when temperature is in range of from -5°C to -15"C, salt heavingin many samples increases a little or does not increase.From Fig.5 one can see: under same initial water content, withinitial dry density increasing, salt heaving ratio decreases when initialdry density docs not surpass a certain point and increases wheninitial dry density is over that point.Mechanism Analysis of Salt HeavingDuring cooling, expending of sulphate salty soil is mainly d&pended on crystallizing of NalS04 - IOH,O. Its crystallizing equationis :. "_Na,SO,+IOH,O==Na,SO, - 10H,OAfter crystallizing, Na,SO,becomes Na,SO, IOH,O, and value ofNa,SO, * 10H,O is 3.1 times over that of Na,SO,; sometimesNa,SO, - 10H,O crystal in soil can cause more expansion, which isdepended on the position that crystal lies among soil grains.Solubility of Na,SO, is affected greatly by temperature.When cooling (higher than -5°C). lot of Na$O,bccomes Na,S04 -(I)Initial dry density (B/ cm')Pig5 Salt heaving ratio vs. initial dry density(initial water content (%) is: 1"10.0,2--13.5, 3"15.5,4--17.0)salt heaving increases fastest is between +15OC and -5'C. It is because that Na,S04 10H,O fills into gaps among soil grains ormakes soil grains displace when temperature is higher than +lS°Cand causes samples expanding when tempcraturc is lower than thatvalue.It should be pointed that: during cooling, as temperature ofsample being negative (

Photo. 1. Test condition is :rd= I.dg/ cm', Photo. 2. Test condition is : rd= I.8g/cm3, Photo. 3. Test condition is :magnified I500 rimes magnifid 6500 times rd = 1.89g / cm', magnified 4000 timcsIt brings about solution in adjacent little or a little and crystallizingis the same. At the same time, interaction between soil grains restrains salt heaving also. Therefore salt heaving value is smallerthen.3). When rd over than critical range, no matter whether poresare filled up by solution or not, Na,SO, crystal crstallizcs from solutionin pore first with lowering of temperature. If pores are notfull of solution, crystallizing of crystal will make it fill up soon.Then crystal crystallizes continuously and cau8es the volume ofcrystal as well as solution increasing incessantly. Since pores are filledup, increasing of volume will cause pores bulk and lead to oc.curing salt heaving of sample.After crystal crystallizing from solution in pore, crystal beginsto crystallize from solution combined weeker to soil grains and adjacent(then distance of soil grains has been expanded) with furthercooling. Its principle of.salt heaving is similar to that of above. Itshould be noticed that a part of crystal makes progress from portto adjacent following farther cooling (Photograph 3). It intensifiesdegree of salt heaving. Though rd is very large and moreinteractions between soil grain then, these forces can not resist theaction of salt heaving force and result in salt heaving ratio rising. Ifpores in Soil have been filled up before cooling, its salt heavingprinciple is similar to above, the different is crosses the process thatcrystal needs to fill pore up.CONCLUSIONIn simulating test, salt heaving value versus tcrnperature infrozen proceeding, S-T curves show that the quickest increasingsalt heaving value range of sulphate salty soil is from +1S0C to-5'C. Salt heaving ratio will be a little when temperature is higheror lower than that range. As temperature is lower than -5'C, espelsially, salt heaving alrnostly does not be occured. Total bulk of icevolume which is crystallized by water and Na,SO, crystal causedby rising concentration because of ice crystallizing in the lowertemperature is very small.As the other advantages, water content,salt content and coolingrate is same, the rclatio between initial dry density and the finalsalt heaving ratio of samples is approximately parabolic, with aminimum salt heaving ratio interval. For sulphate salty Soil made , 'by $ding 3% NafiO, in Xian-loess, the lowest value is in therange of 1.80g / cm3 of initial dry density when water Content isthe best water content (this dry density is about 95% of the best drydensity).By analysis of test results, the mechanism of salt heaving hasbccn revealed, that shows why salt heaving ratio of sulphate saltymil changes with initial dry density changing. When rd is smallerthan critical value, principle of salt heaving reducing with the increasingof rd is that: position of solution in soil changes, from adjacentto pore, and so docs the crystallizing, thus bulk action ofcrystal in soil is reduced. The increasing restraint forcc in soil restrainsthis action again. As a result, salt heaving ratio decreaseswith increasing of r,, in macroscopic. As r, being critical valuerange, generally, crystal crystallizes in pore only, it is little or a littlecrystal crystallizing in adjacent. and restraint forcc is bigger .than.So salt heaving ratio becomes smaller. When r., is over that. range,most part of pore is filled up with the solution, the principle of saltheaving changes a lot with the decreasing of temperature. Crystaloccurs in pore first, it makes the POFC bulk and part of solution inpore transfers to adjacent. Further cooling will get the crystalcrystallizcd in adjacent, volume of sample bulks again. A part ofcrystal in pore, at the same time, goes ahead to adjacent, salt :Ieavingvalue increased once more. These cause salt heaving ratio increasinggreatly with adding of rd.The author wish to express his appreciation for the financialsupport of the State Key Laboratory of Frozen Soil Engineering,LICG AS, China. Thanks also given to the staffs of the State KeyLaboratory of Frozen Soil Engineering for their help.

A.C. Fowler mcl C.G. NoonMathematical Institute, Oxford UniversityW.R. KrantzDepartment of Chemical Engineering, University of ColoradoFollowing ptwious work by Holden and co-workers, we have carried out an asymptotic reduction forthe O'Neill-Miller model of frost, heave, which is a quantitatively accurate representation of the originalmodel, but which only involves the solution of two first order ordinary differential equations. Thisreduction call be made on the Imsis that the frozen fringe is thin, conduction of heat dominates advectionof heat,, a.nd (import,ant,ly) thc permeability of the frozen fringe is strongly sensitive to the water content.Solutions of t,lw reduced model give predictions of heave versus time and of time of formation andthickness of discrete ice lenses, which agree with O'Neill and Miller's numerical results. Furthermore,the results can be interpreted in accord with the idea that clays heave large loads, but slowly; silts heavesmall loads, hut fast; while sands may not heave at all. An explicit heaving rate parameter can bedetermined, whose size depends on the soil type through its permeability, characteristic suction, and onthe applied load. The method of reduction can he applied to generalised versions of the O'Neill/Millermodel, and ill particular we have ca.rried 0~11, similar analyses in respect of compressible soils, saline soils,and we have also initiated work on the extension of the model to allow for differential frost heave.1 BACKGROUNDFrost heaving occurs when moist soil, particularly clavor silt, is frozen from the surface downwards (fig. 1). 11is manifested by an upward heaving of the surface,which is due not so much to the expansion of watkr 011freezing, as to an upwards suction of the subsurfacegroundwater towards the freezing front. This suction isthought to be due to various effects which contri1)ute LOcapillary-like forces; these include surface tensionbetween ice and water phases, a chemical potelltial dueto an electrical double layer on clay particles, anddisjoining pressure efiects due to short range forces inthin water films. A general description of frost he;lviugis given in the article by Miller (1980).As the water is drawn upwards to the freezing frollt,.it for,ms a series of discrete ice lenses in the soil. 11 isthe formation and growth of these ice Ienses whicllcauses the heaving at the surface, which can besubstantial. Frost heave is a natural phenomenol-l whichoccurs in most northerly regions. It is responsible forwidespread damage to roads and pavements (even inEngland, where the construction of the M 1 in 1959 wasf 1100

Region 2, frozen fringeH = qI0.1 0.2 0.3 0.4 0.5Figure 1.: Schematic picture of a heaving soil2 THE MILLER MODELMiller (1972) distinguishes between primary heave alldsecondary heave. In the latter, a thin paltially froze11fringe of ice exists below the lowest ice lens. This regionof soil contains both pore ice and pore water, and isconsidered to be in thermodynamic equilibrium.Moreover, the freezing temperature TI within tllr Frillgris given by a generalisation of the Clapeyroll relatio~.such that TI is related to the pore pressure exertd I.)yeach of the ice and water phases; this relation ca.11. hetheoretically derived (Loch 1978).In allowing separate ice and water pressures pi andpw, a constitutive relation between them n-lust beprescribed. Miller (1980) takes this ill the formpi - p, = f(W), where W is the volumetric waterfraction in the fringe (fig. 2). As mentioned, t,llis call heascribed to surface tension as well as ot1tc.r elk:t,s,although the physical interpretation of. what this 11ttw1sfor ice-water re-distribution is less clear. In fact, Millvr's'rigid ice' concept assumes that the pore ice is rigidlyconnected to the lowest ice lens, and in order for lleavcto take place due to water freezing on to the IowesL Irns.this requires the rigid ice to move past the equally'rigid' (i.e. undeformable, in the model) soil particles.This is accornplished by the process of tllernlal. 1101

3 MATHEMATICALFORMULATIONThe Miller model was implemented mathematically byO’Neill and MilUer (1982,1985). They treat the frozen,unfrozen and partially frozen fringe w three separatecontinua, and formulate conservation laws of heat andmass transport in each region. Because of the finespatial and temporal scale of the model, it is, as Black(1991) says, “hardly ready for use on practicalproblems. Computational difficulties are unusuallyformidable ...”. Both for this reason, and also’ because,notwithstanding its complexity, the Miller model isconceptually too simple (it considers soil to be rigid alldsaturated, pore water to be chemically pure; and isessentially limited bo-one-dimensional heaving), O’Nrilland Miller’s formulation needs simplification. The wayto do this lies in making judicious approximations,which simplify the model without surrenderingaccuracy. Initial work by Holden (1983), Holden el al.(1985) and Piper e2 ai. (1988) has been extended byFowler (1989) and more recently Fowler and Iira~~tz(1993), using four basic approximations: heat transportby advection is small, gravitational effects ar? small, t.befrozen fringe is thin (due to the fact that the freezingtemperature variation in the Clapeyron relation issmall), and the permeability variation with waterfraction is large (specifically k a WY, with -y (x 9)(O’Neill and Miller 1985) being large).These four approximations allow one to reduce thecalculation of frost heave in the O’NeiH/Miller Inodcl tothat of solving two simple first order differentialequations for the freezing front position and the soilsurface elevation. We now detail the effect, of thesesimplifications.Firstly, the fringe is thin compared to the length ofthe soil column, and is consequently seen as a surfaceby the rest of the soil. Secondly, convective transport ofheat is small compared with the conductive flux (i.e.the Peclet number, Pe < 1). The energy equat,ionsoutside the ‘fringe then reduce to:J - 2T = f -(Tb - To)-,f - (3)*tbin zf > z > zs.Using these expressiorls to evaluate the temperaturegradient on each side of the fringe, the fringe equationsmay be solved to tieternline the locations of the movingboundaries z, and 2,. This c m he ~ done using the factthat the effects of gravity are negligible, and theexponent in the permeability function is large. This factgives rise to a bourdary layer below the lowest ice lensin which the wat,er pressure c1lalrge.s rapidly.Asymptotic methods applied to Barcy’s equation leadto a uniformly valid espressiolr for LIE water pressure asa function of water contellt. On the formation of a newlens, water conterrb within t,lw boundary layer relaxesrapidly to’a steady sI,iiI,t’ illltl SO I,IW water pressure canbe considered t,o l)c quasi-st(.al.ic‘.The remaining diC[iwnbial equations in the fringeexpress conservatiuir of IMSS ant1 energy. The smallPeclet number anti thirr rringc allow these equations tobe written as a sct of algcl)raic quations for variables Ievaluated at each aitl(, or ~h Irillgc. Using the formsfound for the ta~Ip(~raI~tIrv a1rt1 prtwure profiles and alsothe remaining frilly(: rrlat.iolls, it is possible to reducethese algebraic q u a 1 iolls I,O ~lle 1.w0 (dinlensional)orchary differenlial rquatiorrs:1:, =-- +--.13(4)4 RESULTSZS - 2, ”1 - :b(5) ,in 2, > z > %f and zj > z > Zb with boundaryconditions T = T, on z z,, T = To on : 2 25 andT = Tb on z = za. The solution to this eqbation ill OIICdimension is then’ 1102

heave (mm.)0.40.3-.-.---.-P-20 kPaP-30 kPaP-40 kPb/there is an asymptot,e tit H finik ,value of N = nT,,beyond which Iwaviltg i n this ~nodt! does not occur,since no lens forrIl;l1icm is pussiblc. On the basis of fig.5, it is possible IO ulldws1.wd IJIP absence of heave insands, on the bais ha1 No > N,., while if NO < N,,then one expects heave LO occur, but not for arbitrarilylarge loads. These observatiwis are qualitatively similarto reported heaving characteristics, but furtherquantitative validatory work is necessary.0.20.11 2 3 4time (le seconds)Figure 3: Heave against time for different overburdensFigure 4: Heave paranwt,rr cy versus effective overburdenN (log-hear st:i1le), clay-type soilpermeability and the characteristic function f(W) infig. 2. LY is proportional to (saturated) permeability,which explains (obviously) why clays should heave moreslowly. than siltg.The dependence of a on load N is more complirat.rrl.Firstly, since N is an effective pressure, zero lo;tdcorresponds to a positive value of N = No = pa - yw,where pa is atmospheric pressure, and pm is the far fieldgroundwater pressure. In fig. 4, we plot the variation ofa with N for values N - O(1) bar, for a characteristicrelation f 0: (1 - S)P/S9, where S is the saturation. aridp = 0.3, q = 0.3. This choice of p and q gives a steeplyincreasing suction as S is reduced, and .may beappropriate for finer soils. Depending on the size of No,we see that heave decreases more or less exponentiallywith increasing load, but is maintained for arbitrarilylarge load. Fig. 5, on the other hand, shows cy versus Nfor p = 0.3, q = 0.1, the lower value correspording tolower suctions at low S, and thus a coarser soil. Here1NFigure 5: HKLW paratllrtvr LI versus effective overburdenN (log-linear scal~), mrd/silt soilI

5 CONCLUSIONS 9. Kooplnans,It.\Z'.IC. atld lt.13. h'liller 1966 Soilfreezing a dO'Neill and Miller's model wm grid locked because its Sci. Soc. AIII. l'roc. 30. (380-695.exhaordinary numerical complexity did not allow thepossibility of making any realistic extensions to it. Inparticular, it is conceptually wrong, insofar &s theassumption that the ice lens velocity vi is spatiallyconstant is (generally) inadmissible for athree-dimensional process. Our previous work not onlyrenders the model tractable, it allows the possibility ofextensions by adding new physics to the simplifieclmodel, and thus obtaining generalisations of this simplemodel. As an example, frost heave of saline soils lcadsto a similar reduction as for pure pore water, thedifference lying in'that the coefficients in the result,ingeiuations are more complicated.The application of the model simplificatio~ls out.lilletlhere to saline and compressible soils is outlined in t . 1 ~thesis by Noon (1993), and further developents toinclude unsaturated soils, as well as three-dinre~~sio~~;lmodelling of differential frost heave, are in progress.6 REFERENCES1. Beskow, G. 1935 Soil freezing and frost heavingwith special application to roads and railroatlx.Swed. Geol. SOC. C, no. 375, Year Book No.3.Reprinted in Black and Hardenberg (1991).2. Black, P.B. 1991 Historical perspect,ive of frost.heave research. In Black and Hardeuberg (lWl),pp.3-7.3. Black, P.B. and M.J. Hardenberg 1991 Historicalpewpectives in frost heave research. The earlyworks of S. Taber and G. Beskow. CRREL spwiitlreport 91-23. Hanover, NB.4. Fowler, A.C. 1989 Secondary frost heave ill f~~rrzi~~gsoils. SIAM J. Appl. Math. 49, 091-1008. '5. Fowler, A.C. and W.B. Krantz 1993 Generalizrdsecondary frost heave model. SIAM J. Appl.Math., submitted.6. Fowler, A.C. and C.G. Noon 1993 A si1t1plili(dnumerical solution of the Miller Inodel of suco~~rl;~~frost, heave. Cold Reg. Sci. Teclurol., in press.7. Hioklen, J.T. 1983 Approximate solutions forMiiler's theory of secondary heave. Proc. FOII~I,IIInt. Conf. Permafrost, Fairbanks, Alaqka,pp.498-503.8. IIoldcn, J.T., Piper, D. and R.H. Jones 198.5 SnlI-I(bdevelopments of a rigid ice model of frost. hravr..Proc. Fourth Int. Symp. on Ground Freczing,Sapporo, Japan, pp.93-99.soil-wnt.er clIaract.crist,ic curves. Soil

GEOCRYOLOGY IN MT. TIANSHANA.P.GorbunovKazakhstan Alpine Permafrost Laboratory of The PermafrostInstitute of Siberian Branch, Russian Academy of SciencesThe total area of pcrmafrost region in Tianshan is 180-200 thousands Km’, which the masses ofcryolithozone -- 90-100 thousands Km2 in total area and 16 thousands Km3in total volume. Usually,the alpine permafrost is located above 2700-3300 m a.s.l., but in some special cases it can occur at 2000m a.s.1. or lower. The lower limit of continuous permafrost lies at 3500-3700 m a.s.1.; there might not befrozen ground under some large glaciers.The permafrost in Tianshan started fornling no later than Pliocene. The rock glaciers are commonseen in Tianshan and can be divided into two categories: the near-glacier and the near-slope area.Thefbrmcr are transformed from moraine material and the latter-from talus or rock-avalanche bodies.Tianshan, as an unitary geocryological region, occupies thearea in 40-46”N. latitude and 67-95’E. longitude. The mountainregions, distributing in the tempcrate zone, elongate along thebaundary between thc temperate and subtropic zones. This determinessome of thc characteristics of the geocryological region.The major ridges and ranges that constitute this mountain systemare: Junggar Alatan, Borohoro, Yirengabirtag, Bogdashan,Karliktag, Xeliktan, Kakshaaltoo, Maigstag, Fergan, Talas andtheir offshoots. Some scientists considered the Mt. Junggar Alatanas an independent mountain, but, geocryolopically, it should be includedinto the Tianshan system.The first infbrmation of the cryogenic phenomena in Eastern(Middle) Tianshan occurred in the paper written by V.I.Roborovski, who visited the Youldos Basin in 1893. He decribedthe frost cracks and frost mounds. The first information ofprmaliost in Tianshan was reparted by the pedologist A.I.Rezsonov, who visited the AK-Sai Basin in Inner Tianshan in1913. However, the systematic and purposeful research onpermafrost and other cryogenic phenomena in Tianshan started onlyat 50’s and 60’s of this century.Nowadays, the area of permafrost regions estimated by us is180-200 thousands Km’, while the masses of cryotithozone -about 90-100 thousands KmZ in total area and no less than 16thousands Km3 in total volume. The following pattern could bexcn in Tianshan: the permafrost area is 10 times of the modernglaciers (about 18 thousands Km2) and two times of the total areaof masses of cryclithozone. This pattern is very localized, it depcndson the helgnt of mountain and climatic factors dependentupon the geographical position of the region.If the volumetric ice content is 1 % of the whole cryolithozone,then the total volume of ground ice ( excluding the buried glacialice) in cryolithozone should be 160 Km’.The permafrost zone could be divided.into three’subzones, i.e.,[he island ( cxtremely unstable), discontinuous (unstable) and continuous(stable) oncs.The altitudinal limit of permafrost distribution, dependent up-On the geographical latitude, local climate and lithological charaGteristics ( thickness of’ snow cover, composition of loose sediments,c~c), might be at 2500-3600 m a.s.1. with a negative mean annualair tempc:ature and above the upper limit of forest belt. However,under a special microclimatic condition, for example under the coverof coniferous trees, the small permafrost masses could be formedunder a positive mean annual air temperature, even down to the altitudeof about 2090m. Such kind of permafrost masses were fountin Northern Tianshan, Inner Tianshan and Eastern Tianshan,down to the latitude of 42”N. Thus, the Tianshan might be thesouthmost mountain systemin the Northern Ramispherc that thepermafrost could be found in forest zone. This should be a noticeableproblem in forestry.The lower limit of the continuous or stable permafrostsubzone varice from 3500 to 3700 m a.s.I., there might not bepermafrost undcr the large glaciers, even at a considerable high absoluteelevation. Comparing the altitudinal position of snowlineand the lower limit of continuous permafrost in Tianshan, someimportant conclusions could be made:The distance between snow line and the lower limit of continuouspermafrost was named the continentality index of permafrostsubzone (Gorbunov,1976). Later, S.A. Harris suggested thecontincntality as the characteristic index, made it clearly in definitionand used it to evaluate the geocryological condition in themountains of Northern America (Oral communication). In the paper,it is called the index of geocryological zone ( AA ).The above indices determine the cryogenic state of the endedmoraine of modern glacier.The index in Western Tianshan shows that many endedmoraines are in the discontinuous and island permafrost subzones.

, - T ITable 1. index of geocryological zone in TianshanInner Region TianshanAA 0.4-1.5 (Km)such an altitudc, the wind has no a determinant cffect on snow cov.cr as seen at high elevation, so that it can carry out its effect of protectingthe soil against cold. All of these result in the occurrence ofpcrmafrost masses down to the elevation as low as 1800 m a.s.1. inthe region with little snow, while in the region with heavy snow,Northern Tianshan 0.3-0.4Eastern Tianshan 0.4-1.5there would be no permafrost even at 2700-2800 m. The peculiarityof landform will aggravate the difference of geocryological condi-Junggar Alatan 0.2-0.5tion in different altitude. Below 3500-3700 m in Tianshan, thereare many intermount depressions characterized by the plainlandform. There, the calm weather is dominant in winter, where thesnow cover, if it is formed, will not be distroyed by wind. At theThus, it could be said that the moraines might be completely orsections in mountain valleys with a hardly flowing cold air (partially in an unfrozcn state. The formative condition of glacialUrumqi river valley in the area of the Chinese Tianshanmud flow there would considerably differ from that morainesGlaciological Station) or in the depressions of Sysumir, Arpin,bound by permafrost. Such a condition is dominant in most partsAK-can, Great and Small Uldous, the anormal cold winter couldof the Northcrn Tianshan, Inner Tianshan, Eastern Tianshan andbe encounter, At those depressions with continuous and thick snowThe Mt. Junggar Alatan. In these rcgions, the large glaciers runcover, permafrost is absent, even at an elevation as high as 3000 m,down to the discontinuous or island permafrost zones, the endedbut at the places without thick and continuous snow cover, themoraincs might be unfrozen or only a small part of the endedfrozen ground could be seen at relatively low altitude.moraines could be subject to the perennially freezing. So, theseIn the intcrmount depressions higher than 3500 m, becausemoraines arc distineguishcd in their poor stability and worse prestheyarc inside the mountain area (Chatir-Kelskaya,ervation in comparison with those moraines bound by permafrost.Kek-Ala-Chan, Kym-Terskyi Sirt), the trifling snow cover has ncThe frozen and unfrozen moraines also differ significantly fromany influence on the altitudinal position of the lower limit of theeach other in the internal structure. The former is characteried bythe existence of intermoraine cavities filled with the considerablecontinuous permafrost subzone, even at the plain condition.amount of melt water. Their burst is usually the origin of the greatThcre is another circumstance that forms the distinctive feaglacialmud flow.tures of perennial frost in the lower and upper gcocryologicalThere is another point to be noticed. On the surface of thesubzones of Tianshan i.e. the alpine landforms dominant in thefrozen moraine, the condition is far more favourable to take formcontinuouspermafrost subzone.ing, because practicallySince the predominance of steep bedrock slope, the snow coverno water could infiltrate through the frozenloose-block sediments, this is quite different from the unfrozenhas no effect of protecting the ground against cold.moraine.All of these, as mentioned above. also determine the stabilityGeocryological data in Tainshan should that, at a same latiofthe boundary of the continuous permafrost subzone and thetude, the elevation of lower limit of the continuous permafrostunstability of the boundary of the lower geocryological subzones.didn't change significantly from placc to placc. The snowline, how- Thermal measurement in Tianshan demonstrated that the seauver,displaces very obviously with the position of the mountain sonal fluctuation of ground temperature could penetrate down to aranges. Even at a same geographical latitude in Tianshan, thedepth of 20-30 m in bedrock. The minimum value of temperaturesnowline varies from 3500 to 5000 m, while the lower limit of conwasrecorded to be -4.Y°C in bedrock (420Om a.s.I., 42'N latitude)tinuous permafrost subzone -- only within 3500-3700 m a.s.I., bc- and -6.S'C in moraine (4050 m, 42'N latitude). The maximum valcausethe snowline is sensitive to the total amount and regime of . uc of thickness of cryolithozone was observed to be 270 m (4160matmosphere precipitation, while the limit of the permafrost subzone a.s.l., 42'N latitude).is mainly dependent upon the thermal condition changing littleThe permafrost in Tianshan started forming no later thanfrom the periphery of Wes!ern Tianshan to the Inner Tianshan or Pliocene, before the first occurrenw of mountain glacier. Duringthe Eastern Tianshan.the period of Pliocene, Pleistocene and Holocene, the permafrostThe law decribed above doesn't deny the influence of precipi- area becamelargcr, especially in the ice ages of Quaternary. In Latetation on the thermal regime of ground.PIeistoccne, the lower limit of permafrost extended down to theThe thing is that, at B high elevation (higher than 3500-3700 hillfoot of Northern Tianshan at about lo00 m ad. Unlike the Mt.m), the atmosphere precipitation will reduce in winter, and the Alps, there is no any recorded unquestionable case to prove thestrong and constant wind will form the discontinuous and corn- degradation of permafrost in Tianshan, nowadays.pacted snow cover that could not protect the ground from freezingpractically.By the island permafrost subzone, the situation is different atthe altitudinal tang6 from the region with little snow to the rcgionSpecial attention should be paid to rock glaciers. According totheir origin and position, the rock glaciers can be divided into twotypes, i.e., the near-slope ones and the near-glacier ones. Accordingto their age and dynamics, rock glaciers can be divided into a0with heavy snow the geographical environment changes tive, inactive and ancient ones. The active rock glaciers in Tianshansignificantly. This is beduse of that in the region with heavy snow are no less than 4000 in number, or rather than 4000 far more.the permafrost could not form under a thick and continuous snowThe near-slope rock glaciers are transformed from talus andcover, while in the region with little snow, the absence or thinness rock - avalanche bodies near the slope feet, the near -glacier rockof snow cover could promote ,the perennial frgczing of soils. At glaciers - mainly from the mordine materal. The formation of the1106

near-slope glacicr is less dependent upon the evolution of' glacier,and the lalter, however, -- directly upon the ice advance and rctreat.When ice advance, it could be completely or partially destroyed.Thus, under the other conditions being equal in thoseplaces, where the glaciers distinguish themself in high mobility thenear-glacier rock glaciers will be seen less than that in those placeswhere :he glaciers arc morc stable. This should be noticed when therock glaciers are used as the paleogcographical indicator.

THE FREEZING AND FROST HEAVE REGULARITIES OF BASE SOIL .FOR ARBITRARY SLOPE DIRECTION AND GRADIENTGuo Dianxiang, Wei Zhengfeng and Ma YijunInstitute of Hydraulics, Shan 'Dong ProvinceAfter calculating the solar radiation energy for-arbitrary slope direction andgradient and analyzing the effect of udnerground water table on freezing depth,relationships of frost and frost heave with mean annual temperature, solarradiation energy and underground water table are set up. Mathematical models forpredicting freezing depth and frost heave of base soil are established forarbitrary slope direction and gradient. The calculation is checked by usingobserved data of the temperature, freezing depth and frost heave of the basesoil.I. CALCULATION OF DAILY SOLAR RADIATION FOR ,ARBITRARY SLOPE DIRECTION AND GRADIENT' In unit time, the solar energy flux projectingon unit area id defined as the solar radiation'intensity (Si, the unit is W/m'). The planeperpendicular to the solar-ray has the maximumradiation intensity, but, the plane parallel tothe solar-ray has the minimum radiation intensity.The solar radiation intensity of horizontalground is directly proportional to the sinus ofthe solar height (h). Considering the distancebetween the sun and the earth and the influenceof atmospheric transparence, the solar radiationintensity reaching the ground is:SSi = 7 Pm sinh (1-1)PWhere S is solar constant, about 1326 W/m':p is vector radius of the earth, in January, pais about 0.97; Pm is atmospheric transparence; his solar height.groundFig.1 Calculated figure of the solar radiationintensity of canal slopeWhen the slope angle of a. canal bed is a,from the triangular relation of Figure 1, weknow that the angle Y between the solar ray andcanal slope'can be written as Y=h + a. Therefore,when the slope direction is perpendicular to thesolar ray, the radiation intensity of the canalslope can.be written as:SSN 7 P, siny- 3 P, sin(h+a)P 4 P, (sinhcosr +.coshsina) (1-2)PThe solar radiation intensityvfor an arbitraryslope can be divided into the vertical andhorizontal quantity. The vertical quantity ofsolar radiation doesn't depend on any direction,its projection on arbitrary gradient slopeequals P,,sinh cosa, The horizontal quantityof solar radiation has to do with the varyingsolar azimuth (y), therefore, it has the strictslope directionality, and the projection of thehorizontal quantity on arbitrary gradient slopePm coshsinacos(y-6). At any time, ona diffeeent gradient slope, the solar radiationintensity can be expressed as follows:SSB.,~ = ;;" Pm(siny)= -% pm[ sinhcosa+coshsinacos(y-B)]Pwhere: y is the arbitrar.y solar height onarbitrary gradient slope:(1-3.sinh = sin8sih6 t cos@cos~cosw (3-4)sinycosscosw=cosh(1-5)cos~sin~cosw-cos~sin6cosy -cosh(1-6)Where f3 is slope direction angle, it equalszero in the north, clockwise rotation is 360"G;y, w are respectively ,solar azimuth and hourangle, at high noon, all of them are zero, inthe afternoon positive, in the morning negative:3 is gradientangle: @ is latitude; 6 is equatoriallatitude, when the sun is to the south ofV1108

Ithe equator, 6 is negative, to the north, it ispositive.Putting (1-4), (1-51, (1-6) into (1-3) we canobtain:Spras-r Pm sinyb7 S ~~[sin~sin~cosa-cos@sinssin~cosSP+cos~cos6cosucoswtsin$cos~~inacos~oos~+cos6sin~sinBsintul (1-7)From sunrise ( ~ 1 ) to sunset (WZ), we Integrate(1-7), this is daily solar radiation:Y'sinhdw(1-16)WlOn the .horizontal ground, the sunrise and.sunset hour angle can be calculated as follows:The solar radiations of the arbitrary slopedirection an$ gradient and horizontal groundhave a relation as follows:(1-18)+si~~cos~sinocos~cosw+cosssinusinasinw)d~ Result are listed in Table 1.(1-8)Where,wl, w2 are sunrise and sunset hourIT. CALCULATION OF FROZEN SOIL DEPTH FORangles. The calculation method of sunrise andARBITRARY POSITION ON THE SLOPE OF ARBITRARYsunset hour angle is the following: in theSLOPE DIRECTION ANTI GRADIENTformula (1-7), let siny=O. we can obtain:1. Calculation of Geothermal Difference on thesin~s~n6cos~-cosQsin6sin~cosB+cos4cos~cos~cos~ Slope of Different Slope Direction and GradientThe formula 121 between the hear flux fromtsin$cos6sin~cosBcosw+cosmsinssin6sinw-0solar radiation ground heat interchange is asfollows:*Ve solve for w, these are sunrise and sunsethour angle. The calculation expression is:nwhere:C1-cos~cos6cosa+sinQcos6sin~cos!3 (1-10)."Cz=cos6s-inusinB (1-11)C~=cosmsinssinucosB-sinQsin6cosa ,(1-12)where b,6,8 are known numbers, a is the shadedangle of a point on the canal slope, it isdivided into the leaveslope angle a and thesunslope shaded angle a! The sunslope shaded'angle U 'is under the influence of the desertedsoil height of the canal top on the oppositeshore, canal bottom width, and slope point, wecan calculate it.with the following formula:slope bottom: a'- +g -1 1(M+N'(1-13)1 1slope middle: a'= tg-(3M+ZN'(1-10)the leaveslope shaded angle equals the canalgradient angle, that i 5: \1a - +g- (+) (1-15)Iwhere M is side slope coefficient, N 1s thebottom width.'When we calculate the sunrise and sunset hourangle for different gradient and slope with theformula (1-9), at first, we should determinewhether the sunrise (or sunset.) of the slope isthe leaveslope shaded angle (a) or the sunslopeshaded angle (a'). If we put U=O into theformula (1-8). we can obtaln the daily solarradiation of horizontal ground:Where Te is the earth's surface atmospherictemperature (regarded as ground surface temperature);Td is the ground surface temperaturewithout solar radiation; B is groupd heat emissioncoefficient; As is coefficient of thermalconductivity of SO'i1 body: he--viscous layerthickness. (Te-Td) is the increase of groundtemperature on the horizontal plane under thesolar radtation, that is (Te-Td)"At.. From theformula (2-I), we can write:The calculation formula of the increase ofground temperature for arbitrary slope directionand gradient can be written as follows:Atg = KtG %Where G is the rate of ground heat absorption;G-l=q (q is reflection ratio), other symbols arethe same as above.In winter, the relation between Lhe canalbed temperature and slope direction is veryobvious, the negative temperature ar'ea. isapproximately symmetrical with the distributionof slope direction. On the north slope, theduration of negative temperature is the longest,and the negative number is the largest. Centralto the north 'slope, the distribution of negativetemperature on the two side slopes are symmetrical.With the,rotation of slope direction, thefreezing time'ts lagged, the thawing time isadvanced. On the south slope, the negativetemperature is the highest, the duration ofnegative temperature is the shortest. The temperaturedifference of 5 cm deep soil layer is0 to 5'C, the temperature difference of 20 cm.deep soil layer is 2.5 to 3.0"C.

0Tahle 1. The rativ Kt of lining canal slope and horizontal groundin seasonally frozcn ground regionsLatitudeGradientSlope directiunS SE(SW) E(W) NE(NW) ElN37"Cr=c)O"45"33.7026.6"21 .8"I. 5501.8411.7321 .6271.5400.9141.3361.3601 .3291.2960.4030.7773.8iO0.8900.9160.0680.2470. 3660.4680.550000.0190.1890.318N40"9 = 9 1)45O3 9 . 7 O26.6"21.8"1 .6781.9551 .8251 .70b1 . hOh0.9741.42%1.4081 .3701 :3320.4110.7790.8690 * 8880.9140 ~ 0 30.2180.3280.42'10.5120000.1130.254N430'L=YO"45"33.7"2h.h021.8'1.8242.0801 .9291.7921.6821 .n351.4831.4601.4161.3720.4200.7810.8480.8860.9110.0580.1890.2830.3860.6700000.0460.182N46"a=90045-33.7026.6"21 -8"1.9362.2132 .Ob5I .8921.7681.0391.5461.5171.4661.4170.4270.7830.8570.8840.9080.0520.1610.2460.3380.42200000.104RemarkTo calculate the sunslopc shaded anglebottom width and canal depth.fa') according t.o t.he ratio 3:1 of the canal2. NeRative Temperature Index Calculation forArbitrary SlopeIn China 'on the study of canal base soilfreezing, according to the St-efan formula, manyresearchers provided that the relation betweenfrozen soil depth (X) and negative temperatureindex (F) can be written AS follows:2 = Efibut, because it is difficult to solve the negative'temperature index of arbitrary slopedirection and gradient. and the formula can't bedirectly used to calculate the frozen soil depthof arbitrary slope direction and gradient, thus,we regard the negative Lemperature indcx as aconstant. According to the prartically observeddata of frozen soil, we inversely solve a seriesof canal trend coefficients, canal bed positioncoefficients, etc., thus, we calculate thefrozen soil depth for different slope directionand gradient plane (Table 2), but, the computingwork is very complex, and the theoreticalevidence is not complete.According to the regression analysis ofgeothermal materials, we know that geothermalannual change regulation can be approximatelydescribed with the sinusoid. The axis of thesikusoid is the annual mean ground temperature(Te), the half of the difference between thehighest and lowest of the curve is the georhermalamplitude (A). Fig.2 is the regressedgeothermal change procedure line of the slope,rhe shaded area is the accumulating value ofmean daily negative temperature, therefore weintergrate the shaded area from TI to TZ andwe can obtain the negative temperature index.The slopes under the solar radiation, inwinter, the ground temperature of the slopeincreases At, the negative temperature index ofthe slope will decrease. Therefore, we integratetIFig.2 Calculated diagramture indexof the negative tempera-the ghosted shaded area from Ti to r:, and canobtain the negative t.emyerature index of theslope under the solar radiation. Fpy arbitraryslope direction and gradient, the generalexpression of the geothermal annual change is asfollows:T =(Te tAt ) t A sin-raBAfter being integrated we can obtain:(2-5)7. is conversion coefficient of days and degrees,3d5/360=1.01, we can consider that one dayequals one year.

ITable 2. The comparison of the measured and computed frozenground depths for different slope directions:;nfii;h;;EFrosheavingfieldLubei inShangdongSlope CalcumiddlelatedSlope Calcubottomlated0.80 1 .Oh 0.96 0.04 . 0.69 0.510.71 0.94 0.86 0.74 0.61 0.50Kz 1 .OD 1.32 1.20 1.05 0.86 0.71Calcu-SloDe latedmiddle M ~ ~ ~ -ured0.35 0.62 0.53 0.39 ' 0.23 0.090.36 0.61 0.47-0.51 0.31 0.15-0.20 0:10-0.08Kz 1 .oo 1.76 1.51 1.13 0.67 0.263. Correction Coefficient of Ground FreezingDepth Considering the Sunshine FactorThe relational expression between the freezingdepth of arbitrary slope direction and gradientplane and horizontal ground can be written as:Where Zg is frozen soil depth of arbitraryslope direction and gradient plane: X. is frozensoil depth of horizontal ground: F 'is negativetemperature index of arbitrary slofe directionand gradient plane; Fo is negative temperatureindex of horizontal ground.AKT changes with the value of -, whenATe+Atg AT,+At -1, there is no frozen ground: -TetAtg >'the sfope yields the negative tempera.ture index,because of the different values of AtB, itcauses the difference of slope directionality,with the increase of A , the difference ofT tAtslope directionality ggts imall. In the highlatitude regions where solar radiation intensityis smaller, the difference of slope directionalityof KT is very small, but, on the south edge 'of seasonally frozen ground regions, the differenceof the slope directionality KT is larger.When the frozen ground depth of the horizontalground is known, we can calculate the frozenground depth of arbitrary slope direction andgradient plane with the values of KT, under theconditions of the same soil properties andburied depth of ground water.4. To Determine the Influence Coefficient ofthe Buried Depth of Ground Water on the FrozenGround DepthTo the freezing depth of the base sol1 ofcanal bed, not only is there the difference ofslope directionality on the same height, butalso there is the differehce of the up-downposition in the same section. The research foryears "about frozen ground shows, under theconditions of the same soil properties andnegative'temperature, the freezing depth ismainly under the influence of the water contentof soil. The general regulation is: the moreshallow the buried depth of underground water,the more the bottom moisture will migrateupwards, the phase transition heat which themoisture yielded resists the development of thefreezing front. The freezing rate is smallerthan that of the soil layer with a low watercontent. !n fact, the deeper the buried depthof Eround .water,,the more the freezing depth ofthe soil body is deep, and the more shallow theunderground water table, the more the freezingdepth of the soil body is shallow. The influenceranEe (I1) 06 undernround water on the freezinndepih is.the aum OF the capilla r y elevationheight (hm) and freezing depth ( z) of soil body,that is:LSh,+zWhen L>h+z, the freezing dep t h isn't underthe influence of underground wa t er table, soKd=l. According to the material of the frostheaving test o? Zhang Ye, Gansu Province, withinthe range of LShmtZ, with the decreaae of thedistance between the underground water table andthe freezing depth, the influence of the burieddepth of underground water on the freezingincreases. The influepce coefficient of the"buried depth of underground water on the freezingdepth is Kd:- 0.96 t 0.65e-D (clay and ,silt ofdhigh and middle liquid limit) (2-10)1 = 0.94 + 0.5e-' (clay and silt of.Kdlow liquid limit),hFor the above re sons, the freezing depth ofarbitrary point on a bitrary slope directionand gradient plane ca be calculated a8 follows:(2-11)Where Z is the freezing depth of horizontalground around the canal: KT ia correctioncoefficieat considerin4 the conditions of thesunshine of different slope direction andgradient; Kd is the influence coefficient ofthe buried depth of underground water on the,f.reezing depth.1x1. FROST HEAVING PREDICTION OF ARBITRARY SLOPE IDIRECTION AND GRADIENT PLANEAccording to the frost heave mechanism, thethree main factors causing the froat heaving are

the soil propertiea, soil moisture (it standsfor the influence of underground water table),and the negative temperature idnex. Accordingto the research achievements on canal frostheaving for years in China, the expression ofthe 'calculation about frost heaving is:Ah6 - y .ZB (3-1)Where ZB is frozen ground depth on thearbitrary slope direction and gradiedt planeunder the condition of the arbitrary burieddepth of underground water, it can be calculatedaccording to the formula (2-11); y is the meanfrost heaving strength of soil (X).The relationship between the mean frostheaving strength and water,content of soil canbe calculated as follows:1. For shallow buried underground water, thecapillary elevation height reaches the frozenground layer, the water content is larger, andthe complimental source is full. The main factoraffecting on y is the buried depth of undergroundwater, that is:-blDy = ale (3-2)Now we put the'tested values of a, b intoTable 3.2. To deep buried underground water, thecapillary elevation height can't reach thefrozen layer, there is no compliment of theunderground water. So the relation between yand the water content before freezing is close,that is:IV. CONCLUSIONS1, The annual mean ground temperature (T,!)and the annual geothermal amplitude (A) determinedthe negative temperature index in thisregion. The increased ground temperature At6with the daily radiation is an important facLor..causing the differences of the negative temperatureindex.2. The buried depth of underground water andcapillary properties of soil are the importantfactors yielding the differences of frozenground depths and frost heaves in the verticaldirection.3. The annual mean ground temperature (Te),the annual geothermal amplitude (A), thetemperature difference bf slope direction (AtB),and the buried depth of underground wa'ter (D)are four basic factors on the freezing of thebase soil and the prediction of calculation ofthe frost heaving.REFERENCESYao Zhensheng, (19 ) The principle of climatology,The Press of Sciences.Zhu Bofang, et al., (19 , ) Temperature stressand control of hydraulic engineeri.ng concretestructure, The Hydraulic and Electric Press.Pan Shouwen, (19 ) The theoretical basic andapp1,ication of small climatic investigation,The Meteorological Press.Central Meteorological Bureau, (19 ) Theclimatic atlas of China, The Press of Map.y = aa(W-baWp) (3-3)where W is the mean water content (X) in therange of t.he freezing depth before freezing:Wp is the plastic water content (X); a,, b.are coefficients, solved in accordance withthP tested materials for years.Table 3. The statistic list. of the tested values of ai, bl in our country ~District , .Gansu Liaonin Heilongjiang Xinjiang ShangdongCoefficient40.0 (silty soil) 37 -0 27 .D 16.9 (silty cl-ay?a1 14.3 (sandy soil) 18.5 20.0 16.24 (soil) 46.3914.58 (sandy soil) (sandy soil)1.25 (silty soill 1.21.47 (sandy soil) 1.40.85 0.62 (silty clay)0.5 0.37 (soil)1.0 (sandy soil)0.963(sandy soil)1112

PERIGLACIAL PERIOD AND PLEISTOCENE NATURAL ENVIRONMENTOF WESTERN MOUNTATNS OF BEIJINGSeveral fossil periglacial phenomena found on the Western Mountains of Beijing enabled us to divide thePleistocene periglacial period into 6 epochs in the area. During the periglacial epochs the surface temperature waslower than that at present by 12-15°C: precipitation in winter was higher than that at present; depth of permafrostreached -3- -20m or more; and the natural environment was a savanna or tundra. The southern boundary of theperiglacial area extended to 38'N.1 IniroductionDuring the last years research on the Quaternarq periglacialphenomena has been rising to a new he1 in the world, lhe scopeand depth of the research increased significantly Intensiw researchof various recent petiglacial processes in polar and sub-polarregions and in some high-latitude regions has promoted theQuatcrnaty perlglacial rewnrch (Kaisser, 196'). Pewe.1075,Wnshhurn, I')?o) Alan! sclmtists time compared the environmentof formation of the Pleistocene periglacial phenonlena with that offormation of the recent periglac-ial phenonlena of sunilar activity.estimated the paleoclimatic conditions, and reconstructed theancient natural environments (Washburn, 1980; Bra& 1976,Mears, 1981) Since 1976 when periglacial involutions andcongelifolds were found in China and the perigdcial climate wassuggested, the problem has attracted the attention of scientists, thenthe research works on thc periglacial processes and periglacialepochs in northern China were successively reported (GuoXudong, 1984. Liu Tungsheng et al ,1087. Ciuo Yudong. 1988. SunJianzhong, 108 I )(1) Ice (soil) wedges. They are well preserved in ZhaitangVillage (i e. Dongrhaitang) and on a platform by a hill east ofQiansangyu BrickfieldLithologically a layer of secondary MaIan loess and yellowishbrownloess-like soil intercalating with sand-gravel layer areexposed on the profile The sand and gravel are mostly angular andcomposed hasically of debris after weathering of diabase on thehills It indicates that this layer of loess is indigenous deposits Itsgeomorphic position corresponds to the second. step terrace on theQingshuihe kver 4 iarye ice (soil) wedge was found on theprofile Hole of the wedge was filled with sand and soil and formeda soil wedge (Fig 1) The ice wedge is I O6m wide at its top and11 Discovery of Fossil Periglacial PhenomenaIhe Western Mountains are situated in the western and northcmparts of the Beijing Plain, i e In a trmsltion zonc between TaihangMowtams and Yanshan hlountalns 'The trend of the mountains ICapprownately northeast 1111: hlghest peak of the WesternMountains. Dongling Mountam. 1) at an elevation of 2 303 m andthe second peak, Baihuashan Mountain. at I 991 m most of theother mountains are Inner mounts and hdlb at ele*ation lower thanI 000 m The main wcam, In thc rl~ca arc Yungdinghe Riber andits tnhutarie,. Qmphulhe K~tol-. ctc Quaternan depos1:s oremostly distributed on both banks of mer \alleys and inintermontane basins. The deposits are mainly loess and loess-likesoil and less alluvial-colluvial sand-gravels The fossil periglacialphenomenafound by LIS are most11 developed in late PleistoceneMalm loess and middle Pleistocene Lishi loess These periglacialphenomena ari ice (soil) wedges, ice (sand) wedges, periglacialinvolutions, congelifolds, congeliturbations, etc.1113

0.2m wide at its bottom in an U-shape, extending to a depth of 1 60m. The fillings in the ice wedge represent coarse-grained sandgravelin its lower part (A). fine-grained sand-sail in middle partwith inclined bedding (B), and fine-grained sand-soil in upper partwith apparent involution bedding (C). It indicates that the icewedge after its melting was filled with congeliturbattion materialsfrom slope of hill(2) Ice (sand) wedges The fossil periglacial phenomena are welldeveloped and preserved on the Western mountains at QiansangyuA field observation in 1977 indicated that the hlalan loess west of0ainsang)u Brickfield contains a browish-red fossil soil layer in itsupper part and a sand-gravel layer about zm thlck In its middle pari4 groups ofjuvtaposlted different.size fossil Ice (sand) wedges werefound at the bottom of the sand-gravel layer The depth of frwcracking of the ice (sand) \\edges IS usually 0 5- I 0 m I'his profilewas destroyed by diging soil for brick nlaliiny(3) Periglacial involutions ,A profile exhibiting the periglacialinvolution IS found at a loess exca\atlon pi1 north of Lhaltang BrickFactory in ecident periglacial rn>olutlon was found in the. Lishiloess at eastern side of the protila (Fig 2) It can he seen in F I 1 ~that the cryogenic periglaclal process caused the homogeneouswater-saturated loess with horizontal hedding to be strong bcnt anddeformed. exhibiting well structural form of per~glacial in\olutionOn the left top of Fig 2. an ice (sod) wedge ehout l m long wasfound It deeply indents Into a periglacial rnlnlution, It indicatesthat the first cold-frost climatic event was folloued immediately bya next more severely cold climatic eventfluvium or usual colluvium, but is a congeliturbate characteristic Ofmudflow terrace. It might be formed as a result of the downwardslow creeping of active layer of permafrost or seasonally frost soilduring their melting111 Geologic Time of Formation of the Fossil PeriglacialPhenomenaI Geochronology and Stratigraphical CorrelationOn the Zhaitang profile the periglacial involution and ice(soil)wedge that ue found are located just at the bottom of Mdanloess to the upper part of Lishj loess. Thus we infer their formationtime must be the late stage of the Middle Pleistocene. A lithostratigraphicalcorrelation indicates that the stratigraphic positionsof the ice (soil) wedye and ice (sand) wedge swarm at both easternand western sides of the Qiansangyu Brickfield are the upper andlower part\ 01' the Malan loess, and hence their formation time isthe early and lata stages of Late Pleistocene. The congelifoldsystem in the Lishi loess at Yanchi is located in the middle andlower parts of the whole protile and hence its formation time is theearly-middle stage of Middle Pleistocene l-he congeliturbationfound at Guanting Forest Center is inferred to be synchronous withthe Shizhuang Formation of early Pleistocene from its contact withunderlying Pliocene red clay and higher geomorphic position of thefourth-step terrace on Yongdinghe hver at which it was found.2 Paleomagnetic datingIn order to determine geologic aye for formation of the fossilperiglacial phenomena on two profiles at Yanchi and ChantingForest Center. he have collccted 24 specimens for palecrnagneticdating ,211 the specimens were demagnetized at an alternative fieldIblth a peak of 250 Oe. their magnetic parameters were determinedon an English rotational magnetometer and the data were processed:md mapped on a computet The obtained results together with the,rbo~e-de>cnbed stratigraphlc positions and age data of the fossilpenglaclal phenomena are shown in Fig.3 It can be seen in Fig. 30 2 m3BdQsFig 1, Protile across perigl:lcral In\olutlon: 11: i,:sh: low. ,IILhaitangI Colluvium with sandy soil. 2 Sand! IoE;< -3 1.141: Ioc):.-3 Periglaclal involutron. 5 Ice (bod) \\edge(1) C'ongelifolds Thev resulted liom congelation ;intideformatlon of soil layer in the Pleistocene perlglacial area underthe severe cold climate During our field Investiyatlon thecongelifolds were lbund in Lishi loess at mlddk lower parr of aloess profile east of Yanchi Railway Station Ab II 15 shown 111 Ihcfigure that the homogeneous, more compact loess laver hasundergone sharp bending and deformation due to tnultlplccongelations. and became loosc In structure charactemtlc ofperiglaciatlon of the loebs(5) Conyel~turbate 4 profile ,iC't(lES the congeliturbate I\ locatedat the site by the hlyhway about one kilometer north ofGu:lntingForest Center on t1.c northern toot of the Western 'IlountamsLower part of the profile consists of Pliocene red clap and middlelowerpart of a layel- of lithologically mlxed loess..lrhe depwth withwave relief and slightly tune beddlny. Mhlch occur 111 ;I hharplyunconformable contact with Pliocene red clac Thc occcrrence ofgravel layer and thin layermg of the loess-like >oil ale ltnhrrcate Indescending order, and gravels arc angular It Indicates that this la$etofdeposits uas transporied at a small dlstance and is neither normal

of'o\erall posltive polarity there are 4 polarltl ~merd c>' en!%. !heirrnagnetic Inclination is negatike and !vgnltlcantl\ vane4 1 hr.represent ! rewrse polanty ekcnts. name11 (rctht.nburg eirnt k . '"I ( I clOOa I3 P Lacharnp event of23 OOOa H I> 13lacL elent n! 1' 11.)\la B I' and Biua event of. 0 2-0 5U 313 U P respectncll I hemiddle apper pan of the profile at GuantinS Cc:crt Cenlcr -how115 ney;r:l:e polarity and unly the loner pati h)\.x positit pkrttyIt Indlc;ltcs that the polarity column along {he\.!:ole proli!e '5 lustat the iwtom of Matsuyama recersc p~llar;~~ epoch. and 11sIhreshc4d ;\ge value is 2 43 Ua B P thc earh P!t.!:,tocene3 I hernroluminescent datingI'he tlrermoluminescent age of llalan It*c\\ at Lharlans %asreported by Lu Yanchou et al(lQ87) Mc had dSturtnrnedthemdummescent age of fossil ice wdses found( (lr! theOian\;insq,u profile The samples wet'? collected ti on^ llllinl:n1ater1aIs in ice wedges. loess in which he I ,ire found,and Its overlying loess (see Fig I) 'The result sho\$s that tillingmaterials from ice wedges are mosth debns crl d~abass,thus no agevalue mas determined I'he age of ICI~SS in i4hich ICC wedpe'uasfound is 35 000*3 SO0 a B P The age of bottom of loess owlyingthe ice wedges is 14 20011 800 a B.P It IS lnierred from thesedata that the formation time of these ice wedges is a very severecold climate stage after 35 000 a B P and before I4 200 a B P andmav correspond to the time of lowest world sea lekcl in a prexailinpstage of the last glacial period, i E I8 100 ;t R PIV .4 Prelimnary dtclsloll of Periglacial pcrirdA preliminary divlslon of thi periglacial period on the WesternMountains of Beijng is shown In Fig 3 6-7 periylacial epochs canbe divided in the area including those at Guantmg Forest CenterThey are described in ascending order as follo~s(I) Chanting periglacial epoch It IS represented by thecongeliturbstlon at Guantinp Forest Center and is a reflection of theEarly Quaternary periglacd cold climate Its age corresponds tothe bottom of Matsuyama reversal, i e. 2 43 Ma B P(2) Yanchi pcrigalcial epoch It is represented bv the congelifoldon Yanchi profile. The loess deposits have undergone sharpbending and deformation under the periglacial cold climate Its agecorresponds to Biwa event (I). about 0 5 Ma B.P But the overlyingloess has undergone the congelation The loose texture of the soillayer may represent another cold climate e\ent.(3) Zhaitang periglacial epoch (1) It is represented by the largerperiglacial involution beloh the second layer of fossil soil onLhaitang profile and its age corresponds to Riwa event (1). about0.3 Ma B.P(4) Zhaitang periglacial epoch (11). It is represented by the Fossilice (soil) wedge indenting into the abo\e-mentioned periglacialinvolution and its age corresponds to Riwa event (II), about 0 2 MaB.P(5) Qianwngyu periglacial epoch (1) It is represented try thefossil ice (sand) wedge swarm at the bottom of the gravel layerbelow a fossil soil layer in the MaIan loess on a profile west ofQiansangyu Bxickfield The penglacial climate was more severelycold than that in the prwious four epochs. Its age corresponds toBlack event, ahout 0 13 Ma B P(6) Qiansangy periglacd epoch (11) It is represented 'bq alarger fossii ice (soil) wedge on the profile east of QiansanbyuBnckfield The ice (soil) wedge was filled ~hb~ously with the/conpchturhation materials after its formatlotl which Is the bestewdenr*r. for the warming of periglacial cold climate Its ,agecorresponds to (jothenburg ekent Inferred to be 18 200 a B Prepresenting the most cold per~ylar~al climate in the area\' \arum1 Ihlronrnent in the PleistoceneI L rtnnat~on of PaleotemperatureIt 15 :walll; considered that the depth of ice wedges or sandnedpcs IS an Indicator for the desree of' cold and frost periglacialclinwe and the width of ice wedges is an indicator for thev cquency of cold and frost periglacial climate The depth of ice~cdges In Malan and Lishi loesses at Qiansangqu (41'N. 115'8'E)ani Lhaitang (39"8'N. 1 15"1'E) is 1 0-1 '3 m and width of them isI 04 m The temperature at the time of their formation atQiansanyyl and Zhmtang must be lower than thc present-day 10°Cby 16- I SL C', as calculated from the temperature of -6- -8°C of theisotherm indicator for the southern boundap of recently active icewedges in Alaska oi' horth America suggested by Pewe (1966). ButWashburn (1980) suggested that the temperature of -6- -ROC istaken as a standard for comparison to be too IOU and temperatureof-S"(' may be ;I more reliable upper limit at the formation time ofIce wed~es We suggest that the temperature during the time of theice wedge formatton at Qiansrmgyu and %haitany may taken to belower than that at present bq IS"('. as the standard taken to be -5°CIn regard to the temperature condition during the formation timeof the periglacial in\olutions and congelifolds, it was rarelydiscussed up IO no\\ But some scientists suggested that thetemperature durlng the formation time of chaotic periglacialinvolutions and plngos was -2°C'. For example, the prevailing epochof the Weichsel glacial period In Netherlands (52-5374) of NorthEurope was Inferred to be -2°C from the data of chaotic periglacialinwlutions and pmyos. that is lower than that at present by IS"C,and is consldered as a southern boundary of permafrost(Washburn.1WQ) I!' ne take the temperature of -2'C ah Btemperature standard during the formation time of periglacialinvolutions and conyelifolds. than the temperature during theformation time of periglacial involutions and congelifolds atZhaitang and Yenchi may bc lowet than that at present by I Z T Itcan seen that the Western Mountains of Bejing and the southernfoot of Yanshan Mountains (38"N) can be the southern boundary ofI'leistocene permafrost or ofperiplacial area in eastern China.2 Estimation of Pennafrost Depth7 he observed meteorological data from Zhaitang MeteorologicalStation indicate that the mcan tcmperature in Zhaitang area duringthe last IO years is 10.1"C. louer than that in the plain area by14°C' and down to -19°C in minimum. The averaqe depth ofpermafrost is 0.76 rn and reaches 0 Ob m in maximum If during thet'leistocenc perigalcial period the temperature at Zhaitang droppeddown to -5°C. lower thati that at present by 15T, than the depth ofpermafrost at that time must be similar to that at present in thenorthetn area of Pa Hingga Range and reached 20 m OF more (XieYouyu,,tYIl) If the temperature during the periglacial perioddropped down tu -2 'C, than the depth of permafrost may belikened to that at present In the northern area of' Songliao Plain.reachins 3-10 M below the surf'ace (Guo Dongxin et a1.,1381)3 ksumation of PrecipitationA certain moisture 'of soil layer's, like the temperature. i$ one ofthe necessarv preconditions for congelifraction and other plasticdefbnnation of the soil laycrs. Was the precipitation duringPleistocene periglacial period hlgher or lower than that at receotthe" An annual mean preclpitation in the last ten vears in Zhaitangarea IS 451 9mm and 508 8mm in maximum. 80-70% of thisprecipitation is concentrated in June.August and September andsnou in winter. that is rare solid precipltation 'Therefore. the soillayer is usually more moist in summer and more dy in winter.Under this climatic conditmn mcentlb no congelation can occur it11115

the soil layer and no ice wedges can be found. In the melting seasonfrom the late spring to early summer only a small-scale congelptionand frost heaving on road surface, can be found in the area wherelocal lowlands, higher groundwater table and higher moisture in soillayer are present. As shown in Fig.1, the textural and structuralflows and plastic deformation of the filling materials indicate thatsoil layer could undergo larger frost-cracking in the periglacidperiod and larger congelation, deformation and congeliturbationduring melting seasons It suggeststhat the moisture of the soillayer was higher than that at present. For this reason, the soil layercould be water-saturated and could undergo plastic creep. Itfollows that the climate at the beginning of the periglacial periodwas comparatively dry and cold, but in the middle and late epochsof the penglacial period the climate could become more cold andhumid, under which the precipitation might be higher than thepresent-day 452mm, especially solid precipitation in winter wasmuch higher than that at present.4. Fossil VegetationThe sporo-pollen composition and species in the filling materialsfrom periglacial ice wedges at Qiansangy are similar to those ininter-periglacial Malan loess at Moshikou (Go Xudong etd.,1981) and Zhaitmb (Yan Fuhua et a1.,1986). The main flora isArtemisia and ('hcr~opdiuceuc. less Cbrnpnsitae, Humdus,Sela~tnclla and hphedru. Among the woody plants Pinw andindividual Arics are found. The sporo-pollens in the loess aroundthe ice wedges are approximately the same as in their fillingmaterials, mainly Artemisia, 'hcnopndiaceae and Sc6ugincllu andindividual Pitnt.9. It indicates that the vegetation did not signiticantlyvary from the periglacial period to the inter-penglacial period.In general, from a viewpoint of ecologic environment, the abovedescribedplants are tolerant of dry and cold climate. It indicatesthat during that time the surface plants were sparse under the moredry and cold climate and windy winter. The natural features weresimilar to those of a savanna steppe landscape. But the structuraldeformation of filling materials in the ice wedges shows that theclimate in middle and late epochs of the periglacid pwiod was morecold and humid. Therefore, the savanna might be in a tundraenvironment.Kaisser, K.,1969, The climate of Europe during the QuaternaryIce Age. Quat- Geology and Climate, NAS, 1-12.Liu Tangsheng, Guo Xudong. 1987, The periglacial phenomenaon loess plateau, China. Loess and periglacial Phenomena, edited byMarton Pecsi and Huth M.French, Academia Kido,Pudapest, 141-149.Mars 8.,1981, Periglacid wedges and the late Pleistoceneenvironment of Wyoming's intermontane basin. QuaternaryResearch, 15(2), 17 Z - 195.Pkwe,T.L., 1966, Ice wedges in Alaska-Clnssificntion,distribution and climatic significance. in: Permafrost Intematiod<strong>Conference</strong> Proc., Natl. Acad Sci., Natl. Res. Counc. hrbl., 1287,76-81,P&.d,T.L.,1975, Quaternary geology of Alaska. U.S.G.S.ProfPap., 835,145.Washburn,A.L.. 1979, Geocryology, a survey of periglacidprocesses and environments, 3 13-318.Washburn,A.L.,1980, Permafrost features as evidenceof climatic change. Earth-Science Review, 15(4), 3 17-393.ReferencesLu Yenchuo et al.,1987, Thermoluminescent dating of coarsegrainedquatz frbm Malan loess. Kexue Tongbao, Journal ofSciences, (1),47-50 (id Chinese).Yan Fuhua, Ye Yongying, and Liu Yuexia, 1986, Sporo-pollenanalysis of loess at East Zaitang of Beijing. Quaternary Research inChina, 7(1),39-43 (in Chinese with English abstract).Guo Dongxin,Li Zuofu,1981, Historic evolution and formationtime of permafrost in northeastern China since late Pleistocene.Glaciology and Geocryology, 3(4) (in Chinese).(30 Xudong et al., 1981, Possibility of Quaternary glaciation inWestern Mountain area of Beijing from "glacial traces" atMoshikou. Glaciology and Geocryology, 3(2),63-66 (in Chinese).Guo Xudangl984, A preliminary study on Quaternary climatein China. Glaciology and Ceocryology, 6( l), 49-60 (in Chinese).Guo Xudong, 1988, A discussion on Quaternmy glacial climateand environment in China. Scientia GBographica Sinica, 8(2), I 14-126(in Chinese with English abstract).Xie Youyu, 1981, Periglacial geomorphology of NortheastChina and its zonation.GlaciologyandGeocryology,3(4),17-23(in Chinese).Brack R.F.,1976, Periglacial features indicative of permafrost:ice and soil wedges. Quaternary Research, 6( 1),3-26.

Haiying FU, J. G. DASH, and L. WILEN,Department of Physics,and €3. HALLET,Department of Geological Sciences and Q.uaternary Research Center,Unersity of Washington, Seattle, WA 98195, U.S.A.We have studied the dependence of the unfrozen water content of porous media at subfreezingtemperatures on the mineralogy, size and surface properties of particles. A seriesof well characterized monodisperse powders and natural soils were examined by a variety oftechniques, including neutron diffraction, quasi-elastic neutron,scattering, time domain reflectometryand mercury porosirnetry[l] [2][3][4]. The melting process and, hence the amountof unfrozen water remaining unfrozen in porous media at subfreezing temperatures, are gavernedby surface melting, curvature effects, and impurity effects. Current theoretical treatmentsof these effects are consistent with experimental results.'JNTRODUCTIONIt is widely recognized that the persistence of waterin the liquid form in porous media as temperature3 dropbvlow O°C underlies many important natural phenomenainvolving freezing, including frost heaving[5][6] andfrost weatheriag[7] [8]. It is also related to a diversityof other processes ranging from ice particle sintering tocloud electtification[9][10]. The amount of unfrozen waterin frozen soils and rocks has been well documented,but there is no consensus as to its causes.Through an interdisciplinary research program weexamine the properties of unfrozen water in porous mediausing of modern experimental techniques and newtheoretical developments in condensed matter physics11 11.We relate the persistence of unfrozen water at sub-zerotemperatures to the tendency for ice to premelt at thecontact with foreign substrates below its bulk meltingpoint and for the freezing point to be lowered due tointerfacial curvature and solute effects. The currentstudy focuses on how much ice melts at the contact ofa substrate and how this melting depends on the propertiesof the substrate.EXPERIMENTSTo ensure maximum experimental sensitivity, finepowders are used as substrates in order to have a largecontact area with ice. Powders of spherical particleare used for a concise theoretical treatment. We usetime domain reflectometry(TDR) to mearjure the waterfraction in such frozen icilpowder mixtures. TDRemploys an electromagnetic wave (EM) pulse and measurathe dielectric constant of a medium by analyzitlgthe reflected EM pulse against the incident pulse. Thetechnique provides a sensitive indicator of minute quantitiesof water because the dielectric constants of waterand ice differ greatly at the observing frequency.We tested monosized spherical powders of varioussizes and diverse cornpasition ranging from polystyreneto silica beads. The measured unfrozen water contentwas found to be consistent with a theory[l2] accountingfor surface melting, and the effects of curvature andsolutw (Figures 1 and 2; the apparent discrepancy inFigure 2 below lo from melting point is due to the factthat the beads are porous). Despite the wide range ofsurface properties of these powders, we found no detectablevariation of the unfrozen water content withsurface property.Powders consisting of irkgularly shaped particles werealso tested, including some natural soils. The simiiaritiesin nonwetting behavior of ice and mercury on mstknown materials make it possible to employ mercuryporosimetry(MP) to characterize the melting behaviorof ice in the porous media[3]. The volume of mercuryvapor is equated to that of water and the pressure P ofthe mercury is related to the temperature AT = T, -Tof ice by1117

where T, ia the melting point, p, and qm the densityand fusion beat of ice at Tm and < the surface energyratio of ice-water to mercury-vapor. The unfrozen watercontents deduced from mercury pormimetry are incle accord with direct TDR measurements for a diveisityof artificial pawders(31. For samples such as naturalsoils, mercury porosimtry and TDR yield similar results,provided apparent impurity effects are taken intoaccount (Figure 3). In a freezing ice/powder mixture,impuritia are expelled into the unfrozen water a telativelypure ice crystal form giving rise to significantmelting point depression even for pore waters that areinitially very dilute.The temperature dependence of the unfrozen watercontent for diverse porous substances agrees quantitativelywith descriptions of both curvature effects andsutface melting, a phenomenon common to many typesFigure 1. Water fraction in frozen mixtureof water/polystyrene beads(5pm in radius):0: TDR; -: theoretical calculations[ll].1of mlida, and ~lt) actively studied topic in condensedmatter pbyaics(ll1. A simple practical model haa beendeveloped to determine the amount of unfrozen waterbzwd on measutements of pore size distribution andsimple theoretical atirnates of Van der Walls interactionsaffecting the water, which are insensitive to theporous media composition and rnineralogy[3]. In addition,novel experiments are currently under way to determinewater transport rates over a wide temperaturerange in single film betwen ice and foreign surfaces,to provide detailed data for comparison with theory.REFERENCES1 Gay, J.-M., Suzanne, J., Dash, J.G. and Fu, Haiying(1992) J. Cryst. Growth 123, 33.2 Maruyama, M., Eienfait, hl., Dash, J.G. and Coddens,G. (1992) J. Cryst. Growth 118, 33.3 Fu, Haiying and Dash, J.G. (1993) J. Colloidal interf.Sci., 159, 343.4 Fu, Haiying (1993) PhD Thesis, University of Washington,Seattle.5 Anderson, D.M. and Morgenstern, N.R. (1973) inPermafrost, Proc. 2nd Int. Conf., Natl. Acad. Sci.,Wash. D.C., pp.257-288.6 Dash, J.G. (1992) Conf. on Quantum Fluids andSolids, Penn State U.,,June.7 Walder, J. and Hallet, B. (1985) Geo. Sac. Am.Bull. 96, 336.8 Ballet, B., Walder, J.S. and Stubbs, C.W. (1991)Perma. Perigla. Prm., 2, 283.9 Gilpin, R.R. (1980) J. Colloid Surf. Sci. 77(2), 435.10 Baker, M.B. and Dash, J.G. (1989) J. Cryst. Growth,97, 770.11 Dash, J.G. (1989) Contemp. Physics 30,89.12 Cahn, J.W., Dash, J.G. and Fu, Haiying (1992) J.Cryst. Growth, 123, 101.k2 .01; .003Figure 2. Water fraction in frozen mixtureof waterlquartr beads (1 35pm in radius):0: TDR; 0: deduced from MP; "-: theoreticalcalculations[ll]..oo 1.001 .01 .1T,-T1 10(K)tooFigure 3. Water fraction in frozen mixtureof water/Faubanks silt: 0: TDR; 0: deducedfrom MP; -X-: deduced from MPwith impurity correction[3].

A STUDY OF THE THERMAL STATE IN THE PERMAFROSTAT THE SEJONG STATION, ANTARCTICAUk' Han andH.C. JungDept. of Env. Sci., Korea Military Academy, Seoul, Korea 139-799Borehole temperature measurements at Sejong station were made by geothermaldatalogger which was designed by the investigator. During December 31, 1991 -February 1, 1992 six temperature data (at the depth of 28cm. Scm, .-12cm, -32cm,-52cm, and -70cm) were obtained by resistive sensors of CRlO and SM716 every oneminute. Fast Fourier Transformation was made on seven temperature data includingsurface air temperature of meteorological center at Sejong Base every thirty minutes.Profiles of surface and subsurface temperature variations represent freezing, thawing,and heat transfer mechanism at the boundary between active layer and permafrost tableThe thermal diffusivities are determined by the Angstrom method, using undergroundtemperatures. The thermal conductivities of the drilled cores and outcrops are measuredby the transient method. The thermal diffusivitv and conductivity measurementsof rockand soil samples give a significant signal on the inferred climatic temperature of thepast millenium at Sejong Base by the inversion technique with well-documentedmeteorological data. The geothermal data in ice-bearing permafrost at Sejong scientificstation are interpreted in terms of the temperature history on a time scale of 2,000years. Two theories are developed: a "forward" theory to calculate the response ofice-bearing permafrost to a surface temperature disturbance, and an "inverse" theory tocalculate parameters characterizing the surface temperature history from suitablemeasurements in the permafrost.INTRODUCTIONThe construction of the King Sejong scientificstation at King George Island, Antarctica, hasprovided an unprecedented opportunity for the strdyof the subsurface thermal regime in an area of tpin,ice-bearing permafrost as shown in Figure 1.Korean Antarctic Research station, King SejongBase is located near the southern shore of MatianCove which is one of the tributary basins ofMaxwell Bay. Barton Peninsula is free of snowduring austral summer and the exposed area isestimated to 15km': At the coast, raised beachs andmoraines are developed 5 or 7 m above sea level.The raised beaches of Fildes Peninsula whose heightis less than 20 m are known to be formed inHolocene. So all the raised beaches around KingSejong station are considered to be formed inBolocene(Y ang and Jeon, 1990).Han(1991) and Lachenbruch et a1.(1982) analyzedthe data to obtain the characteristics of the globalwarming during the last century and an estimate ofthe equilibrium heat flow. Harrison(l991) studiedsurface temperature change and its implications forthe 40,000-year surface temperature history atPrudhoe Bay, Alaska.The meteorological data are collected fromDecember 31. 1991 to February 1, 1992 at KingSejong station. The near station level pressure wasrecorded as 989.8 mb during the period ofobservation. The annual mean air temperature was-2.0 C and the mean wind speed was 8.0 m/s.Predominant wind direction was northerly and themean relative humidity was 85 96. In 1991, B lowest,temperature of -24.4 C was measured at the.5th ofAugust and the greatest gust of 46.6 m/s wasobserved on the 11th of September.In King George Island, 95 96 of land surface iscovered by glaciers and snow all the your round. Thethickness of the glaciers is estimated 100 m. Theymove slowly and creep over their own bed. Ingeneral, sea surface is frozen frequently in the coldseason. The thickness of sea-ice reaches 25 - 30 cm.1119

ce' 5t I-z04-0'0 4 8 12 16 20 24 28date (lS91.12.31-1992.2.11ANTARCTICA4- SOUTH POLEFigure 1. Location of the Korean Antarctic Researchstation at King Sejong Base.Temperature measurements from one hole atsejong station were made by geothermal dataloggerwhich was designed by the investigator andCampbell Scientific Inc. Thermal properties from coresamples were measured. Six temperature data at thedepth of 28 cm, 8 em, -12 cm, -32 cm, -52 cm, and-70 cm were obtained by resistive sensors of CRlOand SM716 during every one minute. The subsurfacetemperatures which were relatively free of thedisturbance from drilling are shown in Figure 2.The thermal diffusivities were determined by theAngstrom method using ground temperature data andlaboratory measurements. The thermal diffusivity atthe depth of 52 cm are calculated as shown in Figure3. The thermal conductivities from drilled samplesand outcrops were measured by the quick thermalconductivity method. The typical values of rocks areused to calculate the response of the permafrost tosurface temperature disturbance.Fast Fourier Transformation was made on seventemperature measurements including surface airtemperature of meteorolpgical center at Sejong stationduring every thirty minutes. Profiles of temperaturevariation represent freezing, thawing, and heattransfer at the boundary between the active layer andpermafrost table.Figure 2. Subsurface temperature variations at depthwith 52 cm. 'The time constant controlling the response of theSejong station permafrost thickness to a surfacetemperature change is of the order of 8,000 years.The results indicate the evidence of a complexclimate history in King George Island, Antarctica.The tentative conclusion drawn from geomorphic andgeothermal measurements can be reconciled.This study is devoted to the development of aforward theory of the response of the permafrost to Hsurface temperature changes, and the development ofan inverse theory for the estimation of thermalparameters characterizing climate history.FORWAFQ PROBLEMSince the permafrost at Sejong Base is close tobeing in equilibrium with the surface temperature, itis adequate to use a simple analytic approximation(Harrison, 1991).thermaldiffysivityfl depth : 32 ant depth : 52 cm016 IO 14 16 22 26 30time (day)Figure 3. Thermal diffusivityby Angstrom method.' 1120

0,. We assume that the heat transport is purelyconductive and that the thermal conductivity isconstant except for a discontinuity at the base of thepermafrost.In forward theory, we find the response of thepermafrost to a surface temperature To of the formTo=T,Il+e$s(t)) * (1)where Tm is the mean surface temperature, Bo is asmall nondimensional parameter characterizing theamplitude of the time-dependent surface disturbance,and fs(t) is the generalized periodic functionwhere the C. are Fourier coefficients, the n areinteger, and w is the angular frequency of thedisturbance. If Bo f the equilibrium permafrostthickness L, is determined bywhere kl, kp are the thermal conductivities of thepermafrost and of the material ben'eath it, and ktVT,is the equilibrium geothermal heat flow.For a small 90, the permafrost thickness L and thetemperature gradient VT immediately beneath thebase of the permafrost behave as follows:(11) '* -xl and aa are the thermal diffusivities of thepermafrost and of the material beneath it, and I isthe latent heat per unit volume of the permafrost.eo is real, but q", P' are complex because thethickness response and the temperature gradientresponse are not in phase with the surfacetemperature disturbance.We examine the response of the permafrost atSejong Base to the simple harmonic surfacetemperature perturbation and calculate the dependenceof the response on frequency. The numerical valuesused are summarized in Table 1. The results are inFigure 4. In the case of the Sejong Base, thecharacteristic frequency wc, at which the gradientresponse parameter goes through a broad resonance,has the value 5.46 X IOL4 yr-'. For o I* w., thepermafrost remains almost in eauilibrium with thechanging surface temperature.We also calculate the permafrost responses for thesimple harmonic surface temperature distrubance withangular frequency 0 = 7.85 X 10 yr"(the 8,000 yearperiod). The thickness and gradient amplitudes are0.54 and 0.1, respectively. In addition, the thicknessand gradient are phase shifted: they lag the surfacetemperature by 1,150 yrs and 150 yrs.If we assume the disturbance function fe(t), thenthe inverse problem is to solve the equations (l), (31,Table 1. lnputnu, I* are nondimensional parameters charactekzingthe thickness response and the 'temperature gradieptresponse, respectively.whereParameterkl 1 .o W m -'K -'ka 2.0 W m -'Ka1 52.0 m'yr"aa 22.0 m'yr"1 1.2X108 Jm'3T.'" -2.0 KL (P) 30.0 mVT 30.0 Ckm-ltl 4 8.7 yearsta 20.5 yearsts 1712.3 yearsValuekl, k2 and al, RZ represent thermal conductivities andthermal diffusivities of the permafrost and of thematerial beneath it. Volumetric latent heat 1, surfacetemperature the permafrost thickness L(D) andthe temperature gradient beneath the permafrost VT(D)are listed. tl, ta and ts are time constants controllingthe response of the permafrost.1121

Timet ( to3yrs)Figure 4. Frequency response of the permafrostthickness and of the temperature gradient beneath thepermafrost for a simple harmonic surface temperatureperturbation.Figure 5. Profiles of the responses of the permafrostthickness and the temperature gradient for the simpleharmonic temperature perturbation function with aperiod of 8,000 years.(4), (5) for Bo and T,, which are the parameterscharacterizing the surface temperature history, at t=O.After some manipulations, one obtainswhere Ais defined by(13)A is disequilibrium parameter. If the permafrost isin equilibrium, then 5 = 1.Since 0. a 1, equation (13) is accurate only if I *1.When Bo has been determined, T,, L, VT,,, arefound, therefore, we can determine the permafrosttemperature history.In the case of the Sejong Base, we coulddetermine the temperature histories for the simpleharmonic and the sawtooth disturbances. Th.e resultsare in Figure 5 and Figure 6, respectively.Ifjp =A' +fy, (14)Bo, and the surface temperature history, areindeterminate. But we have fe'" = 0.707. fl(P) = 0.533,fu(r) = 0.077 for the simple harmonic disturbance andfe") = -0.738, fltP) = 0.313. f.'D' = 0.063 for thesawtooth function.The duration of the memory of the permafrost canbe expressed in terms of t3, the time required for thepermafrost thickness to adjust to surface temperaturedisturbance. In the case of the Sejong Base, weemphasize the 2,000-year average temperature. The-10 -a -6 -4 -2 QTime t ( IO3 yrsFigure 6, Profiles of the responses of the permafrostthickness and the temperature gradient for thesaw tooth temperature perturbation function with aperiod of 8,000 years.2,000-year average temperatures are -2.05 C, -2.13C for the simple harmonic and the sawtoothdisturbance, respectively.This study is to analyze the geothermal data andsurface temperature at Sejong Base on a 8,000-yeartime scale. Two theories are developed, a forwardproblem for calculating the response characteristics ofpermafrost to surface temperature perturbation, andan inverse problem for estimating parametersx122

~ thicknesscharacterizing the surface temperature and thermalhistory from the permafrost.The effects on the permafrost response of thethree time constants tl, tr, and ts are calculated, andcontrol the rate of permafrost response to the surfacetemperature perturbations. The time constant tl 8years is 'a measure of the temperature within thepermafrost to adjust to surface temperature change.The time constant tr = 20 years is a measu,re of thetime required for the temperature beneath thepermafrost to adjust to the position change of thebase of the permafrost. Two constants characterizethe effects of the sensible heat within and beneaththe permafrost. The time constant ts 1,712 years isa measure of the time required for the permafrostto adjust to surface temperatureperturbations. The third constant characterizes theeffects of latent heat at the base of the permafrost.Although there are three time constants (tl. tp, ts),the duration of the memory of the permafrost can berepresented in ts, 1,712 years. The information thatcan be estimated depends on the form and the phaserelative to the present time. Therefore it is important'to consider the uncertainty. This uncertaintyincreases with increasing time into the past. Thepermafrost remembers best what has happened in thelast 1,712 years. In our analyses it seems desirable toemphasize the surface temperature history over thelast 2,000 years.The sawtooth and the simple harmonic oscillationwith a 8,000-year period are used to a reasonableapproximation to these data. We determined thetemperature history for the sawtooth. The memoryby the permafrost is the subsurface temperaturebeneath the active layer. The comparison of theperiodic sawtooth and the simple harmonic is anexample of how the choice of fe affects the results.The present thermal regime at Sejong Base putsconstraints on the subsurface temperature over thelast 2,000 years. On this time scale there has beenvery small change and the subsurface temperatureover the period was roughly the same. The 0.1 Cwarming of the past century is calculated.Han. U. Il99l) Global warming from boreholetemperature evidence. J. Kor. Earth Sci. SOC. 12,93-99.Harrison, W. D. (1991) Permafrost response tosurface temperature change and its implicationsfor the 40,000-year surface temperature history 'atPrudhoe Bay, Alaska. J. Geophys. Res. 96,683-695.Lachenbruch, A.H., H.H. Sass, B.V. Marshall, andT.H. Moses, Jr. (1982) Permafrost. heat flow, andgeothermal regime at Prudhoe Bay, Alaska. J.Geophys. Res. 87. 9301-9316.Yang, J.S. and D.S. Jeon (1990) A study forenvironmental impacts assessment on naturalenvironment in the new construction area aroundthe Korean Antarctic station. Kor. .J. Polar Res. 1,25-34.

TWO-DIMENSIONAL STEFAN PROBLEM AROUND A COOLED BURIED CYLINDERH. Haoulani A.M. Cames-pintaw and J. Aguirre-puenteLaboratoire d'aerothennique du C.N.R.S4 ter, route des Gardes, 92190 -Meudon, FranceA two-dimensional finite element numerical method using the enthalpy as thermal variable is used totreat the Stefan problem around a shallow cylinder placed at a finite depth in the ground.Systematric calculations for a certain numbcr of specific cases conduct to:the knowledge of the thermal behaviour of shallow buried pipes; a particular study of sensitivenessof the thermal systems to the pricipal parameters as boundary conditions and geometrical andthermophysical characteristics; a proposed empirical model, based in rigorous calculations to predict theevolution of the freezing or thawing front.The investigation on the heat transfer phenomena in soils with change of phase leads to either asimple method to easily predict the movement of the first line around a buried pipe or a rigorous numericalcalculation.INTRODUCTIONThe "LABORATOIRE d'AEROTHERMIQUE DUC.N.R.S" has conducted studies on cooled buried cavities andpipes. This group has devcloped several numerical models to solvephase change problems allowing the determination of the thermalregime in freezing or thawing grounds.Two relevant models have been used in many geotechnicalproblems. They make possible the determination, as function oftime, of the progression of the interface seperating the frozen andunfrozen zones, and the thermal field within the domain.In this paper, we focus our attention on shallow buried pipes.Axi--symmetrical ModelAn axi-symmetrical, one-dimensional model with phasechange was developed by Cames-pintaux et al. This model,adapted to cylindrical coordinates, allows to locate the front bysolving the Neuman equation on the interface accompanied by theequations of heat diffusion.For the case of deep buried pipes, the position of the groundsurface does not seem to have a significant influence upon the generalevolution of the thermal process. Thus, the Stefan problem canbe investigated in an axi-symmetrical domain.In order to solve the problems encountered in the case of shallowburied cavities, the axi-symmetrical model presents some limitations.The use of a two-dimensional model then becomes necessary.The minimal depth at which the-problem may be treated withthe axi-symmetrical scheme needs to be defined.Enthalpy ModelA finite element enthalpy method, developed byCames-pintaux and Larnba, was used to solved the phase changeheat transfer problem by considering either a discontinuous surfaceor a transition area, seperating the frozen and unfrozen zones.This model has particularly contributed:to the study of the thermal bchaviour of the ground around cylindricalcavities for cryogenic Storage,to the determination of the domain where the axi-symmetricalmodel may be accurately used to describe or predict the thermal regime.APPLICATION OF THE TWO-DIMENSIONAL MODELTheoretical AspectEnthalpy formulation are well adapted to the treatment ofproblems where the change of phase occurs.Two mathematical changes of variable combined with theenthalpy formulation are used in finite element enthalpy methodproposed by Cames-pintaux and Lamba. This two-dimensionalapproah was exploited to solve enthalpy equations for different geometry's of the domain under different boundary conditions andfor different thennophysical characteristics of soil.Geometry of the domainA cylindrical cavity with a radius Roand at buried depth R', isconsidered. In every case presented in this paper, R,& equal to 0.1m.The symmetry condition reduces calculations to only a half ofthe system Figure 1 I T,is the plane of symmetry, r,the cylindricalsurface of the cavity and r,the surface of the ground and the otherjudiciously chosen limits of the domain. d is the width of the domainalways chosen equal to &R,. .Boundary ConditionsThe heat flux is null on r,(d@ / 6n= 0), the temperature in thecavity @=@,,and the temperature at the surface and at theother boundary r,is 8 = Q,

The variation of density comapanding to the transfonnationwater / ice is neglected. Thus:p,=p, = 2.01~10'kg.cm". calculationshave been made for four hypothetical soils defined by combiningthe two extreme values of diffusivity a and two extreme Valuesof the latent heat L.Table 2 shows the four soils characterised by the product axL.hTable 2a* L latent heat latent heatFig. 1. Underground cavity scheme.The following numerical values of temperatures have beenchosen to be studied.'a1 =0.401x1o6(m2.s")thermal diffusivitya2 = 1.729~ 10"(m2.s-')23.86 79.12102.87 341.13These temperatures arc suddenly imposed at t = 0.It is assumed that the initial distribution of temperature@,within the system is linear (Pigure 2).~-RESULTS3.7These geometrical and thennophysical characteristics wereused to solve the Stefan problem by the finite element enthalpvmethod. For the four soils, thc buried depth R'= 1.10. m and thetemperature in the cavity op, = -I 5C are considered . For the foursoils, the evolution of the upper vertical and lower vertical frostlines arc represented respectively'oo Figure 3 and Figure 4.Fig. 2. Temperature distribution at t =O.Thermophyaical Characteristics of the GroundTo study the effect of the thennophysical properties, particularlythermal diffusivity a and latent heat L, we have usedgeotcchnical &ita employed in previous papers. The chosenthennophysical characteristics correspond to extreme values ofthose considered as probable in a group of soils presented bySawada and Ohmo. These values are given in table 1.Domaintable 1frozen clay unfrozen clayCIcx0.60.30.03.a, * L,100 200 300a1 a2 a1 a2 fig. 3. Frost front progression along the upper vertical plane ofsymetry.heat capacity-103 ~~~0.686 c,=0.978 c,= 1.254 %= 1.291(J.K~-'.K-')k thermal(W.rn-'.K-')Time tIn previous papers, we have studied the thermal behaviourk,=O.554 k,=3.400 k,=0.236 kl=2.0m around a cooled urban pipe but only with help of the axi-symmet- .rial one-dimensinal model.(day)1125

t x 106 (9)8 ' 16 24Io A,-L, x A,-L, o A,-L, A,-L,Yx a2* L,I 100 200 300Timet (day)fig. 4. Frost front progression along the lower vertical plane ofsymetry.To numerically study the Stefan problem around a buried pipewith an onedimensional scheme, it is necessary to make a fictitiousradious R' (Fig. 5). R'delimits the domain and the sameboundary limits are utilized on r,for the tempraturc Tp,,in thepipe and on r,for the temperature TI& the surface of soil. The influenceon R'and @,,position were studied for the four soil (Fig.6).0' 100 200 300 330R*+AR* TD;,=-lsOCTie t (day)Fig. 6 Incidence of the "R 0,' couple on the frost line positionupon vertical axis as a function of time.* Intluencc of latent heat L:The influence of latent heat L on the freezing front behaviourtakes place only for transient portion of solution.This influence is negligiblc for materials with a high thermaldiffusivity, at every time t, but important for materials with a Iowthermal diffusivity a,during the transient portion of the problem.PRACTICAL METHODfig. 5 Axi-symmetrical buried pipe.The two-dimensional and one-dimensional numericalsimulations have shown the influence of the principal parameterson, the thermal behaviour around the pipe during the freeze / thawcycle.1). Incidence of the pipe depth R':For a given temperature in the cavity @,and for a given time t,it is noted that the upper vertical frozen zone thickness increaseswhen the pipe depth R' increases.2). Variation of the temperature @,within the cavity:For a given buried depth R', it is interesting to remark that theupper vertical freezing front movement increases when the temperature@,,in the cavity decreases. Nevertheless, a limit value is im.posed by the boundary conditions at the qround surface.3). Effects of the thermal properties a and L.* Effect of thermal diffusivity a:The establishment or absence of establishment of steady-stateis directly conditioned by the diffusivity a. Thus,thc permanentstate is rapidly reached in media (a2,Ll) and(a2,12)presenting ahigh diffusivity.The use of an elaborate two-dimensional model to solve thenumerical problems for shallow buried pipe implies expensive calculations.Itis interesting to have a simple model to quickly predictthe interface behaviour during a time period T.The interpretation of results leads to empirical expressions andgrhhs permitting a valuable synthesis of the results in the exploiteddomain.The results arc presented ai.dimensionlcss variables with rbgard to curve of frost line obtained for:(T), = 300 days (referenced time),(e),=-1% (referenced temperature within the cavity),(It),= 1.lO.m (referenced cavity depth),(a,,L,) (referenced soil).Variation of Pipe Buried Depth R'The referenced value Ais chosen equal to 0.71 m.For a given values of Q,,and T,, we can predict about the table3 the upward freezing front location with the help of equation (I):A(R)(Q,,T,)I=-(R)(O,,,T,)I.A,where a(R) can be calculated by a linear interpotation'function:('11126

”-a(R)(0,,,Tl),=(0.9286.R+0.0714) (2)for 1.05m

PERMAFROST MAPHNG USING GRASSRichard K. Haugen, Nancy IT. Greeley and Charles M. CollinsU.S.Anny Corps of Engineers, Cold Regions Research and Engineering Laboratory, 72 LymeRoad, Hanover, NH 03755-1290Knowledge of the spatial Occurrence of permafrost is critical for hydrologic and engineering purposes.The site for our study is the Caribou-Poker Creeks Research Watershed, a 37-square-mile areanear Fairbanks, Alaska. The monitoring of air, surface and subsurface ground temperatures since 1986in this discontinuous permafrost upland taiga environment has provided ground truth data for proximalpermafrost and non-permafrost underlain terrain. In our initial analysis, we found a significant correlation(rZ=0.68) between observed mean annual surface temperature and calculated equivalent latitude foreach of the scven drill hole sites. Equivalent latitude is a theoretical index of direct solar radiation incidenton a surface, which serves as a measure of thermal energy received at that point. The use of a GIsto provide a spatial distribution for the equivalcnt latitude index was an obvious next step in mappingpermafr4st. Arc-Info DLG files of elevation, soil and vegetation developed previously were translatedand imported into GRASS.An equivalent latitude map was developed in GRASS using the equivalent, latitude algorithm and an algorithm derived from the observed variations of mean annual air tempcraturewith elevation within the watershed. Further development of our permafrost mapping approachwill include the location of all temperature recording sites with a global positioning system (GPS) for en-. try into the CIS and the refining of mapping algorithms to reflect differing surface energy balance regimeswithin the soil and vegetation mapping units.~ MAP " . DEVELOPMENT " I -AND CIS PROCESSES USEDThe data used in this projcct was translated from an Arc-Infodatabase of the Caribou-Poker Creek Research Watershed developedunder the supervision of Leslie Morrisey for NASA. The datawas translated into a GRASS4.0 geographic information systemdatabase on a SUN Sparcstation 330 system. The data resolution is30 meters. The projection used is univcrssl transvets mercator.The trnnslatcd clcvation map was used as input to create slopeand aspect maps using the GRASS r.slopc.aspect command. Thereapcearcd to be a few placcs where slopc and aspect data wereabcrrant. These are visible in our final maps as small horizontallinear patterns. An attempt was made to smooth the linear data byusing GRASS'S smoothing filters, however we felt that the resultswere not worth the loss of data that occurred throughout the rest ofthe map after the filtering process.To create the equivalent latitude map we input the followingequation into GRASS'S rmapcalc command:equivalent latitude = sin"(sin(s1ope) * cos(aspect) *cos(actua1 latitudc)+cos(slope) * sin(actua1 latitude).GRASS's r.mapcalc command didn't have a sin function, so thederivative equation was used:equivlat= atan((sin(@slope.rec) * cos(@aspect.values) *cos(65.18288))+cos(~slopc.rec) * sin(65.18288))/(sqrt(I-exp((sin(@slope.rec * cos(@aspect.values) *cos(65.I 8288))+cos(@slope.rec) * sin(65.18288)),2)))The rmapcalc command processed this formula for each 30 meterdata cell of our study area and output a map of the results for eachdata cell, an equivalent latitude map. The "@"-sign is required inGRASS to enable the use of the category value instead of the cat&gory number in each of the map layers used.To create the mean annual surface temperature (MAST) map,a regression cquarion was developed using elevation, cquivalent latitudeand temperature data from data sites in the study area.R.mapcalc was run using this equation:mastactual = 21.704-0.003 * elevation-0.29 * equivlat.Temperatures of iryersion zones were then added to the MASTmap using the GRASS command r.infer to extract the inversionzones from elevation and slope maps. Docket inversion zones at elevationsfrom 250-599 meters were assigned a MAST of -2.O'Cand other inversion zones from 195-249 meters were assigned aMAST of -4.O'C.Decision-making rules were developed that would define areas with underlying permafrost. We then used these rules in a scriptfile with GRASS'S r.infer command to create the pcrmafrost maplayer. The following is the inference Tules table script file used tocreate a map predicling permafrost according to the predictedMAST map, clcvation, slope and equivalent latitude:IFMAP mast-plus104-10THENMAPRYP9(mast permafrost zone)IFMAP elevation 195-249THENMAPHYP l0-4.6(inversion zone)IFMAP elevation 250-599ANDIFMAP slope.rtc0-5THENMAPHYP 1 I(pocket inversion zone)IFMAP equivlat 67-90THENMAPHYP 12 (permafrost).GRASS evaluates each data cell of the map layers named inthis table (MAST, elevation, slope and cquivalent latitude) andcreates a new map layer in which the cell is placed into a category1128

according to the "rules" of the table. In this caw cells which fall intothe following categories would be put into the permafrost category in the new map Iaycr: MAST Categories from -6OC to O'C, elevationcategories from 195-249 meters, elevation categories above67'north latitude. (Notice that the category numbers into whichthese cells were placed were originally 9 through 10. These were laterreclassilied into category 2 while every other data cell in the newmap was reclassified into category I).The predicted MAST and predicted permafrost maps shown inthis study have areas of solid black%oloration which are not de-Lined in the legends. These are actually areas,in which the patternsare so dense that the area outlines or patter'ns create the blackeffect.FUTURE RESEARCHAlthough the Caribou-Poker Research Watershed has beenextensively studied by many researchers since it was first designatedas a research area in the early 1970's. this study is the first systematic,longterm attcrnpt to measure ground temperatures within thewatershed with the objective of defining permafrost distribution relationships.The sites selected for ground temperature monitoringin Caribou-Poker Creeks Research Watershed were selected to;ample a wide diversity of atmospheric, vegetative, and geologiccharacteristics, all of which influenct ground temperature patternsand therefore the distribution of permafrost. The ground temperatureinformation from this study combined with detailed informationon site characteristics will permit the analysis of the complexdistribution of permafrost in thc watershed.-The equipment neededto continually record air temperatures and near-surfacc groundtempetiatures at each of the borehole sites is expensive, and only recently have we been able to begin acquiring and installing thisinstrumentation, Three four channel data loggers were installed atthree sites(T-I, K-25, and K-20A) in 1990. They are recording airtemperature, temperature at two levels in the organic layer, and atthe organic/mineral soil interface. In August 1991 a 12 channellogger with senors ranging from "150 cm above the mineral / organicsoil interface to -140 cm below this surface, was installed at avalley bottom site near T-IA. These instruments, together withadditional sensors designed to measure other components of thesurface energy flux which will be installed as funding permits, willprovide a more comprehensive analysis of permafrost distributionwithin thc watershed.Future plans also include a deep (100 to 200 m) bore hole inthe valley bottom of Poker Creek is planned for the future. The intentis to penetrate the bottom of the permafrost in this area. ThePredicted Mean Annual Surface TemDerature

orehole will be Equipped with a multi-channel datalogger tocontinually record ground temperatures and complete surface energybudget instrumentation..The spatial delineation of permafrost / non-permafrostboundaries will require a complex model which includes the terrain,vegetation, and energy balance components projected to afine-mesh grid over the watershed. We plan to do this utilizingGeographic Infromation System (CIS) technology. The initial attemptswill be donc utilizing the more detailed air and surface temperatureinformation obtained beginning in 1990, together with thesubsurface temperature data which have been acquired beginningin 1986.."REFERENCESBilello, M A(1974) Air masses, fronts and winter precipitation incentral alaska. U.S. Army Cold Regions Research and EngineeringLaboratory, Research Report 319.Brown, R.J.E. and T.L.Pewe (1973). Distribution of permafrost inNorth America and its relation to the environment; a review.Proceedings, Second <strong>International</strong> <strong>Conference</strong> onPennafrost,pp.71-100.Collins,C.M., R.K.Haugen,and R.A.Krcig (1988) Natural groundtemperatures in upland bedrock terrain, Interior Alaska. Proceedings,Fifth <strong>International</strong> <strong>Conference</strong> on Permafrost,Trondheim, Norway, p56-60.Dingman, s.L. and Koutz F.R. (1974) Relations among vegetation,permafrost, and potential insolation in Central Alaska. Arcticand Alpine Research.(6),1,37-42.Haugen, R.K. and J.Brown (1978) Climate and dendroclimatic indicesin the discontinuous permafrost zone of the centralAlaskan uplands. Proceedings Second <strong>International</strong> <strong>Conference</strong>on Permafrost, pp 392-398.Haugen, R.K., C.W. Slaughter, K.E.Howe, and S.L.Dingman(1982) Hydrology and climatology of the Caribtiu-PokerCreeks Research Watershed, Alaska, CRREL Report 82-26.Haugen, R.K., S.1.Outcalt and J.C. Hark, (1983) Relationships between estimated mean annual air and permafrost temperaturesin north-central Alaska,Proceedings, Fourth <strong>International</strong><strong>Conference</strong> on Permafrost,pp.462-467.krgenson, M.T., C.W.Slaughter, and LA. Viereck (1984) Reconnaissancesurvey of vegetation and terrain relationships in thePoker-Caribou Creeks Watershed, Central Alaska.Unpublished report and map. Institute of Northern Foreltry,U.S.Forcst Service, Fairbanks, Alaska, 99701. On file at INP.Koutz, F.R. and C.W. Slaughter (1972) Geologic setting of theCaribou-Poker Creeks Research Watershed, Interior Alaska.Unpublished memo, CRREL.Morrisscy,L.A., L.L.Strong and D.H.Card (1986) MappingPermafrost in the Boreal forest with Thematic Mapper satellitedata. Photogrammetric Engineering and Remote Sensing.V.52, p.1513-1520.Rieger, S., C.E. Furbush, D.B. Schoephorster, H.Sumrnerfield, andL.C. Gieger (1972) Soils of the Caribou-Poker Creeks ResearchWatershed, Alaska.CRREL Technical Report 236.Viereck, L.A., C.T.Dyrness, and A.R. Batten (1982), the 1982reas of Predicted Permafrost:aribou-Poker Creek Research Watershed, Alaska, USA1f-J ....,.:: Predicted Permafrost ' @ NO Permafrost1130

Revision of Preliminary Classification for Vegetation of Alaska.Unpublished report. Institute of Northern Forestry, U.S. ForestService, Fairbanks, Alaska. 72 pp. On file at INE.Vicreck, L,A., C.T. Dyrness, and A.R. Batten (1986) The 1986Revision of Alaska Vegetation Classification. Unpublished report.Institute of Northern Forestry, US. Forcst Scrvicc,Fairbanks, Alaska.141 pp, On file at INF.Vogel, T.C. and C.W. Slaughter (1972) A prclirninary vegetationmap of Caribou-Poker Creeks Rcsearch Watershed, interiorAlaska. CREEL Technical Note(unpub1ishcd).Wahrhaftig, c.(1965) Physiographic Divisions of Alaska. U.S. GeologicalSurvey Professional Paper 482.1131

CIRCUMAKCI’IC MAP OF PERMAFROST AND GROUND ICE CONDlTIONSJ.A. Heginbottom’, J. Brown2, E.S. MclnikoG, and O.J. Ferrians Jr4,’Geological Survey of Canada, Ottawa, Canada,2Chairman, IPA Editorial Committee, Arlington, VA.Slnslitulc for Hydro eology and Engineering Gcology (VSEGINGEO), Moscow,‘US. Geological Survey, Anchorage, AK,The <strong>International</strong> Permafrost Association undertook to compile and publish this circumarctic permafrost map, inrcsponse to a recognized need for a single, unified international map to dcpict the distribution and properties ofpermafrost in the Northern Hemisphere at a scale that would be useful to both permafrost and non-permafrostspecialists conccrned with global climatic change, resource development in polar regions, and protection of thcenvironment. The map shows the estimated permafrost exlcnt by percent area (90-lOO%, 50-90%, 10-SO%, 20%, 10-20%,

continuous and discontinuous zones (e.g. Brown, 1973). Some mapsin this group show areas of alpine or mountain permafrost (e+Brown, 1967; Gorbunov, 1978) and subsea permafrost (Mackay,1972; P&v< 1983). A second set of maps show specific attributessuch as thickness and temperature of permafrost (Judge, 1973;Ershov, 1989) and the distribution of geomorphic features indicativeof the Occurrence of ground ice, of frozen ground, or of the formerextcnt of frozen ground (Popov et al., 1966), including pingos(Hughes, 1969), ice-wedge polygons or ice wedges (Shumskiy andVtyurin, 1966), and ice-wedge casts (Williams, 1969). A third groupof maps relates permafrost conditions to environ-mental conditionsincluding temperature (Crawford and Johnston, 1971), extent ofglaciation (Hughes, 1973), and geology, hydrology or vegetation (e&Fcnians, 1965; Fotiev, 1978, Bliss, 1979). Complex maps containenvironmental, permafrost and ground ice information. Such mapshave more often been prepared in the former Soviet Union (e.gBaranov, 1956, 1965, 1982; Kudryavtsev et al., 1978; Melnikov, 1966;Fotiev et al., 1978; Vtyurin, 1978; and Ershov, 1989).PRINCIPLES AND METHODS OF COMPILATIONScandinavia and central Europe, the Cordillera of North Americaand mountainous regions of southwestern, central and eastern Asia.A Lambert Polar Azimuthal Equal Area map projectian, centred onthe north pole, was selected for the map, so that regions of similarlatitude would have comparable areas. Th~s feature of theprojection was considered important for global changeconsiderations related to potential changes in areal extent of permafrostand ground ice distribution and volumes. A base map wascomputer generated by the U.S. Gcologicsl Su~vey, Reston, Virginia,USA, using existing World Data Bank I1 data bases for coastlines,drainage, the latitude and longitude grid, and internationalboundaries. Glaciers and ice caps for North America, includingGreenland, are based on digitized files from Canadian sources.Bathymetry was handscribed aftcr Perry and Fleming (1986).Although not currently available in digitized form, it is hoped tohave the map attributes available in a digital form in the near future.The Canadian contribution was prepared from computerised databases (Heginbottom and Dubreuil, 1993).MAP DFSIGN AND THEMATIC CONTEEA major problem in the compilation of maps of complex naturalphenomena over very large areas is the variation in the level andaccuracy of the available data and information. The principles andmethods used in compiling this circumarctic map recognize andaccommodatc this difficulty. The variations encountered relate bothto disparities in the level of field observations, and to variations inenvironmental and geological factors which control the distributionand attributes of the permafrost and ground icc.The mapping strategy originally proposed was to delineatelandscape units, employing a common physiographic classification ofthe Northern Hemisphere, and to assign legend attributes in eachidentifiable unit. This approach was not feasible, however, since nophrjiographic map of the northern fegions of the earth, at thedesired scale and level of detail, was readily available and it wasbeyond the scope of this project to prepare one. It was decidedtherefore to use existing physiographic or landscape maps in each ofthe three major national permafrost regions (Russia, Alaska, andCanada). In preparing this map no attempt was made to conductnew field studies and every attempt has been made to use all readilyavailable published and unpublished information.Map units in Alaska are based on the 1965 map "PhysiographicDivisions of Alaska" (Wahrhaftig 1965) and contain informationbeing used to revise the 1:2,5500,000 map of Alaska (Fenians 1965).The Canadian contribution utilizes the 1967 map "PhysiographicRegions of Canada" (Bostock, 1970) as a base map and containsmuch of the information presented on the new permafrost andground ice map prepared for the 5th edition of the National Atlas ofCanada, (Heginbottom and Dubreuil, 1993, in prep.). Russian mapunits are derived from the geosystems or landscape approachdescribed by Melnikov (1988), in which natural geosystems are delineatedaccording to common relief, vegetation, soil and soil-formingmaterials and climate. Units for China and Mongolia are based onrecompilations of the maps by Shi and Mi (1988) and from Sodnomand Yanshin (1990), respectively. Existing information for theNordic countries, Greenland, and other mountainous regions ofEurope and Asia were modified and compiled from numerouspublished and unpublished sources, with the assistance of regionalspecialists, as listed in Table 1.A preliminary legend wis agreed to by the principal authors inAnchorage, Alaska, in September 1991 and revised by the sameauthors in Ottawa, Canada, in April 1992. The map scale of1:10,oMl,oM) was selected so that all regions of permafrostoccurrence in the Northern Hemisphere could appear on a single,map sheet. The map extends southward to WN latitude andihcludes mountain or high altitude pcrmafrost conditions in Tibet,The map is a comprehensive summary of permafrost and groundice conditions in the northern circumpolar region. The basic mapunits are described in terms of the extent of permafrost, the quantityof ground ice, and the relative abundance of large1 bodies of groundice: pingos, ice wedges and bodies of massive ice. Note that no distinctibnk made between massive ice of intrasedirnental origin(Mackay and Dallimore, 1992) and massive ice rcculting frum theburial of ice formed at the ground surface, such as buried glacier orriwr ice. Information on permafrost thickness and ground temperaturesis given for selected localities across the region.Permafrost ExtentThe general distribution of pcrmafrost and ground ice is dividedfint into two broad classes, based on regional elevation,physiography and surface geology. Group 1 comprises areas oflowlands, highlands and intra- and inter-montane depressionscharacterized by thick overburden, wherein ground ice is expected tobe generally fairly extensive. The semnd group covers areas ofmountains, highlands, and plateaus characterized bv thin overburdenand emed bedrock, where generally lesser amounts of ground iceare expected to occur. For the purposes of this map compilation,thick overburden is defined as being greater than 5 lo 10m.The estimated extent of permafrost in each map unit is presentedin four classes, based on the percentage of the ground that isunderlain by permafrost (continuous, 90-1M)%; discontinuous, SO-90%; sporadic, 10-501; and isolated patches of pcrmafrost, 0-10%).Areas generally free of permafrost are also indicated. For areas ofphysiographic class 1, the colour scheme uses tones of purple forcontinuous permafrost, blue for discontinuous permafrost, green forspradic permafrost and yellow for areas where permafrost occurs inisolated patches. For areas of physiographic class 2, the colourscheme uses tones of brown for continuous permafrost, orange fordiscontinuous permafrost, gold for sporadic permafrost and yellowfor areas where permafrost OCCUIS in isolated patches.Areas of subsea and relict permafmt, based on both direct andextrapolated measurements, are shown by stippling. The 100 anduxhn submarine contours arc shawn to indicate possible Occurrenceof subsea permafrost on the continental shelves of the Arctic Basin.For Russia, map units known to contain cry~pegs (layeis of unfrozenFund with high salt content), are rnapptd beneath land areas.Ground IceThe relative abundance of ground ice in each map unit ispresented in the form of qualitative estimates of thc percentage ofice in the upper 10 to U)m of the ground. These estimates include1133

Table 1: Regional Contriiutms to tbe proiectDRegionRussia, MongoliaGreenland, North Atlantic Islands,Fenno-ScandiaCentral EuropeSouthern EuropeRumaniaMountains of Central andSouthwestern AsiaChinaContributorsGF. Gravis, LA. Konchenko, and L.N. Kritzuk, Committee of Geology; and K.A. Kondrat kva andS.F. Khrusky, Faculty of Geology, Moscow State University, Russia.H.J. Akerman, University of Lund, Sweden, and Matti Seppiji University of Helsinki, Finland.W. Haeberli, Swiss Federal Institute of Technology, Zurich, Switzerland, and L. King, Justus-LiebigUnivrrsity, Giessen, GermanyF. Dramis, University of Camerino, and C. Smiraglia, University of Milano, Italy.A. Kotarba, Polish Academy of Sciences, Krakow, Poland.A.P. Gorbunov, Alma-Alta, Kazakhstan.Guo .Dongxin and Qiu Guoqing, Ianzhou Institute of Glaciology and Geocryology, Lanzhou, China.the volume of segregation ice, injection icc and reticulate ice. Threeclasses are used for ground ice contcnt (high, >2070; medium,10-20% and low, < 10%) in arcas in physiographic class 1, that is forareas of generally thick overburden. For arcas of generally thinoverburden (physiographic class 2). only two classcs of ground ice 'arc mapped, medium to high (> 10%) and low (.:lo%), due in partto paucity of data.Gradations in the map colours reflect these distributions, withshades of each colour denotihg map units with more ground ice andtints indicating map units with less ground icc.'The distribution and relative frequency of known occurrences oflarge identifiable underground ice bodics are treated separately andshown by symbols. Ice bodies included in this manner comprise theice cores of perennial fmst mounds, especially pingos; ice wedges;and bodies of massive ice, generally tabular in shape. A simple,three step scale of "abundant, sparse, and absent' is used. Surfaceice features, including ice caps, glaciers and very large icings, areshown by patterns and symbols.Ground Temaerature and Permafrost Thickness,V:llues and ranges of mean annual ground temperatures (Cclsius)and permafrost thicknesses (metres) are shown for selccted localitiesacross Ihe mapped area. These are based either on measured valuesor extrapolated observations. Ihe placcmcnt of the values in themap unit generally corresponds to the geographic location of themeasurements.Landscape ClassificalionFor Russia, six principle morphogenetic landscape groups areidentified: lowland plains, high plains, inlra- and inter-montanedepressions, plateaus or flat highlands, ridges, and nmantains.Overall, for the Russian portion of the map, 19 morphogenetic typesof landscapes are identified including, 15 within the firs1 thrcegroups, where accumulative sediments of diffcrcnt origins are welldeveloped. No landscape units are shown withln the categories ofplateaus and mountains. Erosional or denudational landscapecategories have four types.Major litlrological classes present in the upper 10 to 20m of theground are divided into unlithificd and lithified material; the formerincluding peat, clay and silt, sand, and coarse claslic deposits ordebris and the latter comprising soluble rocks (eg: limestone ordolomite), insoluhlc racks and undifferentiated rocks.Transects of the Permafrost Region,Eight north-south oriented transects of the permafrost region areshown as insets to the map. The transects illustrate the major characteristicsof the permafrost body and its ground ice conditions on ahemispheric basis, and in an idealized but representative manner.Alaska: The 1280-km long trans-Alaska pipeline mute providesextremely useful data from a series of shallow borehole extendingfrom Prudhoe Bay in the north to Valdez at the southern terminusof the pipeline. Additional data from deeper holes has been used,where it is available - mainly in the vicinity of Yrudhoe Bay.Mackenzie Valley: The various Mackenzie Valley pipeline andhighway routes, both proposed and constructed, have provided a vastdata base for mapping permafrost. These data have been used incompiling this transect. Deeper boreholes, in the Beaufort Sea andMackenzie Delta areas were also relied on.Central Canada and Arctic Islands: In the Arctic Islands, thetransect is based on data from numerous boreholes drilled forhydrocarbon exploration. On the northern mainland, it is based ondata from the few mineral exploration boreholes which have beendrilled, along wjth a limited number of shallow geotechnica1,borings.Eastern Canada: Thk transect is based on limited data fromboreholes drilled at a few sites for mineral exploration andproduction purposes.European Russia: The transect has been compiled from datafrom a number of mineral exploration holes.West Siberia: Many borcholes have been drilled for purposes ofhydrocarbon exploration throughout west Sibctia. This transect isbased on data from these boring, supplemented with data fromdeep geophysical soundings.Fast Siberia: The transect is based on data from boreholes drilledfor hydrocarbon exploration, mineral exploration (especially fordiamonds) and geological structural research on the SiberianPlatform.Qinghai-Xizang (Tibet) Plateau: This transect is based onresearch carried out along the line of the highway from Golmud,Qinghai Province, to Lhasa, Xizang Province.Boundaries, Imend and SourcesBoundaries of permafrost and ground ice map units are shown bya solid line where they are well defined and follow a physiographicunit boundaly. Where unit boundaries are gradational or are estimated,a dashed line is used. The approximate position of thenorthern limit of trees (compiled from several National GeographicSociety nlapsj is ,shown, since this major change in vegetation has

important implications far ground temperatures and other ecologicalparameters.The explanation of the conventions of the map, the colour schemeand the symbols is given in the map legend. A subsidiary legendprovides information on the landscapc units used for the Russiansector of the map. The principle burces relied on in thecompilation of the map are listed.REGIONAL DISTRIBUTION OF PERMAFROSTAND GROUND ICE CONDITIONSFor the first time, a permafrost map of the entire circumarcticregion has been compiled using a common legend, so thatpermafrost and ground ice conditions can be accurately evaluated,thus enabling regional and global comparisons to be made. Thedistribution and characteristics of permafrost and ground ice arcbriefly described in thc accompanying report, according to majorregions of the Northern Hemisphere Gable 2).The map illustrates how the regional distribution of permafrostand the nature and extent of ground ice within the permafrostregion of the northern hemisphere vary not only with latitude andaltitude, but also in response to differences in climate, topography,bedrock geology and surficial geology. Quaternary history, withalternating episodes of glaciation and deglaciation, and phases ofmarine and lacustrine submergence and emergence of the land, alsohad profound effect on the nature and distribution of bothpermafrost and ground ice.As has long been known, there is considerable variability in thequantity, distribution and reliability of basic data on permafrost andground ice. Compilation of the map has made this particularlyapparent to the authors. While these variations reflect realdifferences in geocryological conditions, they are mainly the result ofthe different amount of research undertaken in different regions,differences in accessibility, natural exposures, resource developmentactivity, and different national philosophies in the conduct ofnational surveys of natural phenomena.Plans to make the map and data available in digital form arebeing developed.Many individuals and organizations have contributed personneland other resources to the compilation of this map. Several individualsdeserve special recognition for their thoughtful involvementearly in the project, including, in particular, Ray Kreig and YuriShut, Kreig Associates, Anchorage, Alaska; Vladimir Solomatin,Moscow State University; Stanislav Grcchishchev, VSEGINGEO,Moscaw; and F.E. Nelson, Rutgers University, USA. The assistanceand support of J. Akerman, W. Haebcrli and L. King in obtainingdata has been especially helpful. Notwithstanding the valuable inputfrom those named, plus many others, the presentation of theinformation and mapping approach utilised are, the sole responsibilityof the authors.REFERENCESTable 2. Regions Used for Descriptions of Pcmdnxt and Groundla ConditionsBaranw! I.Ya., ed. 1956. Geocryological map of the USSR MainDepartment for Geodesy and tsrtography (GUGK), Moscow,North Americascale l:lO,MH),OOO (Russian).Canada (7 regions)Earancw, I. Ya. 19S9. Geographical distribution of seasonallyUSA - Alaska (3 regions)fmen ground and permafrost. National Research Council ofConterminous USACanada, Technical Translation N0.1121; Ottawa.MexicoSub-Sea PermafrostNorth AtlanticGreenland IIcelandSvalbardFenno-SEandiaAsiaBaranov, I. Ya., ed. 1965. Principles of geocryological zonation ofthe permafrost region. Nauka, Moscow, 152 pp. (Russian).Baranov, I. Ya. 1982. Geocryological map of the USSR. MainDepartment for Geodesy and Cartography (GUGK), Moscow,scale 1:7,500,000.Baulin., V.V., cd. 1982. Map of geocryological regions of the WestSiberian Plain, USSR Ministry of Geology, VSEGINGCO, scale~:1,500,000, 4 sheets (Russian).Russia (8 regions)Black, R.F. 1954. Permafrost a review. Bulletin of theMongoliaKorea and JapanCentral AsiaSouthwest.AsiaCentral and Alpine EuorpeGeological Society of America, Vol. 65, pp, 839-856.Bliss, L.C. 1979. Vegetation and revegetation within permafrostterrain. i~ Pqeedings of the Third <strong>International</strong> <strong>Conference</strong>on Permafrost. National Research Council of Canada, Ottawa,Vol. 2, pp. 31-50.Bostock, H.S. 1970. Physiographic regions of Canada. GeologicalSurvey of Canada, Map 1254A, scale 1:S,OOO,ooO.CONCLUSIONSBrown, R.J.E. 1967. Permafrost in Canada. National ResearchCouncil of Canada, Publication 9769, and Geological Survey ofIn preparing this new map, conclusions were drawn about our cur-Canada, Map 1246A, Ottawa, scale 1:7,603,200.rent knowledge on the distribution of permafrost and ground ice inBrown, R.J.E. 1973. Permafrost. National Atlas of Canada.the Northern'Hemisphere and some future information needs. TheDepartment of Energy, Mines and Resources, Ottawa, Plate 11-present map differs from earlier maps of permafrost, first, because it12, scale 2:15,OOO,oM3.provides information for the whole. of the northern circum-polarBrown, HJ.E. 1978. Permafrost = Phgelisol [Canada].region and, therefore, for effectively all of the northern hemisphereHydrological Atlas of Canada, Department of Fisheries and thqpermafrost region. Secondarily, the map shows for the first time,Environment, Ottawa, Plate 32, scale l:lO,OOO,~.information on the distribution and nature of ground ice in aCrawford, C.B,, and G.H. Johnston. 1971. Construction onsystematic manner. All this information is presented in relation topermafrost. Canadian Geotechnical Journal, Vol. 8, pp. 236-251.landscape or physiographic units, which should facilitate its use inBrshov, E.D., ed. 1988,1989. Geocryology of the USSR. Nauks,other, global, hemispheric or regional studies of the interactionMoscow, S volumes (Russian).between the lithosphere, the crymphere and other environmentalparameters.1135

Ferrians ,O.J., Jr. 1945. Permafrost map of Alaska. U.S.Geological Survey, Miscellaneous Geologic Investigations, Map1445, scale 1:2,5W,M)(3.Fotiev, S.M. 1978. Effect of long-term cryometamorphism of earthmaterials on the formation of ground water.Proceedings ofthe Third <strong>International</strong> <strong>Conference</strong> on Permafrost, NationalResearch Council of Canada, Ottawa, Vol. 1, pp. 181-187.Fotiev, S.M., N.S. Danilova, and N.S. Shevleva. 1978. Zonal andregional characteristics of yrmafrost in central Siberia. fiPcmafrost -- The USSR Contribution to the Second<strong>International</strong> <strong>Conference</strong>, National Academy of Sciences,Washington, D.C., pp. 104-110.Gorbunw, A.P. 1978. Permafrost investigations in high-mountainregions. Arctic and Alpine Rcsearch, Vol. 10, pp. 283-294.Hcginbottom, J.A. 1984. The mapping of permafrost. CanadianGeographer, Vol. XXVIII, pp, 78-83.He@nbottom, JA., and L.K. hdburn. 1992. Permafrost andground ice conditions of Norlhwestern Canada. GeologicalSurvey of Canada, Map 1691A, scale 1:1,ooO,M)O.Hcginbattom, J.A., and "A. Dubreuil. 1993. A new permafrostand ground ice map for the National Atlas of Canada. @ProceedinF, Sixth <strong>International</strong> <strong>Conference</strong> on Permafrost,Reijing, vol. 1, pp. 255-260.Heginbottom, J.A., and M-A. Dubreuil. (in prep.) Canada --Permafrost. National Atlas of Canada, 5th edition, scale1:7,500,000, Plate 2.1 (MCR 4177).Hughes, O.L. 1969. Distribution of open system pingos in centralYukon Territory with respect to glacial limits. GeologicalSurvey of Canada, Ottawa, Paper 69-34,8 pp.Hughes, T. 1973. Glacial permafrost and Pleistocene ice ages. hPermafrost -- the North American Contribution to the Second<strong>International</strong> <strong>Conference</strong>, National Academy of Sciences,Washingtan, D.C., pp. 213-223.Judge, A.S. 1973. Deep temperature observations in the Canadiannorth. Permafrost - the North American Contribution to theSecond <strong>International</strong> <strong>Conference</strong>, National Academy of Scicnces,Washington, D.C., pp. 3540.Kudryavtsev, VA., KA. Kondrat ha, and A.G. Gavrilov. 1978.Geocryulogical map of the USSR. General Permafrost Studies;Materials for the Third <strong>International</strong> <strong>Conference</strong> on Permafrost.Nauka, Novosibirsk, scale 1:2,500,000 (Russian).Mackay, J.R. 1972. 'he world of underground ice. Annals of theAssociation of American Geographers, Vol. 62, pp. 1-22.Mackay, J.K. and S.R. Dallimore. 1992. Massive ice of theTuktoyaktuk area, western Arctic coast, Canada. CanadianJournal of Earth Sciences, Vol. 29, pp. 1235-1249.Melnikov, E.S. 1988. Natural geosystems of the plaincryolithozones. & Permafrost, Fifth <strong>International</strong> <strong>Conference</strong>,Proceedings. Tapir Publishers, Trondheim, Nomy, vol. 1,pp. 208-212.Melnikw, P.I. 1966. Geocryological map, Yakustkoi A.S.S.RAkademia Nauk SSSR, Moscow, scale l:S,MW),Mw) (Russian).Nikiforoff, C. 1928.. The perpetually frozen subsoil of Siberia. SoilScience, Vol. 26, pp. 6161.Perry, RK., and H.S. Fleming. 1986. Bathymetry of the ArcticOcean. Geological Society of America, Boulder, scale1:4,704,075.PBX4 T.L. 1982. Geologic hazards of the Fairbanks area.Division of Geological and Geophysical Surveys, Fairbanks,Special Report 15, 109 pp.P&< T.L. 1983. Alpine permafrost in the contiguous UnitedAlaskaStates: a review. Arctic and Alpine Research, Vol. 15,p ~ 145-156. .Pop, A.I., S.P. Kachurin, and N.A. Grave. 1966. Features of the 'development of frozen geomorphology in northern Eurasia. & ' ,Proceedings -- Permafrost <strong>International</strong> <strong>Conference</strong>, NationalAcademy of Sciences, Washington, D.C., NRC Publication 1287,pp. 481-487.Pop, A.I., et al. 1985. Map of cryolithology of the USSRFaculty of Geography, M.V. Lomonosw University, Moscow,scale 1:4,OOO,oM), 4 sheets (Russian).Pop, A. I., et al. 1990. Cryolithological map of North America.Faculty of Geography, M.V. Lomonosov University, Moscow,scale 1:6,000,000, 4 sheets (Russian).Rapp, A., and L. Amersten. 1969. Permafrost and tundra polygonsin northern Sweden. & The Periglacial Environment, T.L. P&

Permafrost Studies in GreenlandHenrik Mai and Thorkild ThomsenArctic Consultant Group,Sortemosevej 2 DK 3450 AllerCd Denmarka Greenland Power CompanyPilestrffde 52 P.O.Box 2128 DK 1015 Copenhagen K DenmarkINTRODUCTIONData collected for mainly hydro-power purposes since the late seventieshave been used €or a new permafrost of south-west map Greenland and topermafrost related studies to hydraulic studies.Mean yearly air- and rocktemperature have been calculated. A linearregression of airtemperature have been established for an overall temperatureprofile on the westcoast. Rocktemperature have been used in measureddepth., The drawn permafrost map have been drawn with a limit to discountinouspermafrost with airtemperature -0,5 to -O,l"C, and continous permafrostwith airtemperature -4 to -5°C.A study of hydraulics structures in permafrost at the location ofPakitsup Aku. is presented with respect of temperature in rock and watersurroundings of water filled tunnels.-:The long-term energy political intentionare to base a major part of the energysupply system in Greenland on hydropower.This objective have resulted inconsiderable hydrological and constructionallyinvestigations also in the field ofpermafrost.The investigations started in the endof the seventies and had for the time beinga culmination in the middle of theeighties. Data collected since that periodeare the basis for detailed temperatureand permafrost studies.PERMAFanST_"-ENLANDAs discontinous permafrost occur interrain with a mean air temperature ofless than about -0,5"C, it is obvious thatit plays a significant role in the terrestrialportion of the hydrological cycle.The precipitation and the radiation .energy balance are importent controllingaspect of the arctic bydrological/permafrostregime. It has influence on the supplyof energy to evaporation and throughits supply of heat to the ground that istantamount to spreading of the arctic layer.Because of the tilt of the axis ro- oftation of the Earth in high latitudes solarenergy is received a at low angle, sothe net energy on each square meter isrelative much less than in the tropics. Atlatitudes higher than 66,7'N (the Arcticcircle) no direct energy is received ata11 from the sun for at least of part theyear, at the same time heat is radiatedfrom the surface perpendicularly. spa- intoce.As a result, there is a strong loss netof energy from high latitudes. This lossmust be made by transport of heat from Lowlatitudes through ocean current and atmosphericcirculation.These factors make the regional climateof polar areas very dependent ontransport of heat from low latitudes- the heat trapping effect of clouds andgreenhouse gasesChange in global energy balance or circulationwhich have an effect on these mechanismsof heat delivery or heat loss,can thus have an exaggerated effect on theclimate and environment of high latituderegions.The various parameters will be documentedin the following:T e e - .Available temperature data are shown intabel 1. The stations with temperaturevalues in different depth are locationswith rocktemperature thermisters. The calculatedyearly mean temperature are nottime consistent for all stations but theshortest time period is 5 years.The inland stations along the westcoastwith airtemperature censors (40 stations)are plot with yearly average temperatureagainst northern latitude, figur1. The databasis are not for all stationsconsistent in time.The regression curve for airtemperaturewith height along the west-coast "AIRcor"curve is iterated found with a correlationcoefficient of 0.91. The resulted temperaturegradient (average lapse rate) is0.6'C/100 m.1137

LOCATIONcj PaarnlutH 62'13'N.W13'WIJ62'13'N,J9'17'W63"15'N,50'21'WK PhgelluarsaruscqL M'OI'NJO'WWM Nu&N 65'09N.SO'WW0 M'IR'N.51'1B'WP Man~ilsoq0 SlsirnfulR KangcrlusruaqS 67'1lI'Nj3'lh'Wv ~asiglantiuarW IlulissatX OeqenarsuaqY h9'2Y'N50eI?'WZ h9'27N.50'r'WTabel 1IVALUE("C)+1.7+o 7+l.3-I 64.7+I 1+I I-4 0-1.8-1 8-1.7-1.9-0.5-2 4-u. I+o I-2.5-0 n-n 34 4-6 540 3+[I5-1.0-4 Y-3 5-3.Y-1.3-6.8-4 2-3.0-9 d-h.?-5 44 1-6 9-I h-3 9-5 Y-5 4-2 2-l.Y-1 9-2.2-4 8-7 0-3 8-3.5-3 5-3 Y-3 h-1 1-4 n-4 0-6 4-3 3-4 d-4.5-Data basis for permafrost map onfigur 3.Data from Denmark MeteologicalInstitute and Greenland HomeRule Hydro-climatological databaseby Greenland Fieldinvestigations.4 4. .-4 ' 60 62 64 66 68 70 712Latitude O NFigur 1.Airtemperature regressioncurvefor inland stations in West-Greenland.The present precipitation conditionshave for Greenland as whole been presentedby Ohmura and Reeh (1991), se figur 2.Alltogether, 251 pits and cores havebeen used to calculate the distribution ofthe annual accumulationthe ice sheet.The precipitation data at coastal siteswere collected at 35 meteorological st+tion.The main features of the precipitationdistribution are a strong longitudinalgradient exist in southein Greenland,south of 6S"N on the west and south of70'N on the east coast, Ohmura and Reek(1991)The mean annual precipitation for a11of Greenland is 340 mm w.eq. The mean accumulationon the Greenland ice sheet isestimated at 310 nun w.eq.For more detailed local variability onthe west coast it is nessary to establisha more accurate map taking into accountthe dominant south-north and west-eastprecipitation gradient.n balance.Measurement of components af the radiationbalance equation for the arctic areaare only found to a limited extent, Prowseand Onunanney ed.(1990) and Thornsen (1991).1138

Figur 2. Yearly total precipitation in m, after 0-aand Re& (1991).’ 1139

PERMAFROST MAP

50'20' ' Mol0On the basis of the shown temperatureand precipitation distribution and also agreat deal of subjective knowledge of hydrologicalinformation a permafrost maphave been drawn, as shown on figur 3.Besides the contour of temperature parameters,there are shown a calculated inlandairtemperature at sea-surface basedon the temperature/latitudes correlationon figur 1.The shown permafrost boundaries aredrawn with a limit to discontinous permafrostwith airtemperature -0.5 to -1.0-Cand to continous permafrost with airtemperature-4 to -5'C.xL!lwsA study of hydraulics structures inpermafrost have been carried out on theexistingcknowledge from temperature andconstructionally investigations. Any hydraulicstructure in permafrost area isbased on knowqedge of the initial temperatureconditions in the surrounding rockprior to establishment of the structure.Scarce data requires drastic assumptiansto be made. In recent years, finiteelement and finite difference models havebeen used to investigate temperature distributionresulting from freeze-thaw cyclesIn rock and steady state and transientsolutions have been derived for stratifiedrock/soils, Irmen Christensen and Mal( '88).For the purpose of having waterfilledtunnels in permafrost detailed studieshave been carried out to investigate temperatureconditions in water-rock interfaceboth in an arctic lake and in a tunnelfor hydro-power plant, Mal (1987 - a)and Mai ed. (1993 - first draft).Scheme for a hydro-power plant in permafrozenrock have made it necessary tostudy the thermal regime around the watertunnels and in the water body In the reservoir.The studie have been carried outat the basin Paakitsup Akuliarusersua(69"29'N,5O0l8'W), figur 4.The hydrological regime is dominated byablatlon water from the Icecap. The lakesin the area 187 and 233 (the level abovesea) will effect,the temperature in thesurrounding rock and give the initial conditionsfor resulting temperature developmentin the water-tunnels.The lakes have a relative stabilizedtemperaturecondition, and will never gobelow 0°C. Compared to the average airtemperatureat the location about -6 to -7'C .this will be a much warmer body. The lakesare characterize as "arctic lakes", wheretemperature never get over +4'C, gee figur5."-Figur 4,50° 20' 50- 10Drainage basin at Paakitsup Akuliarusersua,after H.Thomsen,Geological Survey of Greenland1141

0"""_Active layer*""SeasonoilycryoticFigur 5.Lakg temperature profile in thearctic lake 187 at PaakitsupAkU .Figur 6. Rocktemperature gradient at PaakitsupAku, surface level 237 m.above sea.Fresh water have its maksimum densityat +doc, which means that the warmest wateralways will remain in bottom and thecoldest water in the top. Lake 187 separateaa little from lake 233 due to the glaciergoes direct into the water.in lake187. This means that lake 187 in generalare colder than lake 233.In lake 187 the summertemperature inthe surface will be around 1.5 to 1,8'C.,and in lake 233 around to 3 4°C. In wintertemperaure drop at the surface to O'C, butstill a little warmer in Lake 233.The bottomwater in the lakes reflectsthe summertemperature and can be 1 to 3'Cwarmer than the'surface temperature. TOgetherwith temperature profile measurementain the lakes intensive temperatyremeasurements in the rock have taken place.On one location situated 237 m abovesea level temperature has been measureduntil 250 m below surface. Rocktemperaturemeasure at the atation is shown on figur6. The gradient is calculated to0,0167'C/m ll'C/60m).With the known temperature and rockcondition the hydro-power scheme was chosenwith a lower level headrace as shownon flgur 7 to get as little impact onrocktemperature below O'C as possible.With the topographic information and ' ithe temperature condition in the water androck a dynamic model was runned to simulatethe temperature field close to the tunnel,with the power plant in operation.The model should also be able to describepossible of ice in the water conduct duringdiscontinous operation Mai (1987 - b)and Mai ed. (1993).The tailrace of the power-plant terminateswith a submerged outlet in the fiord.1142

1REFERENCES.Fiyur 7. Hydro-power scheme at Paakitsup Akua, with lowlevel headrace.Irmen Christensen, M. and.Mai, H. (1988):Subsurface temperature predicitationfor hydropower developments. The 7thNorthern Research Basin symposium/workshop. Ilulissat, June 1988.Mai, H.(1987 - e):Hydropower tunnels in Permafrost, Conf,Underground Hydropower Plants Oslo,june 1987.Mai, H., Stigebrandt, A. and Bergander, 8.(1987 - b):Submerged outlet into,a fiord from apowerstation in Greenland. Conf. Hydropower in Cold Climates, Trondheim, july1987.Mai, H. ed. (1993) - draft:Permafrost studies for Hydro-Powerpermafrostgruppen: ArCtiC ConsultantGroup,LIC-Consult end VBB-VIAK.Prowse, T.D. and Ommanney, C.S.L. ed..( 1990) :Northern Hydrology Canadian Rerspective,NNRI Science Report No. 1. EnvironmentCanada.Thornsen, T. (1991):Arctic Hydrology in Greenland - Perma-frost hydrology and data technology.Arctic Hydrolcgy - present and futurotasks. Seminar Longyear byen Svalbard.Norwegien National Committee for Hydrology.Report NO. 23 - Oslo 1.991.1143

SNOW AND PERMAFROST IN THE TIAN SHAN MOUNTAMSHu Ruj and Ma HongXinjiang Institute of Geography, Chincue Acidemy of Sciences, U mqi 83001 1, ChinaThe existence of snow and permafrost is a result of a cold climate in high latitudes and high mountainareas, and they affect on the economic and social developments dircct1y.h the Tian Shan mountains, theseasonal snow cover within the permafrost mne plays an important role in controllin8 he growth andthaw of unstable permafrost, and it's distribution exhibits almast a sama pattern in areal extent with thatof permafrost. Snow and permafrost are interdependent, co-existmce, and play a positive role in balancingthe ecological environment in the mountain areas.INTRODUCTIONSnow and permafrost are extensively distributed in the TianShan mountains. As for snow, reliable records were kept in literatureas early as 1,200 yeras ago. Since late the 1950%. the ChineseAcademy of Sciences has organized many integrated surveys, inveatigations,and fixed-position observations of snow and permafrostin the Tian Shan mountains and has achieved a lot of importantscientific achicvments.DISTRIBUTION OF PERMAFROST IN THE TIAN SHANMOUNTAINS"It can be seen from Fig.1, the distrihtion of permafrost in theTian Shan mountains shows almost a same pattern with the distributionof SHOW cover. The lower permafrost limit in shady slopes isgenerally lower than that in sunny slopes, i.e. about 2,700 m inshady slopes and 3,100 m in sunny slopes, with a difference of 400m in altitude. The lower limit of permanent snow is about 3,600 ma.s.1. in shady slopes and 4,400 m in sunny slopes, the dimerencc inaltitude is about 800m. Thc altitude differences between the lowerlimits of permanent snow and permafrost varies grestly from onelocation to another, with a distance ranging from 870 to 125h.Those suggest that altitude is a dominant factor controlling the dietribution of snow and permafrost in the Tian Shan Mountains.The area of permafrost is approximated at 6.3 X 10 Km (QiuGuoqing ,1983), while the area of permanent snow cover is only 1x 10 Km (Hu Ruji; 1989). The large difference in coverage area between them indicates a more dependenoe of existence ofpermanent-snow upon the raising altitude.Seasonal frozen soils are also widely distributed in the TianShan mountains, and they are almost totally covered by snow coverduring the winters. The existence of them affecbs on landscape andrsological environment of the area significantly, and this effect isFig.] Distribution of permafrost in the Chinese Tian Shanmountains1144

largely determined by the intensity of interaction between snow SEASONAL SNOW WITHM THE PERMAFROST ZONE IN0cover and the seasonal frozen soils. Kii " TIAN SHAN MOUNTAINSDISTRIBUTION OF PERMAFROST WITH VARYING LAT- The ground tempcratures near the lower permafrost limit are~T~SDE AND LONG~TUDE". ,,. I " usually higher and permafrost layers are shallow with a depth nodeeper than 20 m. It was recorded that fhe depths of active laycrshave been increased to 3.5 m due to increased meltwater from snowThe distribution of permafrost and snow cover in the Tianand precipitation in the mountain areas (Qiu Guoqing and others,Shan mountnins exhibit an identical trend along the latitude andlongitude dircctions. The higher the latitude is, the lower the lower 1983).permafrost limit. Moreover, the descending magnitude of the lower Between the lower permafrost limit and stable permafrost arelimit for each degree of latihdc increment increased with increasing as, an unstable permafrost zone existed. The depth of permafrost inlatitude. ,Table 1. Lower permafrost limits in different areas of the TianShan mountains(from Qiu Guoqing)1Qiu Guoqing and others (1983) proposed an empiricai cquationfor calculating the lower permafrost limit in the Tian Shanmountains:II = I1089.5-10.6~-171.2~ (1)where, H is the altitude of lower permafrost limit (m 11.s.I.),x isthe east longitude, y is the north latitude.From this equation, the lower permafrost limit wil decrese171.2 m for each degree of latitude increment, and decrease 10.6 mfor each degree of longitude increment towards the east. Therefore,the lower permafrost limit shows a pattern of being higher in westand south parts, and lower in cast and north parts of the Tian Shanmountains,The snow cover is extremely uneven in the Tian Shan mountains.From Fig.2, for example, a snow depth of 70 cm occurs in theYili River basin in the west part of the Tian Shan mountains,'Whilein east part, a snow cover with such depth can only be found in the,top of the Haerlike mountain, the difference in altitude betweeneast and west ends of snow depth contour can be as large as 2,200m, while the altitude difference betw8cn the lower permafrost limitin the east and west parts of the Tian Shan mountains is only about220 m.this zone varies greatly, depending on the integratd effects. of externalconditions, especially the effects !now cover and snowfallevents. In the west part of the Tian Shan mountains, the depth ofpermafrost will incream SO m with an increasing altitude per 100 m,and in the east part of Tian Shan mounatins, that will increase only20 m with each 100 rn increase in altitude. In addition, severalgeomorphological landscapes formed due to the inpctions betweensnow and permafrost in this unstable permafrost zone. Thctypical landscapes are as follows:Mountain TerrEThe snow erosion landforms are widely distributed in unstablepermafrost zones in the Tian Shan mountains. Snow erosion depressions are also commonly obsericd in this area. The mountainterraces are usually steps of 1 to S meters wide and the surface materialsdiffer from one location to another. This landform oftenstretches into the stable permafrost-zone. Professor Qiu Guoqinghad oncc observed the multi-stage mountain terrace in the upperreaches of Alaxigongjing Valley in the Tian Shan mountahis.Mudflow and Mudflow TcrracesMudflow action zones are usually dispersed below the* 1145

Fig.2 Distribution of snow covers in the Chinese Tian Shanmountainsfrost-weathering and scattered gravel slopcs. Under the action ofrepeated thaw-freeze cycles, the meltwater from snow can accumulateabove the permafrost layer, the mud stated saturated soil thusleads to mud sliding and forms a variety ofmicro-geomorphological landscapes, such as, mudflow tongue,and mudflow terrace etc. These geomorphological landscapes arewidely distributed within the unstable permafrost zones, espccialyin east slopes of Watecrdegen and Xile mountains, south slopes ofYuximolegai mountain and east slopes of Aiken mountain pass.Thaw depressions are also formed in this area and can be obviouslyobserved. Thaw depression mainly occurs in the area wherethe ice content of permafrost is relatively high, and especially inareas with deep layers of under ground water, especially in the are+as near the Kuixian and Haxilegen mountain pass.Stone Stripes, Stone Circles,and Block Streams ~Stone stripes, stone circles, and block streams are also com-ly much thicker and can be as deep as several meters. In the TianShan mountairis, snowfall in this permafrost zone mainly occursduring the summers. Influenced by local topography,avalanchesare common in this area,and exhibiting a alpine avalanche landscape-ecologicalzone during the summers.THE SNOW COVER WITHIN THE AREAS OF SEASONAL~~ . "_"~ "" ,.."FROZEN SOILSThe snow distribution in the areas of seasonal frozen soil varieswith altitude and changing climate conditions and thus followsa typical vertical zonality of landscape in the mountain areas. FromFig,3, a vertical zonality of mountain ecological landscape exists inthc Tian Shap mountains. Thc lower permafrost limit in the mountainareas of the Tian Shan mountains is generally above the forestmne. In the south slopes of the Tian Shan mountains, the forestbelts are discontinuously distributed, and the lower permafrost limitthus lies in the meadow gravel zone of high mountains.1146'.

II I IT---"-I80VIII0v1NFig.3 The vertical natural belts of Tianshan mountains northern and southern slope in China95'8508 0 r 4 1 I 1 -.Dec. Jan. Fcb. Mar.REFERENCES, ."Qiu Guoqing and Huang Yizhi (1983) Basic Characteristics ofPermafrost in The Tian Shan, China. In Proceedings of SecondSymposium on PerrnafrostLanzhou, Gansu People's PublishingHouse p.22Hu Ruji and Jiang Fenqing (1989) Avalanche and it's control inthe Tian Shan Mountains, China. Bei,jing, People'sCommunucation Publishing House, P.14.1147

fFHOS'T-ACTTON DESlGN AND AFF1,TCATlONS OF ENLARGED TYPEPI LF, FOUNDATTON BRlDGE IN WATERLOGGED AREA- OF SONGYOUNGIluang Junhcng: Xu Zhenghni:Gc ~Iu~nyou~ and Zuu Li4'Cr~rnmissio~~ for Urban .& Rural Constructimn of qingdau Municipality, Chinazl'eoplc's C,uverrrment of Hci longjiang Province, China3Watcr Resources Burcou of Rayan County, Heilongjiang'W~t.er Resourcrs Dvpartnlent of Hei longjiang Province1'hc princ iyle nf sel f-anchorage by. frosL heave reaction force were usr-d to design *and constr~ct. the 15 enlargca type pile foundation bridges in the heavily frozensoi I i3r'e.a of Heilongjiang Frovince. 'The buried depth of the pile foundat.ion wasrrducetl to 1.8-2.0 mcters. Which should have heen 12-18 meters according to ther~rjrmal rules. 'I'hc 10 yl+nr's operational uti lization shows Chnt the constructionis in good condition ; + 1 8 d no pheno.mcna oi frost heave occurred. In addition. Thefrost hcavc I(~r(tt Tor the pile found;+Licln was measured for the firsl time inthis country wi1.11 reinforcemcnt S L ~ C R Y meters laid in splittcd seams. Aftcr themcnsurcmtlnt. h:is heen carried out for '3 years. i.c. 2 frozen cycles, the relat.iveerror Iretween the measurcd values and the rlc.signcd values for the maximum Unitfrost heave forcc lies i n the range of 4.4-12.52.I N'!'KOT)UC.TTONSinr-e 1082 wc' havv USI!I~ the principle of srl f-anchoragc. by trost. henvt, reaction forcc forprcvl-nt i ng frost hcavc occur,renr.e, and const ruc1.-CII 15 rlllarged typv pilc iounda~.ion bridgvs wiChr,r~~~forced cuncrt!te. The huried dcpth of !.he pileiollndntion is 1.8-2.0 meters. Our aim is to (:onductthc trost-oct.ior1 design of the cnlrrr.ge~l pileT I I I I ~ I ~ ; ~ I . 011 ~ I ) a ~ quantitative basis, which nevds1 0 IJSP the in-sit.r~ [measured convention:tl d ~ t a i njr(~7yrl soil and to follow thr theory of antilrost-:i~.t.ion:ohtain thc frost heave reactionIur'i c quantitatively; cc~lc.ulnl.e the size ofcavcs on [.he enlarged foundz t.i on rlecessary fort>t:;jr'ing t.he reaction furc.e, det.ermine thereasonable huried depth of pile foundation, andchcck over t.he strength of anti-irost-damagestructurc required for the cnlargrd pile founda-L i on .FROST HEAVE REACTION AND PRTNCIPLE OF SELF-ANCHORAGE OF PT1,E FOUNDATIONThl? frost heave reaction force is a reactiunforce produccd by t.he lrost heave stress orshear force. Tt. acts on the lower laying soilan? t.hus produces thj s ext.rH faundation rcac.tionforce (scc Fig.1).Tf the pile is bo be built -as an enlargedfoundation, i.e., a foundation with TI anchoredshect at depth Z, to make the eaves bear thereaction fnrce, it will result in a betterstabi1it.y 1.0 resist thc effect of frost heaveon t.he pile foundst.ion.FROST-ACTlON T)ESIC,N OF ENLARGED TYPE PII,EFOIINDA'I'IONStructure Configuration on? Rasis for SizeSelecriona) The b~lried depth of eaves of enlargedfounrlal.ion should be chosen to be close to orequal to the maximum frozen depth of soil as farNFig.1 Forces and their distribu.tions acting ona pileHfrozen depth;h, - frost heave amount on the ground surface;Z buried dcpth of anchored sheet;1 length of eaves:7 - tangential frost heave force;Nupper load;L, distribution scope of T;Pa - frost heave reaction force;G - deight of pile;Fl,Fz-- frost heave frictional force;Ft soil weight on eaves;L - distribution scope of reaction farce.as possible in order to retrain from stressattenuation in the course of transfer of frostheave reaction force in the thawed soil layer.b) The eaves length 1 determines the bearingproportion of the frost heave reaction force.It should be as small as possible on conditionthar. the eaves can meet the requirement for thestability of anti-frost-heave.

einforced3) The enlarged type pile foundation shall d) Computation examplesmeet the requirements for the load-bearing capa- The stabilityof anti-frost-heavebility and the stability of pile.We chose Yongchangdong Bridge and Yongchangxi K=(P,+N+C,tFtPt+G,)/CrBridge as the experimental works. The upperDarts of these two bridRes are made of reinforced ANTI-FROST-ACTION EFFF,CT OF CASE HISTORYhollow flat slabs and their lower parts areenlarged type pile foundations. The eaves are A. The Number of Enlarged Type Pile Foundationmade of reinforced concrete base slabs wi.th BridRes Built in the Waterlogged Area ofdimensions of 2.0x2,0xO.6 and 1.8x1.8xO.h meters SonRyonR Totals 15 (see Table 3)respectively for the two bridges, the diameterof pile is 0.5 meter and the buried depths of B.. TanRential F.H. Force Measurement at Yongchangfoundations are 2,O and 1.8 m respectively. Experimental Bridgesa. Layout of instrumentsComputation of Anti-Frost-Heave Stability3 sets of reinforcement stress meters ModelKLa) Criterion of stability20 manufactured in Nanjing were buried respective-Fa + N + G + F(1) ly in the midpiles of Yongchang Experimental.t Pt2Z.IRridaes (see Fig.3). The reinforcement stressb) Frost heave shear forcemeteis wcre welied to the force-bearing reinforcedbars of the pile column on a level of 1.8 m aboveET = TIDH~T~the ground and the reinforced concrete column wasblocked with an iron plate to ensure that thewhere D - diameter of pile, and I bars were solely acted on by the forceunit frost heave $hear force, referring during frost heaving of the pile. Variations ofto Table 1.resistan.ces and resistance ratios were measuredThe distance h from the acting point of the with a proportional bridge for hydraulic engineerresultantforce to the top surface of theing use and the values of tangential frost heaveenlarged foundation can be calculated as follows: force wereh=z --"inb. Measurement results3 for the type of frost heave whlch is (i) Tangential frost heave force (Fig.2,3 andbigger at the upper part and none at 41. AccomDarison between the calculated frosth=z --Hm2for other types,where !Im is the maximum frozen depth of soil,c) Frost heave reaction forceAccording to the recommendation by HeilongjiangInstitute of Hydraulic Engineering, thefrost heave reaction force can be obtained byusing the formula:Pa= -nlslr(0.5h-0.31)h'when lSh and h=i-- H2given in Table 4,The error of unit frost heave shear forcelies between 4.48 to 12.5X, which is within thepermissible range of errors; while the error ofmaximum frost heave force ranges from 3.3% to29.2%. The reason for this bigger error is thatthe magnitude of frost heave shear force isalso dependent on the frozen force between thepile column and its neigltbouring soil. At theend of November, the frozen depth is below 50mm, and the unit frost heave shear force willreach its maximum, moreover, the frozen surfacewill experience a,distruction process; afterTable 1. Design reference values of unit frost heave shear force in kg/cm'(from Hcilongjiang Institute of Hydraulic Engineering)Class of Weak frost Medl.um frostf rosr heave heave heaveStrong frost heaveGrade 1 Grade 2 Grade 3Frost heaveratio X1.0-3.53.5-6.06.0-10 10 - 1515 - 20Unit frost heaveshear force0.3-0.50.5-0.80.8-1.2 1-2-1 *a1.6-2.0Table 2. Computing results for the stability of anti-frost-heave7 0.9, 25.4 10.97 1.83 steady10 1.2 33.9 14.66 1.48 steady13.3 1.4 40 18.56 1.36 steady1149

~'Table 3. Sta'tistics of enlarged type pile foundation bridgesbuilt in waterlogged area of Songyong-tPile foundationBuriedSame ofLocationTime ofBridge completj un 'pan depth Diameter Enlarged , Thickness(m) of pile foundation of base Remarks(m) slab (m) slab (m)TongchangdongBr itlgeYongchangBrigadeJuly, 82 2x6 2.00.52x20.6Ynngchangxi Yongchang AUK., 83 2x6BrldgeBrigade1.80.51..8x1.80.55 samebridgesSanmenxangBridgeSanmeniinngjiaAug. ,' 83 2x6 1.80.51.8x1.80.60.52x20.68 samebridges'2 -20-0.5EFig.4 Observed results of tangential frost heave'tig.2 Observed results of (.ungent.ial lrost heave force for Yongchangxi Bridge in 1983-84force for Yorlpc hiingriong Dridgc in 1982-83that, the frost heave shear force tends to besmaller than the theoretical calculated value,thercfore, the measured frost heave shear force .,ris; mgsmaller than the calculated value. The\stronger the frost heaving of the foundationMsoil is, the closer wil.1 be the calculated value2.0 -20-* of shear its force measured to value.(i.i) Deformation observations (see Fig.5).1.5 -15-Two measoring points (3) and (4) ( see Fig.5)were set on thei.r mid-plles at GongchangdongRridge and Yongchangxi Rridge. The measured1.0 "10-results obtained in the two freezing and thawingcycles between 1982-1981, indicated that the0.5 - 5 -defurmation of the mid-piles due to frost heavewas less than 1 centimeter, no phenomenon ofNO". Dec. Jan. Feb. March frost heaving on the bridge surface occurred.-.during the process of frost'heave, and theEtime operation of the whole works was in good condi-Ytion, thus achieving the aim of anti-frost-heave2 50-4) for these works.m- CONCLUSIONS.Q 1000e A) The operational practice of these works150 -.frlr 10 years has demonstrated t-hat the enlargedtype pile foundation, on which the frost heave 'Tr.g.3 0l)servrri rcsults of Langrt~t.ii

Table 4. Comparison between calculated and measuredfrost heave valuesTangential Unit tangehtialFrost hcavcName ratio . frost heave .frost hcavef orccforce( X )(ton)(kg/cm2)YcngchangdongBridge-1982-1983DesignedMeasuredError ( X )1983-1984DesignedMeasured .Error (7)77101025.42021.333.')2429.20.90.9h4.b1.21 .n512.513.31'3.34038.73.31.41.5510.7- 1983-84 (4).*I& 1"E 5 t Nov. "eY/, --1Dec. Feb. MarchIYongchangxi Bridge"10 - NOV.1983-841982-83(4)W -c. Jan. Feb. March1983-84Yongchangdong Bridge1151

THE SHALLOW COVER DESIGN AND CONSTRUCTION TECHNOLOGY OF BUILD.ING FOUNDATIONS IN DAQING REGIONJiang Hongju Cheng knyuanDaqing Academy of Oilfield Design, ChinaThe perennial observations show that Daqing soil has different frost heave properties from freezingshrink to very strong frost heave. Because of differential heave and thaw collapse in foundation soil, theoilfield building structures are destroyed and the engineering quality is heavily influenced. So, foundationsin saesonally frozen ground regions, shallow buried building foundations, and frost damage preventiontechniques were studied. Meanwhile, the results obtained from research are applied directly tothe engineering design and construction. The results were very good.TWE SHALLOW BURIED FOUNDATION OF BUILDINGSAND DESIGNS FOR PREVENTING FROST DAMAGEThe Design of Shallow Buried Foundations of BuildingsBecause of the different properties of frost heave on the basement soil this will result in the deformation of external wall of thebuilding during freezing. During thawing the difference in meltingtime and velocity. of the basement soil under the external wall bottom,the wall produces uneven settlement deformation. If thetangential stress or tend stress of external wall which caused by thetwo kinds of action is bigger than the limit of material strength,cracking and deformation would merge in the wall body. In orderto guarentee the normal operation of the building structure, the followingwork was done:%e correct determination of the frost heave grade of basement soil-For the frost heave grade of basement soil applied in design,two main factors should be considered. One was the frost heavegrade in investigation. The other was the change of the frost heavegrade of basement soil produced in the building construction or usage.The frost heave grade of basement soil in researched years wasprovided by Chinese design standards of building basements andfoundations (GBJ7-89). The best way of determination was on thebasis of locally observed results.Howcver, if the frost heave grade of basemcnt soil determinedby the method mentioned above was considered as the design basis,two problems would appear:First, the annual air temperature and precipitation in the areais random. If the year of observation'and investigation wag aridwith high temperatures and a low frozen depth, the frost heavegrade of the soil is lower. On the other hand, if the year has heavyrains and low temperatures the grade could be improved.Second, if the original ground level was decreased during theprocess of construction, or since the other buildings affected drainageof the ground water, or the buildings were constructed in regionsof low temperatures and water weren't drained properly, thisresults in rise of content of the basement soil and rise of theunderwater level, the property of frost heave in the soil would increase.When determining the applied frost have grade in engineeringdesign, we think that if only one of the two factors mentionedabove exist, the frost heave grades of three kinds of soilwhich are initially determined as non-frost have soil, less frostheave soil and frost heave soil should rise respectively. The reasonis that the critical values of the frost heave ratio in the three kindsof frost heave grades are on the low side )Le. the average frostheave ratio of non-frost heave basement roil was q < = 1 %. Theaverage frost heave ratio of less frost heave basement soil was1%

~The geologic st-mcture of basement soil, the physical and mechanical index of soil, the hydrological engineering and geologicalstate. As well as considering the homogenitity and stability of theshallow layer soil.The sizeof the building, the important degrees, grades andStates of use, with or without a basement, ahd the foundation types,etc.The conditions of various diskes, hollows and holes, as well asthe cross-pipes under the building foundation and the structuralconditions.The buried depth of foundations neighboring the building.The condition of frost heave and thaw settlement of basementsoil, etc.The six main factors listed above show: it is one of six elementsthat seasonal frozen soil influences the buried depth of foundationsbut it isn't only the elements. Through research and practicc it canreach purpose of reliability and safety, if the buried depth of build.ing foundation is above the freezing depth of soil. The methods canbc divided into two stages: the first stage was before 1980, undersatisfying the conditions of the strength and stability of basementsail, thc buried depth of the building foundation was determinedbetween 0.8 and 1.4 rn from no frost heave to strong frost heave accordingto classification of the frost heave grade of the basementsoil and characteristics of the building. However, some importantbuilding foundations should be deepened. After 1980, through testsand research of various elements for several years, a more rationalcalculating method was obtained to consider the minimum burieddepth of building foundations in frost soils (Jiang Hongju, 1988).After the minimum buried depth of the building foundation wasdetermined by this method and considering the influence of otherelements, the suitable adjustments (total or partial deepening ) wereconsidered as design values. Through experiments and observationfor Several years, we think that the building cracks were not producedby tangential frost heave force and caused by an unevenfrozen soil layer under the foundation.During the thawing stage of foundation soil, due to,the diffcrentdirections of the wall body and the different irradiating degreefrom the sun, as well as the different thinkness of the frozen soillayers and different deformations of frost heave, different thawingtimes and velocities of the frozen soil layers under the foundationwould occur. This resulted in an unequal subsidence and bendingdeformation of the wall body. The deformation in the two wings of'the wall body,was the most obvious and caused cracking when thedeformation exceeded the allowable deformation value.The analysis stated above was proved by a large amount ofobservations and testing of local practical engineerings. The influenceof tangential frost hcave force couldn't be considercd on thethree kinds of soils( no, weak and normal frost heave). With soil ofstrong or very strong frost heave, prevention measures should beconsidered for safety against the tangential frost heave force of thefoundation sidewall in the range of 3 meters from the convex cornersof the exterior wall ( Fig.]) The tangential frost heave forceneedn't be removed at a range of total frost depth in unheatedbuildings and the prevention measures against tangential frostheave should be only considered at half the depth of the outdoorground.Prevention measures for frost heave damage in building foundationsThe effective and economical methods in Daqing region wereintroduced as follows:' The foundation depth of an unheated building was determinedaccording to the method mentioned in sectioned above. But withthe heat building, whether one or multiple stories, was deepened by20-30 cm in the corners and convex comers. The deepened lengthwas prolonged 3-4 m ( especially at the convex corners on the basisof the coner point). Non-frost heave riplacement material could beused in the deepened parts.For the basement soil of heated buildings with strong and verystrong frost heave, the deepencd length at the corner points (convexconers) was prolonged 3 m in the direction of the length and width.Non-frost heave replacement material was used at the exterior wallof the foundation. In unheated buildings; nonfrost heavereplacement material was used at the coner points and internal andexternal walls of the foundation. The replaced depth was half of thefoundation buried depth along the foundation. On the other hand,the foundation could be made with a ladder-shaped cross sectionwith a slope degree of 1:7 (see Fig.2 ) in order to eliminatetangential frost heave force.Thc apron slope of the wall root should be done properly andit would prevent side wall of thc foundation or bottom soil from being permeated by excess water. So as to reduce the action of frostheave and thaw settlement of soil.The foundation wasn't designed on the first storey balcony ofthe building and the balcony floor was used in the beam structure.Under condition of shallow foundation determined by methodsof this paper, for safety resons, the concrete periphery beamshould be added on the top of the foundation and the upper part ofthe wall body. The reinforcement of the periphery beam should bealong the longitude is needed to increase the diameter of the rein-AA 2!r 'A31IFig.] A stretch map of the building planeNote: A-convex corner, E-concave corner, I-longtitudinalwall, 2, gable, 3-internal wall.forcement within the convex walls of building and the two ends ofthe longitudinal walls from 10 meters. The reinforcement of otherparts were done with the basis being reinforcement of other partswere done with the basis being reinforcement bars of the structurein order to improve the ability of preventing against freezing-thawingdeformation of the wall body.P1153

~~ -.~r\Pig.2 A sectional drawing of trapezoidal foundationTable 1. The thickness of the allowable remaining frozen soillayer beneath the foundation durin8 sprinrthawingThe frost heaveStrongNo frost Weak Frostgrade offrostheave frost heavebasement soil heave heaveThe thickness oflimitrcmainingfrozen 70 50 30 15lesssoil layer (cm)StrongerfrostThe foundation of outer step8 bf the building should be separatedfrom the foundation of the wall body. The soil under doorwaysteps should be replaced with non-frost heave material andthe depth should be equal to the buried depth of the wall foundation.The designed temperature of the ends at the corners should beraised properly. ,The Design of Frost Damage Prevention in The Field of ConstructionIn order to not add or reduce the grade of frost heave in thefield of construction, the drainage desipp in the field of construetion was important. The following issues should be payed close attentionto:In order to have a good verticle design at the site, the ground..should have a proper slope degree which allows rainwater or waterseepage from pipes or taps to be drained away quickly so as not topermeate into the ground. Or a blind ditch should be planned usingthe materials of coarse sand and gravel, etc., at a depth of 0.5-0.7m to drain ground water.The comprehensive drainage system for domestic sewage, industrialpolluted water and atmospheric precepitation should beconsidered and linked with the urban drainage system. Meanwhile,drainage measures such as a cable, centrally heated dike and variousother pipes and dikes, and a cutting well should be irnplemented.Water can’t be allowed to permeate into the structures toprevent against frost damages.A BRIEF INTRODUCTIONTECHNOLOGY- ~ ~ .OF CONSTRUCTION ANDIN SEASOGE-EZTGOIL REGION^.On the basis of the climatic features and the freezing-thawingprocedure on basement soils in Daqiri region, the constructionstage of engineering could be divided into three stages:spring-thawing, summer and winter. The summer constnrction wasthe normal construction, but in winter it was contrary. Measuresused during spring-thawing construction (from March 11-20 toJune 11-20) will be mainly introduced.The landform, land vegetation, cave ~ etc. have effects on thefreezing state of basement soil. The grade of frost heave of basernent soils, freezing depth and important physical and mechanicalparameters of frozen soil ( Which were the water content, dry density,thaw settlement factor and thaw compression factor, etc.)should be surveyed and the calculation of thaw settlement shouldbe donc.The bench mark of measuring the site altitude couldn’t be act-ed by the frost heave. The design should use the height of sea level(altitude) and couldn’t use a relative height.Because the natural griiund was not even, if frozen soil layersremained under the bottom of the basic trench when the depth ofthe basic trench was dug to the designed altitude, the allowable remainedthickness of frozen soil should be determined in accordancewith the table 1. If the basement soil of one building has differentgrades of frost heave, the thickress of the frozen soil laycr would beadjusted in various parts and make the amount of thaw settlementas equal as possible. The total amount of subsidence would be restricted to the range of 10-1 5 mm.When the construction field had a complicated landform andmore caves or a large area of moved buildings, and there was afrozen soil and warm soil areas and a great difference of frozendepth under the new building, all of the frozen soil should be removed to assure the safety.If the building had a large volume and the grade of the frozensoil layer under the bottom of the basic trench was also high, theremaining frozen Roil can be exposed thoroughly by the method ofa thaw trench and the construction could begin when the conditionsallowed.Because of the thawing of the frozen soil layer during processof construction, the water content in the soil was large and a largersubsidence could appear. So to ensure the safety the followingpoints should be payed close attention to:a), When constructing the foundation, each part should beconstructed simultaneously so as to avoid unequal subsidence.b), For preventing constructing water from flowing into thebgsic trench of frozen soil, when the foundation was constructed,the quality should be ckecked and the soil of the side wall should bkfilled into each layer.c),When concrete was poured, we should avoid the concrete tobe poured on the frozen soil layer the measures of maintaining thetemperature should be done well in order to prevent the concretefrom freezing.d), Air temperature during March 11-20 drops sharply, so themeasures of maintaining the temperature should be done well duringprocess of the construction mortar and hardening.REFERENCESJiang Hongju and Cheng Enyuan, 1988, A method for calculatingthe minimum buried depth of building foundations, V ICOPPermafrost Proceedings, V01.2, ~1253-1255.

THE PREVENTION AND TREATMENT OF FROST DAMAGE ONBUILDINGS AND CANALS IN PERMAFROST REGIONSJim TianbaoYitulihe Branch of Harbin Railway Brueau,Heilongjang,ChinFIn forest regions of permafrost,thc frost damage of buildings and canals is caused by bigger water contentsof,mils,lower bearing capacity and air tempcraturc,uneven frost heave and thawing settlementctc.This paper introduces mainly how to protect and utlize the bearing capacity of frozen soils, removeunderground water,strmgthen the whoie strength of construction and coldproof and heatproof propertyetc.ta prevent and treat frost damage.INTRODUCTIONDa Xinggan Ling fortst regions located from 119'40'to127'00% and 46'30'to S3'21'N are pcrmafrost regions where meanannual air temperature is about -4'C. In these regions,thc buildingsand canal arc often subjmttd to frost damage in varying degree.So,in this paper,we put forward some measurements of pmventing and treating frost damage.save the heating fuel. Since the total practical transmitting heat re.sistance of pad R,(1 .6m2h "C/Kcal) is bigger than the necessarytotal transmitting heat resistan& R, (I.4m2h 'C/Kcal),So itsheatproof property completely satisfies with requirement of groundsurface.ITHE PREVENTION AND TREATMENT OF FROST DAM-AGE ON BUILDINGSThe Base of Chassis Connected with Buildings of Filling Instead ofRigfig and The Prevention and Treatment of Frost Damage ofThe Old Wooden Buildings.Since the wooden buildings arc fairly convenient in forest regions,sothis is a main subject to study, prevent and treat the frostdamage. The main advantage of new wooden buildings arc:Tcobble,slag,lime slagI IThe smaller weight and pressure and the well-adaptabilityFrom structure drawing,we can obtain that the pressure underchassis is only 0.098MPa. When transfercd to original ground surfacethrough pad, this pressure becomes to 0.02MPa. and to thebottom of suffered pressure layer,it is 0.04MPa, which only equalsto minimum allowable bearing capacity of silt which has maximumwater contents. So,this type of building can be built everywhere.Forold'wooden building which need digging base to lay chassis andbury pillars,it can not be built in regions of silt and poor-geologicalcondition etc.(Pig.l and Fig.2).The well-heatproof and radiative property6. The heatproof property is good,so,this type of building canFig.1 The sectional drawing of house connected with chassisCompared with the ground surface of red brick or yellowday,the ground suiface of lime slag may reduce the heat-low61.5Kcal/ m'h. So,for one ye& (heating 7 months),the house(30m') can save coal 2.1 T.b). Protecting original geological conditions arc preventingthawing settlement and frost heave. Namely protecting soil can&tionr of common haating buildings in the ranges of the thawingplats of f 20% span not to change.In aspect of absorbing heat: Since the apoorbing heat paramat-k155 0,

lar strip of outside wall,this method can prevent damage of -buildings.The lower cost andconvenient constructionSince slag and sawdust are fairly universal in forestregions,and the base filling instead of digging is not limited by con-.ditions,so,thc cost of this type of building is lower, and the con-'stmction is also fairly convenient.But it is difficult for the oldwooden building.rubble base'of inside wallFig.2 The sectional drawing of old wooden houseer of lime ground (0.29) is smaller than that of brick ground (0.7),So,the thawing settlement can be reduced 4.13m.In aspect of heat conduction: Since the heat conductivity ofpad is smaller than that of clay ground surface,So,tha thawing settlementcan be reduced 1.98111.In aspect of heat diffusion: Since we lay the apron slope ,protectionof 3m in every sides of building,which increases theradiative volume of pad,the thawing settlement also can be reduced0.85m.For old wooden buildingqthe buried depth of wall pillar is onlysituated in place of maximum frost heave force.The tangent frostheave force of a pillar is about 8.4 T,and the normal frost heaveforce is 18 T. So,the join forces (26.4T) is much bigger than theprcssure of a, pillar(2.2T). In addition,according to local climaticconditions, for average 3m frost depth, the amount of repeatedfrost heave and thawing settlement is about 18 cm, and, for 7mthawing layer inside house,the minimum amount of settlement is20cm over,So,the difference of 38cm between freezing and thawingcan cause serious damage of buildings.The well-entirety can,prevent uneven frost heave and thawing-set.tlcment ofbuiidings.According to theoretical calculation, the padcan prevent completelyfrost heave and thawing settlement of frozen ground layerundcr buildings. Even some uneven frost heave and thawing settlement,the loose pebble layer can be atljusted,and the floor of reinforcedconcrete and the chassis conneyted with buildings can alsobe balanced. Since the chassis of inside and outside wall are connectedtogether and do not produce relative displacement, so, thefrost damage of buildings will not occur.For old wooden buildings, since different Conductivity of insideand outside wai1,the inside wall is often produced crackle anddamage.So, we design a settled crack with felt to inlay two triangu-The foot of-wall do not produce damage and form frost,and the insidehouse is dry and do not produce seepage.We design a hopper of 45' on top of wall.The hopper maynot only increaw thickness'of coldproof layer but makc coldproofmaterials fill into space of wall at any time so that it can avoid disadvantagesof ventilation in winter also.For the old wooden building,wealso use this method to improve its disadvantages.The action of eaves, dado, apron and pad are that they may eitherkeep room drying and foot of wall not breaking etc. or preventthawing settlement of ground surface and forming frost of outridewall etc.. But they are just serious disadvantage of old buildings.The method by which frost damage of' building can be preventedthrouth filling instead of digging is only suitablc for woodenbuildings but also for brick buildings.." _- ~~The Base of Plug-in Pile of Reinforced Concrete of Protecting.-".".." . . -. - . . . .. . .Original Vegetation.-. " . -. . .-. - " " . . . -. .Tn the ice-rich zone ofswamp and tussock hummocks, therearc humus soil, silt and layered ice under most of' ground surfacewatcr,so,the common built method bf digging groove and liningbase is difficult to construct and prevent frost damage. From198O's, according to < > in 1977,wedesign the turret pile of 30cm corner radius.Meanwhile,we still laytussock hummocks or slag of' 45cm thjckncss on orignial groundsurfact(Fig.3) to prevent damage of vrgetation. The buildingswhich we adapt this methodto build arealwaysgood through perennialuse.'Fig.3 Ventilated and insulated base with reinforced concreteplug-in piles and whole circle beam1156

~ ,- ..Concrete Cone-section Mound Base for Advoiding Frost HeaveEn? Reducing Compressive Pressure.-In permafrost area, Tussock hummocks grow prosperly onslope, and surface water is found everywhere, while clayey frozenground is found under tussock hummocks. It is impossible to useideal reinforced concrete plug-in pile base, and it is uneconomicaland can not avoid frost damage to use general strip base.So, acone-section mound base with linking circle beam is selected in anew school -building. Good effects are obtaincd, and the maincauses are as follows:Cement apron with long eaves, high doda and asphalt groutingcracks can protect not only rain water from permeating intobase but also apron from cracking due to cohesion with wall duringseasonal frost heaving and thawing Settlement.Raising the the ground surface by 0.45m, thc heating wouldnot affect the thawing depth.Half filling and half digging not onlyrcduccd loss and digging soil, but also was used to intersect upperslope water and to remove ground water.Two circle beams increased the whole rigidity. and thiscanprevent the influence of uneven freeze-thaw on building.Grains beside the bottom of beam can fill the soil settlementand prevent cold air from flowing into room from the bottom ofthe circle beam.Thc air layer between floor arrise can prevent either the heatinside room from transfering to increase thawing depth, or floorand beam from decaying during air flowing.It is more importantthat when uncvcn frost heave occurs beneath circle beams, thegrains below beams squecze up cement apron outwards and pushtoward air laycr so as to prcvcnt circle beam from damage.Rase boundary, blind canal and seepage wcll are filled withgrains, and no water and damage on basc boundary occured:Using elliptic cone-section mound concrete base is more cconomicalthan using strip base.The tangent and normal frost heavingforce bctween ground and flanks of base could bc cleminatedbecause of the smooth surface and the shape of the elliptic base,meanwhile,the normal frost heaving fofce and campressive pressureon the bottom of base was reduced.(Fig,4)Complete. Prevention and Treatment of Frost Damage on FdBuildings.A forest bureau built a four-floor living building including aclosed balcony and a brick-concrete structure in 1985. Cracks appeareddue to settlement in summer of 1986.By the begining of1988, more,than ten large cracks showed on the south wall from thebottom of the wall to the arrise of the room. Doors and windowspassing toward balcony were compressed to be damaged bycrushed bricks. The widthes of some large cracks are more than50mm. Although most of the widthes of the cracks on inside wallwere less than 20cm, but dense cracks caused an unsafty state.The building was originally desingned as a 2.55m nigh rubblebase-top' with a 150 reinforced concrete circle beam(40cm width,40cm height), 14 reinforcing bars of 12mm, and 2.95m deep base.However, during construction, a reinforced concrete circle beamwith 49cm width and 90cm height was set up on a rubble base-topwhose height was only 0.7m, and22 reinforcing bars of @ 14mmFig4 Profile of house with double-circle cone-section moundbase..were in$talled.The depth of the basc is I .6-1.7m, which is just In theranjic 01. the maximum frost heaving within 2m from ground surfjce.Thcre was not measurementto prevent cooling, drainage, andrefilled insidc the room.A closed storage warminc room u x cformed around heater pipe under floor. and this rapldc the raw itnddepth of thawing of undcrground liozen layer insidc the room. Inaddition, three 30-50mm soil-icc layers existcd within the 2 m ofthc base-bottom, which caused heavy thawing scttlcmcnt andcracks on the outstde walls. Thc soil around base boundary. whichwas wetted by rain in autumn, resulted in frost heave inwintcr.Therefore. about 20cm of seasonal frost heavc and settlcnlentoccured yearly, which caused veriical cracks on rhe outsidewalls. On the other hand, the horizental norm:ll frost heanng forceon the flanks of' base reached to 10T / m'. unfilled so11 inside roomcould not support rcsistence. This resultcd in hnr17cntal cracks 011thc outside walls. In later construc~lon, it was found that manyparts of under the outside circle beam\ wcre hanged in air owing tothawing settlemcht of soi1,and the mcximum hanged depth at bothsides of southwest corner was up to 2-3m.Combining investigation with local hydrogeological condition,the following treatments are suggested:Avoiding-frost heaving form around base boundary.In order to prevent rain water from permeating into base tocause frost heaving force, besides raising ground surface and addingconcrete apron with grouting a*>halt cracks: dry slag, sandygrain, pebble and cobble are filled in order from the depth of0-0.3111 to the bottom of base. This treatment play a role of coldprevention and insulation so as to reduce the frost Forcb on theflank of base. At the same time, underground water is permitted topermeat into blind canal to flow to hidden well.There is no water in

the dry base boundary,so there is not any frost.damage.Avoiding the thawing settlement of base bottom.Firstly, heaters under floor are moved floor,so the heat sourcereleasing down iseleminated. Secondly, insulated felts, asphalt feltsand new floor are paved. Water on floor are not allowed premeatingdown to wet thcbelbw insulated materials, and the heat in roomcan not conduct down. In order to prevent floor from decaying andreleasing residual heat down to floor, a few vantilating holes arebuilt in addition to retain a air layer of 25cm beneath floor. A sawdustlayer of 30cm with lime is paved on the top of filled dry slag toinsulate heat and prevent thawing. By use of these measurements,the frozen layer does not settle and crack does not appear on theinside wall.. - ." "Repairing" __ crushed brick walls.The crushed brick. walls are commonly rebuilt. Large cracksare firstly mended using high level cement mortar.Aftcr the groutingwater is dry, cracksc, whose widths are mare than Smm, aregrouted into mixed glass-ccment mortar,and small cracks of lessthan 5 mm are compressed into water glass. Some small cracks onwalls and ceilings are smoothed with dilute cement mortar.Atiet'doors and windows are paintcd,the building looks like a new one.The measurements for preventing frost heave, drainage, insulation,and preventing thawing settlement are carried out well duringconstruction. Since'the project was finished in 1989,the buildinghas been worked well.THE PREVENTION AND TREATMENT- OF FROST DAMPAGE - ON ". CANALS ."-The common buried depth of local canals is 3.3m, which liesbetween thc minimum frost depth(2.9m) and maximum frastdepth(3.5m).Although the damage of seasonal ,frozen layer isavoided within the depth, the menace of permafrost layer of 2.5mstill exists. Therefore, the shallow buried insulated pipeline isused.This kind of pipcline is made mainly by: packing polystyreneof 7cm thickness around tht pipeline of @ 1 SOmm; packing asphaltfelts and placing it on the wooden block (one cach pipe) with thesize of 12cm width,20cm height and 70cm length; covering rcinforcedconcrete semicircle pipe with the size of Im diameter and7cm thickness;and paving slag of 0.Sm thickness and clay of 0.2mthickness(Fig,5). The 600m length pipeline,which was built accordingto the above method, has been running well for six years, especiallyin rock layer zone. This confirmed that these measurementsare effective. The better methods are:(]) Replacing the connectionof screw of flange plate with specific hoop, the construction wouldbe rapidly carried out, and the tenacity to resist the uneven freezingand thawing could be strengthened.(2) Replacing alphalt felt andiron line with plastic film and nylon rope to pack polystyrene, costcould be reduced, and tenacity could be strengthened.(3)Replacingreinforced concrete semicircle pipe with reinforced concrete pipe of0.6m diameter, construction becomes easier, and the rigidity issterngthened when vehicles pass above the canal. A layer of 15 cmthickness (more than common ground surface frost heave 147mm,and slight less than the average thawing settlement of base)of slag isfilled on the bottom and the flanks of base. This can adjust unevenfrost heave and thawing settlement. The clay layer of 0.3m thick-ness is slight higher than ground surface. This not only prevents thetop of pipeline from breaking out by behicles but also makks thedrainageeasy. (4) Hanging the pipeline on the wall of reinforcedconcrete pipe,a good air layer is maintained and penetration ofground water is avoided..The feature of this kind d pipeline facilitates the constructingFig.5 Cross-section of buried shallow pipelineand repair of pipeline besides preventing frost damage.In swamp zone with tussock hummocks; is the reinfomd con- 'mete pipeline is changed into the shape of groove, and placed onthe pad of refilled grains of 0.3m thiclncss, and if dag is paved aGcording to thermal requirement. and rainproof layer of clay is covered.the pipeline would be more effective.REFERENCEDesign Standard and Criteria of Base and Foundation inPermafrost Regions, in USSR.Moscow. The House of Amhitecture,1977.Translated by Northwest Institute of ChieseAcademy of Railway Scienccs.Proceedings of Architectural Design(l-3).Beijng Institute of IndustrialConstruction, Architecture Engineering Ministry. TheHouse of Chinese Arehitecture Industty(1973-1978).1158

DETERMINATION METHOD FOR THF. COEFFICIENT OF THE DEGREEOF SUNSHINE AND SUNSYADE ON CANALSLi Anguo and Chen QinghuaNorthwest Aydrotechnical Science Research Institute,Yangling, Shaanxi. ChinaThe coefficient (K1) of sunshine and sunshade extent is a basic parameter forpredicting frost penetration in engineering design. In order to get an accuratevalue for the coefficient tallying with theiactual case, the total radiation oneach part of the canal section and level ground in a freezing period (ZQ, and ZQ)are determined respectively based on geographic latitude, elevation of the canal,local topographic conditions, trend of the canal and shape of the section. Thenthe coefficient on each part of the section may be determined byK~ - (CQaB-CQ)*asltF..BINTSODUCTIONThe coefficient Kd of the degree of sunshineand sunshade is a basic parameter in forecastingthe engineering-design frozen depth (ProfessionStandard of Water Conservancy and Electricity inP.R. China, 1991). At present an experimentalmethod is used to obtain the value, which israther limited and is too rough. It has beenproved in practice that the degree of sunshineand sunshade in each engineering site variesnot only with geographic latitude, sea levelheight and topographic conditions at theengineering site but also greatly with differentengineering patterns. Therefore, according tothe above factors and the strike of the canaland the shape of the cross section, etc., theclimatology method is used to calculate respectivelythe total radiant amount of the sun oneach part of the canal and horizontal ground inthe frozen period and the increment of accumulatingtemperature on each part of the canalcorresponding to the horizontal ground. Thecalculated formula of the value Kd can beobtained and the Kd can be determined accurately.1. Determination of the -Time of Sunrise andSunset on SlopeLet - w ~ and +WH be the time angles of sunriseand sunsei under the conditions 0: no terrainsheltering and at the height H above sea level.h cr and Sfia are the height and radiant flux of" t8e sun on the sl.ope with slope direction B,gradiant a. When there is no other sheltering,the necessary and sufficient conditions for thesun to shine onto the slope are:-WH 5 w 5 +w H (1)S B ~ 2 . 0 (or hBuzO) (2)can be deduced:W~=COS "1 (-tan&tan6-0.0177di secQsec6)(3)SBa=I(~sinG+vcos8cosw+sinBsinacos6sinw)~ (4)where, 4 geographic latitude;6 - the sun longitude;H height above sea level (kilometer);I - the radiative strength of the sun1.94~10- Kcal/cma .min.u.-sin&cos~-cos&sinacosBv=cosQcosa+sinQsinacos8Since the necessary condition of sunshine isSB~ZO, then obviously the time angl7.w~ withS a-O is the Aitical angle when Sga turns fromtie negative to the positive or turns oppositely.Let the right side of (4) be equal to zero,. ,the expressed equation of the critical timeangle ws can be obtained as follows:From the first part of (6) the two absolute'values of the ws can be determined, and thesymbols for the second half can be given.Supposing that the two values of ws are us! andwsz separately, and wS2>wSl, since SB~LO 1snecessary for the slope surface to be exposedto the sun. we can easily see that the timeangle w1 of sunrise and w2 of sunset on theslope should satisfy the following conditions:where w is the time angle of the sun.Using the astronomical formula, the following1159

At the same time, since it would be possiblefor slo e surface to be exposed to the sun onlywhen lwPS 1 ~ ~ wl, 1 , wp should also satisfy thefollowing condition:2, Determination of the Time of’Sunrise andSunset on Each Part of the CanalAs shown in Fig.1, choosing a free point Aon the trapezoid canal with the strike being y,width of the bottom being B, depth of ‘the canalbeing H., gradient of the sideslope being a, ifthe heights of both canal banks a, b do notvary greatly and the canal, is very long, pointA is sheltered from the sunlight by the canaland is equal to that of the two imaginary slopesurfaces Aa and Ab. The gradients of them arerespectively equal to the elevation angle aa,observed from point A to the bank top in thedirection of the perpindicular bank s-(that is,the maximum angle of sheltering from bank a topoint A or the maximum angle of sheltering onthe horizon made by point A), and the elevationangle ab, observed from point A to the bank topin the dfiection perpendicular to bank b. There.-fore, the time angle WA) of sunrise at pont Ashould be equal to the later of the two imaginedslope surfaces Aa and Ab, and the time anglewA2 should be equal to the first of the two. Intwo other conditions, such as the very longcanal in U shape and arc shape and the canal onthe edge of the plateau, a method similar tothe one mentioned above can also he used todetermine the time angle of sunrise and sunsetor the amount of sunshine in every day.ZFig.1 Sketch of the canal and related parameters(see text for details)In Fig.1, the coordinates of points a, b, andA are: a(Xa, Z,)-a(B/2+Ho/tanu,Ho), b(Xb, zb)=b(-B/Z-Ho/tana.Ho) and A(XA, ZA), respectively.The gradients of slope Aa and Ab are separatelyaa=(Za-zA)/(Xa-XA) and “b=(Zb-xA)/(XA-Xb). Theazimuths of slope Aa and Ab are Ba=Yt90° and6b=Y-9Oo.Time angles of sunrise and sunset.^,,, waZon slope Aa, Ub2 and b on slope Ab, can becalculated by the rule described in the firstpart of this paper by substituting the givendata into the relative equations, and then,according to the possible condition that pointA could be exposed.to the sun.The time angles of sunrise and sunset can beobtained at point A.I_ 3. Calculation of the Total radiant Amountnirectly ReachinR Each Part of the CanalThe daily total radiant amount of the sun oneach part of the canal and horizontal ground(corresponding to the condition that theatmosphere is completely transparent or thereare no obstructions on the earth) can becalculated with the following:’on each part of canal (Fu Baopu, 1983):3n horizonta1 7vo = rR’1Iground:where, WBao’ io - total daily radiant amount ofthe sun on each part. o r thecanal and horizontal r;rnun-’.T. - constant of the sun, 1.94~10”~~ 11;cm’.min.R - distance between ZIIF bun and gt,.~.with the sun-ground averagedistance as the unit.* - the length of time in B d;jy f=L.’60 minutes).w1,W~ - time angles of sunris~ andsunsut on each part of tht? , Idetermined by the methoddescribed in the second pjr-t o !this paper.the meaning of the other signs is simi1.1r tothose in the previously equations.In any period of time, the total astronomicalradiaht amount is the sum of total dailyastronomical radiant amount in this period..Tne radiant amount directly reaching thehorizontal ground can be calculated by theexperience formula (Gua Guodong and Lu Yurong,1982).So = a + b.Qo.Sl (10)where, So - total monthly radiant amountdirectly reaching horizontalground, unit (Kcal/cm’.month).Qo -- total monthly astronomical radiantamount 0: horizontal ground, unit(Kcal/cm .month),S1 - average daily sunshine ratlo in amL)n th .a, b ere coefficients related to the transparencystate of the atmosphere, which can beconsulted from Fig.2 (Gao Guodong and Lu Yurong.1982).From (10) the following can be obtained:,So/Qo = s/Qo t b.S1 = f (illwhere f is the function of atmosphere transparencyon horizontal ground.With the research of Fu Baopu, 1983, it canbe known that in the rano.? of our country, thefunction of atmosphere transparency above a

11 - average total cloud amount in a month.HI, h, are the coefficients indicating thetransparcncy degree 1)f the atmosphere, which canbe found Frunl Fig.3 (Cao Guodvng and 1,u Yurong,1982).Fig.3(a) Distributiun graph of coctficient a1calculting formula of ~c~t.teringradiation in winter (Kcallcm')inEig.?(h) Distrihu~ion graph of coefficient bl incrllculting formul~ of scatteringradiation in winter (Kcal/cm')('sing f' AS the transparency degree of theatmosphere, (13) can bc changed into:where, SAa - total monthly amount of radiationdirect I y reaching each part of thecanal, unit. (Kcal/c.m'.month).QBao- total monthly amount of astronomicalradiation on each part of thecanal, ut) it. (Ycal/cm'.month).4. calculation of the Total Scattered RadiantAmount, on Each Part of the CRnalTht. scattered radiant. amount on the horizontalground can he calculated with the experienceformula of Lao Guodong and Lu Yurong, 1982.Do = a1 t bl.Qo.n (13)where, 1, - thr! total monthly amount of ntat-Lel-ed radiation on horizontalground, unit. (Kcal/crn'.month).Qo - total monthly amount of astronomicalradiation on horizontal groundun't, (Kcal/cm'.month).Do = f'.L),=(L t b,n).L),00(14)SuylJosiuR that the scattering and the radia--Lion are similar in all directions, the caltulatingformula of scatter~d radiation flux, byFu Baopu, 1981, can be referred to.- -D =& ~(cosa,tcosa~)cosa-(sina,-sinu~)s~nalRa 2-where, DRa-the density 'of the average daily(15)flux of scattered radiation on slope- surface, unit, (ca~fcm'.min).no- the density of the average dailyflux or scattered radiation onhorizontal site, unit. (caZ/cm',min),.a - gradient of slope surface.1 a, ab are thc maximum angles of shelteredarea respectively from ridges a, b to thestudied point.The flux formula of scattered radiarion at 'each part of the canal can be writtep in theform of:

-where, asai - the density of.average flux ofscattered radiation on any tartof the canal, unit. fcal/cm .min).manab - maximum angles of sheltered areaseparately from the canal banksa. b to 8 certain degree.If considering the model different in alldirections, referring to the method put forwardby many scholars, that the models are differentin all directions is used to calculate thescattered radiation amount on slope surface andare an improvement to the models that are similarin all directions, and approximately using thecorrected value given by Li Zhangqing and WengDuming:A(ho.u.n,fJ)~F(n).cosl.09ho.sinl.42a.cos~35 1=0.~503F(n).cos1.09ho.sinl.42u.cos~(17)where, ho - aveiage height angle of the sun atnoon in the period of calculation.F(n)-l-(O.En+O.O2nL).A(ho,u,n,$)- correcting value, cal/cm'.min.a the gradient of slope.6 - the azimuth of slope direction tocalculating part.n - the average total cloud amount ina month.nL - the average low cloud amount in amonth.The equation to calculate the average dailyflux density of scattered radiation on each partof the canal can be derived as the following:+0.0503~(n).cos1.09ho.~~n~.L2a.cos~From (18). the calculating equation of themonthly summary amount of scattered radiationat each part of the canal can be found easilyusing:where, t - is the number of days in a month,5. Calculatinn the Total Radiant Amount on EachPart of the CanalTotal radiant am0unt.Q is the sum of thedirectly reaching gradiant amount and scatteredradiant amount D, that is:Q - S + l l (20)The monthly sum of total radiation on each partof the canal:The monthly sum of total radiation on the hori- ., /zontal ground:Q'=So t DoIn a long period the sum of the total radiationis the accumulation of the total monthlyamount of total radiation for every month in theperiod.6. Determination of the Vale Kt! for Each Partof the CanalFrom the beainninn .? - of calculating the frozenindex to the day when its maximum value appears,with the following equations (21) and (22),separately calculating the total radiant amounton each part,of the,canal and. horizontal ground,day aEter day, then accumulating them separately,the total radiant sum CQ and CQ can be obtainedon each p,prt of the canafaand horizontal groundin the frozen period. In the period the incrementof accumulating temperature on each part ofthe canal corresponding to the horizontal groundsurface is:(23)where, 8- thermal diffusive coefficient jete'-mined by average wind velocity (ZhuBaifang and yang Tondsheng, t97h),unit. Kcal/m .h.'C.as- thermal absorbing coefficient, taking0.65 for concrete surface.Therefore, the degree of sunshine and sunshadeis considered by formulation, the-value Kd Ofthe revised coefficient of the frozen depth ateach part of the canal can be computed as:where, Fo -- frozen index.H6,,H0 - the value of frozen depth of eachpart of canal section and groundsurface.For example, in Yingchuan frost-heave testingfield, Its geographical latitude is 38.4872,height above sea level (H) is 1.113 km, thestrike of canals in the field are different inE-W, NE45', N-S, NW45O,, the size of the canalsin the cross section ail are 3.10 m in bottomwidth, 1.8$5 m in canal depth, 1:1.35 in sloperatio of side slope. According to the analysisof atmospheric data obtained in the last 30years, the aierage frozen index is hhh.7"C day,the average frozen depth is 105 cm, the frozen 'period is from Nov. 20 to Mar. 13, the averagewind velocity in the freezing period is I.R4m/s.On the basis of the calculating methoddescribed above, we can get the value Kd ateach part of the canal, as shown in the table.REFERENCESFu Raopu, (1983) Atmosphere in Mountains,Science Press, 2-55.Gao Guodong and Lu Yurong, (1982) Balance ofradiation and balance of heat en groundsurface in China, Science Press, 9-35.G.Y. Partridge and C.M.R. Platt, (197h) Developmentsin atmospheric science, 5, radiativeprocesses in meteorology and climatology,Elsevier Scientific Puhlishing Company,Amsterdam-Oxford-Ndw York, 37-41.L1 Zhanqing and Weng Duming, (198,) The distributionand computing model of \he diffused.

adiation on slopes, Acta MeteorologicalSinica, Vo1.46, No.3, 349-356.Zhu Baifang and Wang Tongsheng, (1976) Thetemperature stress arid temperature controlmentof hydraulic concrete structures, WaterResources and Electric Power Press, 28-35.1163

SIMILARITY ANALYSIS OF MODELING TESTOF FROZEN SOIL UNDER LOADLi Dongqing Zhu LinnanState Key Laboratory of Frozen Soil EngineeringLanzhou Institute of Glaciology and GeocryologyChinwe Academy of SciencesThe article is based on the principle of similarity theory , analyzed the static questionof thawed and frozen soil on Similarity analysis systematically, derived the similarconditions of mechanics which modeling test of frozen soil under load must be satisfiedwith ;and on the basis of the conditions, we propose a approximate method of modeling' te$t of freezing and thawing settlc ment under load.INTRODUCTIONAccording to the similarity analysis of modeling test offrozen soil without load(Zhu Linnan etc. 1993) ,we canwrite the followifig similar ctiteria of thawing and freezingmodeling test without load :si2Dotand -Adwhere ( q = Lpw(B, - 0. + A0 )), in whichBU,AQ, 0,respcctivcly are the unfrozen water content , water contentof migration , and initial water content . L is the potentialbeat of phase change . p, is the density of water, 1 is thethermol conductivitj of soil , D is the diffusive coefficient, t is time , I is the geometric length . In order to study, the interaction between engineeting building and basementsoil, and the engineering stability . the article , on the basisof conclusions above , applied the similarity theory to drawthe similar conditions of mechanics of modeling test offrozen soil under load , at two angles of dimensional and' equational analyses , and we proposed a concrete appliedmethod of modeling test of freezing-thawing settlement .SIMILARITY ANALYSISThis paragraph applied the exponcntal method of thedimensional analysis (Xu Ting 1982) to derive the staticsimilarity criteria of the question of unfrozen and frozensoil and constrvct the similar relations . Although the stressand displacement in thawing-freezing soil. are not linearlyproportional to the external load , In this case , the stress uhas a relation to the load P (supposed, here only is aconcentrating force P) , the basement dimensions andthickness I of soil layer, elastic coefiicients E and p ,and volume weight y . therefor , the stress in the soil can bewritten as :o=f (P, 1, E, N, 1 (11According to the theory of dimension, when thephysical quantity y is a function of physical quantities x,,xt, x,, ,.., x, , the dimension Cy3 ofy isequal to themultiplication of dimensions CxJ of x,, x2,. x,, ..., X,.thus, ifand theny= f(xI,xz,xq...,xJtherefor, the dimension of the stress u can be written as[m] = Cp]' 9 [IIb [E]" bId [yl0 (2)where , a ,b ,c ,d ,e are unknown exponental constants .from the dimensional tdble (MLT) (Xu Ting 1982) ofquality, length , and time, we find all the dimensions ofphysical quantities above and put them into the formula (2),we can obtain[IilL",T-2] = [MLT-'Ja 9 [Lib ~L"T-']'O O O d[M L T ] [ML-*T-']'Making a comparision between same dimensions oftwo sides of the above formula, we can obtaina = l-c-e,b = 2&3e-21164

we introduce them into (2) and rewrite it as2[$I =[$IC. [$I 3 0a12We respectively regard - ,- El2,El2(from-P P Pandvi3 - ) ,p as a quantity, according to equation(3)Pand taking c = e = d - 1 , we can write their functionalrelationship (4) , and then (4) is a dimensionlessexpression of the formula (1).c,= 19.1, in order to satisfy x;=lr, . in terms of thereferences(An Weidong 1990; Zhu Yuanlin and Carbee1984): tests results between the relations of the stress-stmincurve and initial elastic module E of frozen soil andtemperature T , we choose the case in which the soil ofmodel and prototype have the same conditions of watercontent and temperature change, thus, the initial andconditions of modeling test without external load(Zbu Linyan etc. 1993) ?re satisfied, in this case, we regardE'== E, p - p, then x4- x4 is also satisfied.From n2 = x,, we arrive at c, =.,cy = 19.1', thus,the concentrating form (to pile basement) on model,' P p =-- =- P ; 'and according to (4) , we arrive atc: 19.1the areal load(to buildings) p = p , the line load(to long basement) q' = 4 = -'q.CI lY.lc, 1thus (4) is rewritten as From n, = x , , wc arrive at cy = - = - , thus,c, 19.1This is the equation of similar criteria of the nonlinearthawing-freezing soil , where xz, z,, n4 aredeterminational criteria .We supposed above, there is only a concentrating force. if there are a linear load q , a areal p and a moment m ,we also consi&r them as factors influenced on the stress ,and introduce them into (1) . thus , we can arrive at similarcriteria comprised of them , which is to say , the criterionElx1 = - is respectively inverted into :PMODEL DESIGNthe volume weight on the model y =c,y = 19.1~.According to the,equation (5) of similar criteria, it isimpossible to satisfy n3 = n3, because we can't find thematerial, the volume weight of which is 19.1 times as weightas that of natural soil, but other characters (for example E,I(, etc.,) are unchanged. we known, in the procedure ofthawing-freezing and heaving of soil, the weight can't beomitted ,which is regarded as the key control factor causedthe settlement of thawing. in order to satisfy the similarrelations between the gravitational stresses of the model andprototype, we can approximately regard-the partial gravityof the nfodel diminished when the length scale of the modelis diminishing as a additional stress of a external load. atthe same time; according to the result of similarity analysisof thawing-freezing heat and mass migration, we write thesimilar ratioes of temperature , moisture, geometry andtime:According to the paragraphaiteria are:x,=-,zz-p,z3=-,)x4=pa12 ElZYlPEWhen designing model, the model mustwith similar criteria of determination:two, mechanic similarin other words, the following criteria are equal to 12be satisfiedAccording to the factors of thc conditions of addingload, test field, workmanship of making model, etc., we takeTHAWING-FREEZING MODELING TESTOF SETTLEMENT UNDER LOADTo state the application of similarity theory andmodeling test in the frozen soil, at first we use thecalculation of thawing settlement to state the principle ofabove those paragraphs. as far as the soil of basement,whether it is thawing settlcment of frozen soil or settlementof unfrozen soil, the soil body gives birth to two kinds ofstresses: one is the gravitational stress; the second is theadditional stress within basement, resulted from theexternal load. so far as undisturbanced or unfrozen naturalsoil, it have experienced prolonged geologic eras, and havefully consolidated under thc gravity, in this case, only theadditional stress under the external load gives birth to tbecompression and consolidation of soil body; and in thefrozen soil, since there is ice, its characters ofphysics,mechanics and thennology arc obviously not the same these1165

I -of unfrozen soil, with the result that although there is notany external load on the frozen soil(exactly speaking,under a finny load), the thawing soil body also yields aconsolidation and settlement. thus, when solving thethawing settlement of frozen soil, at first we shouldcalculate the additional stress of basement soil underthe external load,According to the theory of elastic mechanics(Xu Zhilun 1978) and geotcdhnics (Qian Jiahuan 1988),under a uniform load, as far as a rectangular and circle Ibasemcnt, below the central point, the additional stress ofarbitrary point can be written as the formula= K.P (7)where p is the uniform areal load of the bottom ofbasemcnt, K, is the tidditional stress' coefficient relativeto the dimensions of basement, depth@) and theirproportional dimensions, undcr the conditions whitch thedimensions of basement and the thickness of soil layer aresatisfied with the similar proportion of geometry, theadditional stress cocfficient o$ the basement of model andprototype are cqual, Kf=.K, . in this case, according tothe conditions of mechamc similarity c,,- cg= 1, then,c - 1 . that is to say, to the satisfaction of the conditionsu*of mechanic similarity, the additional stress in the modeland prototype soil are both equal.We have known the additional stress nz thenaccording to references and documents(An Weidongctc. 1990 and Tsytovich, N.A. 1973) , under uniform load,the'thawing-settlement of frorxn soil can be expressed asS=AoH+autH (8)Here & is the thawing settlement coefficient; u, is theaverage additional stress of thawing compression layer;-a is the thawing compression coefficient; W is thethickness of thawing compression layer.Obviously, if the, model has the same initial conditionsof soil moisture and tempcrature change as the prototype,then A, and a are unchanged, at the same time, thecondition of mechanic similarity is satisfied, that is to say,C - 1 , ck =c,, thusc, =c,.urnTherefor, when thc external load is satisfied with thesimilar condition of mechanics, the ratio of the main settle-.ment values of model soil and prototype soil is equal to thegeomctric proportion.As far as the settlement value of the nonuniform frozensoil under load, we can use tho method of geotechnics(QianJiahuan 1988) to write:- -S= ZS, - zA,Hi f ~ a,p,Hi (9)1-1 1-1 1-1Since the procedure of analog simulation is the same asthe uniform soil, herc we do not state it detail.. inCONCLUSIONSUnder the conditions content with similarities of heatand mass migration , we added the following similarconditions of mechanics , that is :are three similar unchanged quantities(1)The linear load(q) ,the moment (m) ,and the arealload (p) of the prototype and the model should be satisfiedwith :(2)The stresses and the displacements of the prototypeand the model have the following relations:is,- U=c,sIn which ,c, is the geometric !tale between theprototype and the model , u and S are the stress and thedisplacement of the model , u and S are the 5hSs and thedisplacement of the prototype .These conditions above, gave the practicabletheoretical proofs of experiments in order better to studythe interaction between engineering buildings and basementsoil, and the engineering stability, at same time, accordingto these similar conditions , we proposed a concrete appliedmethod of modeling test of freezinp-thawing .REFERENCESZhu Linnan and Li Dongqing(l993), Similarity analysisofmodeling test without load, Journal of .Glaciology and Geocryology 15(1) ,189-192.Xu Ting(1982) , Similarity Theory and Modeling Tat,Agromachinery Press of China, 34-62, Pecking.An Weidong etc.(1990), Interaction Among Temperature,Moisture and Stress Fields in Frozen Soil,Lan Zbou University Press, 244-246, Lan Zhou.Xu Zhifun (1978), Mechanics of Elasticity,Higher Education Press, 105-108 , Pecking.Qian Jiahuan (1988), Mechanica of Soil,He Wai University Press, 53-69, Pecking.Zhu YuanLin and Carbee ,D.L.(1984), Wuiaxialcompressive strength of frozen silt under constantdeformation ,Cold Region Science TechnologyVoh, No.1 , pp.3-15.Tsytovich , N.A.(1973) Mechanics of Frozen Soils.High Education Press, Moscow, 250-276.1166 *'

A COMPOSITE MODEL OF MULTIPLE ACTIONS FORFORMING.PATTR9NED GROUNDLi Guangpan and Gao MinLanzhou Institute of Gla'ciology and Geocryology,Chinese Academy of Science, Lanzhou 730000,ChinaThis paper paints out that the patterned ground is a spatial ordered structure 'which naturally exists and is called a "dissipative structure" in the branch ofnonequilibrium thermodynamics, During freezing and thawing, the soil-watersystem is in a state far from equilibrium and the ordered convection in nonequilibriumis formed at critical conditions because of the various actions indynamics such as the buoyancy, the gravity, the force of migratory water causedby frozen field, the diffusion, and the surface tension, etc. The continuousactions of convection change the combination between soil particles and slightlydepict spatially structural patterns of convection in ground. Each convectiveaction alone is very weak. However their effect.^ are enhanced when frost heaveand desiccative shrink are superimposed on the processes of ordered convection.And these processes repeat again and again, finally resulting in the formationof macroscopic visible patterned groundINTRODUCTIONThe regular polygons in permafrost ground andin periglacial environments or near water bodiesin cold regions has attracted geoscientists fora century, Recently, mathematicians andphysicists have also undertaken this study andhave made the research range be expanded andshe research content be d.eepened. The researchincludes (1) the existence of polygonal net andits characteristics (Guo Dongxin, 1990, Li Shude,1990); (2) the envirunmental conditions for .theformation of patterned ground, the property ofpolygons and its regularity (Troy I,. PPwP, 1969);(3) the frozen polysons.appeared $gain inlaboratory (Edwin J. Chamberlain et al., 1979);' (4) the origin and the formation mechanism ofpolygons (R.Y. Ray, 1983); (5) Climate andenvironmental change information recorded bypolygons (Wang Baolal , 1991),R.Y. Ray and W.B. Krantz reviewed the historyof formative mechanism research on patternedground and evaluated the .thesessuch as the frostcracking, the desiccative and shrinking crackingand the Rayleigh convection. The crack theseshave not gripped the basic problem--wheie theregularity originates from? R.J. Ray and othersproposed their Rayleigh convection model andpointed out that Rayleigh convection is thecause for the formation of patterned ground.Through Linear stability analysis they obtainedthe critical Rayleigh number and the ratios ofthe width of polygon to the depth of seasonalfrozen layer (W/L), which is 3.18 for polygonand 2.7 for parallel stripes. Sut they did notclearly show the process from convection cellsto patterned ground and only emphasized theundulatory pattern forming by convective actionat the interface between the seasonal thawedlayer and permafrost. Though Lhe physicalprocess bf this model is very simple, it isstill of significant meaning.In this,paper, the stresses are payed to theformative mechanism of patterned ground usin,gmodern physics theories and a composite model ofmultiple actions is put foruard.THE THEORETICAL BASIS OF THE COMPOSITE MODELFOR MULTIPLE ACTIONSThe main thermodynamic processes in groundare the heat transfer, the migration of particlesand water and the convection of water.Although these processes belong to the categoryof nonequilibriun, but it can be determin.ed withquasi-equilibrium when the temperature gradient(s-) or the concentration gradient of particles(+) is not large. By adopting the principleofsuccessive stable state and the Stephen boundarycondition, we can solve equations of thermalconduction and calculate the depth of permafrost,the temperature field of the active layer (G.M.Feldman, 1973) and the temperature field of nanmadefreezing walls (Gao Min,'1989). But theequilibrium theory of thermodynamics is incapableof solving p.roblems of the patterned groundand st.rong moisture migration. While the nonequilibriumtheory devdoped recently can playan Y.mportant role.The nonequilibrium theory in thermodynamicsaffirms that there is a kind of macroscopicspacial and temporal ordered structure existingwidely in nature, Prigogine called I.t "dinsipativestructure". Tt is different from orderedcrystals at a molecular level. Under the conditionswhen open gystem is.far from equilibrhmthis ordered configuration of nonequilibriumwill appear (for example, i € the temperaturegradient is large enough). The system exchangesmass and energy with the outside environment,and it can form and keep B certain macroscopicorder through the mechanism of nonlineardynamics and the dissipative process of energyin the system itself. For example, the Benard'spattern (Fig.1) the circle in granite (Fig.2)and the order of Rio:organization. In quantita-

the reeedrch the nonequilibrium theory is basedon the mas~-balance equations, the equation ofmomentum conservation, the equation of energyconservation and entropy-balance equation. Therelationship bf force X and fluxes J in thermodynamicsis deduced from these equations:J = LXa2Jt +( 7 ) o X'+ ......axaJwhere I. = ( E )oWhen the system is far from thermodynamicequilibrium i.e. the force of thermodynamics isnot weak, the high power of X can not beneglected and the equation is a nonlinearpartial differential equation. The main workof this theory is to analyse its stability. Forexample, Benard's pattern, starting from massbalanceequation, momentum conservation andenergy conservation and adopting the method oflinear stable analyses, the Rayleigh number isderived:R, = na = 657.51g,(To-Td)d'R =YkWhen R>Rc, the Benard's pattern can be produced.Another school of thought of nonequilibriumIn thermodynamics is syneryetics founded byH. Haken (H. Haken, 1982; Jin Baishun, 1982),It studies the system which is constituted by ELlot of subsystems. Under certain conditionsthe macroscopic ordered state will be formedcoordinately by the interaction and cooperationof thk subsystem. The parameters desc'ribing themacroscopic order degree of the systea arecalled order parameters. Through analyzing thestability of equations of order parameters forvarious processes, the results and evolutionaryprocesses at the critical state can be obtained.ACTUAL EV~DENCES AND THEIR CHARACTERSThere were ma'ny data of patterned ground inthe literature about permafrost and periglacialenvironments. We selected some photm as areexample, the polygons in natural environmentsare shown.in Fig.3, 4 and 5, the frozen polygonsof simulating tests in laboratory are shown inFig.6. Though',there are a few photos, they areenough to identify the existence of the patternedground.The photos obviously show that the patternsare very -regular. The patterns are generallyhexagons or other polygons or parallcd stripes,the shape mainly depends on environmental conditions,According to literature, these patternsare formed on seashores and lakesides or at thedrying lake bottoms or in, regions which have hadrich water historically. These areas not onlyare severely cold but also have warm seasons.It is a miracle for the desert field to havesuch regular patterns. It is obvious that thesepatterns arc not man-made, so they are calledself-organization patterns. It Is undouhtfulthat the reasons for their formation are theopening of the system or the unbalanced tomperatureand water content. This coincides with theformation conditions of the macroscopic orderedconfiguration in nonequilibrium and nonlinearthermodynamics. Therefore, we need not feelpuzzled by this phenomenon, it is just Bdissipative structure existing in nature,.SOME MAIN TYPES OF ORDERED CONVECTIONS 'Soil is a cdnplex mixture mainly made ofparticles of silicate minerals, water and void.There are many physical and chemical actionsbetween soil and water. Soil influences thefrozen threshold and conductibility of water, ,and the surface soil particles can absorb water.The freezing and thawing of water can compressor loosen the soil-.body. Compared with soil,water is a very active fac,tor. Because thebulk of water can move (cpnvection), the watermolecule can diffuse and water phase change can.occur. Thus various phenomenons of the watersoilsystem have strong relation to the wateractions.The types of water convections in the soilwatersystem are different in dynamics becauseit is caused by different acting forces underdifferent conditione. But ordered ca'nvections.in physics have a common point which is theaonequilibrium phase change (second phase change)and the mathematical models are similar. Whenthe temperature gradient is larger and thesystem exceeds the critical state, the.regularstructure of convection would occur mainly bythe action of buoyancy-gravity. That is onlyone kind of the ordered convections. Similarto that in R.J. Ray's literature, when thefrozen ground is thawing, it is assumed thatthe ground surface is at or qear 277°K and theinterface between frozen and thawing ground isat 273°K. According to the character of water,the density of upper water is large and thegravity is larger than buoyancy, the upper waterwill tend to sink, The density of lower wateris small and tends to rise. But it cannot movein both opposite directions at the same placeand time. Uniier the influence of other factors(such as permeability) the water movement wouldappear in a regular distribution (Rayleighconvection). As shown in Fig.7(a), the neighboringwater flows of the convection circlehave an opposite direction. According toanalysis of Bernard's pattern, hexagonal convec-,tions should be in the form shown in Fgg.Y(b).The striped pattern of convection is shown isFig.7(c).The convection patterns can also appear infreezing process. The frozen polygon shown inFlg.6 was formed in a horizontal direction inlaboratory, this study showed ,increased permabilityin a vertical direction during freezingand thawing (Edwin J. et al., 1979). Theseresults make known the exi,stence of orderedconvection. It isn't doubted that during thefreezing period, the soil-water system is in anonequili.brium state and may be in a state ofstrong nonequilibrium. It had been known for along time that water migrates to the freezingfront (Kinosita seiiti, 1985; N.A. Cytovich,1973), No matter what the mechanism of migrationis, we can think that a kind of forcesacts on the water in the thawing region withdirection to the freezing front. In literature(Li Guangpan, 1982) it is calculated that thevalue of binding energy of ice is larger thanthat of water, Binding energy takes a negativesign, so the potential energy of water is higherthan that of ice. From the water of hightemperature to the water near the freezing pointand then to the ice, the potential energy formsa varying potential field. .According to therelationship of potential Q and force F,

IF"K , the force should point to the icesurface. Under the action of this force, thewater. migrates to the freezing front and accumulatesthere and as a result the water increasesthere. Taking the downward freezing from theground surface to underground in winter as aexample, Eirst, the water migrates to the freezingfront, and after water far from the freezingfront had migrated, the water migration becomesdifficult. The water content of the freezingfront may be larger than that at the bottom, thediffusion from top to bottom will appear. A tthe same time the gravity is a downward force,therefore the feedback system of dynamics willbe formed. These factors as a whole constitutethe basic conditions for the forming of amacroscopic ordered structure and the hasicconditions are proved by the theory of nonequilibriumthermodynamics. IJnder the nfluences ofother factors, ordered convection 4 will be formed.The third kind of ordered convecdions in soilis caused by surface tension. In lahoratory 'test were conducted by heating the thin liquidlayer with a thickness of 50 micron and ahexagnol pattern was also observed. The orderedand dried cracking polygon at the bottom ofpools could appear in the field after the poolsdried up in summer. In the evaporating processof the soil-water system, the water rises alongthe ca'pillary because, of the action of surfacetension. This leads to a change of.watercontent at various soil layers. Accompanied bythe downward gravity action, it constitutes afeedback system of d:ynamics in nonlinear andnonequilibrium stares and the ordered convecbiontherefore appears. The actual cases could bevery complex, Such as the decreasing of surfacetemperature during the 'night, the buoyancy ofwater is smaller than gravtty which may lead toan increasing downward force.At present the theoretical analyses are toassess critical conditions for the formation ofordered patterns, that is to analyse the stabilityof stab1.e state solution for the differential.equationsand to find the critical conditionsof losing stability. What kind of realisticmacroscopic scructure will appear is relativeto the boundary conditions and the realisticdynamic processes, and it is very difficultto calculate. Tf the principle of synergeticsis utilized and the consider that we.. patternsin the nonequilibrium system are relative tosynergic consistency of the subsystems, then itis easy to explain the reason to the formationof hexagonal nets of the water convection. Thewater molecules link in the tetrahedral strutturethat belongs to the hexagonal system (Pig.8). The molecules of laminary silicate are alsojoined by hexagonal nets (Fig.9). These hexagonsin the subsystem may be relative to the hexagonallin the soil-water system in a certain way, justas the snow-flake is always hexagonal.ACOMPOSITE MODEL OF MULTIPLE ACTIONSFirst, the multiple and composite actionsmean that the action of formed patterned groundnot only is the ordered convection hut also mustbe accompanied by other processes of physicalactions. Second, the ordered convections ofwater are caused by buoyancy or by the.actiona1force of freezing field or the surface tensionas mentioned-above, When the soil-water systemis far from equilibrium, the ordered convectionof water is the origin of the formation ofpatterned ground, but its .not the sole reason.Before the ordered convection appeared, the waterdistribution can be regarded as nearly homogeneous(at least it is true at the Bame la-yer). thewater distributi.on becomes panuniform- after theconvection started. In Flg.Tl(b), the water flowin the boundary of the hexagon centralized moreand joined the strips. Though the water flowin the centre of the hexagon is also centralised;it is only an isolated region and the action ofwater flow is not.stronger than that at theboundary. The water flow slightly scours Outthe soil. In general, the convection not onlyexists' in transient, but continues for a timeand is repeated mqny times. The results ofprolonged action make the combinations betweenparticles in soil became relaxed and the structuralpatterns of nonuniform distribution ofwater are slightly carved in the soil body. Butthis action is very weak, the obvious andmacroscopic patterns can not appear. When other,actions or processes superimpose on the aboveaction in the soil-water system, the originaland cutting traits can be enhanced. The superimposingaction may be the freezing or the dryshrinkage. During the freezing period theordered convection and phase change in the equilibrium(ice up) are simultaneously. In placeswhere the water flow is centralized there i s alot of ice. The volume of ice expanded thedistance betweet1,particles of soil there. Thelarge voids can contain a lot of water. Whenenvironmental conditions do not change basically,,the convection of water will appear at the sameplace the next year (next freezing period). The Imore the non-uniformity of water, the more waterin the boundary. Year after year, this processcycles several times to form more large cracks.the macroscopic ordered pattern then appears.The freezing'force may be large and then .soilbody can be sunk in thawing. Through severalcyclical processes of freezing and thawing therock underground can be moved to the groundsurface. In the drying processes, the soil. bodygradually loses water content and will shrink.The shrinking and the ordered convection arealso in progress simultaneously. The watergreatly influences the soil body in the boundaryof the convection pattern, therefore the combinativeforce between particles of soil isrelatively weak. So in shrinking processecl itis easy to form the tiny cracks in phe boundary.According to the principle of fracture mechanics,once the cracks appeared, it can be developedquickly. The cracks developed along the boundaryof convective patterns can form the o'rderedpatterns i'n the macrocosm. The above-mentioned,the buoyancy-gravity, the acting force,offreezing-difftision and grqvity or the surfacetension-gravity etc. caused various types ofordered convections, these convections and frostheave or the dried shrinking superimpose on eachother and repeat their action on the soil eyertern,these actions repeat again and again and finallythe ordered ground patterns in the macrocosm are formed.That is the composice model of multiple actions.ACKNOWLEDGEMENTSThe works of Prof. Cheng Guodong, AssociateProf. Guo Dongxin and Mr, Li Shvde helped .ua toput forward this model. Mr. GUQ Dongxin andMr. Li Shude and Ph. n. Zhao Xiufeng providedphotos for this paper. The authors wish toacknowledge their support here.1169

REFERENCESEdwin J. Chamberlain and Anthony J. Gow (1979)Effect of Freezing and Thawing on the Permcabilityand Structure of Soils; EngineeringGeology V01.13, 73-92.G.M. Feldman'(1973) The Calculation Method ofTemperature Field on Frozen Ground (inRussian)(>uo Dongxin, (1990) The Permafrost of China,Gansu Education Press.GRO Min, (1980) Mathematical Model of the Variatiunof Freezing Wall with Double Circle ofFreezing Pipes and its Numeral Calculation;Proceedings of the Third Chinese conferenceon Permafrost, 395-399.H. Aaken, (1983) Synergetics and its RecentRealms of Application; Nature Journal Vol.6,No,6, 1983, 403-410 (in Chinese)Jin Raishun, Zhang Jiyue and Guo Zbian, (1982)Synergetics-A New Subject; Nature Journalv01.5, No.3, 1982, 189-195.Kinosita Seiiti, (1985) The Physics of FrozenGround (in Japanese)Li Guanypen and Gacl Min, (1982) Study on theCalculating Method for Bound Water; SoilsV01.14, No.1, 20-24, ,Li Rusheng, (1986) Non-equilibrium Thermodynamicsand Dissipative Structure; Qinghua UniversityPress.Li Shude and He Yixian, (1990) Features ofPermafrost in the west Kunlun Mountains;Proceedings of Fourth National <strong>Conference</strong>on Glaciology and Geocryology (selection),1-8.N.A. Cytorich (1973) The Mechanics of FrozenGround (in Russian).Ray R.J., W.R. Krantz et al., (1983) A Model forSorted Patterned Ground Regularity; Journalof'Glaciology, Vo1.29, No.102, 317-337.Ray R.J., W,B, Krantz et al., (1483) A MathematicalModel for Patterned Ground: SprtedPolygons and Stripes and Underwater Polygons.Troy L. PbwL (1969) The Periglacial Environment,Montreal, McGill-Queen's University Press,Wang Raolai, (1991) Some Advances in PeriglacialEnvironment Studies: Journal of Glaciologyand Geocryology, Vo1.13, No.3, 273-28b.Fig.4 Patterned frozenground in the Aksaygin L(photo taken by Li Shude000IS)Fiy.5 Sorted frozenpolygon in Qinghai-Xizang Plateau (phototaken by Zhao Xiufeng)Y.(photo selected fromLiteratdie of Edwin J.Chamberlain et al. 1979)Fig.7 Schematicdrawing of the circleof water convectionin soilFig.1 Benard's patternFig.2 The circle ingraniteFig.8 Schematic drawing ofhexagona cell (The corn- Fig.9 Schematicbination of water molecules drawing of microstruci-sa tetrahedral arrange- ture of laminaryment and belongs to a silicathexagonal system)1170

CONSOLIDATION OF DEEP LAYERFROZEN SOILS IN TRIAXIAL TESTSLi Kun Wang Changshmg Chcn XiangshengCentral Coal Mining Research Institute, China ,Before triaxial shear test. the results of the deep layer frozen mil tests arc greatly effected by the consolidationor non-consolidation of the sample. The papcr qivtr the preliminary conclusions of the consolidatedand non-consolidatcd samples, and Comparison bclwttn the consolidated mmples &fore and afterfreezing. According to these rcsuots, the authors propose thet before the triaxial test the samplesshould be tcatcd with hydrostatic pressure consolidation. so that the earth pressure is molded as much aspossible after the sample reached the original density, sa thar the test results are more reliable. ClayeySoil samples should be consolidated before freezing and samdy soil samples should be consolidated afterfreezing.TNTRODWCTIONThe method of artificially fazing ground has been widelyused in sinking .shaft engineering in soft and friable soil layer regions. Under ground the frcczini wall is in a three dimensionalstress state.In order to understand frozen soil characteristics in athm dimensional stress state it is necessary for frozen samples atdifferent depth to bc taken for mechanical test. As soon as thesamples arc taken out of the ground, the volume will expand as thethree dimensional pressure disappears, the unit weight will decreasewith the volume expansion. Generally speaking, the fromu soiltriaxial shear strength increases with the unit weight increasing. If asample can not recover its original unit weight, the triaxial shearstrength will be lower than the "real" strength. This is more evidentwhen the original sample is replaced by a remolded sample. It isimprobable that remolded samples co;taining air can be avoided,the original structure of these samples has been disrupted. Due tothe air content, the triaxial strength of the remolded samples is low- .cr, it can not reflect the teal strength.The deeper the soil layer is, the larger the geostatic pressureand unit weight are..If the samples are not consolidated, the errorbetween the triaxial strength of the remolded sample and the "real"strength will increase. The above mentioned tests arc different fromthose in the shallew layer.To reduce the error between ground samples a@ remoldedsamples, consolidation tests should be conducted, so that the originalpressure state can bc molded and the unit weight can be as closeas possible to the oiginal. Hydrostatic pressure (P) of consolidationtests is usually calculated with the heavy liquid theory, Le.:P = 0.13H (1)where, P - lateral earth pressure; MPaH -- depth 01 sample; mAnother method to attain the value of P is to measure the lalcralearth pressure with a prcssure transducer, the method is wore expensiveand takes more time. In these tests the anthors found theconsolidated hydrostatic valuc by calculatiog with Formula (1).PRELIMINARY RESULTS OF THE TEST AND COMPARI-SONSample size: diameter rp + 61.8mm. height h = 1 Sem, Naturalunit weight:r= 19.012KN/rn3 , Water content:W=32.03%, Soiltype: day.The tests were conducted on the FS-I triaxial sheer (creep)machine, the machine was made in Bcijng Research Institute ofMine Construction. The samples were taken from Jining,Shandong Provinm, The samples were fmm clay and the depth ofthe samples was 170 m. The tests were conducted under the followingthree conditions:(1) Freczin~unconsolidatccJ (FUC) triaxial shear test;(2) Freezing bcfore consoiidution (FC) with exhausted andundrained triaxial shear tesc(3) Consolidation before freezing with exhausted andundrained (CF) triaxial shear test.Thetriaxial shear testswerc conducted rtspcctively for conditions(1) FUC, the samples werc frozen at -30°C for 48 hours, thenwere stabilized at the appropriate test ternpcraturc -10°C for 48hours; (2) PC, the first two steps of sampla preparation arc thesame as in (I), after this the samples were consolidated under constantpressure; (3) CF, the samples werc consolidated at room tempcrature.then were frozen at the test temperature under canstantpressure for 48 hours.The consolidation time depends on the arrival of the principalconsolidation point, during tests the consolidation condition of the1171

specimen was demonstrated by the liquid volume entering the prepsure cell under consolidation pressure, The relationsthip betweenthe liquid volume (AL) entering the pressure cell adtime is shownin Figure 1. The principal consolidation point and principal consolidationtime can be obtained from Figure 1.It can be seen from Figure 1 that the regularity of the two consolidationcurves is almost the same or Similar Curves 1 and 2 indicateconsolidations under conditions (I) and (2) respectively. Thevolume changes of the two curves are different. Curve 2 is belowcurve 1, thisshows that the volume compression of curve 2 is morethan that of curve 1, but the difference is very small, about 2-5 ml,the principal consolidation time is close as well. After the principal.I- CFFig.2 Triaxial shear yield line of frozen clayt (mi4Fig.1 Clay consolidation curve(1-consolidation after frozen)(2-frozen after consolidation)(start from A)CCF reaches a maximum value under three conditions, one ofthe main reasons is the unit weight under-this condition. From Figure1 it can be seen that the unit weight of sample FC is less thanthat of sample CF.In trfaxial shear tests the difference of the test conditions affectsonly the cohesion C. with the Drucker-Prager law it doesn'tarfect the value of 8, but affect the value of K, i.e.:consolidation point of the curve 2 appears for 1-2 hours, the samplewas frozen under constant pressure at the temperature of-1OOC. From A the curve starts to bend upwards, i\ shows that theliquid is squeered out from the pressure cell, and indicates that thesample was frozen and the volume also expanded. Curve 2 is belowcurve 1, the unit weight of sample CF is slightly larger than that ofsample FC. If we want to obtain the failure curve of frozen soil intriaxial shear tests, 2-3 Mohr's circles are usually needed. To attainthe goal there are several types of loading. In tests a group ofMohr's circles are obtained with changing confined pressure. Eachsample was tested under three different confined pressures, themiddle one of the three confined pressurcs is the pressure born bythe sample under ground. Under conditions FUC, FC and CF, thetriaxial shear test results are plotted in Figure 2.From Figure 2 it is clear that the difference of the test results isdue to different test conditions. This difference is reflected mainlyin the cohesion C of the frozen clay, the angle of internal frictionhardly changes with the test conditions. The cohesion C is of thefolliwing rule: C, > C, > Cwc In the artificial freezing sinkingengineering, the earth is frozen under earth pressure, howeverthe condition CF is similar, so that triaxial test result is clise to thereal value under this condition. The condition PC and FUC don'tconform with the condition of the artificial freezing earth layers,there is a significant difference between the test result and the realvalue.Drucker--Prager law iswhere I, is the first invariant of the stress, I, is the secondinvariant of the stress, fi is a parameter, K, is a yield value.From Equation 4 it cag be seen that the difference of the testconditions affect directly the yield value.Recently, triaxial shear tcsts were conducted in frozen clayfrom (No.305) Chen Si Lou Mine, the clay lay at a depth of 305 m,through the test comparisons the tendency was found to be thesame as those mentioned above.In addition, the difference of the test conditions affects therelationshipof stress-strain as well, however an essential changedoesn't occur, only the constants differ.SUGGESTIONSFrom above it is known that the yield values of the triaxialshear strcngth change with the test conditions, When the frozen soilwas measured in the laboratory, the testing conditions (includingstress state, freezing pattern, water content, etc.) in situ should besimulated as much as possible, so that more reliable data inaccordance with the in situ testing are obtained, which can bemore1172 *

dependably applied to engineering. The condition CF accords wellwith the test circumstanccs of the ground sample, but the conditionFC and FUC have a large difference, for this rsason it can be deduced that the other two conditions are not quite in accord with theFa1 circumstanccs. Therefore, for the triaxial shear strength offrozen soils in deeper qround, the test conditions should be cansid- .ered. The most suitable condition is the condition CF.For flow sand in ground layer the condition CF can not reflectthe real value, if the test is conducted under the condition CF the *water will bc squeezed out from the sample, thcreforc thc test results will be affected. So, the authors suggest that for sand samples,freezing should be prior to consolidation, in case the test resultslose their true value.This paper gives only the preliminary test results. and providesthe information for scientific discussion.The authors hope that the triaxial shear test can be perfectedand more unified.REFERENCESChengXiangsheng,' 1988, Mechanical Characteristics of ArtificialFrozen Clays Under Triaxial Stress Condition, Ground Freezing88, pp 173-179.Huang Wenxi, 1983, Engineering Properties of'Sul11173

~~ ~ ..~ cryotextureIPERMAFROST AND PERIGLACIAL LANDFORMS INKEKEXILI AREA OF QINGHAI PROVINCELi Shude and Li ShijieLanzhou Institute of Glaciology and Geocryology,Chinese Academy of SciencesKekexili area is located in the hinterland of Qinghai-Xizang Plateau, where themean elevation is more than 5000 m. Climate is arid and cold, the freezing-period is as long as eight months, and mean annual air temperature ranges from-&.lac to -1.O'C. Mid-low latitude, high elevation permafrost is special toChina, the formation of permafrost is controlled by the height, the thicknesschanges from 1 m tu 128.5 m and much more in rock mountains. The geologicaltectonics and surface water body induced taliks, permafrost was formed duringthe late glaciation and neoglaciation, and periglacial landforms are widelydistributed in this area.1. PRRHAFROSTThe Qinghai-Xizang Plateau is called thethird pole of the globe and is famous for itslofty heights, special geological and geomorphologicalconditions. and natural environment.Permafrost, is a product of the natural environmentevolution during the forming process of theplateau, and is widely distributed. Climate andthe natural environment have brought aboutvarious changes with the uprise of the plateau.Kekexili area, with mean elevation of 5000 m,is located in the hinterland of the Qinghai-Xizang Plateau, where the climate is arid andcold and terrain is lofty. The height rangesfrom the maximum peak Bukaleike summit of 6860 m(also called Xin,qing summit) to the lovest sitein the south piedmont of Kunlun mountains,transcending Rekaleike mountain to HongshuiRiver bend, where elevation is C200 m.The studied area extends northward, to themain ridge of mid-Kunlun mountain, southwardsto the Geladantong glaciers in Tanggula Mountains,east as far as the Qinghai-Xizang highway,and westward to the boundary of threeprovinces or regions: Qinghai, Xizang andXinjiang (Tig.1). The area is approximately 0.83million km .Kekexili Is the most intact area of preservedplateau surface on the plateau and also is themost developed area for permafrost. The cold andarid environment and thermal conditions providea good condition for the formation and developmentof periglacial geomorphologics andpermafrost.zones from north to south.(1) Ice, snow and perennially frozen rock inthe lofty mountains of Kunlun ridge, where thereare many modern glaciers, and the mean annualground temperature is below -3.5-C. The thicknessof permafrost exceeds 120 m and approaches400 m in the rock mountains.(2) Permafrost zone within the drainage basinof the high plain within Chumer River on thesouth piedmont of Kunlun mountains, where, themean annual ground temperature ranges from-1-.2'C to -3.5'C and permafrost is from 40 to100 m thick.(3) Permafrost zone in hills and high mountains,(more than 5000 m) such as, Kekexili hill,'Wulanwula, Wuerkewula. Dongbule, Fenghuo Shanand Wudaoliang, where mean annual. ground temperatureis -1.4 to -4.O'C. and permafrost is36 m to 120 m thick. Ground ice is a welldeveloped periglacial form and appears widely.(4) Permafrost zone in the expanse of thevalley and basin in Tuotuo River, where meanannual ground temperature changes from zerodegrees to -l.OqC and permafrost is I m to 50 min thickness. Thefe are fluvial and permeabletaliks occurring within the zone.(5) Perennially frozen zone in lofty mountainsof Geladandong in Tanggula ranges, wherethe height is more than 5000 m. mean annualground temperature ranges from -1.7'C to -4.5".There are a lot of mcidern glaciers and welldeveloped ground ice. The thickness of permafrostchanges from 10 m to 128.5 m and reaches300 m on mountain rocks (see Table 1).1.2 Ground Ice and Crvotexturk1.1 Distribution..~Temperature .and Thickness ofPermafrostGround ice forms an important but variableof frozen grThi .S area is more than 500 km lonn from south in the upper laver of permafrost in this area.to north, about 400 km wide from east to west, such as the density, network and layer cryotex-Permafrost is continuously. distributed andture, and layer cryotexture of them is common..occu'pies about 90 per cent of the investigated The variable ice layer is from a few millimetersarea, and permafrost can be subdivided into five to several meters thick, alternately occurs in. ..1174

the upper layer of permafrost, ice layer withthe level bedding regularly develops in thelacustrine sediments of Chumerhe high plain, andchanges from a few centimetres40 tens of cent$metres(Zhou Youwu, 1981). The ?ce layer, 10 to12 cm thick was found at the depth of 20.6 m inthe lacustrine sediments near No.68 maintenacestation of Qinghai-Xizr-..g highway. The volumetricice content of permafrost approximates 30to 50 per cent. The ground ice can'be subdividedinto three types in this area, by their forma,-tion conditions and distribution features. Thatis cemented, segregated, buried glacier ice andfissure ice.1,3 The Formation and Development of TaliksThe causes of the formation of taliks in th isarea are complex. The formation and evolutionand characteristics of taliks are controlled b Ythe climate, environment. geological tectonicsand hydrogeologic and surface cover conditionsGeological tectonics is an underground factorthat always influences on taliks. Based on thepredominant factors determining the appearanceand existence of taliks, the taliks in thisarea can be divided into three classificationsas follows:1.3.1 Tectonical talikThis type of talik is related to the tectonicalfaulting and magmatism, it is formed whenheat ground water arises along the fault zonesand brings about heat influences around thestratum. It often occurs in the faults in thesouth of Kunlun Mountain ridge. For example,there are a lot of hot springs in the southpediment of Buhadaban Peak (temperatures up to90"C), it is estimated that the talik extendsmore than 200 m from south to north: others,such as, the talik,extending from Tanyang Laketo Hongshui River.1.3.2 Talik of surface waterThis type of talik was formed due to thethermal influences of surface water bodies onpermafrost, it is also subdivided into fluvialand lacustrine taliks by the category of waterbody and the features of talik. For perennialrivers,the river surface is frozen during winterbut river water still flows. The heat transmissionand insulation from the water body causeslinear talik to form in the beds and lateralsof rivers. There are fluvial taliks under TuotuoRiver, Tongtian River and Buqu River, and theone in Tuotuo River is as wide as 875 m, and isthe largest one on the plateau (Qiu Guoqing,1982; Shang Jianyi, 1982 and Guo Dongxin, 1982).1.3.3 Permeable and radiation talikThe formation of this type of talik is aresult of atmosphere precipitation permeationand solar radiation. Such as those in the terracewith sandy sediments on the north bank of 'TuotuoRiver, crescent dune near Wuxijin lake and inbare gravel surface in the valley of Buqu River.The presence of the taliks mentioned abovecauses the integrity and stability of permafrostto he damaged, the continuity, temperature andthickness of permafrost to be disturbed andcomplicated.1.4 The Development' History of PermafrostMany data (geologic, geographic, paleontologicalpalso-glaciers and palao-preglacial) haveshown that the mean height of the plateau wasabout 2000 m in early Pleistocene,. 3000 m inmid-Pleistocene.and 4000 m in late Pleistocene.Evidence of glaciation in the high plain ofChumer River, and involution layers in lacustrinesandy clay in Kunlun Shan pass, dating to aboutthe early Pleistocene, indicated the glacialand preglacial environment existed on theplateau in mid-Pleistocene, when permafrost wasexpansive. After glaciation of mid Pleistocene,air temperature rose and deglaciation commenced.Air temperature at that time was higher by 11°Cthan at the present, in Qingshuihe area, permafrosthas been disappearing mostly, with theexception of the high mountains. The air temperaturedecreased again in the b,eginning ofthe late Pleistocene when the glaciers advanced,for example, the Zhufeng glaciation. The secondor third glaciation had occurred in Kunlun andTanggula Mountains. Although we found noevidence of Pleistocene glacier advancing inKekexili area, it is a fact that periglacialenvironment covered this area at that time. Asa result, permafrost expanded, other evidencefor this are sand wedges in the No.2 terrace ofTuotuo River (dating 23500*1200"C B.P.), andsand and gravel wedges were also found in theNo.2 terrace of Chumer River, Gangqiqu, Mazhangcuoqing,Tongtian River and Buqu River. About3'000-1500 year ago (Zhou han cold stage) neoglaciationcommenced on the plateau. The grcuundwas refrozen aad coincided with the perenniallyfrozen layer formed during the late Pleistocene.The hole in Xidatan indicates that the permafrostlayer and underlain humus 4,4 m thickwere formed during neoglaciation: on the No.1terrace of Nachitan, involution developed inthe upper layer of sandy soil with charcoalfragments which gives a date of 4910f100'4Cyears B.P. indicating it as the north lowerboundary of permafrost during neoglaciation(Pu Qingyu, 1982),1.5 CoolinR SoilThere is a special soil occurring in thedried and seasonal lakes of this area. A grey 'or grey white salinized soil, underlying thesilt and sandy clay of the lacustrine orfluviolacustine facies, with 30-40 percentmoisture content and high salt content, measurementsshow that it will freeze at,temperaturesbelow -4°C - -7°C.2. PERIGLACLAL LANDFORMS 'Due to the circulation of freezing and thawinga lot of cryogenetical phenomena were formedin the active layer and upper part of perma-frost'in the expansive permafrost region. The developmentand movable process of periglacial landformsare mainly controlled by the exogonicforce, whose property is related to the specialheight and geological-geographical factors inthis area. Based on,the dominant action formingthe varied periglacial landforms, which can be, divided into seven classifications includingmore than 70 forms which are presented asfollow (FIg.1).2.1 Frost MoundsBoth perennial and seasonal frost moundsdegelop in this area. Such as those in the basinof Kunlun Mountain pass, in the headwater ofSema River and in the gorge and divide betweenGangqiqu and Mazhang Cuoqin. The famous perennialfrost mound in the basin of Kunlun Mountainpass, is up to 18 m high, 140 m long, and 45 m .Iwide. There are two depression from the mound1175

Map of Periglacial Landforms in Kekexili Area1. rock streams bl.ock field: 2. block slape rock glaciers: 3. periglacial pillarperiglacial loess; 4. debris cone icing; 5, sand dune soli.fluction tongue;6. macro sorted circles micro sorted circles: 7. frost mound earth hummocks;8. patterned ground palsas; 9. frost block wedge shaped body; 10. explosivefrost mound involution; 11. settlement lake settlement depression; 12. glaciofluvialterrace massj.ve ice: 13. periglacial castle hot spring; 14. glacier.General Table of the Seasonal Thawing Depth and the Temperature and Thickness ofPermafrost in Kekexi.1i Area, Qinyhal Province , .SiteLatitude Elevation MAAT (oc)Thickness of Seasonal thaw(NO 1 (m) MA' ("C), .permafrost depth (MIKunlun Shan 35040' 4800-5000 below -3.5 -2.8 - -3.5 7 5- 100 1.5ChumerRiver 350207 ~480-4500 -6,2 -1.2 40 2-3Wudaoliang 35"15' 4610 -6*5 -1.4 36-h0 3-3.2Fenghuo Shan 34'20' 4700-5100 -0.6 -2.0 - -4.0 60-120 , 1-2Tuotuo River 33"50' 4500-4700 -4.4 0.0 - -1.0 1-50 0.8-6Tongtian River 33"30' 4500-4600 -4.4 -0,3 - -1.0 25 1-4Tanggula Mt. 32-57' 4900-5300 -6.4Tangquanguo 32"40' 5000 below -6 4- 10-1.20 1-3- 128.5 2.8Zhuonan Lake 3.5'18' Area 4800 about -6.5 74 .a 2.4-

subsidence occurring in the north anfi southsides of the mound enclosed by a soil ridge,which is more than 10 m high. Groundwater overflowson the top of the mound to form theseasonal icing which is 1.5-20 m high.2.2 Earth HummocksThese dome shaped hummocks were formed by theair circulation of frequent freezing and thawing,small plactic moun;:s are usually covered withgrass 20 to 40 cm in diameter and 10-30 cm high.2.3 RockglaciersIt is a special geomorphological form in thecold permafrost region, it is also called amovable permafrost body, and consists of clasticdebris, breccai and ground ice. Such as therockglacier in Kunlun Mountain which is uniqueto China (Cui Zhijiu, 1981).REFERENCESGuo Dongxin and Li Shude, (1982) The formationand history of massive ground ice inFenghuo Shan on Qinghai-Xizang Plateau.Proceedings of <strong>Conference</strong> on Glaciology andGeocryology (Geocryology), Science Press.Pu Qingyu, (1982) History of permafrost alongQinghai-Xizang highway. Proceedings of<strong>Conference</strong> on Glaciology and Geocryology(Geocryology), Science Press.Shany Jianyi, (1982) The development featuresof permafrost along Qinghai-Xizang highway.Proceedings of <strong>Conference</strong> on Glaciology andGeocryology (Geocryology), Science Press.Cui Zhijiu, (1983) Discussion on rockglacier,inKunlun Mountains. Proceedings of SecondNational <strong>Conference</strong> on Glaciology andGeocryology, Gansu People's Press.2.4 Settlement Depressions or LakesThe landforms in the handwatcr of ChumerRiver, near Xinxing Lake and between XijinwulanLake and Yonghong Lake are surface disturbancesresulting from disruptions of the thermalequilibrium of underlying permafrost due to thevariation of surface conditions.2.5 Periglacial Loess and Sand DunesWith a lofty and expansive terrain, cold andarid climate and the effects from violentlyaerological west wind currents make this areabecome one of the areas with maximum wind speedwithin the plateau or in China*The loess deposited under the periglacialenvironment is called perlg.lacia1 loess, itwidely occurs in pediment, terraces and valleys.Sand dunes, crescent dunes also'occur widely inthe north side of Xijinwulan Lake and alongQinghai-Xizang highway, and impedes on thehighway.2.6 Frost Jacking of StonesA lot of frost jacking of stones are'oftendiscovered in this area. Some of them stand ingreat numbers like trees in a forest on theslope. Such as those on the slope on the bankof Kekexi1.f Lake, where sandstone and slateare widely distributed.*Information from investigation for Kekexili.

REGIONAL FEATURES OF PERMAFROST IN MAHAN MOUNTAIN ANDTHEIR RELATIONSHIP TO THE ENVIRONMENT'Li Zuofu, Li Shude and Wang YinxueLanzhou Institute of Glaciology and Geocryology,Chinese Academy of SciencesDrilling and ground temperature measurements at some sites in the study areademenstrated that the presence of permafrost within the loess plateau in Chinais an irrefutable fact. Permafrost distribution is sporadic and island.,varyingin extent from tens of square metres to l.5x105m2 and in thickness from about3 or 5 m to 30 m. Sporadic permafrost was first discovered in China. A key topermafrost occurrence is high elevation, through 2 range of above 3500 m, thespatial dis'tribution of permafrost areas is complicated by the local conditionsSTUDY AREANahan Mountain, (35"45'N and 103"45'-104°00'E)3670.4 m above sea level, located about 40 kmsouth of the city of Lanzhou, is the highestpoint within the loess plateau. The summit is aplanation surface formed in the Pliocene epoch(Liu and Li, 1991) and has a generally subduedrelief. There are scattered rock outcrops andpoorly developed Quaternary sediments includingremnant blocks formed by frost weathering.The climate of Mahan Mountain is subhumid andcold although it is located in the expanse andarid loess plateau due to the lofty height andair temperature abruptly decreasing with theincrease of height. Mean annual air temperatureand precipitation are -2.3-C and 494 mm (observedin 1961, after by Re Binghui, 1981). The freezingduration is as long as seven months. Especiallylapse rate of air temperature in winter (fromDecember to February) is only 0.3"C/100 m, lessthan the normal value, based on the relativestatistics between air tempwature and heightat 24 weather stations near the study area.PERMAFROST DTSTBIRUTIDNIn this a'rea, mean air temperature is -2,3'Cyearly and 9.2'C in July, the difference betweenthe ground and air temperature is 2.8'C (observationin 1991). Permafrost was found in Oct.1985. Therefore, four sections, including 12sites, where holes and pits were dug wereselected for the programs of the investigationof permafrost distribution and measurement ofground temperature in 1991. At least, eight ofthem have been identified to have existingpermafrost (Fig.1).Of the four sections, Section A (termed Xiaohutan),3560 m above sea level, an area of"The project supported by National NaturalScience Foundation of China. ,approximately 0,5 km', situated on the northfacing slope of Mahan Mountain, is an ellipticalhollow with poorly drained and well developedhummocks. In winter it is free of snow pile,only thin snow fills in the trenches of theearch hummocks. Four holes were drilled on thebottom of the hollow in July and September 1991to obtain information on perennia.11~ frozenground and bedrock and to install thermometersfor monitori.ng ground temperature. The results(Fig.2) indicate that permafrost occurs on the 'bottom of the hollow varying in extent of about1.5x105m2, almost cpincident with the occurrenceof the well developed earth hummocks in thicknessfrom about 5 m on the edge and 30 m in thecentre of the hollow. The depth of the activelayer is 1.5 m on the edRe and 1.2 m in thecentre of the hollow. No permafrost occurs onthe slope around the hollow, where the depth ofseasonal freezing varies from 3.5 m to 4.0 m.Section B, adjacent to Xiaohutan, situated onthe back well of the paleo-cirq'ues of Doulingtrench, Three holes were drilled on the solifluctionsediment and nivation holloy. Of the three ,holes, the first was selected in R nivationhollow with well developed earth hummocks and. ,solifluction, Prevailing north westerly winter .,wind accumulated a thick snow cover (thickness:40-50 cm), near the surface soil moisture wasvery high, the depth of seasonal freezing was2.5 m, The ground temperature at a depth of 5.lmwas 0.8'C during early July to early August. Nopermafrost exists in this site. The second andthird were all selected on the solifluctiontougues with an average gradient of 12". Thereis no redistribution of snow in winter, thedepth of snow varies only from 10 cm to 15 cm.Although they have a steeper gradient and arewell drained, near the surface soil.moisture insummer is very high. As a result, the maximumdepth of seasonal freezing reaches to 4.0 m.the ground temperature at a depth of 3.5 m wasfrom -0.1'C to -0.2'C during early July tu, .1178

hFig.1 The map of permafrost and periglacial Landforms in Mahan Mountain1. Bush upper limit; 2. Island permafrost; 3. Sporadic permafrost(amplified): 4. Icing; 5. Earth hummocks; 6. Nivation hollow:7. Glacial deposits; 8. Solifluction terrace; 9. Block fields;10. Cirque; 11. Boreholes and pi.ts; 12. Roads: 13. Investigation sectionTemDcraturc PC) . .-2-1 0 1 2Temperature ("C)-3 -2 -1 0 1 2 3O 7guessedmeasuredeHYb: June 176- c: Nov. 228-d DGC. 151012--IIFig.2 Ground temperature envelopes at islandpermafrost in Xiaohutanearly August and up to 1.2'C until late Sept,There was also no permafrost in these sites,However, permafrost was found between thesesites in October 1985, where a 1.2-m-thick layerof clay with debris overlying dolomitite withdeveloped fissures in which ground ice, primarllyin the form of fissure ice up to 2 m thick(average thickness beimg from 5 to 12 cm), wasencountered at a depth of 8 m below Eroundsurface (Li Shude, 1986). Based on the abovementioned and hydrochemical analysis of icesamples taken at the depth of 8 m below theground surface, permafrost exists at the bedrockin this site (Fig.3).Fig.3 Ground temperature curve of sporadicpermafrost in bedrock'Section C, from Honghalang trench to thesummit. At Hutanliang, one hole was drilled ata modern nivarion hollow at 3640 m elevation ona scattered meadow, with coarse grained materialand a snow patch 0.3-3,4 m thick. Although meltwaterfron patches of snow accumulated in thehollow flows near the hole, the soil moisture inlate Autumn is low owing to the steeper gradient(gradient of 15 to 20°), the maximum depth ofseasonalfreezing was 3.2 m, the ground temyeratureat a depth of 3.6 m was 0.1OC in early Julyand increased to 2.2"C in early August; At3520 m elevation on the solifluction tongue atthe bottom of Douling trench has an average

gradient of 3" to So. Vegetation consis~s ofalpine azalea and c.ryptomeria up t.ij SO cm highwith' scattered growth. Recaust. thc surface ru11-off from the upper slope flows arounrl thv solifluctiontongue therc was no organic laycr butsoil moisture was very high or the soil remainr.11essentially saturat-ed through the thawing season.The runoff has no thermal affcct on the chitrlgeof soil temperature. Ground tempcr,oture at adepth of 3.5 m was -0.Z'C from esrly July tolate October. The maximum depth of seasonalthawing occurs until mbd-December (depth: 2.Om).A t this time, however, thc ground temperarure nta depth of 3.5 m was -0.1"C (Fig.4). It isestimated that permafrost is about 3.0 m inthickness and a few tens of square metres inextent.4tFig.4 Ground temperature curve of sporadicpermafrost at Honghalang trenchSection D, situated at a paleo-cirque withsubdued terrain in Tudixian at 3520 m elevation,and plant cover consists of an alpine meadow.Local swamp or solifluction tongues with welldeveloped hummocks also occur at these sites.Snow cover in winter varies from 10 cm to 20 cmin thickness, two holes were drilled at theswamp and the no swamp solifluction in July 1991,respectively, to contrast the freeze and thawprocess at different microterrain units, but themeasurement results of ground terrpurature aredifferent between them.The distance between them is unly 30 m, oneat the swamp solifluction tongue with permafrost(Fig.5) and at the other with none.4tTemperature ("C)Fig.S Ground temperature curve bf sporadicpermafrost at swamp solifluctiontongue in cirque of Tudixian'The above mentioned ~'erntafrcrst, distributionin this :.lrc>a is quite complex.Alpine pttrmnfro.;t i n M;lhon Yuuntain can alsohc suhdivided inio isl;i~~d rind sporadic pcrmafrost.'The former is similar 1.0 those in othfrregions ot China,' the lnean ar~nual nt[,un(l t1.mper;lt.urt!is between -0.Z"C and - 0 . 3 O C . 'lhe~r(~un(I tempcraturc grorlicnt prufilc is almostzero, so, i t. is wnrm prrmafru5t which usually issusceptible to the environmrnt;~I nhanges but itstill exisLs uit.hir~ t.he luess p:3t.(!:ru with onarid or scmi-arid climate in mid--Latituderegions rrnd is about 30 m in thickness. Thedepth of' annual zero amplitudc was e11r:ounteredat a depth'of 12 m below t,he grourld snrl;i(.t+.The thickness of the active layer and the depthof annual zcru amplitude arc all lower thanthose, in uther alpine permafrost regions ofChina. These at-e largely due to t,he originallayer mear the surface (thickness; 30 to 40cm)which ~-crntoins much morc moisture, and acts asan insulaiing c(~vcr, just 8s those above Klunc1,akr (Var,ris, 1987), partly due to the present-eof supra-perm.Jfrost war-er in !.he hi11 lnw. Tnarid i t. i on, massivc ice and ice-rich soi 1 develnfre~lin the 11pprr layer ut pcrmafrost which retardspermafrost degradation. As a result, i t can bepreserved i.1 the loess plateau as secn atpresent.The sporadic permafrost was first suggestedhy Gorhunov (1978) who found perennially frozenhurlies, that occur in the debris and clasticsediments and exist. for & few years in the mid-Asian mountains, its occurrence has heen notreported in China. So it is unique, that one ofthem was encountered in bedrock with fissures,it has B thickness of about 10 m, and its meanannual ground temperatqre approximates zero, thedepth of seasonal thawi.ng probably exceeds 4 m,and mean annual air tcmperaturE is below OOC.The reason for scattered distribution appearsto be that it is located near the lowcar limit of'alpine prrmafrost and in thcrmal fquil'ibriumwith the present surface temperature or nlter~ativelymay represent remnant permafrost degradingin response tu an incrcase of surface temperaturcto above O'C. Especially snow patches insulatethe ground from the cold in winter, so that nopermafrost occurs in the sit.es mentioned above.Other sites with permafrost are at solifluctiondeposits at Hc~ngh~long trench and swampsolifluction tongue at the cirque of Uudixian,which have zero gradient in ground temperatureprofile and mean annual ground temperature isabove 0°C. Permafrost is about 2 m to 3 m inthickness. They may he largely a function oflocal conditions. For example, there is thinsnow cover and much moisture fur fine grBinedsoil in summer in these cases. They may he inthermal cquilibrium with the local surfacetemperature, so that sporadic permafrost onlyexists in the microterrain.EROUNI)ICEUp to now, epigenentic ground ice has beennot found in the study area, although someresearchers think the sediments at. Doulingtrench to be rock glaciers. Syngenetic groundice, huwever, including fissure icc in bedrock,and segeretion ice, is controlled by localgeological and geomorphological conditions.Fissure ice in bedrock develops on the summit or

"Isections along the ridge. For example, thoseencountered in the exposure dug at the backwallof the cirque of Douling trench, Hydrochemicalcomposition for the Ice sample taken in a depthof 8 m below surface is different from those ofgroundwater from the solifluction tongues, icingand snow samples taken around the site. Tritiumcontent of 11.63 TU indicates that the fissureice was formed before the 1950's (Li Zuofu andLi Shude, 1986). In addition, fissure icc inbedrock occurred on a mining exposure in e3rlyJuly and was preserved until late July 1991.Ice stalactite formed at a man-made cave in thesummit and was preserved until late August.Duration is as long as eight months.Segeregated ice mainly develops in the microterrainwith thickly fine gralne'd soil. So inthis area, Xihuotan is an unique site wheresegregated ice develops, Borehole data shows thevertical variability of ground i.ce content.Total ice volume decreases from values of morethan 90 percent at depths of 1.2-1.7 m belowsurface to 45-55 per cent at depths of 1.7-3.5111,and frozen sediment at depths of 3.7-6.2 mcontaining no visible needle ice. They may beclassified in terms of the field observation andhave a relationship to enclosing scdiments.Lens ice or an ice layer with silty particles53 cm thick. Grey silty particles were apparentlyin suspension within the ice, roughly 3-7 percent by volume, lens with uniform coarse crystal.size was simllar to columnar. Vert-ically orientedand elongated crystal structure i s perpendicularto the overlying sediment: Layer tce occurs at adepth of 1.7 m below the surface, the totalthickness is 2.0 m (each layer being 1-2 cmthick) layer cryotexture and short columrrar icecrystals. When it was melted :he core lookswhorled.M A ~ N FACTORS EFFECTING PERMAFROSTa ElevationPermafrost found in this area exipted at thelowest heights of ahovc 3530 m and at the highestelevation of 3620 m (on the north facing slope),which were determined by many investigations inthe s~te. Especially, the reconnaissance studyof permafrost distribution in detail and measurementof ground temperature which was carried outin 1991. Through a range of the heights, onesite has permafrost but another will have none,even though they are only a few meters apart.Meanwhile, on the north facing slope permafrostoccurs at lower heights, but no permafrostexists at higher elevations. Thus:, the highclevation is a key factor to determik~e whethcrpermafrost occurs or no^, through Lhe range ofthis height, regional factors have a great dealof effect on the occurrence, Lemperature. andthickness of permafrost and Rround ice." FacinK SlopeFrom statistics, within the alpine permafrostregion in the Northern Hemisphere, the lowerhoundary of permafrost on the south facing slopeis high, 300-400 m more than that on the nort-hfacj.ng slope (Qiu, Huang and Li, lYR2). Based onthe difference between the south and northfacing slope mentiorled above. The 3600 m heightappears to be a critical one for permafrostoccurrence in this are3 from Lhe geogruphiclocation, mountain height ancl cryogenic phenomenadeveloped on the north facing slope and on. thesummit. Through a range of the elevation, steeprelief or. the north facing slope is advantageousto the blocking of the cold 'current in winter,so that cold duration becomes much longer on thenorth facing slope than that on the south facingslope, In summer, it makes a vertical currentdevelop, so t.hat on thc north facing slope cloudcover increases, and incoming radiatiorl reduc.es,which are also advant.ageous to the decrease ofmean annual surface temperature, meanwhile, itis much wetter on the north facing slope.' Forexample, there js alpine azalea up to above3600 m above sea level, and only on t-he alpinesteppe on the south facing slope. These differencesbetween them have a great.effect to thermalinterchange between the air and the ground.There is no possi.bility of existing permafroston the south facing slope in this area.Terrain and Soil""As mentioned above, the loosened sediments instudy area are thin and overlifs the coarsedebris 0.3-1.0 m thic.k. 'Thus, under thp sameconditions, the thlckness of line grained soilis very import.ant for the absence or presenL ofpermafrost. in the sites with a thin layer IY'?fine grained soil, Pcrrneation from water, makesheat in soil qujckly migrat.e downwards, so thatthe rate of thawfng of t.he sol1 would be fasl,owing to suitable permeabil.ity and steepergradient.Measurement3 of gruund teaperature at thev.atious sites near Hutanlieng show that isogeothermsuf 3.2"C at e depth of 1.2 m occurted on.July 15 at tlfe site wifh coarse debris. Sut atthe sites with fine grained soil it occurs at adepth of 0.5 m until. 1at.e September and hasncver been at the depth of 1.2 m annually. Thisvariation was largely duc to the difference inthe thickness uf finr gr~ined soil with time asthe water permeates through the fine gratlnedlayer to the coarsc debris.Tn tlddition, island pormefrust exists inXiaohutan and its active laycr varies from 1.2 mto 1.5 m in thickness. Thr rcasirn for thiv isprobab1.y partly due to organic rich muck andpoor drainage 111 summer and partly due to thccool jng of the snow cover in trenches UT hummockstcl the ground surface 'in winter. Thus, under thesame conditions, the thickness of fine grainedsoil. .in 1.he study 8rt-a appcars LO be the majortact.or controlling the absence or yresencc 05permafrost.Snow CoverSnow C O V ~ J inrluenccs t.he hcat t.ransfcrhetwecn the .lLr and thrt ground, and, hcnce,affccts the distrjhution 01' permatrust, due turedistribution of snow (-over Ly wind. The snowcovcr dl. the study area exhibits cons~derabicspatial vatia1,ion in deyt.1~. Sump snow patches33-53 cm thic-k per-sist. on the nivation hollowsthrough wi.nter to retluce winter heat loss. 'I'hr.hollows havc relatively thick loosened deposits,hut a thin snow cover filled in t.he trrnches ofearth hummocks rema1113 for u long t,imu, most ofthe area studied is exposed LO wind. 'Thus theeffect ot the snow covcr 1.0 grou~~d surtace isboth insulat.i.ng and cooling i n this arva.Insuloting; insulation of snow cdver to groundsur,fal:t? is with a relation to heat c~rrles insoil, because the alca studied is In thc iangt'of 3500-3650 10 elevation whicl~ is a criticalheight for permafrost tu exist a1111 which I?RSviulent heat cirtles in soil. Therrfore, snawCOVF.~ insulates the ground from thc wint.er cold.For rxt+mple, on the Lack wal I of thc cirque inDouling trench, a snow patch on t.he hollow was

50 cm thick, the depth of seasonal freezing wasabout 2.5 m. But Snow cover on the solifluctiontongue varies from 15 to 20 cm in thickness. Thedepth of seasonal freezing was about 3.5 m toh.0 m, observed at Xiaohutan on December 15,1991 showed that the temperature at noon was2.0"C in ground surface, -0.6'C in snow surfaceand -1.6"C under snow; and that temperature inafternoon was -1O'C in ground surface and -13.2'Cunder snow.In addition, snow patches on the north faci.ng-slope cause the thawing data to be postponed for20 days. It appears that the snow cover acts asa cooling cover when it is thinner than 20 cmand as an insulating cover when thicker than30 cm.In summary, on the north facing slope inHahat1 M0~11taln the occurrence of permafrost iscomplicated by the general influences of soilmicroterrain, and gradient, and the alternativeaction fron insulating and cooling of snow coverto the ground.REFERENCESRe Binghui, (1981) A study on Quaternary glacierand periglacier in mountains near Lanzhou,Journal of Glaciology and Geocryology,vo1.3(1).Liu Yun and Li Jijun, (1991) Late Quaternaryglacier and environment in Mahan Mountains,The Quaternary Glacier and Environment ofWestern China, Science Press, Beijing, China.Li Shude, (1986) Permafrost was found in MahanMountain near Lanzhou, Journal of Glaciologyand Geocryology, Vo1.8(4).Li Zuofu and Li Shude, (1986) Permafrost on theloess plateau in China, GeographicalKnowledge, No.9.Qiu Guoqing, Huang Yizhi and Li Zuofu, (1982)Features of permafrost in Tianshan of China,Proceeding: 2nd National Permafrost <strong>Conference</strong>,Gansu Press, p.21-29.Gorbunov, A.P., (1978) Proceedings Third <strong>International</strong><strong>Conference</strong> on Permafrost, Vol.1,p.349-353.Harris, S.A., (1978) Arctic, Vol.40, No.3,p.179-183.

A SOLUTTON FOR THE ICINGHEAVE OF FOUNDATIONS INPERMAFROST REGIONS BY LOWERING THE GROUND WATER TABLELiu Shifcng and Zou XinqingThe Institute of Forestry Design of Da Hinggan Ling'The frost heave of foundations is one of the most important problems damaging buildings in permafrostregions. Prom many years of surveying the damage to the buildings in Da Hinggan Ling region, particularlyin dealing wity the heave of foundation of residential buildings in Gulian mining area, we propose amethod of lowering the ground water table to solve the frost heave of foundations in permafrost regionsand to obtain some results. Through two years of observation it was found to be cfftctive.ANALYSIS OF FEASIBILITYBased on the freezing principle, the frost heave that ocam insoil must be qualified with three conditions: 1) susceptible soil, 2)preliminary water contcntand externally supplied water. 3)optimum freezing conditions and time. The frost heave can be r esisttd by the weakening of one of the three conditions, so that thefrost damage is protected.When the soil properties arc constant, the amount of watercontent is one of the basic factors affecting the frost heave. In aclosed system, the water capacity determines the frost susceptibility,In an open system, the cxternally supplied water can greatly raisethe susceptibility of soil, although the preliminary water content issmall. Both the water content of soil and externally supplied waterare related to ground water. There is rirrely rain in the developingpermafrost area in Da Wiggan Ling. The surface soil begins tofreeze from October to April, the water supplied from the surfaa isreplaced gradually by ground water. The groundwater table affectsdirectiy the water content and the supplied water during soil freesing. As a result, the frost susceptibility of soil is decided. The moreshallow the ground water table, the lower the frost heave. Thus it is *feasible to decrease the frost heave by lowering the ground watertablc.SPECIFIC IMPLEMENTSAfter the structure of the foundation of the four residentialbuildings is Gulian mining area were completed, a water zone webled from the foundations in October that year and ice formed later..The exit was frozen in the middle of December, and an icingformed about 60m ling in size, 20m wide and 1.2m in averageheight. When the surface ice was dug out, a column of cinfincdground wadr was spread which reached to O.6m high and its fluxwas 1st / h. The coatcr column changed into welling water afterhalf an hour, and its flux was 6t/ h. The icing expandedcontinuously with time, andthe girds ofthe foundations might ofken damaged at any time. In order to determine an optimum solvingmethod, we investigated the hydrogeological condition.The area is located on the north slow of Da Hinggan Ling,where it experiences a continental climate in a tempera1 frigid zore,and altitiude is 444)"--6oom. The summer is short and hot and thewinter is long and mld. The annual air tcmwrture is -ST,theminimum air tcmmrature is +PC, and the maximum air ternperatureis 36.8"C.The residcntial buildings were built on the low ridge and gentleslope. The topography declines from northeast to southwest. Thereare geomorphology units of denudated a&umulation and frozensoil pcriglacial forms which arc in the expression of icing in winterswhich forms swamps after thawing in summer. There ir no obviouswelling water, and streams exist. The lithological character of thestratum is simple. The ground surface is clayey sand and loam withstones formed of Quatmary slopc wash and deluvium. Thebedrock is tuffy gravel rock on the granite is variscian.REGIONAL HYDROG-EOLOGYOwing to the artificial influences and the destruction of vegetation,thc permafrost in the upper fringe of the :UW ridge decreases,but there is permafrost in most parts of the ama. Therefore, theclassification and motion of the ground water is limited bypermafrost. Based on this, the ground water is divided into twotypes: supcrpcrmafrort water and subpcrmafrost water.1183

The SUpcerpcrmafrost water which is affected greatly by mercntSEasons is stored in the pores of the Quaternary unconsolidatedlayer and in the pores and crevasse of strong weathering gravel. Aftcrthe seasonally frost soil thaws, the water is directly supplicdfrom the rain and the meltcd water from ice and snow."he subpermafrost water is stored in the pores and ccvaw ofbedrock. The result of hydrogeological exploration illustrated thatthe crevassc of badrock is developed and provides the conditionsfor the storing and flow of water. The subpermafrost water powsa confined character, because the fmzing layer is a water-mistinglayer. Moreover, because there is no permafrost in some areas, anda hydraulic relation exists between the two layers. Analyzing thefigures of the contours of the water table and the iaoline of the ds~RC of mineralization, the main part of the two layers ofgroundwater ara drained away to Gulian River, another part to theground nurface from springs and forms icings in the winter.The distribution of ground water is homogeneous due to theUneven distribution of clay content. Some run off channels formwhere the day content is small and the water is rich. Most of thechannels have a linear distribution from south to north. So that dishibution of ground water has the character of linear enric%ent.SOLVING - METHODn.zI *- 1-N4Based on the hydrogeological conditionsand analysis of variousfactors, the reason that the icing forms in the foundation is thelinear enrichment of the ground water due tQ the unevonly distributed clay content. The amount of the run-off of the ground wateris large and confined. At the same time, the excavation of the foundationof the residential buildings changes the layer into a weakbelt. This provides the drainage condition for the ground waterthat directly flows to the basel pit. The superpermafrost water issupplicd continuously from the subpermafrost water because of thehydraulic connection between the two layers of groundwater.Afterthe temperature becomes negative, the icing gets larger and larger.In view of this situation, we proposed a treatment method oflowering the ground water table. In July of the next year, anunwatering hole was excavated which was about 15m apart fromthe icing. The beginning diameter of the hole was 420mm, the enddiameter was lROmm and its depth was 120m. From the explora.tion results, the distribution of the ground water along the depth ismainly pore phreatic water of the Quaternary unconsolidattd layerabove 3m, pun and crevasse water in the weathering laycr from 7rnto IOm and ground water in tho bedrock crevasse and tectonic crevasse from 70m to 90m. The ground water began to bc withdrawn12 houn each day on the 13th of August. The amount of wellingwater was 33 tons each hour. A recording of the unwatering is liat.ed in Table 1 and the coresponding figures are shown in Figure 1.The sketch figure of lowcring the ground water table is shown inFigure 2. Three months latcr, a linear withdrawing hopper formedthat was 300m from south to north and loOm from cast to west.D e decreased depth in the center was 2Sm.It was found during the withdrawing that phreatic water in thepres of the Quaternary uncon~~lidatcd layer was rich in.Septmbcr.and October. and the ground water in the poroa and crevasse of theweathering layer between 7m and 1Om became the main supply afterNovember. The large hopper formed around the residential; ; ; , *,Fig.2 The scheme diagram of the loweringground water tableTable 1 Record of withdrhinglastingtime(min.)groundw a t e rIevel(m)0.0 0.06.0 8.2620.0 14.2840.0 15.8060.0 16.90120.0 19.40180.0 21 .OS360.0 24.PO540.0 26.3 I1184

uildings was controllcd effectively by drainage of ground waterand by cutting the supply water of the icing, so that the icing didn'tform. At the same time, due to the decrease of the ground water inthe soil and the water content reduction from 17-24% to about12%, the frost heaving greatly decreased.EFFECTS AND - PROBLEMS"" ".Since the foundation of the residential building in Gulian rniningarea was treated by withdrawing, there has not occurred anywelling water and icing. Therefore, the engineering geological conditionsin thc area, located on the footslopc that easily causes frostheave, are greatly improved. The susceptibility of the frost heave ofthe foundations decreases clearly and the safety of the residentialbuildings is ensurcd. Bccause the method is a radical treatment, theeffect is stable.Higher expense is the disadvantage of the method. But aftermaking the well better, it becomes a tubular well to supply waterfor boilers and for drinking.CONCLUSIONS ...Through treating thc frost heave of foundations in DaHingganLing and other regions, it is considered that the frost heave causedby the rich groundand surface water can be solved ty artificiallylowering the ground water table. The extent of the lowered groundwater table should be in the state that the water content of soil inthe range of the capillarity is less or slightly higher than that of theinitial frost heaving. In general, the ground water table is 2rn lowerthan the frost depth. Through many contrast tests, it is known thatlowering the ground watcr table should begin a month before thesoil freezing, which is from the beginning to the middle of Scptcmbcrin Da Hinggan Ling region. Therefore, there is enough time todecrease the water content after the fround water table is lowered,the small water content is ensured when the soil freezes and thewasonally frozen soil freezes, which runs from the middle of Deamber. to January.The problem that is has a higher expense is solved by using thewithdrawn water €or boilers. Therefore, a well can be used formany aspects. It not only treats the frost heave of foundations, butalso keeps the water supply. The cost of the well will be decreasedand a good economic benefit is obtained.REFERENCESTong Changjiang and Guan Fennian (1985) Frost Heave of Soiland Prevention of the Frost Damage to Building. Published byHydroelectricity.Jin Junde (1988) IceLake and Protection Information on ForestryScience ane Technology (Translated Proceedings), Vol. 10.Yang Rongtian and Lin Fentong, 1986 Hydrogcology and EngineeringGeology in Permafrost Regions, Published byNortheast Forestry University.1185

THE GEOGRAPHIC SOUTHERN SOUNDARY 0F.PERMAFROST IN THE,NORTHEAST OF CHINA," 1Lu Euowei', W&dg Binlin' and Guo Dongxing2'Da Hinggan Ling Institute o f Fqrestation, Inner Mongolia'Lanzhou Institute of Glaciology and Geocryology,Chinese Academy of SciencesPermafrost has wide distribution on the Hulunbeier Plateau and the Songlengplain, and in the forest regions of the Da Hinggan Ling and Xiao Hinggan Lingranges, in the northeast of China. It's geographic southern boundary has bee-ndecided. based on the abundant data obtained. The.southern boundary distributionof permafrost is characterized by the integrated influences of geographicallatitudes and regional natural conditions. However, the southern permafrostboundary has been moving northward because of the serious destruction of naturalflora due to the frequent human activity during the last century. The onlyrelative stable northward-movement of the boundary occurs in the permafrostregions inside forest areas where the frozen ground is in a natural environmentwith most favourable forest restoration.ISTOBIC OUTLINE OF THE DIVISION OF THE SOUTHERN{ERMAFROST BOUNDARYScientists have tried to decide the southernboundary of permafrost in the Da Hinggan Lingand Xiao Hinggan Ling ranges since the 19thcentury. For instance, A.Q. Migandolf (1864),M.N. Shumkin (1940), and others have given theThe frozen ground in the northeast of China,has a total area of 38 to 39x10'km2. The formation,distribution and development are influenced. southern boundar.4es in the geographical map Ofby the geographic latitudes and regional naturalconditions, and is the southern extension of whichpermafrost zones in the northern hemisphere, ofsix boundaries were outlined through thepermafrost on the high latitude Euroasian continent.It can be divided into three types, withDa Hinggan Ling and Xiao Hinggan Ling rangesa range of latitude from 44' to 52'N (Fig.namely: large-sheet, island-melting, and islandl), which have important differences from thepermafrost, with latitude variati.ons and accord- real situation in these regions.ign to the continuity of its distributionAfter investigations of frozen ground in(Table 1). In the island permafrost zone, it ' these regions, Chinese geologists, e.g. Xing'QqVelops better in lower lands than on slopes Kuide and Ren Qijia(1956), indicated that theand it hardly exists on mountain tops. This southern boundary of permafrost in the twoshows Chat the regional natural conditions have ranges traces through Artika, Hudukeand Tat0a significant influence. There are very sparse Aili in Mongolia into Nan Hipgan, Buteha Qi,. distributions or no distribution of permafrostsouth Paligen, Dedu and Shajindi of Dulu Riverin the area of the southern boundary (Zhou Youwu in China, and then crosses the Aeilong Jianget el., 1981). River into the Union former (Fig.1). This SovietTable 1. Distribution Types and the Characteristics of the Frozen GroundType of petmafrost zona Division characteristics Continuity ( X ) Coefficient of interruptionLarge-sheet permafrost Large-sheet 65-75 continuous0.65-0.75Island-melting permafrost continuousDiscontinuous50-6540-500.50-0.650.40-0.50Island 20-40 ' 0.20-0.40Island permafrost zone Sparse island 5-20 I 0.05-0.20Very sparse island (5 (0- 051186

(l-0.3x104years B.P.), the permafrost retreatednorthward to Jinling and Gangyuan on the northpert of the Da Hinggan Ling range, and to thenorth side of Yilihuli Mountains. The permafrostat present was formed in the Neo-Ice Age, 3000years B.P.. which was a result pf the expansionof the remnant permafrost and the restorationof frozen ground in the melting areas, as wG11''as climate colling (Guo Uongxin, rpt al., 1981).These factors demonstrats that the southernpermafrost boundaries have advanced southwardand retreated northward several times due toclimate changes since the permafrost formationin the Late Pleistocene (Fig.2).Fig.1 The Historic Outline Diagram of the SouthernPermafrost Boundary in Da and Xiao HingganLingboundary c,orresponds to the annual average O°Cisotherm or to the January mean atmospherictemperature -24°C isotherm (Xin Dekui, et al.,1959). which is significant as a reference, aswell as showing the slight difference from theactual situation.In the areas of Jiayin, Detu, Arshan andXinbarhu Youqi. dozens of separate field surveyswere carried out during Sept.-Nov. 1973 andduring May-July 1974. *The studies includedmeteorology, hydrology, geology, geomorphology,soil, paleobiology, botany, geocryology, glaciology,and periglaciology. At the present timethe southern boundary of permafrost has beeninitially based on the large amount of informationobtained in the above mentioned expeditions.It has also been adjusted after two furtherexpeditions to the areas of the Da Hinggan Lingand Xiao Hinggan Ling ranges, Hulunbeir grassland,Charigbaishan Mountains and Hua'ng Ganglingterritory, etc. These expeditions were carriedout with practical research in mind, to studythe long boundary of 1300 km and complex variationsof natural conditions, as well as the needto understand the basic features of permafrostdistribution and fluctuation tendencies of thesouthern boundary in the last century.THE BASIC CHARACTERISTICS OF THE DISTRIBUTION OFTHE SOUTHERN PERMAFROST BOUNDARYSince the Late Quaternary, the permafrostdistribution has been expanding and shrinkingas a result of climatic changes in the Ice Agesand Interglacial period, For example, during theGuxiangtun cold period of the Late Pleistocene(7-lxlo'years B.P.-Wisconsin Ice Age), thepermafrost expanded southward in the Da HingganLing and Xiao Hinggan Ling ranges and has beenproven to be close to the average present annualaverage isotherm of 7-8°C along the boundarylinked with Aohan pi, Ganqika and Shuangliao,to the south 6f Xila Mulun River, and extendedeastwardly to Panshi, Huinan, Dunghua, etc,After that, in the warm period of the HolocenedFig.2 The Historic Evolutional Giagram of theSouthern Permafrost Boundary in the Northeastof China Since Late Quaternary Period1. The southern boundary of large-sheetcontinuous permafrost now:2. The southern boundary of island-meltingpermafrost now:3, The southern boundary of island permafrostnow:4. The high-mountainous permafrost now;5. The southern boundary of permafrost inthe Late Pleistocene:6. The southern boundary of permafrost,in the' holocene high-temperature period;7. Rhinoceros tichorhmus fossil site:8. Mammoth fossil site;9. Fossil ice wedge:10. Hemiaric grassland:11. County,The western section of the southern permafrostboundary in the northeast of China iscomposed of southern boundaries of island meltingpermafro t, in Arshan Mountains, and in theisland permaTrost on the east slope of the DaHingpan Ling range and on the Hulunbeier Plateau.It was shown, after the expeditions, that thepermafrost is sparsely distributed to the northside of Xingbarhu, Youqi, and Zuoqi regions,and no permafrost develops in the region betweenKelulun and Wurson Rivers, on the island perma-1187

frost of Hulanbeier Plateau. Therefore, thesouthern boundary of permafrost east of Mongoli'acrosses the upper course of Kelu River, andshould reach the boundary (Number 34) in theHulunbeier Plateau of China through Ba.Yangwulaand through Xinbarhu Youqi and Zuoqi along theHalaha River and come invertedly into the ArshanMountainous r'egion. But, because the relief ofthe Arshan Mountains (above 700-1200 m) ishigher than the surrounding areas, the climateis very cold, the frozen ground is more developedand shows the vertical regional change. Thelower limit of the frozen ground links with thesouthern boundary of frozen ground in the HulunbeierPlateau,, as well as being to the south ofArshan Mountains and the east ern slope of DaHinggan Ling. That is to say, it extends fromthe Halaha River, the Arshan Mountains, Wuchagou,through to Chaihe, Namu. and to the north ofthe Songlong Plateau.The middle section of the southern permafrostboundary in the northeast of China is composedof the sduthern boundaries of the northern Dermafrostin the Songlpng Plateau. The altitude islower (200-400 meters), and has a plain relief.The soil is fertile and there is a river runningthrough the region. The permafrost and thenatural flora have a closely interdependentrelationship. In the last decade the lumber andagricultural production ranges have expandedand have caused the southern boundary of perma-'frost to cro8s south of the Longjian River and *has developed to the south of Xiao Hinggan Ling.The boundary also,extends from the eastern slopeforest district into this district. The range 'isfrom the. north of Chaihe, Nanmu, Aronqi, throughLaochai, Detu, etc., then into the southernmountainous regions of Xiao Hinggan Ling.The eastern section of the southern permafrostboundary, in the northeast of China, is composedof the southern boundaries of the low mountainsand hill regions of Xiao Hinggan Ling through tothe southern permafrost boundary of Da HingganLing. It has shrunk northward by two latitudesmore than that of the southern section of DaHinggan Ling. But, because the relief of thesouthern section of Xiao Hinggan Ling is higher(above 403-700 m,eters), the southern permafrostboundary in the north of Songleng Plain extendsto the southeast. It also extends to the southernsection of Xiao Hinggan Ling, crosses the PingHeishan Mountains, Hei Longjian River and intothe former Soviet Union. That is to say, theregion is from the south of Detu. Qingan,Tianshan, Nancha, through the south of Jia YinRiver, to the opposite shore of the Hei LongjianRiver to the Buleya Mountains.Because of the factors stated above, thesouthern permafrost boundary in the northeastof China is about 1300 kilometers long, itehasa western distribution and is under the influenreof global climate, flora, relief, human activity.etc. Generally speaking, the western amplitudeof the southern boundary is the range of theannual average isotherm between 0 to '-l.O°C, themiddle section is in accordance. with the isothermO'C, and the eastern section crosses between theisotherm 0 to l.O°C.THE RECENT CHANGING TREND OF THE SOUTHERN,PERMAFROST BOUNDARYIn the'last century, the change of thesouthern permafrost boundary of Da Hinggan Lingand Xiao Hinggan Ling has had a close relationshipwith the frequent human activities and the- .increasingly warm climate. According to the-...statistics of Han Sen and others (1987), sincethe last century the global mean atmospherictemperature has risen 0.5'C -every year, from1880 to 1940. From 1940 to 1965 it decreased by0.2'C and fror 1965 to 1980 it rose by 0.3"C.From the observational data -it was found that1981, 1987, 1988 and 1990 were the four warmestyears, with a progressive rise in temperature(Li Peiji, 1991). This is in accordance with themeteorologic observation data obtained fromSheng Yang and Harbin since 1906. In the lastcentury, disregarding the decreasing fluctuationtrend in the atmospheric temperature in 1969,generally speaking there has been a warmingtrend (Zhu Kezheng, 1972b. Thus.. because of thewarming trend of thh global climate and thefrequent human activity, such ai. the expandedrange of agricultural reclamation, excessivegrazing.on grasslands, large scale lumbering,etc., there has been a'profound influence on thesoil, flora, climate, water conditions, etc..of the region. These factors have caused then0rthwar.d' trend of the southern permafrostboundary in the last century.The Hulanbeier Plateau is one of the semiaridregions in China. Because of the 8-9 monthsof dry wind from Mongolia the region is semiaridfor long periods. For examp$e, from 1958to 1962 the annual mean rainfall?.was 308.3 mm.in Xinbarhu Xiqi. From 1963 to 1980 it was220.5 mm. From 1980 to present, the yearly rainfallhas had a decreasing trend {in 1981 it wasOnly 141.5 mm). This seriously influences thegrowth of grass, particularly in Kerlun, Saihatala,Hanwula and in southwestern regions. Thearidity is very serious and because of theexcessive grazing the grassland regression andthe desertization phenomenon is increasing.Thus, in the permafrost of the-low ground and,in the moist zone, thb temperature has risen(0 to 0.5"C) and the permafrost thickness hasdecreased ( 5 to 10 m) and the permafrost hasgradually vanished. The southern permafrostboundary has shrunk from the semi-arid regionsto the semi-moist forests of the north and theeast. Therefore,, the long term aridity is oneof the main causes of the northward movement ofthe southern permafrost boundary in this region.The Songleng Plain and the Hulunbeier Plateauextends to the south of Da Hinggan Ling, the zarea is about 17x104km2. The region has atemperate and semi-moist continental monsoonclimate. The rainfall can be as high as 400-500mm. From 1841 until 1897, in this region, theQing Government had three policies, no reclaiming,no hunting and no lumbering, so the naturalforest environment was maintained. From 1897until 1945 there was a large influx of immigrantsfrom inner districts. the Bin*zhou Xailwaywas constructed, and destructive lumbering wascarried out in this region. These factors causedthe lumbering range between Qiqihar and Nongjiancounty to be expanded and the natural flora was,destroyed. As well, the natural covering actionon the permafrost layer vani5he.d causing theseasonal freezing and thawing aepth to increase.The very sparse island permafrost started toshrink and vanish. In a few decades the southernpermafrost boundary of the region retreatednorthward more than two latitudinal degrees.The southern permafrost boundary in thesouthern section of Da Hinggan Ling has beenrelatively more stabilized in recent times, moreso than that of the Hulunbeier Plateau and theSongleng Plain. The main reason for this is the

controlling and influencing factors of thealtitude height. Thus, to some extent, thedistribution of frozen .ground abides by the lawof vertical zone properties. For instance, thelowest altitude height of the permafrost distributionis about 800 m in the Artaishan Mountainsof Da Hinggan Ling, and there is also biennialfrozen ground on the slope bottom with a thickclayey soil layer in Wuchagou and Bailan. Withan increase in the altitude, the climate of theregion gets colder and the frozen groundevelopsbetter. the frozen ground area is 40 to 50% ofthe total area of the region. The mean annual.ground temperature of the frozen ground is -0.5to -1.O'C. The thickness of the frozen groundia more than 20 to 30 m. Xinan Larix gmelini isthe main wood seed of the forest flora, whichis advantageous because the natural regenerationis above 90X (Lu Guowei, 1989). We can assertthat, under the construction policies offorestry the forest area will be graduallyincreased in China. To have the optimum naturalenvironment, the rate of the northward movementof permafrost must decrease and the southernpermafrost boundary must be stabilized.REFERENCESGuo Dongxin, et a1.;(1981) The PermafrostHistoric Evolution and Formation Period in.the Northeast of China Since Late Pleistocene,, Glacioloiy and Geocryology, Vo1.3, No.4.Li Peiji. (1991) Greenhouse Effect and ClimateChange, Glaciology and Geocryology, Vo1.13,N0.3.Lu Guowei, (1989), Xinan Pine Vegetation andVariation of the Ecalogical Environment ofPermafrost in Northeast China, The Proceedingsof.the Third Chinese <strong>Conference</strong> on Permafrost,The Press of. Sciences.Xin Dekul et 81.. (1959) The Geological KnowledgeAbout the Permafrost Distribution in theNortheast of China, No.10.Zhou Youwu, et al., (1980) Permafrost and IceEdge Geomorphology, the Natural Geographyof China (Geomorphology) The Press ofSciences.Zhu Kezheng, (1972) The Preliminary Study on theRecent Five Thousand Year Chinese ClimaticFluctuation, Zhu Kezheng Monograph, ThePress of Sciences.

ROAD DESIGN AND RENOVATIONS OF THE NORTH SLOPEIN DA HINGGAN LING!,uoWeiquanDesign House of Management Forest Bureau,Eer Guna Lift, InnerMongolia Autonomous pegion, ChinaResearch and practice have been carried out for many years to determine a methodof selecting the road line, renovation against frost heave and thaw settlement,determination of embankment height, foundation depth for bridge culverts,foundation type, and preventing pingoes and icings. These methods have beenused in recent constructions of roads in cities and towns, espe'cially in thedesign and construction of concrete pavement in sections with massive ice,measures are taken to choose a reasonable embankment height, dredging, pavingsoil material, protecting the'natural environment and berm, enhancing thestrength of the pavement layer and base course, and to ensure that the 12 mwidth of concrete road is safe for transportation in cities and towns.INTRODUCTIONMengui Forest Bureau is on the north slope ofDa Hinggan Ling, the climate is frigid and theyearly- average temperature is below -5'C. Theextreme temperature is -54'C and th.e frost seasonis as long as 8 months of the year, Permafrosthas an extensive distribution and is from severalmeters to around 100 m, but the large riverbedis talik. The Quaternary covering layer iscomposed of stones with loam gravel, cobble soiland peat loam,stratum. The surface vegetationis mainly a heavy growth of eriophorum tussock,and swamp ground is distributed in the basin andwide valley of the region. The largest depth ofseasonal thawing is from 0.5 m (peat section) to3.2 m (cobble and gravel, stone section). Thepermafrost includes a large amaunt of ground ice.More than 350 km of road and 200 bridge :ulvertswere constructed from 1960 to 1983. and mosthave some extent of frost damage. The damageis mainly shown in the non-uniform subsi.denceof the roadbed, wrenching and inclination ofbridges and culverts, pingo and ice erosion ofthe road and the formation of potholes in thespring. This causes the ecological equilibriumto be destroyed as the buildings do not adaptto the natural environment of frozen soil, aswell as the causes of unreasonable design andconstruction. In general, in several months oryears, deformation or destruction of the buildingaoccur. Roads change their original lineand culverts have to be renovated. The lightdamage needs to be repaired and prevented andthe cost of repairing 180 km of road is motethan two million Reminbi Yuan a year.ROAD DESIGN AND RENOVATION OF FROST DAMAGERoads and bridge culverts in the north slopeof Da Hinggsn Ling suffer damage due to thelarge amount of ground ice in the permafrost.plentiful surface water and groundwater,suitable freezing conditions, and loam sto'neswith stronger frost heave sensitivity, etc. Roaddesigns and methods for renovating the frost 'damage should take into account the abovementioned factors, For the north slope regionof Da Hinggan Ling, the renovations of frostdamageshould stress the aspects of drainage ,water, heat preservation and the filling soilused in the past. Combined with choosing theline, design and renovation can often be moreeffective.Choosinn the LineThe north slope of na Hinggan Ling is mainlya hilly land of middle-low mo,;ntains; The roadoften crosses the basin,and valley. Whenchoosing the line in the permafrost region withthese natural conditions, the basic designprinciples are: sufficient filling and littleexcavation, selecting a high site and avoidinglow sites. The-line in the,hills should have ahigher site, and cross from the top of the lowpitch in front. of the mountain, and a sunexposed site' is preferable. When built along .the river valley, a high terrace and river talikshould be chosen. Based on selecting the lineand investigation, poor geological sectionsshould be avoided when possible. The lineshould cross the embankment in the upper frontof the massive ice section or in the lower partof the thaw slumping body. The possibility ofchanging hydrogeological conditions in the roadshould be prevented in the regions where groundwater exists, surface water should besdredgedand the roadbed peripher-y should be kept dry.Design and construction should protect thenatural environment of permafrost and theecological equilibrium. Designs of bridgeculverts, except for the flood table area withthe largest flow amount, jammed height of pingoand icing and foundation handling should beconsidered. In recent years, the above mentfonedprinciples have been used when choosing the

line and design and the frost damagegreatly decreased.of roads hasRoadbed EngineeringFillinn heiRht of embankment: A larRe amountof investigations and testing have shoin thatthe minimum filling height of the embankment inthe region of the north slope of Da Hinggan Lingis determined by using the following experienceequation:H = h, t hc (1)where H - embankment height (m)hi- dry soil thickness of embankment (m)in general hl=0.6 mhc- dangerous capillary height of roadbedso.il. (m)- ,The relative soil types and degrees af compactionare given in Table 1.THE MEASURES OF WATER DRAINAGE AND STOPPAGEA water stoppage ditch, water drainage ditchand retaining waLl are often used to drainsurface water in road engineering. A blankditch and water penetration ditch are used todrain ground water or decrease the ground watertable. The drainage ditch should be kept adefinite diatance from the slopefoot of the roadbed.The di'stance can not be less than 1.5-2.0min general conditions and the drainage ditchmust be consolidated to prevent penetration andto have enough vertical gradient. A taperedgutter to preserve heat should be set up toavoid pingo and icing from formilg. When theground water table is higher an insulating layershould be set up to insulate the ground water(see Fig.1)Table 1. Dangerous capillary heightof types of roadbed soil (m)Soil typeCompacted soj.1approaching optimunwater contentSoil of airSoil of~ ~ non-compacted~ ~ ~ c ~SandSandy loamSi 1 t,LoamClay0.100.200.500.400.400.200.300.20 - 0.600.30 - 0.601.0 - 1.20 0.80 - 1.500.80 1.50 - 2.000.80 1.50 - 2.00W-e,tiv1c,O I layer of seperutinglayer of penetrating waterI original ground surfaceFig.1Insulating layer structure of roadbedROADBED WITH FROST HEAVE AND THAW SETT1,EMENTThe handling of roadbeds with frost heave inthe north slope of Da Ilinggan Ling often usesthe method of changing the soil, gravcl soil,cobble gravel and medium coarse sand are oftenused instqad of frost heave soil, the replacementsoil should be less than 809: of the maximumdepth of freezing. The road after the soil hasbeen changed must meet the demands of strengthand water insulation. Thc heat preservationmethod is chiefly used in the sections withmassive ice, eriophorum tussock at the locationcan be used HS the heat yreservat.ion material.One layer is paved in the original surface andberm with heat preservation is constructed (Fig.2).Except for increasing the heigh~ of the roa&bed (minimum value is 1.2 m), the measure ofdraining insulating water is used in the concreteroad i n the town of Yengui. Three layers ofsoil are used as a cushion in the bottom of thepavement, and undcrneath, 15 cm of soil materialis added. This method not only enhances theaction hut also quickly diffuses water so thatthaw settlement. cannot be produced in thesections of massive ice (Fig.3). It is alsoconsidered impurtant in protecting the naturalsurface beside roadbed. The ground surfacewithin thc range of 10-15 m of the ruadhed cannot have its general condition destroyed. Zheng 'Yang road was constructed with this method andbasically no changes have occurred from 1987 topresent,PINGO AND ICTNG PREVENTIONI'xcept. for the naturally forming pingoes andicings, pingoes icings and some small risingpingoes are caused by the changing movement lawof ground water after the road consti-uction clrrthe nor~herrr slope zone uf lla flinggan Ling.The drainage ditch and freezj.ng ditch aregenerally installed above the embankment toblock and drain ground water, meanwhile enoughfield accumulated ice remained for the emhankmcntor blank ditch for penetrating water, torlec.rease the pressure and hea~ preservation, re1191

}xi/berm of heatembankment of loam\ preservationinverse putting tatou tussockFig.2 The sectionof heat preservation embankment300 1200 1\50Fig.3 Sketch mapof cross section on concrete roadbed in YenguiembankmentFig.4 A sketch map of a pingoinstalled in the upper part of the embankmentto divert ground water to the lower part of theembankment, Engineering renovations at manysites show the effect of the blank ditch forpenetrating water, decreasing pressure and heatpreservation is very good, Success or failureof this measure depends on the heat preservationand drainage conditions.BridRe Culvert EngineerinqMost of the bridge culverts were provisionalwood structures and-have been changed toconcrete structures to meet road standards.Frost damage of the bridge culvert is mainlyshown to be an upfreezing of objects causingsubsidence'and wrenching inclination. A deepfuundation is used for the pit foundation. Formeeting the design demand for the bridge pileof roads its embedded depth H is determined bythe following equation:H - 2nrhoT-PZtrrfThe expanding column foundation of reinforcedconcrete exploded short pile and pilot pile areused in shallow foundations. Practice shows ''that the effect of preventing frost heave withpilot pile foundation is better. One layer ofheat preservation needs to be paved, as well asthe sand cushion undet the foundation, in orderto prevent hear transfer of the foundation andthe prevention of frost heave of the foundationis enhanced, as with soil changing the chemicaladdition is used to enhance the hydrophobicityof the subsoil.CONCLUSIONTo summarize that mentioned above, the prevcntativesteps against froat damage and comprehensiverenovation of damages are used with thelocation conditions for a better effectiveness.1. The main methods of protecting the roadbedstability are that the environment in the permafrostregion should be protekted, and theembankment height should meet the criticalheight.2, Effective drainage, the better form ofdrainage ditch is with a shallow width andmultiple dams.3. Choosing coarse soil for constructing theroad and using soil material to quicken thewater drainage and enhance the roadbed.4. Reasonably choosing the line and constructionmethod ensures roadbed stahility in permafrostregions.where r pile radius (cm)h - maximum thaw depth (cm)+ OT tangential frost heaving force (kPa)'fnormal fr.ost heaving forces (kea)P - surcharge load (kN)1192

PRACTICE OF REINFOHCED CONCRETE STRIP FOUNDATIONIN PERMAFROST REGIONSMen ZhaoheAmuer Design Housc Forestry Rureau, Management Bureau ofDa Hinggan Ling, Heilongjiang, ChinaIn A~nuer permafrost region, various of building foundation, for example, rubblefoundation, rubble-sand pad foundation, concrete foundation with sand pad, piletoundation, and reinforced concrete strip foundation have been used. It is provedthat the former three types uf foundations are not approp'riatc to be built inpermafrost region. The designing principal of reinfore c.oncrete strip foundationpermils t.he foundat.ion ground t.o thaw, that is the irost state of foundationground is unconsidcrable. 'l'he mechanical properties, such as, bend, shear,Lensiorl and deformation, can be complete1 y contributed. According to variousfrost henving soil and the depth of maximum thawing plate, the depth of sand padand the section size of base (mainly the height of base) are designed andcalculated. Meanwhile, hy adopting relevant: technical measurements,'suEficientrirlidity and strcngth of the structure of base and upper soil are assured, so asto resist various stress and uneven settlement of frozen ground foundation duringfrost heaving and thawing.based on thc engineering information, correspondingmeasures are taken such as reinforcement andAmuer Forestry Rureau is located in the north t.rying to restrengthen and restabilize theof na Hinggrrn Ling, about 52"15'-53"34'N,destroyed frozen soil subgrade. If the subgrade122"39'-124"14'E. The total area of the regionhas normal or strong thaw settlement, fissuresis 5576.06 km'. It belungs tu a hilly regionfor settlement and girds are set up to reinforceand the altitude is 80011000 m. Permafrost hasthe integral rigidity of the building in thean extensive distribution and the largest thick- sand cushion in designed thickness under theness of permafrost is about 50-80 m. The yearly bottom of foundation. The calculated thicknessuvcrage temperature is -3.8 - -6.3"C, the freez- of the sand cushion is based on the followinging index is 3810 day OC, The extreme lowestformula:temperature is -49.7"C, the highest temperatureis 35OC. The average ground temperature is -1.0'12- -4.2"C. 'Thc winter is as long as 7.months andthe summer is about 5 months. 'The yearly preci- where - additional stress of sand cushionpitation is 397.70 mm. The maximum depth ofbottom (kg/cm')accumulating snow is 36 cm. The seasonally.. I'C - self-weigRt stress of sand cushionthawed depth is 1.0 m in the permaErost regionbottom (ky/cma)with vegetation,IR1 - bearing allowance of soft layer atcushion bottom (kg/cm')REINFORCED CONCRETE STRIP FOUNDATIONAccording to the abservation of base in thepast, the causes of fissures and deformationare due to three aspects.1) The foundation is not stable, and thedeformation of freezing and thawing is no.tuniform.2) The rigidity of the base is not strongenough to prevent various styesses'from deformationin the frozen soil foundation, such asbending stress, tensile stress, shear stress,etc.3) The foundation type and embedded depth arenot reasonable.The t'hree aspects are analyzed as fo.llows:1. DesiRn Principle of Frozen Soil Subnra"deThe subgrade is designed without consideringthe depth of frozen soil under the foundationand the effects of the freezing state of thesubsoil and the effects of gradual thaw. Then2. Desidn of Foundation RigidityTn order to increase and reinforce the rigidityof the foundation, the following steps aretaken:1) Material with the best mechanical propertiesis chosen for the foundation to reinforcethe concrete.2) The strip foundation is chosen because thepressure of the subsoil is small and uniform andthe thickness of the bearing stratum can bedecreased for this fouhdation type.3) The foundation sectional shape chosen isthe inverse "T" shape or the "I" shape sectionwhich have a large bending rigidity.4) Checking the calculation of wall ridigity.The height of the concrete foundation (h) ischanged into the height of the brick bondingbody H., and is the following:H, = 0.9h'- (2)1193

After the converted height is Calculated, theintegral rigidity of the wall is accounted foraccording to L/H

thus, Hp15.51 m, Hs=11.91 m (Fig.2). LF=22 m,Ls=27 m, then:L~/H~=1.42

thus: As-z'-28.4-20.4=7.6 cmP=AslLs=7.h/2703

A MICROSTRUCTURE DAMAGE THEORY OF CREEP rN FROZEN SOILMia0 Tiandc'p2, Wei Xuexia ', Zhang Changqing''State Key Laboratory of Frozen Soil Engineering, LIGG,CAS,China'Lanzhou UniversitvIn the paper damage mechanics was used into mechanics of frozen soil to study the infer relation bctweencrbcp deformation and it change of microstructure. Based on a series of creep test of frozen soil,and also observated its change of microstructure by "Duplication-electronmicroscopd' method. Weconsider that the sticky force of ice rcmaincd in frozen soil, the dcgrcc of directional of mineral grain andthe area damage (include the produstion of microcrackle and its later extension control) the whole creepprocess. So analysis in theory, we start at the basic equations of continuum mechanics, introduce the icecontent in frozen soil. The factor of damage area and the factor of arca .occupied by directional mineralgrains as the three internal variables which characterizc the change of microstnrcturc of frozen soil.NI"I'ODUCTI0~Crccp is one of the most important mechanical behavior offrozen soil. Sincc 1930, when Tsytovich (1930) published the firstpaper on the subject, many investigations have been reported. Sewera1 theories to explain creep in frozcn soil were forwarded,most ofthese studies were based on the macroscopic tests. However, because the complexity of the microstruction changes, and sensitity toternpreture, these very facts make it very difficult to study the creepmechanics at microscopic level.In 1967, S. S. Vyalov(ACPEL,1952) reported his first observationof the microstructure changes of frozen soil in creep process.He used "area" and "orientation' factors -I- to characterize themicrochanges. This might be the initial considerration ofmicrostructure damage in studying behavior of frozen soil. Sincethen, no significant progress was made in the area. there isn't anyrelated published work in china.In past few decades darnage mechanics as one of the disciplinesof solid mechanics was been firmly founded, and has beensuccessfully applied in many scientific fields. In light of the damagemechanics, we have studied the relationship bctween crecpdeformation and the microstructure changes of frozcn soil. Limitedby the paper length, only part of theoretical analysis is given in thispaper. Experements results will be reported in separate paper.Through a series of creep tests and microstructure obvervationby so called "duplication-eletromicroscope" technique, we foundthat the cohesive force created by ice, the orientalization of soilgrains, and the creatation with irs late extension of microcrack controlthe whole process of creep in frozen soil. Starting from thesebasic points, we dcvcloped a 3-dirncnsional damage theory for thecreep deformation in reference to the principles of continuum mechanics and the frame of damage mechanics.A detailed analysis has been made for the case of uniaxialcompression, a prediction cauation for the long-term strength hasbeen dcrived. The theoretical curves correlate with the corrc.sponding test ones.~-BASIC " EQUATIONS OF CONTINUUM .DAMAGE MECHANICSwhen we treat frozen soil as a continuum medium, the followingequilibrium and thermodynamic restrictions should be satisfied:9 + (PV,), = 0 Conservation of mass (2-1)atuk,J + P(f, - G t) = 0 Balance of momenta (2-2)uk, = nk Balence of moment of momentum (2-3)Pi - gktdlk - qhk == 0 Conservation of energy (2-4)P4 - (qk @,a 0 Principle of entropy (2-5)wherep-- the mass density;vk--- the component of displacement velocity, k= 1.2,3;ukI--- the component of stress tensor; .:'fk---- the cornponcnt of the body rorcc per unit mass;dk,-"-- the component of strain velocity tensor;q,---- the component of the heat vector per unit area;e--- the intcrnal energy'density per unit mass;1"- the entropy density;8" the absolute temperature.dIn addition, ()=-( ) indicates the differentiation with 10-dispect to time t.Introducing the free-energy function1197 9

10)INTERNAL VARIABLES DESCRIBING THE",l.__l_.-_"IMICROSTRUCTURE DAMAGEand thinking rp is the function of elastic strainei, , temperature0 and internal damage variables uicp = rp(e;17@yi) (2- 7)Substituting (2-6), (2-7) into (2-5), and using (2-4), we obtainBy the method of "duplication-eletronmicroscope", themicrostructure of frozen soil has been observed at different crcepstages. Examining result of our observation and refering toVyalov's work, we chose three internal variables which are mostimportant to characterize the microstructure change in the CrCCpprocess.1) The cohesive force of ice in frozcn soil plays a key role in thedclormntion of frozen soil. So we chboseHere we uscwhere e: is the plastic strain.Because rp, 1 , u,,and q,arc independent ofinequality (2-8) can't be maintained unless:b >i;k and i,,, theto be one of the internal damage variables.Where:p---- thc mass density of frozen soil;P(i)" the mass of ice in unit volume of frozen soil.Following Vyalov's work, we introduce(3 - 2)the factor of damaged area.(2 ~(3 - 3)andAssuming(2-1 I) becomes(2- 11)(2 - 12)oklk: + Aji 20 (2 - 13)One of the methods to handle (2-1 3) is to choose a functionaldissipativcpotcntial,F=W,,,A,) (2 - 14)The hypersurface expressed by (2-14) should be convex, thisleads tothe factor of oriented area of mineral grainas the other two intcrnal damage variables. where w definesthe area damaged by the microcack and its extension, while n represents thc area occupied by the oriented grains.A,---- the effective area on which force acts;A,----- the area which mineral grain are not orientalized;A,---- the initinal area.Let a, = t"', = a, a,= w, then a,(i= 1,2,3) are theinternaldamage variables, and Ai(i = I ,2,3) which mentioned before, are thecorresponding thermodynamic force of ai.A YIELD CRITERION FOR SOIL / ROCK"" ~~ ~A yield criterion suitable for sail / rock is suggested. The criterioncan be expressed by genernal stress components with advantageof that it has asingle expression, and it coincides withCoulomb yield criterion at thespecial stress condition---purecompression and pure stretch. It's yieldfunction G is expressed asG==ii+hJ~/u3+uJ,-H (4 - 1)(2 - 15) wherewhere I is a scolar factor.(2-15). gives the evolution equations of the plastic strain.8:and the internal damagc variables c1;:119%

a is a material constant which characterize the compressionbehavior.PARTICULAR EXPRESSION OF THE DAMAGE EVOETION EQUATIONSThe function of dissipative potential F is usually composed oftwo parts: one part P,(G) is to express the effect of viscoplasticity,while the other part F,(x) relates to effect of internal stress x, x isthe strength of x .Bacause the viscous flow occours before the plasticdeformation, thusF - F, (G(u- X))H(G(u - x)) + q3 (5- 1)where H( ) is Heariside unit function.F, and F, adaptable extensively to creep deformation ispowerfunction:(5 -4(5 - 3)where N, M, n, m are all constants depending on temperature.There are some additional assumptions:1) The cohesion C of frozen soil depends on ["and w (basedon test), we simply assume,(5 - 4)Fig.1 Forms of yieid criterions on c planeI. Coulomb yield criterion2. The suggested yieldcriterionwhere j? is a material constant.substituting (5")-(5-7) into (2-15), we arrived at a set of3-D damage evolution equations:where (j3)k,indicates the algebraic complement of J'3.(indicates that cot plus with k, 1, and ( .,shows cr is all displacedby u-x in every expression.whereC,---- the initial value of C;#--- the original icc content;b a material constant dependent on temperature.2 ) The orientation factor fl leads frozen soil to be anisotropic,which relates to vector x of the center of yield hypersurfacex = x(n), also causes change of the friction angle 4, thusdl = dm = (5 - 5)3) The effect of the thermodynamic pressure x on F can becontained in the yield function G by the following equation:where(5 -I(5 - 9)10)4) to simplify the calculations, we suggest particular functionsof thethermodynamic forces A, with ai by(5 - 7)(5 - 13)1199

(5 - 14)As the creep developes, x is absolutely relaxed, so the effect of*Fl(x) is neglected in (5-13).THBCASE OF UNIAXIAL COMPRESSIONIf uj, =-n and all other uk1= 0, then (5-S)-(S-14)to one dimensional for whichdegenerate1E.' = - A(1 - u){- [(I - U)(U - X) - - Te]\' (6- I)NaH am+ aw)G(6 - 3)Fig.2 The variation of e(t) and o(t) in'the whole creep processThe expressi of 7ah aH ah H,T,- ,nnd-are given bya{ a< aw am(+lO),and R({"),o) will be determined in lateranalysis.I ;p/;I= - T ",(2" io)

Thus we consider Q is the long-tenmstrength of frozen soil,while t is the shortest time when the rate of plastic strain homeszero. For a certain material and t * are all constants.c) phase 3 (tz< t < t3)-e' =C,r -I- C6,(r2= 4, = Const.< r < r 3)' 2 2(I- w)2 - (I- = - ~ c,T~ (I - rl)(6 - 16)(6 - 17)(6 - 18)where w2= w(tl).d)phase 4 (t t,)-eC=C,t +C, (6>t,) (6719)I$ = 4, = Const. (6- 20)' 4 4(l-~)'-(l-~~)'=-2C,Tc,(t-t,) (6- 21)where a, = w(t,).Considering smoothness of creep curves, the continue conditionsat each ti(i = 1,2,3) must be satisfied.c'(t+)=eP(t,-) (6- 22)in(t;)=i#(t,-) (6- 23)The variation of w(t) in the whole creep process is shown asFig.2.Matcrial Constants &(i = 1,2,3)-wi = a(tJ (i = 1,2,3) corresponding to the time when one phasetransit to another is determined by temperature, but not related tothe level of load, therefore, ai = w(ti) are material constants. Substitutingit into (6-12), (6-14) and (6-20), we getCl(r: -2r rl)=A, (6 - 24). C3(t2 - t,)=A, (6 - 25)A js also material ,constants only depending on temperature.One can see thatA,is equal to the increment of plastic deformationat every corresponding phase.The Equation of Long-Term StrengthBy substituting (6-14) to (6-25), and letting $=b, thelong-termstrength equation is obtained asTime (h.)Fig.3 Comparison of calculated strain with test dataREFERENCESTsytovich,N.A.(1930)?Petmafrost as a bast structures. matcrial ofthe permanent Commiossion for the study of the Nature ProductiveForce of the USSRW,(1980):USSR Academy of SciencesPress.ACFEL(I952):'lnvestigation of description classification andstreigth properties of frozen soils".Vol. 1 and 2.WSA arctiaConstruction and Frost Effects bboratpry (ACPEE)TwhnicalReport 40.AD721745 and AD721746Vialov,S.S.(1978),Rheological principles of soil wchanics, HighmEducation Publisher.Wu Hongyao(l990),Damage Mechanics.Tstovich N.A.(1973), The Mechanics of Frozen Soil, Higher EducationPublisher.(6-27) 'If uo = 0, t, = 0, 1 / n = a, N;Al = T , then (6-27) degeneratesto the long-term strength equation given by S. S. Vyalov-.(6 - 28)which have been extensively used in cdd engineerin'g.Predicting The Creep Curves and Comparing to Test DataMaterial constants Aiand t can be fixed from a nondeclinecreep curve done by test. N,, n and have to be determined by threetest curves with different load level each. After that we can predictthe crecp behavior at any load level.Fig.3 shows a comparison of the predicted curves with the testones at temperature of -5C. A good agrement exists bctbrecn thetwo types of curves.

Pan AndrngDepartment of Cicography,Lanzhou University,l,anlho~J 730000,C'hinaThis paper advances the arguments on the correspondence of cold-wet or warn1-Jry periods under rhespecial geographical environment 01 Xiniinnp Gnce the late Quatcrnnry. Evtdcncc is cited and approachedfrom varlous aspects, the paper pmnts out that the monsoon was thc main f'actor in formingthe combination of warn-wet and cold-dry periods in eastern China. With thc typical continental climate,Xinjang had a contrary combination of temperature and humidity since"the late Quaternary. Theauthor does not deny the possibility that some short periods of warm-humid or cold-dry climate existedin Xinjiang since the late Pleistocene. But the correspondence ol'cold-we: or warm-dry should bethe main features of climatic type in Xinjiang, the special arid region.FTRODUCTIQN .".Under thc domains of monsoon climate, the combination ofdry with cold or warm with wet climate is common knowledge inenstern China,in the past and present.But it is not so in Xinjiang.The cold glacial epochs were not extremely dry and the warminterglacial epochs werc not wet since the late Pleistocene.On theother hand, in line with the sandards of the aridity index in the naturaldivisions of China, there almost are no humid areas inXinjang. Picea schrenkiana forest belt is located in the greatestprecipitation belt of Tianshan Mountain and represents the mosthumid, environment of Xinjiang with a yeurly precipitation ofabout 400-800 mm. While this tree species should be included inxerophilous undoubtedly by the standards in eastern China, andoften is used as the representative of a dry-cold environment of aglacial epoch in paleoclimatic research. For this reason, humid,thcword used in this paper, is just a comparative concept under thespecial situation of an arid area, which is different from the humidclimate of eastern China. But this relatively humidity benefits theconservation of soil moisture and the decrement ofevapo-transpiration, and causes the great difference of ecologicalenvironment in the extremely dry region." CHARACTERISTICS-. OF -...I SPORO-POLLENASSEMBLAGESThere were a series of stronger temperature dropping epochsand remarkable high temperature epochs since the late Quaternaryin northern Xinjiang,bascd on the analysis and computation ofsporo-pollen identification results and 14C dating.The characteristicsof sporo-pollen asscmblages in the cold periods were similarwith that of the surface samples from various grasslands. In whichthere were variety of mesophilous herbs such as Gramineae,Labiatae, Alhagi, Ranunculaceae, etc. Still more pollen ofhygrophytcs or hydrophytes Typha, Acorus, Sparganiaceae,Cyperaceae, ctc. could be found and the pollen of Salix, Ulmus,and Rosaceae appeared occasionally in some samples. It reflectedthc climatic environment with low temperature and humidity,which was favorable to form Grasslands and Marshlands, at thesame time, the altitude of forest belt of picea was lower so that theconifer pollen often added to the deposits at the foot of the mountains.The climate conditions then was more humid and colder thanpresent. In the hot period, the content of xerophte pollen asChenopodixceae I11 and IV, Ephedra, Artemisia, etc. rose abruptly,which often reached to the highest value in each of the cores or seetions, taking the ZK-024 drilling in Balikun Lake as ancxample.The sporo-pollen assemblages were divided by visual andcomputer inspection into six zones (Fig I). Zone I(89.05-65.09 m)is the zone of Chenopodiaceac-Cuperassaceae-Gramintae, thatreflected the cold-wet climate of grasslands or grassy marshlands.And from 74.63 to 79.01 is a special cold-wet stage with a highersporo-pollen influx. Zone I1 (65.09-23.49 m) is the zone ofchenopodiaceae-Artemisia-Alhagi-Ephedra. The vegetation belongsto deserts-grassland, the climate was becoming warm anddry. This zone can be further divided to two warm-dry and acold-wet stages. Thje third zone (23.49-10.76 m) is that ofChenopodiaceae-Rosaceae-Gramineae. The ancient environmentrecovered to grassland in the cold-wet climate, the sporo-polleninflux increased to the highest point. The forth zone (10.76-7.84m), Chenopodiaceae-Leguminosae-Acorns Zone, reflected the flo+ra of shrub-grasslands in the cold-wet climate. Zone V (7.84-2.96rn) is Chenopodiaceae-Artemisia-Ephedra zone,the assemblagesare similar to the modern desert in a warm-dry climate, and thesporo-pollen influx is low here. Zone VI (2.96-0.75 m) isChcnopodiaceae-Artemisia-Gramineae Zone, which can be furtherdivided into two parts. The lowot part has some pollen ofshrubs and trees, Such as Picea, Betula, Cupressaceae etc., and rep

pnts the slightly cold and wet climate. The higher pollen percentageof Ephedra, Artemisia and Gramineae in the upper part reflectedthe desert environment of warn-dry climate. The wholesection shows four cold-wet and four warm-dry climatic epochs.Each cold-wet stage shown by sporo-pollen assemblages and influxcorresponds to the fresh or slight salutes water layer providedby mussel-shrimp analysis, and the watm-dry stage to the saltwater layer. The paleo-ecologic environment evolution reflectedthe paleo-climate changes. When the arid region cntcrcd acold-wet stage, the lower air temperature dropped the evaporationpower so that the relative humidity rose, and the vegetation didwell. The sporo-pollen assemblage represented flora in a great variety,thesporo-pollen influx was raised to a hightr point, and thepercentage of xerophyte and salt herb decreased in the assemblagesat those times. When the arid region returned to $e warm-dry orhot dry clirnate,the plant's growing was blockcd,the sporo-pollenassemblages displayed the dry features and the influx in lower valuesbecause of the higher temperature, the stronger evaporatingpower and &he less effective precipitation. Then the percentages ofxerophyte and salt herb increased. The characteristics! ofsporo-pollen assemblages corresponded to that of surface samplesfrom the general or driest deserts.PARTICULARS OF THE PALEOCLIMATE IN XINJIANGThe correlation of warm-dry and cold wet climate seems contraryto the traditional ideas on cold-dry stages in glacial epochsand warm-wet stages in interglacial epochs. In this connections theauthor makes the following approaches:I) The research on the climate of Xinjang had showed clearly,that the relative humidity is smaller even though there is more water-vaporcontent in the air, if the air temperature is higher (ChenHanyao,1963). Table 1 shows the contradictory situation inWrumqi as an example. The water-Vapor content of the airUrumqi in July is 5.6 times and the water-vapor pressure is 7.4times as much as those in January. The relative humidity in July ,is40 percent lower than that in January, for the mean monthly temperaturein july is 39°C higher than that in January. The statisticaldata on the monthly relative humidity reported (Chen Hanyao,1963), that the relative humidities were higher in winter than insummer in all of the meteorological observation stations ofXinjiang. The rower relative humidity in high-temperature periodsincreases the evaporation power and transpiration, hastens the vanishingof surface water and soil moisture content, and restricts thegrowing of mesophyie. 'Tablc 2 shows that the increasing water-vapor pressure causeda variation of the precipitation, which rises by a greater proportionin eastern China. In Xinjang, however, the water-vapor pressurqin Urumqi is greater by 640 percent in July than in January, but theprecipitation increased only 150 percent in the same period; thcwater-vapor pressure in Turpan, where it is hotter and dryer, isgreater by 710 percent in July than in January;nevertheless, the prbcipitation, is increased only by 50 percent. The facts illustrated theincrement of water-vapor to a great extent in hot periods. But thestrong evaporation power at the same time,however, counteracted,,the effect of water-vapor greatly, and made the little difference inprecipitation between summer and winter become more slight.Thus it can be seen that the dry environment is not followcd bycold but by a hot stage at the present time.2) We can see the similar situation from the research on thedimate periodicity in Xinjang (Zhang Jacheng,l974) that around1970 could be taken for a transition in periodic variation. Table 3shows the extreme minimum temperature that took place in thecold period before 1970, and the extreme maximum temperature inthe warm period after 1970, which reflected the feature of theperiodicity. In the warm period, the precipitation decrrased by7.4-16.4 percent, evaporation increased by 1.649.9 percent. Theevaporation-precipitation ratio shows the air became dryer in thewarm period.3) The archaeological evidence. The 1,000-year cold periodFrom the end of the Yin Dynasty to the beginning of the ZhouDynasty (about 1,000 B.C) had been divided on the base of thephonology and historical records (Zhu Kezhen,1973). and was providedand received by many scholars. Xinjiang was also in a coldenvironment in 3,000-2,000 B.P. Li Guangli, a general of HanDynasty,attacked the Xiongnu pun) at Tianshan Mountain withthirty thousand cavalry in the autumu of 99 B.C. General Li orderedthe soldiers to take big pieces of ice as a beverage in the assault.It is thus shown that the weather was cold. Now the lakewater can not ice up to such a thickness, even on Tianshan Mountaina height to 1,000-2,000 meters in the autumn. The chief ofXiongnu wcnt on a punitive expedition for Wusun, the ancientKazakh, in the winter of 72 B.C. Being caught in heavy snow withdepths to 240 cm, the personal and domestic animals died in largenumbers. This period was named the New Ice-epoch (WangQingtai,l981). At that time, the lakes had larger areas in the TarimBasin and there were many hydrographic nets around it. The famousSilk Road was rather brisk at 2,000 B.P. According to thenarratives in the Shan HAi Classic and the Annotations on theWater Classic, the books on ancient Chinese Geography, therewere 8 rivers distributed over the east, the south, and the northwhich flowed into Luobubo Lakee. Based on the American satellitephotographs and the textual criticism(2hou Tingru,1978), the areaof Luobubo Lake was indeed larger during the former Qin Dynastyand Han Dynasty. Which confirms the 'records in the ancientbooks. Some researchers (Yang Huairen,l981) considered thestiuation wkich "did not accord with the common law of dry glacialepoch", reflected " the size of the lakes and the rivers in Tarim Basinwere not absolutely dominated by the dry or humid period".The author thought that stiuation is just another piece of evidencefor the coexistence of cold and wet climates.4) The formation and development of most interior lakes inthe northwest of China related to the Quaternary glacial action.Each glacial action made up a depositional cycle together with therelevant new deposition. When the glacier had the strongest action,the lakc was provided with the highest water level, the largest area,and the thickest deposition layer. The contraction of the lake surfaceand the thinning of the deposition layer was accompanied withthe reccssion of the glacial action. The relationship between coldclimate and precipitation was also reflected therefrom.5) The evidence of annual rings of trees. The restrictive factorof the naturally survived trees is not temperature but moisture inXinjiang. In the dryer years, the photosynthesis was weakened andthe cell division of cambium was restrained, thus the annual rings

~ 107)'0 50 1~~~~mirabilite oozc clay silt sand humusFig. 1 Sporo-pollen percentage diagram,Balikun Lake CoreZK-024were narrow. The annual rings widened at the year with more hu+midity.6) The research results on mean winnd-field and the transportedvapor quantity in the atmosphere a%ove China showed (QiChengyu,l986),the compound quantity of the whole atmosphereabove China comes from three annual air flows. The vapor carriedby the southwestly airflow from the Indian Ocean and the Bay ofBengal made the greatest contribution. We can see from the 5,500meter uper-air charts, the thermodynamic action of theQinghai-Xizang Plateau was strengthened in summer, it thenintensified the wet monsoon, which carried a great quantity of vaporand made the precipitation increase largely in the area whichwas effected by the monsoon in summer. The motion height of thewet monsoon, however, is only about 3,500 meters (XisXuncheng.l985), little of it could go over the Qinghai-Xizang PlatcaUof 5,000 meters in mean height. In the northwest of theplateau, the upward air current from the surfacc of the plateaucomes down and the subsidence heating effect caused the air to bedryer in summer;Xinjiang could not get cnough moisture to form plentifulsummer rainfall as a conclusion.There was hardly any change in theswelling state of the Qinghai-Xizang Plateau since the lateQuaternaryJt is deducible that the great moisture increase were notable to make a significant influence to Xinjiang in the warminterglacial epoch.In summary, the high temperature was bound to cause.aridity,and only low temperatures enable the humid climate trend. Thecorrespondence between dry and warm or humid and cold periodshas been a general law in Xinjiang since the late Pleistocene.ON THE FORMING MECHANISM I.""HOT-DRY CLIMATES-OF THE COLD-WET ORXinjiang is located at the middle-latitudes, the movement ofthe planetary frontal zone does not constitute the most importantinfluence on its humidity and precipitation for many resons, whichwas greatly different from the situation in eastern China, where themain cause of the greater rainfalls in the warm period was thesouthwest monsoon. Thc beginning and the end of the rainy seasonand the motion of the, rain-band were also closely related to theadvance and retreat of the summer monsoon there. The research on18,000 B.P. climagraph showed that the precipitation decreased 20percent in the northern hemisphere in the glacial epochs (YangHuairen,l981). But this decrement was not homogeneous everywhere.InChina, this situation might find expression in the easternpart with the monsoon climate.Xinjiang was not affected by rhe1204

, Table 1 A comparison of the monthly mean temsetaturc andhumidity betwtcn:January and July at Umqi,XinjangTimes ' Jan. Jul. 'Water content inthe air (g / cm')Water-vapourpressure (hpa)Precipitation(mm)Temperature("C)Evaporation(mm)Relativehumidity (YO)0.49 2.711.6 11.88.7 21.4-15.4 23.510.1 357.178 38monsoon, so the increment and decrement of the precipitationcaused by the monsoon had hardly any effect on Xinjaing. Merelythe incrcment and decrement of evaporation in the global sea-sur.face might cause the variation of the water-vapor transpirationvolume in the prevailing westerly jet stream above Xinjianght thescale of this variation, as illustrated by the facts cited in table I, wasminor, compared with the change scale of the temperature andevaporation power in the'arid area. Therefore the warm interglacialepoch could not bring about the humid environment. On the otherhand, the cyclone actrivities were enhanced in the middle latitudeareas,and the depression of the dew point in upper-air droppedgreatly during the glacial epoch, which enabled the water vapor tosaturate easily and contribute to the formation of precipitation.Though the water vapor in the air current decreased to someextent, the precipitation did not decrease distinctly.Fig.2 shows the relationship between the saturation vaporpressure (Ew) and the air temperature. The lower the temperature,the smaller the satuation vapor pressure, which decreses by the indexrate. When the temperature decreses from O°C to -lO°C , Ewdecreases from 6.1 hpa to 2.61 hpa; and when the temperature doceases from 3OoC to 20°C, Ew deceases from 42.4hpa to 22.9hpa. Itis the same in a decrease by 10 'C, but the Ew value decreased withheat is over five times as much as with cold. Thus it can be seen thatthe lower temperature greatly contributed to the condensation andprecipitation, not dryness. Because of the great disparity of watervapor content between summer and winter, the promotive effect ofTable 2 A comparison of mean monthly vapour pressure andprecipitation of the typical months of winter and summer betweeneastern and western China+Placcname Warbin Peking Nanjing Chcngdu Yaxian Taibei Ururnqi TurpanMcanmonthlyvapour(Jan)Pressure (hpa) (Jul)1.1 I .9 5.3 7.2 18.3 10.9 1.6 1.821.0 25.4 30.6 27.8 32.0 21.6 11.8 14.5Mcan monthly (Jan) 3,,3 .O 30.9 5.9 7.2 87.8 8.7 1.5prccipitation160.7 192.5 183.6 235.5 142.9 227.5 21.5 2.3(mm)(Jul) . ,Thcmultiplc orvapourprcssurc 19.1 13.4 5.8 3.9 1.8 2.0 7.4 8.1of Jul/ JanThe multiplc ofprccipitation or 43.4 64.2 5.9 39.9 19.8 2.6 2.5 1.5Jul/ JanThc incrcmcnt ofprccipitation in 157.0 189.5 152.7 229.6 135.7 ' 139.7 12.8 0.8fuly (mm)* The data cited from Zhang Jacheng et al, 1985* The data cited from Zhang Jaeheng et ai, 1885i

Table 3 A comparison of climatic fluctuatidn in Borne regionsof northern Xinjiang within thirty years , .place Dabancheng Ururnqi Paotai Mucuuowan Buolcclimate periods cold warm cold warm cold warm cold warm cold warmthe years when the extremetemperaturc happened1969 1973 1951-i977 1956 1974 1969 1975 1960 1974decrement ofprccipitaion 7.4 15.0 14.7 16.412.4in warm periods(%)mean annualcvaporation (mm)2700 2745 1508 2262 1802 1865 1894 2003 1534 1592\increment ofcvaporation inwarm periods(%)1.6 49.9 3.5 5.8 3.9mean annualcvaporation- 40.6 44.6 5.1 9.1 12.0 14.5 15.0 18.9 7.9 9.4prccipitaion ratioincrement of E-Pratio in 10.0 18.4 20.0 26.0 19.4warm period(%)predipitaion anomaly perccntagein the year whcn thc 3 -23 -13 -42 19 -55 20 -13 46 -44cxtrcrnc tcmpcralurc appcarcd(%)* The data were sorted out from the records of surface observationlow temperature to humidity is concealed in eastern China. But inXinjiang, with an exceedingly low content and smaller changing ratioof water vapor, the humid effect of low temperature can manifestitself. It can be seen from the research results on the advanceand retreat of ancient glaciers. It was provided by the research onthe glaciology that the development of glaciers must have two ofmain conditions.One islow temperature, and the other is precipitation. Theglacier is unable to advance by a wide margin if it is in lowtempcraturee but the accumulated snow is insufficient. It is providedby the research on tho relationship between the ten-years periodicclimate fluctuation of China in the last eighty years and theprocess of glacial advance or recession on the Mount Qomolangmaand Tianshan Mountain, that the scale and moving speed of theglacier could increase only in the periods when cold correspondedwith humidity. Therefore, the periods when the ancient galciers advancedby a big margin should be the cold-wet environment.(hpa)Fig.2 A diagram of interrelation curve of saturation vapour CONCLUSIONS AND DISCUSSIONS- ..I .-~pressure with respect to water and temperatureIn summary, under the spccial geographical environment of9 1206

Xinjiang, the correspondence of a cold-wet or warm-dry period isthe general climatic law in the arid region since the late Pleistocene.Though the vapor concentration of the upper air current was less incold periods, the relative humidity near thc ground was higher thanthat in warm periods.The interior arid region got dryer and"dryer 8s a result of theQinghai-Xizang Plateau getting higher since the late Pleistocene.Humidity, the word used in this paper is just a comparativeconcept. The vegetative cover rate was lower regardless of the dryinterglacial epoch or humid glacial epoch. It is for this reson thatduststorms or sandstorms were able to be formed in any epoch inthe arid region. In the warm inter-glacials, the higher vapor con.centration of the upper air current makes the tiny dusts into thecondensation nucleus and drops back to the ground with the rainsdespite the fact that the rainfall still cannot contend with the strongevaporation power. On the other hand, the vapor concentration ofupper air currents was lower in cold periods. The dusats, failed tocollect vapor to form raindrops drifted with the air currenteastward to the monsoon region, then condensed and dropped toform loess deposition.me author does not deny the possibility that some short periodsof warm-humid or cold-dry climatc existed in Xinjiang sincethe late Peistocene. But the correspondence of cold-wet orwarm-dry periods should be the main features of climatic type inXinjiang, the special arid region.Wang Qingtai (1981) Ancient Glaciers at The Head of UrumqiRiver, Tian shan, Journal of Glaciology and Cryopedology,Vo1.3, Secial Issue.Xia Xuncheng and Fan Zili (1985) On The Environment and ClimaticChanges in Luobubo Region, -Collected Papers onQuaternary Research in Xinjiang, Xinjang Press, Urtlmqi,8-26.Yang Huairen and Xu Xin (1981) The Influence of QuaternaryIce-Age Climate on The Loess and Desert, Selcctcd Papers onQuaternary and Glacial' Geology of Xinjiang, XinjiangPress,pp.6-18:Yang Ruairen and Xu Xin (1981) The Physical Evolution of Chinain Quaternary, Geography Science and Technical Information,Nanjiang University Press,pp. 4-6.Zhang Jiacheng et al (1974) Preliminary Approach of The ClimateChanges of China, Science Bulletin, 4.Zhang Jiacheng and Lin Zhiguang (1985) The Climate of China,Shanghai Science and Technical Press, Shanghai, 178,579-588.Zhou Tingru (1978) On'The Remove of The Luobunor Lake,Journal of Beijing Normal University, 3.Zhu Kezhen (1973) Premary Research on The Climate Changes ofChina in Recent 5,000 Years, Science Sinica.ACKNOWLEDGMENTSI would like to thank Dr. Zhang Xuewen, the Senior Engineerof Xinjiang Meteology Service, for helpful SuggeFtios and critiquesof the manuscript,REFERENCESChen Hanyao et a1 (1963) Xinjiang Climate and The Relationshipbetween It and The Agriculture, Science Press, Beijiang,pp.62-118.Qi Chengyu et al (1986) The General Expression of Chinese Climate,Science press, Beijiang, p.34.. 1207

THE INTRODUCTION OF APPLICATION METHODS FOR CTIN FROZEN SOIL EXPERIMENTAL RESEARCHPu 'fibinState Key Laboratory of Frozen Soil EngineeringWGG,CAS,ChinaThis paper introduces the basic theory and method for experimenting with frozen soil samplesnondestnrctively, quantitatively and dynamically using CT (Computer Tomograph). Accurately ob.serving the digital imagea inside the samples in experiments, we can find a new technique for studyingfrozen soil in engincciing.THE SIGNIFICANCE OF INTACT QUANTITATIVE EXAM-INATION SAMPLESAlong with intensively developing scientific study, computertechnology has beeen widely applied to analyse expcrimental processesfor many experimental equipment. On the other hand, intactexamination of frozen soil and other samples is a very efficaciousmethod in theoretical studying and engineering applications in differentenvironments (tempcratute, pressure, composition of samplesand their changes). We hope to find some way of connectingthe two fuctions of reflecting the physical shape and properties ofmatpials in experiments. The CT equipment is an efficaciousmethod. Using data analysis, We can find the regularity inside exchangesamples, and get new theorys for qnginccring applications.0THE TECHNOLOGY INDEX OF'mDifferent material has different absorbing cocfficicnt p ofX-ray. The image of scanning and reconstruction shows quantitativep maps. Professor Housfield has given a definition of CTcount, air is -1OO0, water is 0. The biggest examinable materialI I+1H.VoltngeTHE THEORY OF CT EXAMINATION"CT" is the abbreviation for the computer tamography. It usesX rays to penetrate the object slice which is revolving. Detectorsgather the X-rays which have bccn attenuated from the object. Thecomputer gets this information, does A/D transformation andcalculates according to algorithms. Then the absorbed coefticient p,which corresponds, with the physical properties of everypoint ofthe object, is quantitively displayed as data images and the instrubtions ate dynamically analysed, quantitatively and dynamically.Because the samples are not damaged in the whole process, thesamples can be examined many times for every slims, the image8can be recorded, processed and analyzed. CT is ordinarily used toexamine the human body, but it also can be used to check non-metallic materials. With some improvements, the instrument may becomean efficient examination tool for samles. The principleschematic diagram bclow shows the system.Dig.1 CC system working whemitic diagramdensity is about 4 grcm / cm', and a munt of 3000. This densitymeasurement range can fully cheek every non-metallic material.The CT munt caculation contents linear equation:In the equation, CT-CT count, p*-the X-ray absorbing coemcient1208

of examined ob#& pW4e X-ray absorbing coefficient of pur^water.In 20 years of development , there are 5 designed types of CT.The equipment of our laboratory may complete examination of aJice within 5-10 seconds, the diameter of the sample extqnda to 42an, It has 523 datectors, 576 pulw exposures for a slice. The detected data is calculated by computer. Spatial resolution reaches 0.8x 0.8 mm and smaller linear objGCts and 0,065 mm diameter metalpoint can also be found. The density contrast resolution reaches to0.5%. Since calibrating has been done by using standardpolyethylene phantom, which has the same shape as samples, theCT counts have absolute quantitative significance, more accurateray and display data. The images would be enhanced, distinguishedand compared with other data. So we shall get better results of experiments.THE EXAMINING METHOD OF SAMPLESAdding some equipment to the common CT, making somesample containers and models which comply with experimentstandardization, using programmed software for data calibrating inthe scanning process, making the apparatus of pressure and temperatureinside the sample container for examining constructionchanges in different environmental conditions, pre-burying somephysical quality materials, we can analyse temperature and stressfield in samples according to the changes of form position. We alsocan use new software to display 3 dimensional images and to analysesample quality more clearly. Combining physical principlesand other techniques, we may design many efficient experimentsdepending on the purpose, and so that extensive research may bedone.THE SIMPLE EXAMPLEThe photo 1 is an image which shows the layers of a sample. Itindicates that there is great differance of CT data among ice, water,air, and soil. The split of ice appears very clearly. In the differentareas, the statistical data shows the expansion or contraction of thesamples. The position change of image point reflccta movemtnt ofthe matter in the examining process. The display control of the imagesis very useful for spectators to distinguish different materials.Because the examination study of CT is just beginning, thenew method is only introduced so that it may be of usa in krturescientific studies.1210 '

" Preliminary Data for Permafrost Thermal Regime and itsCorrelation Meteorological Parameters near the SpanishAntarctic Station"M. Rarnos. Department of Physic. UAH. 28871- Alcal6 de Henares. Spain.During 1991/92 Antarctic summer a turbulent atmospheric parameters measuring devicewas installed near the Spanish Antarctic Station (SAS), in order to register theexchanged energy flux between the soil and low the atmosphere.Temporal evolution of permafrost was also registered using several temperatureprobes.Base on the energy exchanged flux data and the movement of the free boundaryin permafrost. We will study the permafrost evolution as function as the energy change inthe soiVatmosphere surface.INTRODUCTION.Soil thermal evolution in coldregions where permafrost exists dependson the energy exchange between soilsurface and low atmosphere and this willbe a boundary condition in the heat andmass transfer process in the permafrostactive layer. This frost and thaw 1processes could generate differentgeomorphy surface structures which canbe quantified.In the study area, Livington(South Setlhand Island. Antarctica) seefigure 1 where the Spanish AntarcticStation is, the active layer of permafrostis about 0.7 to 1.0 m. deep Hall (1992).The aim of this study is tomeasure, using the aerodynamic method.Dyer(1974), Cancillo (1991) with atwo levels micro meteorological towerthe energy flux exchanges between soiland low atmosphere in differentturbulents regimes in 1991192 summerpolar conditions.The measured flux energyexchange in the soiVatmosphere will beuse as a condition in the heat transferpermafrost problem Sorbjan (1989).Theoretical thermal evolution ofthe permafrost active layer study will becompared with experimental thermaldistribution data measured with severalthermal probes Ramos (1 993).Figure 1Experimental area.1211 '

EXPERTMENTAL METHOD. 'Characteristic parameters, of theturbulent atmospheric flux inatmospheric surface layer, air speed,temperature, relative humidity areramdon hnction of space and time. Anysoihtmosphere energy exchange modelhas restriction and there is not exactsolution to the problem.To study thermal evolution inpermafrost a necessary condition is toknow the energy exchange flux insoiVatmcrsphere interface. And it will.define boundary conditions in the heatmass transfer permafrost process.' The most important energyexchange flux between soil and thelower atmospheric layer, are:- The soil (as a grey body) visibleradiation absorption.- SoiVsky infrared radiation exchange.- Airhi1 sensible heat flux.- Airhail latent heat flux' (less important)Lunardini (1 98 1).A two levels micrometeorological tower was designed andinstalled near SAS in the Antarctic1991/92 summer to measure wind ,speed, temperature, relative humidity,incident and reflected radiation into thesoil. So base on this and using theaerodynamlc method we studied thesoiVatmosphere energy flux evolution.The tower was equipped withtwo common anemometers (180,,560cm. height) and three thermocouples fortemperatures measuring (90, 180, 560cm. height ). Data was registered everyminute and average value was storedevery thirty minutes. Local airtemperatures were also measured (0, 5,10, 20 cm. height) at the same rate.Figures show the evolution offollowing parameters in a typical day:fFiguie 2,- wind speed evolution in bothlevels,-176Meteo 8/11/9 20 5 10 15 20 25ti- ull (H)Figure 3.- Daily tower temperaturesevolution.9008007008005004003002001000-100 0 , 510 limp0 w11 (H) I5 20 25Figure 4.- Incident and reflected visibleradiation.

-12IO8,E2 6Coef.Pelicula 8/ll/9 2thermal distribution in the soil and thequasiestacionary method for sine wavepropagation.1 4Suelo V 8/11/9200 5 10 IS 20 25Ti-W (H)Figure 5.- Daily temperatures of the close tosoil layer.With data acquired and usingaerodynamic method we will find outsensible and latent heat flux evolution insoillatmosphere boundary.Thetd distribution in thepermafrost active layer was measured atdifferent deeps (between 0 to 1 m.).Steady-state analyze of the thermalevolution measures will allow us tostudy the fluxes inside sail arid itscorrelation with the soillatmosphereexchanged energy flux balance.Figures 6 and 7 show typicaldaily evolution on two of the group offive temperature probes used in ourmeasuring system.+ l q(90) 4 llrp("10) + Tlrp.(-50)Figure 7.- Daily active permafrost layerevolution in probe V.CONCLUSIONS.1 .- Aefodynamics method is suitable forstudiing turbulent flux balance in the lowatmospheric layer.2.- Thermal spatial evolution inpermafrost is suitable for analyzing theactive permafrost behavior and itscorrelation with the surface energyfluxes.IO9- 68,Suelo IV 8/ll/9 2REFERENCES.Cancillo. N,L; Baludio e iosfluyo8 de energfa @n lacapa lieite de suparfieieITesis. Dnv Igrtraaadura . (1 991 ) .A.J; A Review of FiuxProfile Relatimsrhips *.Boundary Layer Meteor.pp 363-372. (1971).Figure 6.- Daily active permafrost layerevolution in prove lV.Data analysis will allow to findout soil roughness. So, thermaldifusivity, could be obtained usingIJc,L Xechrpn.ica1Weathering on Livicstonxslma south metlandIslande, AIlttU-ctica*.Recent PragresB inAntarctica Earth Scieace .)757-762. (2992).Lunardini, V.J; Heat Transferin Could Climates a. mwYork. VNR. (1981).1213

R~UZOS, I; "An Exact Solution forthe Finite Stefan Problem Sorbjan, 8; Structure of thewith Temperature-dependent attmsphexic Boundarythermal Conductivity and Layer* Prentice HallSpecific Beat* Int. J. (1989) IRefrig. Vol 16, n* 5(1993) 91214

AN EXPERIMENTAL STUDY OF CANAL LIWING PREVENTEDFROM FROST DAMAGE AND SEEPAGERen ZhizhongNorthwest Hydrotechnical Science Research Institute ,Yangling Town, ChinaThis paper prcsents the observation results during the two freezing periods of 1985-1987 for an experimentalcanal with a ltngth of 400m and the operation investigations from 1988-1991 for the lined canalwith a length of 49.8 Km. The results show that the canal should adapt, limit or partly eliminate thefrost heave and cut off seepage loss. For the canal section types, the best structural type of lining slabs isthe trapezoid with an arc at the down slope angle. The best material type of lining is the concrete slabcombined with a layer of plastic film. The better structural types of lining slabs are the concrete slabswith 18 and 10 cm thick, the "II" shape concrete slab (6 cm thick), the 6 CM thick asphalt concrete slab,and the cast-in-site ribbed concrete slab (8 cm thick, 20cm rib hcight).MTRODUCTIONIn Dongying of Shangdong province, an arterial canal wasbuilt with a 49.8 Km length, 16.5-11.0 m bed width, 3 m depth,1:1.5 slope coefficient, 2 m water depth, 1:7000 in water slope ratio,35 m'/ s in design flow capacity. In order to find reasonable typesof materials and structures for canal linings which can be preventfrost damage and seepage, a length of the arterial canal, when constructed,was used as an experimental canal with a length of morethan 400 rn. In the arca, the mean annual air temperature is about12.2'C. the days of mean daily air temperature below O'C are about76-87 days. The negative surface temperature indices are250-350'C day. The frozen depth is about 55 cm. More than 80%of the area distributed under the canal bed is composed of sandyloam with the ground water table ranging from 0.7 to 1.9 m. Fineand loam particle content in the sandy loam is about 68.6-87.7%.The capillary water is up to 1.7-1.9 m. The dry density and plasticlimit water content are 1.4-1.5 g / cm3and 17.2-20.3%.DESIGN PRINCIPLEFrom a technical and economical point of view and due tostrong frost.damagc in the area, the experimental canal design principlewas that the structural type of canal lining should adapt, eliminateand limit partly the frost heave, and cut off seepage loss.From the design principle, the structural typcs of canal liningsconsisted of:(A). (I) Prefabricated slab (8-10 cm thick) and a layer of plasticfilm (0.2 mm thick), (2) "II" shape prefabricated concrete slab (6an thick) and plastic film;(B). Cast-in-situ ribbed concrete slab (8 cm thick, 20 cm ribheight) and plastic film;(C). 6 cm thick asphalt concrete slab;(D). Prefabricated concrete slab (8 cm thick) and sealcd subsoil(40-60 cm thick) with plastic film;(E). (I) Prefabricated concrete slab and plastic film, (2) prefabricated reinforced concrete slab and plastic film ( the two types arewidely used in canals ).OBSERVATION AND RESULTSAfter the experimental canal was built, systematic observationswere conducted from 1985 to 1987. The distribution ofobservationpoints along the canal section is as shown in Fig.1. The mainresults are decribed as follows:In the two freezing periods of 1985-1987, the lowest and highestair temperatures were -16.4'C, -20.4OC and -12.6'C, -19.4'C,the lowest and highest mean daily air temperatures were -12'C,-10.3'C and -ll°C, -8.S'C, and the negative surface temperatureindices were 282.2"day and 141.4'C day, respectively. The distributionproperty of ground temperature and frozen depth are shown inTab.1, and Fig.2. It is clear that the lowest ground temperature was-6.8OC and the maximum frozen depths was 83 cm for structure A.The ground water tabla underlying the canal bed was about I rn in1986, and 0.1-0.5 m in 1987. The subsoil water content was about27-35% for the shaded slope and bottom, and about 20-25% forthe sun exposed slope.The maximum frost heave amount of the observation points isshown in Tab.2. The greatest and the lowcst displaccment were 200rnm and 29 rnm respectively for structure I3 and D.The structures of A, B, C and I3 had been opcratcd well during1988-1991. They cpuld effectively prevent the canal from frostdamage and seepage after undergoing several freezing-thawing cyclcs,except for a small crack on the shaded slope for C structure(see Fig.3). Rut E type or lining slab suffered from frost damage.The damage circumstance are shown in Fig.4, 5, and 6.

1 canal sectionFig.] Distribution of observation points along canal sectionFig.2 Distribution of frozen depth in A structureFig.5 Frost heave stagger in structure EFig3 Sliding circumstance of test sectionFig.4 Frost heave circumstance in structure EFig.6 Slide and collaps in structure E1216

ITable I. Maximum frost depth (cm) and lowest ground tern- heave and high ground water table areas, it is the best canal sectionperaturc ('(2) at 10 cm depth undcr slabshape.sexionground map.frozen depthground tcmp.frozen depthground temp.frozen depthground tcmp.frozcn depthground temp.frozen depthground tcmp.frozcn depthground tcmp.frozen depthasphalt EM.l slab 8 cm slab scald slab cretc1986-5.8I1-4.454-3.653-3.249-2.040-1.2291987 1986-6.873-5.0 -6.859 83-4.0 -6.844 65-2.9 -4.237 56- -4.049- -2.842- -1.4281981 1986 1987-4.853-6.5 ' -6.058 50* 50-6.2 -5.852 so* 45-3.8 -3.834 50. 44--5041 3931slab1986 1987-5.2I1 53-5.012 55-3.662 38Effective Methods of Preventing Canals from Frost DamageThe methods to prevent canals from . damage - are follows: as(1) In order to adapt canal to the displacement offreezing-thawing, the trapezoid section canal with an arc at thedown-slope angle and bcd can be selected.(2) For materials, the structure consisting of a concrete, slaband plastic film is the best one. It combines the advantages of bothconcrete and plastic film. ,(3) Through the contrast of structure and cost analysis, thebetter structural types of lining slabs are arranged as A(I), A(2), C,and B. For D, if the construction problem, can be solved, it is also abetter one. For the structures of A(2) and B (see Fig.7), because ofthe effect of air insulation between slabs, and films and ribs aroundthe slab, frost damage is reduced. In addition, using link expansionpints, concrete slab failure is avoided.Design of Frozen Depth and Calculation of Frost Heave Amount"The relationships between negative temperature index of theconcrete slab and of the air and frazen depth are shown in Fig.8.The subsoil frost heave properties and corresponding structures aredetermined by the maximum value of frozen depth. The calculationmethod is described as follows:ground temp.frozen depthground temp.frozen depthground temp.froLen depth-0.621-0.623-1 .o22-1.0 4,821 14-0.5 -0.619 19-0.3 -0.621 11-1.5 -1.221 20 18-1.0 -1.2I8 21 18-1.3 -2.221 23 21-1,414 18-0.520 20-1.217 20Design ,.- . Of frozen aepthS,observation point for A(l) structure:h = 34.45 + 4 . 2 8 zS,obscrvation point for A(2) structure:h = 45.72 + 3 . 3 2 cground tcmp.frozcn,depth-4.948-2.8* The maximum depth measured is 50 cm.41RESULT ANALYSIS" plastic film0n slab rib I slabaSubsoil Frost Heave PropertyBecause the frost heave amount is from 5 to 10 cm, accordingto (SL23-91), the subsoil engineering classification offrost heave belongs to the TI1 class.Frost Damagc Types for Differett Canal Scction.~ShapesBecause of the restraint of the down-slope angle to the frostheave displacement, the displacement failure often occureed in thecentre ot the bed for narrow canal, and the displacement drack OGcurs near the down-slope angle for the wide bed canais.The trapezoid canal with an arc at the down-slope angle,compared with trapezoid section canal, it has some sepcial advantages,such as eliminating the restraint at the down-slope angle, decreasingthe inequable frost heave, better anti-heave capacity, betterhydraulic properties and low cost. Therefore in strong frostFig.7 Local effect of structurewhere hlis frozen depth (mm), T,,is surface negative temperatureindex of lining slab ("C day). For A( I):Tal =-3.09+0.67Tb, r~0.999. (3)For A(2):T,, --2.97+0.55Tb, r =0.999. (4)Tbis daily lowcst negative air temperativc index ('C day), r isregression coefficient.Frcst heave amount (the displacement of lining slab)S,observation point for A(1):bl=-1.73+0.15+(hl * W,), r=0.97 (5)S,obcrvation point for A(2): '6, = 1 .04+0.13+(hl- Wl), r = 0.84 (6)1 1217

Table 2. Maximum frost heave amount of observation point(mm)ycarstmcturc 10 cm lI 8cm scal asphalt stmcturce slab slab slab slab slab B86 12.5 0.83 19.0 13.0 3.0 12.0 6.7a7 -20.0 17.0 23.0 15.0a6 33.3 12.5 58.0 31.9 4.0 32.0 39.1 8 cm slaba7 44.0 41,O 43.0 60.0a6a786a1868748-200 46.6 7a 69 14 45 77.447 51 79 a3 Negativ temperature index on slab37.4 58.2 69 16 54.0 72 68.2 ' ("C day)20.0 5330.8 53.2 63 73 29 101 70. I368660.0 70.022.0 74.086SI 167.0 62.08637.4 9.2 61.0 64.018.0 4.0 44. I8625.a 4.2 7.0 29.02.0 6.0 3.387868786878687-13.0 -I6-5.0 5.0 2.0 6.07.0 -10-1.7 8.3 7.0 2.03.0 -6.0I .7 6.7 14.0 6.04.0 5.0-I8 -26 negative air temperature index(OC day)8.0 9.0 0.838.0 -IO4.0 6.0 -0.83-14.0 5.00 5 ' 0-7 -5Fig.8 Relationships between the frozen depth and negativetemperature index on concrete slab and of thc airwhere S,is frost heave amount (cm), W,is soil water content atS,point (mean value of IOOcm soil depth).REFERENCESThritovize, H.A., 1985, Mechanics of Frozen Soil, PublishingHouse of Science.Mathematical Research Institute, 1977, Chinese Academy of Sciedce,Common Mathematic Statistics Methods, PublishingHouse of Science.

THE! IMPACT OF SALT TYPE ONDEFORMATION OF FROZEN SALINE SOILSRoman L.T.',Alifanova A.A.', Zhang Changqing''Department of Geology, Moscow State University, Russia2State KeyLaboratoryof Frozen Soil Enginccring,LIGG,CAS,ChinaThis paper contains data on long-term deformation of frozen soils contaminated with solium chlorideand magnesiam and sodium salphates. The results show that testillg of samplcs with similar physicalpropcrties which differ only is salt content gives a family of creep curvcs.Relativc deformation increascsproportionally to salt content rise and the time necessary to achieve the same magnitude of deformationreduces. Thus, higher salt content accelerates crccp. The impact of salt contcnt on deformation is idcnticalto the effect of time.This reduces the time requircd for detormation analysis and to apply salt-timcanalogy to predict long-tcrm deformation of frozen saline soils based on shorter experiments.It is gcncrally rccognized that salts present in soil water reduccstrength of frozen soils and increasedeformation since during dissolutionsalt ions dissociate into anions and cations thus contributingto water coherence. The freezing point of soil water drops and theamount of unfrozen waterincreases. The above processes dependon the quantity of dissociated ions which, in turn, is affected by thechemical composition of salts and theirconcentration.Hence, theextent to which salts affect strength and deformation of frozen soilsdepends on the same parameters, i.e., salts somposition and poresolutionconcentration.Soilsgencrally contain a complex of salts and it is difficult toidentify the effect of each salt type. The concentration of pore solutionis unstable and varies with moisture content. That is why to assessthe impact of salt type on soil strength and deformation totalsalt content is most commonly used which is the ratio between saltand dry salt masses. The experimental evidence available establishesonly relationship between mechanical properties of salinesoils and total soil content. Such an approach does not permitinvestigators to establish the nature of impact of salt type on thestrength and deformation of frozen soils andobtainreliable quantitativevalues for design.characteristics.Only purposeful investigations based on extensive experimentalevidencc can eliminate this gap. This paper contains data onlong-term defoqnation of frozen soils contaminated with sodiumchloride andmagnesium and sodium sulphates.Tests and data processing were carried out following salt-timeanalogy method. The validity of time analogymethods to predictlong-term dcformation of frozen soils, including saline, has beendemonstrated before (Roman,l987,1990). Frozen saline loam hasbeen tested for uniaxial compression under isothermal conditions.The test samplcs were 89.8-92.5 mm high and 44.8-45 mm in diameter.Major physical properties were assumed to be constant.Given below are theirmean values: frozen soil density = 2.0g/ cm3; soil particle density = 2.71 g / cm3; moisture content =30%. The salts were NaCI, Na2S04 and MgS04 and salt cOntCnt= 0.2, 0.4, 0.6, 0.8, I .O, 1.5%.Three tests were conducted for eachsalt content case. Mean deformation values were used in the analysisof test data.The results show that testing of samples with 'similar physicalpropertics whichdiffer only in salt content gives a family of creepcurves. Fig. 1 shows such a family of'MgS04 frozen loam.Itfollows from Fig. 1 that relative deformation increasesproportionally to salt content rise and the time necessary to achievethe same magnitude of deformation reduces.Thus, higher salt content accelerates creep. The impact of saltcontent on deformation is identical to the cffcct of time.This rcducestime required for deformation analysis andapply salt-timeanalogy to predict long-term deformation of frozen saline soilsbased on shorter experiments.Both the results of short-term experiments and expected valuesof long-term deformation modulus yielded by salt-time analogywere analysed to determine the impact of salts on frozen salinesoil deformation, Themajor task of such an approach is to definetransmission coefficient, i.e., correlation betwecn times t,and t,duringwhich similar deformations offrozen saline soils are ,achievedfor each given salinity.Creep test data permit one to define transmission coefficientthrough time-dependent compliance which is a relativedeformation-stress Relation ( Ji= E, / a). A family of curves J,-lntis plotted for each series of isohetric constant-stress tests. Heresalt content is the only factor which accelerates compliance. Therest physical properties of samples remain stable. According totime-analogy theory(Wrzhumtsev,l982) determination of transmissioncoefficient is much easier when opmpliance curves of such afamily arc similar. Our investigation$, revelled that compliancecurves of saline soils are observed if stress is attributed to the area- 1219

D ~ 1.5 ~ % ~ =1 .oTable 1 Long-term deformation modulus(50 years) or frozen saline loam, MPaF-T4 2-t (min)Fig. I. Creep curve family for frozen loam with differentMgS04 content. Uniaxial compression test.Stress = 1.26 MPa,temperature = -3'c (p=2.0g/cm3;p,=2.71 g/cm3; W,=0.3).occupied by solid components (soil particles and ice) instead of thecross-section area of the sample. With 'accuracy acceptablcfbrpractical purposes this area (Fk) can be assumed to be proportionalto the content of solid components in volume unit of soil (k) whichcan be easily calculated provided basic physical properties areknown (soil particle density, p,; dry soil density, pd; ice, p,;totalmoisture, W,and unfrozen water content, W,,1 (W - W ),f' = I- - 2-c 1p, p,With this assumptiontaken into account the area occupied bysolid components (Fk) will bc F, = F K and mess attributed to thisares uk= u/ K . Here compliance is the relationshipbetween therelative settlement at a given instant of' time t (E,) and stress uk,J,, = el / u,(where F is area and u is stress attributed to the samplecross-section).The obtained crcep curve families for saline loam are shownon Fig.2. It follows from Fig.2 that compliance of loamcontaminatedwith NaCl is scvcral timeshigherthan for Na,SO,andMgSO,, all other things beingequal.The effects of the two latter salts on the compliance diffcrslightly. Thc transmission coefficient is also salt-dependent. NaUlhas a more pronounced cffcct on the transmissioncocfficicnt in.crcase with increasing salinity than two other salts. Thcsc cxamplcsalso illu

-" 2.0p 1.5k 1.00.50 . . . , . -i0.5 1.0 1.5 2.0Fig 2. J,vs Igt,. Uniaxial compression test data for loam withTable 2 Freezing point of loamy sand(p =2;p,=2.71 ; Wa=0.3)M~SO, i -0.2 -0.6 -0.8 - -1.0 -1,2 -1.68- -- "."-0.81 - -1.01 -1.21 -1.7" - - "" ,- ,-NaCl -0.2 -0.68 -1.08 - -1.61 - -2.21 -3.21"" -REFERENCES"Fig, 3. Jp / uq vs T / Tblfor frozen loam contaminated withdifferent salts (salt content between 0 and IS%, temperature = -3,-5, -7'c), time = 8 h.Glinka N.L.,1976, General Chemistry.Chemistry PublishingHouse, Moscow,p.712.Roman L.T.,1987, Frozen peat as foundation soil of engineeringstructures.Science Publishing House,Moscow.p.229Roman L.T., Kuleshov Yu.V.,1990, Prediction of' long-termdeformation of saline soils with time analogies. In: Frozen SalineSoils As Foundation of Engineering Structures. SciencePublishing House, Moscow.p.73-83.Roman L.T., Artyushina V.T., Ivanova L.C.,1992, Relationship betweenfrozen saline soil strength and soil water freezing point.Moscow State University Publishing House,Moscow.Urzhumtsev Yu.S.,1982, Prediction of long-tcrm resistance ofpolymeric meterials.Scienccs Publishing House,Moscow.p.22IFerry G.,1963, Visco-elastic properties of- polymcrs.LiteraturePublishing house, Moscow.p.535.1221

..GEOPHYSICAL METHODS OF CRYOLOGY ECOLOGICAL MONITORINGB.M. Sedov, Yu.Ya. VashchilovNorth-East Interdisciplinary Research Institute, Russia Academy of Sciences, MagadanGeocryological wells are peculiar as they can be drilled without muds, and so-called “dry” drilling allowsto avoid thawing of rock. There are some difficulties in making geophysical investigations of suchwells due to impossibility to use electrical logs. But this can be easily done by using a series of radiactivcmethods, seismic log and caliper log, which allows to take into account the effect of change in a well diameter.The series of radiactive methods iucludes gamma-log, gamma-gamma density log, neutrongamma-log and X-ray radiometric log. The combination of all these methods allows to makelithological division of the section to distinguish ice including very thin ice interbed, to determine the totalice content. Seismic log method also allows to investigate the space between wells and to make soundingof well-surface rock mass. The sounding can be made to obtain three dimensional seismicgeocryological models.Cryology zone depth to 20-30 m is permanently effected bySeason temperature variation. Human activity influences this partof section and this can cause the damage of the natural temperaturcregime. When rocks containing ice and water are higher or lowerthan O’C, cryology ccological situation changes. The major monitoringobjects of cryological ecological situation in cryozone can bedivided into two groups according to the rate of the influence onthe temperature regime of the rocksJhe first group is connectedwith warmth addition which cause the temperature increases tillOT. In contrast, in the second group the temperature decreases todegrees below zero. In the first case ice thaws out in frozen rocks, inthe second case water freezes. The process of transition of temperatureto degrees above or below zero is accompanied by an abruptchange in some physical.and mechanical properties of rocks.The sharp change in seismic electrical properties of friable andfractured rocks, which contain water and ice, are the most interestingfor geophysical methods. This allows to use electrical andseismic methods of geophysical investigation to determine physicalcondition of frozen or thaw rocks, to find out their location inmanydimensional space including time coordinate.In north-east Asia the geophysical methods of geocryologyecological condition monitoring is used to do the following: toidentify taliks and to watch their development during the cxploita.tion of dambs and during the filling of water collecting areas; todiscover the places of water leakage; for open and undergroundmining of natural resources; to search for abandoned mines, for agriculturalworks; for engineering-geocryological investigation; farprediction of possible change in geocryological condition and forsome other purposes. The human influence is accompanied by therock temperature decrease which cause increase in permafrostthickness or the beginning of permafrost. For instance, whenground dambs are being built, there is natural freezing or artificialcooling of their bodies; when show is removed from aerodromesand roads this is accompanied by drop in annual average rock temaperature; the permafrost appears in the places of taliks of drainedlakes, which are used for haymowing.As compared with other gcophysical investigationsgeophysical methods of cryology ecological monitoring have someadvantages, mainly, they don’t have additional influence on thetemperature regime change. For instance, the study of temperatureregime of perennial rock warming, up to its thawing out in thewell’s walls at some depth. To restore temperature regime it takesmuch time during which it is impossible to distinguish bctwccn thestudied cryological changes and the influence of drilling.When wells are used during a long period of time they need tobe properly equipped to be always in the working condition.The practical USE of geophysical methods of cryologyecological monitoring in cryologyzone evidences its effectiveness. Ithelps to make detailed and rapid investigations, it may be moreeconomical than other methods.1222

PERMAFROST IN THE SELENGE RIVER BASIN(ON THE MONGOLIAN TERRITORY)A.SHARKHU,WInstitute of Geography and Geocryology, MongolianMongoliaAcademy of Sciences, Ulaanbaatu,210620,In this paper, the basic results cf permafrost investigation in the Selenge River are considered, includinga territory of more than 300,000 Km'. The study of the thermal region in the rock and ground, dcpendingon different natural factors, serves as a scientific methodical basis for discovering qualitativeand quantitative parameters of permafrost conditions, On the basis of this analysis and field rcsearchsmall scale permafrost map has been complied of the Selenge River Basin, in which basic parameters ofpermafrost distribution, thickness, temperature and composition are shown.INTRODUCTION".The zone of permafrost occurrence embraccs all thc territoryof the Selenge River Basin which is a basic economical region ofMongolia. The circumstances required a comprchcnsive study ofpermafrost in the given region.The first summarized characteristics of Mongolian permafrostwere determined by N.Longid(l969), G.F.Gravis (1974) andN.Sharkhuu (1 975). In the last 20 years, permafrost investigationshave been concentrated by the author in the Selengc River Basin.As a result all materials obtained were generalized and compiled intoa number of large and middle scale geocryological maps of industrialareas, and a small scale map of this Basin.(1978,1982,1989).The main purpose of this paper was to discovcr the gencraland regional regularities of permafrost occurrence on the basis ofanalysis of the changing ground temperature regime depending onvarious factors ornatural conditions.NATURAL CONDITIONS- .-The Selengc River Basin is situated in the central part ofMongolia and is surrounded by the high ridges(2,OOO-3,500 m) ofthe Hubsugul, Hangai and Hentei mountains from the west, southand east. Betwecn these mountains is the Orhon-Selengemiddle-low mountain depression.The basin has a sharp continental climate. Mean annual air' temperature changes from -22°C to -30°C in January and from12OC to 18°C in July. The mean annual value is from -6OC in thehigh mountains to 3'C in the depression.A large part of the Hubsugul and Hentei areas is charactcrizedby a taiga zone and the Hangai area - by a forest steppe zone. Forestcoverage occurs predominantly on the north facing slopes of themountains. Only the south part of the Orhon-Selenge depression ischaracteristic of a steppe zone.GROUND TEMPERATURE REGIME"" .. . ~An important indicator of the ground temperature regime isthe mean annual ground temperature at the level of zero annualamplitude. The penetration depth of zero annual amplitude temperaturesin a large part of the given territory is about 10-15 m.However it decreases to 5-10 m at the swamp sites of valleys andwatersheds and reaches to 15-30 m in the sand - pebbles depositsand in the fissuring solid rocks.The analysis of the.permafrost map sections of this basin hasgiven the quantitative characteristics of general regularities in thechanging of mean annual ground temperatutes. In particular, themean annual ground temperature by the mountain altitudinal beltdecreases about 0.4-0.6"C for each 100 m rise above the absolutesurface height, and by the latitudinal zonation it increases0.9-I.0"C for each 100 kilometers moved from north to south (N.Sharkhuu, 1975).The regional regularities in the changing of ground temperatureregime have been studied in the sites of geocryological investigationssuch as the sites of the Hatgal, Erdenet, Wlaanbaatar andother areas. On the basis of approximate calculations and factoranalysis carried out by using data of geothermal measurements inmore than 450, boreholes with depths from IO m to 200 m, the authorestablished a number of quantitative parameters of changingmean annual ground temperature depending on basic factors ofnatural conditions.As on ground composition and moisture, slopeaspect, vegetation and snow cover, rain infiltration water and otherfactors. Therefore, the mean annual ground temperature in theSelenge River Basin changed within the large ranges of plus 6OC tominus PC, including the predominant ground temperature of the1223 -

magnitude of plus 2'C to minus PC.The value of geothermal gradient at depths exceeding the layersof yearly temperature fluctuations ranges from Ioc to 3°C forevery 100 m. The geothermal gradient in the transition from valleysto watersheds under similar geological and geographical conditionsdecreased by almost 1.5-2.0 times. Therefore, the gradient on thewatersheds and slope of mountains averages 0.01-0.02 deg / m indIn the valley bottoms it is 0.02-0.03 deg/ m. Consistent with this ,the thickest permafrost is observed at the mountain uplands andthe relative cancellation of permafrost thickness is characteristic ofthe valley bottoms.PERMAFROST FEATURES' Permafrost is determined by the features of it's occurrence,thickness, temperature, cryogenic structure and evolution history.These features are closely connected with ground temperature.Especiallyoccurrence, thickness and temperature of permafrost whichhave a direct relationship not only between each other and but withgeneral and regional regularities in the changing of ground temperatureregime( Gcocryological Conditions of The MongolianPeople's Republic, 1974).As the result of the above established regularities, the authorcompiled, in 1982, an engineering geocryological map of theSelenge River Basin on the scale 1:1,000,000. On a generalizedbasis, this map had been made the permafrost map or the map ofgeocryological regionalization of this territory on the scale1:6,000,000, in which are shown only interrelation indicators of occurrence, thickness and temperature of permafrost. (see figure...).This map has three geocryological sections, which is a sufficientlyillustrated altitudinal belt of permafrost occurrence. Besides , moregeneralized characteristics of latitudinal zonation and thealtitudinal belt of permafrost occurrence in the Selenge River Basinis presented in summary tahle I. ~The permafrost map of this Basin is divided into the followinggeocryological two zones and five areas, in particular : the areas'ofcontinuous (>85%) and discontinuous (50-85%) permafrost arecharacteristic of the zone with predominant permafrost and theareas of widespread (lO-50%) and rare spread (1-10%) islandpermafrost and sporadic permafrost are Characteristic of the zonewith predominant seasonal freezing (of thawed grounds.) As wellthc map is subdivided into ten geocryological sub areas (or sites),distinguished,by the magnitude qradations of occurrence and mcanannual temperature of permafrost sequences.The above represented map and table shows clearly that Ihcoccurrence, thickness and temperature of permafrost are submittedto gcncral rcgularitics or to latitudinal zonation and the altitudinalbelt of natural landshafts. So according to the latitudinal belt, witha rise of absolute hight of the earth's surface in the mountains ofthe Alpine type and in the mountain taiga and forest steppe zone ofthe Selenge River Basin there is a regular increase in the continuityand thickness of permafrost and decrease in temperature. Thelatitudinal zonation of permafrost occurrence is characteristic ofthe steppe and forest steppe zone of this Basin.The territory of the Selenge River Basin is charactcrizcd byunextendcd permafrost. Thercforc, thc depth of seasonal groundthawing or thickness of the active layer corresponds to theFig.1 Permafrost map of the Selenge River Basinpermafrost table or to the upper boundary of the permafrost sequence. The average thickness of seasonally thawing or active layersis about 2-3 m.An important feature of permafrost is thc cryogenic structureof permafrost sequence characterized mainly by freezingtype,cryogenic texture and ice content of rock a-nd ground. In general, overwhelming masses of permafrost sequence in the zone withpredominant permafrost are confined to solid rocks and in the zonewith predominant seasonal freezirlg which is confined to loose depositsor grounds.Epigenetic frozen solid (magmatic, sedimentary andmetamorphic ) rock have inherited cryogenic texture, which correspo,nds to the character of rock fracturing and to the type of undergroundwater, in particular ice is arranged in rock fissures, fractures and joint chacks. They are characterized by cold rocks. However,in the upper (0.5-3.0 m) layers of the rock weathering zone,the ice content is 3-10%, until the depth of20-30 m -- 2-5% andlower than 100-150 m it is less than 03%.Predominan,rly epigenetic and partically syngenetic frozenloose (alluvial, lacustrine, glacial and their mixed ) dcposits havecryogenic or ice connection which depends on facial, cryotextureand ice content of loose deposits. In most cases loamy frozenground has layered and nctting cryotexture, sandy frozenground-massive cryotexture and gravelly frozen ground - ring visiblecryotcxturc. One may approximately account that permafrostwith massive and ring cryotextures has a poorly (5-15%) icc con-

Table 1. Characteristics of the latitudinal zonation andaltitudindl belt of the permafrost occurrencepermafrostaltitudinal beltoccurrcncc areas of and latitudinal areasnumberzonationpermafrostoccurrence .1golets and mountainon golets andcontinuousmeadow belt of 11.6 meadows ofpcrmafrostalpine typcmountains2mountaln taiga,rneadowdiscontinuouson all typcand steppc beltoftaiga 85.6permafrostof landshaftand Carcstslcppc zoneWidespread is mountain Corcst in valley bottoms3 lands of and steppe belt 64.5 and on north-facingpermafrost of forcst steppe zone dopes of mountains4mountain-forestrarespreadlandssteppe belt ofof permafrostforest steppe zone93.15mountain steppeon swampsporadicbelt of 15.5 sitcs oroemafroststeppe zonevalleys1-5forest in bottoms vallcyislandstcppc 336.3 and on northpermafrostzonefacing slopesof mountainstent, permafrost with netting cryotexture - a greatly (2040%) icecontent. In general, the Selenge Rivcr is characterized relatively bypoorly and middle ice content Ioosc deposits with predominant ofmassive tcxture. Lacastrine and lacusrine alluvial deposits are usuallyrelated to the more ice content permafrost.At present, a stability of permafrost evolution is characteristicof all the territory of the Sclenge River Rasin. Relic permafrost hasnot discovered here. Howcvcr, on some local sites of the basin degradationand agradation tendencies of permafrost evolution arenoted, which have only local characteristics. In particular, thepermafrost agradation is noted only on some sites situated in thetaiga'zone of the Hentei mountainous region. The permafrost degradationis observed comparatively often in some local sites locatedin the other territories of the Basin. The notcd aggradation anddcgradation or permafrost in the Selenge River Basin is closelyconnccted with a large dynamic evolution of relatively high temperature(minus 0-I@) and low ice content (about 8-15%)permafrost scquences, which is characteristic of territory in thesouthern fringe tone of permaliost occurrence. Especially the largedynamic evolution of permafrost predominantly in the direction ofdegradation is rapidly manifested under economic development ofncw territories in the Basin.PERMAFROST ASSESSMENT.... ~The degrcc of' complicated permafrost conditions of theSclengc River Rasin may be divided into threc ycocryological rc-pions:1. The region of continuous and discontinuous permafrost ischaracterizcd by complicated permafrost conditions for economicdevelopment of terfitory caused by widespread comparatively lowtemperature (minus 1-4'C), large ice content (about 10-40%) andthe thickest (100-200 m and more )Qermafrost sequenks. ..2. The region of widespread and rare spread island permafrostis characterized by middle complicated permafrost conditions foreconomic developmcnt. The main cause of,difficulties for economicdevelopment is a large dynamic evolution of high temperature (minusO-IOC) permafrost sequences.3. The region of sporadic permafrost is characterized by nocomplicated permafrost conditions for economic development becausethe ice content of permafrost sequences is practically absent.CONCLUSIONSThe most rational method of geocryological investigation forstudying the occurrence, thickness, dynamic and other indicators ofpermafrost may be based on the discovery of general and regionalregularities in the formation of the ground temperature regime.A compilation of permafrost maps on small scales is advisableto carry out a united (table form) legend determined by theinterdependent characteristics of occurrence, thickness and temperatureof permafrost i s. as in thc compilation of the pcrmafrost mapof the Selenge River Basin.Further investigations, especially on the dynamic evolution of

permafrost in the Selenge Rivcr Basin situated in the southernfringe zone of permafrost occurcnce, are presented for scientificand practical interest for geocryologists and other specialists. Theymay be at a high scientific level, and carried out only under closecollaboration with scientists-geocryolists from interested countrieswho arc adhering members of the <strong>International</strong> Permafrost Association."- REFERENCES"Geocryological Conditions of The Mongolian People's Republic,1974, Transactions of Joined Soviet-Mongolian Scientific-ResearchExpedition; Vol.10, Moscow, pp30-48,74-91.Sharkhuu, N., Lubsandagva D. and Zhamsran S., 1975, Basic Featuresof Permafrost in Mongolia, Ulaanbaatar,pp22-66,90-103.Sharkhuu N., 1982, Engineering Geocryological Conditions of TheSelenge River Basin, In Proceedings: Materials of VIII-th <strong>Conference</strong>of Young Scientists and Aspirants on Gcocryology ofGeologic Department of Moscow State University by M.V.Lomonosov.Sharkhuu N., 1989, Geocryological Conditions on Territory ofHubsugul Phosphorit Deposits, In Bulletin of MongolianAcademy of Sciences, 4, Ulaanbaatat.

DEFORMATION OF THAWING DISPERSED LARGE DETRITAL ROCKS OF CRYOLITE ZONEShesternyov D.M.Institute of Permafrost, ChitaThe paper is concerned with the composition and structure peculiarities of largedetrital rocks of cryolite zone. Classification diagram on basic structures oflarge detrital rocks has been suggested, which is put into the basia of analysisof deformation changes of thawing dispersed large detrital rocks of cryolitezone. It has been established, that for large detrital rocks with contactcryogenic textures and frame structures cryogenic textures have greater influenceon thawing coefficient values than ice saturation (total moisture) af rocks asa whole, and the curyes of thawing coefficient variations of different genesisrocks are identical.An intensive national-economic exploitationof cryolite zone of platform and mountain foldedregions of Russia and other countries is beingdone during the last decades. According to ourcalculations the cryoliie zone in these regionson the depth of annual temperature variationsconsists of 60-90X of seasonally and perenniallyfrozen large detrital rocks (LD). LD are consideredto be the rocks which contain more than 10%of detrital rocks and minerals, the diameter ofwhich is more than 2 mm.Comparatively few papers are devoted to thestudy of deformation of thawing LD. (Vedernikov,1959; Tsytovich, Kronik, 1973;Ilshkalov, 1974; Votyakov, 1975; Vyalov, 1979:Davidenko, 1981; Ziandirov, Kilbergenov, 1987;et al). In these works some regularities ofthawing coefficient variations (A,) and compressionCoefficient variations (a) for LD withdifferent genesis, composition, structure andproperties are shown, the methods of field andlaboratory investigations are offered: correlationregressionLD models of regional importance arebuilt. However, though the authors have made agreat contribution to the study of thawing LDdeformations it should be noted that someproblems in their works have been paid littleattention to. The main of them is the absenceof profound investigations on the influencelithogene structures (L S L D) and cryogenictextures (C T L D) on deformations of thawingLD. Rut satisfactory solution of the problemcould not have been solved because profoundresearch into structure and cryotexture formationin LD has not yet been done. That is whywe had to pay great attention to the solution ofthis problem when studying the deformation ofthawing LIj*Using methodical approach offered by V,I.Osipov (1965) and taking into account thepeculiarities of LD composition and structurewe shall treat the LD structure in regard tospace distribution, interrelation,>,size andgeometrical form of LD-forming Branulometric* elements and their groups (Table).Classification of Structures ?f LD.Class Group Subgroup/50>KvL10 MixedFrameless (LD partic les,sand, dust andclay parti cles)90>K,klO90>K,~50FrameSimple(only LDparticles)KVr90Sand,clayUnf ullycontact50’KV

"hot stamp uf an area of 5000 cm in pits rindshafts up tu 10 m in dept-h. A t thc same timcdetailed description of the distribut,ion oflarge-detrital particles i n LJI was performcd,the int-errelation of definite I. S 1, D and C ?'L 1)types was revealed, identification of the lattcrwith the A, and "a" coefficient values was done.The regularities of ice inclusion diStrihutionin a finely-dispersed LO component and the characterof ice cuvering of large detrital [)arLicleswere paid attenlion to. The research wasdone by the author sirice 1976 to 1986 under H.A.Kydryavtsev and R.D. Ershov at the geocryologychair of the Moscow State llr~iversity, then hythe author himself or under his direct guidanceat the hydrogeology and e'ngineering geologychair of the Chits Politechnical Institute, andsi.nce 1986 up to now in t.he laboratury of engineeringcryogeodynamics of the Chita Departmentof the Permafrost Institutc of t.he RussianAcademy of Sciences Siberian Department.The processing of thc dat-a obtained has beendone in two 'stages. A t the first stage thranalysis of the change of granulometric L1 cumpositionof differellt genesis was performcd andtheir inf1uenc.e on litho- and (;ryogenic Lnstructures was eslahlished. A t the secunrl stagestandard values KtoL, p,, p (total moisture,density of Lll skeleton and 1.D density i u naturalstatel, A and "a" were established for rockswith different I, S L n and C '1' 1, 0.It made possible 1.11 establish the following:1) nispersity incrc:ise in the series sandsandyloam-luam, fur one subgroup of I,0 st.ruc-, ./tures with LD particle content, practicallyalways leads to the increase of values Wtot;2 ) LD particle content. increase with one andthe same finely dispersed component type reducesWtot and increasr F, PC (there is no considerabledifference bet.ween Wtott 3nd PC for theframe 1,D structures, irrespective of the FCdispersity increase, but we could nor. say thesame ahout fhe r-ange of their variation series).Clearly, the changes of P M P (physicomechanicalproperties) of large detrital rocks,1, S I, r) and C '7 L 1) affect. in a definite waythe formation of A and "a" roefficicnt v~lues.'rhus, fur example, in'layer 3 (Fig.la\, gravelpebbleLO with sandy SC: is charart.erized byframe and unfully contact L S L 0.Similar L S I, n types are typicnl for separatelevels of gravelpcbble I,D, the scctiuns ,of them are shuwn in Fig.lb and LC. Sut C 1' L Dand \Jtot differ greatly from each other inevery concrcte case. Contact-fully crust-massiveC 'I L 4 (W~~)t=0.20-O.I8) are dcveloyrid in therange from 4.0 t u 6.0 m (Fig.la), conLa~tunfullycrust-massive C T L D (Wtnt=0.12-0.08)-i n the range from 5.[) 1.0 8.0 m. In thr firstcase the ice crust thickness un the contactsbetwccn rock dehris equals 1-2 mm, in the s~~cond- 2 -3 mm. In the process of thawing ht'al settlcmcntor rocks A, with less values Wtrrt werc1 .5-2.0 tlnles highc,r than in the rucks w ~ t hgreater values Idtot. Tt. shows that the pussibi-l i t y 1.0 evaluate A, arcording to Wtt,t chongre,even i n the quantitative respcct, is rRt.herproblematic fur 1.n. That thc ahuve-mentioned-a P,"1228

example is not the only one is seen from thedata in Pig.lb and Fig.lc. Tn Fig.lb contactcrust-lentiformC T L D are developed along thesoil profile of gravelpebble LD with loamy FC(layer l), and they differ only in the thicknessof ice crust on rock debris, In the rock profilefrom 9.0 and down and from 3.0 to 4.0 the ice -crust thickness on large detrital particlesequals 1-2 mm, in a layer between these depththe ice crust thickness on rock debris hardlyaverage 1 mm, but on the contacts between themit equals 1-3 mm. That is why A, values arehigher in LO with the same C T I, T) t.yyeu butwith lower values of their '{tot.Geologo-genetic anti regional peculiaritiesof formation greatly influence deformationvalues of thawing larpe detrital rocks of80-cryolite zone. A, dzpendence on large detritalparticles content (K,) for large detrital rocksof different genesis is shown in Fig.2.There were no^ less than 5 separate definitionswhen calculating A, and Wtot in everycase. Generalized curvcs were obtained for LDof diffcrent genesis wit.h sandy (Fig.2a)., sandyloam (Fig.2b) and loamy (Fig.2~) finely dispcrstdcomponents. It follows from the deeendenccsindicated in the figure that when Kv content inLD increases from 0.1 to 0.9, A, values- decreases. In the first part of the segmentKv=0.1-0.5, the rate of the decrease is t4t'highest for all tD genetic types and increasr.~with the increase of FC dispersity. Its largestvalues are typical for. alluvial, fluviolacialand diluvial-soliflual LD deposits: i.c. deposits$-la-18060i404012020100080i2010080604020"6040200.1 0.3 0.5 O?Fig.2 Standard thawing coeific lent v~luc chanu~s [Ao\ of diffcrent genesis large detrilsl ror;hs forseparate regions of Russia. Conditional symhols: l-Udokall range, 2-Vit. im-Patorn plateau, 3-Stanovoyrange, 4-Chulman hollow, 5-Chita-lngoda qullow: ?,(Is, fg, H -correspondingly LD of eluvial,diluvial-soliflual, fluvio-glacial, di luvial arlri alluvial genesis, Yto,:=24-10 - st.~nrlard values oftotal moisture (the first number - when K,=O.55-0.15, t.hcSCCOII~ - whcn Kv=0.85-0.90j.-- 1229 -

that are characterized by comparatively highvalues Wtot. great variations of C T L D, weaklithification or good roundness of largedetrital particles, its least values are typicalfor eluvial and diluvial LD. An exception iseluvial LD having been formed in weatheringgranites of Stanovoy range, the variation rangeWtot (Fig.2a) is similar to variation rangeWtot for diluvial-soliflual LD deposits withloamy FC (Fig.Zb). Tn the segment of Kv from0.5 to 0.9 the rate of A, changes drastically.nut differentiation of its values for LD ofdifferent-genesis is practically invariable.Here the Wtot influence on the rare of A,changes reduces drastically, because though Wtotvalues are practically the same for LD ofdifferent genesis, the differences between 7.are rather great. Tn this connection ourhypothesis about the predominant influence ofthe distribution character of ice inclusions(C T L Dl over Wtot on A. changes becomes, inour opinion, more evident.Thus, based on the results of this study,it is concluded that:-in the segment of ,changes of Kv contentfrom 0.5 to 0.9 1ithogene.structures andcryogenic textures of LT) have predominantinfluence on A, value:-the largest A. values and the highest rateof their decreasing with the Kv increase from0.1 to 0.9 are typical for alluvial, fluvioglacialand diluvial-soliflual LD deposits.-changes of A,=f(Wtot, Kv, Kfc) in thesegment KV=O.1-O.9 irrespective of LD genesisand FC type are descibed by the followingcquat.iont:A. - wtot . KFC (n,K, + "2KFC)where nl, n2 - empirical coefficient; K~c=l-K,bi,e KFC - volume composition FC in LD.The results of the research are tested andput into practice in desitning and constructingof large mini'ng complexes of Eastern Siberia.REFERENCESnavidenko V.P., (1981) Classificat,ion of Complexityof Permafrost-Engineering-GeologicalConditions on the Basis of Setting Evalua~iunin Frozen Large Detrital Grounds Thawing (onthe example of central districts in theMagadan region), In the hook "Thawing Groundsas Strdcture Foundations". -M.: Nedra,pp.72-92.Osipov V.I., (1985) Concept "Ground Structure"in Engineering Geology. In the Journ. Eng.Geology, No.3, -M. Nauka, pp.4-20.Shesternyov D.M., (1986) On the Problem ofC1,assification of Large Detrital RockCryogenic Textures. In the book: Eng. Geol.Investigations in the Permafrost Region.Rlagoveshchensk, pp.331-333,Shesteryov 9.M., Yadrishchensky G.E., (1990)Qock Structure and Properties of iJdokanCryolite Zone. Novosibirsk.: Nauka, p.126.Tsytovich N.A., Kronik Ya. A., (1973) Physicaland Mechanical Properties of Frozen andThawing Large Detrital Grounds. In the book:TT Tnternational Permafrost <strong>Conference</strong>. Rep.and inform., Vo1.4, Yakutsk, pp.52-62.llshkalov V.P., (1974) Building Properties ofPerennially Frozen Rasis and Quick Methodsof their Determining. Novosibirsk: Yauks,p.84.Vcdernikov L.E., (1959) Research into FrozenLarge Tletrital Grounds. Proceedings ofAll-llnion Research Institute-Magadan, Vol. 13,pp.43-267,Votyakov T.N., (1975) Physicomechanical Propertiesof Frozen and Thawing Grounds in Yakutia.Novosibirsk: Nauka, p.2715.Vyalov S.S., (1979) Engineering "ioblems inGeocryological Investigations on sites ofRAM, In the book "Engineering Permafrostology"M.: Nauka, pp.149-156.Ziangirov R.S., Kilbergenov R.G., (1987)Deformability Evaluation of Large detritalGrounds. Engineering Geology, No.3, M.:Nauka, pp.107-109.1230

PALSA FORMATION IN THE DAISETSU MOUNTAINS, JAPANToshio SONE1 and Nobuyuki TAKAHASHI'Institute of Low Temperature Science,Hokkaido University, Sapporo, 060, Japan'Hokkai-Gakuen University, Sapporo, 062, Japan ,Palsas and peat plateaus, in different stages of development, exist in a mire(1,7201~1 a.s.1.) in the Daisetsu Mountains (43'37' N). They consist of peat cover about 1 mthick and permafrost core of sand and silt. Many segregated ice lenses were visible inthe silt layer. The permafrost base is at a depth of 5m. Ground temperature observationsin the palsa indicates that temperature is in equilibrium with the present climate, andthat permafrost can develop under present climatic conditions. On thebasis of two tephralayers in the peat, the palsa was initiated to heave in around A.D.1830.INTRODUCTIONPalsas and peat plateaus are common features in manypermafrost regions. According to the definition of palsas, theyare peaty permafrost mound possessing core of alterationlayers of segregated ice and peat or mineral soil material(N.R.C.C.,1988). They are considered to be the only reliableindicator of the discontinuous permafrost (Brown,1974).However, palsas or palsa-like mounds are recently reportedalso In continuous permafrost (Washburn, 1983a).In Japan, palsas were first discovered in 1986 and theirfOrmS.Si28S and distributions were reported (Takahashi andSone, 1988:Sone et a1.,1988). The authors attempted tomeasure the ground temperature profiles of a palsa andinvestigated its internal structures by drilling. In this paper,we describe the results and discuss the type and age of palsasin the Daisetsu Mountains.ENVIRONMENTAL SETTINGS OF THE DAISETSU MOUNTAINSThe Daisetsu Mountains are located in central Hokkaido,northern Japan. Thay are.cornposed of Pleistocene pyroclasticor lava plateaus with several younger strato-volcanoes. Manykinds of periglaciai phenomena and landforms, such as earthhummocks. sorted polygons, block fields and slopes, and frostcrack polygons have been reported. Discontinuous permafrostis distributed mainly of the windward bare ground above 1650m (Sone.1992). Palsas were dtscovered in a mire. Theyindicate geornorphologicaly the existence of permafrost ataround the lower limit of permafrost in the DaisetsuMountains.The mire at the south of Mt.Hira$a-take IS located at analtitude of 1720 rn(43'36'54"N, 142'54'06"E). on a broad passof the lava plateau between Mt.Chubetsu-dake and Mt.Hakuundake(Figure 1). The length of the mlre is about 650m fromeast to west, and 350m from north to south. The surroundingvegetation of the mire is Pinus pumila community, The mireis characterited by palsas, peat plateaus and string bogs. Inthe Daisetsu Mountains. palsas and peat plateaus develop onlyin this mire.Figure1 Location of the palsa mire in the Daisetsu MountainsThe elevation of the timber line around the mire Is about1,500 - 1.600m a.S.1. The mean annual air temperature, and thefreezing and thawing indexes at the mire are estimated to be -2.0'C, 2,OOo"Cdays and 1 ,50OoCdays respectively (Sone,l992).1231

2PAULAND PEAT PLATEAUS IN THF MIRE AT M E SOUTH OFKSIIRAGATAKEAbout twenty palsas and peat plateaus exist at thislocation with SiZBS ranging from 4 to 80 m in diameter andfrom 0.2 to 1 m in hoight. Their plan figure is generallycircular ar elliptic. Mostare poat plateaus in morphologicalclassiflcation, and mineral.cored palsas in structuralclasslfication. These are characterized by flat upper surfaceand relatively Steep side slopes, sometimes accompanied withponds just around thern(Figure 2). Permafrost is observedonly under the palsas in the mirs. Severe westerly winds inwinter sweep away the snow deposit of the ground surface,and the snow depth of the mire is less than Im. The top ofpalsas is often exposed partly during awhole winter season(Takahashf and Sone,1988)..For an Investigation of internal structures and monitoringthe ground temperature. a palsa was chosen. Palsa 6. namedby Takahashi and Sone(1988), is 80cm high and 1 Om long,elongated in north-south direction. The palsa is composed ofa peat layer of 60 - IlOcm thick underlain by silty sand andgravel layer. The permafrost table is about 70cm deep and wasnearly parallel to the palsa's surface in late September. Thepermatrost table is nearly cotncident with the boundarybetween the peat and the silty sand and gravel layers. Thesnow cover thickness was measured as much as 40- 60 cmdeep or! the east and west side of palsa on 29th December,1986. However, the snow was less than 40 cm deep on the topof palss D, and some parts of palsa were exposed. A small pondon the south part of palsa indicates that part of this featurehas begun to decay.BORING CORE ANALYSIS OF PALSA BDrilling of the top of the palsa was attempted in late June,1988 and 1989. Core samples with a diameter of 4.5 ctn wereollected from the ground surface to a depth of 423 crn foranalysis of stratigraphy, water content, bulk density andignition loss tn 1988. The depth of permafrost base wasobserved to be 500 cm in 1989.Figure 3 shows the internal structure of palsa 6. It's topwas peat with a thickness of 80 cm, the lower part of whichcontained sand and gravel. The peat layer was underlain bysand and gravel with a silty matrix (80- 170 cm). At thedepth of 150 cm. a thin ice lens was visible. The sand andgravel layer was underlain by silt with sand and gravel (170- 220 cm). Thin ice lenses (1 ~ rnm) were found in this siltlayer. This silt laye: was underlain by sand and gravel withsilty matrix (220 - 250 cm). From the depth of 250 cm, siltlayer extended to the bottom of the core sample. This lowersilt layer includes alternations of ice-lenses and frozen soillayers indicating rhythmic ice-lens formatlon. The thicknessof the ice-lenses in this silt was 1 - 3 cm. These ice layerswere corkentrated in the lower part at.the depth of 320 cmfrom the ground surface.0 500 1000 0 1.0 '2.0 0 50 100WATER CONTENTIWt %) BULK IGNI'IION@ENSITYlg/cm31 LOSS(Wt %IFlgure 3 Water content, bulk density and ignition loss of the blring core of palSaB

Care samples were transported to a low temporaturelaboratory In frozen state. Each sample was cut and sectionedwith thickness of 3 - 7 cm. In analysis of bulk density, eachsectioned sample was welghed in air and In kerosene at -10°Cin the laboratory. After drying for 2 hours at a temperature ofll0"C. the alr-dried weight was measured. The results of theanalyses are alsa shown In Flgure 3.The water content of the upper peat layer was more than200 %,while those of the mineral layers down to 2.6 m werelower than 110 %.The high water content of the upper peatlayer, does not, however, indicate the richness of ice. Sinceorganic materials are lighter than the mineral ones, theformer show higher water content of a percentage dry weightbasis, In fact, hlgh ignition loss of the peat layer Indicates ahigh organic content in the upper peat layer than the lowerpart of the core. The bulk density of the peat layer was lowerthan 1.0, while that of mineral layers until 2.6 m deep fromthe ground surface was higher than 1.5.Water content in the lower sllt layer was variable, partlyvery hlgh and partly low: 438.3 %, 891.2 %and 624.8 96 atdQpthS of 341 -345 cm, 345-351.5 cm and 398.5-404. 5 cmrespectively, while 60 %, 63.9 %and 69.4 % at depths of285.5-291 cm, 310.5-361 cm and 367-371.5 cm respectively.The bulk denslty was also variable in the lower silt layer.A high percentage of water content and relatively low bulkdenslty suggest that the mineral samples are rich in icelenses. Silt is usually very frost-susceptible. In fact, the icelonses or layers develop mostly in silt layors, eepeclally inthe lower silt layer below the depth of 320 cm, where'theyconsist of pure ice.These ice-lens layers are surely formed not by ice injectionbut by ice segregation, because they are parallel to thefreezing surface and alternate with frozen silt layers.THERMAL -- REGIME IN PALSA EThe authors carrled out a continuous monitoring of groundtemperature at the top of palsa €3, where the ground surface isoften exposed in winter. As a result, nearly year-round groundtemperature data were obtained.(A) TEMPERATURE(.C) (B) TEMPERATURE('C)-30 -20 10 0 10 20Figure 4 (A) Every fivo day profiles of mean daily groundtemperature (E>) Maximum, minimum and mean groundtemperatures at palsa B from September 1987 to SeptemberI 9aaOn palsa B, the temperature sensors (Pt 100) were insertedin another bore hole at the depths of 0 cm, 30 cm. 70 crn;100cm and 150 cm. Ground temperature records were obtained atevery two-hour interval from September 14, 1987 toSepternbar 21, 1988 except for 10 days in June. Figure 4Ashows every five day profiles of mean daily groundtemperature for this duration. The temperature ranges andmean values of mean daily ground temperatures at eachposition of palsa are indicated in Figure 48.The seasonal fluctuation of alr temperature affected theground temperature variatlon at each position. From October,toDecember in 1987, the ground temperatures of the paisa atdepths of 70.100 and 150 cm were neably constant, and veryclose to the freezing point; these periods should correspond tothe "zero curtain" at each depth. The ground temperatureremained below 0°C at below a depth of 150 cm. The depth ofthe permafrost table is estimated to be 105 cm on the basis ofFigure 40. The temperature measurement of palsa indicatesthat the temperature of the upper portion on the ground isadjusted to the present climate, and that permafrost candevelop at sites with thin snow cover, even under presentclimatic conditions in the mire.The thermal properties of peat contribute to thepreservation of permafrost (Brown and Pewe, 1973). Theannual mean ground temperatures are calculated to be -3.5"C,-3.1 and -2.9 at 70 cm, IOOCm and 150 cm deep respectively(Figure 48). The mean annual temperature of the groundsurface of palsa is calculated as to be ;2.6'C. while themean annual temperature of the ground at depths of 70, 100and 150 cm are lower than that of the ground surface. Thisfact reveals that the thermal properties of peat affect thetemperature profiles at this site.DSAND AGE OF PALS.A FORMATIONWashburn (1Y83b) proposed that palsas comprise twodifferent forms: one is an aggravation form due to frostheaving by growth of ice; the other is degradation form due tothe disintegration of an extensive peaty deposit. The twotypes of palsas are difficult to identify by shape (Washburn,1983b). Whlle the degradatlonal type indicates thermokarstprocesses, the aggrzdatlonal ono indicates the growth ofpermafrost undorneath.On the basis of the comparisons of air photos taken in1955, 1966, 1971, 1978 and 1982, the changes in site andareal extent of palsas and peat plateaus during the period of1955-82 were examined (Takahashi and Sone, 1988). Theirtotal area was reduced by 36 %over 27 years. However, thereare several palsas that were hardly reduced.On the contrary, a palsa appeared on the photograph takenafter 1971. It was not visible on the photographs taken in1955 and1966. This indicates That the paisa grew after 1971,In addition, another palsa. which was discovered in the fieldsurvey, seems to be iwan early stage of devetoprnent. Hence,some palsas seem to be In the growth stage.While string bogs around the palsa develop in harmony withthe present slope direction, traced on the string bogs remainon the surface on palsa 8. Therefore, the palsa began to growafter the string bogs were formed. This shows that the palsawas upheaved recently by permafrost aggradation, becausethere is no permafrost underneath the string bogs, developingat lower places than surroundings. The palsa contains thesegregated ice layers in the lower silt layer, and the groundtemperature profiles of the paisa indicate that at least theupper part of the palsa was affected the present seasonalalternatlGns of air temperature. Therefore, ice-lens layers ofpalsa can bo formed under present climatic conditions.. 1233

ACKNOWLEDGMENTSxcrophilous pcathydrophilous pcatTa-a ashKO-cz ashThe authors wish to thank ProJessors M. Fukuda and Y.Ono(Hokkaido University) and Dr.M.Sumita (GEOMAR, KielUniverslty) for their suggestions and her instructions ontephra layers. They are also indebted to Mr.Y.Watanabe(Photographer), Mr.N.Yasuda (Curator, Sounkyo Museum),Mr.H.Daimaru, Mr.T.Sato, Mr.H.Muto, Mr. K.Yoshikawa,MrONagaoka, Mr.DtOgawa, Mr.S.lshimaru, Mr.A.Ozawa (graduatestudents, Hokkaido University, at that time) for theirencouragements and assistances in the field.Figure 5 Columnar section of surfacial peat layer of palsa BThe peat layer in palsa B consists of two types of peat;xerophilous peat in tho upper part and hydrophilous peat in thelower part (Figure 5). Two tephra (volcanic ash) layers areembedded in the peat layer at depths of10,0-12.5 cm and 14.0-15.0 cm from the surface. These layers are "Tarumai-a ash(Ta-a) and "Komagatake-c2 ash" (KO-cP), deposited in A.D.1739and A.D.1694 respectively (Endo et a1.,1989). The boundarybQtWeen the xerophilous and hydrophilous peat is at a depth of6 crn from the surface. On the basis of the rate of peataccumulation between the tephra layers, the hydrophilous peatceased to accumulate in A.D.1829. The change from thehydrophilous to the xorophilous peat indicates anenvironmental change from wet conditions to dry conditions atthe site. Therefore, it is considered that the upheaval of palsabegan in around A.13.1830.The authors conclude that some palsas and peat plateaushave been formed under present climatic conditionsassociated with permafrost aggradation in the DaisetsuMountains.A considerabla thickness of peat is required for palsaformation at thls location, because the thermal properties ofpeat contribute to the formation of a palsa (Zoltai andTarnocai,l971). According to Seppala (1988), the minimumthickness of insulating peat layer, nlwded for paisa formationin a region with a mean annual air temperature of -2.O"C, isabout 40 cm. In the mire, a peat cover of more than 60 cmthick is necessary for the growth and persistence of palsa(Takahashi,l990). The14Cage of the peat samples at 40 cmbelow Ko-cl layer (A.D.1694), was measured at 4,520*130B.P.(NUTA-455) (Takahashi et al.,l988). The mean rate of peataccumulation iT! the mlre Is calculated to be about 0.1mmlyear. The thickness of the peat of palsa is 80 cm.According to the mean rate of peat accumulation, 6,000 yearsare required to accumulate the peat thicker than 60cm, in thethermal condition for peat glowing is constant. The peataccumulation In the mire of the Daisetsu Mountains beganaround 7,000B.P. (Takahashi, 1990). The beginnlng of the palsaformation In this mire, therefore, is calculated to be after1,000 B.P., though more data are required for thereconstructlon of the developmental history of palsas and peatplateaus in the mire.From a view of the internal structure, palsas and peatplateaus in this location are not easy to grow in their earlystageof development. Because frost-susceptible silt, inwhich ice-lenses grow effectively, is under 2.5 m deep from?he ground surface, the paisas and peat plateaus do not beginto heave untll freeztng penetrates to the silt layer,REFERENCESBrown, R.J.E.(1974) Some aspect of airphoto interpretation ofpermafrost in Canada. National Research Council of Canada,Divlsion of pullding Research, Technlcal Paper, 409,20p.Brown, R.J.E. and PbwB. T.L. (1973) Distribution of permafrostin north America and its relationship to the environment:Areview,..l963-1973. Permafrost North AmerlcanContribution, Second <strong>International</strong> Permafrost <strong>Conference</strong>,Yakutsk, USSR, National Academy of Science Publication2115, 71-100.Endo, K., Sumita, M. and Uno, R. (1989) Late-Holocene tephrasequences in eastern Hokkaido and their source volcanoes.Journal of Geography, 98, 506-510.National Research Council of Canada (1988) Glossary ofpermafrost and related ground-ice terms: TechnicalMemorandum No.142, p.156.Seppala M. (1988) Palsas and related forms. In "Advances Inperiglaclal geomorphology" edlted by Clark M.J., 247-278,John Wiley 8 Sons Lid.Sone, T.(1992) Permafrost environment of the DaisetsuMountains, Hokkaldo, Japan. P.P.P., 3, 235-240.Sone, T., Takahashi, N. and Fukuda, M. (1988) Alpine permafrostoccurrence at Mt.Talsetsu, central Hokkaido, in northernJapan, Permafrost; Proceedings vol.1, Fifth <strong>International</strong><strong>Conference</strong> on Permafrost in Trondheim, Norway, August1988, Tapir Publishers, Trondheim, 253-258.Takahashi N. (1990) Environmental-geomorphological study onthe Holocene mire development in the DaisetsuzanMountalns, central Hokkaido, northern Japan. Environ.Sci.,Hokkaido University, 13(1), 93-156.Takahashi, N.and Sone, T. (1988) Palsas in the DaisetsuzanMountains, central Hokkaido, Japan. Geographical Review ofJapan, 61, 665-684.Takahashi, N., Nakamura,T., Sone, T. and Igarashi, Y. (1 988) 14Cdates around the bottom of peat layer in the bogs on theDaisetsuzan Mountains, central Hokkaido, Japan. TheQuaternary Research, 27, 39-41,Washburn, A.L. (1983a)Palsas and qontinuous permafrost,Permafrost, Forth Inter?. Conf. Proc.. Nat. Acad. Press,Washington, D.C., 1372-1377.Washburn. A.L. (1983b) What is a palsa ?. Akad. Wiss. GottingenAbh., Math.-Phys. KI. Folge 3, 35, 34-47.Zoltai. S.C. and Tarnocal, G. (1871) Property of a wooden palsain northern Manitoba. Arctlc and Alpine Research, 3, 115-129.'1234 -

A STUDY ON CHARACTERISTICS OF ICE-DAMAGE ANDPRVENTION OF HYDRAULIC PROJECTS IN NORTH CHINASu Shtng hi and Zhang TiehuaTian Watat Reoourccs and Conscnjancy Bureau-heed OD investigation and analysis, thiu paper introduces th6 characteristics of ice damage, the effect ofiw damage on hydraulic projects, the counter measure and design for preventing icc damage, and someobtained result8 in the North China. Thaai prcscntntions could contribute to a reference of assuring asafe operation and denim of hydraulic projbets.INTRODUCTIONTht North China is about 0.32 million square killometers inuta. It is close to the Bohai ICP on the eaat, naar by Taihang moun-Win on the west, attained to Yellow river on tha south and againstthe her Mongolia platenu on the notth. It bohp to thelemi-dry rnonqoon climata of tefipcruc =ne. In winter, the (1~tkg north wind runs cumntly, and along aeaohore, river and mr-voir etc. the hydraulic projseto have auffcmd from acrioua ice damageand brought about certain effect on the living of people.In North China, Northeast and Northwest area, there are differentdcgrce processes of freezing-thawing in water areas of river,mcrviors etc. in winter. In Northeast area, there are a longer freezeand a thicker ice layer, such as, the thickness of ice layer is1.8-2.8m in Heilongjang river and the frcczc is 180 days. or so.The characteristic of ice damage is that the damage of frorcn soil ismore serious than thrt of the running ice. Northwest area has ashorter freeze bccause of the effect of alpine air flow, BO the periodof running ice is longer before freezing. The characteristic of ice' damage ia that the damage of running ice is more serious than thatof frozea soil. In region of North China, the freeze is about 120days. Thickness of ice layer is 0.3-0.5m and the maximum thicknessof icc layer is 1.2m. The characteristic is double damagcs offrozen soil and running icc:The characteristic shows that the icedamage in North China is more serious than that in Northeast andNorthwest ma.Tianjn is a typical region of ice damage in North china . Bocause of affections of dynamic, static and frictional force etc produdby ice layer and ninning ice during the process offreezing-thawing in winter, the hydraulic project along river andseashore etc. suffered from serious damage.According to records of Feb. and March in 1936,1947,1957,1969 and 1977, the coart of Bohai sea bay on the north-ern region was formed continuous icc layer. This caused many facilities(e& ship and platform of petroleum ctc.) suffered from seriousdamage. ,.In water arcas of river, canal and rescrvior etc, because of iceresistance produced during the process of ict growing, there are affectionof static ice pressure after ice tha wing and affection of dynamiclash and friction when ice thawing. particularly, there is anaffection Qf obstruction during the period of running ice in spring.So, thcse actions will bring violent hit to bucpier of bridge andagueduct and result in serious loss of projxts.In short, the damage of ice, and hydraulic projxts arc fairly seriousin winter in North China. Therefore, we must investigate andanalyzc eeriousiy conditions, sum up these experiences and take efficientmcnsurcs to ensure a safe operation of hydraulic project.THE EFFECT OF ICE DAMAGE IN WINTERThe Obstruction of Canal and Trashrack Produced by IceAlong hydraulic project line, there are trashracks, which arccomposed of steel belts weldded and has a interval of 7-8cm betweenbelts, in the front pool of every pump itation. Since the thinice is unsteady and broken easily in early winter, so, thcsc hrokcnices flow to the front of trashrack with water and mass to form II icestack besides a fraction of broken ice ipto front pool throughtrashrack. Meanwhile, the bottom of ice stack dives with rivers andnestles closely against trashrack. This results in the rcduction rfflowsection and decrease of water level. So, it often causes accidentof cuttinl off water when water level of the front pool drops dnanto critical submerged water level of.the flow entrace. When icc laverof canal thawes in the yeany middle of Febrary, there will be manyrunning ice of different size massing in front of trashrack: In thisstage. since the body of running ice is thickness in size, softer in tex-1235

ture and bigger in buoyancy, so they do not dive generally and notbring about the the phenomenon of ice stack also. But when theymeet the pile of grass in front of trashrack, they will form a "mixtureof ice-grass". This will not only result in reduction of flo,w, butalso form an ice-dam in front of trashrack. During proecss of therunning ice advancing in open canal in spring, if there is ice stack toresist flow in front of the backward siphon tunnel. This will causeincrease of water level of upoer reaches and form ice flood. -Damage Under Static Ice Pressure of Contionuous CloyIce Cover- , . . . ". . ", -In early winter, the ice cover becomes more stable with thechange of air temperature in day and night, the expansion of icecover takes on periodic change. Meanwhile, the static ice pressurealso increases with increase of expansion. vice versa; In thawingstage of ice, the expansion of ice layer increases continuously withgradual increase of air tempcrature. The value of static ice pressure /which put itself brokenness of ice layer as a maximum limit relateswith thickness of ice layer, boundary conditions, allowabledeformation and rigidity. That is, the thicker the ice layer, the biggerthe boundary rigidity, the shorter the affecting, span, and thehigher the air temperature, the bigger the static ice pressure.Otherwise, it is oppsite. So, the damages of different degrees occuron the bank-wall of front pool, both side slopes of canal, slope ofreservior and wall of anti-wave along Yinluan line.Since the slope of canal or reservior is pushed by the static icepressure or'ice layre, so, the slope which has slope of 1:3 anduneven surface can be damaged. Its law of damage is that slopeprotection which has slope protection built by laying stones isdamaged and surface of dyke without slope protection is damaged.Position of damape generally occurs onto sun slope or corner ofdyke. In thawing stage of ice layer, the temperature is different atday and night. At night, the ice layer contracts forming continuousor discontinuous crack, then is filled with water forming a new contlnuousice layer. From 8 to 14 o'clock on the next day, with thein.:rease or air tcmpreature, the ice layer expands and climbs upintermittently along sun slope of dyke,. As a result of climbing, thepush force of climbing ice is produced. This push force can bringabout serious damage IO the wall of anti-wave or sub-cofferdam.For cutting offpassagewavwhich the ice layer climbed up, thethawing zone was cleared out by artificial way to protect'the wall ofanti-wave, but the cost of artificial clearing needs 6000 Yuan everyyear.Because of continuous action of static ice pressure of ice layer,the front pool of pump station often suffered from damaged. Suchas the front pools of pump station in Yinluan project which arecomposed of two sidc upright walls of pump room and canal of enteringwater, the damaged place was mainly on the two side uprightwalls of sheet pile of reinforced concrete. Under static ice pressure,the expansion joint of sheet pile wall was moved 3-5cm. On thecontact place of ice layer and the side wall, the epidermis of con-crete fell off and had a devoloping rate of I-2mm every year.COUNTERMhASURE FOR AVOIDING ICE DAMAGETake the project of Diverting Water from Luan River toTianjnDWLT project)as an example. In order to enssure a normaloperation of water transport in winter, electric heat conductingpipe, high water pressure pipe and high air pressure pipe were stUP on edges of sluice gate respectively. It was confirmed that beforea continubus ice layer was forrned,it was effective to disturb thcfreeting water surface not to form a frozen area by successivepasses of electricity, warter and air, except for a greater consumption of power. Once the ice layer was formed, it would be very dimcultto melt the ice using the method.According to lilrious features in DWLT project, the followingmethods were proved to be effective.Countermeasure for Preventing Runing Ice From Blocking up Ca-"nal and Trashrack.r-The failure of icC layet is the reason that the trashrack isblocked up 'in &rry wih't'er. TherCfore, oh6 of the following threemethods may be effective: avoiding' freezing in 'front of trashrack;transportink water at a 'sped of less thad0.67m / s, at which icclayer can't be brokdn; trad6ySbrting dater normally afer temporarystop in order to form a continuous ice layer.The key to prevent trashrack from blocked up is to avoid theformation ofice-grass mixture in front of the trashrack in spring.In order to clear out the ice-grass mixture, the grass grown in winterin front of the trashrack has to be cleared out. As long as nograss exists, the runing ice floaded on water wmld'rnelt with thewarming of climate in spring.In the aspect of the blocked up siphon with runing ice, afterobservatios fori ten siphons; only the siphob in Bcijn drainage riverwas blocked up in 1984. It is considercd that the solution is to renewthe' iceproof pier at the entrace of tunnel. Meanwhile;trashrack should be settled within IOm outside the entrace oftunnel. The trashrack should be able to intercept the runing icewhich has"r width of a quacer of the width of tunnel,,resist thepunching of runing ice, and split runing ice.Countermeasure for Preventing The Ice Damage on Slopes of Canaland Reservior CofferdmThe k10pes of canal and reservior cofferdam, where the sunshines heavily on the North, are pushed off by ice layer. Accordingto obsentation of ground coffer in the North China, the limit ofslope without dope protection is 1:5, that is, when the Slope isgreater than 15, the slope would be pushed off. The slow whichhas a siope of 1:2 was most heavily pushed off by icelayer, and amound was pushed up by about 1-2~1.The limit of slope, which has a smooth-faced stone lining is1:3. Th&slopc which has a slope of )mort than 1:3. an uneven surfa&.or B unsteady base, would be suffered frqm failure in differentdegrees. Therefore, in the future design, slope protection must beaccepted. and it is better to use reinforced concrete or COnCerteslope protection rather than stone. Fabric filteT layer beneath theslope protection is. the best, The surface of slope protection shouldbe smooth as possible as it could be.. Considering the unhinderedclimbing up of the interface Ween ice layer and dop protection,safe slope should be adapted. Is possible, slippy coating can bepainted on the surfacc of lining or slope protection, SO that the icelayer can climb easily. This is an effective measure to reduce theslope protections of canal and reservior cofferdam.1236

If.anti-wave wall, sub-cofferdam and other constructions arcsit up on the top of the cofferdam, the capacity to resist thecompressive strength of ice layer must be calculated. The actingpoint should be checked at 1 / 3 or I /.2 of the height of the wall.This is an effective measure to controll tile anti-wave wall on theslope protections of canal and rcservior cofferdam not to bc damagedbv the compressing stress of static ice layer.Countermeasure for Reducing the Static.!ce ,Pressen the Front. " ."" .--"" -...Pool of Pump Station. ." . ""..Aker continuous ice layer is formed in the front pool in winter,the static ice pressure would be acting from stable stage to meltingstage. The ice pressure is defincd as the combined stress of thecompressive strength of the maximum thickness of ice layer. If theside wall has a greater stability and rigidity, the ice pressure is statedin accordance with the fracture or heave of ice layre. If the capacityto resist static ice pressure is lower, the ice prcssurc on theside slope wall is stated in accordance with its deformation orunsteady.Su shengkui's formula is recommended for the calculation ofstatic ice pressure. Based on icefield stability thocry, the calculatingmethod for static ice pressure is illustrated as follows:'Combining practice with the concept of lever stability in themechanics of material, authors proposed the icefield stabilitythocry in accordance with the research of the maximum horizontalpressure on hydraulic buildings at the moment that the ice layer becomesunsteady. Basic assumptions are as follows:Ice layer is elastic and istropic material, and its stress-strainbehaviour is satisfied with Hooke's law.The compressing stress reaches to its maximum value at therncrnent that the icelayer changes to unsteady.Connection between ice layer and building is regarded ashinp,d joiht.&sed on above assumptions, before the compressed lever isout ol'stability, the differential cqution of the bending deformationol'the let-er in the axial direction is:EIY"P,Y = 0 (1)where: pptatic ice pressure(T / m');Y:vertical deformation of ice layer(shown in Fig.1);E1:rigidity of ice layer.Solving the above equation .with a hinged joint boundarycoundary condition, we obtain:p =- z2 EZ (2)l2Where, 1 is the distance from building to the opposite bank ofice layer. Generally, the value of 1 is selected as 15 times of the netwidth in calculation, but it should not be greater than 150m.If the thickness of ice layer h and the calculating width of theextruding of icc on building b are obtained, equation (2) can betransformed to:Where, 1: the distance from' front pool to opposite side wall; E:elastic modulus of ice, 88-98 (T/ m2); bl.Om; h:thichncss of ioclayer (m).In order to diminish the action of static ice pressure on the sidewall in the front pool of Yinluan Pump Station, it turns out to beeffective that water is pumped from bottom by dive pump andsprayed out from a guiding pipe on the surface of front pool, 80that a artificial spring is formed, which keeps water from freezingand a voids the action of static ice pressure. This kind of method isusedin a few hydraulic buildings, to eliminate the action of icepressure, for instance, Erwangzhuang reservior, canal and frontpool, At the same time, the natural surounding around pump stationis improved.CONCLUSION^wIn short, the prevention of ice damage for a long water transportis an important subject to ensure safe operation in winter. Aftertransporting water of 8 winters, it is proved that the original design is still not perfect. In addition to popularize above successfulcountermeasures, management and scientific research should bedevoloped; field observation carried out; better combined measurestudied; and true operation run, so as to ensure absolute safty oftransportion of water in wintir.REFERENCE. "-Su Shengkui, 1980. Action of Ice Load on Hydraulic Buildings.Water Transportion Engineering, No 5.Xu Bomen, 1985. Inflate Pressure and Calculation of Ice Layer inReservior. Hydraulic Electric Tcchniquc. No Y5.Su-Shengkui, 1980. Problems of Analysis and Design of Anti-iceDamage in Yongdinxin River. Hydraulic and Electric Technique.No6.Snib, B, 1982. Load and Effect of Water Wave, Iw and Ship onHydraulic buildings. Haihe River Water Cconservancy (Supplement). Nol.Xu-Jianfeng and Xu Caihua, 1986. Ice Damage and Prevention inYollow River. China Water Conservancy. No3.Ice fieldFig. 1 Ice field calculation sketch1237

THE LATEST PLEISTOCENE CRYOMERE IN THE REGION OF "KOPJES" AND THEBIG MESETAS, PATAGONIA, ARGENTINADario Trombottol & Bernd Stein2Centro Nacional Patagbnico (Cenpat), Boulevard Brown 3000, 9120Puer to Madryn, Ch., Argentina.Universitat Bamberg, ,Physische Geographie, Am Kranen 1 D-8600Bamberg, Germany.In the northern area of south Patagonia, near - he 46' S. L. and atdifferent sites we find clear cryogenic \ structu es that prove the,activity of freezing and thawing of the latest cryomere during thePleistocene in Patagonia. The profiles show ice wedge-casts. At theflanking sides of the "Pampa del Castillo" the head penetrates intothe Tertiary forming droplike subhorizontal involutions. inter- Anmediate clay-layer could indicate a possible interstadial. Paleoclimaticand sedimentary indicators express conditions of coldnesswhich aIe ameliorating towards the Holocene and Age a predominantlydry environment with a strong participation of the wind.7The paleoclimatic reconstruction ofthe Quaternary Patagonia, or the.Neogene, has brought up many questionsand different hypothesis. The idea ofenoEmous glaciations during the bigcryomeres contrasts with r,atherrestricted glaciations near theCordillera or minor local"glacioblastos" (GROEBER, 1950, etc)associated to the big mesetas orPatagonian mountains of over 2000 ma.s.1. The Patagonian glaciation helpedto explain theories about the origin anddispersion of the famous "RodadosPatagdnicos" and certain hypothesis ongeomorphology. Until now it is generallyassumed that the glaciation coveredmainly the "Cordillera Austral1' reachingpartly into the extense Patagonian area.In some cases the glaciatioqs approached70' W.L. but it: is only neax 52' S .L.where they reach the Atlantic Ocean(CALDENIUS, 1932, etc). Step by step newinformation has been achieved and somecold periods in the past have alreadybeen documented. Geochronometric datingstold as about old glacials towards thelimit of the Messinian/Tortonian(MERCER, 1985) .During all these years the pastgeocryology of the area has beenneglected. CALDENIUS (1940) holds thatthe fluvial deposits of the Eormation ofthe "Rodados Patag6nicos", embedded withfine sediments or a kind of "loess",were influenced by'solifluction and heundeIlines an important denudation cyclein East Patagonia. CZAJKA (1955)mentions rests of structures caused bythermal contraction during thePleistocene at the "Pampa del Castillo"and on the 'IRodados Patag6nicos". Hesupposes that permafrq'st must havereached at least the mouth of the riverRio Negro. Ice wedge-cast's were placedat SE of the locality "Las Heras"(CORTE, ,1982). The important coldperiods however, left their traces offreezing and thawing through the courseof time. buring the Quaternary cryomeresand possibly even earlier the biggestpart if not the whole area was coveredby permafxost.Periglacial phenomena imprinted theirmarks during the Pliacene-Pleistocenewhen the IIRodados Patagbnicos" (shol tRP) were depositing and also aposteriori. It analyses indicators ofthe Late Glacial and proves relevantclimato-stratigraphical changes of theenvironmental conditions of selectedsites before the arid or semiaridcondi tons of Patagonia today.srunYBBEBThe study area corxesponds to thetypical Patagonian landscape: bigmesetas , such as "Pampa del Castillo"(fig. 1) which have been classified intodifferent levels accordi.ng to theirheight and degree of erosion. Apart from* 1238 .

Fig. 1: Location of the Study sites in Patagoniathe continental mesetas and approaching The absolute min. temp. registered inthe coast we find mesetas of marine Sarmiento is -33' C (incomplete data) .origin. The mesetas, surrounded by bad The absolute min. temp. for PeritoLands, generally dry valleys' and Moreno is -17,5' C and the mean annual"bajos", together with mountains, table- precipitation is 116 mm (CABRERA, 1976) .tops or Itkopjes1' and of course frequent Towards the Southern Andes cool..and different Neogene basaltic layers temperatures and their influence gainconstitute the Patagonian table land of more and more importance.the -study area.The study area is characterized byThe analysed profiles are situated at luvic yermosols in two typical layers.three characteristic sites close to the The upper layer consists of sand.and46 0 S.L. (fig. 1) where excellent silt mixed with gravel (-RP)) of aexamples of ice wedge-casts to,gether greyish-brown colour. Below we find awith other characteristics -of a thin chestnut-coloured clay-layer.periglacial paleoenvironment are found. The vegetation belongs to theThe sites axe : "Holdich", "Kensel" and "provincia patagonica" (CABRERA, 1976) ."Las Heras". The first site is located Thorny plants and species that grow inin the SE of Chubut and related to the form of "cushions" adapted to an aridflanking sides oE the geoform called and windy environmgnt prevail."Pampa del Castillo" (appr. 700 ma. s,. 1. . Site "Kensel", near "Ce~roKensel", is located in the NW of Sta.Cruz (appr. 720 m a.s.l'.), east of 'the"Meseta del Genguel" on the nat. road no40. The third site is located in the NEof Sta. Cruz, at a height of 350 ma. s. l., near the settlement "Las Heras".Here the level of tesraces is lower thanat site "Ho1dic.h". According to TROLLand PAFFEN (1969) the climate belongs tozone 111, e.g. cool temperate, type No12: steppe climate: sernidesertic anddesertic climate.AGeologically the study area belongs to"Chubut extraandino", north of the"Nesocraton del Deseado" . WINDHAUSEN(1924) carried out a very importantgeological and topographical investigationin the area of golfo San Jorge".Quarternary geology and stratigraphyhowevex, are relatively unknown.In the whole study area a depositionOf evaporites on top of the RP, volcanicashes and eolian sediments which wereretransported and then modified byfluvial, cryogenic and sedimentaryagents, can be observed.1239

During the expedition different studytranssects in EW and NS direction weretraced in order to investigate theNeogene stratigraphy and drawgeomorphological conclusions. We madeprofiles at different sites whichrepresented cryogenic phenomena of the'past in and exemplary way and alsoproved to be related to those levelsexposed to paleoclimatic changes. Theprofiles were layed out along roads, inquarries or excavations of wastedeposits. Whenever this was impossible'drillings and excavations were made. Foroux analyses we applied classicalmethods as well as visua.1 and tactilecharacteristics. RepresenCative samplesof important layers and levels weretaken. In the laboratory mainlygranulometric analy'ses and analyses ofcarbonates were realized and wherepossible the sand-quar tz grain sur facetextures were analysed in order tosupport the paleoenviromental genesis.DISCUSBIQNWRESULTSAt site "Holdich" we foundpseudomorphs of ice wedges (fig. 1)which penetrate up to 90 cm into themarine Oligocene at the flanking sidesof the "Pampa del Castillo". Theinterior of the ice wedge contains 7%silt, 9% clay and over 84% fine andmedium sand. The ice wedges were filledand later on covered and protected bythe head which reaches a thickness ofover 1 m. This deposit (sediments > -2phi) is composed mainly of fine sand(41%), appr. 13% of silt and 17% ofclay. It does not contain CaCO, and isrich in RP. At the inferior part of thehead a clay layer (49%), with athickness of 10-15 cm can be identified.This layer displays a polyhedralstructure, clay skins, 37% sand and 14%silt, a strong carbonatic reaction and acontent of gypsum.Site 14Holdichl' is evidently influencedby solifluction. The head penetrates insome cases into the Tertiary formingdroplike subhorizontal involutions.Solifluction can be identified with thehelp of the following pattern: 1-general characteristics of the depositwith sorting of certain levels and aspecial orientation of the clasts; 2-continuity of the covering layer even onthe slope and spatial distribution ofthe deposit; 3- geomorphology of thearea with characteristic soft relief and4- textural characteristics of thequartz grains.The eolian origin of the sediments canbe verified through quartz gIain surfacetexture analyses with SEM. The abundance(over 25%) of textures with high relief,angular outline, conchoideal fracturesof different size, striations, parallelsteps and arch-shape,d steps are due tocryogenic phenomena which added to theinherited textures, probably related tointerstitial material of the RP. Insidethe casts, the quartz grains showeolization and a higher frequency ofchemical textures due to differentenvironmental characteristics and theage of the se&iments.Near "Cerro Kenselu1 the pseudomorphsvery frequently appear in form ofpolygons (with sides over 1 m) along theroad. The sediments of the casts are ofa darker colour than the host material.They are rich irl fine and medium sand(37% / 30%) and contain 15% clay and 10%silt. The RP appear in the filling aswell as in the host material and on thesurface. The host material containsCaCO, (appr. 50%), very much clay (55%)and silt (19%) and contrasts with thefilling because of its lighter colour.Site "Las Heras" is an example of thecharacteristic succession and thelithologic model described for Pto.Madryn in Northern Patagonia (TROMBOTTO,1992) . The upper layer consists of sandand silt of a greyish-brown colour. Itsthickness varies between 10 and 20 cmdepending on human influence. This layercovers another layer of brown sand witha thickness of 10-15 cm, polyhedralstructuIe and clay skins. Below thesetwo layers we find ice wedge casts witha depth of up to 70 cm. The pseudomorphspenetrate a sandy, calcareous layer withRP. A transversal cut reveals thecharacteristic polygonal structure. Justas in the case of site "Kensel" there isa contrast between the colour of thehost material which is less dark, sandy(44%) with CaCO, (27%; 10 yr 7/3 - 8/3)and the filling of the structure whichis darker and only slightly calcareous(10 yr 4/2 -5/3) At site "Las Herasq' thesediments of the filling are composed ofmedium sand mainly (52%), with 14% clayand 6% silt. At the lowest part of theprofile we find again RP impregnatedwith CaCO,, d'iisplaying a "nougatstructure" ( 50 cm) and subsequentlyforming "columns'! and subhorizontalstructures as well as "windows1'(TROMBOTTO, 1992) . These last carbonaticdeposits, or "tosca" clearly prove therelation between structures and thesubsuperficial water drainage and thewash-out of sediments and minerals.The sediments of the fillings of the1240

pseudomorphs at the thlee sitesmentioned can be very welJ compared.Analysing the granulometric cumulative ,curves, we obtain a Qd between 2.1 and2.5 phi, e.g. fine sand tending towardsmedium. The STD values (2-2.5) indicatea poor sorting of the material. Whilesite I1Holdich" displays an extremelyleptokurtic curve, the other two sites,very similarly, indicate leptokurticcurves. The symmetry of the frequencycurves gives us values of 0.3-0.6 in allcases. For "Kensel" and "Las Heras" weobtain exactly the same values. The peakin sand may be due to the transport ofcolloids during the pedogenesis.Comparing "Holdich" and "Las Heras" we ,find a remarkable continuity andsimilarity with the dif Eerence thatI'Holdich1' is strongly affected bysolifluction and influenced by itsgeomorphological position. Solifluctionmust have participated in thedecarbonization of the deposit.coNcLuBIQNsIn this first attemp.c and based on theinvestigation of characteristic sites wedare a paleoclimatic and sedimen,tary,reconstruction of, the area near' the 46'S.L in Southern Patagonia. For thispurpose we developed the followinggenexal pattern of the paleoenvironment:1- The ice wedge casts we foundrepresent the latest glacial andprobably the coldest period of the LateGlacial appr. 20.000 years ago. Theexistence of those structures requiredpermafrost and a mean annual temp. of atleast 14' C below,the present temp.2- S.ome casts are filled withsolifluction eolian material. Apaleoenvironment of this kind wouldindicate still cold, but slightly morete.mperate conditions and aridity orsemiaridity.3. A warm impulse allowed the creationof the clay level through pedog'enesisfor which we suppose a win. mean annualtemp. of 5' C and a mean annualprecipitation of 250 mm. ?\his momentmight be related to an interstadial.4- A posteriori this clay layer i scovered by solifluction which again isrelated to a cold impulse and a semiaridpaleoenvironment.5- Finally the predominating climateduring the Holocene was cool temperateand dry; conditions which hardly allow apedogenesis and which are generallyvalid until today.-We would like to .thankDr. FrankSchdbitz (University of Barnberg,'Germany) very much €or the chance ofpartic.ipating in the expedition andsharing with him a month of adventures.all over Patagonia which enabled us torealize this study. We are indebted toALUAR, Puerto Madryn for offering us itslaboratories, to THEM-CONICET, Mendozafor its cooperation with the SEM and weare especially grateful to SabineHerfert for the translation.REFERENCESCabrera, A. (1976). Re'giones Fitogeogr6ficasArgentinas. EnciclopediaArgentina de Agricultusa yJardinerla, Tomo 11, 1, 85 pp.,Buenos Aires.caldenius, C. (1932). Las GlaciacionesCuaternarias en, la Patagonia y Tierradel Fuego. Geografiska Annaler, Haft1-2, 1-164.- - - (1940). The Tehuelche or PatagonianShingle-Formation. Geografiska Annalel,160- 181.Corte, A. (1982). Geocriologia general yaplicada. Revista del Instituto deCiencias Geolbgicas, UniversidadNacional de Jujuy, Nr . 5: 1-33 ps.Czajk'a, W. (1955). Rezente undpleistozane Verbreitung und Typen desperiglazialen Denudationszyklus inArgentinien. Acta Geographica 14, Nr.10, 121.- 140, Helsinki.Groeber, P. (1950). Quartare VereisungNordpatagoniens. Sonderdruck derZeitschrift "siidamerika", 6 pg., 'Buenos Aires.Mercer, J.H. (1985). Las VariacionesGlaciales del Antiguo Cenozoico ensudamkrica a1 SUL del Ecuador. ActaGeocr iogbnica, Nr . 3, 86 - 206,Mendoza.Troll, C. und Paffen, KH. (1969). Karteder Jahreszeiten-Klim'ate der Erde.Erdkunde, Band XVZII, 28 pp.Trombotto, D. (1992). The CryomerePenfordd, Patagonia. <strong>International</strong>Workshop: "Permafrost and PeriglacialEnvironments'in Mountain Ayeas", 33pp. , Calgary.Windhausen, A . (1924). Lineas Generalesde la.Constituci6n Geol6gica de laRegi6n Situada a1 Oeste del GoLfo deSan Jdrge. Boletin de la AcademiaNacional de Ciencias, Tom0 XXYTI, 3,167 -320, C6rdoba.a1241

SEASONAL FREEZING AND THAWING GROUNDS OF MONGOLIAD. TumurbaatarInstitute of Geography and Geocryology. Mongolian Academy of SciencesThe seasonal freeeing and thawing grounds of Mongolia are subdivided into threeregions: the first is a seasonally thawed ground region. the ground thaw depthin the loam is 1.8-2.8 m and in the loamy sand with gravel is 2.8-3..6 m. Thatthaw begins from June and continues until October: the second is seasonallythawed and frozen ground region, the depth of seasonally thawed and frozenground is 2.6-2.8 m in the loam and 3.5-9.5 m in the .loamy sand with gravel.Ground frost' begins from the middle of October and continues until April. Thethird is ground freezing region, permafrost does not occur here. Ground isfrozen seasonally. Depth of frozen ground is 1.7-T.5 m in the loam and 2.5-3.2 mi'n the loamy sand with gravel, Ground freezing is from the beginning of Novemberto March.INTRODUCTIONThe seasonally freezing embraces all of theterritory of Mongolia. The first summary ofMongolian seasonally freezing was given byP.I. Gukov (1961). N. Longid (1969), S.I.Zabolotnuc (1974) and D. Tumurbaatar (1975).In the last 25 years, seasonally freezinginvestigations have been concentrated by theauthor in the territory of Mongolia. All maierialswere generalized and complied into large andsmall scale seasonally.frozen and thawed groundmaps of Mongolia (1977, 1985).*In some industrial areas maps were compiled,of large and middle scale seasonally frozen andthawed ground (1971, 1976, 1989).The main purpose of this report is the discoveryof the general and regional regularitiesin the seasonal freezing and thawing of ground,of changing of temperature and composition ofground and natural conditions.NATURAL, CONDITIONSIMongolia Is situated in Central Asia, fromthe sea, to the border of the Russian Republicin the north and China in the south. The southwesternand western part of the territory iscomposed of the Altai, Khangai and Khubsugulmountains in the eastern-Khentei mountains. theeastern-north-Khingane mountains, the easternand south-steppe and Gobi desert.Mongolia is on the whole a mountsinoue countrywith an average absolute height of 1580 metresabove sea level. The lowest point is the basinof Khukh-Nuur in the east, which lies a,t analtitude of 560 metres above sea level and thehighest point within the Republic is Khyiten inthe Mongoli'an-Altai. which stands at 4374 metres.Average air temperature of the coldest month-January in -30°C in the north of Mongolia and-14'C in the south. The value of the warmestmonth-July is 12'C in the north, and 24'C in thesouth, Average annual air temperature in thesouth part of country is -5.2'C and in the north4°C.Mongolia has a sharp continental climate andthese conditions promoted the development ofpermafrost in 15% of the territory and thedeepening of seasonally freezing ground on thewhole.THE THAWING OF'GROUND IN MONGOLIAThe main permafrost regions are in the Khangai.Khubsugul, Khentei and Mongol-Altai mountainterritories end'are characteristically seasonalgrounds.The predominant absolute altitude of Khubsugulmountains country ranges from 2000'to'3460 m(a.s.1.). But absolute altitude in the Darkhatdepression ranges from 1500 to 1800 m (a.s.1.).At Dood Tsagan Nuur in.the Parkhadin depressionon the left bank of the Shishkid river, thedepth of seasonallthawing Is 1.3-2.3 m in theloamy lake grounds, 1.5-2.5 m in the loamy sand,and 2.6-3;3 m in the sand. In the beginning ofthe valley, in Sharga at the river, it is 3.3-3.5 m in the gravelly ground with sandy fill.Seasonally thawing of ground in the northfacingslope of the mountain is 2.3 m in theloamy sand. On the south bank of Khubsugul lakeis situated the city of Khatgal (1650-1700a.s.1.). in which seasonal thawing is ;3.2 - 3.5min the gravel ground with loamy fill.The above research materials shows that thedepth of seasonally thawed ground ranges from1.3 m to 4.0 m in Khubsugul mountain territoriea.. Deepest ground thaw is 3.3-3.5 m in the sandand gravel. Shallowest thaw of ground is observedto be 1.3-2.3 q in clay and loam.The river-heads Sharga and Tes are located onthe lake plateau of the Khandai mountain region.Here the depth of the seasonal thawing is 2.2-3.1 q in the boulders with 'sandy fill.The depth of the seasonal thawing ground on1242

'the Tarbagatai raugc at the altitude of 3000-3200 m (a.s.1.) is 0.9-2.3 m in the loam andloamy sand ground, and at the height of 2000-2300 m a.s.1. it is 2.8-3.8 m in the same typeof ground.In the lake hollow, 'Terkhiin Tsagaan Nuur,the seasonal thaw of ground is 1-61 m on thehidrolaccolite in the loam.Seasonally thawed ground (Northern Khangairange) in the region of Shiluustei at the altitudeof 2200 m a.s.1. is 3.8-4.5 m i n the loamysand and sand with gravel, in the settlementRajan Oboo somon near the lake valley (countrv)Tsagaan Nuur (1935 rn a.s.1.) it is 1.7 m in theloam, in the center of the settlcmmt Galuut somon itis 3.5 m in the sand wi~h gravel, and in thecenter of the settlement Bajanbylag somon (2200m a.s.1.) it is L.2 m in the sand with gravel.The above mentioned data shows the depth ofseasonally thawing changed from 0.9-4.5 m, inthe Khangai m0untaj.n country.The smallest thaw located on north-facingslope and lake and river valleys is 0.9-2.3 min the clay and loam, the deepest thew on southfacingslope is 3.5-4.5 m in the sand with gravelground. Absolute altitude of Khentei mountainouscounrry is 1500-2000 m.In Bajanzurkh somon and Nalaikh, located inthe south-west part of Xhentei mountains, theseasonal thaw is 2.h-2.8 m i n loam and 3.0-3.5 min sand.In th'c southern part of Khentei range, in theregion of the settlement Manyunmorit somon, theseasonal thawing ground is 2.8-3.2 m in loamwith gravel.The abscllutc altitude 1620-1730 m i n locallowlands.marshlands and hollows, and the seasonalthaw is 1.2-2.3 m in the loam and locallyis 2.3-2.8 m. In the eastern part of Khenteirange, in the region of Batshereet on the northfacing slopes, the depth of thaw is 3.0 m in lowmoisture (7-137) sand with gravel.The smallest depth of thaw in Khentei Mountainregion is 0.6-1.3 m and occurs in the areasof the north-facing slopes, marshland, wet sandwith thick continuous moss cover and water shed.The deepest thaw (3.5-4.5 m) occurs in theareas of the south-facing slopes and of thediscontinuous permafrost zone and water shed.As well as i n the river deposits with the sand,loamy sand pebbles and gravels.Altai mountainous region is situated in thewest ddge of Mongolia, and the absolute altitudeof Altai mountain is from 3000 m to 3474 m.There is continuous and discontinuous permafrostin the region.The valley "Khonghor ulen" located betweenthe Altai mountainous region is situated at theabsolute altitude of 2420 m and the seasonalthawing of ground is 2.6 m in the gravels orgravelly sand. And near to lake "Tsagaan Nuur"which is at the absolute altitude of 2100 m thethawing depth in the clay and loam is 1.8-2.2and in the sandy ground is 2.8-3,O m. Nearm,"Tsenghel" soman at the absolute height of 2150m, the thawing depth of sandy ground is 4.3 m.Lake "Tolbo" is at 2100 m and the thawing ofmuddy loam ground near the lake "Noghoon" is2.8-3.0 m.On the north facing slopes of the Altaimountainous region the thawing of broken stoneground is 3.0-3.8 m and on the south facingslopes the thawing of the same ground is 3.5-.0.5 m.The conclusion is that the thawing' of thelake mid clay and loamy ground in the valleybottoms between the mountains of the Altai mountainsregion is 1.8-2.2 m, and in the sandyground is 2.8-3.0 m.Thawing of gravel and sandy ground on Lht.south facing slopes oE mountains is 3.5-4.5 rn.And the thawing of north fecing slnpcs of mountainsi n the same ground is 3.0-3.5 m.It was easier to reveal the altitudinalregularities in the changjng seasonal thawingground than to reveal the latitudinal rtigularities.The deepest thawing of ground in loam is2.8 m and in sand is 4.8 m under 1000 m heightfrom the latitude of LS in Mongolia.A t the absolute altitude of 1700-2300 m inKhanghai. Altai, Khubsugul and Khentei mountainousregion the deepest scasonal thawing is from2.0 m in loam and tu 4.5 m in sand. 4 t theheights of 2300-3000 m the thawing depthdecreases 1.5 m in loam, 3.5 rn in sand and atthe heights above 3000 m it is 1.0 m in loamand 2.0 m in sand. In the direction from southto north the thawing depth of ground decreased.For example: the thawing depth,of ground nearGhalut somon situated i n front side or Khangaimountains ridge is 2.5 m in loam and 3.5 m insand. But 500 km to the north in the Tsagaanlake va'llcy of Khubsugul province the depth ofthawing is 2.0 m in sand and 1.3 m i n loam(Fig.1).The change of the depth of seasonal thawing ;Iin loam ?x) and sands (0) in Khannav-Khuysg8l'rnountain regions of Mongolia ITHE SEASONAL FREEZING OF GROUNDThe south part of Mongolia is predominatedby the steppe and Gobi semidesert zone.The.seasona1 freezing of ground in this zoneis tightly connected with the natural complex.In the southern part of Mongolia the precipitationis low annually, 40-100 mm, snow cover isthin 3-5 cm, and is cold i n winter and warm insummer.The seasonal freezing ground consists genera1l.yof clay, loam, loamy sand, gravel, sandand different solid rocks of the Tertiary andQuaternary period.In the southernmost point of Mongolia, ZaminUud, the amplitude of surface temperature is21-22'C, mean annual ground temperature is6-7"C, ground moisture is 8-12X, depth ofseasonal Freezing is 2.3-2.4 m in loam, and, 2.4-2.5 m in sandy ground. To the north near SainShand, moisture of sandy ground is 10-157, meanannual ground temperature is 4.5-5.0"C. anddepth of ground Preezipg is 2.2-2.4 m in loamand 2.4-2.6 m in sand (Table 1).Near Dalanzadghad the mean annual groundtemperature is 7-8'C, aid depth of freezing is1.7-2.1 m in loam and 2.2-2,4 m in sand.1243

Table 1.Depth of seasonally freezing ground of MongoliaPopulatedarea(settlement)AmplitudeMeanoscillationDepth Composition ofannualof Moisture of lithology seasonalgroundtemperature of ground of the frozetemperatureI O F \ of ground groundground\ b jsurf aceSouthern part of MongoliaDalanzadghad 8.0' 18.6 10.0 loam. 2 .oZamiin Uud 7.1 21.5 8.0 sand 2.5Sain Shand 6.2 20.9 9.0 sandMandalgobi 5.0 18.6 5.2 loamy sand 2.5Central part of MongoliaB a j ii n 1 .(l 19.9 16.0 loamy sand 3.4Choir 3.1 20.0 15.3 loamy sand 3,OTsetsetleg 2.8 15.9 14.9 loamy sand 3.4Ulaanhaatar 0.6 25.0 18.0 loamy sand 3.7Northern part ofKharaa 2.1 22.0Errlenet. 1 .o 15.8Bulghan 1.6 19.3Ulaantolgoi 1.8 16.0MonRolia . .16.5 sand13.4 loamy sand 3,O20.0 loamy sand 2.919.7 loamy sand 2.5The conclusion is that the seasonal freezingof ground in the southern part of_Mongolia beginsfrom the beginning of November and is completelyIrozen,at. the end of Fcbcuary. The thaw of seasonalfreezing ground beklns from second week ofYarch and ends in April.The. ground frcezina ?n the southern part ofMongolia' is 1.7-2.4 m in loam 'and 2.h72.6' m insandy ground.Examples of the ground seasonal freezing ofsteppe .and mountainous steppe eone of the centralpart of Mongolia can be seen in the regions nearto Bajan somon, Ulaanbaatar, Tsetserleg, Rajanhonghorand Arcaykher.Rajan somon in the Central province issituated in the steppe zone which is 90 km tothe south of Ulaanbaatar.The main st-ructure of seasonal thawing andCreeziny ground are deluvium break stone andgravel, lacustrlne loam and sandy loam.The fluctuation of suriace temperature is10.1-14.2°C and the mean annual ground temperatureis 0.8-1.9OC, and the ground'moisturecontent is 3-152. Ground freezing is 2.7-4.5 min loam and 3.5-3.9 m in the sandy ground.The ground freezing near Ulaanbaatar occursin the proluvial and alluvial sands with gravels,and deluvial loamy sands with broken stones. 'The fluctuation of surface temperature is20-25'C, mean annual ground temperature is 0.5-2.5'C; ground moisture is 6-2OX, and the depthof seasonal freezing of ground in the gravelground with sand in the river and lake valleyis 2.6-3.3 m.The depth of ground freezing in the back sidesof the mountains 'in loamy sand with broken stonesis 4.2-4.8 m. The fluctuation of surface temperaturenear Arvaikher town i s 15-18"C, meanannual &round temperature is 0.3-2.OeC, groundmoisture is 5-10? and the depth of freezing ofground is 2.2-3.2 m in sandy ground and 2.5-2.9min loamy ground. The freezing of ground is fromthe end of October until the beginning of April.The seasonal freezing of ground near Bajanhonghortown is 2.9 m in loam and clay, and 3.4min sa,nd and loamy sand. The depth of seasonal.freezing of ground near Tsetserleg town is 2.9 min clay and loam, and 3.4 m in sand and loamysand. The seasonal freezing of ground in thecentral part of Mongolia is from the end ofOctober until the beginning of April. The meanannual ground temperature is 0.3-1.O'C in theloamy sand with deluvial broken stones in theback sides of mountains, and 2.0-2.5OC in thealluvial sand and lbamy sand with gravel. Thedepth of seasonal freezing of ground is 4.0-4.8111in the back sides and 3.5-4,O m in the frontsides of mountains, and 2.6-3.3 m in the riverand, lake valleys. It's possible to determinethe seasonal freezing of ground of the northernpart of Mongolia or the region of Orhon andSelenga on the basis of the studied materialsobtained in the territories, of Selenga, Bulghanand Khubsugul provinces.The state farm of Zun Khara is situated 120km to the north of Ulaanbaatar. Here the seasonalfreezing of ground begins from the first tendays of November and finishes at the end ofApril. The mean annual ground temperature is1.5-2.1°C, depth of seasonal freezing is 3.2-3.4m in clay and loam and 3.6 m in sand and loamysand of Khara and Shara river valleys (Table I).The depth of sandy ground freezing in theeast-north facing slope of the mountain nearDarkhan city is 4.0-4.3 m, the smallest freezingwas 1.7 m in the,loamy ground with 16-25% moisturenear the high bottomland of Khara rivervalley.The depth of freezing ground in the valleynear Erdenet hill (12b0 m) is 2.3 m in lakesand with moisture of 254, and is 3.5-3.7 m inlight loamy sand on west facing slope of Erdenethill (1300 m a.s.1.). Ground freezing is 3.9 min the north facing slope of Erdenet hill(1450 m a.s.1,).The seasonal freezing of ground in theKhanghal river basin begins in the beginning ofNovember and finishes at the end of April.There is sand with loamy fill and with moistureof 15-202, mean annual ground temperature is1244

3.0-3.2'C, and depth of seasonal freezing is 2.4-2.7 m.The mean annual ground temperature of the loamand loamy sand with broken stems on the southfacing slopes of mountains is 2.5-3.OoC, moistureis 9-llX, and depth of seasonal freezing is 2.3-3.7 m. And on the facing slopes of the mountainsthe moisture i s 10-12% in the loam and loamy'sand, it is mean annual temperature is 0.5-1.OaC,and the ground freezing is 3.7-4.0 m deep.Near Bulghan town (1220 m a.s.1.) the groundfreezing begins from the beginning of Novemberand finishes at end of April. Mean annual groundtemperature is l.6"C and it is seasonal freezingis 3.2 m deep. In Murun town, which is situatedin the northeast point, the seasonal freezing ofground begins after October 20, and finishes atthe end of April. It's seasonal freezing is 3.9mdeep.The conclusions taken from the above mentionedevidence shows that there is a fluctuation of1.7-4.0 m in the seasonal freezing of ground inthe Orkhon-Selengha region.The depth of seasonal freezing in the groundincreases from the south to the north untilreaching the central region (Ulaanbaatar). Andit confirms fully the conclusions of Russianscientist S.I. Zabolotnik who discovered thataccording to the natural regularities there is areduction in the Orkhon-Selengha region.The changes of the thickness of seasonalfreezing is tightly connected with the factorsof the rise in the winter duration in the directionof the north and the increase of absolutealtitude of the earth and also the influence ofsnow cover.Beside these factors it is also connected withother factors of freezing lake ground structure,moisture and temperature. Under the increasingabsolute altitude from 900-1030 m in Dalanzadghadto 1200-1703 m in Ulaanbaatar the mean annualground temperature reduces slowly from 6.7OC inDalanzadghad to 0°C in Ulaanbaatar. But thethickness of snow cover does not change.On this condition the depth of ground freezingis increasing from south to north.The surface absolute altitude is reducingfrom Ulaanbaatar to the border from 850 m to 600m, and the thickness of snow cover increases.And snow cover reduces the depth of seasonalfreezing by 0.2-0.7 m . Therefore the thicknessof seasonal freezing in the northern part of theOrkhon-Selengha region is 0.8-1.4 m less than theregion of Ulaanbaatar city (F'ig.2).- 81THE REGIONALIZATION OF THE SEASONAL FREEZINGAND THAWING GROUNDThe seasonal freezing and thawing ground ofMongolia is subdivided into three regions (Fig,3) :1. Region of seasonally thawed ground. Thedistribution of permafrost is continuous. Groqndthaw in the loam i s 1.8-2.8 m and in the loamysand with gravel is 2.8-3.6 rn. That thaw beginsfrom beginning of May and finishes in October.2. Region of seasonally thawed and frozenground. The distribution of permafrost is dXscontinuous.The region of seasonal thawing andfreezing grounds are subdivided into subregions:a. Subregion: Depth of seasonal.ly thawed andfrozen ground is 2.8-3.4 m in the loam and loamysand and 3.5-4.5 m in the sand with grsvel.b, Subregion: Depth of seasonally thawed andfrozen ground is 2.6-2.9 m in the loam and loamysand and 2.9-3.5 m in the sand with gravel.Ground frost begins from middle of October ar.dfinishes at the end of April.3, Region of seasonal frozen ground. Permafrostdoes not OCCUF here. The region of seasonalfreezing grounds are subdivided into two subregions:a. Subregion: Depth of seasonally frozenground is 2.5-2.7 m in the loam and loamy sandand 2.8-3.2 m in the sand with gravel.b. Subregion: Depth of seasonally frozenground is 1.7-2.1 m in the loam and loamy sandand 2.4-2.5 m in the sand with gravel.CONCLIJSIONSThe territory of Yongolia on the whole hasseasonal freezing and thaw.The depth of seasonal thawing in loam is 1.8-2.8 m and in gravel is 3.6-4.6 m. The depth ofseasonally freezing in loam is 2.h-3.2 m and insand with gravel is 2.8-4.5 m.REFERENCESBrown R.J.E., Permafrost in Canada, Universityof Toronto Press, 234p.Zabolotnick S.1. (1975) Seasonal freezing andthawing grounds. Geocryological conditionsof the Mongolian People's Republic. The joint.Soviet-Mongolian scientific-research geologicalexpedition, transations, Vo3.10, pp.49-73, (Text in russian).Zhuu Youwu, Qiu Guoqing anrf Guo Dongxin (1991)QuatErnary permafrost. in China, QuaternaryScience Reviews, Vol.10, pp.511-517.Longid H. (1969) Permafrost of Mongolia. ('l'extin Mongolian).Tumurbaatar D. f1975) Seasonal freezing andthawing grounds of the Flongolian, Academy ofSciences pp.39-74. (Text in Mongolian).Tumurbaatar D. (1933) Classification .of' seasonal-1.y freezing and thawing grounds, Geographicproblems of Mongolia, Vol.23, pp.25-29.Fig.2 The changc of t-he depth of seasonal Ireezingin loam (x,) and sand (0) in central Mongolia1245

Fig.3 Map of regio~l of seasonal freezing and thawing ground of Mongolia1246

HIGHWALL STABILITY IN STRIP MINES IN PERMAFROSTJalal VakiliNinyo & Moore9272 Jeronimo Road, Suite 123AIrvine, California 92718 USAThe instability of highwall slopes, exposed to thawing,surface coal minesin permafrost regions is described. Remedial measures are proposed to preventshallow as well as deep slope failures in these regions. A practical designconfiguration, which involves introduction of a ber.ch wide at the bottom of thepermafrost layer, is proposed to mitigate highwall instabilities.INTRODUCTIONCoals, ranging in age from carboniferous totertiary, are found in many areas of permafrostregions. Although predominantly lignite andsubbituminous in rank, the range is completefrom lignite to anthracite, Reserve6 ofstrippable coal in Alaska, for example, areconsidered as 5 billion tons subbituminous and2 billion tons bituminous coals, Reserves forthis area include 120 billion tons of inferredresources and probably more than 1.5 trilliontons of estimated reserves (Hammond 1976). Thecoal is mined at present by the strip miningmethod. Draglines and scrapers are primarilyused to remove the overburden. Coal seams aremined using front-end loaders, shovels, andtrucks, or easy miners.In designing large excavated slopes. in stripmines, the most important decision of theengineer is the selection of a slope angle.The purpose of the geotechnical study of theslope is to, ensure reasonable stability of theslope' in the most ecdnomical way. Because therock mass composing each highwall slope isunique, there are no standard solutions forslope stability analysis. A practical solutionis formulated from the basic geology data,engineering properties of soil and rock,geometry of the slope, the ground water andenvironmental observations.. , 'This paper.describes the factors affectingcut slope stability in permafrost regions and .the alternatives available for improving thestability of slopes.FACTORS AFFECTING STABILITY OF SLOPESen Characteri t?l%%erburden of czali?eds, in permafiastregions of Alaska, and particularly the stratabetween the'coal beds, consists largely ofpoorly consolidated sandstone and claystonewith pebbly layers (Wahrhaftig 1954). Theresponse of cut slopes in fa:ozen ground isdirectly related to the nature.of the soil andthe distribution of ground ice. It is difficultto determine the conditions from a siteinvestigation program alone, and a clearpicture of the anticipated response may notemerge until actual ground ice conditions areexposed during mining operations.Minina MethodStrip mining usually is accomplished byremoving the overburden from a strip across-onedimension of the deposit. A parallel strip isthen excavated usually in the oppositedirection and the overburden is placed into thestrip previously mined. This cycle is repeatedas long a6 the deposit allows, The pitconfiguration is determined essentially byconsidering all the important factors: i,e.terrain, ratios, overburden, shape of deposit,thickness of seam, production requirement, andequipment available. Where the overburden iswell over average thickness, or is of such anature that is has a low angle of repose, thelonger reach and other characteristics ofdragline may make it the cheapest strippingunit in the long run (Bhattacharyya, 1982).If the coal is pitching, is as in theanthracite region, the dragline makes itpossible to operate by casting, work deeperdown the pitch at a lower cost, and in someinstances, to recover the coal without thenecessity of men or equipment entering the pit.Method of mining determines the height andangle of highwall and spoil slopes, benchwidth, location and specifications afequipment. Dragline exerts extra load an thehighwall and contributes to the overallinstability of slopes, and can often generateshallow failures in the surface alluvium andthawed rocks.11247'.

mWANISI OF SLOPE FAILUREOnce the factors affecting slope stabilityare evaluated, it will be necessary to developa model that represents the particularconditions found within the pit. For theselected model, the ehear resistance will becalculated along the potential failure surfacethat would be necessary to bring the materialenclosed by the slope profile and the failuresurface to the point where the movement isabout to occur, as shown on Figure 1. If thecalculated shear resistance at limitingequilibrium is less than the available shearstrength, the slope will be stable. Thecomparison of the available and computed shearresistances is carried out by means of an indexknown as the factor of safety. In mathematicalterms :FLfr- c + o tan @ = shear strengthwhere: F = factor of safetyrtalong some shear surface? = equilibrium shear stress alongthe sane shear surfacec = cohesionu = normal stress.$ = angle of internal frictionTo determine the effective normal stress aand the shear stress r along the failuresurface, various analysis methods are used,1.e. method of slices for circular arc analysis(Bishop 195s).The analysis of failure in permafrost groundrequires the specification of the in-situ shearstrength of frozen ground. The strength offrozen ground deVeelops from cohesion,interparticle friction, and particleinterlocking, much as in unfrozen soils.Cohesive forces are attr.ibute8 to adhesionbetween soil particles and the ice in the soilvoids and to surface forces between particles.The strength of .frozen soil decreases withduration of- load application and exposure tothe atmosphere (Vyalov 1959). Therefore, longtermestimates of permafrost soils strength areneeded in order to be able to apply classicalmethods of analysis such as method of slices(McRoberts, 197413). When a frozen soil issubjected to a lpad, it will respond with aninstantaneous deformation and a time- andtemperature-dependent deformation called creep.At low solid concentrations (ice-rich soil) thestrength may approximate that for ice and the'long-term strength will be very small since icecreeps under extremely small stresses. At highsolid concentrations (ice-poor soils)interparticle friction and dilatancy play amajor part in determining the soil strength.The great difference in strength between frozenand unfrozen soil is derived from addedcohesion due to t%e ice component and the creepeffects of ice. Ice-poor ground exposed tothawing will lose its cohesive bonding andlong-term fstrehgth of this kind of soil can bededuced from triaxial testing of unfrozensoils. In coal mines, when overburden isremoved, the highwall slope and the pit .bottomwill be exposed to thawing. Permafrost materialwill, therefore, lase their cohesive strengthwhich may result in highwall slope failures.Failures may be in the form of a shallow slopeface failure or a deep-seated overall slopefailure.Two kinds of shallow face failures are commonin permafrost terrain:If permafrost soils are not sufficiently icerich,a thawed layer begins to form on thesteeply inclined cut slope. This thaws to somecritical depth, at which time the face failurestarts by development of a skin flow in thethawed layer which exhibits characteristics ofa viscous fluid in its downslope motion, asshown on Figure 2. Skin flow has been reportedby many investigators (McRoberts 1973).FIGURE 1 - PROPOSED HIGHWALL MODEL FOR PERMAFROST REGION

, ,FloU~ 2 - SKIN FLOWFIGURE 4 - BLOC? SLIMRotational slides usually occur in completelythawed soil. These failures are similar to theclassical circular type of failure common inclays in more temperate regions as shown onFigure 5 (McRoberts 1973).AThe third type of slide of icy permafrostslopes is Creep movement along the beddings(McRoberts, 1975a). Creep movement might beinitiated by a warming up but not thawing offrozen ground (Figure 6) caused"by removing theinsulating vegetation cover, or by changingstress levels from additional load applied fromequipment, or stress re'lief from cutting'theslope. If such a process occurs, one result maybe long-term creep rupture of the frozen soil,which would Pesult in a failure like form blockslides. The creep stren'gth of firozen soils is ,defined as the stress level,,'after a finiteLess thaw-susceptible soils can be subject to time, at which rupture, instability leading torapid desiccation and after undermining byrupture, or extremely large deformatibnsmelting ice, may fail as a fall failure. Itwithaut rupture occur, and eari be determined bysimply involves the downward movement oftests on undisturbed permafrost soil samples.detached blocks falling under the influence of8~gravity as shown on Figur.e 3 (McRoberts 1973;1974a).J"0 ' = SHEAR STRAIN~''5 SHEAR srwssFIGURE 6 - CREEP MOVEMEhNT-# =. A$ "kcff4 B.m,a PARAMETERS TOBE DETERMINEDEXPERiMMENTALLYREMEDIAL MEASURES TO PREVENT $T,OPE FAILUREFIGURE 3 -The feature complicating the design bf cutsFALL FAnuREin permafrost soils is the instakiility that mayoccur when"exposed frozen soils are allowed toThree types of deep-seated overall slopethaw, If the soil is fine-qrained and ice-rich,failure have been encountered in permafrost ablation processes cah generate su'bstantialterrain.rates of surficial failures. The magnitude ofBlock slides involve the movement of a large the movements produced by actively retreatingblock that has moved out and down a with cuts is substantial and could be deliberatelyvarying degree of back tilting. These types of induced to speed stripping process,slide are common where glacio fluvial sands and The performance of actual cuts in permafrostgravels overlie glaciolacustrine silts and soils has been studied by several authorsclays. Slumps along bedding plane could involve (Pufahl 1975, McPhail i976). Experihce to datelarge areas, ashown on Figure 4. Detailedsuggests that the following dedign , , 'Iobservations of block slides are given by configurations are appy-opriate:.(McRoberts 1973, 197433, and Isaacs 1972).Cuts in frozen id-poor'doil, or bedrock- Cuts in soil and bedrock 'which -are-'stable upon thaw can be designed so,that the ^slope angle is compatiblewith khe,imfrozen properties. ,* Cuts in ice-wedge terrain . .- If a cut in terrain in which theground-ice condition exists as amatrix of ice-wedrje' in ice-poor2FIGURE 5 - ROTATIONAL SLIDEmineral soil, the most ehitable designdebends on local conditions. Steepcu$s are usually desirable becausefewer ice wedges are encountered.Cuts ih ice-rich pefmafrbst-. Various design configurations can beconsidered for cuts in which the I 'ground-ice conditions are horizontallycontinuous. Because it can bedifficult to preverk degradation, thedesign of ice-rich cuts ih'warmpermafrost must necessarily allow forsome thaw in the ice-rich soil.Introducing highwall benches willprevent thiwed material from movlnginto the pit,as shdwn on Figure 1. Analternative method for,improvihg slopestability, in this case, is to operatestripp'ing operations during winter ,time when the ambient temperature,isbelow Ereezing'point and-there is nodanger of thawing.

TU reduce the danger of along-beddingslides, the strip mining operation shouldbe planned so that all bedding planes onwhich sliding could take place would dipinto the highwall face as shown on Figure1.When the mining depth is more than thedepth of permafrost layer, it isrecommended to introduce a bench in thehighwall slope at the bottom of thepermafrost layer, as ahown on Figure 1.The lower slope of the highwall is notfrozen and will not thaw, so there is ascarce possibility of instability of thelower slope. Additional water fromthawing permafrost and ice wedges shouldbe collected at the toe of the upperslope by introducing a drainage ditch asshown on Figure 1. A wide bench will helpto improve the overall stability of theslope and prevent the thawed material ofthe permafrost layer from moving into thepit. To prevent face failure in the upperslope after thawing, the slope angleshould be designed on the basis offrictional strength of the thawedpermafrost materials only.Finally, in no other area of slopestability design do constructionactivities offer such an importantinfluence as in permafrost terrain.Inappropriate construction activities,' such as premature removal of topsoil oroverburden, may disrupt slope stabilityand should be avoided by careful advanceplanning of mining methods and miningsequences.McRoberta, E. C. and N. R. Morgenstern, 197413,**The Stability of Slopes in Frozen Soil,IaMackenzie Valley, N.W.T., Can. Geotech.J., 11:554-573.McRoberts, E, C., 1975a, vvSome Aspects of aSimple Secondary Creep Model forDeformations in Permafrost Slopes, Can.Geotech., J.,. 12:98-105.Pufahl, 0. E., 1975, "The Stability of ThawingSlopes,*@ unpublished Ph.D. thesis,University of Alberta, Edmonton.Vyalov, S. S., 1959, "The Strength and Creep ofFrozen Soils and Calculations €or Ice-SoilRetaining Structures,t* U.S. Army cold Reg.Res. Eng. Lab. Trans. 76, Hanover, N.H.Wahrhaftig, Clyde and J. H. Birman, 1954,"Stripping - Coal Deposits on LowerLignite Creek, Nenana Coal Field, Alaska,"Geological Survey Circular 310.REFERENCESBhattacharyya, K, K., J. Vakili, S. Y. Chi,1982, "Geotechnical Considerations for theamsign of Highwall and Spoil Slopes inLignite Mines1" Proc. Mini Symposium,"Lignite Mining and Stability, Sept. 1982,Honolulu, Hawaii.Bishop, A. W., 1955, "The Use of Slip Circle inthe : tability Analysis of Slopes",Geotazhnique, 5, 1, pp. 7-17.Hammond, S., G. R. Martin, and R. G. Schaff,1976, "Biennial Report 1974-75," Divisionof Geological and Geophysical Surveys,State of .Alaska.Isaacs, R. M. and J,. A. Code, 1372,, "Problemsin Engineering Geology Related to PipelineConstruction". National Res. Counc. Tech.Nem. 104 pp, 147-179.McPhail, 5. F., W. B. McUullen, and A. W.Murfitt, 1976, "Yukon River to PrudhoeBay:n Lessons in Arctic Design andConstruction, Civ. Eng. (N.Y.), 46(2): 78-82.HcRoberts, E, C., 1973, "The Stability ofSlopes in Permafrost," unpublished Ph.D.thesis, University of Alberta, Edmonton.McRoberts, E. e. and W. R. Moryenstern, 1974a,"The Stability of Thawing Slopes," Can.Geotech. J., 11:447-469.

DESTRUCTION AND REHABILITATION OFWang ChangshengSHAFT LINING USED IN FROZEN SHAFTandLiu RihuiBeijing Research Institute of Mine Construction,Central Coal Mining Research InstituteBeijing, 100013, ChinaIn recent years, the shaft linings of more than 10 shafts sunk with freezingmethod in China coal mine were destructed in varying degrees; the regions ofdestruction occurrence are mainly in Huaibei, Xuzhou and Datun coal minedistricts,and the major expressions of shaft lining destructions are peeling andcoming off of shaft lining concrete and crooked reinforcing bars.thus leading toshort-term stop production of mines and seriously menacing the security ofmines. In this paper .the process of shaft lining destruction of these shafts isexpounded and a simple analysis is made of the reason of shaft liningdestructions. In accordance with shaft lining destructions, rehabilitations ofthese shafts have been carried out many times by mine construction departments.The'%tate of the rehabilitations is also given briefly in this paper.BRIEF ACCOUNT FOR DESTRUCTION QF SHAFT LINING IN Inspection after the sound stopped showedFROZEN SHAFT AND EMERGENCY TREATMENTthat there was severe circumferencial peelingoff around shaft at the depth of 241 m , theSince 1955 when freezing method for sinking destruction surface wa6 smooth, and gravel wasshaft was adopted first in China, more than 320 cut off into a plane, with peeling-off depthshafts have been sunk, totalling to 50 km. In being 200-250 mm and height being 500-1500 mm.China freezing method has already become a major The reinforced bars was revealed and bent underspecial shaft sinking method in unstable water- force. There were several seepages in the shaft.bearing strata. After 1987, the accidents of Afterwards, it was also found that crackspeeling and coming aff of shaft lining concrete appeared in hoist tower with the longitudinaloccurred , early or late, in 17 shafts sunk with cracks in the lead and some cracks -looked like,the freezing method in Huaibei & Datun the character /\ with a dip angle of 50-60. .coalfields, China. In some shafts, there were There were altogether 14 cracks(Fig.1).water flowing into shaft, carrying sand with it, The method for emergency treatment is thatthus menacing the security of mine production. three lines of 18# channel steel rings are usedThe mine had no alternative but to stop every one meter. 58 channel steel lagging isproductian to make emergency treatment. This employed between the ring and shaft 1ining.Therepaper briefly describes the destruction andis iron wedge for tightening between the laggingrehabilitations of Linhuan Colliery auxiliary and shaft lining. 22 pieces of 180 N/m rail areshaft, Haizi Colliery auxiliary shaft and placed longitudinally around each circle forZhangshuanglou Colliery auxiliary shaft.reinforcing the rings.Rails are connected to therings with clips and then gunite is used forLinhuan Colliery Auxiliary ShaftThe coal annual output of the mine wasdesigned as 1.8 million tons.The mine was handedreinforcement.Auxiliary Shaft of Haizi Collieryover and put into production in 1985. The The design annual output of Haizi Colliery isauxi1ial.y shaft of the mine is 508.7 m in net 1.5 mt. The mine was handed over and put intodepth and 7 .2 m in net diameter. Freezing method production in 1987. The Auxiliary shaft of thewas adopted in overburden and double-wallmine has a net diameter of 7.2 m.Freezing methodreinforced concrete lining was used with the for sinking shaft was adopted in overburden andbuter wall and inner wall being 150 mm and 900 compound lining was used with its thicknessmm thick respectively.The outer wall was closely being 1100 to 1400 mm. Where the destruction ofintegrated with the inner wall. The design shaft happened, triple-wall lining was used.,Thestrength grade of concrete is 40MPa and the outer wall is 400 mm thick and is made of preactualcube strength is generally SOMPa or so caet concrete blocks. The middle and inner wallswith the maximum- value being 59.5 MFa. so are 300 mm and 700 mm thick respectively and areconcrete is very good in quality and is an consisted of in-situ concrete with a plasticexcellent product.plate placed in between. The concrete grade ofIn April, 1987, there happened severe burst lining is designed as 40MPa and concrete qualitysound from inside the shaft and then some scraps was rated as an product up to the mark.of concrete dropped from the shaft, the whole An overall inspection of lining was made inprocess lasted for half an hour.September, 1987.1t was found from the inspection1251 8

I 233m1 249m *I 269mN E s241mFig.1 Strengthening scheme chart .of Linhuan Colliery auxiliary shaft' 1252

that ruptures of concrete were observed in twoplaces of shaft-.The upper place of rupture,is232.8 m deep in the direction of NE with acircumference of about 7.5 m and a maximumheight of 0. 5 m. The lower place of rupture islocated in 237.5~ depth in the direction of NWwith a circumference of about 7 m and a maximumheight of 1 m. The strata in which the. shaftlining ruptures happened are clay , fine sandand water-bearing sediments of the QuarternaryPeriod, and are about 6 m from weathered zone.Atdestructive part, 9 reinforced bars in thedirection of NW could be revealed.Als.0 there wasa seepage of water. The emergency treatment ofthis shaft was same as that of Linhuan Collieryauxiliary shaft.Zhangshuanglou Colliery Auxiliary ShaftThe mine was designed to have a annual outputof 1.2 mt and put into production in 1986. Thenet diameter of the mine's auxiliary shaft is6.5 m; depth, 565.7 m; overburden thickness, 242m;freezing depth, 315 m. Double-wall lining wasadopted with the outer and inner reinforcedconcrete walls being 600 mm thick each. A layerof compressible polystyrene plastic sheet 20-40mm thick was filled between the

longitudinal coordinate, Z'=O at theborder between overburden and base rock;lining height abov,e the border;-coefficient, p= j6,proportional coefficient between liningsurface shear stress and,liningdisplacement;linink thickness.calculation values show that whv shaftlining destruction happens mostly in Sl,lmmer ismainly because of additional :compressive forcearising from temperature increase in summer.BACK GROUTINGFor above-mentioned destructed frozen shaft,we have, inside shaft, adopted such methods foremergency treatment as hanging channel rings onshaft with vertical rail reinforcing andgrouting inside shaft for control of water etc.,but the problem has not been solvedfundamentally. The shafts still continued tosubside, causing shaft installations twisted andmaking production unable to conduct properly.Thechoices for solving it are placing a lay of insitulining inside shaft or consolidatingformations behind shaft lining thus transformingtheir structure from loose state to halfconsolidatedor consolidated state, the ultimateaim of which is making shaft itself bearingcapacity greater than external forces andkeeping shaft stable. The latter choice is usedto strengehen formations behind shaft lining atpresent.Surface Grouting PlanBoreholes are put down around the shaftlining destruction from the ground surface awayfromtheshaft by means of the directionaldrilling. This method, not having an effect onthe production of mine, is an effective and goodmeasure. But owing to same reason, this plan isstill in preparatory stage.Beck Grouting Inside ShaftThe purpose of back grouting inside shaft isstrbngthening the shaft lining in the vicinityof the border between base rock and overburden;forming a ring of grout diaphram wall of somethickness surrounding the shaft and shaping awater-absent and pressure-absent consolidatedregion, in order to share vertical pressure andhorizontal strata pressure which the shaft bear;blocking up and consolidating original liningcracks and fissures,controlling water-make,water-drenching and water seepage, increasing theentirety and stability of shaft, and sopreventing deformative destruction from furtherdeveloping and happening again (Zeng 1991).Take Haizi Colliery as an example. Groutingbegan with weatherig zoon in base rock. Everylayer of boreholes were put down from bottom totop. Grouting was mainly in the cracked regionand in the formation of grit gravel and mediumto fine sand of Quarternary Water-bearingstrata. For the sake of construction, the firstrow of boreholes were arranged in weathered baserock.As grout is liable to diffuse upWaKdS,it ispossible not only to build a good basis at thebottom, to block the border surface between thebase rock and the overburden,and also to preventlater grout from diffusing downwards.Grouting EffectOver 1000 t concrete was injected in the mainshaft and auxiliary shaft of Haizi Colliery. Thethickness of grout diaphram and water-absent andpressure-absent consolidated region are allgreater than 2 m. Basicly there exists no waterin cracked region of lining after grouting.During one year following grouting, measurementsmade through dlsplacement and strain indicatorsshowed that there was little changes indisplacement, especially in radial displacement.so that by means of back grouting , thedisplacement at ruptured part of shaft has beengot an effective control.CONCLUSIONSAlthough lining cracking has got controlledeffectively after treatment, it is not able tobe eradicated. Various kinds of shaft liningstructure have been adopted in the constructionof frozen shafts in China (Yu & Wang 1988). Whenthe strata which the shafts pass contain gritgravel formation in the lower part, it ispreferable to adopt flexible shaft liningstructure. Placing a layer of bitumen slidingsheet between the outer and inner wall,having anadvantage of being flexible and seepageresistent,helps to reduce the effect of stratasubsidence on interneal lining and to changeuneven loading condition of internal lining.REFERENCESSu Lifan (1991) Damage of lining in frozen shaftand reason analysis. Mine ConstructionTechnology No.1,33-37.Yu Xiang and Wang Changsheng (1988)Structure andstress analysis of seepage resistant liningsinshaftssunk with thefreezingmethod.Nottingham: 5th Int. Symp.of Ground Freezing.Vo1.1,311-318.Zeng Rongxiu (1991) A new technique for backwallinjection for water-bearing sand shaft, WorldCoal Technology No.9, 3-6.1254

PRESSURE INFLUENCE ON PORE CHARACTERISTIC OF FROZEN SOILSWang Jiacheng: Xu Xiaozut Deng Yousheng' and Zhang Lixing'1U.P. Lebedenko'and E.M. Chuvilin''State Key Laboratory of Frozen Soil Engineering,LIGG; .AS*Faculty of Geology. Moscow State UniversityBy using a Poresizer we determined the pore characteristic of saturated frozenNeimong clay under the conditions of constant temperature and pressure and ofconstant pressure and temperature gradient. Results show that when thetemperature is the same, the tatal pore volume and the amount of greater poresize of frozen soils are the minimum under the atmospheric environment, and themaximum under the vacuum condition and increases with increasing air pressureunder the environment of air pressure higher than the atmospheric pressure. ,When the air pressure is the same, the total pore volume of frozen soilsincreases with decreasing temperature and the structures in frozen soils changefrom big ice strips to the small:INTRODUCTIONThe pore characteristic of soils before andafter freezing is of great significance forphysico-mechanical properties and cryogeneticstructure and for mass transfer in soils.Chuvilin, E.M. and Lebeden.ko, 1U.P. (1988)described the formation and classification ofcryogenetic structures of frozen soils. Viyalov,S.S. and Chuitoviqi N.A. (1953) pointed out thatthere is higher pressibility €or the frozen soilwith higher temperature under overburden pressureso as to decrease porosity and to movetexture elements. Wang Jiacheng (1989),studiedthe physical process and crvogenetic structureduring soil freezing. All of the descriptionmentioned above are of great significance €orthe understanding of pore characteristic changesduring soil freezing. So far, there is littleinformation about the pore characteristic offrozen soils under overburden pressure oratmospheric pressures. By using the porometer,we determined the porosity of frozen soils underoverburden pressure and atmospheric pressures.The purpose of this paper is to present the testresults.TES? PRINCIPLE AND NETHODThe sample used in the test is NeimongoliaClay. Its grained size composition and physicalproperties are shown in Table 1.The test conditions can be divided into twogroups: one is pressure loaded on the sample,including atmosphere and vacuum. The sample isquickly frozen and the temperature kept constant.In other words, the influence of water migrationduring soil freezing is negligible. Another ispressure loaded on one end of the sample and thetemperature is kept different at both ends ofthe soil column. i.e., the sample is frozenslightly and water migration and ice segregationoccurred,.. .Sample preparation Ls as follows:For the first group, the saturated clay ballwith a diameter of 5 cm is put into a containermade of stainless steel. Nitrogen is filled intothe container or air extracted from the container(vacuum) and then the pressure is kept constantfor about two hours. Afterwards the containerwith the sample is put Into a lob temperaturecirculation bath to let the sample quicklyfreeze for four hours. Finally, the sample isTable 1.Physical properties of claySample Grained size compositionname >0.05 0.05-0.002 CO.002 limit, X limit, ZG;avitySpe. Surf.area m'/gClay 2.1 50.8 47.1- 32.8 20.4 2.73 28' 125s '

taken aut of the container and air dried at zerodegree centigrade and the porosity is determinedat the side and the center locations, respectively.The test conditions are shown in Table 2.For the second group, the saturated clay iscompacted into a soil box made of plexiglasswith a size of 15 cm in diameter and 5 cm in...height. The atmospheric condition is kept at thetop of the sample with different values of 0,-600 mmHg and 0.2 MPa, respectively. The temperatureat both ends of the sample is kept at 0 and-4 degree centigrade, respectively and thesample is frozen from the bottom upward. After3 days of testing, the sample is taken out ofthe soil box and air dried and the porosity isdetermined.kESULTS AND ANALYSISPore Characteristic of Frozen Soil UnderConstant Temperature and Pressure and QuickFreezinR ConditionsFrom Table 2 it can be seen that there is agreat change in porous volume before and aftertesting under non-pressure condition. The factof the increasing of porous volume in thesample after freezing indicates swelling occursin the sample due to phase change of water. Thefact of porous volume in the center of thesample greater than that in the sides indicatesthat water migrates from the side to the centerduring sample freezing. If the temperaturekeeps at -30 degree centigrade, we can see thatthe total porous volume increases with theincrease of the atmospheric pressure no matterwhere the sample is located and is greater inthe side than in the center. But the totalporous volume is the maximum and greater in theside, too, under vacuum condition. This meansthat the pressure gradient is the main mechanismfor the change of the porous Characteristic inTable 2. Porous characteristic oftemperature and pressuresoils. If the atmospheric pressure is kept at8 MPa, we can see that the total porous volumeof the sample'increases with the decrease oftemperature. 14 implies tha4 frost heave is theessential mechahisr for the change of porousvolume. The changing regularity of the porosityunder different temperature and. pressure conditionsis the same as that of the total porousvolume. The pressure not only changes the totalporous volume but also changes the compositionof porous diameter. With the increase of pressureor in the vacuum condition, the number oflarge porous diameter is increased, but smallporous diameter decreased.Figures 1 and 2 show the,'distribution curvesof porous diameter of samples under differenttemperature and pressure conditions, respectively.From Figures 1 and 2 it can be seen that thecontent of porous diameter greater than 0.03 Dmof the sample before freezing occupies only 3X.But as a result of frTezing and pressure, thecontent of porous diameter in the range from0.03 Um to 10 um is increased from 1X to 15X andthat of diameter greater than 10 pm increased toabout 20X. The changes of the pore characteristicunder the action of the pressure mentioned aboveinduce the recombination and re-orientation ofthe structural elements of soils and result inthe difference of micro-cryogenetic structuresof frozen soils.Photos 1 to 6 show the micro-cryogeneticstructures of frozen soils under differenttemperature and pressure conditions. From thosephotos it can be seen that when temperature iskept at -3O'C and there is no pressure loaded onthe sample, the quasi-layered and networkcryogenetic structure is formed under the conditionsof multi-direction freezirg. With anincrease in pressure, the range of quasi-networkstructure is shrunk from the side to the centerof the sample and ,the branchy structure issoils frozen under constantconditionsNo. Total porous Specif. surf. Porosity of Cant. poro. Temp. Pressurevolume, mlfg area, m'fg x dia. (0.03, X "C MPa. . . ..... . . . .. ... .10-1 0.2091 I1 I 785 31.75 72 C10-2 0.2057 15.369 75 S5-1 0.1802 14.454 26.48 84 C5-2 0 I 2047 13.589 30.55 73 S6-1 0.2094 35.31 11.86170 C6-2 0.228212.426 36.27 65 S7-1 0.2058 d 12.330 34.91 71 C7-2 0.1947 /12.755 32.01 76 S8-1 0.1896 16 782 32.14 92 C8-2 0.2352 12.145 37.89 815-1 0.2320 14.638 36 06 66 C15-2 0.2385 14.418 37.65 62 S-30 0, .-30 4-30 a-10 /s-5 8-30 11.59-1 0.2393 14.677 38 a 93 69 C -30 -600mmHR9-2 0.3888 12.972 48.12 42s12-1 0.3297 21.211 43.51 63 C -5 -600mmHg0.1776 12.980 32.66 97 20 0. .- . . .- ...*wherec - ccnter; s - side.1256

100 10 1 0.1 0.01pore diameter,p mFig.1 Pore size distribution curves underdifferent atmospheric pressure conditionsPhoto 4 X P=11.5MPs T--30BC" 1loo 10 1 0.1 0.01 'porediameter, rnFig.2 Pore size distribution curves under 'diferent temperature conditionsformed instead of the quasi-network. When pressureis kept at 8 MPa, the cryogenetic structurechanges from the network structure to the branchgradually and the network changes from big tosmall and ice straps from thick to thin with theincrease of freezing speed.. ..."Photo 6 X P=8MP> T=-jQCPore Characteristic of Freezing Soil. UnderTemperature nradizt and Constant PressureTable 3 shows che determined result of thepore characteristic of soils under unidirectionalfreezing conditions. It is shown that the porecharacteristic along the soil profile is changedafter testing due to water migration and icesegregation. If there is an overburden pressure,the total parous volume and the amount of porousdiameter greater than 0.03 urn are increased inthe ice segregated zone and decreased in thedehydrated zone after testing.In the case of vacuum condition, the porousprofile is still controlled by water migrationand ice segregation. The overburden pressurecan reduce the differences of total porousvolume and diameter distribution of porousprofile, but the vacuum environment can enlarRethe differences. The cryogenetic structure ismainly in the shape of horizontal layers. Whenthe overburden pressure exists, many vertical1257

Table 3, Porous characteristic of soils frozen underunidirectional pressure conditions" . . . . . . .3-10.183814.28529.023-20.175616 * 54326.34,972-3 cm03-314-10.19850.172714.85113.77552 .QO27.3285953-4top-4014-20.175414.39729.9897middle0.214-31-11-20.17790.21210.207615,56616.29117.18132,1236.3136.95986481bottom0-1 cm1-2 cm-401-30.218716.86436.28732-3 cm-6OOmmHg1-4 0,1940 17.14133.70 86 3-4-4cm1-5 0.1940 15.282 35.28 90 4-5 cm0.1776 12,980 32.66 97 20 0900-1 cm0fissures occur in the profile and the networkstructure is formed.CONCLUSIONS1. Porous characteristic of soils frozenunder constant temperature and pressure conditionsis controlled by the temperature andpressure. When the temperature is the same, theatmospheric pressure gradient is the mainmechanism for the formation of the porouscharacteristic. The total porous volumeincreases with the incr&asing of atmosphericpressure. When the atmospheric pressure is thesame, frost heave is the main mechanism for theformation of the porous characteristic. Thetotal porous volume increases with the decreasingof temperature.2. Water migration and ice segregation arethe main mechanism for the formation of theporous characteristic of freezing soils. Whenthere exists overburden presaure, the differencesof the porous characteristic in freezingsoils become less because of densification.3. Under vacuum condition, the total porousvolume and the amount of greater porous dianeterare greater than those in the conditions withoverburden pressure.4. The cryogenetic structure of frozen soilis influenced by temperature and pressure. Inthe case of multidlrection freezing, the networkstructure gradually changes to' a branchystructure with the increasing of btmoapheric1 pressure. In the case of unidirection freezingwith overburden pressure. the vertical icestraps increase along the direction of thetemperature gradient.REFERENCESChuitoviqi, H.A.. (1986) Mechanics of FrozenSoils, Translated by Zhu Yuanlin et al.Chuvilin. E.M. and O.M. Yazulin, (1988) FrozenSoil Macro- and Microtexture Fo mation,Proceedings of 5th Internatlona4 <strong>Conference</strong>on Permafrost, 320-323.Lebedenko, 1U.P. and L.V. Shevchenko, (1988)Cryogenic deformation in fine grained soils.Proceedings of 5th <strong>International</strong> <strong>Conference</strong>on Permafrost. 396-400.Wan8 Jiachens and Liu Jimin. (1989) PhysicalProcess and Cryogenetic Structure of FreezingSoils, Proceedings of 3rd National <strong>Conference</strong>on Frozen Ground, Press of Sciences, 78-84.1258

PERMAFROST CHANGE FOR ASPHALT PAVEMENT ALONG QINGHAI-XIZANG HIGHWAYWang Shaoling and Mi HaizhenLanzhcu Institute of Glaciology and Geocryology,Chinese Academy of SciencesDue to the thermal. absorption of black asphalt road surface, 60% of the man-madetable of permafrost under roadbed has been decreasing. The formation of the thawcore and uncoincidently frozen ground in the vertical profile acceleratesdifferential settlement of the roadbed and damage of road surfaces. Besides theabove mentioned, either types of frozen ground tend to change, or perenniallyErozen ground disappears completely under some of the, roadbeds.Qinghai-Xizang highway with asphalt pavementhas been used for 7 to 15 years, differentalsettlement appears in most of the roadbeds withinthe permafrost region and damage of roadbedheavily inEluences transportation. There aremany reasons for the damage of roadbeds, inwhich the thaw settlement of permafrost is themost important. IThe authors worked in the study area fromJune to August of 1991. 380 holes at the roadbedand at natural sites were drilled tu contrastthe permafrost changes ac sites between naturaland man-made conditions.1. A CONTRACTED LENGTH OF HIGHWAY THROUGH THEPERMAFROST REGIONThe investigation along the highway in the1970's showed that the northern boundary ofisland permafrost was at K 2877+600 of Xidatanand the southern boundary of continuous permafrostwas at K 3424. north of Anduo. From Anduosouthwards to No.125 highway maintenance station,permafrost occupies 20 per cent of the length of90 km of the highway within the island permafrostregion. Exploratory data for this time demonstratedthat the southern and northern boundaryof permafrost have been changing indeed, especiallythe southern boundary uf continuous permafrostwhich has moved northwards 16 km. Within theisland permafrost south of Anduo, permafrost wasonly found in the roadbed at an accumulated.length of 2 km. The length of highway throughthe permafrost region at present that has beencontracted is about 18 km.2. TYPE OF FROZEN GROUNU UNDER THE ROADBEDUnder natural conditions, with the exceptionof a few sections, frozen ground is mostlycoincided in the vertical profile for bothla tera 1s of the highway wit :hin is land uerma frostand it all coincided within the continuous permafrostregion. The holes on the roadbed fromXidatan to Anduo (K 2880 to K 3424) indicatethat the type of frozen ground under roadbed isdifferent from that under natural conditions.Frozen ground under the roadbed all discoincidedwithin the island permafrost region, and mostof them also discoincided.2.1 Coincided Frozen GroundThe aan-made table of coincided permafrostunder the roadbed is 2.8 to 4.2 m deep, thisdepth is close to or less than that undernatural conditions. Ice layer with soil or icerichsoil occurs below the table of permafrost,it is distributed on the hills with high elevation(Table 1). Where mean annual air temperatureis below -h.O"C, the depth of seasonal freezingunder the roadbed exceeds or equals the depth ofserisonal thawing each year. Frozen ground iscoincided in ,addition. Coincided permafrost alsoappears in the low roadbed (height 150 cm). Thedestroyed asphalt pavement section, whereasphalt pavement zero fill cut sections, andsubgrades of part of culvert are coincided permafrostit is estimated that distribution lengthof coincided permafrost approximates to 80 km,occupying 15 percent of

~~Table 1. Mileage and distribution section ofcoincided frozen ground under roadbed alongQinghai-Xizang highwaySection Road mileage Length (km)Kunlun Shan K 2893-2904 11Kekexili Shan 1 K 3009-3020 11Qusuihenan Shan K 3035-3037 2Fenghuo Shan K 3064-3080 16Kanxinlin K 3171-3176 5Tanggula Shan K 3316-3348 32Total length 77was formed between the seasonal frozen layer andthe man-made table of permafrost whose depthchanges from 4.5 m to 10.1 m. The thickness ofthaw core changes from 0.2 m to 6.6 m, most ofthem are 0.5 to 2.O'm. The development of thawcore is related to the constraction duration ofthe asphalt road, properties of filled soil,water content and type of permafrost under road.The section, which spends shorter constractiontime, has more coarse grained soil and lesswater content, has a thicher thaw core.3, CAUSES OF TAE THAW CORE FORMATION AND ITSHARMFULNESS"3.1 Causes of FormationSince the asphalt pavement has been constructed,the thermal condition under the roadhed haschanged obviously due to the specific solarradiation and thermal balance conditions on theplateau and thermal absorption of the black roadsurface. Contrasting with the natural conditions,the difference between mean annual surface temperatureon the asphalt pavement and meanannual air temperature is hoc, which induces thefrozen ground in the roadbed to thaw 20-30 daysearlier and freezes 20 days l'ater. Annualthermal input on the road surface exceeds outputwithin the areas with a mean annual air temperatureof more than -hoc. Tn the above area, thedepths of the seasonal freezing and thawing are4.2-4.5 m and 4.5-6.0 m, respectively: and thelayer of seasonal freezing disappears completelyduring the period between early July to lateAugust. The heat during August to October entersinto the roadbed and is accumulated, whichinduces the man-made table of permafrost to godownwards, causes the frozen ground to discoincidein the vertical profile and finally makesthe thaw core form.In recent decades there has been a trend ofclimate warming on Qinghai-Xizang Plateau,which is advantageous to the formatjon of thaw 'core.3.2 HarmfulnessIn the forming process of thaw core differentialsettlemeit if- roadbeds developed graduallyand finally induces the damage of road surface,lhe harrnfulne-ss of thaw core to the highway isdifferent due to differences of the type offrozen ground and the lighologic distribution.(1) In the road section of Xidatan. becausethe roadbed is filled with coarse grained soilwith little ice and water tontent althoughthere is a thick thaw core, the total settlementand deformation are still little.(2) Large total settlements and seriousdamage of toad surfaces emerge in the roadbedfilled with lacustrine sediment or fine grainedsoil with an ice layer with soil and ice-richsoil. Such as those in the basin of Kunlun Shan,in both banks of Qingshui River and on the northfacing slope of Kekexili. After the thaw corein the roadbed is condensed repeatedly, thethickness of the thaw core increases slowly withthe development of the man-made table of permafrost.So it is difficult for the roadbed toattain stability.(3) The man-made table of permafrost underthe roadbed is like a pan in which water isaccumulated perennially. The water accumulationmakes the man-made table of permafrost godownwards and the heat balance system in theroadbed becdme sophisticated. On the other handthis also causes a hidden problem in that anexpansion mound emerges in the road surface inwhich water flows continually into the thawcore and pressure water forms in the thaw core.At present, most of thaw cores under road alongQinghai-Xizang highway is forming and developing,and it would spend a long time for them to berelatively steady.4. PERMAFROST CHANGES UNDER THE ROADBEDIn recent decades, degradation of permafrostwithin the Qinghai'-Xi&ng Plateau is under x-aywith world wide warming, esp. in the margin ofthe plateau with island permafrost. The regionaldegradation, the economic activity and sometechnical factors aggravate permafrost degradationunder the roadbed. There are apparentindications that permafrost degradates an,ichanges from quantity to quality in some siteswhere vegetation has been removed and h ith X\IIIVwater pits. Such as in hoth sides of thc I - ~ ~ ~ I ~ ! ! ~ P ~4.1 Permafrost Disappears and Talik EX~~III?~There are various talik salonR the hi

Exploratory data for railways in the 1970'sdemonstrated that r,?rmafrost develops in terraceand high flood plains with the exception of theriver bed on the No.1 terrace of Beilu River.The permafrost was 12.8 m thick and continuous(No.1 hydrogeological team of Qinghai Province),The changes are as follows:In Xushul River permafrost was not found attwo of the 18 m deep coreholes on the roadbedof both sides of bridge; but, permafrost of 8.lmthick only occurs on the high flood plain.Permafrost was also not found at two coreholeof 15 m and 18 m depths on the roadbed of bothsides of the bridge in Beilu River, it is 9.5 mthick in the high flood plain.The above facts show that thin permafrostunder the roadbed has thawed completely becauseof asphalt pavement influence on it. However,permafrost develops under natural conditionssuch as in high flood plains.4.2 Types Change of Frozen Ground Under RoadbedUnder natural conditions, ground ice isdistrihuted widely within depths of 1 to 2 mbelow the table of permafrost and decreasesgradually with the Increasing of depth. Atpresent, the depth of most of the man-madetable of permafrost under the roadbed along thehighway is 1-3 m lower than that of naturalones. The table of permafrost decreases graduallyto induce ground ice near the table to thawpartly or completely. By comparsion of statistics(study group of Qinghai-Xizang highway, 1983).The proportion of frozen ground with high icecontent has decreased under the roadbed alongthe highway (Table 2). The type of frozen groundbelow the man-made table of permafrost appearstu change in some sections of the roadbed, thereis a trend that ground ice content is gettingsmaller, which is advantageous for roadbedstability.Many engineerings along the highway have ,demonstrated that the influences of human act:ivityand technical factors on the permafrostenvironment exceed the influence of climatewarming, esp. the degeneration under the roadbedwithin warm permafrost areas which transformsfrom qualitative to quantitative. Many deformationsare not reversible. Damage of permafrostenvironment in the both sides oE the highwaymay have serious consequences. After the asphaltpavement wa6 constructed it is inevitable thatpermafrost under roadbeds thaws year by year byadding to the height of roadbed within mostareas where mean annual air temperature is above-6°C. In order to gain an ideal effect, generalmethods, such as to make alterations of thematerial and colour and the property of heatabsorption of the roadbed, to improve on theinsulation in the roadbed and the drainage in.both sides of roadbed and to protect the permafrostenvironment, must be adopted for therenovation projects along the highway.REFERENCESWang Jiachen, Wang Shaoling and Qiu Euoqing.(1979) Permafrost along the Qinghai-Xizan?highway. Journal of Geography, 34(1): p.18-22.Study Group of Qinghai-Xizang Highway, (1983)Distribution laws of frozen ground with. highice content along Qinghai-Xizang highway,Proceedings of.Second National <strong>Conference</strong> onGeocryology, Gansu People's Press: 43-51. -. ITable 2. Type change of frozen ground with high ice content north of Tanggula PassSection of.amount 46.6 52.9 60.2 19.7 53.4in 1979"ercentaae -amount 39.7 50.8 43.7 32.3 16,5 41.8 21.5 36.7in 19911261

ASTUDY ON PREVENTING FROST HEAVE OF THESHAFT-TYPE ENERGY DISSIPATORWan?, ShirongWater Conservancy Bureau of Panjin City,Liao Nin Province, ChinaIn the seasonal frost ground area the breakages of traditional open-type energydissipator by'frosr heave are very widespread and serious. The new shaft-typeenergy dissipator was adopted. The complex prohlem of frost damage was solvedsatisfactorily by means of its overall arrangement and construction (no need ofspecial measures to prevent frost heave). 'The overall arrangement is quite anew type and made of shaft (including shaft dissipating force) and t.unnel. Theshaft and the tunnel adopt a thin-wall structure of a crosy-section of circular,arch, egg-shaped or U-shaped, these shapes are advantage to equalize and todecay the force of frost heave. This type is simple, convenience, economic andsafe in structure. It is a new-important way to prevent frost damage in.middleand small hydraulic engineering.INTRODIICTIONMany open-type fal.1-water structure (includingsteep gradient) of hydraulic cunstructure ofirrigation channel were destroyed by frost.darnagc because its foundation are not at samelevel, high or lpw, anti the amount and theforce of trost heave are non-uniform. One sideof tha open-type rctainjrig wall was filled withsoil, so it was easily toppled down and brokenup by horizontal force of frost heave. The opent.ypeenergy d issipator, which was built out: oftraditional shaft-spillway, has same weakness ofno1 preventing frost heave. Tn recently tenyears, the engineers and techniciance of waterconserv:irlcy project set up crcativ.ely the shafttypeenergy dissipator in cntrance (instead ofthe open-type energy dissipator out of thetrndilional shall-spillway). For widely used inwater ~onservancy prc~ject, it gains goodresults (Wang Shirong, 1990).THE BASICALLY ARRANGING TYPES OF TAF, SHAFT-TYPEENERGY DISSIPATORWe research the shaft-type energy dissipatorsirl over 200 places of Heilongjiang, Jilin,Liaonin, Shuanxi Province and Inner MongoliaAutonomous Kegi,on. There are a lot of types,but hasic type was shown in Fig.1, t.he directionnf water to be flow in entrance changes down tothe shaft dissipating energy to consune energy,the horizontal tunnel is lower to form as apool decaing force so that the bottom currentconsume energy. In engineering practice, thedissipating energy.with shaft is main, sometimesB supplementary means is to set up barrier orhaffle rlecaing force for dissipating waterenergy. The shaft-type energy dissipator iswidely used in the sluice gate with fallingdifferentia, the culvert, the emissary of smallreservoir and the spillway besides.THE STUDY ON FROST HEAVE OF THE SHAFT-TYPEDISSIPATORENERGYWe made investigations on the shaft-typeenergy dissipator to prevent frost heave in overZOO places. From 1984 to now the experimentalstudy of the shaft-type energy dissipator wascarried OUL in Laboratory of HeilongjiangHydraulic Training School (especial experimentson the ratio of dissipating energy and on frostheave) and the outdoor observations were went onin experimental field (the author took part inthe experiments and the outdoor investigations).A lot of scientific data was gained. The crosssectionof shaft-type energy dissipator iscircular, and the cross-section of the horizontaltunnal has complexly curvilineal shapes, such ascircular, arch, egg-shaped and "U"-shaped (WangShirong, 1990). The horizontal stress, thetangential stress and the normal stress of frostwere influenced by shapes of structure, thereforethe concrete calculation of these stressesis complex. Because the geol.ogic condition isnot the same in different region, so it greatlyinfluences on frost heave, therefore it isdifficult to calculate the stresses and theamount of frost heave with a united-constantformula. For this reason, the calculating methodwas not detailed here. In engineering pratice,experiments of the *haft-type energy dissipatorshow that the characters of preventing frostheave are as follows:The Overall Arrannement to be Rinorous isConvenient to Prevent Frost HeaveIn comyarigon with traditional shaft-spillway,the open-type energy dissipator out of shaftwas superseded, the combination of energydissipation of shaft with dissipating energy ofthe bottom current in tunnel was taken andlongitudinal length was decreased. The shafttypefall-water structure has only a shaft withlittle cross-section to decay force (in Ftg.l),1262

einforcement of alongitudinal scction/orizontal ribs of reinforcinnsteel A 1ongi;udinal ribs of \reinforcement of ahorizontal section"100Fig.1 Liuhe fall water structure of Mulan County in Heilongjiang Provincethis shaft replaces open-type multi-pools decaingforce of multistage fall-water structure and thusthe longitudinal length of the fall-water structurewas reduced to half in comparison with thetraditional multistage fall-water structure. Ithas many advantage, such as, the flat area islittle, the parts of structure is a few, the .frozen part and frozen area are little, theoverall arrangement is concentrated and simple,so it is convenient to prevent frost heave(Wang Shirong, 1990).offset of symmetryhorizontal farceoffset frost force in circularlayer of frozen soilIt is AdvantaRes to Prevent Frost Heave thatthe Foundations are in One LevelIt is well stable and advantageorls'to preventfrost heave that the foundation of the entranceand the outlet of the shaft-type energy dissipatorare in one level. When the foundationsare in one level, the soil is uniform, the watercontent is the same and the amount and thestress of frost heave are uniform; It is convenientto keep temperature with water to preventfrost heave for whole foundation and to eliminatedamage of normal stress from the foundation.Because the shaft and the horizontal tunnel areburried in soil, the frost depth is shallow andthe stress and the amount of frost heave arelittle.The Entrance of Water Adopts the Shaft toAntistress of Frost HeaveIn Fig.2, the wall of shaft is circularsymmetry to beat the force of frost heave,horizontal stress is equal in magnitude and isopposite in direction, the forces' transmitedthrough the circle of shaft are equilibrated.The soil around the cylinder of shaft is instate of circular stress, the hard frozen soilwas compressed by himself and offsets the forceof frost heave. Actually the soil around thecylinder of shaft is not frozen immediately,but one layer by one, the first layer is frozento form frozen soil ring, it offsets the forceformed by next frozen layer.In other words, the cylindric frozen wall inevery layer offsets the force of frost heave ipnear layer, the concrete shaft can not bedirectly damaged by frost heave.of frost force compressed by circularfrozen soil layerFip.2 Schematic drawing of the equilibrium orthe decay of frost force in shaft ci.rc2e ofshaft-type energy dissipatorThe Horizontal Tunnel adopts AdvantageousStructure for Offseting and DecainR Force ofFrost HeaveThe horizontal tunnel in the shaft-typeenergy dissipator adopts advalltagcous structurefor preventing frost heave, such as crosssectionof circular, arch, egg-shaped and U-shaped. Fig.3 shows the arch-shaped horizontaltunnel in the shaft-type enerRy dissipator. Inarch action, earth of arch way is in state tobear arch-shaped stress, the arch-shaped hardfrozen soil was compressed by each other and tooffset force formed by frozen soil. Thereforethe frost force does not directly acts the archway. Actually the frozen depth of soil aroundthe arch is not very thick, but arch soil isfrozen layer by layer, first frozen layer formedthe frozen arch wall to offset the frost forceformed by next layef. The bilateral surface ofarch is symmetry to bear horizontal force offrost heave, its magnitude is equal and thedirection is oppsite, these forces keep in1263

equilibrium through arch which has good characterof compressive resistance. The tangential forceof frost heave in a side of arch slips along theangularly curved surface and decays. It keepsthe temperature of foundation with water toprevent frost and to dispel the damage of normalforce in the foundation. A s the same, horizontaltunnel in the shaft-type energy dissipator hasa cross-section circular shape, egg-shaped orU-shaped etc., all of shapes has above advantagesto keep in equilibrium of frost force or todecay frost force."-I-Fig.3 Schematic drawing of the equilibrium orthe decay of frost force in horizontal arch ofshaft-type energy dissipatorCONCLUSIONSIt makes known the advantages to preventfrost heave with the shaft-type energy dissipatorthrough the theoretical study and theengineering application for ten years, Itsoverall arrangement is scientific rarionality.The main part is a cylinder of shaft, and thehprizontal tunnel adopts thin-wall structure ofa cross-sec,cion of circular, arch, egg-shapedor U-shaped shapes. These shapes have advantagesto bear force equilibrium. This structureeconomizes the engineering amount and itsapplication is safe reliability for long period.The style of dissipating water energy hascreative characters. The shaft to dissipateenergy is main and combines with dissipatingenergy of barrier, baffle and the bottom currentin tunnel. The vertical arrangement made upmultirepeated, synthetic consumed energy,although the space of dissipating energy islittle, but the effect of dissipating waterenergy is good. This type of the shaft-typeenergy dissipator has developmental future-REFERENCESWang Shirong, .(1990) An application of compositeconstruction energy dissipator to smallhydraulic engineering in seasonally frost'4 "regions, (Journal of Glaciology andGsocryology.) Vo1.12, No.1.Mans Shirong, (1990) Analysis of preventingfre'ezing property with shaft-type enerpy.dPssipator, (Journal' of Engineering apdPermafrost), 1990 No.2.Wang 'Shirong, (1990) Developmental symmarizationO9:th.e engineering of the fall-water, ' structuqe in China, (Society of Heilongjiadg ',' .Hy,drdulic Training School), 1990 do.4., .,. ,1264

FIELD EXPERIMENT RESEARCH OF WATER AND HEAT TRANSFER WITHINFREEZTNG AND THAWING SILT LOAM UNDER FIXED GROUNDWATER LEVELSWang Yi, Gao Weiyue, Zhang LianghuiWater Resources, Research Institute of Inner Mongolia, HuhhotJnner Mongolia, 010020, ChinaIn order to research the accumulating and declining law of water content within the active layer offrozen soil in the Hetao Irrigation Area of Inner Mongolia, the heat and water transfer experiments werecarried out in the site. The results show that the frozen depth is less than the depth where the groundtemperature is 'C. During freezing period, the water content at the same depth changes in "S" form witha increasing time and insreases fast and slow in earlier and later periods respectively. Change Curve ofwater cantent has a obvious peak value. During thawing period, the thawing speed is from slow to fastwith decreasing groundwater levels. The dissipating curve of water content is shown in opposite "S"form.INTRODUCTIONThe Hetao Irrigation Area of Inner Mogolia is a seasonallyfrozen area. Before frqzen, the groundwater level is high and thewater content is large. After the soil has thawed out, sowing cannot be done on time because of a large amount of water in the soil.Sa it is very important to research the water and heat transfer lawsof frozen-thawing soil under different groundwater levels for irrigationarea management and agriculture production.A lot of research work about water and heat transfer in freezingsoil has been made (Xu Xuezu, 199l),which mainly include theinfluences of soil texture, temperature gradient and freezing speedon water transfer, large part of the experiments were madc in laboratories.There was a little information on the field experiment.Pukuda (1985), Kuntsson(l985) respectively published field observationresults of frost heave under constant and unconstantgroundwater levels,Gao Weiyue (1989) gave forth field observinginformation about groundwater inflow during fretzing period.Unti1now we have not seen more detailed information about time andspace changes of groundwater accumulation during thawing periodand especially about the process of water dissipating within frozensoil layer during thawing period.In order to investigate laws of water and heat transfer in theHetao Irrigation Area during frozen-thawing period, the experimentsof water content accumulation and dissipating duringfrozen-thawing period were carried out from Nov. 1991 to Apr.1992.of 1 m2 in opening area are installed into the natural soil underthrce fixed groundwater level i.e. 1.5 m, 2.0 m and 2.5 m. The soil isput into the buckets. A layer of sand filter of 30 cm thick is put underthe bottom of the soil column. The top of the bucket is at thesame level as the natural ground surface.Table 1. Experiment soil physical propertiesDry unit Saturated Total saltGravity Grain Siz(mm)weight watcrcantent content(E/ (%I (g/ IOOgsoil) >O.OOS

11 12 1 2monthFig.1 Frozen depth O°C depth vs. elapsed timeFrom March 10, the frozen soil thaws downward from the groundsurface and upwards from the frozen soil bottom and the thawingsoil reaches together on the end of April.Soil freezingIn Fig.], the change curves of frozen depth with changing timeand the depth where the ground temperature is o "C are drawn. Atthe beginning of freezing from Nov.20 to Dec. 5, the soil surface isfrozen at night and thawed on day. So a thin frozen soil layer isformed within the surface temperature decreases steadily below 0°Cand the frozen soil layer thickens downward. This period is calledas the steady freezing period.During freezing period, the 0°C soil depth is larger than frozensoil depth. The deeper the depth is, the larger the distance betweenuem is. The reasons are: first, the surface energy of soil grain actson the water of' soil; second with the increase of frozen depth, thegroundwater with salt flows continuously towards frozen layersand the salt content of unfrozen water will be concentrated; so thebelt of more salt concentration is formed in the moist belt of frozenfront. The salt in the concentrated belt moves up with the unfrozenwater and at the same time moves down because of the concentrationgradient. The results are, with the freezing front movingdownward, there are more and more salt Concentration at thefrozen front and the frozen temperature becomes low and IowJhisis a reason that frozen depth is gradually less than soil depth ato°C. The second reason is that the soil temperature is not the sameat different depth. The deeper the soil layer is, the higher the soiltemperature is and the smaller the temperature cbange is.Soil thawing, During thawing in spring, freezing and thawing of soil surfacetake place alternately on day and at night. Affected by NorthMongolia and Siberia cold current, there is also intermittence freesing-thawing. A unsteady thawing layer of 20-30 cm thick isformed in the soil surface. At this time, the temperature at the bottomof the frozen layer begins rising and the temperature of thewhole frozen layer is almost the same. This period is not steady.When the temperature is higher than 0°C. the frozcn soil layerthaws grasually and this period is steady. Seen in Fig.2, the thawingspeed is not the same at different water level. The thawing speed isslow at the high groundwater level. This shows that in high120 I3 4 5monthFig.2 Thawing depth vs. elapsed time under differentgroundwater levelsgroundwater level, the water accumulation quantity within frozenlayer is large and the ice crystal almost occupies whole soil pores; inlow groundwater level, water accumulation quantity is less and icecrystal occupies part of soil pores. During thawing period, whengroundwater is high, the thawing water within the top of soil makesheat energy go down to frozen soil, this is called the infiltratingthaw. When the groundwater is low, thc thawing water affected bygravity seeps down along the pores. At the same time the thawingwater releases heat to heat the low frozcn soil and widen the waterroute.From the above experiment data, the thawing ways are not thesame at the different water levels. The way at high groundwaterlevel is that in energy transferring and the thawing speed is slow.The way at low groundwater levels is that in energy and mediumtransferring and the thawing speed is fast.Water Accumulation within Frozen LayersAs the soil is freezing, the freezing front moves downgradually. The curves of the water content within different soil layersare in "S" form. At the beginning, the water content increasesfast and later the water content increases slowly and finally watercontent does not change (see Fig.3) .The water accumulation within the soil layer has two sources.One is the water content within frozcn layers moving upwards fromthe bottom with higher temperature to top of layer with lower temperature.The other is the groundwater moving from unfrozen milto frozen soil. At the bcginning of the freezing, the groundwatermoves from the unfrozen layer to the frozen layer. The water routeswithin tho soil are unblocked. With the freezing time increasing, thefreezing front moves downward a ccrtain distance, at frozen layer,wafer routes are blocked gradually by ice crystal. The groundwatermoves to frozen layer less and less. Therefore, at thepeginning, thewater content accumulation mainly depends on the groundwatertransfer. The speed of water accumulation is fast. Later, tht watercontent transfer mainly depends on unfrozen water within the soil.The water content rises from the low layer with high temperature totop layer with low tempcraturc and the speed is slow. Thus the "s"shape of water arsumulation within frozen soil is formed.Fig.4 is the sections of moisture content at different time anddifferent water levels. In Fig.4, the water content accumulationwithin frozen layers has obviously peak value. With pushing down,1266

3630 -I "I 1.5m1 1 - 1 I 1c1 1 1 2 1 2 3 4 5monthFig.3 Water content given layen VI, elapsed time during freesing and thaftiiiu20 30O h . . .Water content (%)20 30. . .Nov.20-20 30plained as follows: The thawing water within thawing layer (layera) transmib heat to low frozen (layer b).The result is that layer bthaws.lcc within layer b transform into wa~r.The.layer volume decreases nine perccnt and vacuum is formad (Wang Jiacheng, 1992).Because of the vacuum soak, surplus water content within layer amoves downward to layer b. Later, tha frozen-thawing border surfaceshifts down.Thc water content dissipating within layer + d&pcnds on yporizatioq and gravity mpaga.From a,bve.it can beStcn at the beginning of thaw, tha water shifting maydepends onvacuum soak and the aped io fast; Later* the thawing water shiftingmainly dewnds on vaporization and gravity -age and thesped is slow,,The moisture content sections at different water levels anddate are drawn' in Eig.5, As can be seen,. the water contmt dissipatingam not the same at the different-water lqveb, When thegroundwater level is 1.5 m, the water content within lower frozensoil layer does not change obviously and the groundwater leveldose not rim obviously. When the groundwater level is 2 .h and2.5m,from the thaw beginnin& the groundwater rises and water08 11020 30water content (%)20 30 20 3040-80 * Y120 [ISrnIJm.102.0m2.5m1 20Fig5 Water content profiles under dflerent groundwater levelsduring thawingFig.4 Water content profiles under different grounwater levelsduring frwzingthe peak value increases and the peak value location shifts down.The peak value gets tht biggest valua when the frozen period ends.The location of the biggest moisture content is not the same atdifferent water levels. When the groundwater level is low, the locationof the biggest value moves down and the value becomes small.This is the result affected by seasonal frozen depth and originalmoisture content.Water Content Dissipating whthin Thawing LayerWhen the daily average temperature rises above 0°C. theground surface and the bottom of the frozcn layer begin tothaw.The water content within thawing layers dissipatinggradually.In Fig.3,the curve of thawing water disspating with thetime at different layer are ahawn in roveme 'S'. This can be ex-wntent within lower frozen soil 1aycr.dscmase. Thia shaws that thesurplus thawing water dissipating at high groundwater level mainlydepends on the vaporization. When the groundwater level is lowthe dissipating of the thawing water mainly depends on thevaporization and soak.That is ,when the groundwater level is high,the form of the thawing is in energy transmitting;whcn the level islow, the form of the thawing is in mcnrgy and medium transmitting.CONCLUSIONThe temperature decrease$ along the soil depth and the saltconcentration at the freezing front make the frozen depth be smallerthan the depth where the soil temperature is 0"C.The deeper thelayer is ,the longer two distance is.The thawing speeds are not the same at different groundwater1267

levels. When the groundwater level is high, the form of the thawingis in energy transmitting and'the speed is dow. When the water levdIs deep, the' form of the thawing is in energy and medium transmitting'andthe speed is fast.Duriirg Freezing period, the curve of the water content withineach frozen layers is in-%" shape. kt the early period, tne water ac-&mulation mainly depends on outside water source and increasefast.Later it depends ori inside unfrozcn water and increase sbw.The cWe of w&er cbntent along'the different depths,has the obviouspeak 'Viluc The deupcr the water level is, the biaet the peak, .value is and the location of the peak value shift down.The dissipating curve of the water content within thawing lay-'crs is in reverse %' shape. At the curly period. the dissipating mainlydepends on the vaduum soak and swd'is fast. Later, it dependson vaporization and gravity seepage and speed is slow.REFERENCESFukuda,M. and Nakagawa,S.(1985) Numerical Analysis of FrostHeaving Based on The Coupled Heat and Water Flow Model,hoc. of 4th Interdl. Sympo.on Ground Freczing.71-75Gao Weiyue and et al. (1989) The Observing of Groundwater Inflowduring Freczing Period, Journal of Glaciology andGeocryology,,Vol.l4, NO.^,^ 137.Knutsson,S., et al, (1985) Andy;& of Large Scale Laboratory andin,&itu, Frost Heave Testh, Proc.of 4th Internl Sympo. onGroud Freezipg, 65-70.Wang Jiacheng and et ai. (1992) Experimental Study on Conditionsfor Ice Formation of Saturated Sand in Freezing and ThawingCyclcs, Journal of Glaciology and Gcocryology, Vo1.14,N0.2..k01-106.Xu Xuezu and et al. (1991) Experiment Rekarch of Water TransferWithin Frozen Soil, Science Publication House, Beijng. '1268

THE EFFECTS OF GOLD MINING ON THE PERMAFROST ENVIRONMENT,WUMA MINING AREA, INNER MONGOLTA OF CHINAWang Yingxue R Tong BoliangLanzhou Institute of Glaciology & Geocryology,Chinese Academy of Sciences, Lanzhou, ChinaIt is found that mining is the main factor which caused the changes of the, permafrost environment. After mining, the area of forest ,decreased, the soillayer was reversed and the fertility decreaded, the ground surface and groundtemperature and water content in the soil layer decreased, all of which resultedin the increase of the seasonally thaw depth, the deterioration of permafrost,the disappearance of swamp and a tendency of aridness and desertification, andfinally affected the ecosystem and caused the vegetation to be in an inversion ,succession. We suggest that the mining and environmenfal administration mustbe consi'dered at the same time and propose some suggestions "abdut recovery andadministration of the environment after mining.INTRODUCTIONWuma mining area is situated an the northwestslope of Da Hinggan Ling dominated bypermafrost. Topography of the region is mainlylow ridges wit.h an altitude of about 450,111 anda relative undulation of 200-400 m. The miningarea is lorated between the Yilijiqi Rivervalley, which is one branch of the Eerguna River,and Lajigan valley. The width of the valley is100 m to 800 m. In the lower terraces and lowlyingland there is seasonal running water andthe region is heavily swampy. The minlng sitehad been mined and the veRetation has beendestroyed. Today, the site is prevalent withsuccessional forest which are scattered in thevirgin forest. The vegetation is mainly Pumila,'Betula dehuricaea, Salix hypoda and Populus 'devidiana etc. Except for some sand dunes andbore-holes, the ground is covered by forest andherbace,ous whose canopy density is 60-909.Eriophorum vaRinatum and Carex are exuberant inmarsh-land. The depth of Carex layer is 0.1-0.3m.Because of its high latitude location and thehigh barometric influences of Siberia andMongolia, the mining region is the coldest in.China. From the correlation analysis of 10years of information (1977-1988) of EerRunayouqimeteorological station and the observed informationin Wuma mining area in warm seasons, themean annual air temperature in Wuma mining areais -4,4'C, and the coldest temperature is -46.2-C.The frozen period is about seven months and thethaw index is 1829.3 ('C.d). The total radiaition in warm,season (May-Sept.) ,is 'only 59.7- \67.4 kcal/cm . The annual precipitation is400-500 mm among which 70-907 of it is fromJune to September. The depth of snow cover isabout 20 cm.,The region belongs to a fault-block valley ofMesozoic age. Granite of the Haixi period formsthe bedrock and appears on the tops of the ridges(CUO Dongxin, 1981). Quaternary desposits arewidely distributed. The depth of the unconsolidatedlayer of Quaternary sediments is 4-9 m.The lithology of them is mainly humm'soil qndsandy clay. In the valley, the upper 0.9 m isblack and grey humus soil and peat,, the lowerpart is greyish-yellow sandy clay, gravel-sandand weathering layer (Figel).From the information of physical prospectingin this region and the information of theadjacent regions of Gulian (Guo Dongxin. 1981)and Eluosi which ha,ve the same latitude 0: the0.0-0.3 m: Humus layer including little-0.7-1.2 m: Sandy soil layer with 50-60%clay 30-40% sand. It appears yellowishdrey,1.0-2,5 m: Sand-gravel layer with 55-65%gravel, 30-40% sand and 7-10% clay. Itso appears yellowish grey. The diameter-isfrom 0.25-10 cm. Roundness ig good.\There is thick ice layer with thicknessof 0.5-1.8 m under the depth of 1.8 m.2.5-4 m: Weathering layer with 80-90%4 *Aboelastic rock, 10-15% sand and 5% clay+,!+ +at depth of 0.3-1.5 m. It is mass$. constructure -with rich ice.-+++++Bedrock: Granite with rich crevas,se.+\It is frozen.Fig.1 The Bar-graph of Lithologic Characteristicsin Wuma mining Areastudy area, we estimated that the depth of permafrostin this region is 50-80 m. The maximum1269

L C L " 0llllY-l -lUyrr.c"u= IJ -&." I.> b. &,ICobserved temperature at a depth of 2.5 m is -3.0and 0.8"C in swamp and non-swamp sites respectively.The seasonal thaw depth is 0.3-0.9 mand 0.5-1.0 m in the eastern footslope and swamprespectively, which is the minimum in the region,0.8-1.0 m in the front of the diluvial fan, 1,2-2.0 m in the,eastern slope and 2.5-3.5 m in thewestern slope. Eastern and western slopes arerespectively 1.2-2.0 m and 2.5-3.5 m.123456I80.4 lunI.W ~ L ~ : I cullccliL, n~uuuu UIIU 81uunu sul~act. remperaturesin different geographical positionsin stripped and un-stripped sites. The resultsare as following.(1) After trees, herbaceous and bryophtacovers are stripped, a lot of water appears onthe ground surface. One year after the stripping,water content in the frozsn layer reducedabruptly in comparison with the un-stripped site(Fig.3). The water contents of humus and peatdecreased 50% and the content of sand-loamdecreased 30-4OX. The main reason is that thearound accepts a large amount of radiationdirectly. In the stripped site, the radiationintensity increased more than 100 times incomparison with natural sites only in the mosscovered site and non-moss covered site, thedifference of temperature at the depth of 20 cmis 11-15'C (The Branch of Siberia Institute ofPermafrost, Russian Academy of Sciences, 1988).Relative wntcr coirtcnt (%)200Fig.2 The Distribution of the Maximum SeasonalThaw Depth in Wuma Mining Area, ChineseMongolia1 - River talik; 2 - Swamp (0.5-1.0 m);3 - Alluvial flat (1.3-9.0 m);4 - Western'footslope (1.3-0.9 m);5 - Eastern footslope (0.3-0.9 m);h - Front of diluvial fan (0.8-1.0 m);7 - Middle parr. of eastern slope (2.5-3.5rn);8 - Middle part eastern slope (1.2-2.0 rn).Thick layered ground ice is well-developed inWuma Mining area. In swamp, the thickness ofground ice is 0.3-0.5 m within the depth of 0.5-1.0 m. In 'same section. there is massive groundice whose thickness is between 0.5 and 0.8 m andthe length is more than 100 meters stretchedalong the valley of the region at the depth of1.8 m under the first terrace of the westernbank of Yilijiqi River from a mining crosssection of the western side of Wuma mining area.In the Ice there are thin clay layers betweenpure ice lavers which form a layered structure.From the abandoned pits ground ice is also foundwh.ich appears to be pingoes with diameters ofabout 2-4 m.THE EFFECTS OF MINING OR 'THE PERMAFROSTENVIRONMENTWuma mining area is about 300 m wide and20300 m long. Around the area there are nineother mining areas. All of them are stripmining.mechanized mining has destroyed largeamount of vegetation and the layer of soil.The mining includes two periods of prestrippingand mine-choosing. The pre-strippingmeans to strip forest, the layer of herbaceousplants. bryophyta, humus and peat before thenext years' mining. In the process of layer-to-Fig.3 The comparison of water' content of frozenRround in pre-strippedFrom the observed information of stripped andun-stripped sites during July (one year afterthe stripping) and un-stripped sites in Wumamining area. to August of 1990 (Table l), wefound that one Year after the stripping, theair temperature, ground surface and groundtemperatures and daily ground surface temperatureranges in stripped sites have increasedobviously in contrast with un-stripped sites.It is evident that the veRetation and peathave a strong protection effect on the permafrost.When they are removed, the layer of sand-gravelis exposed and has a higher temperature,simutaneously, the water on the ground surfaceand in soil and the water from the thawed groundice dischatge easily, which had made soil drierand swamp disappear.(2) Thar depth increased after the vegetationcover was removed. The contrast observationbetween stripped and un-stripped sites indicatesthat the thaw-depth in stripped sites is 2-3times more than un-s'tripped sites one yearafter the stripping (Fig.4). This is caused bythe decrease of water content in soil and heatused for evaporation, and the disappearance ofswamp.(3) Environmental changes after mining.Before mining,.tM forest in mining sites must1270

Table 1. The contrast of temperature between strippedand un-Stripped sites in 1990 ('C)SiteMrrnthDaily air Daily ground Daily ranges of Ground temp.temp. surf. temp. ground surf. temp. depth of 2.5 mun- 7 16.4 15.9stripped 8 14.6 14.6Stripped 7 17.68 15.222.318.318.08.729.218.3-3-2.4-0.8-0.6may be never recovered or recvoered for a verylong time.Time (month) As permafrost and vegetation each dependother in permafrost regions, the vegetation4 5 6 7 8 9 10 1112recovery is very slow in permafrost regions,which affect the steady state of permafrost. But1 1 1 1the destruction of vegetation is one of the mainfactors to cause the changes of environment.Because of severe destruction to the environmentin the'process of mining we must take somemeasures to decrease the degree of destruction.We suggest that, first, we must do our best toreduce the destruction of ground surface and150 -natural landscape: second, rationally despositthe used ore body and reduce its cover in the200-natural surface; third, pre-store the organicsoil layer and replace the layer after mining;fourth, plant trees on the mining sites aftermining.Fig.4 The comparison of thaw depth between, .stripped and un-stripped sites one year afterREFERENCESthe stripping in Wuma mining areaGu DonRxin and WanR Shaoling, (1961) Division of,be cut or buried, but the forest around thePermafrost Regions in Da-Hinggan Lingmining sites was destroyed differently* In theNortheast of China (in Chinese). Journal ofmain mining sites forest was destroyed complete-Glaciology and Geocryology, 3(3), p.1-9.ly, all of this made of forest destroyed The Branch of Siberian Institute of Permafrost,Russian Academy of Sciences, (1988), Generalaround the river valley area.Geocryology, p.280-283, (in Russian).(4) VeRetation succeedsslowlv. Owing to thedestruction of vegetation and sbil laye;, the.environment has changed greatly. In some placesforest has been destroyed forever or has beenmade to be in a reversed succession. No vegetationis in the main mining sites two years aftermining. Only in both sides adjacent to the mainmining sites is there greeh moss and somepatches of grass growing there.The gold mining history in the region can bedivided into two periods. the first is from Mingand Qin Dynasty to the beginning of the 1940's.The second is from 1988 to present. There is80 years between the two periods. From fieldinvestigation the destroyed forest in the firstperiod has not recovered completely consideringthat mining in the first period was manual andore deposit came from boreholes, the destructionof vegetation is mainly man-made. Thus, thevegetation is in the process of recovering. Thepine recovery is slow, but the birch is rapid.The difference between virgin forest andsuccessional forest is obvious. The diameterof virgin forest Is 40-60 cm, but the succeesionalforest is 10-20 cm. The area of abandonedboreholes is now mostly occupied by flats andsand dunes, only very sparse vegetation hasgrown there. It is evident that when the vegetationis destroyed, the recovery is very slow.It needs more than 100 years for the successionto reach its climax. But the modern miningtechnology destroyed veRetation severely which' 1271

~ .RECENT DISCOVRRY OF PERIGLACIAL PHENOMENA ONTU WE1 BA SHAN (BROKEN TAIL WILL) IN ZWALAINOERJNNER MONGOLIAWang Zhenyi’ and Lin Yipu2‘Department of Geology and Surveying Zhalainoer Administrationof the Ministry of Power Industry of P.R.C. in China21nstitution of Paleontology and Paleoanthropology of Academia Sinica, Beijing,ChinaSince 1957,two terms -“WAN0 SHAN Solifeuction“ and “ZHALAINOER Solifluction” were coinedand their tentative age corresponding to “Riss“ and “Wurm” subglacial in Europe rcspectly by Prof. PeiWenchung, being discussed continually by scholars who intercstcd in p~riplncial phcnbmena.This paperdeals with a new Kind of solifluction showing on broken tail hill which nwy be comparable to that of”WANG SHAN Solifluction”. Beside, damage caused by freeze are to be rcported.IIIINTRODUCTIONSolifluction (literlly “soil flow”), a term first proposed by J. G.Andersson(l906) while studying pcriglacial phenomema on Bear Is-land, is definitely one of the most significant processes of soilmovement in periglacial areas,Zhalainocr,(N.49’20’, E.117’35’) is one of the famousperiglacial areas in China. Since 1957, two terms --WANG SHANSolifluction and ZHALAINOER Solifluction were coined and theirtentative age, the former corresponding to Riss the latter corre.’ sponding to Wurm by Prof. Pei Wenchung while studying thepxiglacial phenomena in Zhalfinoer, Manchouli, Inner Mongolia.The agc of WANG SHAN Solifluction is one of the debateingfocus among scholars who interest in studying the periglacialphenomena in Zhalainoer , The former author (Wang) discoveredrecently a solifluction on TU WET HA SHAN (Broken Tail Hill),itmay be comparable to that of WANG SHAN Solifluction. therefore,the authors wish to report the new find, and discuss its ape.Broken Tail Hill(Tu Wei Ba Shan in Chinese) is a littee solitarypromontory situating onto East-North of Zhalainoer station, beingnot far away from the open-pit of Ling Chuan coal mine inZhalainoer, a well-studicd locus studying WANG SHANSolifluction and ZhALATNOER Solifluction.A few years ago, a program of paving tube across throughtthat Hill was carried out, in the course of digging a series of outcropswere revcaled,and some fossil bones were collected by theworkes.The former author Wang, as a senior engineer, went to thespot investigating after he heard without hesitate.GEOLOGICAL COLUMN OF-qROK~-N.T~IL-H~LI~The Visible Layers seen”-The visible layers from the outcrops of the sections on brokentail hill nre as fblliwlngs, front !op to bottom:1) The layer of silt-sands thlckncssl-2 MThere were scveral f‘os~~l skulls of Ros primlpcntui .IndCoelodonta . . antiqitatisuncarth~J from thc lowcr portlcns ,>I. ihlslayer.2) The Inycr of involuted struclliresthickness 2-3 MInvolutions (cryoturbatlons, Ger. Tanchenboden) arc majorperiglacial features of periglacial phenomena, thcv are characterizedby lkdding distortion and interpcnetration of different layers.It may be explaned as resulting from thc squee71ng ol moist, plasticlayers between rigid parts of soil. Refore 1957. both geologists andfreshmen saw the involutions already, but could say no morc than”I dondt known”. Instead. Prof. pei came here and saw them. hewas excited by reminiscenting of the. involutions which he saw previouslywhen he was studing the periglacial phcnornena In French.“It seems to me that I meet an acqualntancc again, cerr:linly, hemust be called Solilluctlon.Fs:lid Prof. PCI.3)The Layer of Gravels [hicknew unknownThis gravels may be thc remains (01 wdtrnents 01 :Inctcnt river.4)The white sands-stone layerIhickness unknownT)ISCUSSION OF THE SOI.,IFI.CCTTONAs visualized catographically by K.Kaiser, cryoturbations(involution) are widesprcad in the pcriglacial areas,particularly inSlberia. Zhalainoer situates in the pcrmd‘rost arca Of easternMongolian Plain being in the neighhourhood of East Stheria, it isnot strange that cryoturbations alwaks occur, the question arisethat what’s agc of WANG SHAN Solifluction? and what’s the ageof the new find of Solifluction on broken tail hill?are they bdth in1272

the same subglacial stage?"YES, They are both of Wurm subglacia! stage". After studing.The authors give such answer.The age of WANG SHAN Solifluction was to bepresumptively Riss subglacial stage, it is under the circumstancewithout any C-14 dating; However, up to present, there are twoC-14 daings particularly for.WANG SHAN Solifluction:1) 28,900 1,300 years before present(PV-I 72); and2) 33,760 1,700 years before present (PV-170).Hence, the authors answer that both WANG SHAN and TU WE1BA SHAN Solifluctions are of Wurm subglacial stagc.For years, the authors studied WANG SHAN Solifluationthoroughly, in the Upper Pleistocene ~cological column, from topto boffom, there are 8 dating for it, they arc:1) 3,080* 80 (PV-201) ....... 540 M2) 5.2702 80 (PV-167) ....... > 540 M3) 6,710 4 200 (F'V-ZK-825) ..... 3 540 M4) 7,070k 200 (PV-166) ....... > 540 M5) 11,460 f 230(PV-15) ...... 535-536 M6) 11,600f130(PV-171) ,7) 28,900k 1,300 (PV-172)...... 538 M8) 33,760k 1,700 (PV-170) ...... 538 MThis is the standard section seen in the openpit of Ling Chuancoal mine. Both WANG SHAN and TU WE1 BA SHANSolifluction might be comparable to that of the Solifluction occurcsin the level above sea water 538 M , therefore , their ages arc of:28,900 k 1,300 years before present.REFERENSELTage Nilssov (1982) , The Pleistocene Geoligy and Life in theQuaternary Ice Age: PP:45-48.1273

UNIAXIAL STRESS RELAXATION OF FROZEN LOESSWu Ziwang, Ma Wei, Chang Xiaoxiao and Sheng ZhongyanState-Key Lsboratory of Frozen Soil Engineering, Lanzhou Institute of Glaciology *":and Geocryology, Chinese Academy of Sciences, ChinaThis paper analyses and discusses the relaxation law and affect factors of loessunder constant strain conditions. It was found that the larger the inital stress, or higher temperature was, the stronger the stress relaxation was'. Its relaxationequation is:u(t) A(B)E:(t + l)-'Meanwhile, this paper intrdduces a new method of determining viscose coefficie,nttl of frozen soil: 0-Tr.G.INTRODUCTIONSince ice and unfrozen water are present infrozen soil the frozen soils have very obviousrheological behavior. under loading, in thistype of soil there will simultaneously takeplace deformation (creep), stress weakness(relaxatiqn) and strength reduction (Thritovize,1985). The strength red.uction is due to theproduction of the rheological process of stressrelaxation in frozen soil. So the importantfactors affecting the strength of frozen soilare re-direction of mineral grains and ice andstress relaxation. In view of this problem, alarge amount of tests were done. This paperdiscusses and analyzes the studied results ofstress relaxation of frozen loess under anuniaxial stress state.SAMPLES AND THEIR PREPARATIONIn the Lanzhou loess sample the size was lOlx200 mm; the water content was 14-15?; the dryunit weight was 1.72-1.78 g/cm'; the test temperatureswere -2, -5, and -1O"C, and its precisionwas 0.1'C. All of the tests were done on aMTS-810 Vibrational Tri-axial ExperimentalMechine. The basic physical parameters are shownTable 1:EXPERIMENTAL RSSULTS AND ANALYSES1. Relation of Sttess-StrainThrough tests, the curves of instantaneousstress-strain were obtained under differenttemperatures (Fig.1). From Fig.1, we can knowinstantaneous strength and failure strain offrozen loess under dirferent temperatures(Table 2).Fig.1 Stress-strain curves of frozen loessTable 1. Physical parameters of samp,leSoilnameComposition of grain size ( X ) Liquid PlasticGravity limit limit>0.1 0.1-9.05 0.05-0.005

Table 2. 1nst.nntuncous comprPssivc strengt.hand failure strainTernperaturc("C)-2 -5 -7 - 106o Iof (Mpa) 2.914 5.770 7.727 8.003Ef ( X ) 6.32 9.18 11.5 8.63"Load rate was 12.5 MPaImin, Failurc time1 min.was2. Rclaxatinn LowThrough tes1.s. wc obtainerl thc relaxaliorrcurves shown in Fig.2. From the I.u'~ downwards,the constant strain of eac.h curve is rcspcctive-I y 7X, &X, and 3X. It is seen that relaxaI.ior1yruccss obviously hos two stsgrs: the strongrelaxrrtion stage anrl thc slow relaxation stage.The first. stage is the main relaxation stogc,it is about 30-407, of t.he initial value orhigher, and t.his process generally is comp1et.etiwithin 1-2 hr.50 r40 U60r30E, = 5%Fig.2 Relaxation curves of frozen locss ( - 5 O C )From Fig.2. .it is seen that difftrcnt constantstrains are clqsely related to the c~~rrespondihginiLial stress. The init.iaL stress increaseswith an increase of constant strain. With aninc.rease of constant. strain, the initinl stresswill be close to the instantaneous strength, andthe stress relaxation will bet-orne strong. Wedefine the relaxation degree S as:'tWhere Uo is initial stress; Om is strrblc stress:Of is instantaneous strength. Fig.3 are t,hccurves of S VS. initial strain E,. This relationcan be described by eq.(2):1S A€: t B (7-1Where A~2.174, R=-0.6029.0.5 13. Effect of Temperature on Stress Relaxation of- Frozen SoilWe know that the lower the temperature offrozen soil, the bigger its strength is. Our-e ("C)relaxation tests also reflect this phenomen 011(shown in Fig.4 and Fi.g.5). Both initial stressand relaxation stress increase gradually with adecrease of temperature. Fig.5 The relation bctween t.emper:>tur'ristress and1275

- s100-80 -60 -m 40-20 -greater Tr value, the more materials approach asolid state.From eq.(5), we may obtain:So, through Fiy.7, we may determined Tr value ofour tests.1 , . . . . 10 2 4 6 8 1 0 1 2-0 ( OC )Fig.6 The relation between relaxation degreeand temperatureThrou'gh synthesizing the above-mentionedproblems, we. may obtain the following equationof stress relaxation:a(t) = A(O)E! (t+l)-c (3)Where O(t) is stress at an arbitrary time (MPa);c0 is constant strain: t is time (hr.): 0 istemperature; m and < are test paramctcrs, m=0.46,

CONCLUSIONS1) Under the same condition, the bigger theconstant strain or the higher the temperature,the stronger the stress relaxation of frozensail.. 2) The relaxation law of frozen loess can bedosc,ri'bed by the following equation:U(t) = A(e)E!(tt I.)-'3) Once we know the shearing modulus offrozen soil, it is easy to obtain its viscosityfactor from relaxation tests..REFERENCESThritovize, H.A., (1985) Mechanics of frozensoil, Publishing House of Sciences, pp.127-135.Vyalov, C.C., (1987) The rheologic principle ofsoil mechanics, Publishing House of Sciences,pp.130-137.1277

APPLICATIONS OF DATA BASE TECHNOLOGY IN FROZEN SOIL RESEARCHXia ZhiyingLanzhou Institute of Glaciology and Geocryology,Chinese Academy of Sciences, Lanzhou 730000,China'The paper analyzes in .detail the features of the permafrost data, staRes themethodes of how to process them, and classifies the data into several types.Furthermore, the data was handled according to the theory of standardization andtheory of a relational data base, and developed a data base system suitable forthe characters of permafrost data and practical application. The system ismainly composed of the following three parts. The first is information acquisitionand systematic design of the data base system, the second i s retrieval ofdata, the thitd is output of information and its technical utilities. Testsshow that the system is very effective for registering, management and technicalmanipulation. Therefore it yosscsses important practical values.1. INTROD,UCTIONData base technology which concerns datasharing is an important branch of computersciences and is the core of information systems,its application has expanded in geoscienceresearch in recent years in 1983, a conferenceon geographic information systems was held bythe geoscience department of the Chinese Academyof Sciences and the developing centre of geographicinformation systems was established soonafterwards. In 1988, the glaciology and geocryologydata base centre (WDC-D) was set up inChina by the organization of World data BaseCentre (WDC) and the works of data standardlzationand system development were graduallycarried out, Rased on the research done by theauthor in recent years on data arrangement,evaluation and standardizing, especially for thedata or materials collected along the Qinghai-Xizang highway, the author provides a method forestablishing the data base application system ofpermafrost environment and forecast of Qinghai-Xizang Plateau. This paper discusses someproblems related to establishing a frozen soildata base, and it is possible for the data baseto cover gradually aXI the permafrost regionsin China, The model development methods (PanJinping, 1985,) and principles are adopted in thesystem development according to different softwaredeveloping periods to establishan informa-Lioo system of permafrost environments andforecast with completed functions and perfectproperties.2. THE RESEARCH,CONTENTS OF GEOCBYOLOGY AND ITSPRESENT DATA MANAGEMENTThe requirement analysis for any softwarebefore d.eveloped is of important meaning, so itis necessary to discuss the research contents,data types and the applicable management of dataand so on.Geocryology is both an independent disciplineand an inter-discipline science. It studies thecryosphere, its spacial and temporal development,its relationship with the lithosphere, hydrosphereand human economic activities stc.Geocryology emphasizes the freeze-thawprocesscs of soil or rock, including (1) Thethermodynamic condition in the freeze-thawprocesses; (2) The physical and physiochemiatryprocesses occurring in freezing and thawing soil;(3) The frozen soil and its components, itsstructure, its state and properties; (4) Thegeographical phonemena and processes of frozensoil and their determining conditions.From the development viewpoint, geocryologystresses several points. The first is thesynthesis research which is based on amounts ofreliable information. The second is the transitionfrom qualitative to quantitative analysis.The third is the transition from static descriptionto dynamic forecast, etc. All the abovefeatures ask for scientific data management andtreatment which in turn are needed to rule outthe data collecting regularity to detetmine thecollecting range, quantity, accuracy, timeperiods and means, to establish a unified datamanagement system-the data base system, toincrease the data utilization ratio and decreasethe unnecessary repetition in investigationalwork and to find a suitable mathematical toolfor quantitating the qualitative information andfor transforming them into numerical, comparable,and measureable values (Cheng Guodong, 1988).The data in permafrost research covers a widerange and is the foundation for the establishmentof a data base system. But the dataadministration work has been in a manual periodfor a long time and the data is so dispersed,non uniLiedly formulated and nonstandardizingthat establishment of permafrost data base hasto be in 011r agenda. The more important thingis that no relationship has been founded amongthe data. The data base system is designed to

solvc this prohlcm of hockw;lrdlv d:~t;l a;lnilyl*m(-lltIr!- ;ldopti~lg tllr d.1ti1 h;ls;l- ;I~III~II~SII-:I~it111 t

Password s~fexuardII 'I/+ Reset recordes items - IDelete records+"KT"-]-"- Synthetic outputDate base output7 Partial file output- Output by classificationSynthetic statisticsTest data statisticseteorological data statisticsata statistics of ground temp."f Data safeSua+-"-kurehole information safeguardTest date safeguard,Yeteorological data safeguard- Ground temperature safeguard>Borehole information inputTest data input- MeLeorolvgical data inputGround temperature data inputFig.1 System appearanceI block) IInquiry blockFig.2 Partial program structure

GraphingIt is the function of choosing data based ondifferent application models to calculete, forexample, the numerical calculation of naturaland man-made upper table of the permafrost andthc temperature gradient, etc. Simple figuresca,j be drawn.5.6. Providing Various Data ServicesThe system provides operating instructionsand other services like inquiry, coping, dataunlpading, etc.5.7. The Measures of System Maintenancc andSecurityThe system has security characters of differentsecurity classifications to guarantee eachuser operates the data base system within thelimits of authority. When necessary, it crinreset recorded items so as 'to pr0tec.t thesecurity, independence, reliability andintegrity of the system.6. THE OUTI,TNE OF THE SYSTEM AND ITS PROGRAMSTRUCTUREunified formulation and normalized principle.Applied in the frozen ground data base systemdomestically or from abroad, the authorci+nonly carry out the basic system develvpment workbaser1 on the various materials gathered mainlyalong Qinghai -Xizang highway. It needs grcrrtimprovement and recommendations and suggestionsare welcome. We'll bcnefit from them by beingahle to est.ahlish a more perfected informaLionsystem of permafrost environment and forecasting.REFERENCESCheng Guodong (1988) A review and prospect forthe regional permafrost researches in China.Journal of Glaciology and Geocryology,Vol.10, No.?.C..J. Data (1980) An introduction to base systems,Vol.. I, 11.Pan Jinping (1085) Softwarc Development Technique,Shanghai Science Publisher.Sa Shixuan, Wang Shan (1983) Concept on DAta13ase System, High Education Publisher.Yao Qingda (1987) Data Base and Application,Science Publisher.The outline of the system is the operationflow diagram appearing in the system design anddata exchanging (sec Fig.1). It is the foundationof the program design and the guide ofoperation. A part of the structure of thesystems program is presented jn Fig.2. Thcstructural progr,am deslgn method used inprogrammihg and program design can be scen fromit. It is unnecessary to divide the module indetail for this can make the system dispatchtoo frequently and decrease the operating speedof the system under foxbase. When the controllingflow logic is clear, less modules aredivided, when complex, more modules are divided.Tn that way, the maintenance and operatingspeed of the system arc both considered at thesame time (Sa Shixuan, Wang Shan, 1983).7. CONCLUSIONSBased on detailed analysis of the vatiousdata of frozen gkound, this paper concentrateson the establishment of the frozen ground database system, which is suitable for frozen groundresearch, engineering construction in permafrostregions and the data exchange either internationallyor internally. It also presents thebasic function and system consti,tution thesystem can realize. The following probL-s arcdiscussed or solved:1. The system design bas ',' on microcomputerhardware;2. Realized, normalized stot-age an3 managementof data: '3. Realized data-sharing add eliminating thephenomena of personal occupied materials:4. Avoiding the data loss caused by manualdata management:5. Providing reliable, integrated dat.a ormaterial about frozen ground for theWorld Data Centre (WDC):6. It is convenient for the user because rhcdata statistic, analysis, calculation andthe outputing of results and simplefigures are all realized on computer:7. The data compilation and changing can becarried out at any time.It should be pointed out that this system isnon a perfected application system especiallyin the normalized data research for there is no1281

'HOST NRAVF PROPERTIES OF NONSATIIHA'I'EI) CCIMPACTEII COHESTVR SOILAND ITS APPLICATTON Ihi WINT?R CONSTRUCTION OF CORE DAMS, Y.'Xie Y ingqi and Uang JianguoHeilongjiang Hydraulic Research IustituteThe authors introduce the frust heave regularity of nonsaturatcd compacted cohesivcsoil irnd evaluatc quantitatively the irlf.Iucnce oC the saturated' degree anddensity on thc frost. heave. Meanwhile, wc have an idea that. the compacted soilwit.I-I the low saturated degree and high density could be usc.d to pake ,th"cimpervious barricr-which would be safe in !.he winter ul-ithout insulation'xn ,.scason~l irozen regions. The practical engineeri.ng method 'IS givcn, and in:~ddition, thc formula'.of estimating Lhe needed t.ime of melting the 'ffozen layert:hi1r.,ughly wi I1 be mentiolied, in thc nIet.hod of thermal qquilibrium.IN'IRODIJCT~OYIn or~ltfr. t'r> :;olive cnginecring problems in 1.11eseasonal froz?n rr!gions, wc' regard thc imperviousbarrier of r:,.ohesbve soil which is buill. tolast the winI.er under frcezin$; without irlsulation,allrjw no cxternal water recharge, paps nlicezi.ng BIIII mrltirig cyclc and meeting thephysict~l nnd mechan.ica1 index that is thcpcnctratioll I.11efficient.. compressibi lity-coefficicntand she,nring coefficient, etc. So west.udir.d, the unrehI.r';iincd frost heave regulat'i t yof nonsa!.ur,rtcd compaL.tnrl cohesive soil under ac.onfined sysLerrl, with cone di rcctioncll freezing,no additional Iootl, as we1 I os the changingsit-uation of soil permeahiliby (Paggy, 1980) andshearing sI.tength after one freeze-t-haw cycle.THE TEST CDNlITTlONS ANT)ME'I'H(IDSclearly show that the ilmount of 'Crost heaveincreases with the decrease of dry unit wcight.Fig.1, Fig.2 and Fig.? s,haw the curves of theCrost heave amounL (Ah), frozen depth (Hm) VS.elapsed time. Rased on these f igurcs, we callsee: When the degree 6f saturation is cunstant,the amount of frost heave will redu,ce. with theenlatgcmcntf of dry unj t: weight. Frotu.Fig.4, theAh-W curve with parameter Pd, ve,c+rl see thatwh'rn the wqLer cnntena.lp) ig con'ptant, theanlourlt of frast hqav.@ will i,ncrep?e with thecnlargemept of the dry unit weight. So thecontradictions between the omounts of frostheave red,uition with thc enlargehent of dry unitweight and frost heave increasing w i t h th.cenlargemeot of the soil skeleton differelices ofdensity in some refereqccs "(Orouf, 1980)reflected the practical'conditions respectively.'The urlidircctinllal frost heave instrument. wasopere\.erl in the l o w t.i.mpe:atur,t!. laboratory.The s;implc tube is 131) mlm in -inner liamrterat the top, 120 mm at. t.hc buttow ill 1,'23, internalwall taper 200 rnm in " ' ight.Thc sample's t.c!mpcratm II olpng the depth each70 mm was mehsurcd hy uric )i- two thermocou,nleswhich were nadc 'ftotn r.npp'cr..Tht~ ircczitlg Lnmpcratuit. at the t.op was -20'C.and at L!I~ bottom it was t2'IC. Thc melting. 4 'tcmperaturv at the t.op was t15'C - t20°C,, and. t;tthe hrrLt.um wn

SamplescollectingsiteTable 1. The analysing results of soil for the teatPercentage of followingsize (mm) by weight ( X )Specific'P 'L 'P gravi.ty Classification. ->0.05 0.05-0.005 (0.005 x xglcm'. .Wanjia, Harbin,Heilongjiang Province 16.4 64.0 20.0 21.5 34.6 13.1 2.69CIGeibehe, Wuchang County, 16.0Heilongjiang Province63.0 21 -0 23.0 35.0 12.0 2.66 CILinxong CountyHeilongjiang ProvinceCharseng ReserviorJilin Province11 .o 50.0 39 .O 23.5 43.4 19.9 2.71 CI27 .o 42.0 31 .O 21.7 38.8 17.1 2.71 CITable 2. The results of test dataTest ItemDry unitResults Satura- Water weighttion content (glcm' )Number Sr(X) w (%)The maximumfrost heaveamount(mm)Frostheave Noteratio(X)F-IV-1F-I 11-12345623456707070707070ao80ao80808017.820.222.724.126.328.617.820.222.724.126.328.61.891.831.771.741.69.1.651.971.921.871.851.811.761.601.521.441.401.341.281.63I .601.521.491.431.37-0 e 270.711.201.553.5312.231.120.952.176.079.9814.28-0.2* 0.5' 0.01.02.48.90.70.61.44 .O6.79.5F -v-123456 '90909090909017.820.222.724.126.328.62.062.021.961.941.901.871.751.681.601.561.501.45* 0.58,1.203.448.409.0813.960.40.82.35.66.19.3pa = 1.37In equation (l), the enlargement of Pd willresult in the reduction of W when Sr is a constant.Under the circumstances that other condi-10cions weren't changable, the absorbed water contentof soil was relative to the surface energy- 8 of particles.' 6The reduction of water content means thereduction of migrated water content whicha 4 provided for the phase change and ice gatheringQ 2and produced heave. From the macroscopic scale,the amount of frost heave, will decrease with theenlargement of pd' Generally speaking, the3 amount of frost heave would increase with therising of Pd and the frost heave property ofE 6soil was directly relative to the degree oEx 9saturation in soil,12 On the other hand, property the of the compactedwarm soil can be known as: When the workof compaction was :a constants, the dry unit weightwas the function of water content (see Fig.5).Fig.2 The test result of Ah-H,-tOn the contrary, when the work of compaction waswhen Sr equals to 80%not constant, a different dry-unit weight could1283

IP* = 1.45/.75degrees of saturation were shown in a mathematicalequation, it would be:where, pdi - the density value which actedobviously on the frost heaveability, unit: g/cm:S, - the degree of saturation which waslisted in decimals.Fig.6 was the curve of the frost heave ratiowith dry unit weight with a parameter Pd. Theyare a group of approximate logarithmic curveswhich met at the point of the horizontal axissr=70X. According to the suggestion of somedata thatnon-frost heave soil was qSlZ, thecritical dry unit weight value of the frost andnon-heave soil was:Sr=70% Pd

Table 3. The permeability and shearing strength o f warm soilShearing strengthpdK(g/cm') (cm/sec) C(kg/cm')0,("1Notei, 75 i .86 32.5 non-permeability *.1.70 5.9~10" D. 90 c16. 5 in test1.65 3.5~10" 0.89 13.01.60 2.8~10" 0.79 7.51.55 0.74 4.0Table 4. The permeability and shearing strength of frost boil soil1.75 8.7~10-~ 1.65 22 .o1.70 6. S X ~ O - ~ 1.31 Ilr.51.65 5.5~10-~ 1.07 13.51.60 5.4~10-~ 0.83 8.51.55 4 e 7~10-~ 0,60 5.5Fig.7 The ideograph of calculating the shawingtime in frozen layersimple evaluation method is suggested.If the cold quantity inside the frozen layerheight H in one unit area was Q, the quantityof heat inside the frozen layer transferredthrough the new filled warm layer in one unittime and area was q, the quantity of undergroundthermal heat in one unit time and area was q'and the time needed 'in melting all the frozenlayer was T, the thermal equilibrium equation is:Q = (q + q'V (4)Because the melted layer depth produced byunderground thermal heat in the seasonal frozenground made up about one-tenth, so we couldcontinue to assume: q'=O.lQ. The equation 4could be simplified as:0.9Q P qT (5)and Q=HH(Cvlt- ItL-pd*W.i) 0CP(6)where, Cv - the volume specific heat of frozen- ground;t - the coverage minus temperature ofcp frozen ground;L - the latent heat of water phase:Pd - dry unit weight of frozen layer;W - water content of frozen layer:i - relative ice content.q =X't,+pHt2where, h - heat transfer coefficient;t:*- average air temperature during' y melting.After Equations (6) and (7) were put intoEquation (5) the time needed to melt the frozenlayer is as follows:0.45Ht.H~(c,lt,,l+L.Pd.W.i)T - + . (8)X'tcpUnder the condition of natural air temperature,the cohesive soil was compacted in theoptimum water content and its relative icecontent must be less than 1. In order to considersafety and simplify the calculation, we .canassume that i=l.O, so Equation 8 can be written:LcpEquation '(9) was the function that we referredto evaluate melting time of the frozen layer.Using Equatiqn (9), we can also know that fromthe rebuilding to the beginning of the next coldseason, the new filled warm layer maximum depthwithout residual frozen layers inside the imperviousbarrier should be:

The value in Equation (10) was a known numberand its maximum value was the continual time ofplus temperature on a daily average in theseasonally fro+en ground region.CONCLUSIONThe regularity of frost heave in compactedsoil obviously showed that: when soil wassaturated in the special conditions, the frostheave ratio would reduce with the enlargement ofdry unit weight. When the dry unit weight reacheda critical value, the compacted soil which frozewithout heave, i.e. frost heave ratio qrlX couldbe reached. In order to obtain the soil, atfirst, degree of saturation Sr was solvedaccording to the designed dry weight andcorresponding optimum water content and then thecritical dry unit weight pi needed for freezingwithout the heave could be arrived at inEquation 3. If the compacted soil was made toreach pi and S, was kept at one constant, thewater content should be decreased, Equation 9can be used to evaluate the melting time of thefrozen layer and determine a suitable time forrebuilding.REFERENCESC.A. Peggy and V.K. Horsler, (1980) The EngineeringProperty of Compacted Cohesive in theState of Saturated and Frost Boil Recycline,Beijing Water Conservancy Science andTechmoldgy, Complimentary Issue.B.O. Orouf, et al., (1980) The Frost Heave ofSoil and the Application on BuildingStructure, Translated by Li Yongsheng.

OBSERVATION AND RESEARCH OF SORTED CIRCLES IN EMPTY CIRQUEAT THE HEAD OF URUMQI RIVER, TIAN SHAN, CHINAXiong Htigang' Liu Cmgnian2 and Cui Zhi ju2'Department of Geography, Xinjiang University, Xinjang, China'Departmint of Geography, Beijing University, Bei jng, CGnaThe features of the sorted circles have been analysised with the farm of the sorted circle, grain size, watercantent. frost heave stone sirting, fabric and reference to wooden stakes inserted vertically into theground. The results show that the sorting degree becomes much better; the influence of frostheave decreasesgradually; and the center of sorted circles was covered by vegetation increasingly from the upperto lower sorted circigs. Moreover, the general tendency of the frostheave is central frostheave > internalgutter frostheave > gutter frostheave in a sorted circle.THE STUDY AREAThe Urumqi river rises on the north slope of the Kalawchenridge, Tian Shan and is about 150 Km long. Precipitous mountainpcaks, 4.000"4,400 m in altitude, arc covered by snow and glaaers.Snow line is 3,950-4,200 m above sea Icvel. The emptycirque ip approximately 3,820-3,950m (ad) with an area of 1.5Km', located in the north of the river head, and is facingsouth-st (Fig 1). In the empty cirque many sorted circles, sortedstripes, sorted nets and sorted polygons developcd, almost includingall the classifications of sorted forms given by Washburn(1979) The sorted circles, with different diameters and forms, are atypical case among them. Since the place, altitude and developingtime arc different, developmental degrees of sorted circles have anobvious diversity. For recounting conveniently, the sorted circlesdistributed in thrcc different elevations, 3,820, 3,880 and 3,950 m,were called lower; middle and upper sorted circles acparatly. Theconditions of sorted circles am shown in Table 1.STONE STATISTICSSince sorted circles of different developmental stages undergodifferent sorting degree has diversity. Generally, of three principalcharacteristics of stone can use to decided the sorting anddevelopmental degras. Fimt, ratio (P = Dr / Dc) of mean 'size of* gutter stone (Dr) and central stone PC) The bigger P is, the betterthe sorting and developmental degree are. Second, variance (S) tomean (X) ratio (P = S / 100 X X) of stone size in the centre of sortedcircle. The F is regarded as the parameter of the representedsorting degree. The bigger its value is, the worse the sorting degreesarc. Third, percentage (Q = A / 100 X 100) of numbers of gutterstone (A) which ab face is tangent with sorted circles. The bigger Qis, the more mature and sorted the sorted circles are. All theparameters were counted from one hundred stones.Statistical data(Tab1e 2) shows that the ratio (P), increasedfrom 2.5 to 12.5, the percentage (Q) rose by 17 pcrccnt and the varianceto mean size ratio (F) of central stone decreased nearly onetime, three results of the stone statistics suggested that sorting anddevelopmental degree increase gradually from upper to lowcr sortedcircles.The developmentil degree of sorted circles is a sequence ofchanges gradually along the matrixal source region in the sameplaw. If th~ sorted circles arc near the source region, theirdevelopmental degree is worse. Conversely, if they arc far from thesourn region the degree is better. In middle sorted circles. Wemeasured the forms and stone sizes of sorted circle from the edge ofthe talus to the centre of the cirque on a line (Table 3). The ratio (P)far from the talus, is more ten times than that of it near the talus.Moreover, the heigkt between the centre and gutter of sorted circlesdecrease and the microcircles, soil polygons appear on the centre ofsorted circles gradually with the distance increased. These alsoprove that the sorting and developmental degree of sorted circlesmuch better gradually. From a review of field evidence and physicallogistics it has been deduced thit the place, which is far from thetalus,has a comparatively smooth surface and fineness of high content,so it can contain much water, and is influenced by frost- thawEasily, implying that the developing rate of sorted circles is fast.'OBSERVATION OF -STHEAVE-In the summer of 1990 wooden stakes were used to set upobservational posts in lower and middle sorted circles. Three differentmethods were used to measure frost heave. Firet, 15 woodenstakes 3.5 crn in diameter and 25 cm long were Completely driveninto different positions of a sorted circle (Fig 2). observing the frostheave of different positions with stakes of the same diameter andlength.Sccond, 6 wooden stakes 25 cm long and 1 x I-- 5 x 6 cm in' diameter were placed in the centres of 6 sorted circles with a similar1287

E23legendOf cirqueuntil1 of Little IceEl ~ g Stage eLu until ofa Ncoglaciatian Stagesorted circlesa talusglacier. Fig.] Index map of sorted circles in empty cirqueTable 1 . Conditions of the different sorted circles in empty cirquealtitude till age diamctcr(m) in placc (m)formplacecentralconditionupper16th" like a back of nosorted 3,950 19th 1"s utcamcd vegetationcircles centuries bread $quemiddle Z.Oo0" slightly middk nosorted, 3,880 3,500 2-4 high of MnPtY vcgctatioecircles B .P centrc cirquelower l0,OOO" higher mouth covemd bysorted 3,820 20,000 1-3gutterofemptycircle B.P cirquevegetationI after Cui Zhiju 1981. Wsng Qingtai 1981.scale, trying to find out the function of frost heave to the woodenstakes with the same length but different diameter.Third, 8 woodenstakes 3.5 cm in diameter and 5-40 cm long were separatelydriven to the centres of 8 sorted circles with the similar ales, inorder to understand the condition and influence of frost heave tothe wooden stakes of different depths and the development of sort-ed circles. Generally, the layer 40 cm beneath the ground surface a Io .acts very strongly, so the longest wooden stake selected was 40 cm.The accuracy of the survey was tg 1 rnm or better.For discussing the frost heave relationship between the centreCtUFig.2 Position and numbers of wooden stake in sorted circle1288 *

Table 2 . Statistical parameters of stone of sorted circle in threedifferent placesF P Q Altitudeupper sortcd circles 5.57 2.5 49 3,950middle sorted circles 4.96 6.0 60 3,880lower sorted circles 2.82 12.5 66 3,820and gutter in a sorted circle.the observational data which was gatheredby the first method, were divided into 4 groups: (l).Centralstakes;(2).Internal gutter stakes; (3).0utside gutter stakes;(4).Gutterstakes (Tablc 4). -e data shows that action of the centralstakes was strong and that of the gutter stakes was weak. The meanheave of the central stakes was 2.65 times than that of the gutterstakes in the first observational cycle(Ju1y 6,1990 -August 20,1990) and 1.3 times in the second observational cycle (August 20,1990- - August 14, 1991). The general tcndency of mean heave ofthe stakes was central stakes > internal gutter stakes outsidegutter stakes > gutter stakes. This reflected fully that the frostheave decreased from the centre to the gutter of sorted circles.There was also great diversity among the action of every side of thesorted circle. The fastest rising stakes were on the west side and theslowest rising stakes were on the south sideflable 4). The meanheave of the former was 2.1 5 times and 1.83 times than that of lattcr,separately in two obserational cycles. This difference may be related to the water content and grain size of different parts of thesorted circle.The heave of the wooden stakes for thc pcridd 1990-1991ranged from 33 rnm, for 5 X 5 cm in diameter, to 67 mm, for 1 X 1cm in diameter (Table 5). This suggested that wooden stakes withthe least diametcr rose quickly in the centre of the sorted circle. Itmincides with the condition that the small stones rise to thc groundsurface quickly in the process of freezethaw sorting.But the tendencyof frost heave is not clear, this may be related to the maturityand inaction of lower sorted circles.The heave of the stake 20 crn long was 122 mm and stake 40cm long was 45 mm from August 20,1990 to August 14,1991. Theformer is more 2.7 times than the latter. The surface layer above 25cm acted strongly. Three stakes, 5, 10 and 15 cm long rose outcompletely and lay down on the ground surface in the field observationof August 14,1991. This is consistent with thc conclusionthat the frost heave of the surface is strongcr than that of depth,which was also achieved in research of other regions(WangXiyao,1982, Cuitoweiqi,l985).The ground surface was casily influenced by changes of theoutside environment, In early summer the ground surface began tothaw, and the highest frequency of frost-thaw appeared in June.The stakes were pushed upward with ground surface freezing.When the ground thawed from the surfam downward, the stakesfailed to return fully to their pnfrcczing position, to produ& a netof upward heave of the stakm. Repetitive frceze-thaw cyclcs resultin a gradual upward movement of the stake. There was still freezingin the depths of the sorted circles at this time. In summer the thickactive layer was caused by high tcmpcrature, precipitation and lowerfrequency of freethaw. If thc material in the depths moveupward, it must overcome the gravity and cohesive forces of mudin the upper layer, In September the frequcncy of freeze- thaw increased.There arc still very strong influences of freeze-thaw in theground surfacc, the stakes in the surface layer rise quickly.The con-Table 3. Changes of sorted circle with an inmasud distancenumbersdistanceFromtalus(m)5.0’9.514.518.021.524.526.5from talus in middle sorted circler5.25 0.324.5 0.65.6 0.193.2 ’ 1.413.6 1.142.8 1.57’ 2.5 3.44heightbetweencondition ofcentrccentre andwttm (m)25201713 soil polygons10 microcircles and polygons0 microcircles and polygons0 microeircles and polygons1289

Table 4. Heave of wooden stakes in different positions and sides of lower sorted circleinternal outsideccntrid gutter gutter north cast south weststakcsstakes stakes stakcs stakesstakes stakesstakeq(4'5'6) (3,7,12,13) (l,9,10,15)(2AIlJ4) (7,8,9) (IOJlJ2) (1.2.3) (13,14,15)July6,1990 0 0 0 0 0 0 0 0'August 20,1990 5.3mm 4.0mm 4.8mm 2.Omm 4.3mm '3.7mm 2.0mm 4.3mmAugust 14,1991 39.2mm 37.0mm 29.5mm 30.0mm 37.8mm 23.5mm 43.0rnmTable 5. Heave of wooden stakes with different diameter inlower sorted cirdesDiameter of ataka1x1- 2x2 3x3 4x4 5x5 6x6July 6,1990 0 0 0 0 0 0Augurt 20,1990 2mm 5mm Smm lOmm 4mm 3mmAugust 14,1991 67mm 38mm 36mm 41mm 33mm 41mmTable 6. Heave of wooden stakm with different longth in middle sorted circleslengbt . Of #take8 (cm).July 6,1990 0 0 0 0 0 0 0 0August20,1990 20mm 47mm 26mm 13mm 8mm 7mm 4mm. ImmAugust 14,1991 * 122mm 8lmm 39mm 30mm 45mmstake rose out completely and lay down on thc~ounddition is similar with that of June.In autumn temperature decreases,the active layer begins Odouble freezinglfrom both surface and bottom. Conditions of frost heave are very complicated and thefrost-pull theory and frost-push theory are used to explain thefrost heave of stones by downward freezing which can not bedirectly applied to upward freezing by reversing signs, because thevertical direction of ground surface easily influenced by the outsideenviranment is higher than that of depth.Comparing the data of lower sorted circles, the heave ofwooden stake in middle sorted corcles is a markedly larger. FromAugust 20, 1990 to August 14, 1991, the heave of stakes, with thesame lengths, diameters and at the same position in middle .sortedcircles, is 3 times than that of stake in lower sorted circles.CONCLUSIONFrom the upper to lower sorted circles (from infant stage toold stage) the sorting degree becomes much better; the intluencc offrostheave decreases gradually; the centre of sorted circles was cowcred by vegetation increasingly; and forms of sorted circles changefrom the high to low centre. Preliminary conclusions of their faturesand developmental backgrond which have been drawn fromfield and laboratory analysis is in Table 7.Wooden stakes with small diameter heave higher than that ofones with a large diameter in the same pcriod. Furthmore, the frostheave of surface is stronger than that of depth. The stronmt activelayer is above 25 cm from the surface. In the same plm the general1290

Table 7. Featurc of sorted circle in different developmental stages at empty cirquedcvelop- condition condition glacialsortingmentalof frost of circle stagc indegrcestage hcave ccnvc plaw:upper very Littlc smallsorted infant lower strong sorted ICCcircles frosthavc circlcs Agemiddle strong small sorted Ncosortcdmature mcdiurn frost circlcs and glacicirclcshcavc polygons anonlower weak no small wpcrsortcd old higher frost sorted- Wan8circlcs hcavc circles Fentendency of the frost heave is central frost heave > internal gutterfrost heave > gutter frost heave in a sorted circle. The maturityand sorting degree of sorted circles increase along with the increaseof distance of the material source region.Finally, additional data and research are necessary on thedcveloprncntal time, microclimate changes in different heightswhich control the developmental degree of sorted circles.ACKNOWLEDGMENT”_Financial assistance for this study was provides by the NationalNatural Science Foundation of China and Science Foudation ofTian Shan Station, Prof. Liu Chaohai(LanZhou Institute ofGlaciology and Cryopcdology, Academia Sinica) supplied varioushclp with our work.REFERENCECui Zhiju,l981. On the glacial cirque at the head of Ururnqi River,Tian Shan, Journal of Glaciology and Cryopedology,Vol. 3,Special Issue.pp 57--63 (in Chinese). Cuitoweiqi,H.A, 1985,The mechanics of frozGn ground, Sciences Press, Beijng. pp80-99(in’Chinese).Mackay;J.R.1984, The frost heave of stoncs in the active layerabove permafrost with downward and upward freezin,Arcticand Alpine Research, Vol, 6,NO. 4, pp439-446.Wang Jingtai,l98IyAncicnt glaciers at the head of Urumqi River,Tian Shian, Journal of Glaciaology and Cryopedology, Vo1.3,Special Issue pp 24--35 (in Chinese).Wang Xiiao, 1982, Frost heave and its distribution in different layersinfluenced by shallow ground waterJourna1 of Glaciologyand Cryopcdology, Val. 4, No.2, ppSS”WnWashbq,A.L, 1979,c3eocryolog~,Edwa~d Amold, ~ ~22- 277:1291

STUDY AND DEVILOPHENT OF THE TECHNIQUESAGAINST FROST DANAGB OP HYDRAULIC STRUCTURESBomeng Xu Anguo Li Lijun ShaoInformation SysteH of Antifrost Techniques of HydraulicStructures of China,74 Beian Road. Changchun Citu.ChinaThe progress of the study on the techniques against frost damage of hydraulic struc--tures in China is introduced in the aspects of foundation soil heaving,frost heavingforce and the design measures against frost damage. The remaining problems to besolved are simply discussed and the study direction and tasks in the future arealso presented.key words: Hydraulic stucture, frost damage, frost heave prevention and treatment.INTRODUCT-IONOur country has a vast territory. quite a larRepart-of which- is in the cold zone of the NorthernHemisphere. The area of frozen so i I in our countryis 68.6% of the whole territory In which theseasonally frozen soil vith frost depth more than0.5~--46.3% , the permafrost--22 3% . In thesereglons, not only a few hydraulic st.ructures wereseverely damaged by the freezing a nd tharlne action.The investigation data indicated that in Lishu andseven other irrigation districts of the province ofJiling, 71.4% of the irrigation canal structureswere damaged by the frost action: in the Chahayangirrigation dostrict there, were 93 structures damp--ed in the 112 investigated structures; the canallining in the different regions were also damagedto a certwln extent (Xu Shaoxin, 1986). Therefore,it is necessary to study the various problemsconcerned with soil frost and ice action.DEVELOPMENT OF THE RESEARCH WORKOn the whole, the research work in this fieMin China may be divided into three developmentstages as follows(Xu Bomeng.1988).The first stage approximately ran through the1950's. In this stage, the research work wasdeveloped mainly around the influence of soilfrost and the physical and nechanical propertiesand earth filling techniques in winter. In theearly 1950's. a number of laboratory and in-sitetests of the eart.h filling methods, earth fillqualities in winter were.carried out and thawingsubsidence characteristics of frozen soil cakesand their fillins bodies in water were studied.These test results had application in the watertl--ght blankets of Longfunshan and other reservoirsin the province,of Heilangjiang, which initiateda precedent of ut-ilizintz frozen soils imn hydraulicconstruction in cold regions. At the same tias,theprevious Shengyana Hydraulic Research Institute ofMinistry of Water Resources had compiled and publi--shed "Construction of Roller Earth Dam in Winter"The second stage was approximately from theearly 1960's to the mid-1970's. The research onfrozen soil developed mainly vith the focuson thehydraulic structures destroyed by the heaveoffoundation soil. Some efficious measures againstfrost damage vere sommarized. At the same time,theinvestiRation of ice injury and observation of icepressure were carried out, the law of ice pressureformation and calculating method were studied,The third stage was fror the mid-1970's topvesent. It is the stage that the study of preven--tion and treatment of freeze damage of hydraulicstuctures and canal lining and soil freezing werequickly developed. Since 1977, the "CooperationGroup of Research,,on Antifrost Technigues of Hydr--aulic Structures and InforBatin System of Anti--frost Techniques of Hydraulic Structures wereestablished. They held eleven academic exchangeand more than thirty symposiums, on vhich over 500papers were presented about the investigation andsummary of freeze damape and its -prevention andthe treatment techniques of hydraulic stuctures,soil heaving and its action tothe stuctures were presented. Forty volumes of theJounal "Engineering and Frozen Soil",fifty volumtsof the Journal "Waterti'ght Techniques of Canalhave been published. The "Experimental Yethod offrozen soil", the "Design Standard of Anti-Frostheavingof Canal System Structures", the writings"Prevention and Tteatment of Freeze Damage of Hydr--aulic Structures , Watertight of Canal and the"Callecting Drawings of Canal System Structuresin Seasonally Frozen Soil Regions" have beenprinted. At present, the "Standard of HydrayljcStructures Design Against Ice and Frost Heave ISin compilation. Besides, not quite a few foreignuseful references have also been translated intoChinese.RESEARCH ACHIEVEMENTThe primary achievements are as follows:- 1292

1. The Freezing and Thawing Properties of Soils,and Earth Dam Filling in Uinter,The test indicated that the stable freezingtemerature of clay is lower t.han 0% and its mini--mum supercooling temperature may reach -5C. atthis timesthe clay is still in a plastic condition,consequently. can be filled and compacted withincreasing speed. The clay core of the earth darof Pinghe reeervoir was constructed in winter. thecompacted density under an air temperature of -12Y:reached 1.68g/cm3 , satisfying the design require--nent.The tests of the effect of soil freezing andthaving on the filled soil properties indicatedthat controlling the moisture of filled soil to benear or lower than the plastic limit can assure thequality, and avoid the effect of freezing andtha.wing action. As a result of the soil being se--para:ted by the ice in the course of freezing, thecompressibility increases, whereas the shear stren--gth decreases(Wang Liang.1983). , These resultsaentioned above have been applied in the clay coresof Pinghe and Dahofong reservoirs and the earth-,rock coffendam with weathering sand of BaishanHydroelectric Station and have a good efficiency.2. Characteristics of Crumbling. Sinking andCons.olidation of Frozen Soils and Their FilledBodies in Water add Application in the UatertightBlanket of Earth DamsThe test results demonstrated that the frozensoil. vhen puts into water, thaws and crumbles intosmall Pieces" from la,yer to layer and drops down.The crumbled amount within several hours is over80% in weight. BY the action of the dead weight"and seepage pressu're, the density of soil piecesaccumulation . gradually increases, appoaching thenatural density of the soil within a short time.This indicates th'at the properties of frozenclarer soils cruabling in water can be used toform a good watertight blanket. These results havebeen applied in several reservoirs of Heilongjiangprovince and have had good results.3.Hoisture Migration and Frost HeaveObservations indicate that the distribution ofthe.frost heaving amount along the depth is notunifor. with different foras. The limit of .thedepth of the water tab'le influencing on the 'dois--ture migration and frost heaving may be evaluatedby the capillary lift of soil, The hvdrau.lic struc--tures are usually located at the sites with highervater tables. Under the condition of an open system,frost heaving of ground is great. According to theobservations,in Heilongjiang and Jiling provinces,having greater frost depth of ground, the maximumheaving reached 55cm and 43cn,respectivel~. conse--quentlu the hydraulic structures were damagedto a larger extent.With a number of laboratory test and in-situ'observations, the design values of frost depth andheaving and their calculation methods are presented.4. Frost Heaving Force of SoilThe observations for years in sever1 field testsites indicate that the tangential frost heavingforce mainly occurs at a certain range of the upperfrost depth; the maximua value is brought about atthe time of the frost depth being about 70-80% ofits maximum, The main factors affecting tangentialfrost heaving force are the heave of soil and rou--8hness of Pile surface.The design values of tangential frost force forpile vith even surface are presented: 20-40kPa forthe soil vith weak heaving susceptibility, 40-80kPafor the soil with medium heaving susceptibility,80-150kPa for the strongly heaving soil.The value of horizontal frost heaving forceacting on the retaining vall under bi-directionfreezing of the back-filling soil behind the wallsis associated with the frost susceptibility of thesoil and the allowable deformation of the wall.Thehorizontal heaving, force follows an increase of thedepth fro. the ground surface and has the maxim,umin II certain height from the wall bottom,Based on the ip-situ observation data: thedistribution of horizontal heaving force along thedepth in triangle or trapezoid form has been deriv--ed, having the maxilum value of about 50-100kPafor weakly heaving soils, 100-150kPa for mid-heav--ing soils and 150-20QkPa for strongly heavingsoils(Sui Tieling,l990).The vertical frost heaving force is dependantupon the heav.ing susceptibility of soil and heav--ing constraint, and 'also associated vith thebottoa area eubedded depth and allowable deforma--tion of the foundatio$(Xu Shaoxin,l989): The ,larger the bottom . area'; the lover the hearinaforce caused by the surrounding soil to thefoundation; the deeper thd foundation. the smallerthe freezing layer of soll'beneeth the foundationand the heaving force,5.Hechanism of Frost Damage'of Canal Linings.The frost heave amount at each position of thecanal is Mainly dependant uponthe moisturecondition of foundation soil. The canals withdifferent runs have diff:eredt';frost depths andheaving amounts. The lower part an&bot$ba of thecanal the heave aaount is great. reversely. at thetop part of the canal slope the heaving is small.The frost depth, heaving amount. and alloiabledisp"1aceaent of the canal )inning are the goveringparameters for canal design to prevet frost damage(11 Anguo, 1987),.,.THE HEASURES AGAJNST FROST DAl(qG.8 ' , ", I I# '-.Based on th'e research ,results.qu'ite 'a few ef-.-fective'. methods against frost damage have beenProvided and havwcreated a considerable amoantof economic and societal benefits in practice.1.Thermal insulationThe one these methods is using a water layer,If possible, a certain depth of water above thebottom plate of the structure in winter can keepthe foundation sot1 unfrozen. The depth of watermay be considered to be equal to 0.6 tiaes that ofthe soil frost depth of the corresponding site,Another method is placing a layer of rigid foamedplastic plate with a certain thickneess beneaththe foundation. lining slabs of canal or on theback of retaining vall. The third nethod ia*torake the concrete foundation plate hollow. usingthe air in the cavity pocket to insolate heat.2. Replacement of soilThe non-frost-heaving material of. sand andgravel is used to :replace the frost susceptiblesoils. The particles smaller than 0.05 mm In thereplaceuent meterial do not excaed 10% in weight.The replacement thickness usually is 0.75-1.0 ofthe frost dept and dependent on the distributionof the horizontal frost heavine force,, alonr thedepth for retaining walls, Also. the impermeable!meterial is used and emb,edded into the foundationsoil to interrupt moisture migration a d eliminatefrost heaving.3.Structurl measures'

Different structural types that are effectualin preventing frost damage are applied: such asthe sluice type called the vord "-":the sluice withfoundation in the fore of reversed placing box orreversed arch self anchoring piles involving thepile with the expanded foundation plate, the pilevith exploding-expanding ends, and the pile withbottom beam; the wall consisting of precasthollow boxes filled with non-frost heaving materialof sand gravel. etc: the .buttress retaining vallwith the buttress spans not exceeding the maximumfrost depth and the retaining structure withanchored slabs, etc. For the canal linings, thestru,ctural measures involve flexible structures,such as the embedded membrane lining and asphaltconcrete lining: the rigid structures. such asconcrete slab, concrete slab with beam, concretcslab in the form of "n", wide-shallov concretccha!nel, arc concrete channel, small rater troughof U" trpe(Li Anguo,l987); reasonable arrangementof deformation seams fllled with flexible sealmaterial such as polyvinyl chloride ointment ortar plastic daub.4,The 'other measuresThe other measures involve evasive methods, suchas making the channel alignment to evade thelocation with sol1 susceptive to frost heaving;adopting the canal i.n fill or cut-and-fill a8possible; adopting the tube the or overhead watertrouhg to transfer water: compaction of clay orclayey loam foundation soil to increase its den--sitr; strengthening the waterproofing of canal andsurface drainage; limiting the date of the canalbeing out af operation prior to 5-10 days of theinitial date of the cold season; the date of waterrunning through canal in spring not being priorto the end of cold season: strengthenfng the main--tenance of the structures, etc.FUTURE DIRECTIONS AND RESPONSIBILITIESOver the years, gratifying achieveuents of pro--tect'ing hydraullc structure and canal Iinlng fromfrost damage have been made. Very many questions,hovever. need to be resolved. The study directionand responsibllitiey in the future are mainly:'1.IntensKve studies on the lav of Soil frostheav and technigues of preventing frost damage, thecharacteristics of frost and frost heaving ofcanals, the methods Tor prediction, classificationstandards of frost heaving for the foundation soils2.Studying the formation and developing mecha--nisms of frost heaving forces to select reasonablemeasures of preventing frost damage.1.Perfecting measuring instruments and methodsof temperature, stress etc, used in laboratory and'in site, drawing unitive test operation rules toincrease the accuracy of test data.4.Studying new materials, nev structures andmechanized construction techniques to reduce workcost and quarantee construction quality.5.To sommarize further the experiences in pre--venting frost damage, spreading presented results,raking them real productive forces.REFERENCE1.Li Anguo and Han Shujian, 1987. Brie-duction of Antifrost Technique of CanaEngineering & Frozen Sei!: NO 4 .f Intro-1 Llnlng.2.Sui Tleling, Ma WenJle and ti Dazuo, 1990,Distribution - .- . .. . of Horizontal Heaving Force on TestRetaining Wall. Proc.4th Hat. Conf on Glaciologyand Ceocrrology(se1ection).3.Wang Siyao. 1982, Frost Heave and It8 Distri--bution in Different Layers Influenced by ShallowGroundwater,Journal of Glacioloey and Crropedology.Vol. NO 2.4.Vang Liang, Xu Bomeng and Vu Zhijin, 1983.Property .of Soil Freezing-Thawing and Earth DamConstruction in Winter. Proc 4th Int. Conf, onPermafrost.5.Xu Bomeng etc. 1990, Prevention and Controlof Frost Danage of Hydraulic Structures, JilingScience and Technology Publishing House, China.6.Xu Bameng,l98&. Study on Prevention andControl Technique of Hydraulic EngineeringinNortheast Region, Engineering and Frozen Soi1,No.l.7.Xu Shaoxin,l987, Revier of the Study on FrozenSoil in Hydraulic Engineering of Our Courtrr,Engineering and'Frozen Soil, Ho.4.8.Xu Shaoxin, 1989, Frost Heaving Force onFoundations in Seasonal Frozen Ground Region, Proc.3rd Chinese Conf, Permafrost (selection).9.Zhu Dafu and Lin Suxlng,l986, A Frost HeaveClassification of Canal Base-Soil with ConcreteLining and the Heasures Against Frost Damage,Journal of Glaciology and Geocryolosr. vo1.8, No.3.10.Zhu Piang,l990, Quantitative Study on BasicRelation between Frost Heave of Canal Base-Soil,Eglneering and Frozen Soil, No.1.1294

UNFROZEN WATER CONTBNT IN MULTWRYSTAL ICEXu Xiaozu', Zhang Lixin', Dcng Youshcng', Wang Jiacheng'IU.P.Lebdenko2and E.M.ChuvilinZ'State Key Laboratory of From Soil Engineering, IGGAS,PRC'Geology Faculty of Moscow State University, RussiaAn artificially large crystal iw was smashed and divided into four groups with the grain sizes of 10,7-3, 3-1, and < Imm, respectively. Tbe unfrozen water content of multi-crystal ice was determined bythe nuclear magnetic resonan= technique and by calorimeter. Four factors influencing the unfrozenwater content in multi-crystal ice, including grain sizd of ice, interfaces between crystals, freezing speedand air content in water, were investigated. Results show that the unfrozen water content increases withdecreasing of grain size of ice. The unfrozen water content between interfaces of ice-water-ice is greaterthan that of ice-water-air. By using deaerated water and a quick freezing, we can obtain greatermulti-crystal of icc than the undcaeratcd water and a lower unfrozen water content if other conditionsare the same.INTRODUCTION ,The basic difference between froztn and unfrozen soils is thatice exists in frozen soils. The ice content and ita property in frozenmils are of p at significance for physical and mechanical propertiesof frozen soils. And the ice content and its property depend onthe unfrozen water content in ice to a large extent. A lot of previouswork has been done on unfrozcn water content in frozen soils(h4.Anderson et. a1.,1974, A.R.Ticc et. a1.,1978, E.D.Ershov, 1979),but seldom dealt with unfrozen water content in multi-crystal ice.Xu Xiaozu determined the unfrozen water content in ice made bydistilled water (1987), but didn't describle the changing regularityof unfrozen water content in multi-crystal ice.SAMPLE PREPARATIONTo obtain multiGcrysta1 ice with different grain sizes, it is ne&essary to prepare large crystal ice first. Therefore, we pour distilledwater into a plexiglass container with the size of 15 cm in diameterand 25 cm high. The temperature at the water surface is ,kept at zeM or slightly below zero degrees ccntigrade. The side of the containeris surrounded by insulalion material. Water is graduallyfrozen from the top downwards. Usually, after one week of freezingthe ice thickness may reach to about 10 cm and the diameter of theice crystal is larger than 10 mm. Taking one sample from this massiveice as the sample with a grain size of larger than 10 mm. Afterthat the massive ice is smashed and sieved and divided into threegroups with a grain size of less than 1 mm, 1 to 3 mm and 3 to 7mm, respectively. Each group of mqti-crystal ice is divided intotwo subgroups. One subgroup is emerged in distilled water of zerodegrees centigrade and quick frozen again to create ice-water-iceinterfaces between ice crystals, and the other subgroup is keptwithout distilled water to create ice-water-air interfaces.To investigate the influence of freezing speed and air contentin water on the unfrozen water content in multi-crystal ice, twoother samples are prepared. One sample is made with Water fromthe previous determined sample with the grain size of the ice crystalbeing larger than 10 mm and another sample made from deaeratedwater and both of them art frozen with a high sped. All of thesamples mentiontd above are frozen at minus 20 degrees centigradefor 5 hours and arc determined by the nuclear magnetic resonancetechnique in a warming cycle and by a calorimeter.RESULTS AND ANALYSISAs a kind of grained material, ice crystals, like soil particles,can absorb a certain amount of unfrozen water under the conditionof the temperature being bclow zero degrees centigade because ofthe existence of free energy at the surface of particles. Figure 1shows the curves of unfrozen water content 1s. temperaturc for theice crystals with the different grain sizes mentioned above. It can beseen from figure 1 that the unfrozen water content increascs withdecreasing of temperature in the power form. From figur 1-a it canbe seen that if the temperature is the same, the unfroztn water contentof multi-crystals of ice changes with the grain size of ice crystaland can be divided into three groups: the minimum a grain sizeis larger than 10 rnm, the middle grain size is from 1 to 7 mm andthe maiimum grain size is less than 1 mm. The difference of frozenwater content curves for ice grain size less than 7 mm is not greatbecause of the difference of ice grain size being less after water is filledand being quickly frozen. From figure l-b it can be seen thatunder the conditions of ,ice-water-air interface and quick freezingthe changing regplarity of unfrozen water content with temperatureis the same as shown in figure I-a, but the maximum of unfrozenwater content for the case of IWA is less.The curves in figure 1 can be expressed by the regressive equationsshown in table l. From table 1 it can be seen that the relatedcoefficients are greater than 0.94.1295

3.5 1 a'1d= 7-3-II1I\ITemperatureOCATemperature "C31d-3-lmm2{ IWAIIIII.Temperature 'CTemperature "CBd< ImmI1IFig.1 Unfrozen water content in multicrystd i a with differentgrain sizes VI. temperature (A-for the WI case. and B-for the IWAcase). Figure 2 shows the curvesof unfrozen water content vs. temperaturefor multi-crystal ice with grain size leas than 7 mm and interfaceof ice-water-ice and ice-water-air respectively. Prom figure2 it can be seen that if temperature is the same, .the unfrozenwater content of multi-crystalice with interface of ice-water-ice ishigher than that with interface of icewater-air. There is no muchdifference for different grain sizts.Figure 3 shows the curves of unfrozen water content VS. temperaturefor samples of grain size larger than 10 mm and deaeratedand undeaerated distilled water being quickly frozen. Prom figure 3it can be seen that if the temperature is the same, the unfrozen watercontent of t$e sample made by undeaerated distilled water beingquickly frozcn is much greater than that of the sample with grainsizez greater than 10 mm. It indicates the freezing rate has a greatinfluence on the unfroztn water content. Microscope observationon the thin section indicatm that the freezing rate influeces thegrain size of ice crystals. The higher the freezing rate, the less the icecrystals is and the less the ice cryatal, the higher the unfrozen watercontent. Compared with the unfrozen water content of the deacratedsnmple, the unfrozen water content of the undeaerated sample ishigher even if the freezing condition is the same. Microscope observationindicates that the grain size of the deaerated sample is gteaterthan that of the undeaerated sample.CONCLUSIONSThe unfrozen water content of multi-crystal ice is controlledby the grain. size and the interface between ice crystals. If other-20 -15 -10 -5 0Temperature 'C~ig.2 Unfrozen water content of multicrystal ice with different, interfacesTable 1 Regressive equation of unfrozen water mntentof multi-crystal icefgrain regressive coefficient relatedsize,mm equation A B cocffcient> 10-0.05498 1.08953-1-0.04259 1.07233-7 *-0.05224 1.30451-3-0.04568 1.08381-3 * Y -(AX+Br -0.05967 1.3170

,\D>lOmmI vacuum TT- unvacumm8 TtTemperature "CCFig.3 Unfrozen water content in multicrystal ice with deaeratcdand undeaerated water quickly frozenconditions are the same, the unfrozen water content of multi-crystalice increases with decreasing grain size and with increasing icecontent.The freezing rate and air content in the sample are the importantfactos for the grain size of ice crystal formation. If other conditionsare the same, grain size of ice crystals decreases with an increasein the freezing rate and air content.The maximum of unfrozen water content for IWI and IWA interfaceis less than 3.5 and 1.5 %, respectively.ACKNOWLEDGEMENTSThe authors wish to express their thanks to senior engineer,Tao Zhaoxiang and Mr. Gu Tongxin for their determination bycalorimeter. This work is very important for calibration ofunfrozen water content taken from NMR.REFERENCESAnderson,D.M., TI=, A.R. and Banin A., 1974, The water-icecomposition of Clay / water system, CRREL Research Report322, pp.10.Tim, A.R., Burrows,C.M. and Anderson, D.M., 1978, Phase compositionmeasurements on soil at very high water contents bythe pulsed nuclear mapetic resodance technique, reprintedfrom Moisture and frost-related soil properties, TransportationResearch Board, National Academy of Sciences, pl1-14.Xu Xiaozu, Oliphant. J.L. and Tice,A.R., 1987, Factors affectingwater migration in frozen soils, CRRBL Rport 87-9,pp16.Ersh0v.A.D.. 1979, Water phase composition of frozen soh,Moscow Univershy Publishing House, pp189.1297

THE THAW SETTLEMENT OF RAILWAY FOIJNDATIONSTN PERMAFROST SWAMP REGIONSYang Hairong' Liu Tieliang' and Guan Zhifu2'Northwestern Institute, Railway Ministry Academia, China2Gingwu nepartment of Railway Bureau in Harbin, ChinaIn the permafrost swamp areas of northeast China, thaw settlement is one of themain causes of damage to railway foundations, which are very difficult to repair.Since the establishment of the line, sucessive reparative measures have beentaken, but they have not been effective and the damages continue to worsen.Through observation and analysis of thaw settlement of railway foundations, inthe permafrost swamp regions in China, a summary of the theoretical and syntheticcountermeasures is given to prevent thaw settlement using industrial materialfor water insulation.SURVEY OF NATURAL GEOGRAPHYThe permafrost region of Da Hinggan Ling isloc~t.etl in the eastern fringe uf Nei Meng Guand in the northwest of Heilongjiang. Thewinter is frigid and long, with the period ofncgotivc temperatures continuing for around7-8 months, the summer is short. The highesttemperature is about 36"C, in the north, theextreme lowest temperature is -52.3"C. Theyearly,average temperature is -2 - -6.2OC, theyearly precipitation is 450-550 mm, and theyearly evaportion is about 820 mm (Deer Buer).The altit-ude is between 500-1300 m, and therelative difference in elevation is generallyless than 300 m. This region is one'of themain distribution areas of permafros in Chinaand swamp land has an extensive disdibution.The railway length crossing the permafrostregion is about 530 km. Geological phenomena,such as pingo, icing, massive ice, et.c.,-makerailway construction and operation verydifficult.According to observation and analysis ofpreventing settlement in the railway foundationin the past, new methods of keeping ground waterand surface water from penetrating into therailway foundatiun have been researched and theevolutional tendency of settlement deformationin the railway foundation can be forecasted(Yuan Haiyi, 1987).THE LOW EFFECTIVENESS OF RRPAIRMENTSThe problem of settlement in the railwayfoundation in permafrost regions mainly occursjn swamp sections with poor drainage,. In thepast, other than the step of controlling theheight of the embankment, the synthetic methodsof constructing a berm in one or two sides ofthe foundation and the installation of a laddershaped drainage ditch 20 m from the slope bottomwere used (Fig.1).Pig.1 Diagram of Embankment Berm1. Roadbed BermBerms installed in the foot of the slope ofthe roadbed /re most often used to prevent thawsettlement of the roadbed in permafrost regions.According to statistics, new berms were constructedto prevent thaw settlement in swampregions from 1974 to 1990, and the extendeddistance is as long as 126.7 km. Through observation,it has been shown that berms installedin settlement sections are effective at Eifst,but with time, settlement damage continuouslyappears in parts of the roadbed. The causes ofthe thaw settlement and the ineffectiveness ofberm are analyzed as follows:(1) Effect of surEace water: Because of poordrainage in the drainage ditch or in the areaaround the slope bottom of the roadbed, groundwater and surface water accumulates in the slopebottom of the berm which causes heat to successivelypenetrate into the base of the roadbed.This causes non-uniform deformation of theroadbed and parts of the permafrost thaw underthe base of the roadbed. Successive heavyrains are one of the direct causes of fastsettlement in the roadbed.(2) Effect of ground water (suprapermafrostwater). In sections with ground water, when thedrainage ditch cannot prevent surface water fromsuccessively supplied ground water, surface waterpenetrates into the ground and flows into theroadbed from the drainage ditch causing the berm1298

to lose its effectivity.(3) Ef,fect of berm without enough length.Joints are set up at the heavy settlementsections of the roadbed berm but are not setup at slight settlement sections in the bermof .the road bed so they are often not connectedproperly.(4) Effect of ballast puddle ditch. Theballast in the upper-layer of permafrost belowthe roadbed forms a deep potshaped ballastpuddle ditch which isadvantageous for theaccumulation of ground water. Based on investigations,ballast puddle ditches can be as deepas 2 m, Tf the ballast puddle ditch is notdealt with, even if berm is constructed on oneor two sides of the raodbed, seasonal precipitationpenetrates into the base of the roadbedthrough the ballast puddle ditch causing thepermafrost below the base to thaw and causingroadbed settlement.(5) Effect of the filling material used inberm. Tatou tussock that grows locally is oftenused as filling material for berm in the northeastof China. But due to a lack of water thetatou tussock dies and rots, thus causing abiochemical reaction which causes the temperatureof the frozen soil to rise along with therotting process. Meanwhile, its effect on theheat preservation and insulation heat graduallydecreases with time (Table 1). With time, themigration becomes dense, its effect of protectingagainst heat gradually worsens, so afterthe berm is constructed the effect of preventingroadbed settlement gradually decreases.Table 1. Change table'of the coefficient thermalconductivity of tatou tussocklimitCoefficientthertnal 0.1063conductivityw/ mkGrass of Grass of Grass oforiginal berm after berm afterplace 2 years 10 years0.2745 0.5358(6) Effect of berm with enough height. The~urpose OC constructing berm is to decrease theamplitude of the upper boundary in the slopefoot of the roadhed to keep the surface andground water from penetrating into the roadbedbase. So the rule of providing the berm withthe proper height and filled with tatou tussockor peat cannot hp less than 1.0 m. Rut inpractice, some berm do not have the abovementioned stipulated height.2. TlrainaRe Ditch(1) The distance of the drainage ditch farfrom the slope foot of the roadbed. The mainreason for this was to keep the natural surfacenear the slopefoot from being destroyed by thedigging of the drainage di,tch and to prevent ,water from flowing into the frozen soil aroundthe ditch which would cause thawing and affectthe stabilitv of the roadbed. Practice verifiesthat the distance of the drainage ditch fromthe slopefoot should be 10 m. Drainage ditchesshould conduct prevention treatment, if not,even if the drainage ditch is- far from theroadbed, a large amount of settlement occurs inthe roadbed.(2) Shape of the draindge ditch. Aconventional ladder shaped section is continuallyused in drainage ditches in frazen soilregions. There are many problene with thisform, such as it is difficult 'to rnpstruct, itdoesn't have effective slope prevention and itis easily blocked. The Googwu Qe'part.ment ofRailways, based on many years of practicalresearch, used a new shape for drainage ditches,an inverse "T" shape used on the heave section.The self buried principle was used far thepurpose of protecting the slope of the ditch inareas where thaw and collapse occur in thefrozen soil regivn.INDUSTRIAL MATERIAL USED FOR INSULATION AND_.WATER PREVENTION IN ROADBED SETTLEMENTRENOVATIONSRenovations are a basic step for preventing .surface and ground water from penetrating intothe base of the roadbed and for preventingsettlement of roadbeds in permafrost regions.In recent years, railway departments have usedintricate methods of renovation, by buryingindustrial material for preventing water penetrationinto one side of the upper reaches of theroadbed and for preventing ground water fromflowing into the roadbed. Based on the depthof ground water, the embedded depth of thewater prevention material is divided into shallowand deep embedded depth.1. ShAllow depth: in t.he area with moreplentiful surface water and shallower groundwater (less than 1 m), when it is difficult forone side of the upper reaches of the roadbed tovertically drain water, a large amount ofsurface and ground water may cross the roadbedAnd flow downwards causing the roadbed toseriously subside. In that condition, themeasure of dredging and burying industrialmaterial for insulating against ground andsurface water penetrating into the r0adbe.d istaken. For example, in some roadbeds in the .northeast, fhe height at the center of theembankment Is 1.5-1.7 m, the roadbed has alarge amount 6f subsidence every year fromsurface water and shallow ground water crossingthe embankment Erom the upper side of the line.Thus, other than the constructed berm in thesection, a two layered polystyrene foam slab ispaved at the bottom of the berm, and- a onelayered chloroprene rubber slab is paved underthat. The two kinds of slabs are verticallyburied to a certain depth under the aroundwater table 2.0 m from the slopefoot of theother side of the berm and standard berm isconstructed (FZg.2).After the water insulation industrialmaterial is set up in the section, ground andsurface wat.er could not flow and penetrate intothe roadbed, but a large amount of wateraccumulated in the berm slopefoot at the sideof the upper roadbed. In order to eliminateharmful accumulating water on the roadbed or toprevent damage in the nearby roadbed sectionwithout water insulation, a new culvert isconstructed in the lowest area of the roadbedso that water accumulation in the slopefoot ofthe berm quickly drains and makes the base ofthe nearby embankment in a long-term dry state.In the low embankment, when the depth of theballast puddle ditch is more than 1 m, one layerof soil material to prevent water penetrationshould be paved under the roadbed to preventthe damage from roadbed subsidence and forprotecting against potholes that allow rain to

Fig.2 Diagram of a buried shallow shapedinsulating water slabpenetrate into the embankment and the basethrough the ballast puddle ditch.2. Deep depth: whed the depth of the groundwater is deeper or themokarst lakes exis-t nearthe roadbed, because the water penetrates intothe roadbed base and produces a large amount ofsubsidence, the treatment method of using deeplyburied water insulating industrial material isused. An example is that there exist severalthermokarst lakes which are 7 m from the slopefootof the roadbed which has a height of 3-4 m,in order to prevent damage from a large numberof subsidences caused by the lake water pentratinginto the roadbed, water insulating industrialmaterial is buried in one side above the roadbed.The Rround .water bearing bed is dug awayand a chloroprene rubber slab is verticallyburied, with the larRest buried depth being 4 m,then soil is refilled to the surface to form a"water stoppage wall", meanwhile the surfacewater between the slopefoot. of the embankmentand the water stoppage wall is prevented from 'penetrating into the roadbed and this preventssubsidence of the roadbed. On the other hand,a drainage ditch is coastructed and the originaldrainage ditch is renovatbd in the upper side ofthe thermokarst lake so that the surface watersupplying the thermokarst lake is cut off.Before renobation. there is a little water flowin the original drainage ditch and in the nearbyroadbed. After renovation a large amount ofwater flow obviously accumulates in the ditchand drains awly through the bearing culvert.After the above mentioned section is renevatedby dredging and water insulation, the finalresults are being observed at present.which is suited to a permafrost environment haveobviously decreased. Aagiorpermous forest,fallow and herbaceous vegetation which aresuited to extensively distributed talik environmentare incte@sing. Meanwhile the animal lifewhich is found in permafrost regions has beensignificantly changed (Shi Yafeng, 1986).Because the global climate is rapidly becumingwarmer the southern boundary of permafrost willvahish frum the Da Hinggan Ling region in Chinaby the year 2030. The degenerative process ofpermafrost in this region will also be furtheraggravated by the increasing development andconstruction taking place. This process willhappen very quickly so that in the next tenyears the original composition and cryogenictexture of the frozen soil layer will beseriously changed. The bearing capacity ofpermaf,rost and thawing soil will be obviouslydecreased after the temperature rises, thedeformation capacity will quickly increase andthaw settlement action will also increase.Meanwhile a large number of engineering buildingswill be destroyed by many deformations. Inrailway roadbeds in permafrost swamp regionsthere will be produced lasting, undepleated andlarge scale thaw settlement so the question ofthaw sertlement in roadbeds which is originallydifficult to renovate will become more intricate.How this situation will be handled can not bedetermined at present.EXFERENCtSShi Yafeng,(l986) Big Changes in Climate andXnvironment will Appear in Whale Earth,Scientific Newspaper, 3.29.Yuan Haiyi, (1987) Definating Permafrost in DaHinggan Ling Region, Proceedings of the ThirdChinese <strong>Conference</strong> on Permafrost, PublishingHouse of Science.DBVELOPMENT TREND OF SUBSIDENCE DAMAGE OFROADBEDS IN THE FUTURESince the last cencury, permafrost in undevelopedregions in Da Hinggan Ling of China is inthe stage of degeneration due to the gradualwarming of the climate. In fact, not only theclimate has the trend of warming in Da HingganLing region, but the permafrost as well, whichhas produced regional degeneration in the southand southwest regions, these regions have notbeen influenced by human activity in the lastten years according to observations. Theexisting animal and plant life have beengreatly changed by permafrost degeneration, sothe plant coenosium and animal type thatoriginally existed in the reRion have seriouslychanged.In the southern regions with permefrostdeReneration not effected by human activity(such as the upper reaches of the Nanwen riverand Hala river), latch, pinus sylvestris, etc.,

THAW-CONSOLIDATION OF UNSATURATEDFROZEN SOILYong Lifeng, Xu Bomeng and Lu XingliangResearch Iistitute of Water Conservancy Committee of Songhua-Liao Rive Basin, ChinaIn recent years, wc have conducted research on the thaw-consolidation of unsaturated frozen soils. Takingloam obtained from the right tank of Weilong River as an example, the research results arc discussed.The results indicate that the saturation of frozen soils has an important effect on theirthaw-consloidation properties. Consequently, the quantitative evaluation of thaw-consolidation offrozen soil should be determinde in terms of its saturation. For unsaturated frozen soils, thaw-consolidationproperties will be affected not only by their moisture content, but also by dry densities. Comprehensiveeffect of the both factors on thaw-consolidation properties would not be the same as that ofsaturated from soils. Based on the test data, the cquations of calculating thaw-consolidationcoefficients of frozen soils are presented.- INTRODUCTIONThe construction experiencein cold regions tells uq that infrozen ground, cxpecially in permafrost zones. thaw-consolidationof frozen foundation soil is the main causc of the engineering struoture settlement, even damage. Thus, in years of study, the principlesand features of thaw-consilidation of frozen soil is one of the mostimportant research areas.The previous investigations(Wu Ziwang 1981,Chcng Xiaobo1981, Zhu Yuanlin 1982 and Tong Changjiang 1985 'et al.) showthat thaw-consolidation of froun soils were mostly evaluated bythe moisture content. Strictly speaking, it is suitable only for thesaturated frozen soil but not for unsaturated frozen soil. Tocorrectly evaluate the thaw-consolidation properties of unsaturatedfrozen soils, taking loam as an example, we have conducted experimentalresearch and obtained the special law which dcscribesthe multiple influence of dry density and moisture content on thethaw-consolidation of unsaturated frozen soils.TEST CONDITIONSExperiments were carried out on the frozen soil thaw"compressionmeter. The soil samples consisting of medium-loam andhcavy-loam were obtained from the right bank of Heilong River.To bring to light the law of thaw-consolidation properties offrozen soils, besides a fcw experiments made undisturbed frozensoil samples, we mainly carried out simulated tests with the soilstaken from in-site. Firstly, roil samples were prepared to have differentdry densities and moisture contents, then frozcn in single direction.The samples were 2.5cm high with diameter of 6.4cm. Thepreload on the samples was 0,OlMPa. The thaw-consolidationdeformation was directly measured with a micrometer.ANALYSIS OF TEST RESLJLTS- . . . . .. . . . . . . . -Thc magnitude of thaw-consolidation is directly govcrned h!,the decrease in porosity, which is related to the dry dcnsitv andmoisture content.When I'rozen soil is in saturated condition, the moiqturc anddry density or porosity have a quantitative telatyion as Ibllows(YinZhijan,l980) *Gr, w"- I(1)rwG - rlorGw=e (2)where G is specific gravity of the soil grain: W, r,and e are the saturatedmoisture content, dry density and porosity of frozen soil,rcspectively:r,is the density of pure water at 4OC.Thcrefore. the saturated moisture content may be trtken tocharacterize the magnitude of the void ratio, the reverse is ture.Thus the general effect of saturated moisture content and dry density(or porosity) on the frozen soil thaw-consolidatipn may betermed as single effect of moisture content or dry density. namely,thaw-consolidation coefficient A, (relative thaw-consolidationquantity expressed as a percentage) may be expressed as follows.orA, = f~ (W) (3)A, =f, (rd) (4)However, when frozen soil is in an unsaturated condition,there is not a definite quantitative relation between motsture con.tent and dry density ( or void ratio). Thus, depending upon theconditions of dry density and moir,:ure content, the thaw-consolidationcharacteristics of unsaturated frozen soils will be not the

0,97("-same as that of saturated frozen soil. Based on the testdata, the cates that the initial thaw-consolidation moisture content is closethaw-consolidation characteristics of unsaturated frozen soil are to the plastic limit w, of the soil. The data in fig.] show that therediscussed.is a similar conclusion obtained by the author of using unsaturatedsail. Thtkby, in ,pr&tice, when the initial thaw-consolidationThe Relation of Thaw-Consolidation Coefficient to Dry Density moisture content,is unknown, it may be replaceed with wI.and Moisture Content The objective of the initial existance thaw-consolidation moisturecontent means that only the part of the natural moisture con-The test results indicate that thaw-consolidation characteris- tent over the initial.thaw-consolidation moisture content is the ef.tics of frozcn soils depends upon both dry density and moisture ficient moisture causing the thaw-consolidation process of frozencontent, the test data is ploted in Fig.1. The general tendency is that soil. Thus, in general, this part of moisture is called the efficientwith either of dry density or moisture content changing, variation thaw-consolidation moisture content W, (Zhu Yuanlin 1982),of thaw-consolidation coefficient would take place, which is obvi- namelyw, = w-w, (5)The efficient thaw-consolidation moisture content essentiallyreveals the effect of moisture content on thaw-consolidation characteristics.The Law of Combined Effect of Dry Density and Moisture onfheThaw-Consolidation Property of Frozen Soilw (%)Fig. I Relationship of A, with r,and w(heavy silty loam, W,= 17%)ously different from the thaw-consolidation characteristics of saturatedfrozen soils.Simultaneously, the thaw-consolidationcocficient decreases with dry density increasing and moisture contentdecreasing.It is not dimcult to understand the phenomena mentionedabove. It is well kmownthat in geotcchnique, dry density and voidratio are the two indexs describing thc density of soil. The highdensity of soil shows the intErvals of soil grains being small, andccrtainly,the relative displacement between soil grains in thethaw-consolidation process would be small, Thus, the thaw-consolidationcQefficient is in inverse proportion to dry density. Besides,in the process of soil freezing, the compactive action on thesoil grains formed by the ice crystals strengthcns as moisture contentincreases. Such action to thaw-consolidation can not becaught up by the dcad weight. Thus, even if two samples have thesame dry density, while their moisture contents are different, thethaw-consolidation coefficient for the two samples will be not thesame. In this condition the higher the moisture the greater thethaw-consolidation coemcient is. 'Besides, it can also, be seen from fig.1 that undcr unsaturatedconditions, the initial moisture content of frozen soil thaw-consolidationw, is not a definite value as it is under saturatedconditions. At this time, it varies as dry density changes althoughthe range of such variation is smaller. The research.concerned onthe initial thaw-consolidation moisture content of frozen soil indi-In summary, in order to reflect the combined effect dry of densityand moisture content on the thaw-consolidation property offrozen soil, the test data were arranged, taking RW,/ r, as a vari- 'able and replacing the initial thaw-consolidation moisture contentby plastic limit. The results are shown in fig2 It is illustrated in thisfig. That there-exists a good linear relation between the thaw-consolidationcoefficient of heavy loam and variable. It is demonstratedfrom this that taking R = W, / rd as a variable can reallyreflect the inherent law of the combined effect of dry density andmoisture content on the thaw-consolidation coefficient. By analyzingwith the fkst square method, the optimum probable value ofthe thaw-consolidation coefficient A, of frozen heavy silty loamcan be obtained as followsw- wpA ~ - 0.02)(6)rd(Coefficient of correlation r = 0.98)Similarly, using the above mentioned method, the relation ofthaw-consolidation coefficient of medium loam to its dry densityand moisture content can also be obtained as followsw- wvpAD = ].I(- - 0.0039) (7)rd(Coefhient ofcorrelation r=0.99)It can be seen frbm above two .formulas that while A,=O,(W-W,) / rd is a constant, namely(W-W,)/rd=B (8)In other words, under unsaturated conditions, when the relationbetween moisture- content and dry density is satisfied, thethaw-c~nsolidation phenomenon would not occur. In formula (81,B is the parameter concerned with soil properties.

3PPig.2 Variation of A, vs. (w-w,) / r,Experimented Values (Ae, YO)Pig.3 Comparision of computed and tested values of A,Verifying and DiscussingWhile putting the computed values of formulas (6) and (7) tobe compared with corresponding test values(Fig.3). as will be readilyseen, both of them coincide well. Most of the relative errors bptween computed and test values do not exceeds 15%.Having formula (8) simply changed, we asn obtainWe= W,+Br, (9)which verities the above related analysis that under unsaturatedconditions the initial thaw-consolidation moisture content offrozen soil is in direct propertion to dry density,In existing information, the combined effect of dry density andmoisture content on thaw-consolidation of frozen soils is not considered,the belief being that either of then may be used to expressthaw-consolidation characteristics of frozen soils: As a congequance, two formulas considered to be equivalent wire derivedfrom dry density and moisture content, correspondingly. Both experimentsand computations indicate that only under saturatedconditions can the two formulas be equivalent and unitive, that isto say, what is expressed by them is the thaw-consolidation properties in a given condition of saturated soil. Therefore, they are notsuitable for unsaturated soils. Whereas, formulas (6) and (7) haveunitedly reflected the multiple effect of dry density and moisturecontent on thaw-consolidation properties of Frozen soil, eliminatingcalculated deviations from using different formulas.CONCLUSION -. ."The following conclusions are obtained from thisinvestigation:1. Under unsaturated conditions, the thaw-consolidationproperties of frozen soils depend upon the combinative effect of drydensity and density and moisture content, the law of which isA. =A( w- wr*where A and B are parameters concerned with soil properties.2.Under unsaturated condition, the initial thaw-consolidationmoisture content, plastic limit and dry density satisfy following relationshipW, = W,,+Br,3.The calculated values obtained by-the formulas in this papercoincide well with the experimental results, the relative deviationbetwcen them does not exceed 15% in general.4.Through the study of thaw-consolidation properties ofunsaturated frozen soils, it is indicated that the method used in thisstudy has generality without being limitted by soil properties. Itmay also give as a reference to the thaw-consolidation propertiesof seasonally frozen soil.REFERENCES"Wu Ziwang etal, 15181, Preliminary studyof thawing-sinkingcharacte-ristics of frozen soils, Professional .Papers ofLanzhou Institute of Cryopedology, Chinese Academy of SciencesNo.2, Science Press, Bei jng.Chen Xiaobo, 1981, The thawing-sinking and compression Charaoteristics of frozen soils of Muii region, Qilian mountains, ProfessionalPapers of .Lanzhou Institute of Cryopedoligy, Chinese' Academy of Sciences, No.2, Scienq Press, Bcijing.Tong Changjiang, 1985, Thaw-consolidation behavior ofseasonally frozen ground, Fourth Int. Symp. on Ground Free2ing,pp.l59-163.Yin Zhijian et al, 1980, Geotechnics and Foundation,Pub.Constructional Industry,Beijing,China.Zhu Yuanlin et al, 1982, The thawing-sinking of frozen soils, Proceedingsof Chinese Society of Glaciology and Geocryology<strong>Conference</strong> (Selection, Geo-cryology), Science Press, Beijing,China.- 1303

NOTCHED CHARPY BAR IMPACT TEST ON FROZEN SOILIYu QihaoZhu YuanlinINTRODUCTIONIn permafrost regions many engineering were in progress, suchas making foundations of buildings, highway and bridges, water'conservancy projects und exploring minerals and so on. None ofthem has no relationship with excavating of frozen soil. In this case,some problems about the excavating strength of frozen soil weeinvolved. For lacking of cxpermeatal data and theoritical guid. engineerscould not design the machines efftctively and economicalyto process the frozen soil (Zhang S. X., 1992). So the study on suchproblem is of great significiancc for the designing of the excavatingmethod of frozen soil in various propcts.There are only a few materials about the cutting and Smpactingpenetration strength of frozen soil. The earliest studies started in1950s. The institute of exploring mineral in USSR applied a standardcut with a simple ection (flat wedge) to study on the cuttingstrength of fromn soil (a. A. Cuitocich, 1985). At the same time,the institute of road in USSR adopted a standard impacting headto study on the strength of frozen soil. In recent period, with theneeds of projects, Zhang Zhaokiang (zhang S.X.. 1992) from Beijing Agriculture Engineering University did a deep research on theprocess and speciality of the impacting penetration with sclf-designed quick triaxis experiment plate and auto-impactingpenetrator. Meanwhile he did some experiments on the cuttingstrength of frozen soil. In the same period, Wang Shuwen (WangS.W., 1989) studied oh the relations among the impacting pcnctrationindex, impacting penetration resistance and the various influencingfactors and hargot some good results. In addition, Ladanyi(Ladanyi, 1976) has also done some experimcnts on impacting penetrationof frozen soil at different impacting speed. 'The purpose of this study is to investigate the effect of temperature,water content and drydensity on the fracture absorbed energyof remolded frozen Lanzhou sand, Lanzhou silt and Wuxi clay.The test progrom was as follows: 1) the samples with samc dry density(sand and silt arc 1.5g / cm3,clay is 1.4g / cm') and differentwater content (sand and silt from 8.9 to 18%, clay from 23.3 to28.8%) were tested at three different temperature -5, -10, -1S0C,respectively. 2) with same water content (sand and silt are IO%,clay is 27.7%) and different dry density (sand and silt are 1.4 toI.SSg/ an3, clay is 1.3 to 1.45g/ cm?, the samples were also testedat three diffcrert temperature -5, -10, -15°C.SAMPLES AND THERE PREPARATIONThe materials used in this study was rcmolded Lanzhou sand,Zmnzhou silt and Wuxi clay. There main physical propcrities ure asTable 1, Table 2, and Figure 1.E3 :] \20100 10 1Equivalent spherical diameter"4Fig. 1 Cumulative mass percent finer vs. diameter of Wuxi clay

Table 1. Physical Parameter of SiltWSoil5.4 58.6 34.3 1.7 24.6 11.7\hR.2 ' n e Princblc of imDact machine1Table 2. Physical parameter of Lanzhou sandGrainsizc(mm) r0.5 0.5-0.25 0.25-0.1

where CVN is the absorbed energy, A and B are parameters$is minus temperature (°C),B,is a reference temperature (-1°C)The Jnflucnce of Water Content" ".The influence of' water contcnt on impacting test is similar tothat of temperature. The cementing ice content increases with increasingof water content under the same conditions. This makesthe pores become more and more smaller. The strength of the samplesincreases. In terms of the absorbed energy, it also increaseswith the increasing of water content (Fig.6, Fig.7, Fig.8). But thetrends are different for different types of soils. The results can bedescribed respectively as follows:for sand:CVN = (C * W+D) (3)Ffor silt:for clay:CVN ='w / (E. W+E) (4)CVN = GLogW+H * (5)where, CVN is the absorbed energy, W is the water content, C, D,E, F, G,and H are test parameters.% c2001rd= 1.4 g/cm'. 2 4200' -4 i8 -io -iz -14 -16Temperature ("C)Fig.5 The absorbed energy vs. temperature with different watercontentThe Influence of Dry Density~"~~~With increasing of dry density, the distance between soil grainsbecomes shorter. The connecting surface and cementing ice contentincrease. This results in the increasing of the linking strengthamong soil grains. So the absorbed energy increases with increasingor soil dry density. The data show that the frictional angle of sandstarts to influent on its strength at critical dry density. And itsstrength increases quickly when the dry density gets over this criticalvalue (Chamberlain, 1972).In the same general trcnd, there are different laws about theabsorbcd energy changing corresponding to sand, silt, and clayrespectively because of internal reasons (Fig.9, Fig.10, Fig.11). Underthe conditions of same water content and dry density, thechanging scale of absorbed energy of sand is 12% higher than thatof silt. Inparticularly, the curveof the absorbed energy changingwith the dry density of silt is clearly different from that of sand andclay. It increases slowly in non-liticar type when the dry density is1306

lower than 1.458 / crn’and increases tast in similar lincar type whenthe dry dcnsity is higher than 1.45g/ cm3, The grains of clay arevery small and their ratio surface is very big. It’s difticult for thegrains to conncct well each other. Its absorbed energy is mainlyfrom the amount of ccmcnting ice. The internal frictional force haslittle effect on the absorbed energy. Its changing law likcs that ofsand.We work out three equations as follows:for sand:CVN = (K - r,+L)”4 (6)for silt:for clay:C VN = (7)CVN = R - Sin(r,)+S (8)where, r,is dry density (g/cm’);K,L,M,N,R, andparameters.S arc tcstCNCLUSIONSWith thc traditional notched Charpy bar impact test in fracturemechanics, the value CVN of frozen soil were firstly investigated.It shows that the results are resonable and this method canbe uscd in the study of frozen soil, and the methods and theories infracture mechanics can be completely introduced into the field offrozen soil mechanics.Following with the dccrcasing in temperature, thc value CVNincreases continuously. The CVN changing range of sand is notgreat. But the range of silt and clay is bigger than sand .With increasing of water content, the absorbed energy also increasesin non-linear for the three types of soils, The send is biggestin changing range and at the point of 15% it obviously has an in-Ilection point. The changing range of silt and clay is relativelysmooth and there are not inflection points.Following the increasing of dry density, the CVN value alsohas a similar case as the former. Under the same conditions, thechanging range of CVN value of sand is usually greater than that ofsilt by 12%. Moreover, there is a inflection point on We curve ofCVN VS. dry density at thc dry density 1.45g/cm3. Before this.point thc value of CVN increases slowly, and after increases fast,- REFERENCE~; 4400 I1.4 1.44 1.48 1.52 1.56Dry density (g/ cm’)Fig.10 The absorbed energy vs. dry density (silt)Zhang Shaoxiang, 1992, The Mechanics Characteristics and thererelations of Instantaneous Strength of Impacting Penetrationand Cutting, Ph. D paper, Beijing Agriculture EngineeringUniversity. P3-16.Cuitoviclr,H.A.. 1985, Frozen Soil Mechanics, Science Press, Bei)ing. P464.Ladanyi,B., 1976, Use of the Static Penetration Test in Soils, Cana.Geotech. J. Vo1.13, No.13. .Chambcrlain, 1972, The Mechanical Behavior of Frozen EarthMaperialsunder High Pressure Triaxial Test Conditions.P76-78.Wang Shuweng, 1989, Study on the Technology of Field Testing ofFrozen Soil Strength, M.D. paper, Beijing Agriculture EngineeringUniversity.P3-48.3 ‘4014 1.3 1.34 1.38 1.42 1.46Dry density(g/crn3)Fig.11 The absorbed energy vs. dry density (clay)

and presents H spccial rule on the low ridges,tlanji;r\u;tn area (125°.J~'OO"-1Z5051'13."E.Granite is distributed widel!'. High forests are52'o'(j"-5'"' II 12''N) is 1ocatt.d i n Huma c'ounty,exhuberant. In local areas, the bedrock is bare.Heilongji,lnK Provtnce,. Thc area is qhout 32 km'. Because of the increase of global temperature,In ortif!r LI, select the optimum bchemes fur con- pcrmafrost has detcrlorated on the footslopesstructing Hanjiayuan Furest Bureau, our institute and little remains (CUO Dongxin, 1981). In theundkrtook many hydrogeological and ehgineeringvalley areas around the drainage and branches of.geological investigations, physical prospecting tho Weileigeng, River, the topography is flat andand designs from 1985 tu 1992, and has donethere are n lot of tussock hummocks, In thisdetailed research on the permafrost characteri- area. permafrost is well developed and in thestics and the exploitation' and utilization oflocal bare areas there are taliks (Fig.1).ground water. The research provided scientific, information for constructing Hanjiayuan ForestHYDROLOGICAL CHARACTERISTICSBureau,. ftost damage preventinn and reasonableexploitation and utilization of ground waterThe hydrological conditions in this are.a.atetesources.controlled b!' geology, geomorphology, hydrology,geological texture. and yermafros-t. The struc-DTSTR~SllTIO!4 OF TH.E CHARACTERJST,ICS OF PERMA: tural system belongs to i latitudinal structureFROST IN HANJIAyUkN AREA belt uf Yilihuli, the anticlinorium of Jiagedaqiand the southern edge of the weileigeng rise.Hatijiayuan area lies on the eastern slope ofFrom the results of physical prospecting, thereDa Hinggan Ling. The mean annual air temperature are two structure lines in this area. One is thein the area is -2.Cl'C. the maximum dir tempera- main structure line along Veileigeng River fromture is 38'C and the minimum Bit temperature is west to east, and the other is'along the out -- ----48'C. The atea includes the'weileigeng River flow tributary of Veileigeng and fingjiBg%uand ita branches. which include the inflow andRiver. The two lines belong to a compressionoutflow rivers, Changqingou River, Jinjiagoustructure belt (see Fig.2).River, taohuibaogou River and Xinhuibagou River.Quaternary deposits are simple. There is onlyWeileigeng River originates in the northern granite of the Variscan period from the Neopaslopedf Yili,hule mo'untain and crosses Fuxilileozoic era. The granite mainly consists offram the welt .o east, through Hanjiayuan, anddiorit granite, andesite granite, biotite' finally reaches the Huma River. The width ofgranite, e'tc., which is distributed in the lowthe river bed is 50-90.m, the river depth-isrolling ridges and forms the bedrock of Quater-0.5-2.0 m, and the maximum discharge is 229.42nary deposits. Quaternary deposits are mainlyton/s. The rivers in this .area ate controlled distributed in the valley area. The lithologyhv different factors.is course sand, middle sand and gravel. The-I .The area is in the intermontane valleys. The characteristics of hydrology are controlled byvegetation on the muuntaine are Larixgedlinii, geolog)f, geomorphology, hydrology, geologicalBetula sylvestria, Populus davidiana. etc. Instructures and permafrost. The characteristicsthe valleys, shrubs and gtess are exuberant.are listed below:Some regions have been cultivated.The distribution of 1. The Tvpes of Ground Waterperniafrost is controlledby climate, hydrology,There are two types of ground water: Water ingeology dnd geomorphology.1308

oundary in 1982Permafrost=boundary in 19920 I OOO mpermafrostFig.1 Thedistribution of permafrostfaultRock a Quaternary drilling holehole0 IO00 mUFig.2 Geological texturerock fractures and in pores of the QuaternaryThe type of water In rock fractures 1s HCOSdeposits.The water in the rock fractures isCs-Na 'and HCO$-Ca-Mg, pH>i', the mineralizationcontrolled by an expression structure, the degree is 0.4288/1, and the water hardness isquality is not rich. The water in pores (phreatic 5.42 H. The types of water in the Quaternarywater) i.s influenced by permafrost. It only pores is HC0.-Ca, HCOs-SO -CaMg, ph

The supply of water in rock fractures ismainly from precipitation. It penetrates intothe weathering fractures and texture fractures.Because it is limited by the conditions of topography(the lolling and slope degree). Phreaticwater in the pores of the Quaternary depositsoccurs only, like convexity, in some parts 3fthe drainage area of Weileigeng River. The supplycondition is complex. The quality of the verticalsupply is less, because of the lesser area ofwater accumulation. It is mainly the horizontalsupply from the rock fractures.4. The State of Ground WaterFrom 1987-1992's observation information inlow water periods, the phreatic water in thepores of the Quaternary depzsits reach themaxlmum in July, and minimum at the end of April.The time of the two extreme values is close.This is another characteristic of this area.EXPLOITATION AND UTILIZATION OF GROUND WATERAs described above, our institute has undertskenmany hydrogeological and engineeringgeological investigations and designs, and manyboreholes have been drilled for this purpose.In order to see the situation clearly, representativeboreholes were chosen and renumbered. .Thedetails are presented in Table 1 and 2.ks mentioned above, the hydrogeological conditionsin this area are controlled by geology,geomorphology, geological structure and permafrost.The authors of this paper consider thatthe factors of geology, geomorphology and geologicalstructure are comparatively constant andtheir drainages are not obvious. Rut the deteriorationof permafrost is rapidly changing.In the region, permafrost was distributed inWcilcigcng Yiver and i.ts branch valley's before1782. Thus, in the region, there was no phreaticTable 1. Groundwater in rock fracturesHule Kelling Decreased PenetratingChemicalwater depth coefficientSo. tY Pes(t/d) (m) (m/d)Mineralization(g/l)PHTnickness ofaquifer(m)1 107.39 41.84 0.05782 120.53 47. 60 0.os13 68.6049.35 0.024 126.2318.10 0.10H-C-?.I11-C-N11-NH - C - N0.106 7.40.160 7.50.201 8.00.3188 7.8753975.7781.80Table 2. Phreatic water in pores U T Quaternary depositsHoleso.Welling Dccreased Penetratingxater depth coefficient Chemical Yineralizatiun Thickness ofPHaquifertypesft/dj(m)(m/d)(811) (m)3 777.1h O.Lj0 62245 H - c ,- s 0.101 6.0 - 3.30 737. If) 0.303 575.'13 H-U-C 0 .081 6.5 2.757681.7 0. 410 lr17.lr3 H-C-Y 0.073 6 . 3 4.03Y 3 2 . 9 7 1.200 726 .Rh 11-C-Y 0.128 f> . 3 3.158 555.79 0.3 1079 . Y 9 HS-CY 0.087 6.2 2-65cj 485.48 0.15 926.24 11-c-u 0*08H 6.4 2.59Shh . 44 0.8 457.6 H -c: 0.103 h . 4 2 .06588. 3 0.15 1473.42 H -c 0 . I 61 h .8 2.68I rlh2H. 9 1 .I .35 301.72 HS-C:l o . 2 95 R . 6 2.3711 7'37. I h 0 . -5 2 034. 72 H-C-M 0.089 7.1 2.5712Shb.4L 0 . 1 9 875.87 H-N-C 0 . 0 I I) 0 . 4 3 .OJ1054. OX 0 . h 4 h79. 52 11s-CY . 0.137 6.2 2.951310

cause new problems for exploiting and utilizingthe ground water in this region.1. Water in rock fractures is not rich. Theamount of water in boreholes is 100 t/d. It hasa good quality and is not polluted. It basicallybelongs to calcium bicarbonate water and can beused as drinking wat'er in small residential areasand should be supplied by pipes. The optimumdepth of the well is 120 m, but the well mouthneeds to be well covered.2. Phreatic water in the pores of the Quaternarydeposit;$ is rich. The welling water in asingle borehole is up to 1033 t /d. But thedrainage area is small, the thickness of theaquifer is thin and not suitable for the storageof large amounts of water. From this, it can beused as a k'ater resource for large or middlescale industry. A large mouthed well and penetratingchannels can be used to exploit theground water.3. The increased area of gold mining hasincreased the number of inhabitants, which hasdestroyed the original state of phreatic wateri n the pores of Lhe Quaternary deposits, andhas increased the degree of pollution. Thesituation appears to be worsening and due tothis, the k'ater in the pores-of the Quaternarydepusits is not suitable for drinking water inthe eastern uart of Jiniiaaou. but is suitablefor industrial use.4. Ground water i.n the western part of Jinjiagouis not polluted. The area has not beenexploited at present. The water in the Quaternarydeposits can. be used as drinking water , but aprotection he1.t must be established to preventpollution. It can be exploited with a largemouthed well and penetrating channel.1992,'s investigation information illustratesthat the remaining permafrost in the valleys ofWeileigeng River and in some branch areas, hascompletely disappeared,The lowered permafrost base also representsthe permafrost in its deterioratiwg process.From 1982's survey, the permafrost base was morethan 5 rn (drilled into the weathering layer),1987's surveyed results indicated that the basehad risen to 3.8 m, aid only in local areas was5 rn (near Jingjiagou area). According to this,the permafrost in Hanjiayuan area in Da HingganLing is gradually disappearing at an acceleratingrate.REFERENCESGuan Wuanjun, (1983) Trees in China, ForestryPublishing, House of China, 929pp.Guo Dongxin, (1981) Permafrost zonation of theDaxing Anling Range, Northeast China,Journal of Glaciology and Geocryology, 3(3),1-9.Yuan Haiyi, (1989) Degrading permafrost inDaxing Anling Range. In: Proceedings of theThird Chinese <strong>Conference</strong> on Frozen Ground,Science Press of China, Beijing, 54-58.Zhang Wuanru, (1986) Forestry and soils of ChinaScience Press of China, Beijing, 96pp.RAPID DETERIORATION OF PERMAFROSTThe authors of the paper "The DeterioratingPermafrost in Da Hinggan Ling" (Yuan Haiyi,1999), generally described the deterioration.Through field tests, we verified that when plantlife is destroyed, the deterioration speed ofpermafrost, increases. The main reason is thatthe air tcmperature becomes increasingly warmer.From the information of Mehc meteorologicalstatlon, the air temperature was -5°C from 1957to 1Y70, and -4.4"C from 1971-1985. From theinformation of Huma meteorological station, theair temperature was -2.h8"C in 1985, and -2.Ol0'cin 1987. The r.ising air temperature, in additionto human activity, makes the permafrost deterioraterapidly.Hanjiayuan area, Veileigeng River and itsbranches.are the gold enrichment zones. Themechanized mining and local mining arc growingrapidly. As gold mining is developing quickLy,it causes the prospering and development ofhighways, sideroads and farming. All of thesetactors cause the deterioration of permafrost.\de can say with certainty, that the permafrosti n the area of Hanjiayuan will deterioratefurther following the construction of HanjiayuanBureau, and subsequent damage of vegetation.From the information of 1982's investigation,the drainage areas of Weileigeng River and itsbranches were almost all occpied by permafrostand there was local sporadic permafrost.1985 and 1987's investigation informationindicated that the psrmafrvst in the drainageof h'eileigeng River and its hranchos, haddeteriorated on a large scale. Only in localareas was there permafrost.

, SEASONALLY FROZEN GROUND AND ITS BEHAVIOR ON FROST HEAVEIN THE YUMENZHEN.RECIOM, GANSU PROVINCE, CHINA"Yue Hansen and Qiu'GuoqingLanzhou Institute of Glaciology and Geocryology,Academia Sinica, Lanzhou 730003, ChinaThe field observations in the Yumenzhen region, Gansu Province during 1985-1986 'showed that the behavior of the seasonally freezing and frost heaving of groundwas controlled by the local climatic conditions, geological and hydro-geologicalconditions. Under similar climatic conditions with a mean annual air temperatureof 6.9'C and a freezing period lasting about 125 to 135 days, the groundscomposed of well-drained bed rock and gravels didn't have any obvious frostheaving. The €ine-graded soils ha3 a strong-heaved with a ratio of frost heavingq>6X as the buried depth of ground water Z(1.5 m whea Z equals to 1.5 to 2 m,rl-3-6X; and as 2>2.5 m, the q of fine-graded soils might be as low as

Table 1. Major parameters and characteristic indexeset various sitesYumenzhenObser- Meteorolo- Nantan Team 16 Team 17 Team 8 Team 4 Team 20 Bgishueivationgical Huanghva Huanghua Huanghua Yenma Yenma Y enma duansiteStationClayeysoil138.48Clayeysoil29.031.081.29Clayey Clayey Clayey Clayey Clayey Clayey 'soil soil moil soil soil20.75 31.53 37.79 , 1i.62 26.88 27.422.52 1.62 0.4 >4 - 1.18 1 :721.50 1.09 1.07 1.55 1.28 1.401.58 0.85 3.70 0.65 0.55 6.78 0.2111301130637.155.613.31101.80.81010.79a3137018.5770.524.753.845;9199-0890695.268.88.157511.263.817.55JEnd of Mid. Mid.Mid.Mid, Mid.Mid,Mid.Oct. Oct. Oct. oct * Oct. ocr " O K t * oct.K Nov. 22 Nov. 25 Nov. 20 Nov .20 Nov. 20 Nov. 25 Nov.25 NQV. 20L Dec. 8 Nov. Dec 23 .13 Nov .23Dec .28 Nov. 20M Feb .18 Mar. 10 Mar. 20 Mar. 15 Mar. 1 Mar .5 Ma,r. 15 Mar. 10N Mar. 10 Nov.30 Mar,lS Mar. 10 ' Msr.10 Mar.150 Apr .h Apr .15 .25 Apr Apr.25 Apr.13 Apr..SO Apr -25A: Lithological character: F: Thiakness of frozen ground; I(: Beginning of stable freezingB: Water content before freezing; G: Total amount of frost heaving; L: Beginning of etable frostC: Buried depth of ground water; H: Freezing penetrating depth; heaving:D: Dry unit weight: I: Frost heaving ratio; , M: Time when freesing frontE: Salt content; J: Beginning of unstable freez$ng; reaches maxilava depth;Nr Time whpn mqvimurn froetheaving occurmi0: Time when ground 'thawedout thorouahly.Table 2. Tempeyature at Yumenzhen and Yanaa Farm (1974-1976) I), 2)MonthMeanSite ' Jan, Feb. Mar. Apr, May June July Auh. ' Stpt. Oct. Nov. Dee. . annualItemvalueYurnenzhen A -9.5 -8.0 0.3 8.9 15.4 19.3 21.3 20.2 J4,9 6.8 -2.7 -11.7 6.5B -8.8 -5.2 -1.1 6.3' 18.2 22.4 24.9 20.2 18.4 ,9.0 1.3 -9.2 -8.1Yenma A -11.5 -8.5 -1.7 7.3 14.8 19.3 20.9 19.6 13.7 5.2 -3.6 -13.9 5.2B -7.1 -5.0 -1.3 3.8 10.3 16.6 19.7 18.9 14.4 6:8 -0.7 -6.7 5.8At Mean monthy value of air temperature:B: Mean monthy value of ground temperqrure at the depth of 5 fm.1) Depth of ground water table 11.2,m st Yumenzhen. 1.4 at Yenma Farm:2) The Yumenzhen Meterologlcal Station, 1984, the analyrir end division of the climatic data for 'agricuiture, Yulpen City, Gansu Province. Unpublished.1313

period lasts about 125 to 135 days. The groundbeings to freeze in the middle of October. Themaximum freezing depth of the ground occurred inlate- February,. The thorough tha'wing per.iodoccurred from the middle of April to early May.The frozen period of the ground is as long assix to seven months. Generally, the freeAingprocess in the north begins earlier than in thesouth while the thawing process is to thecontrary.The-diff.erent combination of the geological,and geographical conditions leads to a greatdifference in the development process of theseasonally frozen layer in the time to reachthe maximum frozen depth and to be meltedthoroughly (Table 1). The thickness of theactive layer changed largely at various sites,113 cm in Yumenzhen, 101 to 60 cm in the FinegradedSoil Plain. The formation and developmentof frost heaving of the soil was dramaticallydifferent in the region. On the pluvial fanthere was no obvious frost heaving. On the FinegradedSoil Plain at various sections, alongwith different buried depths of ground water,frost heaving of the soil occurred in differentdegrees, weak frost heaving when the burieddepth of ground water was deeper; strong frost,heaving when shallower. Table 2 also shows thatthe distribution of the frost heaving along thedepth was greatly changeable at various sectionsin this region.THE GEOGRAPHICAL DIFFERENTIATION LAW OF'THE- DEVELOPMEYT OF SEASONALLY FROZEN GROUNDThere are many complex natural geological andgeographical factors influencing the formati.on,development and disappearance of the seasonallyfrozen ground. They can be grouped as follows:1) The composition' (grain size, mineral andchenical composition), texture, and buriedcondition of the rock and soil; 2) The initialwater content before freezing in deposits andits distribution along the depth, the burieddepth and their dynamics a's well as chemicalcomposition.and salt content of ground water;.3) The thermal region of soil layer (Cudlevcev V,A.1978), the geographical differentiation of thedevelopment of seasonally frozen ground isgoverned by the sombination of the rock (orsoil), water, and temperat,ure.-The Differentiation R~sultant. f r ~ m GeomorphicUnitsThe landform is the most important factorresulting in geographical differentiation. Tnregards to the above, the researched area canbe divided into 4 parts: Changma Pluvial Fan,Fine-graded Sotl.Plain, Mt. Yenmabeishan andMt. Hanxishan. These geomorphic units aredifferent in.water-heat conditjun,.lithologicalcharacters, supply and drainage of th4 groundsurface water, and burled depth and flow of theground water, governing the diffc'rence and combinationof soil, water and temperature in timeand space, leading to the differences in developmentof the seasonally frozen ground,As shown in Table 2, because the mean monthlyair temperature at Yumenzhen was higher thanthat at Yenma Farm, the freezing of the groundsurface at Yumenzhen started later than at YenmaFarm: but thawing was earller. From south tonorth, along with the gradual change of thegeomorphic position, the air condition, featuresof the soil, and buried depth of ground water,as well as, the developing condit-ions of theseasonally frozen ground, change correspondingly,on the Changma Pluvial Fan, the arid groundcomposed of gravels and sands with the deeperbur-ied depth of ground water didn't have anyobvious frost heaving under natural conditionsduring the observation period, while on theFine-graded Soil Plain, under natural conditionsthe moist fine-graded soi-1s with a shallowerburied depth of ground water exhibited frostheaving to a certain degree. In Nantan, team 16,team 8, team 20 were examples (Table 1). In Mt.Yenmabeishan and Mt. Hanxiashan, the groundcomposed of well-drained bed rock with a poorwater supply didn't have any frost heaving.The Differentiations Resultant from the BuriedDepth of Ground WaterIn the Fine-graded Soil Plain with similarclimatic conditions, the buried depth of groundwater is the difinitive factor: determining thestructure and chemical cnrtoosition of soil,responsible for the magnitude and distributionof moisture content and thermal regime of thesoil profile, leading to geographical differentiationof seasonally frozen ground, As shownin Table 1, the soil tended to densify with theincreasing buried depth of ground water. Undersuch a natural condition, the soil moisturemainly originated from the recharge of groundwater, So, water content and its distributionalong the soil profile will be logically inrelation to the buried de,pth of ground water.As shown in Table 1, the water content and saltc-ontent in the ground tended to decrease withthe increasing buried depth of ground water.The buried depth of ground water was alsoinfluenced by the development situation ofvegetation directly. Therefore, it was theburied depth of ground water 'that determinedthe presence and change of the various factorsinfluencing, the freezing-thawing conditions,and that governed the different developmentsituation of seasonally frozen ground.The influence of buried depth of Rround wateron the around temperatureThe development of seasonally frozen groundis closely related to the ground temperature.The ground water, as a natural heat source, isgreatly influences the soil temperature duringthe frozen per,iod. As shown in Table 2, duringthe period from November to February of 1974 to1976, the temperature at thh depth of 5 cm atYenma Farm was observed to be higher than thatin Yumenzhen, although the air temperature washigher in the latter. This might result fromthe fact that the t.ablc of ground water wasburied as deep as 11.2 m in the latter while1.4 m in the former. A lower ground cemperature,of course, would lead to a deeper penetrationof the freezing process, as a result, the thicknessof the seasonally frozen ground was observedto be 113 crn at the Yumenzhen and 99 cm at theYenma Farm.The effect of buried depth of ground water onthe freezinR process and freezing penetrationDue to the different buried deuths of eroundwater, there was great difference in the freezingprotoss and fr,eezing penetrati,oh depth ofsoil on the various sections, Generally, themaximum freezing depth of soil would occurearlier at those places wtth a shallower burieddepth of ground water, while the time for thesoil to completely thaw was opposite, Also, themaximum freezing depth of soil would decrease1314

'with the buried depth of ground water decreasing.*For example,-the soil at Team 8 of Yenma Farmwhere the table of ground water was bured at adepth of 0.4 m reached the maximum freezingdepth of 53.8 cm on Mar. 1, thawed thoroughlyafter Apr. 25; at Team 20 where the buried depthof ground-water.water was 1.18 m, it reached themaximum freezing depth of 63.8 cm on Mar. 15,thawed thoroughly on Apr.. 20; at Team 16 wherethe,buried depth of ground water was 2.52 m, itreached the maximum freezing depth of 101 cmon Mar. 20, thawed thoroughly on Apr. 15 (Fig.1).MonthNov. Dm, Jan. Feb. Mar. Avr. MayFig.1 Dependence of freezing depth (a) and frostheave (b) on buried depth of ground waterThe effect of the buried depth of ground wateron the frost heavingThe process. magnitude and distribation ofthe frost heaving along the soil profile variedwith the buried depth of ground water. Generally,as the buried depth of ground water graduallyincreased, the aaximum frost heaving of soiloccurred earlier; the total frost heav,ingmagnitude decreased (Table 1).In order to analyze the quantitative relationshipbetween the frost heaving and'theburied depth of groitnd water, the statisticalquations were used, as shown below,h - 41.35 e"0.016z (1)rl - 94.05where: h, Frost heaving magnitude (cm):rl. Frost heaving ratio (X);z, Buried depth of ground water (cm).The correlation coefficient is -0,995 forEqu.(l) and -0.997 for Equ,(2).The Effect of Human activityThe effects of human activity were in theareas of the irrigation of farmland, the supplyand drainage of water by engineering, and inchanging the soil-water-temperature situationof ground greatly. Generally, the irrigation offarmland and the leakage from canals made the',water content in tbe ground increase, impellingthe frost heave especially; In the areas ofTeam 17 and Beishueiduan, because of the effectof canals and irrigation, the frost heavingratio of aoil increased by 14.24% and 13,97%respectively in comparison with that of the soilunder natural conditions. According to investigationon Changma Pluvial fan, the surface layercontaining a large fine content and a littlewater didn't have obvious frost heaving' undernatural conditions; however, where affected bythe leakage of canals, the frost heaving ratiocan be as high as 20X or more.DIVISION AND CLASSIFICATION OF FROST HEAVING OFGROUNDTo reflect the engineering geological conditionsof seasonally frozen ground overall inthis region for the general planning of civilengineering and canals to be.constructed, andfor the prevention and solution of the engineeringfrom frost.damage, it is necessary toclassify the seasonally frozen ground inaccordance with the frost heaving behariour ofsoil.Division PrincipleThe division was based on the comprehensiveanalysis of the natural factors affecting thedevelopment bf seasonally frozen ground and ofthe natural processes, mainly to reflect thefrost heaving of seasonally frozen ground undernatural conditions. Considering the naturalconditions and researched degree, the divisionmay be divided int.0 two orders.(1) First order: Seasonally frozen groundarea. According to geomorphic units, this regionmay be divided into: 1) Mt. Yenmabeishan seasonallyfrozen ground area, 2) Mt. Hanxishan seasonallyfrozen ground area, 3) Changma PluvialFan seasonally frozen ground area, and 4) theFine-graded Soil Plain seasonally frozen groundarea.(2) Second order: Frost heaving section,Based on difference of rock (or soil) preperties,and hydro-geological conditions and mainly the'buried depth of ground water, the seasonallyfrozen ground areas were subdivided into frostheaving sections in accordance with classificationof ground foundation by frost heavingsusceptibility. Due to the randomness of humanactivity, the frost heaving susceptibility ofthe ground near arterial canals was markedqualitatively.The division of frost heaving of ground isshown in Fig.2.Frost Heave ClassificationThe classification of frost heave in seasonallyfrozen ground is an important basic problemin Geocryology. So far, due to the complexityof the question, there is no overall reliablemethod to evaluate comprehensively the frostheaving susceptibility of soil foundation.Considering the aim and task of this project,the authors put forward the frost heavingclassification for natural frozen ground inthis region.The principal guiding thought to classifythe frost heaving susceptibility of frozenground in this region is that the suggestedclassification should be able to reflect theactual situation of seasonally frozen groundunder the natural conditions of this regionand to harmonize it with the existing related1315

ACKNOWLEDGEMENTSThe authors would like to express theirgratitude to Mr. Yin Rongfa, the manager of theGansu Provicial general Company on AgriculturalExploitation, and his colleagues for their helpin field ,works and to Prof. Cheng Guodong, ZhouYouwu and Chen Xiabai for their valuablesuggestions with this paper.REFERENCECudlevcev, V.A. et al. (1978) General Geocryology,End Press, Moscow University Press, p.185-230.Fig. 2 Division of seasonally frozen groundsurroundinn ~~~ ~ ~Yumenzhen1. Mt. Yenmabeishan seasonally frozen groundarea;2. Changma Pluvial Fan seasonally frozen groundarea:3. Boundary of seasonally frozen ground area:4. Unfrost-heaving section;5. Frost-heaving section;6. Boundary of rock and soil:7. Gravelly soil;8. Clayey soil;9. Observation site:10, River;11. Mt.Hanxiashan seasonally frozen ground area;12. Fine-gradesoil Plain seasonally frozenground area;13. Boundarq of frost heaving section:14. Weak-frost heaving section;15. Strong-frost heaving section;16. Bedrock; 17. Sandy soil; 18. Arterial canal.classification. Based on the fact that the burieddepth of ground water and the soi 1 character arethe major factors determining the behavior ofsoils iifrost heaving, the class ification issuggested as shown in Table 3.Table 3. Classification of foundation soils byfrost heaving susceptibilitySoilBuried depthof ground Frost heaving Classwater z(m) ratio (x)BedrockGravel Not considedSandTJnfrost-heaving222.5 rlCl Unfrost-heavingClayey 2.55222 1Cr753 Weak-frost heavingsoil 2>2r1.5 3

CULVERT ENGTNEERING IN THE PERMAFROST REGIONON QINGHAI-YTZANC PLATEAUZhang Jinzhao and Yao CuiqinThe First iurvey and Design Institute of Highways,'The Minj.stry of Communications, Xian, China'This paper mainly introduces the engineering environment and characteristics ofculvert engineering in the permafrost region of Qingha-Xizang Plateau. Based onthe investigation of the effects and state of culverts and the long term observationof representative culverts in the permafrost region of Qinghai-Xizang highway,and by analyzing the damage of culvert engineering and renovating theQinghai-Xiz,ang highway design by, taking into consideration previous examples ofconstruction, the principles of damage protection and engineering measures ofculverts are given with regard to the different aspects of design, constructionand maintenace. This payer can he used as reference in the areas of design,construction and maint.enance of highways,THE SNVIRONMFNT AND CHARACTERISTlCS 0F CULVERTENGTNEEBINGA part of Qinghai-Xizang highway passesthrough a continuous permafrost region of 520 km,the hcight above sea level is more than 4500 m.Heat thaw lakes and ponds are spread throughoutthe region, ground ice is very developed, thesuperpcrrnafrost water is rich, thaw settlement,frost heave, frostchurning and other disadvan-Lageous gvologlc phenomena in engineering commonlyexist.It is high and cold and the oxygen contcnt~inair is only 50X of that in the interior. Theannual mean air temperature is low (about -3.0to -6.OoC.), the freezing period is more than 7-8n1011ths of the year. Even in warm season, minusair temperatures often occur in evening. Theamoutri. uf annual precipitation is 250-400 mm, itis concentrated between June and September andoccurs in the solid forms of snow and hail.Hundreds of culverts have been created in thepermafrost region with a total length of 520 kmalong Qinghai-Xizang highway. The operatingstates of the culverts were carefully investigatedin October, 1990. The results showed thatserious damage occurred in 15% of all culvert.3,middle level damage occurred in 21.17 and littleor no damage occurred in 63.5% of the culverts.The easily and very seriously damaged sectionsare the inlet buildings and culvert bottomlinnings. These two types of damage make up morethan 86X of all damaged culverts.CULVERT ENGINEERING UAMAGF, TYPEThe main types of culvert damage in the ycrmafrostregion of Qinghai-Xizang highway are(1) cracking and scttlement of the culvertterrace, (2) cracking, breakage and settlementof the bott'om lining of thc culvert, (3) seepagethrough ,.he bottom lining of the culvert, (4)cratrking, slupc and settlement of the inlet andoutlet wal Is and flank walls of the culvert,(5) damage of the lining on the inlet and outletand the wa 11 to cut off water. . (6) the collapseof the inlet and outlet. of the culvert. The mainreasons causing the damages arc the changes oftho permafrost table and the repeated freezingthawingeffects of the seasonal active layer.THE PRINCIPLES TO PREVENT AND TKEAT DAMAGE INCULVERT ENGINGERINGIn addition to the fac~ors of general displacement,effective stress is closely rclatedto the temperature acting on the permafrost andthe culvert base stability affected by waterlatent heat must be considercd for base designsin permafrost regions. Simultaneously, the correspondingengineering countermeasures to keepthe demanded thermal regime during culvertconstruction and operation must be taken. Thisis a basic principLe for preventing frost damagein culverts in permafrost regions.Geothermal CharacteristicsThe xeothermal reaimes of the base of t-hetypical-culvert along Qinghai-Xizang highwaywere investigated, as shown in Figures 1 and 2.The figures show that at 0.5 m depth the maximurpground temperature of the culvert mouth ishigher by 4-5°C than that of the culvert centre.But. the difference of the temperatures is onlyabout 1'C at 2.0 m depth between the body andmouth. Figures 3 and 4 also show the same reyulationthat the permafrost table beneath theculvert, mouth is 1.0 m deeper than beneath theculvert centre. This indicates that (1) the airtemperature effect on the culvert inlet andoutlet is greater than on the culvert body, (2)the air temperature effect on the upper layer isgreater than that on the deep layer, (3) in theculvert design,' the base depth at Lhe inlet andoutleL of the culvert should be 0.5-1.0 m morethan at t.hr culvert. centre, otherwise other

T "C6- 0.5 m depth-6 --8 - 'I9 6 7 8 9 10 11 12 I 11989 I 1990Fig.1 Ground temperature curves below the mouth and centre of culvertground temperature curves below culvert mouth;-.-. -. - ground temperature curves below culvert centre.IT c"C)8-6-4-2-0, "-2 * /. "-4--6I I 15 6 7 8 9 10 11 12 11.-1987 I 1988Fig.2 Ground temperature curves below the mouth a'nd centre of culvertGround temperature curves below culvert mouth;-.-.-e- ground temyerat.ure curves below culvert centre.Fig+3 Curves of zero.tempcrature belowthe mouth and centre of culvertbelow culvert mouth;-._.- below cplvert centre.Fig.& Curves of zero temperature belowthe mouth and centre uf culvertbelow.culvert mouth;-.-a- below culvert centre.1318

engineering ,countermeasures should be taken, (4)after understanding the base geothermal regimesduring the operation of the culvert, reasonableculvert engineering design can be determined.Culvert DesignDesign principleBased on the structural types and constructioncharacteristics of culvert engineering and engineeringgeological and hydrological conditionsin the area, two kinds of design principles canbe taken. a) One Is the principal of maintainingthe frozen condition, i.e. the base of theculvert beneath a certain depth is kept in afrozen state during the operation of the culvert.b) Another is the principal of allowing thaw,i,e. the base of the culvert between certaindepths is allowed to thaw during the operation ofthe culvert, the principle is only adapted to theweak thaw settlement soil and non-thaw settlementsoil.The depth of the refilled soil is closelyrelated to the engineering cost and the effectiveprevention of frost damage. It is determined bythe means of the structural type, the degree ofallowed deformation, soil type and ice contentof base soil, The two following design principlesshould be mainly considered.(A) Frost heave deformation of base soilshould be controlled in the range of alloweddeformation of culvert engineering after beingreEilled.(B) During thawing the bearing capacity ofrefilled soil and active layer of permafrostmust satisfy the design requests, the thaw settlementamount must also satisfy the engineeringrequests.In order to reduce the vertical and horizontalfrost heave force, sand-gravel qoil must befilled in the base bottom and around the base.Simultaneously, asphalt 1 cm in thickness shouldbe filled into the positions where base touchesthe embankment. The expansion joint should bealso filled with rubber asphalt that has a largedeformation at low temperatures.Because a large settlement amount in culvertengineering is’allowed, it is effective to set a-settlement joint to prevent and treat culvertengineering cracks caused by inhomogeneous settlement..Afterbuilding the culvert the naturalpermafrost’table beneath the culvert centre willagain rise beneath the inlet and outlet of theculvert and will again descend, as shown in Table1. In order to get a reasonable design forculverts in technique and cost, it is better tobuild the culvert base in sections, the sectionsbase depths are determined by different permafrosttables.Table 1. A comparison between man-made tableand natural permafrost tableMan-made tableNaturalNo. Pile No. (m)permafrostLeft Centre Right table (m)1 K157c373 1.68 0.65 0.83 1.75-2.10 .2 K279t370 2.45 1.50 2.15 2.05-2.103 K287+166 2.23 1.41 1,73 1.604 K337t170 1.75 1.35 2.05 1.00-2.45Choosing the structural tvueIn a permafrost area, the culver! structuraltype is the technical key to assuring the engi-.neering quality and a lower engineering coat.Rased on design priciple I, rapid constructionis the main condition to.select the culvertstructure. And for design priciple 11, thestrength and the ability of adapted deformationcan be taken as the main conditions to selectthe culvert structure. From the investigation ofthe state of the culverts and the long termobservation of typical culverts along Qinghai-Xizang highway, the culverts with a cover slabmade with reinforced concrete, the frameworkstructure culverts with a cover slab made with.reinforced concrete, the culvert box of reinforcedconcrete and the culvert tube made withtin-plated alloy steel are good structural typesin permafrost areas. For the culvert with acover slab made of reinforced concrete. based onthe different base conditions, different structuraltypes can be chosen, for example, theconcrete base of a strip type, concrete base insections, etc.Buried depth of baseThe following value can be used to determinedthe depth.A) For design priciple I, The base depthshould be 2/3-3/& times the depth of the naturalpermafrost table. The base depth at the transitionsection from the body to the inlet and outletmust be deeper by 0.4 m than the depth ofculvert body. The base depth at the inlet andoutlet must be 0.25 m beneath the natural permafrosttable.B) For design priciple If. The base depth atthe culvert body can be determined by the calculationof thaw settlement. The base depths ofthe body and the inlet and outlet must begreater than 1.00 m and 1.25 m respectively.The above methods are only adopted to thebase of the culvert with a corner slab made ofreinforced concrete.Culvert ConstructionCulvert construction in permafrost areas havemany special demands besides the general demandof culvert construction,Construction seasonThe cold season8 after Oct. and before Mayare the better construction seasons for theculverts designed with the principle of permafrostprotection.For the culvert design based on the principalof allowing thaw, the warm season frob June toSept. is the better construction season; otherwise,special designs and countermeasures mustbe taken.Construction methodBecause the base depth is shallower and thesection of construction ,is scattered, the openexcavation method should be taken in culvertconstruction. But the method has the technicalproblems that the labour conditions are difficult,the efficiency is low and the water-heatstate of the permafrost is disturbed greatly.For resolving the disadvantageous problems. themethod of demolition and rapid excavation isadapted to the permafrost along Qinghai-Xizanghighway.Notes on construction1319

(A) In warm season construction, the hase pitcan not be allowed to be exposed for lung periods.The exposure time was not longer than 15days and the construction time for whole culvertproject was not longer than 50 days.(E) After excavated, if the whole or part ofthe base is built on the ice and soil-contentice layer, the design must he changed and thebase depth must be adjusted.(C) Water, stlow and high-water-content soilin the base pits must be cleared up due to itslarge latent heat.(D) The artificial dist.urbance of permafrostmust be reduced as much as possible.(e) Before excavation, all material used inthe base construction must. be prcpareri.Cul.vert MaintenanccPreventinx damaRes in culverts . .(A) The connected section between the culvertand the pavement must be checked frequent-ly. Ifdamages ore discovcred, relevant maintenance andrepair countermt:asures must be taken,(13) A culvert platform base bears the load ofthe culvert and vehicles. The deformation, displacement,cracking and set-tlqement of the baseshould nut. he allowed but when occurring shouldbe treated quickly.(C) The effect of water on the permacrostbase and culvert is great, the unb.lncked culvert.must he maintained and cleared when bl.ockageoccurs.(0) Cr,acking and expansion joint.3 must bechecked to prevent water permeation.CONC1,IISTONS'Two kinds of desipn principles of culvert.cngineering can bc laken. One is the principleoC maintaining the rrozen contlit.ion. Another isthe principle of allowing r.haw. The t.wo pricipleso r ~ l y adapt tu the weak and nun rhaw scttlernenrsoil. Otherwise other countcrme8sure must hetakcn.In !.he psrmafrvs~ area, the selectcd struc-I.ura1 typc oC culvert is very important for thequallty and cost qf erlgineering. The culvcrtwith il cover slab made with rcjnforced concrete,t.he framework structurt. culvert with a coverslab made with reinforced concre!.e, the culverthox made with reinfor-r:ed concrete and theculvert tuhe made with tin-plated alloy steelare gout1 structural types in permafrost areas.The ohservation and analysis for typicalsectiorls ,along Qinghai-Xizang highway show that.the air temperature cffect on thc shallow layeris greater than the dccp layer. In the culvert.design, the base huried depths at the inlct andoutlet of the culvert should be more than 0.5-1.0 m than thaL of the culvert ccntre.The cold seasons after Oct. and before Moyare the better construction SF.RSOIIS for theprinciple of malntaining the frozen condition.For the ptinciplc of allowing thaw, thr? warmseason from .June to Sept. is the betterconstruct-ion ScRson.

..NUMERICAL ANALYiIS OF TEMPERATURE AND STRESS ON THE CANALSUBSOIL DURING FREEZINGZhang Zhao'*'and Wu Ziwang''State Key Laboratory of Frozen Soil Engineering, LIGG,AS,China'The First Survey Design Institute, Lanzhou, Ministry of Railway of ChinaThis paper presents the numkrical simulation analysis of two-demcnsional nonlinear mathematicalmodel for the heat transfer in the siturated canal subsoil using the finite differences mothed. Elementsdivided based on the different isothermers at different moment, are used to calculate the stress of thefreczing canal subsoil by thc finite elements method of plane strain problem. The following main conclusionare obtained: The maximum values of stresses and deformations on the surface of the canal linningoccur at bottom and down slop; The stresses decrease with the increase of depth under the canal bottomand down slop; And the stresses value cyclically decrease and increase with the increase of depth on twosides slop of the canal because of the binding force of boundary. Finally, if the surface deformation ofcanal lining is restricted to zero, the restrictive stresses of lining must reach up to 0.5~ 105N/.m2. Thecalculated result correspond relatively well with the measured results of similar condition in site.rNTRODUCTION"-When the subsoil water freez,es due to low air temperature inwinter, thc subsoil volume will expand. The increment of volumeexpansion will be much more considerable expecially for saturatedsubsoil in open system. While subsoil freezes and thus induces volumeexpansion, the total expansion of subsoil will be confincd partiallyby the presence of canal lining, When the increment ofvalume expansion is more than the allowable deformation of canallining, the damage of canal lining will happen. Therefore, thcchanges of tempcrature and moisture field are the basic reason ofemcrging frost heave and frost heavc force. In this paper, the distributionof two-dimensional unstable temperature field on the canalsubsoil was obtained by the numerical analysis of the finite diffcrencemothed. Based on the analysis results, the analysis methodof linite elements about plane strain problem werc carried out tocaculate the frost heave force. The closc relationship betwccn elasticmodulus of f-rozen soil and minus temperaturc was taker intoconsidcration.Tf1E " STRESS DISTRIRUTION OF CANAL SUBSOIL. "- ~."Thc Determination of Temperature Distribution"" ~" ." .Thc mathematic model to determin the temperature distributionwas based on the bad engineering geologic condition; i.e. thegroundwater table is about 1.73-2.0m; The canal water is suppliedto groundwater when canal operating in summer; The subsoil frezzingbegins when the canal operation stop in Nov. So thesimulntion calculation of saturated subsoil in open system wereconsidered. Assuming the canal to be unlimited long, homogeneoussubsoil, no transport of water and heat by evaporation and no otherpotential energy, taking a vertical section, therefore, this processan bc described as two-dimensional parabolicsl partial nonlineardifferential cquation. For Ihe definite problem, finite dilTerenccs bycnculation were made, the differential equation were differenced byCrank-Nicholson method. Finally thc problem were solvcd bymeans ol-TRM-PC with self-compiled program..Mcchtln$sModel- . - .".and -. Calculation . Method. . . . -In the certain condition of temperature and moisture, whilefrost heave is partially by the canal lining, heave pressure will acton canal lining. When the frost heave pressure is greater than theallowable deformation strength of canal lining, the damagedeformation of canal lining will occur. The finite element methodanalysis to calculate the stress distribution in freezing subsoil isshown below. .ITaking a vertical section as a XOY coordinate plane, canallong direction as 2 coordinate axis, it is obvious that the stress,strain and deformation on the section are only the function of xand y, This.is a plane strain problem. Analysis will be performed, under the following conditions.I). The boundary force on the upper boundary of canal is thewcight of canal lining and the freezing force between canal liningand subsoil in the vertical direction of canal lining. The lowerboundary is back-up roll.2). The subsoil is homogenous and isotropic frozen andunfrozen soil.3). Frost heave occurs in the direction of heat flow by 100%.4). The defromation in the direction of x on the right and leftboundary are zero.Elements.""As zhu Qiang (1988) pointed out, when the freezing depthreaches the one-thirds to two-thirds of maximum freezing depth,frost heavc force rapidly increascs, in the last period, it will be leap

Iincreasing until the frozen depth reaches the maximum value. i.e.The appearance of maximum frost heave force is correspondingwith the stable stage of maximum freezing depth. Therefore, basedon the configuration of isotherm line at 1760 hour obtained by finitedifferences analysis, the caculating range can be decided triangleelement as showed in Pig.].Elastic Modulus and Poisson's RatioThe experimental research (Wu Ziwang, 1983) indicated thatthe relationship behwpn elastic modulus of frozen soil and minustemerature can be described as following empirical formula,E=a+blrl" (1)where, m is a constant with less than I, generally taking 0.6. a and bare experimental constants related with soil type , for loess gmerallytaking 100 and 500.In cah~lation, the temperature of subsoil are from -10°C attop to O°C at lower position. Themfore, taking the average of triangleelement into formula (l), the different elastic modulus of triangleelement of frozen' soil can be obtained. Poisson's ration offrozen and unfrozen soil is 0.3 and 0.4 respectively.Nodal External LoadIn the calculation of finite elements method, external forcewere transmitted at element nodes. Therefore, all kinds of loadmust be considered at element node. In this analysis, there are twokinds of load, One is concentrated load. Another is distributedload.The equivalent node load of distributed boundary force , asshown Fig.1, induced by dead load of lining and freezing force actedon the boundary, 1-2,2-3, ..., 18-19, of the upper boundary clrments,2,4, ..., 18, 20,..23. based on the experimental result of therelationship between leap anti-stretch strength of lining material incontact with frozen soil and temperature (Wang Jianjuen, 1986)',\ and long period strength is about one-thirds leap strength (WuZiwang, 1981), the freezing force will be taking 3 .4~ 104N/ m2 inthis calculation. Taking 0.07 m thick of lining and 2.0 X 104N/ m3density of lining, the dedd load of lining will be carried out. Therefore,using virtual displacement priciple, the equivalent node loadwill be:I(Pl)"(Xf up x; Y; x; Y',)=-p(O PI 10 1 OO)T (2)I ( l y e =(x; r:x; Y;x: Y:)P= 2 Ir(sin(x)cox(x)sin(x)c(ls(x)Oo)'(3)2(PI' =(PI)' + (P*f (4)where, PI is dead load of lining, taking 0.02Kg / cm'; P, is freesin& forcc, taking 0.34 Kg / cm'; 1 is the thick of element, taking oneunit.The node load of distributed volume force, for eachhomogenous element, can be written as:Qy= q,t A (5)where A is area of element, q is specific gravity of element.Based on dead from equivalent priciple and virtual work principle,the node load of distributed volume force can be expressed asload vector type. i.e.The node load caused by initial strain cxpreeses the loadcaused by frost heave ratio in the heat flow direction. The researchshows (Zhu Qiang, et. al. 1988) that, in the condition of highground water table, the frost heave will be occured in the wholefreezing layer and frost heave ratib increases with the increasing ofdepth. The expcrimental research (Chen Xiaobai, 1983) also showsthis requlation, and in the load pressure condition, the frost heavewill be confined especially in low freezing ratio condition.The freezing ratio on the typical position of canal section obtaincd from temperature field analysis is shown in Tab.]. Based onthe background condition and distributed force, the frost heave ratiofor eaEh freezing layer can be obtained as shown in Fig.2.Therefore, taking this as initial strain, the node load can be determined.Assumpting initial strain of anyone element is (e,), the equationfor calculating stress is become,Taking equation (7) into following equation:(F)' = tA(B)T(u) (8)The node force of element will be written as :Based on the relationship between the node force and the node dia-placement,Fig. I The diagram of elements decided* Wang Jianjun, (1986). The finite element calculation for thecanal lining structure in wasonal frbst.1322

w a )L 0 20lo OFig.2 The frost heave ratio of canal sectionTable 1. The freezing ratio on the typical positionlayers bottomI 2.9832 1.1403 . 1.4624 1.1905 0.987(i 0.5837 0.335slop embankmen3.255 3.8272.418 3.6962.155 2.8491.795 2.4001.508 2.0300.843 1.1020.558 0;782where,Finally, based on the relationship between displacement and strain,and the equation (15), the node laad caused by frost heave ratiocan be obtained.Therefore, according to accumulation principle of force, thenode load can be expressed as following equation.and chahging the second item of equation (9) into minus one,which is the node fQrCC caused by initial strain, the node loadcaused by initial strain can be written as:(WL (BT)(D)(8,)tA (1 1)The node equlibrium equation and general rigid matrix, the nodeforces and the node loads of each element must be in equilibriumstate. Based on the equilibrium~ondition of each node in directionof x and y, the equilibrium equation for each node can be set up.where,z: summing for the all elements in circle of node, Using theImatrix pattern, it can be expressed as:Taking equation (11) into above equation, the node equilibriumequation expressed by node displacement can bc obtained.when, w] is general rigid matrix, 6 is tow vector of displacement,F is row vector of load.'- 1323

For the n-th node, the equilibrium equation (21) will be thelinear algebraic couplcd equation of 2n-th rank. Resolving a~(1p'N/m2) this . uy(IO'N/mZ) ~ x ( I O ~ , m 2 ~coupled equation, the node displacement can be obtained. There-Oa3 0.5 0.7 0.9 0.3 0.5 0.7. 0,9 0.2 0.4 Q,6 Q.8fclrc, the element stress can also bc obtained by equation (7).TheResull~_ C@culating9Through different kinds of frost heave ratio and confined con- 160160clition, the stress distribution along the direction of horizontal and(a)(b)dcpth, and the displacement distribution on the canal surface havebeen carried out by means oi' computer with self-compiledprogram. Fig.S The frost heave vs. frozen forceDISCUSSION AND . ANALYSTS.. . . . . When freezing pcriod is 1760 hours, thc Fig.3 shows the calculatedresults. Its regulation is similar with the obscrvatcd data (seeFig.4), that is the maximum deformation occurs at the bottom andslop down of canal, and the deformation of upper part is small andalmost same. But the calculated values of deformation are mor\than the obcrvatcd values because thc most disadvantage backgroundcondition wcrc considered and the resisting heat efi'cct oflinninp nlaterial wcrc not considered in this calculation. The maximumfrost heave l'orcc is also at the bottom and down slop ofcanal. It IS why the failure oftcn occurs in these parts. It is given byc:~lcul;ltion that thc maximum frost heave stresses arc 0.9xIO'S ' nl' :Ind 0.5 x 105N / m' in the horizontal and vertical dirsc11011rcqxctivcly.Fic.5 is thc cnlculnted value< nf frost heavc strcsscs alongdzpth :It the points of' A,R,C,D,E. Rcc;\usc of cl'l'cct of the surfhccrcsrl-;lininp ;Ind thc +ad wight of subsoil. thc lrost deformition issm;111 on thc upper of' subsoil. and the F[rcS

Zhu Qing,1988. The distribution of seasonally frozen soil along thedepth. Journal of Glaciology and Geocryology, lO(1).Wu Ziwang, 1981. The expcrimental study of frozen strength bctweenthe foundation and frozen soil. Anthorogy ofGlaciology and Geocryology, No.2. Sciense Press, Beijing,~129-139.Wu Ziwang, 1983. The characteristics of the strength and failure offrozen soil. The proceeding of Second National <strong>Conference</strong> ofPermafrost. Gansu People’s Press. Lanzhu. p275-280.

THE RELATIONSHIP BETWEEN THE RAILWAY PROJECT CONSTRUCTIONAND ENVIRONMENT PROTECTION IN PERMAFROST AREAZhengQipuThe 3rd Design Jnstitute of Survcying and InvestigationThe Ministry df RailwayIt has been practically provcn that various hazards will occur if the protection to the surrounding cnvlronnlentis ignored during the design, construction and operation of railways. Such as the uncven settlementof roadbed, the frost heaving, cleavage and collapsc of house walls, frozen fold of thewater-supply pipes, cleavage of bridge pier, cracking of bridge surfaces, cleavage of the side walls in theinlctr and outlets tuunels.THE SIGNIFICANCE- OF THE ENVIRONMENTAL.PRO-. .-."ECYION IN PERMAFROST AREASThe naxinanlin pcrnmafrost area in the Northeast of' China, isa high latitude zone of permal'rost, the climate is severely cold. Thesoil is in froten and thawcd states with the change of seasonal ternperaturcs.The cold resistant and water-favouring plants areaboundant. The landscape in the permafrost area and the main basicconditions formcd the ccological features.In the permafrost arca. the strength increases suddenly whenthe watcr i n thc soil changes into ice. The bearingcapacity is almostequnl to zero when the Ice melts into watcr. ThisJeature of water inthe l'ro7cn soil decide whether the frozen soil has a very high bearingcapacity 21s well as deciding if thc capacity decreases sharplywhen the icc mclts. The frost heaving and melting settlement of thissort of' soil-sand rcally forms the basic difference of the mechanicalproperties of frozen soil from other soils due to the water in soillicczing and melting. If there is no freezing and melting of the waterin soil, there is no frost hazard of buildings. However, the enginecringfoundation ol'various buildings in permafrost areas arc hurledin the pcrcnnlal frown soil layer with diltcrcnt foundations. Toprotcc~ thc stahIiLy and prevcnt dclormatmn of the buildings, theq;lbility o f thz foundation hearing stratum has to be considcrcdtil-st. In [hilt wily, it is necessary 10 protect the stability of the environmentsurrounding the perennially froten soil..How to preventenvironnlental dcstuction from thc cconornical activities, not thech,lnges Jur: 10 thc natural conditions and regional climate factors.of hrgc pcriods. will bc discussed.In pt-actlcc It was provcd th:lt the various hazards would occur11. tllc protzction ot'thc surrclttndlny cnvironment during the dcsipn,L'onstructlon and uperallon was ignored. Such as the uncvcn sctrlcn:mt01 ro:ldhcds , the frost hcnvlng, cleavage and collapse ofhouse u.:IIIs. 1'1-07rn I'old 01 water-supply pipes, cleavage 01 brdgcpiers, cracking of bridge surfaces, cleavage of side walls in inlet andoutlet tunnels.In the design of various buildings of railway projects inpcrmafrost areas, the design principle of foundations 0f'"protectingfrozen soil" is often adopted. So we often stress the stability of thetenlpcraturc licld between the contact surfacc of thc bcaring stratumand various foundations and frozen soil to preveqt the frozensoil from melting within the bearing stratum due to the temperaturechanges, which leads to the deformation and destruction of buildihgs.This opcnion is one-sided as provcd in practice. Many buildingdamage are not due to the bad contact position between thebcaring stratum and foundation boundary, but due to the incorrectsurface position of the buildings location and the lack of protectionof the surrounding environment, which leads to environmentalchanges and building damages.For example, deforestation of theforest and vegelationm surrounding the buildings, grazing and cultivation,over cxcavating and borrowing soil without a plan, nodewatering ditch arragement, unreasonable cultivation and irrigation.A11 the above factors will lead to the destruction of theecological equilibrium. The destruction of the heat equilibrium ofpermafrost due to the destruction of the ecological equilibrium, willcause the frozen soil to melt, and building damage, etc.From thc above examples, it is pointed out that during constructionin permafrost areas, not only the environment protectionin surveying, dcsigning and constructing, needs to be payed attentionto, but more importantly is to enhance the measurements toprotect the environment during construction and operation. Somebuildings sufl'ered damage in the lirst tcn years of operation. Thereason was ignoring the environmental protection duringopcration. So it is very important to protect the cnvironmeni surroundingsduring construction in permafrost areas.1326

THE RELATIONSHIP BETWEEN ENVIRONMENTAL PRO-TECTION "AND THE RAILWAYPROJECT CONSTRUC~. .. " ~ ~TIONThe railway pro,mt construction in pcrn~afrost area mainly inDa and Xiao Hinggan Ling Mts. includes the line, roadbcd, apcrtureof bridge, turlnuel, house buildings, water supply and waterdrainage, communication and electric power, etc. For forty or moreyears, sincc the 195O's, the 2000 km railway has operated in thepermafrost areas of Da and Xiao Hinggan Ling Mts., in thenorthcast which has led to various building damages, the most scriousis caused by breaking driving, the lesser is slow driving by ratclimiting. Rased on the incomplete statistics 01' the 'nine lines in thepermafrost area by the Railway Ministry of Harbin, there are 124hazards only for line roadbeds. So, numbers of people are need torenovate and keep the line open. The cost of renovating, curing andmaintenance ate several million yuan each year, 6460 working daysand 5010 working days are respectivcly used to do rushreopennings, there were only two working day segments of Taheand Genghe in 1985. From the investigation information for the 33stations and 230 buildings only 85 buildings are completely stable,36.9% in total, 110 buildings suffered hazards to different degrees,47.8% in total, 35 buildings can not to be uscd or need rebuilding,15.3% in total. I'Thus it is obvious that during railway operation. hazards tothe railway are vcry serious duc to :I lack of cnvironmentprotection. Not only is there an extremc waste in labour power andmaterial resources, but the safety during operation is also aflicted.Now the relationship between the railway and cnvironmental pro.tection is respectively recounted with examples as follow^I) Soil excavation surrounding the buildinEs C:IVCF the th:Iwinghnsin to lntcrally and vcrtic:tlly, the frozen toll foundation melts.settles and causcs damage the buildlnps.The old train-chcck house :It the 4lonpul Sr;\tlon. Yahn line,is located on the 2nd platf'orm 01' nc11-11 trtcr, thc lound;ltlon type ishurlcd plle with carlh lill, and the huarlng stlaturn IS w,tthln thepermafrost Ia'ycr. Recausc FOII ercavatlon ilnd filiinp IS nwr thetrain-check house, forming a man-made pond, hold~ng wiltcr andice-forming for many ycars. Two ycars later after its opcratlon, thcthawing b;rm hclow the bottom ol-the pond I$ lormed i~nd euplndsycar by year, al?d causcs the Iroxn so11 In thc hcarlne

!i1llt1W\Til; llr~ll;~'lon ,)I I.,,cI ,,:!I~;%: h \ ~ r ~ ~ nS~~~~r,l,~~1~t~~~l~n ~ c ~ ~ t 'The k 1 ) ! , !In$\Mll!hlll 1 ' ~ ~ C I U L ,lrL':l ~ ~ 01 O IIL, ~ ~ : 1 ~ ~ , Xl,ltl 1 I I I ~ ~ C! 1V 11~' ~ ~ ~ ~\I!< 1!1 1hs \(>::I:\.,.;, 1hC I11,lIIl ~h~l~;l~tctl~~lL~.111 l!l, 1, !l(>l!l

REFERENCES."" ....Zheng Qipu, 1986, The Harms on the Railway Project of GlacialHill and Its Prevention, The Papers of The 5th <strong>International</strong>Enginecring Geology <strong>Conference</strong> (T.A.E.G.) pl S57-1567. RuenouAircs, Argentina.

REGIJIARTTY OF FROST HEAVE OF THE SEASONALLY FROZENSOIL IN HETAO IRRIGATION AREA, JNNER MONGO121AZhou DcyuanAdministration Rureau of Hetao Irrigation AreaYongji, Inner MongoliaIn Hctao irrigation arca, in comparison with thrcc other northcrn regions in China, the freezing index isnnt very hlgh, but thc lcvcl of undcrground watcr is higher, the frcczing speed is lower, the amount ofwater migration is largc and the action of frost hcavc is vcry strong, comprellensive action of moistureilnd heat results from a strong frost hcavc. With thc theorics of' gray system and similitude and theirrncthods, a random-dctermination model is set up. and thc rcgularity of frost heave ol'seasonally frozenFoil is revcnlcd more nccuratcly.SUIMMA RY OF THE NATURAL CONDITIONS.. ""... ... . ., .-.." .".... . .Y0ng.p w k lield for frazcn soil is located at the center of Hetaoirrigalion ere:,. 2.5 km northeast from Linhe suburb. The geographicalposltion is 40°43'N, 104"44' E, and 1039.55 meters ahovcsca level. The hld belongs to seasonally frozen soil arca.The soil in the irrigation area is sand loam. Size composition;In:llysis and physio-chemistry properties are shown in Table I.In the+ test field, the fbllowing observations were arranged:fro72n depth, rrost amount, amnunf of water migration, frost heaveforce (tangcntial, normal, horizonttrl), meteorology in a small region,ground tcinpcrature and dcvclopment of underground water,CtC.The observation method and the arranged forms were all doneusing gcncral mcthods. An open type steel hoop standarddynarnomcrer was used to rncasure force, thermalcouples were appliedto mcnsure temperatures. The secondary meter is a UJ-33aVolt meter. Other methods werc general,PROCESSES OF FREEZING AND FROST HEAVE OF. .. .SE,W)N..ZI.I,Y FRO~EN SOILThrough observation and test in the Yongji test ficld forl'ro7cn soil from 1987 to 1990, thc following processes werercvcaled, they arc frozcn depth. watcr migration, amount of' frosthcil\*c. ilnJ I'rost hcnvc forcc. In gcncrnl, thcy can be dividcd as follolvingstcps:I. Unstcndy frcczing, from carly Novcmbcr to thc middle ofNovember. The ground temperature fluctuated around O'C, andtllcrc wns nn nltcrnate changc of frcczing in the night and thawingduring thc day. When the temperature was continuously belowzcro, steadv frcczing formed.2. Slow frcczing, from late November to thc late December.The atomosphere rcnlperature decreases, the increase of the freczinpindcv \vas not rhst, the rreezing advance was slow, frozen depthsreached to 45 percent ofthe maximum, and water migration, frosthcnvc amount and frost heave force produced gradually.3. Quick freezing, Irom thc early January to early February.Tncrc was n continous decrease in the atmosphere temperature, theTable 1. Size compositkn and physiochemistry propertiessand loamof\particle annlysis 'h dry Spccilic Void liquid plaslic plastic Saltdcsity gravity ratio limit limit indcx contcnt..,>\,\.r0.1 0.1-0.05 0.054.005 1330

~ --increase of the freezing index was fast with the fast freezing peneuation which reached to 98 percent of the maximum value. At thesame time, the amount of yater migration, frost heave and frostheave force also developed quickly, but the frost heave force declined in the late period due to relaxation.4. Steady freezing, from middle February to late February.The freezing index increased slowly and the freezing entered into a.steady freezing state with the frozen depth slowly reaching its maximum.The amount of water migration and frost heave remainedcomparitively steady and the frost heave force continued to relaxand decrease.5. Thawing, form the early part of March to late April. Theatmosphere temperature increases to zwo, the frozen layer beginsto thaw quickly. Water migration comes to an end. The amount offrost heave and frost heave force decrease gradually until they vanished,CHARACTERISTICS OF FREEZING AND FROST HEAVEI . .- .OF SEASONALyYFR%ZEN ._SOILchxcteristics of FreezingThe freezing index was not very high. The heat flow was greaterin the ground. The depth of the annual fluctuation layer 13.5 wasrnetcr (I-uo Xuepo, 1983), and the average temperature was 10.8T.The gradient of temperature in the frozen layer was smaller. Theaverage temerature gradient of the frozen layer was 0.096"C / cm insand loam. Freezing speed was slower. The average freezing speedwas 0.96 cm / d in the freezing period of sand loam. Frost depth\vas not very deep. Its averge value was 103.3 cm in sand loam.Characteristics - . . - of water migration~.Variance of the underground water table. .- ...When the irrigation area wag irrigated in autumn, the undergroundwater tablc rose to its maximum and the ground surfacefroze. Before freezing thc undcrground water table was high, thenin the frcczinp period, the ground water table was in the decreasingprocess, which developed at almost the same spped as the freezingfront. Thc length of the penetration was relatively steady from thebeginning to the end and was in the range of capillary action.Rcdistributjon .. of water confht.In the freezing process, the water in the soil was redistributed,showing that the accumulated area of water gradually moved downand the total tcnds to enlarge. The water content increased to,7.4percent in sand loam when it was compared with that before freezing.Characteristics of Frost He?ve"""." .The characteristics of frost heave in the laycr of sand loam. Itwas shown that the amount of frost heave took up 21.2 percent inthe upper part, 54.4 percent in the middle part, 23.4 percent in thclowcr part. The thickncss of each part was one-third of the thickncwof frozen Inycr. When the freezing spccd was smaller, the sup.ply of wntcr \vas sul'licient, and a strongcr frost heave was formcd.Thc average ratio was 8.7 percent.Charac,terist@ of Frost Heave ForceThe development of general tangential frost heave force wasparallcl with unit tangential frost heave force, and both maximumforces appear at the same time.Frost heave force increases with the increase of the frost heave.The maximum tangential frost heave force appeared when theamount of frost heave reached to 83 percent of its maximum horizontalfrost heave force reached to 98 percent of its maximum values,then they dccreascd one after another.The distribution of horizontal frost heave force was not thesame along with the hqight of the wall. The outline of maximumforce appeared approximately in a trapezium which used the side ofthe wall as the bottom. The force was zero at the top of the sample,and was similar in the height of 0.3 and 0.7 height of wall reachedto their maximum. The force at the base of wall was 40 percent ofthe maximum-.THE REGULARITY OF FROST HEAVE OF SEASONALLY~" . ,. . ""FROZEN SOILReeularitv of Frozen DeothA linear relationship between frozen depth and the square rootof the freezing index was found, as shown in Fig.1. The equation ofthe relationship is:H,= a+bF0.'where, a and b are coefficients relating to the soil, and they are-5.3310 and 3.7849, respectively for sand loam.The frozen depth HPad a power function with the depth ofunderground water table before freezing, as shown in Fig.2, the followingis the equation:H,= a * Zbwhere a and b are coefficients related to soil, 0.0276 and 0.213 forsand loam.The random-dctermination model of frozen depth"According to the theory of comprehensive transposition ofwater and heat (Ding Dewen, 1983) and the method of similitude(Gao Min,1983), through therical analysis and regression, the random-determinationmodel was cstablished as followings:H 3.29(-JW, + 0.262 "*whcre, H(-- frozen depth (cm), W,--water content before refeeting(YO), 2-avcrage value of underground water table in the freezingperiod (cm).The Regularity of- Watcr migration-. ." -~The amount ofwatcr migration.. ."It had a linear relation with the freezing index F, which isshown in Fig.3, and the equation is:Q = a+bfwhcrc a and b -- coefficients related to soil, -0.2531 and 0.01 11Ibr sand loam.Random-dctcrmination modcl of the amount of water migratio.. ... -. .. .".- .- - . .""Thc thcory and method to build the model are similar to ihe1331

where, Q-amount(cm), 2--averageing period.HQ = 2.4 x 10-6(L)zof water migration (%), Hrfrozen depthdepth of the underground water table in a freez-0 5 10 15 20 25 30 35\iF Frozen index ("C d)'-'Fig.1 Relationship bctween frozen depth and frcezing indexfor sand loamE-5 a-3C8eu.1101000~50 100 150 200 250 300Z,, Depth of underground water tablebefore freezing (cm)Fig2 Relationship betwccn frozen depih and the depth of un-c5 c.d EtLEderground water table before freezinglorThe Regularity of Frost HeaveThe amount of frost heave had a power function with thefreezing index F, as shown in Pig.4, and the equation is :h= a Fbwhere a and b are coefficients related to soil, 0.3347 and 0.850 forsand loam.The frost heave h had a power function with the depth Z of theunderground water table, as shown in Fig.5, and the equation is:h = ae-bzowhere a and b are coefficients related to soil, 0.1445 and 0.0020 forsand loam.Random-determination model of the amount of frost heaveThe theory and the method for establishing the model are similarto the above. The random-determination model of the amountof frost heave is:6H7/I = - 264.25 + 58.551n[ - 784 -t 324.4( W, + 2.4 X 10 - T)]Zwhere h--amount of frost (cm),. The other symbols are the sameas in the previous equations.The Regularity of Frost Heave Force". . . .- .~The outline of the measurcd point of tangential frost heaveforce had a relation with the amount of frost heave, as shown inFig.6, and the equation is:T = a+bhwhcre a and b are coefficients related soil, 0.4858 and 0.0145 forsand loam.The outline of the measured points of the normal frost heavcforce showed a linear relation with the amount of frost heave, BSshown in Fig.7 and the equation is:u = 2hbL0100~0: F, Freezing index ("C d)Fig.3 Relationship between water migration and freezing indexfor sand loam (1989-1990)erzLLi0 200 400 600 800 1000F, Freezing index ("C d)Fig.4 Relationship between frost heave and freezing indexfrozen depth model. Random-determinancyofwarcr miirntlon is as t'ollows:model of the amount

g 201 .\m$ 151 1991-1992Lloo}a=-50Z,, Depth of underground water tablebefore freezing (cm)0 20 40 M) goAmount of frost heave (mm)Fig.5 Relationship between water migration and the depth ofunderground water table before treezingFig.8 Relationship between horizontal frost force and frostheave (1987-1990)Is0 15 30 45 60 75h, Amount of frost heave (mm)'Fig.6 Relationship between tangetial frost force and frostheave8ecY)0t4002001 987- 1990*.*0 15 30 45 60h, Amount of frost heave (mm)Fig.7 Relationship between normal frost force and frost heavewhere a and b are coefficients related to soil, 0.6706 and 0.0546 forsand loam. The outline of the measured points of the horizontalfrost heave force had a power function with the frost heave amounth, as shown in Fig.8, and the equation is:u,= a hbwhere a and b are coefficients related to soil, 0.6706 and 0.0546 forsand loam.CONCLUSIONThe regularity of frost heave in the area was affcctcd and limitedby the conditions of the natural environment and irrigation activity in the Hetao irrigation area, Inner Mongolia. In the area, thefreezing index was not very high compared with the three othernorthern regions in China, but the underground water table wascomparitively higher. In the process of freezing, because of thecomprehensive action of water and hest transfer, the opcn type offrost heave was formed with the decrease of freezing speed, theamount of water migration was large, the frozen depth was notvery deep, but the frost heave was very strong and the frost heaveforce was also large. So these cause the intense destruction toditches and hydraulic engineering buildings.REFERENCE:-Ding Dewen, 1983, Calculation of Frost Depth and Moisture Conditionin An Open System.Proceedings of Second National<strong>Conference</strong> on Permafrost.Gao Min and Ding Dewen, 1983, Research on The Amount ofFree Frost-Heave Soils by Thermodynamic Method, Professional Papers on Permafrost Studies of Qinghai-Xizang Plateau.Luo Xuepo and Ding Dewen, 1983, Determination of Thermal REgime in Annual Fluctuation Layer of Groud Temperature,Proceedings of Second National <strong>Conference</strong> on Permafrost.

FOSSIL PERIGLAC,IAL LANDFORMS IN THE SHENNONGJIA MOUNTAINS, CHINAZhou ZhongminChangsha Normal Univers.ity of Mater Xe,sources & Electric PowerTWO fossil lines of block fields of the Late-Glacial time have been preserved atthe elevations of 2800 m and 2900 m in the region of the highest peak ofShennongjia Mounta*ns, and to the northeast at Laojunshan Mountain, The blockfields, nivation hollows and cryo'planation terraces differ both in distributionheight and in scope within the above mentioned lines, therefore different namesare designated for them, i.e. the early Shennongjia periglacial stage and alater one. This is probably related to the cold and dry prairie climate in EastChina that evolved southward to the riparian basin of the Yangtze River in thelate period of the Upper Pleistocene in the year of 18000 B.P. The fossil blockfield lines represent the geomorphic boundary of the palaeoclimate.DEFORMATION OF TYE FOSSIL PLANATION SURFACE ANDTHE PROCESS OF T!IE UPLIFTING STAGESThe area of the Shennongjia Mo,untains is 3250km', It is the effect of several processes, eg.land formation after the Indo-China movement andsubsequent regression, tectonic folding andfaultinfi ?+Iring the Yanshan movement, and upliftingof tht mountains in stages since the Himalayamovement. 'She long period of planatian actionresulted'in an undulating quasi-plain, that is,the present summit planation surface left overat the highest peak of 3000 m a.s.1. It isequivalent to the ground surface in the Beitaistage, which is also named the Huanghunlingstage.Below the above mentioned fossil geomorphicsurface is another broad and flat denudationbench, represented by the Dajiuhu erosioncorrosionquasi-plain of 1800 m. This is theground surface in the Western Hubei stage. Therelevant deposition is the red stratum coveringthat of the above Cretaceous period. It isequivalent to that of the Tangshian stage andhence belongs to the Miocene Epoch.Since the first act of the Himalaya movementat the end of the Miocene Epoch, the WesternHubei stage quasi-plain started uplifting tothe entirety of 500-600 m. Thereafter a denuda-tion surface of about 1200 m a.s.1. was formedat the foot of the mountain and is ,called theMountain Initial stage. This is equivalent tothat of the Fenhe stage and belongs to the ageof Pliocene (Yang L.K., 1991) to the earlyPliestocene.The second act of the tlimalaya movement wasfrom the end of the Pliocene to the beginning ofthe Quatern&y. The Shennongjia Mountains wereuplifted by faulting and differential processes,leading to the deformation of the planationsurface, The planation surface of the Huanghunlingstage was deformed into two benches of3000 m and 2600 m in elevation; those of theWestern Hubei stage were 1800 m and 1500 m; andthose of the Mountain Initial stage were 1200 mand 1000 m. The adjacent Yangtze River and thelarge tributaries originated from the Shennongjiadeep undercut, which was around 600 m. Thetime of the crust uplifting and deep rivercutting was named the Three-Gorge stage. Basedon the exploration of Huyiatan and Gaojiadianin Yichang city, and to Gulaobei and Luyianchongin Zhijiang county, we discovered that the LowerPleistocene stratum was mainly a set of fluvial-lacustrine depositions. This sug8esred that therivers sharp cutting action mainly occurredafter the Middle and Upper Pleistocene, whilethe terraces developed after the Upper Pleistocene.and only took place at the valley slopesbelow 150 m in altitude. The height of upliftingsince the Upper Pleistocene, therefore. accountsfor less than one fourth of the total amount inthe Quaternary. The Physiographic stage revealedthat when compared with the adjacent YangtzeRiver valley bottom, the Shennongjia Mts.uplifted about 1200 m since the Neogene, and theuplifting height during the Quaternary was ashigh as 600 m. It is, therefore, concluded thatthe climatic changes in the highest peak ofShennongjia Mrs. since the late period of theUpper Pleistocene of the Quaternary wasindependent on the crust tectonic uplifting,but dependent on the global clirr,rtic oscilla-tions. The up and down fluctuations of theperiglacial landforms and the vegetation zonewas mainly due to the influence of the East-,Asian monsoon circulation changes since thelast glaciation in the late period of the UpperPleistocene,OUTLINE OF THE PALAEO-CLIMATE ENVIRONMENTBEFORE THE HbLOCENEShennongjia has an ancient origin of floracomponents. It has many tertiary relic plantspecies. Some examples of such plant species1334 -

are given below.Ravidia involucucrata ball, deciduous wood,grows at moist lowlands ind valley bottomsbetween 1100-1600 m a.s.1.Ginkgo liloba L. grows on hilly land andplains that are below 1000 m in elevation.Liriodendron L. is the tertiary palaeotropicalrelic species.The tertiary re1i.c plant species in ShennongjiaMts. are more than ten according to preliminarystatistics.Up to the end of the Neogene and the heginningof the Quaternary, the highest peak mightreach 2500 m a.s.1. and two planation bencheswere distributed at the elevations of 1200 mand 600 m.During our exploration to the summit of thehighest peak from the northern slope, we foundneither Eossil glacial erosion landforms orfossil moraine. We climbed to the summit fromthe eastern gully along a large block streamwith an elevation of 2500 rn at the terminal,what we saw were all of fossil periglacialgeomorphogenetic scenery.2903 m a.8.. The vertical range of distributionis more than 200 m. The basal volcanic brecciathat made up the block field had well developedjoints. The frost wedged blocks varied from0.5-2 m in diameter and were commonly angular in,shape. The depth was usually 5 m or so (Photol).The region of the highest peak and, about20 km to the northeast, at Laojunshan Mt. had afossil block field line at 2800 m a.s.1. Besidesthe volcanic breccia in the composition of theblock field, there were also dolomite limestone,siliceous limestone and even Silurian greensandstone (Photo 2).CLIMATIC ENVIRONMENT DURING HOLOCENE AND THEPRESENTIn 1984, the author studied the drill coredata about the peat layer in, Dajiuhu regionreported by Zhou Minmin. According to theanalyses of the pollen assemblage and I4C dating,Zhou proposed that the vegetation and environmentdevelopment in Dajiuhu region underwentthree stages. (1) Between 12030-8033 years ago,pollens were mainly of Picea, Pinus, Tauga, Abies,etc. The accompanied broadleaf plants wereQuercus, Juglans, Fagus, Betula, Carpinus, Ulmus,etc. (2) Between 8033-3800 years ago, pollenswere mainly of Juglans, Fagus, Quercus, Ulmusand Carpinus. About 5000 years ago, Keteleeriadavidiana, Tsuga, Podocarpus and other coniferwoods increased markedly. (3) From 3800 yearsago to the present, pollens were mainly ofQuercus, Fagus, Betula, Ulmus, Toxicodendronand Carpinus.'From the first stage, we can see that therewere abundant conifers that resembled thepresent vegetation type at over 2000 m a.s.1.The small amount of peat mingled with the siltclay in profile suggests a dry and humidalteration of the climate. The second stagereflected that the vegetation was dominantlycomposed of warm and humid favoring plantspecies. The Tsuga genus is the inherentcomponent in the vertical zonation of subtropicalorographic vegetation in China. Fagus,Juglanwand Quercus are the major componentsof the present subtropical orographic evergreenand deciduous broadleaf mixed forest. Thissuggested that the Keteleeria davidiana, Tsuga, ,Podocarpus and other broadleaf arbores hadtheir flourishing period around 5030 years ago,ie. the Hypsithermal Interval. Using the presentmeteorological data of Dajiuhu region, we coulddeuce that the mean annual air temperature about5000 years ago was at least 4-3'C higher thanat present. The third stage was similar to thepresent components of vegetation type in Dajiuhuregion.FOSSIL PERIGLACIAL LANDFORMSBlock field: It has a large continuous distributionin Huazhongdin and its periphery area,down to the gentle slope of Feishaiyazhi atNivation hollows: One is situated in thevicinity of the fossil block field line at 2900m a.s.1. in the western side of Huazhongdinpeak. It developed in the volcanic breccia anddolomite limestone, and was usually 10-20 m indiameter (Photo 4). The other is near the fossilblock field line at 2800 m a.s.1. in the northernslope of Dashennongjia, It developed in thedolomite limestone and was characterized bylacking rock basin and ice threshold and alsolacking a complete ledge. In contrast, it31335

ottom was rather flat and was nearly eight onthe flat index. The bottom of the hollow waslarge and the largest could reach 150 indiameter. Some of the hollows connected withthe flat floor valley (Photo 5).Flat floor valley: It is a kind of periglacialwide valley. It mainly developed in dolomitelimestone on the northern slope of Dashennongjia.There existed nivation hollows in the sourcearea of the flat floor valley. There alsosurvived a fossil rock-bar of 20 m high andfossil frost heaving of about 5 m high on thetwo sides of the slopes of the flat floor valley.No moraine waa found in the valley floor (Photo 6).Photo 5. Ni~ation hollow Photo 6. Flat floorin the early ~ ~ e n ~ v~lley o n ~ in j the ~ ~ earlyperiglacial stage. ~ h e perigla- ~ ~ ~ ~ ~ ~ ~ acia1 stage.Cryoplanation terrace: There Is well developedsmall bedrock terrace at 2900 m a.s.1. betweenthe southern slope of Huazhongdin and Feichaiyia.The terrace inclines slightly downward and is50 m in width. It has a veneer of angular debrisresulting from frost weathering. There can beseen apparent bedrock steep treads in the background.There can be seen another old cryoplanationterrace, which is 120 m in width, at 2750 ma.s.1. under the above mentioned terrace(Photo 7).Block slopes: Their distribution is mostobvious in elevations of 2800-2600 m. They areaccumulations of angular and varying sizes ofblocks continuously covering the 30' bedrockslope surfaces. The lithological property offragments on the block slope on the slopes at2600 m a.s.l., the block slope, consisting ofvolcanic breccia fragments, covers the oldblock slope consisting of dolomite limestonefragments. Different block slopes connectednear the toea and thus formed a periglacialdebris fan.Stone streams: Masses of basal volcanicbreccia boulders with poor anRul.arity streameddown the gully bed of the Eastern Gully from anelevation of 2850 m to 2500 ,n. This is a wellpreserved stone stream, which is 3-5 m in depthdnd 10 m in width. The boulders are usually0.5-2 m in diameter and their length axes areparallel to the gully bed, There are Sinarundinarianitida growing in the fissures of therocks. There are no protalus rampart developingin the end of the stone stream (Photo 8 ).Frost weathering collapsed cliff: It wascreated under the complex situation consistingof periglacial agents, glSavity and water flowaction. In the source area of the Eastern Gullyat 2850 m a.s.l., the nearly erected collapsedcliff is as high as 30 m.Thin layered loess-like depositions: They canbe seen at the foot o f mountains, valley slopesor valley bottoms, and are characterized by0.5 m thick loess-like materials mingled withvolcanci breccia, limestone and feldsparquartziferous sandstone debris.In summary, the distribution of the mainfossil periglacial landforms in Shennongjia areshown in Fig.1.THE SIGNIFICANCE OF FOSSIL PERIELACIAL LANDFORMSIN SHENNONGJIAHuazhongdin peak, 3105 m a.s.l., is the secondhighest peak in the eastern Chinese mainland.It is only lower than Baxiantai (33"56.6'N,107"46'E, 3767.2 m a.s.1.). the highest peak inTaibaishan of the Qinling Ridge. Its latitudeis 2.5' southward of the Qinling Ridge and issituated in the subtropical zone. It is thetransitional bridge of the mountain region irthe west to the hilly land and plain region inthe east. It is more than 1000 m higher than Mt.Huangshan (1841 m a.s.1.) and Mr. Lushan (1426 ma.s,l*).During the peak of the Last Glaciation atabout 18000 B.P., there was a wide distributionof grassland and sparse tree prairie, and loessdeposition. It reached southward to Nanjin city,Wuhu Lake to Taihu Lake and the southern areawas along the Yangtze River bank, and to Jiujiangcity in the west. It belonged to an arid

REFERENCESMa Qiuhua and He .Yuanqin, (1988) Features ofmoraine and glacial period on Mr. Taibaishanduring Quaternary, (In Chinese), Journal ofGlaciology and Geocryology, Vol.10, No.1.Yang Liankang, (1991) Dlscovery of relic fluvialpebbles on planation surface in the Three-Gorge section of Changjiang, (In Chinese),Acta Geographica Sinica, 46(3): 373-374,Zhou Minmin, (1985) Vegetation and its environmentin Dajiuhu reRion of Shennongjia duringHolocene, (in Chinese), Thesis of M.S.,Tnst. of Geography, Academia Sinica.Fiy.1 Distribution of fossil periglacial landformsin the highest peak region of the ShennongjiaMts, (1:lOOOOO)1. peak: 2. block field: 3, stone fortress;4. block slope: 5, nivation hollow:6. fl.at floor valley: 7. Ice (soil) wedge;8. rock bar: 9. cryoplanation terrace:10. frost weathering collapsed cliff;11. scone stream: 12. precipice:and cold climate environment, with the meanannual air temperature being at least 6°C lowerthan present and even 10-12°C lower during thecoldest period, The annual precipitation was300 mm. The climate didn't favour the rlevelopmentof glacial morphogenetic features butfavoured the growth of periglacial morphogeneticfeatures.In the highest peak of the Shennongjia Mts.and Laojunshan Mt., many kinds of periglaciallandforms developed in the early Shennongjiaperiglacial stage during the late period ofthe Upper Pleistocene. The fossil block fieldline (2800 m a.s.1.) was representative and thegrowth of the flat floor va-lley was the mostobvious feature. A clear fossil block field linewas preserved at the altitude of 2900 m in thehighest peak of the Shennongjia Mts. The secondperiglacial process retreated to the summit andthe extent shrunk greatly. This can be extrapolatedfrom each single periglacial landformthat was greatly less in scope than the formerone. The well preserved fossil block fieldacted as the striking feature of this secondperiglacial climate.The Qinling-Taibai Glaciation during the lateperiod of the Upper Pleistocene consisted of twosubglaciations (Ma Q.H. and He U.Q., 1988).There were also two occurrence3 of periglacialprocesses during the Late Glacial time inShennongjia. Whether or not they coincided inchronology needs further precise explorationwith age dating. The Quaternary periglacialphenomena in Mt. Huangshan, wjth a southwardlatitude and constituted hy granite, are nottypical, while that in Mt. Lushan are unclear.Thus the fossil periglacial morphogenetic featuresin Shennongjia are of great value.Rased on the analyses of the physiographicstage, palaeobotany, present vegetation andclimate situation, it is evident that all ofthe Shennongjia Mountains were controlled in aperiglacial environment only at the time of theLast Glaciation during the Quaternary, and endedprobably before the start of the Holocene.1337 *

THE RESEARCH OF POROUS SLA3 STRUCTURES FOR PREVENTINGFROST DAMAGE OF ROADSZhu Yunbing' and Guo Zuxin''Yichun Management Department of Roads, Weilongjiang*Harbin College of Building and EngineeringIThe active mechanisms uf porous slab struct~r~s are analyzed from two aspectsof definite quantity and quality in order that the frost damage of roads canbe solved in engineering.INTRODUCTIONFrost heave and potholes cause serious damageof roads in seasonally frozen soil. Siliceousshale material has an extensive distribution InHeilongjiang. Some properties of siliceous shalematerial are tested, The soft and low strengthsiliceous stone cannot be used as a material inthe road structure layer, but in order todevelop and fully use the resources, mingledmateria 1 is designed basehd on the principle ofsuspens i on, and the crush rate and the strengthare tes t ed. The results show that the compressedelastic modulus of the mixed material is 200-300MPa and forms a porous slab structure. Becauseof the 1 ow cost and high protection of lime, themingled material of siliceous stone with a grainsize of 20-30 mm is mixed with a soil with alime co Nn tent of 1CI in the laboratory, andanalyzed. The mingled material is called sili-C ~ O U S stone which is used not only in the baseof secondary highways but also in the subbase ofhighways.PHYSICAL PROPERTIES OF MINGLED MATERIAL1. Hydraulic PropertiesThe porosity of the siliceous stone is ashigh as 35-40% and its pore size is tiny. Table1 lists the water distribution situation ofsiliceous stone and lime soil under optimumwater content of shaping for mlngled material,optimum water content of shaping is about 19%in experiments.The pore structure of siliceous stone has anexcellent water stability (see Fig.1).Siliceous stone docs not expand under waterinfluence. The volume expands after the limesoil absorbs the water and then decreases andvanishes because of the pores of siliceous stone.This can prevent the structures volume fromexpanding due to frost action pro,duced in theroad bed in poor drainage conditions. This issignificant for road structures in cold regions.Tahle 1. Water content of materials-Type ofFinalWater contentof materialmingled water before miximaterial , contentSiliceousLimestonesoil5:5 19% 23% 16.85:3:7 19.6% 23.6% 18.5%t 30 /"Fig.1 The curves of volume expansion andabsorbing rate2. Freezing Stability(1) In order to simulate the upper stratumof the road subjec,t to load, samples are teatedunder the condition of lateral constraint.(2) After the sample was saturated with waterfor hours, the sample went through five freezethawcycles.1338

(3) Freezing temperature was -2O"C, for four ceous stone mingled material of two gradinghours.proportions are taken for the base course ofThe thawing temperature is an ambient oneadvanced road surfaces and the trial road sec-(about 10-15'C), the time for saturation is 2tions are constructed.hours, After the samples are processed by theThe surface layer of the original road is 10above mentioned methods, the volume expands and cm of asphalt concrete. 'The base course has twothe strength loss is shown in Fig.2. After five kinds of 50 cm thick lime of semiaridity andfreeze-thaw cycleb, the mingled material.has an stability with steel dregs of fly ash and.lime.fairly good stability.In consideration of the heat preservation'property of mingled siliceous stone, because itis the base course, its thickness is decreasedby 10 cm in the basis of the original design(Fig.3). The structure of the siliceous stoneizarO - Osection is calculated in a three-layer system.%wThe design requirement is met in the areas ofthe downbending value and pull stress.@* - 40/ 00 - 2- u-4: I t 25cm I 3ocm "t1loCmflyash Isilicon stones I silicon stones ' "r8 I mingle -&rial I mingle materid I- 55 1 I 3:7 I3:7 95 7:3 1silicon stones : lime soilFig.2 The curves of volume expand andlossing strengthSiliceousstones C 325.9 Kcal/rn'.O'C:mingledmaterial(5:5)0.65 Kcal/m.h."CSiliceousstonesmingledmaterial(3:7)" -. "_CV 352.2 ' Kcal/me."CA,1.20 Kcal/m.h.'C-Wet clay W 0.28 l0OXc OT 587.3 K~al/m~.~'C:COTY 511.2 Kcal/m.h.'CXTM 1.301.05Kcal/m.h.'CKcaL/m.h.'CATfoundation4ocmI3. Thermal PropertiesThe coefficient of thermal conductivity forplus and negative temperatures are measured byFig.3 Section of the road surface structureinstalling instruments based on the line of heatsource principle in a transient state,FIELD TESTINGSome thermal parameters are given in Table 2.The measuring results show that the thermalFrost damage of road structures in the northconductivityand heat capacity of mingledern regions is mainly displayed by frost heavematerial are less than that of the base materials. and potholes.So the property of preventing freezing and heatBased on observation of 05 point datum markspreservation is very good and it can prevent the in a trial road section from the beginning offreezing thickness of the road surface.the winter until February, the datum markschanged by 1 cm. Yon-uniform frost heave basi-Table 2. Thermal parameterscally did not occur (Table 3). Subsidence deformationand net fissures did not occur untilMaterial Parameters Value Unitthawing in the spring. The total downbendingvalue is small (see Fig.4, Table 4). It wasCV L73.0 obvious that the froit-damage preventjon ,.lethodKcal/m'.*CAsphaltx 0.75 Kcal/m.h."Cis effective,x' 1.03 Kcal/m.h."CCONSTRUCTION OF TRIAL ROAD SECTIONRased on thc analysis of the ambience, sili-8A-\.,- 8-3 40 -8 \7 - 120 'CI:,.v2 so-\ ,#I 112 4MonthFip.4 Downbending value vs. time- mingled material of siliceous stones;steel dregs;_""-" . . f l y ash.

~ ~~ ~~~~~~Table 3. Observed frost heave of the porous slabFrost heave (cm)Structuretype 8612 87.1 81.2 87.3 87.4Max Min Max Min Max Min Max Hin Max Mi5:5 0,92 0.90 1.10 1.00 1.23 1.21 1.25 1.21 1.24 1.203:T 0.93 0.90 1.14 1.00 1.25 1.18 1.25 1.23 1,.25 1.23Steel dregs 0.89 0.89 1,15 1.13 1.30 1.25 8.40 8.01 8.41 8.40Fly ash 0.91 0.90 1.20 1.13 1.35 I .23 8.71 8.23 8.70 8.53Table 4. Calculated downbending valuesRoadsectiontypesCalculated downbending value (1/100) mmEnd of Autumnconstruction end DecreaseEnd ofspring thawIncrease3:?5:5Fly ashSteel dregsao 42 46285 46 46%143 24 83%106 40 h2W66 43%66 43569 187271 7 89,DISCUSSIONSince the trial road was constructed, drilling,sampling and analysis has been taken three times.The upper dTilllng profile is the thaw layer,the middle freezing layer and a thaw sublayer.This shows that the thermal property of the roadsurface structure is so different that thefoundation below different road surface structureshas a different distribution of tcmperatureand humidity.Fig.5 shows that humidity is basically identicalbelow a definite depth (e.g. below 220 cm).The identical extent of water migration fordifferent distributions of the temperature fieldresults in an indentical distribution of humidityabove the depth. It follows that the humidityof the soil foundation below the base of courseof siliceous stone is at a minimum. The adjustedaction of the water temperature of the basecourse of siliceous stone occurs many times. Thedrilling information in 1988 and 1989 shows thatthe adjustment action is brought into actionmore than one or two times.Siliceous stone mingled material of differentgraded proportions have a different distributionof the temperature field for different thermalconductivities..Fig.G shows the results oftheoretical calculations and observations, the0°C isotherm of the base course of siliceousstone mingled material. using different gradingproportions (3:7, 5:5). Its O'C isotherm iseasily above 30% for heat preservation actionof the base course being 50% siliceous stone inthe winter.It is obvious that the negative temperaturegradient of the road surface structure of thebase course with 50% siliceous stone is lessthan 30%.The greater the negative temperature gradient,the smaller is the amount of water migration,and vice versa. The film water of soil frozen at-3 - -5'C is measured in the tests. The hear+==-0.4 -0.81.2 .1.62.0 ---2.4 LWater content (%)Fig.5 Water content VS. depthx-5 : 5mingled materialo"-3 17- .steel dregso-..fly ashquantity affects the process of the O'C isothermdeveloping downwards. So the O'C isotherm of theroad surface structure of the base course with50% siliceous stone is below 307 in the latestage and the water content of the upper soillayer in the road surface structure of the hasecourse with 30% siliceous stone is Heater than50%.The total Erost heave of the road surface isdecreased in the base course of mingled materialwith the porous and tiny pore,siliceous stonesadjusting the water temperature of the soil -foundation. the soil foundation is kept dryerin the spring and makes the surface structurebe able to prevent freezing.

100 -150 -.200 -86 87Fig.6 Thc results of theoretical calculationsand observations- X- observation A-0- observation 9"-x"- calculation A"-0"- calculation BCONCLUSIONResearch shows that mingled materi.al of siliceousstone not only adjusts the water temperaturebut also has effective slab €unction. Itis feasible to design a structure layer of roadsurface, Resource investigation shows thatsiliceous shale is not only easily dug buteconomical. Because of the excellent propertiesof siliceous stone mingled material, the structurelayer of the road surface has the potential.to decrease the thickness of the embankment andcan reduce the engineering cost. Thus it is ofeconomic significance to use the technique andto develop the material.

~ .DRILLING CHARACTERISTICS OF ENGINEERING GEOLOGY OFPERMAFROST ON DA HINGGAN LING REGIONZou XinqingDa Hinggan Ling Investigation and Design Instituteof Management Bureau, Forestry Ministry of ChinaEngineering drilling of freezing stone soil was done in the. spring and autum seasons, in the north slopeof Da Hinggan Ling. The pore sizes were 130-150rnm, the length of the jackbit was 150mm, theoutblades were 2.0-2.5mm, the main axis pressure was from 800-1200kg when using a low speed andshort time, and the circle footage was 0.1-0.2m. The main axis pressure was changed to 600-1000 kg inthe soft permafrost stratum and the circle foot age was 0.3-0.6m,through practical investigations it wasproven that the above mentioned method met the technical demand and the improved drilling technologycan be used in drilling and sampling in engineering geologywith a favourable effect in permafrost regions.QUESTIONS ASKED ,.The stratum was of a freezing stone soil in the permafrost regionon the north slope of Da Hinggan Ling. The working area ofthe shallow drilling and pit was large, the efficiency was low andthere were many difficulties. When drilling machinery was used, accordingto the normal technical demand of drilling in the past. thetemperature of the frozen ground rose, the ice layer in the freezingstone thawed and even becamedry due to the friction between thedrilling rig and the stone soil of the pore wall. Sample takingcouldn't present the real condition of the freezing stratum, it wasdifficult to determine the depth of the permafrost table, measurethe water content of the freezing stratum and judge the ice extentand cyogenic tcxture type. The geotechnical propertics of'the constructionfield were misjudged which caused the destruction ofbuildings in serious cases. The technical drilling demand which isapplicable in the permafrost region of Da Hinggan Ling is given inthis paper, resulting from extensive research in the last thirty years.It can meet the demand of engineering design and preserve the stabilityof buildings.DRILLING TECHNIQUES FOR ENGINEERING GEOLOGY.. ... .-." . "IN THE PERMAFROST REGION"". .. .. - -. . .Choice of Drilling Season in The Permafrost Region in The North. .,"" "~ . , ~, . ~~S'ope-?fDa Y!!?ggan,Ling.,In general, drilling can be done in any season due to theunique natural environment and climate conditions in thepermafrost region. A feasible season must be chosen to ensure thedrilling demand of engineering geology in the permafrost region.In the permafrost region of the north slope of Da HingganLing, the winter is long and the climate is frigid. The daily averagetemperature is lcss than O°C lor more than 200 days of the year.Daily low tcrnperature of -3OOC has thc continuous time of100-130 days. Daily low temperatures of -4O'C occur for 20-30days. There are heavy rains in the summer and prccipitation is concentratedin July, August and Novemhcr. According IO the demandsand purposes otenpineerinp'it IS important to choose a suitabledrilling time. Drilling seasons for engineering geology arefeasibly chosen in February, March, June and November. Thepermafrost table can be exactly determined by drilling from July toOctober. The maximum depth of seasonal freezing can be determinedby drilling from the middle of March to the last ten days ofApril.Choice of Drillign Pore Site. -" _x . .In order to decrease the thaw outside the undisturbed core ofpermafrost due to friction heat of the $ckbit, the drilling machinetype and pore size of the drilling must be suitably chosen. A largenumber bf comparative experiments in the last few years haveshown that the effectiveness of the DPP-100-1 type of drilling machineis the best, next is the Xy-100 type. Drilling machines with alarge horsepower are suitable for drilling in permafrost regions.According to experiments in drilling in permafrost regions, such asin Gulian, Xilingji, Tuqiang, etc, a 150mm pore size is feasible for afrozen stone stratum, the drilling core is more complete, the soiltemperature doesn't change easily and a 130mm pore size can meetthe demand of the experiment.DRILLING FOOTAGE FOR ENGINEERING GEOLOGY"I- ..."Drilling footage of clear water or slurry is nat feasibly used inpermafrost regions, dry drilling is preferable. Because of the highstone content in the permafrost stratum of Da Hinggan Ling,1342 -

choosing a jackbit of a stiff alloy should meet the following technic:~]demands to gain the best effect.I. Length of jackbit is 7 Amm.2. A k534 type alloyjackbit is selected, with the outblade of theonsideand outside being not less than 2.0-2.5mm.3. Length of core barrel is 0.6-0.8m.TECHNIQUE OF DRILLING OPERATION"." . . ..- . . .. ..There is an extensivc distribution of frozen stone soil with differentice contents and which is more compacted and soft peat soilwith a high ice content in the permafrost region on the north slopeof Da Hinggan Ling. The drilling demand is not identical.Freezing Stone Soil with A LOW Ice. Content And High~ ~ " "... - ~~CompactionThe volume of the ice content in this soil is small, the generalcondition is less than 20% of the ice layer is filled in the pores ofthe stratum and stones, and. it is difficult to take ice core in the processof drilling footage. So drilling footage uses a low speed and amiddle main axis pressure. If drilling footage uses a high speed anda large pressure, the time is longer and the heat of the drilling rigcauses the frozen ground core to thaw. According to experimentaldata, circle drilling footage is not feasible for more than 2-4 minutes,footage is feasible in the rangc of 0.1-0.2m. Center pressure ofthe drilling rig with a 150mm jnckbit is feasible in thc range of800-1 200kg.Frozen Soil with A High Ice Content and Soft Peat Stratum-- . "" . -"This soil is mainly distributed in the basin and sedimental stratumof the lake facies in Gulian coal mine. Most belongs to a stratifiedcryogenic texture and the stone content is less. Drilling footageof a middle speed should be chosen, the circle footage should bc0.3-0.6m, and the main axis pressurc of the drilling rig can be incrcascdin the drilling footage in the peat stratum. Because of thedeeper outblade of the alloy jackbit is pressed into the frozenground, a higher core ratio is gained when the core sample is drawnin the drilling footage. The main axis pressure should be increasedso that the jackbit and core are closely blocked so that,a drop in thecore can be prevented and the core drawing ratio is improved,In order to Iind the destructive causes of the fissures in thehospital of Tuqiang Forestry Bureau in Da Hinggan Ling, drillingwas done around the foundation' of the building in November,1984. Using the method of dry drilling, large axis pressure and highspeed, the frozen ice core was completely thawed, the core producedheat and no permafrost was found. Using the same method,a second drilling was done with a DPP-100-1 type drilling machinepermafrost and the ice layer were still not found in the core ofdrilling pore. Next, a low speed and middle pressure, with a circledrilling footage of 2-3 minutes was used, and the ice layer andpermafrost were found in the core. This shows the importance inchoosing the drilling footage parameter in engineering geology drillingin permafrost. Drilling footage with this type of drilling machineshould be chosen by the following parameters of drillinpfootage. 'Rotational speeds using the first gear, small-middle throttle,and 40-60 rotations per minute.Total pressure of the axis center of the drilling bit is800-1200kg.Circle drilling footage, in which a l5Omm jackbit is used, is 2minutes for a frozen stone stratum and a quick drawing jackbit. Acircle drilling footage, in which a 130mm jackbit is used, is 2-3minutes for frozen loam, and a circle drilling footage,using a130mm jackbit, is 3-4 minutes for frozen cobble stratum, but theaxis pressure of drilling is more than that of frozeng loam andfrozen stone soil.The length of the core barrel is 0.6-0.8m (being suitable fordepths of 0-10m) amd the length of the jackbit is ISOmm. Threeidentical size drillings are prepared, after the first circle end anddrawing drilling, and another drilling is ne'eded. Drilling of the firstcircle can not be continually used until the drilling has cooled.The above mentioned method of drilling and parameters ofdrilling footage are very effective in drilling and sampling processesof permafrost engineering geology.CONCLUSIONS". . "A large number of experiments in engineering geology drillingin tb permafrost regions of Gulian, Xilingji, Tuqiang, etc, showthat drilling in different stratums have different parameters for drillingfootage. A reasonable pore size, drilling and operational techniquesare not only the key factors for excellent cffectivencss in engineeringgeology in permafrost, but are also important factors onwhich the properties of permafrost can be determined. A satisfactoryeffect was recieved using a 130mm pore size, 150mm longjackbit, 9 not less than 2.0-2.5mm outblade of the inside and outside,800-1200k total pressure of the drilling axis center, circlefootage of 0.1-0.2m for frozen stone soil and 0.3-0.6m for frozensoft soil, etc..ACKNOWLEDGEMENTS"The author wishes to thank Assistant Professor TongChangjiang and the Lanzhou Institute of Glaciology for pcrmittingthis paper to be published.

QUATERNARY GEOLOGY AND GROCRkOLOGY IN NORTMERN QUEBEC, CANADAMichel Allardl and Jean A. Pilon2ICentre d'etudes nordiques, Universite csval Clement tremblay,Ministere des transports du Quebec'Geological Survey of CanadaIn Nunavik (Northern Quebec) theorical and practical knowledge of periglacial features and permafrostproperties has increased tremendously over the last five years. We owe this achievement toa joint task force (governments-universities) set up to carry out a comprehensive research programwithin the framework of airstrip construction in 14 villages distributed over six degrees oflatitude, Erom the discontinuous to the continuous permafrost zone. Patterned ground types andgeocryological facies are closely related to the various types of surficial deposits laid by theWisconsinan glaciation and, subsequently, by the post-g'lacial marine inundation and emergence on thecoastal fringe of the peninsula. Tundra polygons and low-center mudboils are the dominant periglacia1features on glacial landforms such as drumlins and moraines, on glaciofluvial landforms likeoutwash plains and eskers and on fluvial terraces and raised beaches. Blth soil wedges and icewedges can be found under polygon sides. Soil stripes and gelifluxion lobes occur on slopes.Interstitial ice dominates in these coarse, sandy and gravelly, sediments as revealed'in numeroussections, quarries and drill-holes. The dominant periglacial features on the post-glacial marinesilty clays are cryogenic mounds and fields of high-center mudboils. In these fine-grained sediments,segregated ice (lenses and reticulated) dominates in warm permafrost (O'C to -2'C) in thesouthern part of the region. Volumetric ice contents, AS measured on cores, are usually 50-60% but'they may attain 809, in the upper layers of permafrost. Massive icy beds occur in cold permafrost(-5°C to -5°C) in the northern part of the region and were observed in retrogressive slumps. Groundprobing radar and electrical resistivity surveys indicate that such icy beds may be up to 11 mthick. Other features such as palsas ih bogs, frost blisters, icings and aggrading permafrost undertidal marshes are also significant in the region. Extensive knowledge of correlations betweenQuaternary geology, ground patterns and geocrynlogical facies (displayed in table form andillustrated with selected pictures) has made it possible to develop a photo-interpretation key whichallows t'o plot high quality, large scale maps of permafrost conditions prior to decision takingrelative to digging, cutting into or moving soil for construction projects (buildings, service andtransportation infrastructurs, etc.).RATIONAL UTILIZATION OF WATER RESOURCES IN PERMAFROSTREGIONS, ARTIFICIAL BECHARGS OF GROUNDWATER STORAGET.V. Burchak and L.M. DemidyukThe All-Union Oil & Gas Research Institute, Moscow 125422, RussiaCharacteristics of surface and groundwater in permafrost regions are given. It is shown that themost typical resources of ground water are those orig'inating from taliks of river valleys: subpermafrostwater of mountain artesian basins; water from structures of limited area or from creviceand creviskarst ground mass and zones of tectonic disturbances, The necessily of due regard forpeculiarities of the formation of water resources is marked, and of the surface and subsurfacedrainage in planning the water resources utilisation and protection from pollution and depletion,and also in designing water-supply systems.Comparative assessment of surface and groundwater as the source of water supply is given withregard for special requirements depending on severe climatic conditions. Possibility and expediencyof wide utilization of grouridwater for supply purposes is shown. Possible water intake for variousgroundwater basins is characterized.The necessity of researches for non-conventional "Know-how" of waver supply is substantiated withregard for permafrost region peculiarities and special requirements of consumers. Particularattention is given to the selection of water supply sources for oil production regions.Expediency is shown of wide utilization of teli.ks of river valleys together with artificial rechargeof groundwater (ARGW) by:1) building water storages: 2) widening of river beds: 3) installation of special infiltrationfacilities: 4) creation of extra negative water pressure in the depression funnel under the riverbed: 5) joining a sub-bed-near-bed talik with that of a near-bed one by means of melting thepermafrost pillar separating the taliks; 6) enlarging the talik capacity by creation of optimumconditions for melting permafrost ground:,'7) taking measures against freezing of the water intakearea of the talik.Principal technical solutions for realization of ARGW method are characterized as well as thedesign technique of infiltration and vacuum facilities, melting procedure of permafrost pillars,and also the water'intake on the whole.Results of the water supply network selection for an oil production site located in continuouspermafrost region are considered.

ICE WEDGE DEVELOPMYYT ON SLOPES, FOSHEIM PENINSULAELLESMERF, ISLANO, FASTYRN CANADIAN ARCTICAntoni G. LewkowiczDepartment of Geography, Erindale College, University ofToronto, Mississauga, Ontario, L5L 1'26, CanadaThis paper describes the cryostratigraphy of sites on the Fosheim Peninsula where either multipleice wedge systems or multiple growth stage ice wedges have been exposed and provides explanationsfor wedge regrowth.Seven sites within a 50 km radius of Eureka were examined from 1987-1991. Current cracking activityof wedges was assesse4 by taking samples across the centres of wedges or in younger growth sta8e.s'and analysing the tritium content. Rates of wedge growth inferr.ed from this method are between 1 and5 mm/year. Tn the headwalls of retrogressive thaw slumps wedge regrowth occurred within formermudflow deposits, but on one low-angled slope, the headwall was in previously undisturbed materialand active wedges exhibited secondary and tertiary stages 29 cm and 10 cm high respectively (currentactive layer of 53 cm), In an exp3sure beside a recent active-layer detachment slide, four stagesof growth were present in an inactive wedge, together with several major layers of segregated ice.The most complex section was at the base of a long slope where colluvial deposits overlay peat.Wedges of at least two and possibly three different ages were present. Wedge 1 was the oldest andgrew syngenetically at least partly during the period of peat accumulation which ended at 6460f70B.P. A secondary vein developed as the surface aggraded further but eventually this became inactive.Wedge 3 may have started epigenetically or syngenetically but existed prior to 540f50 B.P. Sincethen it has grown syngenetically and is still active. Wedge 4 is the youngest wedge: it wasinitiated epigenetically after 540f50 B.P. The rate of sedimentation at this site derived from themean corrected radiocarbon dates is 7 mm/year, but given the standard errors could be between 4-25mm/year. This surface aggrada.tion is attributed to a combination of slopewash, solifluction andactive-layer detachment processes. Although not observed to date, it is presumed that antisyngeneticwedges exist upslope of this and many of the other sites studied'.There is no evidence of recent cooling within climatic records from Eureka, but the coastal locationof this weather atation could .buffer it from changes that may have occurred further inland. Thus theuppermost growth stages at several of the sites could be the result of changes in rhe summerclimate. However, the degree of active layer change required means that this explanation is veryunlikely to be correct for the lower growth stages. Instead, regnowth is attributed to active slopeprocesses which result in colluviation.LAKES AND PERMAFROST IN TYE COLVTLLE RIVER DELTA, ALASKAH. Jesse WalkerDepartment of Geography, Louisians State University3 Baton Rouge, Louisiana, 70803-4105, USALakes in arctic deltas, like lakes in the deltas of lower latitudes; vary greatly in area, depth,shape, seasonal character and formation processes. Lakes in the Colville River Delta on the coastalplain of Northern Alaska provide examples of this variety and reflect the numerous factors,including permafrost, that are involved in arctic lake for ation and maintenance.The delta, only 550 km in area, has innumerable ponds andL41 lakes over 0.0125 km' in area: 5 ofthem over 2 km . Lake area (for those larger than 0.0125 km') is 96.5 km or 17.5% of the totalarea of the delta.The lakes and ponds in the delta include remnants of the well-known oriented lakes that occur overmuch of the coastal plain, abandoned rlver channel lakes, terrace flank depression lakes,thermokarst lakes, intra- and inter-dune lakes, intra- and inter-bar lakes, perched lakes, iceshove ponds and ice-wedge pol-ygon ponds. Although some of the delta's lakes are sufficiently deep(over 8 m) so that a thaw bulb exists beneath them, most are less than 2 m deep and freeze to thebottom during winter. These lakes have thin active layers in their bottoms similar to that of thetundra surface. Mowever, the importance o € permafrost in lake morphology evidences itself best inrelation to shoreline erosion, the eventual breakthrough (tapping) of river to lake and subsequentlake drainage.Lakes associated with ice-wedge polygons, some of which are perched several meters above normalriver level near river banks, frequently drain. Drainage usually occurs on top of an ice wedge thathas been cut into by the eroding river leaving the polygon lake dry. A similar process occurs inother lakes (many of which have bottoms well below river level). Once the tundra separating thelake from the river is suEficiently narrowed, breakthrough can occur. After the lake becomesconnected with the river its level fluctuates with river stage. During non-flood stages (all but3-4 weeks per year) the erosional benches, formed along the margins of the lakes as they enlargeddue tQ lake-bank thaw and erosion, are subaerial and thus subjected to the permafrost formingconditions that are typical of the coastal plain. Further, once tapped, lakes become sediment trapsfor the river's flood water. Such lakes are eventually filled with sediment and the thaw bulb thatexisted beneath them freezes.1345

SORTER CIRCLE DYNAMICS: 10 YEARS OF FIELDOBSERVATIONS FROM CENTRAL ALASKAWalters, James C.Department of Earth Science, University of Northern Iowa,Cedar Falls, Iowa,50614, USAField studies of sorted circles occurring in the periglacial environment of the Maclaren Summit andHigh Valley areas of central Alaska have been conducted over a period of 10 years. Repetitivephotography and surveys of wooden dowels, metal rods, and marked stones provide information on thesurface dynamics of the circles. Excavations provide information on subsurface characteristics.Although much variability was noted in the dynamics of the circles over the study period, somegeneral conc1usio.s can be drawn and inferences can be made within the context of existing modelsof sorted circle development.'The circles examined in this study are found in low spots in the silty till which blankets the area.These depressions become quite wet in the spring because of snowmelt, rain, and thawing ground ice.Sorted circles have been studied at six sites. three where temporary ponding occurs and three whereseasonal wetting but no ponding takes. place, Overall, movement of markers in circles was upward andradially outward from the fine-grained centers to the coarse borders. Circles at sites experiencingtemporary ponding showed the most activity. Vertical displacements of markers were greatest incircle centers with an average uplift of approximately 1 to 4 cm/yr. Horizontal displacements ofmarkers generally increased from circle centers outward but then decreased as they approached withinseveral centimeters of the coarse borders. Average horizontal movements range from 0.5 to 2.2 cm/yr.Circle characteristics differ depending on positiun in the shallow depression in which they occur.Lsrger circles 1-2 m in diameter are found along the outer margins of a depression. These largercircles are slightly to strongly domelike in their centers, closely spaced, and more oval or evenpolygonal in plan view. Fine-grained sediment is abundant here and some vegetation exists. Circlesin the central portion of a depression are smaller, 0.5-0.1 m in diameter, strongly convex upward,and widely spaced with a stony surface surrounding them. Fines are pres'nt only in circle centers,and no vegetation exists. Another common feature in the central area of a depression is a pattern ofsmall stones.surrounded in a circular fashion by larger stones. No fines are present on the surfacein these features, but excavations reveal plug-like areas of fines beneath their centers.Long-term monitoring of these sorted circles indi'cates the diameters of most are becoming smeller.Repeat photography and measurement of circle diameters show the central aiea of fines decreasingwhile the border areas are expanding. The fine-grained centers are also becoming coarser. Observationsshow that coarser material progressively moves up and out away from a circle center toconcentrate along the stable border. Fines also move up in a plug-like fashion into the circlecenter where they dominate the center, but they are slowly removed by deflation, rainwash, sheetwash,etc. Therefore', the diameter of a circle becomes smaller as more stones fill in the circleand the fine-grained sediment becomes separated from its source. If conditions allow for continuedactivity, a circle will progress to an end point where no fines exist on the surface,. and whatremains is a pattern of small stones surrounded by larger stones.- 1346

AUTHOR INDEXAguirre Puente I., 368,611,1124Akagawa Satoshi, IO50Akerman Jonas, 1022Akscnnov V.I., 1Alexeeva Olga I., 855Alifanova A. A., I2 1 9Allard Michel, 5,182,1344An Viktor, 843An Weidong, 11Antonov-Druzhinin Vitaly P., 1054Are Felix E., 436.846Arts R., 286Aziz A., 17Barry R.G., 23 .Barsch Dietrich, 27Bartoszewski Stefan A., 32Baulin V.V., 1060Belloni S., 36Biggar Kevin W., 42Bird Kenneth J., 94Bondarenko G.I., 851Brennan A.M., 23Brewer M.C., 48Brown Jerry, 969,972.1 I32Brudie E.L., 244Bruskov A.V., 1Burchak T.V., 1344Burgess M.M., 54Burn C.R., 60Burns R.A., 66Butsenco A.N., 506Caine N., 1044Ca1deroai.G. ,72Caldwell J.B. ,244Carnes-Pintaux A.M ., 11 24Carter L.David, 48,78Carton A., 36Cater Timothy C., 316Chamanova 1.1. ,1060 .Chang Rudolf V., 855Chang Xiaoxiao, 722,1274Chang Yen, 596Chehovalty A.L., 1060,1062Chen Hongzhe, 105Chen Qinghua, 383,1159Chen Ruijie, 1064,1067Chen Xiangsheng, 1070,1171Chen Xiaobai, 84,143,689,1037,1073Cheng Enyuan, 302, I 152Chew Guodong, 675,965,971,1010Chernyakov Yurii A., 862Chuvilin E.M., 89,160,1255,1295Cohen Tenoudji F., 61 1Collett Timothy S., 94Collins Charles M., 1076,1128Corapcioglu M.Yavuz, 100Corte A.E., 1073Cui Guangxin, 1079Cui Jiqnheng, 105,1082Cui Yongsheng, 407Cui Zhiju, 111,397,1086,1287Dai Baoguo, 116Dai Chtmtian, 1 I6Dai Huimin, 120Dai Pin, 116Dallimore S.R., 125Danilov Igor D., 858Dash J.G., I I 17Demidov V.V., 506Demidyuk L.M., 1344Dew Yousheng, 131,773,1255,1295Devjatkin V.N., 134Ding Jingkang, 138,1092Ding Yongqin, 143Domaschuk L., 149Dramis F., 36Du Chengxian, 1 I6Dubina Mikhail M., 862Duhuil Marie-Andrec, 255Duchkov A.D., 134Dydyshko P.I., 155Ershov E.D., 89,160Esch David C., I64Everett Kaye R., 267Fang Tsung Ping, 586Fediukin Igor V., I70Fedoseeva Valentina I., 865Fei T., 500Fei Xueliang, 1096Feng KG, 768Feng Yanhui, 789Ferrell John E., 471Ferrians Jr.0.J. , 1132Forbes Bruce C., 176Fortier Richard, 182Fotiev S.M., 955 IFowler A.C., I100Frech Hugh M., 482,968Frolov Anatoly D. , 170Frydecki Janusz , 5Fukuda M., 488Gao Min , 1 I67Gao Weiyue , 1265Gao Xingwang , I88Garneau R.R., 286Gavrilov M.K., 987,1006Ge Huanyou , I I48Gerasimov A.S. ,955'Gcrshevich V.B., 506Giardino John R. ,1019Gilichinsky David A., 869Glenn R. ,48Goncharov Ju.M., 875Gorbunov A.P. ,1105Gorelik Yakov ,879Goriainov N.N. ,66Gotovtsev Semyon P. ,891Gray James T. , 192Grechishchev Stanislav E., 54,198Greeley Nancy H. , I I28Gu Zhongwei ,204,388,778,819Guan Zhifu , 1298Guevorkian S.G. ,660Guglielmin M., 72Guo Dianxiang ,1108Guo Dongxing ,210,282,809,1186Guo Xingming ,835Guo Xudong , 1113Guo Zuxin ,1338Guryanov Igor E. ,885Guthrie Robert S. , 694Haeberli Wilfried, 214,272,1014Haiying FU , 1 I I7Hall Kevin J. ,220Hallet B. ,226,l I17HanUK,1119Hansueli Gubler ,332Haoulani H. , 1124Harris Charles, 232Harris Stuart A. ,238,1019Hartzrnann Ronald J. ,574Haugcn Richard K., 1076,1128Hazen Reez ,244,494He Ping, 250He Yixiang ,718Heginbottom J.Alan ,255,1132Heuer C.E., 244Hinkel Kenneth M. ,261Hinzman Larry D. ,267,326Hirakawa Kazuorni ,449Hivon Elisabeth G. ,42Hoelzle Martin, 214,:72Horiguchi Kaoru II 1064,1047Horrigan Timothy 0. , 1076Hou Zhongjie ,556,608Hu Qiheng ,967

Hu Ruji ,1144Hu Shicai ,4 1 6Huang Junheng , I148Huang MaoHuan ,278Huang Xiaoming ,1010Huang YiZhi ,210,282,758,809Huncault P.A. ,286Hunter J.A., 66Iordancscu M. ,286Jakob Matthias, 27Janoo Vincent, 292Jian Gong, 298Jiang Hongju ,302,1152Jiang Weiqiang ,592Jiao Tianbao , I 155Jin Huijun ,307,803Jin Naichui , 3 I2Jin Zhengmei ,278Jorgenson M.Torre ,316Joshi Ramesh C, ,706Joyce Michael R. ,316Jr Bayer John ,292Judge A.S. , 1 I ,66Jung Duhwoe ,648Jung H.C. , 11 I9Kagan A.A., 730Kamensky Rosteslav M. 322,923,969Kane Douglas L. ,267,326Kang Xingchcag ,10 IOKasse C., 643Kella Felix ,214,272,332Kcrshaw G.Petcr ,338King Lorenz ,344,625,1022King Lorenz, 1022Klirnovsky Igor V., 891Rlimowicz Zbigniaw ,350Kolunin Vladimir ,879Kondratyev V.G. , 155Koniakhin M.A., 937Konrad J.M., 550Konstantinov Innokentii P., 322Koster E.A., 987Krantz W.B. , 1044.1 100Kritsuk L.N., 897Krivonogova N.P. ,730Kunitsky Viktor V. ,903Kurfurst P.J. , 54,356Kurilchik A.F., 909Kutasov I.M., 362 'Kwok R., 149Lachenbruch A.H. ,987Lauriol Bernard , 192Lebedenko IU.P., 160,1255,1295Leclaire P. ,368,611Lehmann Rainer ,374Leibrnan M.O., 380Lewis G.C., 1044Lewkowicz Antoni G. ,232,1345Li Bin, 1096Li Dazhou ,592I.' Dongqing ,835,1164Li Gang, 278Li Guangpan , 1147Li Hao ,383LiKun,1171Li Shijie , 1174Li Shude ,1174,1178Li Yi ,1079Li Zuofu ,1178Liang Fengxian ,204,388,744,819Liang Linheng ,204,393,778,819Lilly E.K. ,326Lin Chuanwei ,685Lin Ying ,312Lin Yipu ,1272Liu Fengjng ,738Liu Gengnian ,397,1086,1287Liu Hongxu ,403Liu Qingren , I 16,407Liu Rihui , 1251Liu Shifeng , 1183Liu Tieliang , 1298Liu Xuekui , 1308Liu Yifeng ,429Liu Yongzhi ,122,764Liu Zongchao ,429Lomborinchen R. ,411Lou Anjin , 138Love11 CWilliam , 968Lozcj A. ,7?.Lu Heiyen ,797Lu Xingliang ,416,1301Lunardini Virgil J., 17,420Luo Anjing ,1092Luo Guowei , I 186Luo Minru ,426,622Luo Weiquan , 1190Ma Hong ,429,1144Ma Wei ,432,556,122,1214Ma Yijun , 1108Magierski Jan, 32Mai Hcnrik , 1137Makarov Vladimir N., 91 1Makeev O.V. ,506Mamzelev Anatoly P. ,436Marsh Philip, 443Matsuoka Norikazu ,449Melke Jerzy ,350Melnikov E.S. , 54,356,1132Melnikov Vladimir ,455Meltzcr Liya I. ,914Men Zhaohe ,1193Mi Haizhen ,461,1259Mia0 Lina ,278Miao Tiandc , I197Michalczyk Zdzislaw , 32Michalowski Radoslaw L., 465Migala Krzysztof, 919Moblcy Keith F. ,471Molmann Truls ,477Moore J.P. ,517Moskalenko N.G., 54Murashko A.A. ,89Murray D.F., 48Murton Julian B. ,482Na Wenjie ,592Na Yunlong, 426Nakano Yoshisuke ,750Nakayama T. ,488Nelson Frederick E. , 261,987Nixon J.F.(Derick) ,244,494Noon G.G., 1100Olovin Boris A ,923Osterkamp T.E. ,783,500,987,Ostroumov V.E., 506Outcalt Snmucl I. ,2610zouf'J.CI.. 523Pan Anding. 1202Panday Sor;lh M. . 100Pavlov A.V. , 51 IPerlshtein G.Z. ,909Pewe Troy L. ,966Pilon Jean A. , 5,1344Ping C.L., 517Pissart A. , 523,972Polyakov V.A., 897Popov Viktor A. , 322Poznanin V.L. ,660Prick A., 523Prigoda V.Ya., I55Pu Yibin , 1208Qiao Dianshi ,529Qiu Guoqing 307,533,803,1028,1312Qu Xiangming ,312Ramos M., 121 IRasmussen L.A. ,226Ren Zhizhong , 1215Rivkin F.M., 380,869Roman L.T., 1219Rooney James W., 648Roujansky Vladislav E., 858Sadakova M .N., 155Salnikov P.I. ,927Samarkin Vladimir A., 869Samyshin V.K., 909Sarrelainen Seppo ,539Saveliv V.S. ,380Savitsky Victor A. , 846Schmid Willy , 2 14,654Schmitt Elisabelh , 544Schofield A.N., 1070- 1348 *

sodov BM., 1222Sego Dave C. ,42Seguin Maurice". ,182Senneset Kaarc ,477Shamanova I.I., 1062Shankov Vladimir V. ,198Shao Lijun ,1292Sharkhuu A. , 1223Shcn Mu, 550Sheng Yu ,556,1073Shag Zhongyan ,250,1274Shmg Zhongyan ,1274Shestarnyov D.M., 1227Shi Yafeng ,968,972Shields D.H. ,149Shoop SaUy A., 559Shpolyanskaya N.A. ,930Shur Y.L. ,564Sicgcrt Christine , 569Skaret Kevin D. ,338Skvorpov A.G., 66SlaugL Lr Charles W. ,574Slavin Borovskiy V.B. , 564Sletten Ronald S. , 580Smiraglia C. ,36Smith C.C., 1070Smorygin Gcnnadi ,455Snegirev A.M., 934Sokolova L.S. , 134-Solomatin V.I. ,937Sone Toshio ,488,1231Song Changqing , 11 1,832So0 Sweanum ,586Stein Bernd , 1238Streltsova O.A. , 865Su Shengkui ,1235Sui Tieling , 592Sun Zhenkun ,407SVGC Otto 1. ,596Takahashi Nobuyuki , I23 1Tang Shuchun ,602Tang Xiaobo ,605Tang Zhonghai ,832Tao Zhaoxiang ,608,773Tarasov A.M. ,356Taylor A.E., 125Tellini C. ,72Tenoudji f.Cohen ,368Thimus J.F. ,611Thomscn Thorkild , 1137Timofeev V.M. ,66Todd B.J. ,66Tomita Hsiao ,292Tong Boliang ,617,1269Tong Changjiang ,622Torgashov Y.Y. ,941Tremblay Clement, 5Trombotto Dario ,1238Tschcrvova E.I. ,356Tu Guangzhi ,967Tumurbaatar D., 1242Ulrich Roland, 625Urdca Pet, ,631Uziak Stanislaw ,350Vakili Jalal , 1247Valuyev A.S. , 155Van Everdingen Robert O., 638Vandenberghe J., 643Varhchilov Yu.Ya. , 1222 "Vasil'chuk Yurij K. , 945Vasilyev M.L., 155Vinson Ted S., 648,1031Vlasov Vladirnir P. ,951Vonder Muhll Daniel S. ,214,654 ~Vtyurina E.A. ,660Vyalov S.S., 955Wagner S. ,214Walker HJesse ,1345Walsh Michael ,292Walters James C. , 1346Wang Baolai ,664Wang Binlin , 1186Wang Changshcng , 1171,1251Wang Chunhe ,670Wang Guangzhou ,832Wang Jiachcng 675,734,778,1255,1295Wang Jianguo ,1282Wang Jianping ,678Wang Qiang ,744Wang Shaoling ,461,1259Wang Shirong ,1262Wang Shujuan ,307Wang Wenkai ,685Wang X.L., 120Wang Yaqing ,689,1073Wang Yi ,1265Wang Yingxus ,1178,1269Wang Yinmci ,768Wang Zengting ,678Wang Zcrcn ,730Wang Zhanchen ,755Wang Zhenyi , 1272Wayne William J. ,694Wei Xuexia , 1197Wci Zhengfcng , I108White T.L. ,700Wipweera Harsha ,706Wilen L. , I 117Williams P.J. , 700Woo Mingko 443,112,725,738,987Wu Jinrning , 138Wu Qi jan ,678Wu Qingbai ,718WU Ziwang I I ,432,722,1274, I32 1Xen Zhenyao ,797Xia Zhaojun ,725Xia Zhiying ,758,1278Xie Yinqi ,1282Xiong Heigang ,397,1086,1287 ,Xu Bomeng ,416,730,1292,1301Xu Dongzhou ,105Xu Jingguang ,529Xu Xiaozu 131,734,773,1255,1295Xu Xueyan , 1092Xu Zhmghai , 1 148Yakushev V.S. , 160 ,Yang Daqing ,738Yang Hairong , 1298Yang Lifeng , I 30 1Yang Zhenniang ,738,744Yang Zhihuai ,744Yao Cuiqin, 1082,1317Yen Yinchao ,750Yi Qun ,797Young Kathy L., 712Yu Qihao ,250,1304Yu Shengqing ,416,755Yuan Haiyi , 1308Yue Hanscn ,I3 12Zcng Zhonggong ,307,758Zhw Changqing 432,722,764,1197,1219Zhang Duo, 789Zhang Hengxuan ,969Zhang Huyuan ,768Zhang Jianming ,764Zhang Jiayi ,250Zhang Jinzhao, 138,1317Zhang Lianghui , 1265Zhang Lixin, 131,608,773,1255,1295Zhang Qibin ,204,778,819Zhang T., 783Zhang Tiehua , 1235Zhang Xianggong ,768Zhang Xikun ,529Zhang Xin ,789Zhang Yuanyou ,312Zhang &YOU ,793Zhang Zhao ,1321Zhang Zhaoxiang ,797Zhao Jun, 813Zhao Lin ,307,803Zhao Xiufeng ,210,282,809Zhao Yutain ,8 13Zheng Qipu ,1326Zhou Dcyuan ,1330Zhou Xinqing ,426Zhou Youwu ,204,393,778,819Zhou Zhongmin , I334Zhu Chcng ,826Zhu Jinghu ,832Zhu Linnan , 764,835,1164

Zhu Qiang ,838Zhu Yuanlin , 1 1,250,970,1304Zhu Yunbing , 1338Zolotar A.J., 955Zou Xinqing ,1183,1342Zuo li ,11481350

GENBRAL SUBJECT-SENIORAUTHORINDEXCANALChang Rudolf V. ,855Jian Gong ,298Jiao Tianbao ,I I55Jin Naichui ,312Li Anguo ,383Qiao Dianshi ,529Rcn Zhizhong ,I 215Wang Wenkai ,685Xu Bomeng ,730Zhang Changqing ,764 .Zhang Zhao ,1321Zhu Qiang ,838CHEMISTRY OF FROZEN SOILSDeng Yousheng ,I 31Ershov E.D. ,160Melnikov Vladimir ,455Osterkamp T.E. ,500Ostroumov V.E. ,506CIVIL ENGINEERMGAksenov V.K. ,IGoncharov Ju.M. ,875Huneault P.A. ,286Jiang Hongju ,1152Kutasov I.M. ,362Men Zhaohe ,I 193Sui Tieling ,592Tang Shuchun ,602Tong Changjiang ,622Toxgashov Y.Y. ,941Ulrich Roland ,625Vlasov Vladimir P. ,951Vyalov S.S. ,955CLIMATE CHANGECarter L+David ,78Nakayama T. ,488Pan Anding ,1202Schmitt Elisabeth ,544Zhang T. ,783Zhao Xiufeng ,809DISASTER AND ENVIRONMENTPROTECTIONForbes Bruce C. ,176Huang Yizhi ,282Jorgcnson M.Torre ,316Kamensky R.M. ,322Makarov Vladimir N.,911Olovin Boris A. ,923Su Shengkui ,1235Wang Yingxue ,1269Zhang Qibin ,778Zheng Qipu ,I 326Zhou Youwu ,819ECOLOGYDai Chuntian ,I 16Gilichinsby David A., 869Liu Qingren ,407Mcltzer Liya I. ,914Zhao Yutian ,813EQUIPMENTCollins Charles M. ,1076Cui Guangxin ,1079Fortier Richard ,182Pu Yibin , I208Tao Zhaoxiang ,608Xia Zhiying ,1278FROST HEAVINGChen Ruijie ,1064Chen Xiangsheng ,1070Chen Xiaobai ,1073Dai Huimin ,120Ding Yongqin ,143Fowler A.C. ,1100Grechishchev S. E. ,198Guo Dianxiang,] 108Jiang Hongju ,302Lewis G.C. ,1044Liu Shifang ,I 183Michalowski R.L. ,465Pissart A. ,523Shen Mu ,550Svec Otto J. ,596Wang Shirong ,1262Xie Yingqi ,1282Xu Bomeng ,1292Yue Hanscn ,I 312Zhou Deyuan,] 330GENERALBarry R.G. ,23Chen Xiaobai ,1037Cheng Guodong ,1010Gavrilova Maria K. ,1006Haeberli Wilfried ,1014Harris Stuart A. ,1019Heginbottom J.A. ,I 132King Lorenz ,1022Nelson F.E. ,987Qiu Guoqing , I028Van Everdingen R.0.,638Vinson Ted S. ,103 IGEOPHYSICAL PROSPECTINGSedov B.M. ,1222Snegirev A.M. ,934Zeng Zhonggong ,758Zou Xinqing ,I 342Lu Xingliang ,416HEAT-MASS TRANSFERGao*Xingwang ,188Hallet B. ,226Hoelzle Martin ,272Sheng Yu ,556Wang Yi ,1265Xia Zhao-Jun ,755Xu Xiaozu ,734Yen Yin-Chao ,750Zhang Lixin ,773HYDROLOGY ANDSOURCEBartoszewski S. A. ,32Brewer M.C. ,48Burchak T.V. ,1344Hinzman Larry D. ,267Kane D.L. ,326Marsh Philip ,443Woo Ming-Ko ,712Yang Daqing ,738WATER RE-1351

Yang Zhenniang ,744Yuan Haiyi ,1308MMINGDubina Mikhail M..,862Vakili Jalal ,1247Wang Changsheng ,I 251Wang Jianping ,678I’ERIGLACIAL PHENOMENALiu Gtnnian ,397Allard Michcl,l344Barsch Dietrich ,27Bondarcnko G.I. ,851Burn C.R. ,60Caldcroni G. ,72Cui Zhijiu ,1086Cui Zhi ju ,111Guo Dongxin ,210Guo Xudong ,I 1 13Harris Charles ,232Harris Stuart A. ,238Kritsuk L.N. ,897Kunitsky Viktor V. ,903Leibman M.O. ,380Lewkowicz Antoni G.,1345Li Shudc ,1174Luo Minru ,426Matsuoka Norikazu ,449Murton Julian €3. ,482Shpolyanskaya N.A. ,930Slaughter Charles W.,574Solomatin V.I. ,937So0 Sweanum ,586Tong Boliang ,6 I7Trombotto Dario ,1238Vonder Muhll D.S. ,654Vtyurina E.A. ,660Wang Baolai ,664Wang Chunhc ,670Wang Zhenyi ,1272Wayne William J. ,694Zhou Zhongmin ,I 334Zhu Cheng ,826Zhu Jinghu ,832PHYSICS OF FROZEN soxsChcn Ruijie ,1067Chuvilin E.M. ,89Corapcioglu M.Y. ,100Ding Jingkang, 1092Fcdiukin Igor V. ,I 70Pedosccva V.I. ,865Fci Xueliang ,1096Gorelik Yakov ,879Guryanov Igor E. ,885Haoulani H. ,I 124He Ping ,250Huang Maohuan ,278Janoo Vincent ,292Leclaire P. ,368Li Anguo ,1159Li Dongqing ,I 164Li Kun ,1171Ma Wci ,432Miao Tiande ,I 197Roman L.T. ,1219Shesternyov D.M. ,1227Slcttcn Ronald S. ,580Thimus J.F. ,611Wang Jiacheng ,675Wang Jiacheng ,1255White T. L. ,700Wipweera Harsha ,706Wu Ziwang ,722Wu Ziwang ,1274Xu Xiaozu ,1295Yang Lifeng ,1301Yu Qihao ,1304Zhang Huyuan ,768Zhang Zhaoxiang ,797Zhu Linnan ,835Akagawa Satoshi ,1050Haiying FU ,1117PIPELINEAntonov-Druzhinin V.,1054Biggar Kevin W. ,42Burgess M.M. ,54Cui Jianheng ,1082Ding Jingkang ,138Domaschuk L. ,149Huang Junheng ,1148Liu Hongxu ,403Moblcy Keith F. ,471Molmann This, 477Nixon J. F. ,494Zhang Jinzhao,l317Zhang Xin ,789REGIONAL GEOCRYOLOGYAllard Michel , 5An Viktor ,843Arc Felix E. ,846Aziz A., 17Baulin V.V., 1060Belloni S. ,36Chehovsky A.L., 1062Chcn XiaoBai ,84Collctt Timothy S., 94Dallimore S.R. , 125Danilov Igor D. ,858Devjatkin V.N. , 134Gorbunov A.P. ,1105Gray James T. ,192Gu Zhongwci ,204Haeberli Wilfried , 214Hall Kevin J. ,220RanWk,1119Hazcn Beez ,244Weginbottom J.Alan ,255Hcnrik Mai ,I 137Hinkel K. M. ,261Hu Ruji, 1144Jin Huijun ,307Kellcr Felix , 332Kershaw G. Peter, 338King Lorenz ,344Klimovsky Igor V. ,891Klimowicz Zbigniew, 350Kurfurst P.J. , 356Lehmann Rainer ,374Li Guangpan , 1 I67Li Zuofu ,1178Liang Linheng ,393Lomborinchen R. ,411Lunardini Virgil J.,420Luo Guowci ,I 186Ma Hong ,429Mamzelev A.P. ,436Migala Krzysztof ,919Pavlov A.V. ,51 1Ping C. L. ,517Qiu Guoqing ,533Ramos M. ,1211Salnikov P.I. ,927Sharkhuu A. ,1223Shur Y.L. ,564Siegert Christine ,569Sone Toshio ,1231Tumurbaatar D. ,1242Urdea Petru ,631Vandenberghe J. ,643Vasil’chuk Yurij K.,945Walker H.Jesse ,1345Walters James C. ,1346Wang Shaoling ,1259Wang Yaqing ,689Wu Qingbai ,7 18Xiong Heigang , I287Yu Shcngqing ,755Zhang Zcyou ,793Zhao Lin ,803REMOTE SENSING AND MAP-PINGBurns R.A., 66Haugen Richard K. , 1128Liang Pcngxian ,388

ROADSAn Weidong , I1Cui Jianhtng , 105Dydyshko P.I. , 155Esch David C. ,164Kurilchik A.F. ,909Luo Weiquan , I 190Mi Haizhcn ,461Saarelainen Seppo , 539Shoop Sally A. ,559Tang Xiaobo ,605Vinson Ted S., 648Yang Hairong ,1298Zhu Yunbing , 1338

LIST OF PARTICIPANTS IN VI ICOPAguirre-Puente, J.Akagawa SatoshiAkerman Jonas H. &Ms. Akerman**Allard, Michel**An WeidongAnisimova N.P.Antonov-Druzhinin V.Are Felix E.Aziz A.Balobayev V.T.Barry, Roger G.Saulin V.V.qigga,r K.Y. & NilBradley, P.G.Rrewer Y.C.**Brown Jerry RCella Brown"Cames-Pintaux,Anne-MarieCarlson Robert &Camille Q . CsrlsonCarter, L. David*Chang R .V.'Chen FengfengChen Ruijie+Chen 'XiaobaiCheng Guodong"Cheng YouchangClark, Michael G.Clarke Edwin S. RAlta ClarkeCollett, Timothy S.Corapcioglu M. Yavuz &Carol Y. WongCui GuangxinCui JianhengCui ZhijiuDai RaoguoDai HuiminDallimore ScottFranceJapanSwedenCanadaCanadaRussiaRussiaRussiaUSARussiaUSARGssiaCanadaUSAUSAUSAFranceUSAUSARussiaChinaChinaChinaChinaChinaUSAUSAUSAUSAChinaChinaChinaChi,naChinaCanadaLaboratoire d'Aerothermique du C.N.R.S., Yeudon3-4-17 Ftchujima, 3-Chome Koto-Ku, Tokyo 135Department of Physical Geography IJniversity of Lund,Solvevgatan 13 S-23362 LundUniversity Laval, Ce'ntre D'Etudes Nordiques, Sainte-Foy,Quebec GlK 7P4Centre D'Etudes Nordiques, Universite Laval, Quebec G1K 7P4Permafrost Institute, Siberian Branch, Russia Academy ofSciences, 677010, Yakutus, 10INGEOTES, 626718, Nory Urengoy, 'fyrnen Reg.Petersburg Institute of Railway Engineers, Moskovskp av., 9.St. Petersburg 190331Department of Mechanical Engineerrng Gonzaga University,Spokane, WA 99258Permafrost Institute, Siberian Branch, Russia Academy ofSciences, 677010, Yakutus. 10Ilniversity of Colorado, CLRES/NSIDC, Boulder Co. 80309-0449PNIIIS, Okruinol PR. 18. 105058 GSP. MOSCOWDepartment of Civil Fngineering, Royal Military College,Kingston, K7K SLORox 194900, Dept. of Transportation and Public Facilities,Pouch 6900, Anchorage, Ak 99510U.S. Geological Survey, 4200 University Drive. Anchorage,Alaska, 99508-466<strong>International</strong> permafrost Association Editorial Committee,P.O. Box 9200, Arlington, Virginia 22219-0200C.N.R.S., Lab. d'Aerothermique, 4 , Route Gardes, 92190MeudonInstitute of Water Resources, Ilniversity of Alaska,Fairbanks, AK 99701U.S. Geological Survey, 4200 University Drive. Anchorage,Alaska 99508-4667Permafrost Institute, Russian Academy of Sciences,Yakutsk 677018State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730600Lanzhou Institute of Glaciology and Geocryology. ChineseAcademy of Sciences, Lanzhou 730000Lanzhou Institute of Glaciology and Geocryology, ChineseAcademy of Sciences, Lanzhou 730330Daqing Oil Field Design Academy. 555469306 Geology R Geography Ruilding, Knpxville, TENN 37995-1410Clarke Engineering Company, 1818 So. University Avenue,Suite 9-Fairbanks. Alaska 99709U.S. Geological Survey, Sox 2504-6, MS-940, Denver FederalCenter, Lakewood, Colorado 83225Department of Civil Engineering, Texas A&M University,Collage Station, TX 77843-3136Mining and Technology Ilniversity, Xuzhou 221008The First Survey and design Instituke of Highway, TheMinistry of Communications, Xian 710068Repartrnent of Geography, Beijing Ilniversity, Reijing 100871Hei1ongj;ang Institute of Forestry, Yeilongjiang Province,150040Heilongjiang Institute of Highway and Transport, 40 qingbinRoad, Harbin 150080Geological Survey of Canada 601 Booth St, Ottawa, K1A DE8'Participants in the field trip to Lhasa (A-1)**Participints-in the fiield trip to Tianshan Mountain (A-2)

Deng NanDeng YoushengDevjatkin V.N.Ding JingkangDramis FrancescoDubikov G.1,Dubina M.M.Esch DavidFedorov A.N.Fedoseeva V.I.Fei XueliangFerrians, Osca~Flaate, Kaare &Astrid FlaateFortier R. &Brigitte Dufour"French Hugh M.*Gnmper Barbara"Gao 'deiyucCao XingwangGarty, Jacoh kGarty Rachel"Gavrilova M.K,Ge YuanyouGe QihuaEilichinsky D.A.Gong WangshengGorbunov A,P.Grechishchev S.E.Gryc GeorgeGu ZhongweiGuglielmin MauroGuo DianxiangGuo TingbinGuo XudongGuryanov I.E.Haeberli W.*Hall Kevin RMrs. A.J. HallHallet, Sernard"Han UK**Harris Charles**Harris S.,4.*&P.R. HarrisHe PingHeginbottom J.A.ChinaChinaRussiaChinaItalyRussiaRussiaUSARussiaRussiaChinaUSANorwayCanadaCanadaSwitzerlandChinaChinaIsraelRussia 'ChinaChinaRussiaChinaRussiaRussiaUSAChinaItalyChinaChinaChinaRussiaSwitzerlandSouth AfricaIJSASouth KoreaIJ KCanadaChinaCanadaChinese Academy of Sciences, Beijing 100864State Key Laboratory of Frozen Soil Engineefing, LIGG.CAS, 730000Institute of Cryosphere of Earth of Russia Academy ofSciences p,b. 1230 Tyumen 625000, Russia FederationNorthwest Institute, Chinese Academy of Railway Sciences,LanzhouDepartment of Geology, Universita di Camerino, VialeReeti 1, 62032 CamerinoDepartment of Eeocryology, Faculty of Geology, MoscowState University, ,Moscow 119899Permafrost Institute, Russian Academy of Sciences,Yakutsk 677018Alaska Dept. of Transportation, Research Engineer, Alaska2301 Peger Road,Permafrost Institute, Russian Academy of Sciences, Yakutsk677018Permafrost Institute, Russian Academy of Sciences,Yakutsk 677018Xian Highway and Transportation University, XianU.S. Geological Survey, 4200 University drive, Anchorage,Alaska 99508Norwegian Road Research Laboratory, P.O. Box 6390 Ettertad0604 Oslo 6Centre d'etudes Nordiques, Pavillon F.A. Savard,Universite Lavel, Sainte-Foy, Quebec G1K 7P4Dept. of Geography, University of Ottawa, Ottawa,Ontario K1N 6N5Eawag, Ste. 8600 Dubendor!Water Conservancy qesearch Institute of Inner Mongolia,010020Lanzhou Institute of Glaciology R Geocryology, ChineseAcademy of Sciences, Lanzhou 730000Tel-Aviv University, Department of Botany, Tel-Aviv 69978Permafrost Institute, Siberian Branch, Russian Academy ofSciences, 671010, Yakutus 10Water Resources Rureau of Bayan County, Yeilongjiang 151800The First Survey and design Institute of Highways, TheMinistry of Communication, Xian 710068Institute of Soil Science and Photosynthesis of TheRussian Academy of Sciences,Pushchino 142292Chinese Academy of Sciences, Beijing 100864Permafrost Institute, Siberian Branch, Russian Academy ofSciences, 677010, Yakutus 10All-Union Research Inst. of Hydrogeology R Eng. Geology,142452 Zeleny-village, Noginsk District, Moscow RegionU.S. Geological Survey, 345 Middlefield Road, Menlo Park,CA 94025Lanzhou Institute of Glaciology and Geocryology, ChineseAcademy of Sciences, Lanzhou 730000Istituto di Geologia, Universita degli Studi Di Parma43100, ParmaInstitute of Hydraulics, Shandong Province, 250013The National Natural Science Foundation of China,Academician, Beijing 100083Institute of Geologic, CAS, Beijing 634 Box, 100029Permafrost Institute, Siberian Branch, Russian Academy ofSciences, 677010, Yakutus 10Laboratory of Hydraulics, Hydrology and Glaciology,VA'd-ETH Zentrum CY-8092, ZurichDepartment of geography, University of Natal, P.O. Box 375,3200 PietermaritzburgUniversity of Washington, Q.uaternary Research Center AK-60Seattle, WA 98195nept. of 2nv. Sci., Korea Military Academy. Seoul,Korea 139-799Department of Geology, University of Wales, P.O. Box 914,Cardiff, CF1 3YEDepartment of Geoaraohv. . Universitv of Calgary, Calgary,I. 1Alberta, T2N 1NZState Key Laboratory of Frozen Soil Engineer ing, LIGG,CAS, 730000Geological Survey of Canada, 601 Booth Stree t, Ottawa KlA OE8

Hinkel K.M."Guo Dongxin'Huneau~, Paul A.Hinzrnan, Larry D.1Ioelzle M.'t*Horiguchi KaoruNuang HaohuanHuang Yizhi"Hu Qiheng1ordanesc.u MirceaIstomin Vladmir A.Jiao Tianhao RLong Jiyun*Jia Jiarrhua.I i n Hui juri'$'(Jin NaichuiKurnensky Ros~eslav M.Kline, Jlouglas L."Keller F.*King Lorenz J."Klirnovsky I.V.Koster u,.A.ikqeKrantz William 5. RJune KrantzKunitskiy V.V.Yurfurst P..J. Rllana Yurfurst.Lachenhruch, Arthur REdith 5. J,achenhruchLangager, Hans ChristianLautridou Jean PLfhmann, Rainer"9FLewkowicz A.C.Li AngucLi DongqingLi HaoLi Shijie**1.i ShurleLi Yi1.i YushengJ.i ZuofuLi WeiguoLiang FengxianLi nYipuUSAChinaCanadaUSASwitzerlandJapanChinaChinaChinaCanadaRussiaChinaChinaChinaChinaRussiaUSASwitzerlandGermanyRussiaNetherlandI1 s ARussiaCanadaUSADenmarkFranc.eGermanyCanadaChinaChinaChinaChinaChinaChinaChinaChinaChinaChfnaChinaDepartment of Geograph, Mail Location 131, UniversityCincinnati, Cincinnati, Ohio 45221-0131Lanzhou Instlzute of Glaciology & Geocr.yology, ChineseAcademy of Sciences, Lanzhou 730000500 Boul Bene-Lbvesque Ouest, Place Air Canada, Bareau 600Montreal (Qubbec) H2Z 1W7University of Alaska, Water Research Center, Fairbanks,Alaska 99775-1760Lahoratory of Hydraulics, Hydrology R Glaciology, VAW-ETHZentrum, CH-8092 ZurichThe Institute of Low Temperature Science, HokkaidoIlniversi ty, .Sapporo Oh0State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000Lanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 730000Chinese Academy of Sciences, Reijing 1008641800 Yonte Saiate-Julie, Varennes, P. Quebec J3X IS1Russia, 142717, Moscow Segion, Leninskij Qajon, Razvieka,VNIIGAZ !Yitulihe Branch of Harbin railway Bureau, Heilongjiang 022168The First Survey Design Institute, Lanzhou Branch, Ministryof Railway of China, Lanzhou 730000Lanzhou Institute of Glaciology & Eeocryology, ChineseAcademy of Sciences, Lanzhou 730000Heilongjiang Hydraulic Research Institute, Harbin 150080Permafrost Institute, Russian Academy of Sciences,Yakutsk 677018University of Alaska, Fairbanks, Water Research Center,Fairbanks, Alaska 99775-1760Laboratory of Hydraulics, Hydrology and Glaciology, ETHZentrum, CH-8092, ZurichGeographical Institute, Justus Liebig-Universitat, Dh300GiessmnPermafrost Institute, Siberian branch, Russian Academy ofSciences, 677010, Yakutus 10Geographical Institute, University of Utrecht P.O. Box 80.115, 3508TC UtrechtUniversity of Colorado, Dept. of Chenical Engineering,Campus Sox 424, Roulder, Colorado 80309-0424Permafrost Institute, Siberian Branch, Russian Academy ofSciences, 677010, Yakutus 10Geological Survey of Canada, 601 Booth St. Ottawa,Ontario YlA 088345 Middelfield Road, MS/Y23, Menlo Park, California 94025Greenland Field investigations, Greenland Home Rule Agency16 Rosenvrngets All6 DK-2100 Copenhagen (1,Centre ne Geomorphologie, Ruedes Tilleuls 14000 CaenGeographisches Institut, Universitat Heidelberg ImNeuenheimer Feld 348 n/W-6900 HeidelbergUniversity of Toronto, Oept. of Geography, Erindale College,Mississauga, Ontario L5L 1C6Northwest Hydrotechnical Science Research Institute,Yangling Town 712100State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000Ningxia Hydrotechnical Science Research Institute, Yinchuan750001Lanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 730000Lanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 730000?lining and Technology University, Xuzhou 221008Dong nan San Tiao, No.14, Seijing 100005Lanzhou Institute of Glaciology R Geocryology, ChineseAcademy of Sciences, Lanzhou 730000Heilongjiang Institute of Forestry, Seilongjiang Province,150040Lanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 730000Bureau of Zhalainoer Coal Mine and Institution ofPaleontology and Paleanthropology of Academy Sinica,021412

Liu Gengnian .Liu HongxuLiu XuekuiLovell, Charles W. &Marry TottenLu Guoweitu XingliangLunardini Virgil’Luo WeiquanMa ChunlinMa HongMa WeiMa YijunMa Zhixue*Mackay. J. RossMakarov V. I.Matsuoka NorikazuMatthias Jakob**Melnikov E.S.Men ZhaoheMi HaizhengNa YunlongNakayama TomokoNazarenko Alexander A.& Nazarenko LydiaNelson, Frederick RMargaret WildeXNidowicz Sernard RMs. Carol TrahimOhata TetsuoOsterkamp T.R. RJ.M. OsterkampOutcalt Samuel I.*+Pang WeizhenPerlstein G.Z.Pkwb Troy L. &Mrs, Pew6Phukan Arvind P.E.Ping C .L.Pissar t, AlbertProkop ieva L.V.Pu Yib inQiao DianshiQiu Guoqing**Romanovskj N. N,Rooney, James W. &FlorenceSalnikov P.I.Schmitt EliaabethwChinaChinaChinaUSAChinaChinaUSAChinaChinaChinaChinaChinaChinaCanadaRussiaJapanCanadaRussiaChinaChinaChinaJapanRussiaUSAUSAJapanlJSAIJSAChinaqussiaIJSAUSAUSABelgiumRussiaChinaChinaChinaRussiaUSARussiaGermanyDepartment of Geography, Reijing University, Reijing 100871Heilongjlang Province Low Temperature Construction ScienceResearch Institute; Harbin 150080Institute of Forestry nesign, Heilongjiang Province, 150080Purdue University, School of Civil Engineering, WestLafayette, IN 47907Da Hingganling Institute of Forestation, Inner Mongo 1 ia022150Bei An Road 74, Changchun 130061CRREL, 72 Lyme Road, Hanover NH 03755Design, House of Management Forest Bureau, Eer Guna 1, i ft,Inn,er Mongolia 022363Heilongjiang Institute of Forestry, Heilongjiang Pro V ince150040Xinjiang Institute of Geography, Chinese Academy ofSciences. Ilrumai 830011State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000Institute of Hydraulics, Shandong Province 250013Lanzhou Institute of Glaciology $i Geocryology, ChineseAcademy of Sciences, Lanzhou 730000University of aritish Columbia, 217-1984 Yest Mall,Vancouver, S,C. V6T 1W5Permafrost Station, Permafrost Institute, Rolishoi TeatrStreet, Build. 9, Krasnoyarckiy Krai, Iyarka, 663200Institute of Geoscience, University of Tsukuba, Ibaraki 305Department of geography, University of-.british Columbia.Vancouver, B.C., V6T 1‘45All-Union Research Tnst. of Hydrogeology & Eng. Geology.142452, Zeleny-village, Noginsk Distric-t, Moscow RegionAmuer Design House Forestry Rureau, Management Rureau ofDa Hingganling, 165302, YeilongjiangLanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 730000Da Hingganlirrg.1nstitute of Prospecting and Design,Ministry of Forestry, 165000Institute of Low Temperature Science, Hokkaido Univ.,Sapporo2 Petrousky Streat Yakutsk 677891Rutgers University, Dept. of Geography, Box 5080/KilmerNew Brunswick, New Jersey 08903601 East 57th Place, Anchorage, Alaska;fairbnks, Alaska AK99517Institute for Hydrospheric-Atmospheric Sciences,Nagoya University, Chikusa-Ku. Nagoya 464-01Geophysical Institute University of Alaska 995172466 Trenton Ct Ann Arbor, M, 48105Consultant (Formerly Central Coal Mining Research Institute),Beijing 103013Permafrost Institute, Siberian Branch, Russian Academy ofSciences, 677010, Yakutus 10Arizona State University, nepartment of Geology, Tempe,AZ 85287-14042281 Foxhall Dr Anchorage AK 99504AFES-SALRM 533E. Fireweed St. Palmer, Alaska 99645Universitb de LiAge, Geomorphologie et GPologie duQuaternaire, 7, Place du 20 Aout, 4000 l,i&gePermafrost Institute, Siberian Sranch, Russian Academy ofSciences, 617010, Yakutus 10State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000Suihua Area Hydraulic Bureau, Heilongjiang 152054Lanzhou Institute of Glaciology R Geocryology, ChineseAcademy of Sciences, Lanzhou 730000nepartment of Geocryology, Faculty of Geology, MoscowState University, Moscow 119899Rom Consoltants, Inc. 9101 Vorgvard, Anchorage AK 99507Chita Department of Permafrost Institute of SiberlanDivision of Russian Academy of SciencesGeographisches Institut Di Univ. Giessen Senckenbergstrasse1, 6300 Giessen

Senneset Kaare 8Inger SofieShao LijunSharkhuu N.Shatz M.M.Shen MuShen Yongping**Sheng YuShi YafengShoop, Sally A. &Clayton MorlockSiegert ChristineSlaughter Charles W.**Sletten, Ronald S.*Snegiryev A.M.Solomatin V.I.Sone Toshio**So0 SweanumSteensboe Jorgen S.Sun JinyueTang ShuchunTao ZhaoxiangThomsen ThorkiloTilley Philip**Tong BoliangTong ChangjiangTremblay ClementTu GuangzhiUrdea Petru I.Uziak-t StanislawVan Everdingen Robert 0.Vandenberghe. Jef**Velikin S.A. ,Vinson, Ted. S. &Suzanne VinsonVonder Muhll D.S.**Vyalov Segrey S.Walker H. Jesse**Walters James C.Wang Baolai* &Jian QicenYang Changsheng"Wang GuoshangWang Huan*Wang JiashengWang JianpingNorwayChinaMongoliaRussiaCanadaChinaChinaChinaUSAGermanyUSAUSARussiaRussiaJapanUSADenmarkChinaChinaChinaDenmarkAustraliaChinaChinaCanadaChinaRomaniaPolandCanadaNetherlandRussiaUSASwitzerlandRussiaUSAUSACanadaChinaChinaChinaChinaChinaThe Norwegian Inst. of Technology, Geotechnical Division,N-7034 TrondheimSi Dalin Road 140, Changchun 130012Institute of Geography & Geocryology, Mongolian Academy ofSciences,Ulanbartu 21620Permafrost Institute, Siberian Branch, Russia Academy ofSciences, 677010, Yakutus 10Genie Civil, Universite Lavel. Ste-Foy (Quebec), QuebecG1K 7P4Lanzhou Institute of Glaciology 5 Geocryology, ChineseAcademy of Sciences, Lanzhou 730000State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000Lanzhou Institute of Glaciology i% Geocryology, ChineseAcademy of Sciences, Lanzhou 730000CRREL 73 Lyme RD Hanover, NH 03755-1290Am Wehr 7, 0-8801 BertsdorfLand/Water Interactions Pesearch Program Pacific NorthwestResearch Station, USDA Forest Service, 308 Tanana Drive,Fairbanks, Alaska 99775University of Washington, Civil Engineering. Mail StopFX-10, Seattle, WA 98195Permafrost Institute, Siberian Branch, Rusrpian Academy ofSciences, 677010, Yakutus 10Faculty of Geography, Moscow State University, Moscow 119899Hokkaido University, Institute of Low Temperature Science,Sapporo Oh0Department of Civil Engineering and Construction, BradleyUniversity, Peoria, IllinoisRosenvengets Alle 16-DK 2100 CopenhacenInstitute of Crop Germplasm Resources, Chinese Academy ofArgriculture Sciencea. Beijing 100081Da Qing Oil Field Construction Design Research Institute,163712State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000Pilestraede 52 P.O. 2128 OK 1015 Copenhagen KDeot. of Geography. University of Sydney, NSW 2006tor 10 QueGn ST-Mittagong NSW Australia 2575)Lanzhou Institute of Glaciology & Geocryology. Ch ineseAcademy 'of Sciences, Lanzhou 730000Lanzhou Institute of Glaciology R Geocryology. Ch ineseAcademy of Sciences, Lanzhou 730000Ministere des Transports du 9.. 700, boul. St-cyr illeest, 30e etage, Quebec G1R SH1Chinese Academv of Sciences, Beiiina 100864"University of Timisoara, V. Parvan 4, 1903 Timisoara19 Akadomicka. 20-033 LublinThe Arctic Institute of North America, the University ofCalgary, Calsary, Alberta, T2N 1N4Inst. of Earth Sciences, De Boelelaan 1081 HV AmsterdamPermafrost Institute, Russian Academy of Sciences. Yakutsk677018Oregon State University, Dept. of Civil Engineering.Corvallis, Oregon 97331 'Versuchsanstalt fbr Wasserbau, Hydrologie und GlaziologieETH-Zentrum, CH-8092. ZurichScientific and Researth Bureau "Geotechnique" PodsosenskgPer., 25, LO3062 MoscowDept. of Geography, Louisians State University, BatonRouge, LA 70803Dept. of Earth Science, University of Northern Iowa,Cedar Falls, IA 50hlLCentre d'etudes nordiques, Universite Laval, St-Foy,Quebec G1K 7P4Central Coal Mining Research Institute, Reijing 100013Lanzhou Institute of Glaciology & Eeocryology, ChineseAcademy of Sciences, Lanzhou 730000Beijing <strong>International</strong> Convention Center, BeijingState Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730300Central Coal Mining Research Institute, Beijing 1060131358

Wang ShangliWang ShirongWang XinglongWang YaqingWang Yi**Wang ZhanchenWang ZhengyiWei XuexiaWoo Ming KOWu JingminWu QingbaiWu ZiwangXia ZhaojunXia ZhiyingYXie YanXie YingqiXue ShiyingXu Zhi HongXu ShaoxinXu XiaozuYang HairongYang JinghuiYang ZhenniangYen Y.C. 'Young K.L. &Mar8are.t M.Yu Qihaoyu Xiang and Wrs. YuZhang ChengqingZhang JiazhenZhang JieZhang Jin'zhaoZhang LixinZhang QibinZhang ZhaoZhang ZhaoxiangZhang HengxuanZhao Lin**Zhao XiufengXZhao YutianChinaChinaChinaChinaChinaChinaChinaChinaCanadaChinaChinaChinaCanadaChinaChinaChinaChinaChinaChinaChinaChinaChinaChinaUSACanadaChinaChinaChinaChinaChinaChinaChinaChinaChinaChinaChinaChinaChinaChina'Lsnzhou Institute of Glaciology & Geocryology, ChineseAcademy of Scieoces. Lanzhou 730000Water Conservancy. Bureau.of Panjin City, Liao Nin Province,124010Heilongjiang Institute of Highway and Transport, 40 QingbinRoad, Harbin 150080Lanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 730000Water Resources Research Institute of Inner Mongolia Huhhox,Inner Mongolia 010021Bei An Road 74, Changchun 130061Bureau of 'Zhalainoer Coal Mine and Institution ofPeleontology and Paleanthropology of Academy Sinica,021412Department of Mechanics, Lanzhou University. 730000Geography Department, McMaster University, Hamilton,Ontario L8S 4K1The First Survey and Design Institute of Highways, TheMinistry of Communication, Xian 710068Lanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 730000State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000Department of Geography, McMaster University, Hamilton,Ontario L8S 4K1Lanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 730300Lanzhou Institute of Flacciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 730000Hei Longjiang Provincial Research Institute of WaterConservancy, 150080Chinese Academy of Sciences, Beijing 100864Chinese Academy of Sciences, Beijing 100864Heilongjiang Provincial Research Institute of WaterConservancy, 150080State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000Northwestern Institute, Railway Ministry Academia,Lanzhou 730300Water Conservancy nesigning Party, Shuangliao County JilinProvince, 136400Lsnzhou Institute of Glaciology & Geocryology, Chinese -Academy of Sciences, Lanzhou 730000CRREL, Hanover, New Hampshire 03755Department of Geography, McMaster University, Hamilton,Ontario, L8S 4K1State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730300Central coal Mining Research Institute, He Pingli,Beijing 100013State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000Chinese Geography Association, An Ding Men Wai, Da TunRoad No.917, Beijing 100101Research Society for Chinese Development of Cold Region,150010The First Survey and design Institute of Highways. TheMinistry*of Communications, Xian 710068State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000Lanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 7300C3The First Survey Design Institute, Ministry of Railwayof China, Lanzhou 730000Beijing Argricultural Engineering University, Reijing 100083Chinese Development of Cold Region Research Society, DaoliUistrict, Zhongyi Street 34, Harbin 150010Lanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 730000' Lanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, Lanzhou 730000Institute of Crop Germplasm Resources, Chinese Academy ofArgriculture Sciences, Beijing 100081'1359 -

Zheng DuoZheng DuoZhou DeyuanZhou YouwuZhou ZhongminZhu JinghuZhu LinnanZhu Qiang2hu YuanlinChinaChinaChinaChinaChinaChinaChinaChinaChinaJilin Provincial Institute of 'Water Conservancy Sciences,Changchun 130022Beijing Institute of Geography, Chinese Academ Y ofSciences, Beijing 100101Administration Bureau of Hetao Irrigation Area in Yongji,Inner Mongolia 015000Lanzhou Institute of Glaciology & Geocryology, ChineseAcademy of Sciences, .Lanzhou 730000Changsha Hydraulic Electric Power Teachers' CO 1 lege,Changsha 410077Department of Geography, Harbin Normal Univers i tY 9Harbin 150080State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000Gansu Provincial Research Institute of Water Conservancy,Lanzhou 730000State Key Laboratory of Frozen Soil Engineering, LIGG,CAS, 730000