Chernobyl Nuclear Accident Defense Department Reports
Chernobyl Nuclear Accident Defense Department Reports
Chernobyl Nuclear Accident Defense Department Reports
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
CHERNOBYL NUCLEAR ACCIDENT DOCUMENTS<br />
DEPARTMENT OF DEFENSE DOCUMENTS<br />
BACM RESEARCH<br />
WWW.PAPERLESSARCHIVES.COM
About BACM Research – PaperlessArchives.com<br />
BACM Research/PaperlessArchives.com publishes documentary historical research collections.<br />
Materials cover Presidencies, Historical Figures, Historical Events, Celebrities, Organized Crime, Politics,<br />
Military Operations, Famous Crimes, Intelligence Gathering, Espionage, Civil Rights, World War I, World<br />
War II, Korean War, Vietnam War, and more.<br />
Source material from Federal Bureau of Investigation (FBI), Central Intelligence Agency (CIA), National<br />
Security Agency (NSA), <strong>Defense</strong> Intelligence Agency (DIA), Secret Service, National Security Council,<br />
<strong>Department</strong> of <strong>Defense</strong>, Joint Chiefs of Staff, <strong>Department</strong> of Justice, National Archive Records and<br />
Administration, and Presidential Libraries.<br />
http://www.paperlessarchives.com
CHERNOBYL NUCLEAR ACCIDENT DOCUMENTS<br />
http://www.paperlessarchives.com/chernobyl_nuclear_accident_doc.html
<strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant <strong>Accident</strong> CIA, <strong>Department</strong><br />
of <strong>Defense</strong>, <strong>Department</strong> of Energy, Congressional, GAO,<br />
and Foreign Press Monitoring Files<br />
4,010 pages of CIA, <strong>Department</strong> of <strong>Defense</strong>, <strong>Department</strong> of Energy,<br />
Congressional, GAO, and foreign press monitoring files related to the<br />
<strong>Chernobyl</strong> <strong>Nuclear</strong> <strong>Accident</strong>.<br />
On Sunday April 26, 1986, at the <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant near<br />
Pripyat, Ukraine, reactor #4 exploded. For the 25 years from 1986 to<br />
2011, this incident has been referred to as the world's worst nuclear<br />
power plant accident.<br />
THE ACCIDENT<br />
According to reports filed with International Atomic Energy Agency (IAEA)<br />
on April 25, 1986, technicians at the <strong>Chernobyl</strong> plant launched a poorly<br />
executed experiment to test the emergency electricity supply to one of<br />
its Soviet RBMK type design reactors. The test was meant to measure a<br />
turbogenerator's ability to provide in-house emergency power after<br />
shutting off its steam supply. During the experiment the technicians<br />
violated several rules in place for operating the reactor.<br />
During the experiment, the emergency shutdown system was turned off. The<br />
reactor was being operated with too many control rods withdrawn. These<br />
human errors, coupled with a design flaw that allowed reactor power to<br />
surge when uncontrolled steam generation began in the core, set up the<br />
conditions for the accident.<br />
A chain of events lasting 40 seconds occurred at 1:23 AM on April 26.<br />
The technicians operating the reactor put the reactor in an unstable<br />
condition, so reactor power increased rapidly when the experiment began.<br />
Subsequent analysis of the Soviet data by U.S. experts at the <strong>Department</strong><br />
of Energy, suggests the power surge may have accelerated when the<br />
operators tried an emergency shutdown of the reactor. According to Soviet<br />
data, the energy released was, for a fraction of a second, 350 times the<br />
rated capacity of the reactor. This burst of energy resulted in an<br />
instantaneous and violent surge of heat and pressure, rupturing fuel<br />
channels and releasing steam that disrupted large portions of the core.<br />
The surge destroyed the core of reactor unit four, containing<br />
approximately 200 tons of nuclear fuel. Some of the shattered core<br />
material was propelled through the roof of the reactor building. The hot<br />
core material of reactor 4 started about 30 separate fires in the unit 4<br />
reactor hall and turbine building, as well as on the roof of the<br />
adjoining unit 3. All but the main fire in the graphite moderator<br />
material still inside unit 4 were extinguished in a few hours.<br />
It was a day and a half before the people living in Pripyat were ordered<br />
to evacuate. The residents were told they would only be gone for several
days, so they left nearly everything behind. They never returned. Soviet<br />
authorities made the decision not to cancel May 1, May Day, outdoor<br />
parades in the region four days later.<br />
The graphite fire continued to burn for nearly two weeks carrying<br />
radioactivity high into the atmosphere, until it was smothered by sand,<br />
lead, dolomite, and boron dropped from helicopters. Despite the wide<br />
spread of radiation, Soviet officials at first said very little publicly<br />
about what happened at <strong>Chernobyl</strong>. It was not until alarms from radiation<br />
detectors in other countries, many hundreds of miles away, forced the<br />
Soviets to admit to the <strong>Chernobyl</strong> accident.<br />
Radioactive material was dispersed over 60,000 square miles of Ukraine,<br />
Belarus, and Russia. Smaller amounts of radioactive material were<br />
detected over Eastern and Western Europe, Scandinavia and even the United<br />
States. The accident has left some nearby towns uninhabitable to this<br />
day.<br />
Radioactivity forced Soviet officials to create a 30-kilometer-wide nohabitation<br />
zone around <strong>Chernobyl</strong>, sealing off Pripyat. Still, the power<br />
plant continued to generate electricity until it was finally shut down in<br />
December, 2000.<br />
During the first year after the accident, about 25,000 people, mainly<br />
Soviet Army troops, were dispatched to the site to clean up the accident.<br />
Thousands of workers, called liquidators, were employed during the<br />
following years of the cleanup.<br />
Around October, 1986 the construction of a 21 story high metal and<br />
concrete shelter was completed, enclosing the reactor and the radioactive<br />
material that remained. Almost 200 tonnes of melted nuclear fuel rods<br />
remain within the damaged reactor. This containment shelter was not<br />
intended to be a permanent solution for containing the radioactive<br />
material. Over time, the shelter has weakened; rain entering through<br />
holes and cracks has caused corroding.<br />
By 2006 the plans for a new shelter was about 7 years behind schedule,<br />
with a completion target date of no sooner than 2012. In February of 2011<br />
it was reported that construction of the shelter may have to be halted,<br />
due to a $1 billion dollar short fall in the funds needed to complete the<br />
structure.<br />
A United Nations report released in February 2011 estimates the disaster<br />
caused thyroid cancer in 7,000 children in the affected area. The report<br />
said despite the high rate of cancer, only 15 fatalities in these 7,000<br />
cases have occurred.<br />
THE DOCUMENTS<br />
CIA FILES
215 pages of CIA files dating from 1971 to 1991.The files cover the<br />
Soviet Union's atomic energy program; The effect of the <strong>Chernobyl</strong><br />
accident on the Soviet nuclear power program; and the social and<br />
political ramifications of the accident in the Soviet Union.<br />
A 1981 report covers the less publicized Soviet nuclear "accident" near<br />
Kyshtym in 1957-58.<br />
Media reporting of a nuclear accident near Kyshtym first appeared in<br />
1958. It was not until 1976, when the writings of Soviet dissent Dr.<br />
Zhores Medvedev began to appear, that wider attention was given to this<br />
subject. Medvedev, an exiled Soviet geneticist, claimed in several<br />
articles and books that a "disaster" occurred near Kyshtym in 1957/58. He<br />
alleged that thousands of casualties and widespread, long-term<br />
radioactive contamination occurred as the result of an explosion<br />
involving nuclear waste stored in underground shelters.<br />
The general consensus today is that a combination of events, rather than<br />
a single isolated incident at Kyshtym nuclear energy complex caused the<br />
radioactive contamination in the area. A study of the claims by Medvedev<br />
can be found in the <strong>Department</strong> of Energy section, in the 1982 report "An<br />
Analysis of the Alleged Kyshtym Disaster"<br />
U.S. GOVERNMENT FOREIGN PRESS MONITORING<br />
900 pages of foreign media monitoring reports from 1986 to 1992, produced<br />
by the U.S. government's National Technical Information Service's U.S.<br />
Joint Publication Research Service. They contain information primarily<br />
from Russian and Eastern Block news agency transmissions and broadcasts,<br />
newspapers, periodicals, television, radio and books. Materials from non-<br />
English language sources are translated into English.<br />
The reporting includes firsthand accounts of experiences during all<br />
points of the <strong>Chernobyl</strong> disaster. Topics covering the accident and its<br />
aftermath including domestic and international politics, sociological<br />
affairs, nuclear plant fire, evacuations, sealing the reactor,<br />
cleanup mobilization, health implications, and people returning to<br />
region.<br />
DEPARTMENT OF ENGERY REPORTS<br />
1,244 pages of reports dating from 1982 to 2009 produced or commissioned<br />
by the <strong>Department</strong> of Energy.<br />
The agencies and institutions contributing to these reports include Los<br />
Alamos National Laboratory, United States <strong>Nuclear</strong> Regulatory Commission,<br />
Lawrence Livermore National Laboratory, Savannah River <strong>Nuclear</strong> Solutions,<br />
Oak Ridge National Laboratory, Brookhaven National Laboratory, Argonne<br />
National Laboratory, and the Pacific Northwest Laboratory.<br />
Highlights include:
The 1986 Report of the U.S. <strong>Department</strong> of Energy's Team Analyses of the<br />
<strong>Chernobyl</strong>-4 Atomic Energy Station <strong>Accident</strong> Sequence DOE/NE-0076.<br />
The U.S. <strong>Department</strong> of Energy (DOE) formed a team of experts from Argonne<br />
National Laboratory, Brookhaven National Laboratory, Oak Ridge National<br />
Laboratory, and Pacific Northwest Laboratory. The DOE team provided the<br />
analytical support to the U.S. delegation for the August, 1986 meeting of<br />
the International Atomic Energy Agency (IAEA), and to subsequent<br />
international meetings. The DOE team analyzed the accident in detail,<br />
assessed the plausibility and completeness of the information provided by<br />
the Soviets, and performed studies relevant to understanding the<br />
accident.<br />
The 1987 report Radioactive Fallout from the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor<br />
<strong>Accident</strong><br />
The Lawrence Livermore National Laboratory performed a variety of<br />
measurements to determine the level of the radioactive fallout on the<br />
western United States. The laboratory used gamma-spectroscopy to analyze<br />
air filters from the areas around Lawrence Livermore National Laboratory<br />
in California. Filters were also analyzed from Barrow and Fairbanks,<br />
Alaska. Milk from California and imported vegetables were also analyzed<br />
for radioactivity.<br />
Other report titles include: An Analysis of the Alleged Kyshtym Disaster;<br />
Workshop on Short-term Health Effects of Reactor <strong>Accident</strong>s; Preliminary<br />
Dose Assessment of the <strong>Chernobyl</strong> <strong>Accident</strong>; Internally Deposited Fallout<br />
from the <strong>Chernobyl</strong> Reactor <strong>Accident</strong>; Report on the <strong>Accident</strong> at the<br />
<strong>Chernobyl</strong> <strong>Nuclear</strong> Power Station; Radioactive Fallout from the <strong>Chernobyl</strong><br />
<strong>Nuclear</strong> Reactor <strong>Accident</strong>; Radioactivity in Persons Exposed to Fallout<br />
from the <strong>Chernobyl</strong> Reactor <strong>Accident</strong>' Radioactive Fallout in Livermore, CA<br />
and Central Northern Alaska from the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor <strong>Accident</strong>;<br />
Projected Global Health Impacts from Severe <strong>Nuclear</strong> <strong>Accident</strong>s -<br />
Conversion of Projected Doses to Risks on a Global Scale - Experience<br />
From <strong>Chernobyl</strong> Releases; The <strong>Chernobyl</strong> <strong>Accident</strong> - Causes and<br />
Consequences; <strong>Chernobyl</strong> Lessons Learned Review of N Reactor;<br />
Reconstruction of Thyroid Doses for the Population of Belarus Following<br />
the <strong>Chernobyl</strong> <strong>Accident</strong>; The characterization and risk assessment of the<br />
Red Forest radioactive waste burial site at <strong>Chernobyl</strong> <strong>Nuclear</strong> Power<br />
Plant; Estimated Long Term Health Effects of the <strong>Chernobyl</strong> <strong>Accident</strong>; and<br />
Environmental Problems Associated With Decommissioning the <strong>Chernobyl</strong><br />
<strong>Nuclear</strong> Power Plant Cooling Pond.<br />
DEPARTMENT OF DEFENSE REPORTS<br />
816 pages of reports dating from 1990 to 2010 produced or commissioned by<br />
the <strong>Department</strong> of <strong>Defense</strong>.<br />
The reports include: <strong>Chernobyl</strong> <strong>Accident</strong> Fatalities and Causes; Biomedical<br />
Lessons from the <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant <strong>Accident</strong>; <strong>Nuclear</strong>
<strong>Accident</strong>s in the Former Soviet Union Kyshtym, Chelyabinsk and <strong>Chernobyl</strong>;<br />
Retrospective Reconstruction of Radiation Doses of <strong>Chernobyl</strong> Liquidators<br />
by Electron Paramagnetic Resonance; Neurocognitive and Physical Abilities<br />
Assessments Twelve Years After the <strong>Chernobyl</strong> <strong>Nuclear</strong> <strong>Accident</strong>; Simulating<br />
Wet Deposition of Radiocesium from the <strong>Chernobyl</strong> <strong>Accident</strong>; and Radiation<br />
Injuries After the <strong>Chernobyl</strong> <strong>Accident</strong> Management, Outcome, and Lessons<br />
Learned.<br />
GAO REPORTS<br />
184 pages of reports from the United States General Accounting Office,<br />
whose name was later changed to the Government Accountability Office. The<br />
four reports are Comparison of DOE's Hanford N-Reactor with the <strong>Chernobyl</strong><br />
Reactor (1986); <strong>Nuclear</strong> Power Safety International Measures in Response<br />
to <strong>Chernobyl</strong> <strong>Accident</strong> (1988); <strong>Nuclear</strong> Power Safety <strong>Chernobyl</strong> <strong>Accident</strong><br />
Prompted Worldwide Actions but Further Efforts Needed (1991); and<br />
Construction of the Protective Shelter for the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor<br />
Faces Schedule Delays, Potential Cost Increases, and Technical<br />
Uncertainties (2007).<br />
UNITED STATES CONGRESSIONAL HEARINGS<br />
634 pages of transcripts from three Congressional hearings: The <strong>Chernobyl</strong><br />
<strong>Accident</strong> Hearing before the Committee on Energy and Natural Resources,<br />
Ninety-ninth Congress, 2nd session on the <strong>Chernobyl</strong> accident and<br />
implications for the domestic nuclear industry, June 19, 1986; The<br />
Effects of the accident at the <strong>Chernobyl</strong> nuclear power plant hearing<br />
before the Subcommittee on <strong>Nuclear</strong> Regulation, United States Senate, One<br />
Hundred Second Congress, second session, July 22, 1992; and The legacy of<br />
<strong>Chernobyl</strong>, 1986 to 1996 and beyond hearing before the Commission on<br />
Security and Cooperation in Europe, One Hundred Fourth Congress, second<br />
session, April 23, 1996.
~TI<br />
FLE CY<br />
N<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency<br />
Alexandria, VA 22310-3398<br />
SWES% Ot<br />
DNA-TR-89-45<br />
<strong>Chernobyl</strong> <strong>Accident</strong> Fatalities and Causes<br />
A. Laupa<br />
G. H. Anno<br />
Pacific-Sierra Research Corporation<br />
12340 Santa Monica Boulevard<br />
Los Angeles, CA 90025-2587<br />
June 1990<br />
Technical Report<br />
CONTRACT No. DNA 001-87-C-0104'<br />
Approved for public release; -<br />
distribution is unlimited.<br />
___ __ ___ ___ _A<br />
___ __<br />
0'.
Destroy this report when it is no longer needed. Do not<br />
return to sender.<br />
PLEASE NOTIFY THE DEFENSE NUCLEAR AGENCY,<br />
ATTN: CSTI, 6801 TELEGRAPH ROAD, ALEXANDRIA, VA<br />
22310-3398, IF YOUR ADDRESS IS INCORRECT, IF YOU<br />
WISH IT DELETED FROM THE DISTRIBUTION LIST, OR<br />
IF THE ADDRESSEE IS NO LONGER EMPLOYED BY YOUR<br />
ORGANIZATION.<br />
.N 4
DISTRIBUTION LIST UPDATE<br />
I<br />
This mailer is provided to enable DNA to maintain current distribution lists for reports. We would<br />
appreciate your providing the requested information.<br />
NAME:<br />
ORGANIZATION:<br />
OLD ADDRESS<br />
CURRENT ADDRESS<br />
I TELEPHONE NUMBER: ( )<br />
z5:<br />
SUBJECT AREA(s) OF INTEREST:<br />
DNA OR OTHER GOVERNMENT CONTRACT NUMBER:<br />
CERTIFICATION OF NEED-TO-KNOW BY GOVERNMENT SPONSOR (if other than DNA):<br />
I SPONSORING ORGANIZATION:<br />
CONTRACTING OFFICER OR REPRESENTATIVE:<br />
El Add the individual listed to your distribution list. Please return OTE:-<br />
Plaertr the mailing label from<br />
SD e te cd o n the document so that any additions,<br />
El Delete the cited organization/individual, changes, corrections or deletions<br />
S Cacan be made more easily.<br />
0 E Change of address.<br />
________________<br />
SIGNATURE.-
Director<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency<br />
ATTN: TITL<br />
Washington, DC 20305-1000<br />
Director<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency<br />
ATTN: TITL<br />
Washington, DC 20305 1000
REPO NTATON T D CUM AGEForm<br />
Aproved<br />
CE R DOMB No. 0704-0188<br />
C ubhfreportin burden c toA his collection of Information Is estimated to average I hour per response Including the time for Instructions 62ev7ng searching existing data sources<br />
galhet ng and maintaining the data needed and completing and reviewing the collection of Information Send comments regarding this burden estimate of any other aspect oftlhls<br />
colleclon of Information Including suggestions for reduc:ing this burden to Washington Headquarters Servces Directorate for Information Operatlons and ReportIs 1215 Jefferson<br />
Davis Highway Suite 1204 Arlington VA 22202-4302 and to the Office of Management and Budget Papen ,ork Reduction Project (0704-01881 Washington, DC 20503<br />
1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED<br />
1 900601 Technical 870729 - 881231<br />
4 TITLE AND SUBTITLE 5. FUNDING NUMBERS<br />
C - DNA 001-87-C-0104<br />
<strong>Chernobyl</strong> <strong>Accident</strong> Fatalities and Causes PE - 62715H<br />
PR - RM<br />
6 AUTHOR(S) TA -RH<br />
A. Laupa: G. H. Anno WU - DH026130<br />
7 PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8 PERFORMING ORGANIZATION<br />
REPORT NUMBER<br />
Pacific-Sierrra Research Corporation PSR Report 1865<br />
12340 Santa Monica Boulevard<br />
Los Angeles. CA 90025-2587<br />
9 SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10 SPONSORING/MONITORING<br />
l)efense <strong>Nuclear</strong> Agency<br />
AGENCY REPORT NUMBER<br />
6801 Telegraph Road DNA-TR-89-45<br />
Alexandria. VA 22310-3398<br />
RARP/Young<br />
11 SUPPLEMENTARY NOTES<br />
This work was sponsored by the <strong>Defense</strong> <strong>Nuclear</strong> Agency under RDT&E RMC Code<br />
B4662D RM RI-I 00038 STRP 3500A 25904D.<br />
12a DISTRIBUTION/AVAILABILITY STATEMENT<br />
12b DISTRIBUTION CODE<br />
Appro\ed for public release: distribution is Unlimited.<br />
13 ABSTRACT (Maximum 200 words)<br />
-Based on aailable sources of information, an assessment is made of the Chernob\l accident<br />
fatalities and causes of death that resulted from acute injury effects. <strong>Accident</strong> victims grouped<br />
according to whole body radiation lose and biologic response were examined and reconciled<br />
based on comparing various sources of information. Fatalities are identified with the occurrence<br />
of acute radiation sickness (ARS) syndromes. A maximum likelihood regression analysis<br />
was performed on tWe available fatalit and close data based on five different statistical models<br />
to estimate the L[ i'; LDi.. and LDO values. The estimated LD 50 is about a factor or two<br />
higher than previously published values which is attributed to post accident emergency medical<br />
and clinical care. .<br />
14 SUBJECT TERMS 15 NUMBER OF PAGES<br />
- ('hernobyl <strong>Accident</strong> Radiation Injury - Bone Marrow 34<br />
Clinical Classification, Lethality Causes, Skin Damage, 16 PRICE CODE<br />
Radiation Pulmonitis_ Radiation Doses. Fatalities _ _-_ _<br />
17 SECURITY CLASSIFICATION 18 SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20 LIMITATION OF ABSTRACT<br />
OF REPORT OF THIS PAGE OF ABSTRACT<br />
UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED SAR<br />
NSN 7540-280-5500 Standard Form 298 (Rev 2-891<br />
I<br />
Prec ltw1 by ANSI qla 21PI A<br />
?.qA 102
SECURITY CL.ASSIFICATION OF THIS PAGE<br />
CLASSIFIED BY<br />
N/A ice Unclissified<br />
DECLASSIFY ON:<br />
N/A since Uniclassified<br />
14. SUBJECT TERMS (CONTINUED)<br />
Intestimal Syndromle<br />
Acute Effects<br />
Accession For<br />
NTIS C-PA&I<br />
DTIC TAB~<br />
SECURITY CLASSIFICATION OF THIS PAGE<br />
LIN('I.ASSII 11.1)
CONVERSION TABLE<br />
Conversion factors for U.S. customary<br />
to metric (SI) units of measurement<br />
To Convert From To Multiply By<br />
angstrom meters (nI I. 000 000 X F- I0<br />
atmo.phere kilo pascal (kPa) 1.013 25 X [+2<br />
bar kilo pascal (kPa I.000 000 X F+2<br />
barn meter 2 (iM 2 ) I.000 00(0 X F'-28<br />
British Thermal unit (thermochemical) joule (.1) 1.054 350 X E+3<br />
calorie (thermochemicall joule (09) 4. 184 000<br />
cal (thermochemical ). cm 2 mega Joule/m 2 (M. l 'm 2 ) 4. 184 (000 X [-2<br />
curie giga becqucrel (GBq)* 3.700 OttO X I+ I<br />
degree (angle) radian (rad) 1.745 329 X F-2<br />
degree Fahrenheit degree kelvin (K) 'rK=(I9f + 459.67) 1.8<br />
electron volt joule (.11 1.602 19 X [-19<br />
erg joule (.1) 1.0( 0( X 1[-7<br />
ergisecond watt (\V) 1.000 GOO) X [-7<br />
foot meter (m) 3.048 0009(1 X F- I<br />
fool-pound-force joule (.J) 1.355 818<br />
gallon (U.S. liquid) meter 3 (nm 3 ) 3.785 412 X F-3<br />
inch meter (t) 2.54(0 00) X [-2<br />
jerk joule (.J) 1. ( 000 X F+9<br />
joule kilogram (.1 Kg) (radiation dose<br />
absorbed) (irav ((i ' 1.000 OtO<br />
kilotons terajoules 4.183<br />
kip ((900 hlb newton (N) 4.448 222 X F+3<br />
kip'inch 2 (kqi) kilo pascal (kPa) 6.894 757 X F+3<br />
kiap newton-second/rn 2 (N-sm 2 ) I.00t) 0(00 X F+2<br />
micron meter (n) 1. ((9)0 00)9 X F-6<br />
mil meter (m) 2.540 (t9(0 X F-5<br />
mile (international) meter (m) 1.6(09 344 X [+3<br />
ounce kilogram (kg) 2.834 952 X I-2<br />
pound-force (119 avoirdupois) newton (N) 4.448 222<br />
pound-lorce Inch newton-Ometer (N.m) 1. 129 848 X F-I<br />
pound-lorce, inch nc,%, ,.'nleter (N' m 1.751 268 X [+2<br />
pound-force/foo 2 kilo pascal (kPa) 4.788 (26 X E-2<br />
pound-force/inch 2 (PSI) kilo pascal (kPa) 6.894 757<br />
pound-mass (ibn avoirdupois) kilogram (kg) 4.535 924 X F-I<br />
pound-mass-foot 2 (moment of inertia) kilogram-meter2 (kg. m 2 ) 4.214 (O1 X [-2<br />
pound-mass/foot 3 kilogram 'meer3 (kg.'in 3 ) 1.61 846 X F+ I<br />
rad (radiation dose absorbed) (;ray (Gy) I* ((100 O0t0 X F-2<br />
roenlgen coulomb..kilogram (C'kg) 2.579 7i99 X F-4<br />
shake second (q) I,000 000 X [-8<br />
qlug kilogram (kg) 1.459 390 X F+ I<br />
torr (mam I1g, 00(I kilo pascal (kPa 1 1.333 22 X F-I<br />
The tecquerel (0q) is the SI unit of radioaciviH: lip = I eventis.<br />
S1he (;ra,<br />
((;y) is the SI unit of absorbed radiation.<br />
Iil
TABLE OF CONTENTS<br />
Section<br />
Page<br />
CONVERSION TABLE .....................................<br />
iii<br />
FIGURES..............................................<br />
TABLES .................................................<br />
vi<br />
1 INTRODUCTION .......................................... 1I<br />
DATA SOURCES ON CHERNOBYL VICTIMS ...............<br />
3 CHERNOBYL FATALITIES ................................. 6<br />
4 UNCONFIRMED CHERNOBYL FATALITIES .................... 8<br />
5 CAUSES OF FATAL OUTCOMES ............................ 9<br />
6 FATALITY INCIDENCE VERSUS RADIATION DOSE............14<br />
7 LIST OF REFERENCES................................... 24
FIGURES<br />
Figures<br />
Page<br />
I<br />
<strong>Chernobyl</strong> accident fatalities: normal distribution. 238 individuals.<br />
90% confidence interval ....................................... 18<br />
2 <strong>Chernobyl</strong> accident fatalities; log-normal distribution. 238 individuals.<br />
90% confidence interval ....................................... 19<br />
3 <strong>Chernobyl</strong> accident fatalities: Weibull distribution, 238 individuals,<br />
90% confidence interval ....................................... 20<br />
4 <strong>Chernobyl</strong> accident fatalities: logistic distribution. 238 individuals.<br />
90% confidence interval ....................................... 21<br />
5 <strong>Chernobyl</strong> accident fatalities: log-logistic distribution. 238 individuals.<br />
90% confidence interval ....................................... 22
TABLES<br />
Table<br />
Page<br />
1 <strong>Chernobyl</strong> victim folloXw-up reported by State Committee<br />
119 8 6 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3<br />
2 <strong>Chernobyl</strong> victim follow-up reported by Guskova 119871 ..... 4<br />
3 <strong>Chernobyl</strong> victim follow-up reported by Fry 119871 .......... 5<br />
4 C hernobyl fatalities .................................... 7<br />
5 ARS fatalities at Hospital No. 6 in Moscow ................ 10<br />
6 ARS syndromes identified as causes of fatalities ............ 10<br />
7 <strong>Chernobyl</strong> accident fatalities and causes, Hospital No. 6 in<br />
M o scow .............................................. 12<br />
8 Radiation doses of individual patients ..................... 15<br />
9 Radiation doses of individuals and groups ................. 16<br />
10 Regression m odels ..................................... 17<br />
11 Fatality percentile doses ................................ 18<br />
vi
SECTION 1<br />
INTRODUCTION<br />
This report reviews current available Soviet syndromes, each of which could have been<br />
data sources on <strong>Chernobyl</strong> accident victims an independent cause of a fatal outcome.<br />
hospitalized for acute radiation sickness All patient groups identified in the data<br />
(ARS). Patients are grouped by the whole- sources are correlated in terms of the<br />
body ganmma radiation close indicating the severity of bone marrow syndrome and<br />
severity of bone marrow syndrome and ascribed causes of fatal outcome. Based on<br />
biologic response criteria used to place the available data, all but two of the<br />
patients in the different dose groups. We <strong>Chernobyl</strong> fatalities can be accounted for<br />
compare and reconcile the different data according to patient category, close level.<br />
sources according to the fatalities reported and lethal cause considerations.<br />
as a result of the <strong>Chernobyl</strong> accident.<br />
Included is a brief review of information on<br />
additional unconfirmed casualties attrib- A maximunm likelihood analysis of the<br />
uted to the nuclear accident.<br />
available data on the <strong>Chernobyl</strong> accident<br />
survivors and fatalities was also performed.<br />
We further review the Soviet data sources Five different statistical models were<br />
in terms of the reported causes of fatal applied to develop estimates of the incioutcomes<br />
and the occurrence of ARS dence of lethality with total body close. The<br />
synldromes identified as causes of fatalities, marked effect in survival owing to medical<br />
In many cases, fatal outcomes are ascribed attention afforded the accident victims is<br />
to the effect of several severe ARS dernonstrated.
SECTION 2<br />
DATA SOURCES ON CHERNOBYL VICTIMS<br />
There are three sources of original data oil Dr. S. A. Fry reports on presentations by<br />
<strong>Chernobyl</strong> patients with acute radiation Dr. A. Barabanova, a Soviet physician<br />
sickness:<br />
associated with Dr. A. Guskova at the<br />
Moscow Hospital No. 6. and by Dr. ). P.<br />
" USSR State Committee on the Utiliza- Orsanov. an internationally know\n biophystion<br />
of Atornic Energy. (hereinafter icist who went to <strong>Chernobyl</strong> two to three<br />
referred to as State Committee). The days after the accident to Sulper\ise<br />
<strong>Accident</strong> at tie Chernobvl <strong>Nuclear</strong> Power environmental monitoring activities.<br />
Plant And Its Consequences. August 1986<br />
" Guskova. Earlh' Acute Effects Among the<br />
Victims of the <strong>Accident</strong> at the Chernobvl The main clinical response and medical<br />
<strong>Nuclear</strong> Power Plunt, April 1987<br />
outcome data on ARS patients. as reported<br />
* Fry. Foreign Trip Report. October 1987<br />
in the three references, are summarized in<br />
Tables 1. 2. and 3.<br />
2
co I ri-<br />
CZ (U<br />
4)4<br />
00<br />
E<br />
4--<br />
"<br />
E G-<br />
C<br />
-=<br />
L~)U<br />
0.<br />
-U)<br />
0<br />
U<br />
_0C<br />
C> Cl I5<br />
cz0<br />
" C<br />
O~ ~ E c- -0<br />
C<br />
0 O di-- I0<br />
..<br />
U<br />
U<br />
C7)<br />
Z<br />
-0<br />
0 C> )<br />
CC<br />
CO-z<br />
O~a~ja.'<br />
.;<br />
U<br />
CZ 0<br />
.0 -<br />
0U<br />
w
-) C)<br />
4- o.r<br />
"7Z0<br />
o o cL; -r;<br />
Uri I C , I 1- 0() .<br />
cz<br />
-l - - I rV<br />
a<br />
"C,6 C i 'r 0 C<br />
00 -<br />
> UC.<br />
- I<br />
I-<br />
CL-<br />
ON~l<br />
ON<br />
'00<br />
E ~ ' - 0<br />
Co<br />
C z U C<br />
C> ' cu<br />
C Z<br />
0<br />
0z , 0<br />
00<br />
U E 0<br />
.0 0<br />
C3<br />
0 ON L<br />
0 - C)<br />
C<br />
o . c . U<br />
CU<br />
rM<br />
CV) 0 C<br />
0 co 0 0<br />
C_<br />
-I<br />
f -r l a u 0 v<br />
z
Table 3. <strong>Chernobyl</strong> victim follow-up reported by Fry 119871.<br />
<strong>Chernobyl</strong> Fatalities<br />
Description<br />
No. of<br />
Victims<br />
Initial (on-site and in Pripyat) 2<br />
Died as result of ARS<br />
In Moscow 27<br />
In Kiev 1<br />
Died as result of CVA 1<br />
Total deaths 31<br />
Radiation-induced Skin Injuries. Moscow Hospital No. 6<br />
Description<br />
No. of<br />
Victins<br />
Number of ARS patients 115<br />
Radiation-inducel skin injuries 56"<br />
Skin injuries incompatible with life 19<br />
Waves of Erythema<br />
Wave Onset Remarks<br />
1st 36 h No obvious skin injury for 36 h.<br />
First wave lasted up to 24 h.<br />
2d 5-10 days Erythema more widespread.<br />
later<br />
Involved skin areas covered by<br />
clothing. Coincided with latent<br />
period of ARS.<br />
3d 2-3 months Included 28-30 patients.<br />
8-10 had no previous manifestation<br />
of skin injury.<br />
"T\\o of 56 patients also had thermal burns.<br />
5
SECTION 3<br />
CHERNOBYL FATALITIES<br />
The number of fatalities from the Cherno- Guskova did not include the CVA fatality<br />
byl accident totals 31. Table 4 shows the among the fatalities reported. Only one<br />
reported nLumbers of death and reconciles fatality is listed for Kiev in Table 2. whereas<br />
the different data sources. There were two two fatalities are listed in Table 1.<br />
initial fatalities: one victim died at his<br />
workplace. his body was never recovered:<br />
another victim died from severe burns at<br />
6:00 a.m. in the morning of the accident in<br />
Dr. Barabanova further reported that the<br />
CVA victim was among the fatalities at<br />
the Pripyat hospital. State Committee Hospital No. 6 in Moscow lFri. 19871. This<br />
119861 reports 28fatalities as occurring piece of information appears to be in error.<br />
within the first 50 days (Table 1). Three State Committee 119861 reports that two<br />
people subsequently died from ARS on patients died in Kiev on days four and ten<br />
Days 77. 96. and 91 (Table 2). resulting in after the accident. Guskova 119871 shows<br />
a total of 31 fatalities,<br />
only one death in Kiev. a patient who died<br />
on the 10th day from combined heat and<br />
Dr. Barahanova reported that one transfer radiation injuries. The transfer patient who<br />
patient from <strong>Chernobyl</strong>. with signs and died shortly after hospital admission from<br />
symptorns of acute radiation sickness, CVA must have been the Kiev fatality on<br />
suffered a cerebrovascular accident (CVA) day four. The early death in Kiev has been<br />
enroute and (lied from that cause shortly confirmed by 'Deputy USSR Minister of<br />
after admission to the hospital IFry. 19871. Health, Ye. Vorobyev. who told a press<br />
Hence. the <strong>Chernobyl</strong> fatalities include 2 conference on 9May1986 that a third<br />
initial fatalities. 28 ARS fatalities (27 died person, the first victim of radiation<br />
in Moscow. and I in Kiev), and 1CVA exposure. died three days after the accident<br />
fatality (Table 3). In Table 4. notice that in a Kiev hospital IGoure. 19871.<br />
6
- - 0<br />
LL v<br />
0<br />
4<br />
AC.<br />
o0<br />
- I C<br />
CC3<br />
0<br />
EE<br />
- U<br />
(U<br />
-'~c<br />
-Co0<br />
-C13<br />
1-0<br />
r<br />
C<br />
QU(<br />
0~<br />
o~ o M<br />
7.
SECTION 4<br />
UNCONFIRMED CHERNOBYL FATALITIES<br />
According to Dr. Barabanova's presenta- there (ten or so meters away) is roughly<br />
tion. Fry lOctober 19871 reports the 1000 roentgens. ' ' Later, in referring to Y.<br />
following: "A journalist/photographer who Velikhov, Vice President of the USSR<br />
spent some time at the accident site shortly Academy of Sciences, seen walking in front<br />
after the accident is reported to have died of the camera, the reporter adds: "A little<br />
about 9 months later: the cause of his death bit earlier in the film. we were told why the<br />
is controversial according to Dr. Barabano- fam,,ous scientists had to risk their lives.<br />
va who did not elaborate further."<br />
What about the invisible cameraman? The<br />
other mnembers of the filming crew?"<br />
The Moscow weekly Nedeh*'a cited Noioye<br />
Russko*ve Sloro of 31 May 1987 on the death Three Soviet sources appear to agree on at<br />
of the motion picture director Vladimir least one additional <strong>Chernobyl</strong> fatality. The<br />
Shevchenko two months earlier frorn Moscow News highlights the radiation risk<br />
radiation sickness IGoure. 1987a1. Shev- to the camera crew. Ncdelva reports the<br />
chenko produced the documentary. "Cher- death of the movie director and the hospinobyl.<br />
Chronicle of Difficult Days." which talization of two cameramen. Dr. Barabawvas<br />
filmed in May 1986 at the <strong>Chernobyl</strong> nova mentions the reported but controverpow\er<br />
station. Nedelha also reported that sial death of a journalist/photographer.<br />
two cameramen received large radiation<br />
doses and are Itheni currently hospitalized. Non-Soviet sources report additional fatalities<br />
and radiation injuries among the<br />
Tie documentary filn on the <strong>Chernobyl</strong> cleanup crews according to Radio Free<br />
accident has been described in the Moscow Europe "Rad Background Report." 14<br />
News No. 11. 22-29 March 1987 IGoure, October 1986 IGoure, 1987bl. At the same<br />
1987al. Readers are left with the irnpres- time, official Soviet spokesmen deny that<br />
sion that a camera and its crew are taken more people have died from radiation<br />
on a tour through the damaged reactor, overdoses at the <strong>Chernobyl</strong> atomic power<br />
often close to high levels of radiation by the plant since the accident there last year<br />
plant's chief engineer, who says, "That pipe killed 31 people [Los Angeles Times. 19871.<br />
I. Prohably refers to a dose rate of I(100 R/h.<br />
8
SECTION 5<br />
CAUSES OF FATAL OUTCOMES<br />
All three Soviet data sources discuss the<br />
importance of radiation damage to skin and<br />
basically repeat the same message:<br />
dromes, each of which could have been an<br />
independent cause of a fatal outcome.<br />
These two groups total 19 fatalities, which<br />
is the number of cases with severe skin<br />
injury commonly cited as being incompat-<br />
ible with life. Apparently, the remaining<br />
eight fatalities were caused by combina-<br />
tions of various syndrornes without radi-<br />
ation skin damage being an independent<br />
"In 19 cases, the deaths of patients with<br />
third- and fourth degree acute radiation<br />
disease occurred only as a result of<br />
severe damage, which was incompatible<br />
with survival, to 50-90% of the surface<br />
of their bodies." IState Committee, cause of the fatal outcomes.<br />
19861<br />
-"n the cases of at least 19 of the 56 Various ARS syndromes reported by<br />
patients<br />
paint ses ith aturst the ns we Guskova 119871 as apparent independent<br />
with burns, the burns were causes of fatal<br />
unquestionably<br />
outcomes<br />
fatal."<br />
(or being<br />
IGuskova.<br />
capable<br />
1987j of producing a fatal outcome) are listed in<br />
"Of the 27 patients who died of Table 6. As mentioned before, radiation<br />
radiation injuries at H~ospital No. 6, 19 skin damage was the exclusive or contributhad<br />
radiation induced skin injuries over ing cause in 19 fatal outomes. Lethal<br />
90-100% Isicj of their body surface. It intestinal syndrome, indicated by the<br />
wvas the Soviets's opinion that in these appearance of diarrhea from day four to<br />
cases, the skin injuries were so severe clay eight, was noted in 10 patients. These<br />
as to be considered incompatible with patients were exposed to 10 Gy or higher<br />
life. even in the absence of bone gamma radiation- they all died within the<br />
marrow danmage." IFry, 19871 first three weeks after exposure.<br />
All these statements appear to be true but Acute radiation pulmonitis was observed in<br />
incomplete. Guskova 119871 discusses the seven patients with third- and fourth-degree<br />
interrelationships between individual caus- ARS. It was typified by a rapidly intenes<br />
of lethal outcomes in somewhat more sifying difficulty in breathing, by ventiladetail,<br />
summarized in Table 5. Although all tion failure, and by the onset of lethal<br />
patients with third- and fourth-degree hone- outcomes from hypoxernic coma. "Autopsy<br />
marrow syndrome had severe radiation revealed large blue lungs with marked<br />
burns (Table 2), only five lethal outcomes interstitial edema."[Guskova. 19871<br />
could be ascribed exclusively to radiation<br />
dan,age to vast areas of skin. These five Six cases with lethal outcomes \\ere<br />
cases did not involve radiation enteritis or ascribed to radiation damage causing<br />
irreversible myelodepression. and their irreversible hematopoietic aplasia or to<br />
whole-body (loses did not exceed 6 Gy. In complications caused by bone marrow<br />
14 other cases. radiation damage to skin transplants. Hemophilia achieved thanato-<br />
\\as combined with other severe syn- genetic significance only in one case.<br />
9
Table 5. ARS fatalities at Hospital No. 6 in Moscow.<br />
Causes of Fatality<br />
Number of<br />
Fatalities<br />
Radiation skin damage exclusively 5<br />
Skin damage combined with other<br />
ARS syndromes 14<br />
Combinations of ARS syndromes other<br />
than skin damage 8<br />
Total fatalities 27<br />
Table 6. ARS syndromes identified as causes' of fatalities.<br />
ARS Syndrome<br />
Fatalities<br />
with Syndrome<br />
Radiation skin damage 19<br />
Intestinal syndrome (10 Gy or more) 10<br />
Acute radiation pulmonitis 7<br />
Irreversible hematopoietic aplasia or 6<br />
bone-marrow transplant complications<br />
Iemophilia 1<br />
Thermal burns/internal contamination 2<br />
Radiation induced vascular damage<br />
I<br />
10
In Moscow. 56 of the 115 ARS patients -Patient injury or medical treatment<br />
suffered from radiation induced skin category.<br />
injuries. Two of the 56 patients formed a<br />
unique subgroup: they also had severe Patient categories include one cerebrovasthermal<br />
skin burns produced by steam and cular accident, a patient in the medium<br />
incurred a significant internal radionuclide gamma dose group who died from lethal<br />
close estimated as 1.5 to 4 sieverts on the radiation clanage to the Vascular system.<br />
basis of postmortem radiometry. The radio- Five fatal outcomes are ascribed solely to<br />
nuclide (loses of all other patients did not radiation damage to large skin areas. These<br />
exceed 1 to 3 percent of their external close, patients fall in the severe gannia dose<br />
The gamma dose of these two patients was group: their estimated whole-body closes<br />
on the order of 4 to 5 Gy. The patients died did not exceed 6 Gy.<br />
on days 23 and 18, respectively, from their Further. patient groups ilude bone<br />
cornbined injuries jFry. 19871 . marrow transplants and fetal liver cell<br />
transplants. These patients wvere selected<br />
One patient in NIoscoWv suffered a fatal on the basis of a whole-body close of 6 (\<br />
CVA 2 . probably associated with general- or greater, estimated from the lym plocyte<br />
ized radiation induced vascular damage count in peripheral blood and from cyto-<br />
I Fry. 19871. The patient received an genetic analvsis of chromosome aberraaverage<br />
bone marrow close of 3 Gy and tions. This close level \\as Understood to<br />
suffered severe radiation induced skin represent irreversible or extremely proinjuries.<br />
manifesting as three waves of tracted and deep myelodepression.<br />
erythema. with the third wave developing<br />
almost three months after the accident. The bone marrow transplants include 13<br />
Accordingly. this patient (case 3 presented patients grouped in two subcategories:<br />
b ini the Dr. second-cegre Barabanova)<br />
e<br />
must<br />
ARS<br />
be<br />
group<br />
the one fatality<br />
\vlo died a. Six patients had radiation damage to<br />
on day 96 after the accident (Table 2).<br />
skin and intestines at a level that was<br />
deemed not incompatible \\ith life.<br />
Four patients died between days 27<br />
Based on currently available Soviet data<br />
and 29 after the transplants. Two<br />
sources reviewed. Table 7 provides a<br />
sumnmary of our effort to correlate the<br />
patients survived their bone marrowv<br />
Chernobvl accident fatalities and causes transplants; they had gamma expo-<br />
terms of:<br />
sures sin of 5.8 and 9.0 Gy. Gamnia<br />
grouped ndoses<br />
for the four patients who died<br />
ranged from 4.3 to 10.7 Gy, placing<br />
* Severity of the bone marrow syndrone one fatality in the severe and three<br />
(the whole-body gamma radiation fatalities in the extremnelv severe<br />
doses).<br />
bone marrow syndrome groups. They<br />
(lied from "mixed virus-bacterial<br />
* ARS syndrome identified as exclusive infections." also called as "cornplior<br />
contributing cause of fatal outcomes. cations caused by bone marrowv<br />
2. \ote thal thi ('\' A ltalit v occurred in \1l co\%: a different ARS patient suffered a (''A enroutle<br />
to the ho,,pilal and died on cla\ ltour in Kiev.<br />
11
0 C<br />
C:<br />
- I I I<br />
co<br />
m<br />
C<br />
0 ,<br />
o --<br />
-) 0 : r-%<br />
CO2v 2C 5W<br />
C6-<br />
C. I C I<br />
4,) I I I<br />
rs X-
transplants" jGuskova, 19871. State "unspecified" category. The totals of<br />
Committee 119861 reports that all six individual lethal causes have been taken<br />
bone marrow transplants in this fron Table 6. Adding the numbers of<br />
group had similar disorders that may patients in relevant patient categories, the<br />
have been caused by graft-ver- corresponding totals are not always identisus-host<br />
(GVH) disease. Two of tihe cal with those in Table 6. but are still in<br />
four fatalities ill this group occurred general agreement. For example. we would<br />
prior to the puiblicationi of the State have 13 rather than 10 cases of lethal<br />
Committee 119861 report. It is stated radiation damage to intestines. In cases<br />
that the GVH reactions may have \\here the fatal outcomes associated with<br />
been contributing causes of these two a patient category are ascribed to multiple<br />
deaths. contributing ARS syndronmes ("yes" in<br />
Table 7), all the syndromes may not<br />
b. Seven patients had radiation damage necessarily be present for each and every<br />
to their skin. intestines, and lungs at fatality.<br />
a level that was deemed incompatible<br />
with life. They died from acute In other cases. the relevant patient cateradiation<br />
clamage to the skin. gut. gories identify fewer fatalities than the<br />
and lungs on days2 to 19 after the totals of Table 6. For example, only four<br />
translplants (days 15 to 25 after rather than six total cases of lethal inexrposure).<br />
fections and GV1I- diseases are identified<br />
directly. In these cases. the question marks<br />
In the category of fetal liver cell trans- in Table 7 suggest other possible patient<br />
plants, all six patients died from radiation categories associated with a particular<br />
darnage to the skin and intestines IGusko- cause of death.<br />
a. 19871. Patients for the transplantation<br />
of human embryo liver cells were selected The individual gamma closes of the accion<br />
the basis of extrernely severe clarnae dent victims are generally unavailable.<br />
to the skin and intestines and extremely Exceptions are the cerebrovascular acciunfavorable<br />
prognosis IState Committee dent victim with anl estimated close of 3 G\.<br />
19861. the four bone marrow transplant fatalities<br />
who died from mixed Virus and bacterial<br />
Two patients in the thermal burns and infections with estimated closes of 4.3.<br />
internal contamination category died from 5.0-7.9. 5.8-6.0. and 10.7 Gy, respectively.<br />
their combined injuries. including bone and the two thermal burn and internal<br />
marrow damage and racliation damage to contamination patients with their combined<br />
skin. closes shown in Table 7. Guskova 119871<br />
also states that the patients with lethal<br />
In Table 7 all but two of the 27 fatalities intestinal syndromes received a short-term<br />
in Moscow can be traced through patient whole-body gamma radiation dose on tihe<br />
categories, thus leaving only two in the order of 10 Gy or higher.<br />
13
SECTION 6<br />
FATALITY INCIDENCE VERSUS RADIATION DOSE<br />
A total of 238 <strong>Chernobyl</strong> accident victims 7). Four indiViduals in the fourth close<br />
received estimated whole-body garnma group have specific close estimates. but no<br />
radiation closes in excess of 0.8 Gy. They specific values were available for the<br />
were diagnosed as suffering from acute remaining five individuals.<br />
radiation sickness (ARS). Based on the<br />
severity of ARS. the patients were grouped Based on the method of analyzing binary<br />
in four close range categories. The radiation data given by Cox (1983). a maximum<br />
data published to date include the dose likelihood analysis of the data given in<br />
ranges and the number of patients and Table 9 was performed to estimate the<br />
deaths in each (lose range group.<br />
incidence of <strong>Chernobyl</strong> fatalities with close.<br />
Along with the explicit close estimates given<br />
Although individual case histories have not in Table 9 for both surVivors and fatalities.<br />
been published, individual dose estimates it 'vas assumned that the remaining indihave<br />
been reported for some cases. includ- vidual doses were uniformly distributed<br />
ing II survivors and 7 fatalities. Table 8 over the corresponding lose range. For<br />
lists individual loses and reference fol- example, there are five fatalities in the close<br />
lowed by a number or letter of patient range of 4.2 to 6.3 Gy listed under the<br />
identification. The numbers associated with "3rd- close group in Table 9, thus. for this<br />
Guskova 119871 refer to the original patient analysis. closes of 4.41, 4.83. 5.25. 5.67.<br />
identification system: the letter "x" refers and 6.09 Gy each \\ere assumed for the five<br />
to patients mentioned but not otherwise fatalities.<br />
identified in reports. Some of the individual For purposes of comparison, five different<br />
closes reported by Guskova<br />
been updated in a<br />
119871<br />
later report<br />
have<br />
(UNSCEAR.<br />
model<br />
moefrswrealidtanyzth<br />
forns were applied to analyze the<br />
been updatled i oa ate pted NSC . survivor/fatality binary data: normal. log-<br />
1987): Table 8 contains the updated values, normal. Weibull. logistic, and log-logistic.<br />
Table 9 sunmarizes the available data on Table 10 gives the model forms along with<br />
radiation closes for both individuals and parameter estimates obtained. Based upon<br />
groups of individuals. In most cases, the the X 2 goodness-of-fit statistic, a ranking<br />
patients in a dose group divide into two would fall between the Weibull and logsubgroups:<br />
a few patients with individual normal models, with Weibull the best fit.<br />
dose estimates. and the remainder distrib- However, all models fit the data well<br />
uted over the given (lose range. The (p
Table 8. Radiation (loses of individual patients.<br />
Dose Number of Survivors Fatalitie<br />
Dose<br />
Range<br />
Group (G%) Patients Deaths Dose Patient Reference Dose Patient Relerence<br />
Ist 0.8-2. I 140 0.9 Guskova 119871 - 97<br />
1.4 - 48<br />
1.9 Geiger 119861 - 6<br />
2nd 2.0 -4. 55 I 3.3 Guskova I19871 - 39 3 FrY 119871 - 3<br />
3.9 - 21<br />
3.3 State Committee 119861 - D<br />
2.25 Fr\ 119871 - 2<br />
3.0 Telvatniko% 119871<br />
3rd 4.2-6.3 21 7 5.6 Guskoxa 119871 - x 5.2 (uskoka 119871- 6<br />
4.9 Geiger 119861 - 2 4.4 - 5<br />
4th 6-16 22 21 8.7 Guskova 119871 - x 6.4 c;uskoa 1 19871- 28<br />
1.2 - 16<br />
9 -8<br />
7 Fry 119871 - x 7<br />
Tomals 238 29 II patients 7 patients<br />
15
C13l<br />
I)"<br />
- 0-<br />
Cu<br />
oo<br />
o<br />
C<br />
IIV<br />
0~0<br />
0 0l<br />
I-c<br />
U. c<br />
01
C) 00 00 kfs v-s<br />
(- I 1r) ON C0<br />
\C r- r "0 r<br />
00<br />
ON<br />
C<br />
C<br />
E -0 - \-<br />
C0ON0<br />
CCO<br />
5-j<br />
co "0<br />
5-<br />
oO -a u<br />
E<br />
+<br />
, 6<br />
"<br />
'1)z<br />
a 7)<br />
cu 11<br />
Vt re-<br />
E' r<br />
00<br />
17-
Table 11. Fatality percentile doses.<br />
Percentile Dose (cGy)<br />
slope*<br />
Model LD 1 () LD 50 LD 9 (? (percent/cGy)<br />
Normial 409.9 616.4 822.9 0.194<br />
Log-normnal 391.4 592.0 895.3 0.159<br />
WeibUtl 401.0 631.4 843.3 0.181<br />
Logistic 422.8 606.0 789.3 0.218<br />
Log-logistic 403.3 590.1 863.5 0.174<br />
*slope L 90 -1 D) (percent/c~y)<br />
99.9<br />
Prior non-<strong>Chernobyl</strong> estimates<br />
99.0- (See text)<br />
95.0-Z<br />
90.0-<br />
S70.0-<br />
.. . . . . .<br />
5~ 0.0 - ......<br />
10.0 -......<br />
0... ....<br />
1.0......<br />
....<br />
0.1..... 200.00..0.80.100<br />
... Dose......<br />
Figre1.Clerobl ccdet atliie: nrml isriuton.28 nd.iuas.90<br />
1.08
99.9<br />
Prior non-<strong>Chernobyl</strong> estimates<br />
99.0- (See text)<br />
95.0-<br />
90.0-<br />
70.0-<br />
50.0<br />
V30.0-<br />
3000 402507010<br />
Dose (cGy)<br />
Figure 21. <strong>Chernobyl</strong> accident fatalities, log-normial distribution, 238 individuals, 90%<br />
confidence interval.<br />
19
99.9<br />
99.0 Prior non-<strong>Chernobyl</strong> estimates -<br />
90.0<br />
10.0<br />
CD 5.0 -<br />
70.0-<br />
50.0-<br />
-~30.0-<br />
1.0-<br />
01200 300 400 500 700 1000<br />
Dose (cGy)<br />
Figure 3. <strong>Chernobyl</strong> accident fatalities: Weibull distribution. 238 indlividuals, 90%7c<br />
confidence interval.<br />
20
99.9<br />
Prior non-<strong>Chernobyl</strong> estimates<br />
_70.0-<br />
99.0-<br />
95.0-<br />
90.0-<br />
S50.0-<br />
o30.0-<br />
m ......<br />
5.0<br />
0.1 200 400 600 800 1000<br />
Dose (cGy)<br />
Figu~re 4. <strong>Chernobyl</strong> accident fatalities: logistic distribution, 238 individuals. 90%<br />
confidence interval.<br />
21
99.9<br />
Prior non-<strong>Chernobyl</strong> estimates<br />
(See text)<br />
99.0-<br />
95.0-<br />
90.0-<br />
= 70.0-<br />
a 50.0-<br />
v30.0-<br />
10.0<br />
5.0-<br />
1.0 '<br />
200 300 400 500 700 1000<br />
Dose (cGy)<br />
Figure 5. <strong>Chernobyl</strong> accident fatalities: log-logistic distribution, 238 individuals, 90%<br />
confidence interval.<br />
22
five different models used to analyze the the LD 5 () based on prior estimates with that<br />
data: the 90% confidence bounds are given estimated here from the <strong>Chernobyl</strong> data.<br />
by the two dashed lines. In order to That is on average. prior estimates of the<br />
graphically illustrate the data from Table LD 50 wou*d be about 290 cGy whereas the<br />
9 used in fitting the various models, the value is about 607 cGy according to our<br />
data were grouped according to the analysis of the <strong>Chernobyl</strong> data: Ilis<br />
horizontal lines shown in the plots. The amounts to about a factor of two reflected<br />
data used in the maximum likelihood by medical care.<br />
procedure is treated in binary form. and the<br />
asterisks on the horizontal line simply Also, as is apparent from tile plots of<br />
indicate the dose range midpoints of the fatality incidence versus close, in general<br />
data groups. Accordingly, while the log- there is a significant difference in the slope<br />
normal (Fig. 2) and log-logistic (Fig. 5) magnitude betwveen the <strong>Chernobyl</strong> dose<br />
models appear by inspection to fit the data response curves and that according to prior<br />
better than the other models, they actually estimates of lethality. As indicated in Table<br />
have the lowest X 2 values (see Table 10) 11. depending upon the model, slope<br />
although as indicated above, all five of the estimates based oil analysis of the Chernomodels<br />
fit the data well. byl data range from about 0.16 to 0.22<br />
percent/cGy: the corresponding slope value<br />
A range of incidence for radiation lethality from prior lethality estimates is about 0.44<br />
based on some prior estimates is also percent/cGy. This difference in the steepindicated<br />
in Figs. I through 5 by the broad ness of the close response curves could also<br />
shaded band. The left-most boundary is be Clue in part to medical attention. Albased<br />
oil an LD 5 () midline tissue close of though more than likely, it is due to factors<br />
256 cGy and a slope estimated from close other than strictly ARS in combination \\ith<br />
response curves for large animals (swine. medical attention such as injuries to skin.<br />
sheep, goats. and dogs) which tend to be which as mentioned above \was a ver\<br />
parallel (Bond and Robertson 1957, Cron- significant factor. For example. in the abkite<br />
and Bond 1958. and Cronkite 1982). sence of medical attention, a larger propor-<br />
The right-most boundary is based on an tion of lethalities Would be expected at<br />
1I) 5 ) value of 325 cGy. suggested by lower doses. i.e.. in the second and third<br />
Lushbaugh (1969) for healthy young close groups. This certainly Would tend to<br />
adults, with the slope also inferred frorn steepen the dose response curves. lipowevlarge<br />
animals.<br />
er, because of the imprecision of data<br />
regarding burn injuries and other non-ARS<br />
One means of assessing the efficacy of effects, we have not attempted to perform<br />
medical care including the various treat- any isolated causal analysis with regard to<br />
ment modalities of ARS afforded the fatality incidence utilizing truncated sub-<br />
Chernoby I accident victims is by comparing sets of' data.<br />
23
SECTION 7<br />
LIST OF REFERENCES<br />
Bond, V. P. an(l J. S. Robertson, Ann. Rev. Science Applications International Corp.<br />
Nuc. Sci., Vol. 7, pp. 135-162. 1957. La Jolla, California, October 1987a.<br />
Cox, D. R.. Analv'sis of Binary Data, --.- "Medical," The Chernobvl <strong>Accident</strong><br />
Chapman and Hall Ltd., New York, NY, Data Base, Sec. VIII, Science Applica-<br />
1983. tions International Corp. McLean, Virginia,<br />
October 1987b.<br />
Cronkite. E. P., "The Impact of Estimates<br />
of Human Radiation Tolerance upon Guskova, A. K., Eary Acute Effects among<br />
Radiation Emergency Management," the Victims of the <strong>Accident</strong> at the Chernob\l<br />
Proceedings of a Symposium on the Control <strong>Nuclear</strong> Power Plant. Ministry of Health<br />
of Exposure of the Public to Ionizing of the USSR, LN-818-87, April 1987.<br />
Radiation in The Event of <strong>Accident</strong> or Los Angeles Times. "The World," Part I. p.<br />
Attack. National Council on Radiation 2. December 10, 1987.<br />
Protection and Measurement, Bethesda,<br />
Maryland, May 1982, pp. 21-27.<br />
Lushbaugh. C. C.. "Reflections on Some<br />
Recent Progress in Humnan Radliobiolo-<br />
Cronkite. E. P. and V. P. Bond. "Acute Re Advaes i n Rad o.ol.<br />
radiationgy" Advances in Radiation iology. Vol.<br />
Frceosyr J.. Vol. 9.19. A d 3. Academic Press, New York, 1969. pp.<br />
Forces Med. J.. Vol. 9. 1958. pp. 277-315.<br />
313-324.<br />
Telyatnikov. L. P.. Personal interview with<br />
Fry. S. A., Trip Reort: International Atomic the Fire Brigade Commander, Novemn-<br />
Energy Agency (IAEA) Workshop on the ber 1987.<br />
Aledical Handling of Skin Lesions Follotiing<br />
High Level <strong>Accident</strong>al Irradiation, UNSCEAR (United Nations Scientific<br />
Institute Curie, Paris, France, Center for Committee on the Effects of Atomic<br />
Epidemiologic Research, Medical and Radiation), A/AC.82/R. 473, Annex H,<br />
Health Services Division, Oak Ridge Early Effects in Man of High Doses of<br />
Associated Universities, September 28- Radiation; Appendix 1, "Acute Radi-<br />
October 2, 1987.<br />
ation Effects in Victims of the <strong>Chernobyl</strong><br />
<strong>Nuclear</strong> Power Plant <strong>Accident</strong>." A. G.<br />
Geiger. H. Jack. The <strong>Accident</strong> at Chernobvl Guskova, ed., 10 November 1987.<br />
and the Medical Response, Journal of<br />
American Medical Association, Vol. USSR State Committee on the Utilization<br />
256. No. 5. pp. 609- 612. August 1, of Atomic Energy. "The <strong>Accident</strong> at the<br />
1986. <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant And Its<br />
Consequences: Annex 7, Medical and<br />
Gotre, Leon, "Radiation Situation at the Biological Problems," trans. by IAEA.<br />
AES and in the Danger Zone," The IAEA Experts Meeting, Vienna, Austria,<br />
<strong>Chernobyl</strong> <strong>Accident</strong> Data Base, Sec. V, August 25-29, 1986.<br />
24
DISTRIBUTION LIST<br />
DNA-TR-89-45<br />
DEPARTMENT OF DEFENSE<br />
ARMED FORCES RADIOBIOLOGY RSCH INST<br />
ATTN: AFFRI-DEPT OF RAD BIOCHEM<br />
ATTN: BHS<br />
ATTN: DIRECTOR<br />
EXH<br />
ATTN: MRA<br />
ATTN: PHY<br />
3 CYS ATTN: FCPRT<br />
ATTN: NUC SECURITY<br />
INTELLIGENCE CENTER, PACIFIC<br />
ATTN: COMIPAC<br />
INTERSERVICE NUCLEAR WEAPONS SCHOOL<br />
ATTN: TTV<br />
2 CYS ATTN: TTV 3416TH TTSQ<br />
ATTN: RSD<br />
JOINT DATA SYSTEM SUPPORT CTR<br />
ATTN: SCIENTIFIC DIRECTOR ATTN: C-332<br />
ATTN: TECHNICAL LIBRARY<br />
ATTN: JNSV<br />
ASSISTANT SEC OF DEF (C3)<br />
ATTN: DIR (S&TNFC3)<br />
ASSISTANT SECRETARY OF DEFENSE<br />
INTERNATIONAL SECURITY POLICY<br />
ATTN: NUC FOR & ARMS CNTRL PLCY<br />
ASSISTANT TO THE SECRETARY OF DEFENSE<br />
ATTN: EXECUTIVE ASSISTANT<br />
ATTN: MIL APPL<br />
ATTN: MIL APPL,C FIELD<br />
NATIONAL DEFENSE UNIVERSITY<br />
ATTN: ICAF, TECH LIB<br />
ATTN: NWCLB-CR<br />
ATTN: LIBRARY<br />
ATTN: STRAT CONCEPTS DIV CTR<br />
NATIONAL SECURITY AGENCY<br />
ATTN: CHIEF A4213<br />
NET ASSESSMENT<br />
ATTN: DOCUMENT CONTROL<br />
DEFENSE INTELLIGENCE AGENCY<br />
OFFICE OF THE SEC OF DEFENSE<br />
ATTN: DB<br />
ATTN: NAVAL FORCES<br />
5 CYS ATTN: DB-, RSCH RESOURCES DIV ATTN: STRATEGIC PROG & TNF<br />
ATTN: DB-5C<br />
ATTN: DB-6B<br />
STRATEGIC AND THEATER NUCLEAR FORCES<br />
ATTN: DIA/VPA-2 (FED RES DIV)<br />
ATTN: DR E SEVIN<br />
ATTN: DN<br />
ATTN: DR SCHNEITER<br />
ATTN: DT<br />
ATTN: OFFICE OF SECURITY<br />
STRATEGIC TARGET PLANNING<br />
ATTN: OS<br />
ATTN: JKC (ATTN: DNA REP)<br />
ATTN: RTS-2B<br />
ATTN: JLT<br />
ATTN: JP<br />
DEFENSE INTELLIGENCE COLLEGE<br />
ATTN: DIC/RTS-2<br />
ATTN: JPTP<br />
AITN: DIC/2C<br />
THE JOINT STAFF<br />
DEFENSE LOGISTICS AGENCY<br />
ATTN: ED30<br />
ATTN: J-3 SPECIAL OPERATIONS<br />
ATTN: COMMAND SECURITY OFC<br />
ATTN: J-3 NUC OPERATIONS BR<br />
DEFENSE NUCLEAR AGENCY<br />
ATTN: CD<br />
ATTN: J-5 NUC & CHEMICAL DIV<br />
ATTN: JAD/SFD<br />
ATTN: DFRA<br />
ATTN: JAD/SSD<br />
ATTN: DFSPG ULLRICH<br />
ATTN: JSOA<br />
ATTN: NANF<br />
ATTN: SAGA<br />
ATTN: NASF<br />
U S EUROPEAN COMMAND/ECJ-3<br />
ATTN: OPNA<br />
ATTN: ECJ-3<br />
ATTN: OPNS<br />
ATTN: RAEE<br />
U S EUROPEAN COMMAND/ECJ-5NPG<br />
20 CYS ATTN: RARP ATTN: ECJ-5<br />
4 CYS ATTN: T;TL<br />
U S EUROPEAN COMMAND/ECJ-6-OJ<br />
DEFENSE NUCLEAR AGENCY<br />
ATTN: ECJ-6<br />
ATTN: TDNM<br />
2 CYS ATTN: TDTT W SUMMA U S EUROPEAN COMMAND/ECJ2-T<br />
ATTN: ECJ2-T (TGTS DIV)<br />
DEFENSE TECHNICAL INFORMATION CENTER<br />
2 CYS ATTN: DTIC/FDAB U S EUROPEAN COMMAND/ECJ4-LW<br />
ATTN: ECJ4-LW<br />
FIELD COMMAND DEFENSE NUCLEAR AGENCY<br />
ATTN: FCPR<br />
Dist- 1
DNA-TR-89-45 (DL CONTINUED)<br />
U S EUROPEAN COMMAND/ECJ5-N<br />
ATTN: ECJ5-N (NUC BRANCH)<br />
U S EUROPEAN COMMAND/ECJ7/LW<br />
ATTN: ECJ-7 LW<br />
UNDER SEC OF DEFENSE FOR POLICY<br />
ATTN: DUSP/P<br />
ATTN: USD/P<br />
U S ARMY CONCEPTS ANALYSIS AGENCY<br />
ATTN: TECHNICAL LIBRARY<br />
U S ARMY FIELD ARTILLERY SCHOOL<br />
ATTN: ATSF-CD<br />
U S ARMY FORCES COMMAND<br />
ATTN: AF-OPTS<br />
U S ARMY HUMAN ENGINEERING LAB<br />
ATTN: DIRECTOR<br />
ATTN: DR D HODGE<br />
UNDER SECRETARY OF DEFENSE<br />
ATTN: CHAIRMAN PSEAG<br />
ATTN: CHAIRMAN, DEF SCIENCE BD<br />
2 CYS ATTN: C31 U S ARMY INFANTRY CENTER<br />
ATTN: ATSH-CD-CSO<br />
DEPARTMENT OF THE ARMY<br />
U S ARMY ITAC<br />
ARMY RESEARCH INSTITUTE<br />
ATTN: COMMANDER<br />
COMBAT MATERIAL EVAL ELEMENT<br />
ATTN: SECURITY ANALYST<br />
ATTN: IAX-Z<br />
U S ARMY MATERIEL COMMAND<br />
ATTN: DRCDE-D<br />
ATTN: DRCSS<br />
DEP CH OF STAFF FOR OPS & PLANS<br />
U S ARMY NUCLEAR & CHEMICAL AGENCY<br />
ATTN: DAMO-FDQ<br />
ATTN: MONA-NU<br />
ATTN: DAMO-NCN ATTN: MONA-ZB<br />
ATTN: DAMO-NCN(NUC CHEM DIR)<br />
U S ARMY TRAINING AND DOCTRINE COMD<br />
ATTN: DAMO-ZXAATN<br />
ATTN: ATCD-AO CDO<br />
DEPARTMENT OF THE ARMY<br />
ATTN: ATCD-FA<br />
ATTN: DAMO-NCZ<br />
ATTN: ATCD-N<br />
ATTN: DAMO-OD<br />
ATTN: ATOO-NCO<br />
ATTN: DAMO-ODSO<br />
ATTN: DAPE-ODRE<br />
U S ARMY WAR COLLEGE<br />
ATTN: DAPE-HREATNLIRY<br />
ATTN: SARD-ZCS (MAJ ADAMS)<br />
ATTN: LIBRARY STRATEGIC STUDIES<br />
HARRY DIAMOND LABORATORIES<br />
ATTN: SLCIS-IM-TL (TECH LIB)<br />
JOINT SPECIAL OPERATIONS COMMAND<br />
ATTN: J-2<br />
ATTN: J-5<br />
U S ARMY AIR DEFENSE SCHOOL<br />
ATTN: COMMANDANT<br />
U S ARMY ARMAMENT RSCH DEV & ENGR CTR<br />
ATTN: DRDAR-LCN-F<br />
U S ARMY ARMOR SCHOOL<br />
ATTN: ATSB-CTD<br />
ATTN: TECH LIBRARY<br />
US ARMY MATERIEL SYS ANALYSIS ACTVY<br />
ATTN: DRXSY-DS<br />
USA MILITARY ACADEMY<br />
ATTN: DEPT OF BEHAV SCI & LDRSHP<br />
ATTN: DIR NATL SECURITY STUDIES<br />
USA SURVIVABILITY MANAGMENT OFFICE<br />
ATTN: SLCSM-SE J BRAND<br />
USACACDA<br />
ATTN: ATZL-CAD-N<br />
DEPARTMENT OF THE NAVY<br />
FLEET INTELLIGENCE CENTER<br />
U S ARMY BALLISTIC RESEARCH LAB ATTN: FICPAC, CODE 21<br />
ATTN: AMXBR-VLD-R (DR RAINIS)<br />
MARINE CORPS<br />
ATTN: AMXBR-VLD, DR KLOPCIC<br />
ATTN: CODE PPO<br />
ATTN: DRDAR-BLA-S, TECH LIB<br />
ATTN: PSI G/RASP<br />
ATTN: DRDAR-BLT<br />
ATTN: SLCBR-D<br />
MARINE CORPS DEV & EDUCATION COMMAND<br />
U S ARMY CHEMICAL SCHOOL<br />
ATTN: LIBRARY<br />
ATTN: ATZN-CM-M<br />
NAVAL OCEAN SYSTEMS CENTER<br />
U S ARMY COMD & GENERAL STAFF COLLEGE ATTN: CODE 9642-B<br />
ATTN: ATZL-SWJ-CA<br />
NAVAL PERSONNEL RES & DEV CENTER<br />
ATTN: ATZL-SWT-A<br />
ATTN: CODE P302<br />
Dist-2
DNA-TR-89-45 (DL CONTINUED)<br />
NAVAL POSTGRADUATE SCHOOL<br />
ATTN: CODE 1424 LIBRARY<br />
NAVAL RESEARCH LABORATORY<br />
ATTN: CODE 1240<br />
ATTN: CODE 2627 (TECH LIB)<br />
NAVAL SEA SYSTEMS COMMAND<br />
ATTN: PMS-423<br />
ATTN: SEA-09G53 (LIB)<br />
ATTN: SEA-6631<br />
NAVAL SURFACE WARFARE CENTER<br />
ATTN: CODE F-31<br />
ATTN: G GOO<br />
NAVAL TECHNICAL INTELLIGENCE CTR<br />
ATTN: NTIC-DA30<br />
NAVAL WAR COLLEGE<br />
ATTN: CODE E-11 TECH SVC<br />
ATTN: CTR FOR NAV WARFARE STDS<br />
ATTN: DOCUMENT CONTROL<br />
ATTN: LIBRARY<br />
STRATEGY DEPT<br />
NAVAL WEAPONS EVALUATION FACILITY<br />
ATTN: CLASSIFIED LIBRARY<br />
AIR UNIVERSITY LIBRARY<br />
ATTN: AIJL-LSE<br />
ATTN: LIBRARY<br />
ASSISTANT CHIEF OF STAFF<br />
2 CYS ATTN: AF/SAMI<br />
ASSISTANT CHIEF OF THE AIR FORCE<br />
ATTN: SAF/ALR<br />
DEPUTY CHIEF OF STAFF/XOX<br />
ATTN: AFXOOSS<br />
ATTN: AFXOXFM<br />
FOREIGN TECHNOLOGY DIVISION, AFSC<br />
ATTN: CCN<br />
ATTN: SDA<br />
SPACE DIVISION/YH<br />
ATTN: YH<br />
STRATEGIC AIR COMMAND/ADWN<br />
ATTN: ADWN<br />
STRATEGIC AIR COMMAND/SPD<br />
ATTN:<br />
SPD<br />
STRATEGIC AIR COMMAND/STIC<br />
ATTN: 544 SIW/DI (STIC)<br />
NUCLEAR WEAPONS TNG GROUP, ATLANTIC<br />
STRATEGIC AIR COMMAND/XOXO<br />
ATTN: CODE 222 AIR COM /<br />
ATTN: DOCUMENT CONTROL<br />
ATTN: XOXO<br />
NUCLEAR WEAPONS TNG GROUP PACIFIC<br />
STRATEGIC AIR COMMAND/XPX<br />
ATTN: CODE 32<br />
ATTN: XPZ<br />
ATTN: DOCUMtNT ;ONTROL STRATEGIC AIR COMMAND/XRFS<br />
OFFICE OF CHIEF ir r .,VAL OPERATIONS ATTN: XRFS<br />
ATTN: NIS-22<br />
TACTICAL AIR COMMAND/XPSC<br />
ATTN: NOP06D<br />
ATTN: TAC/DOA<br />
2 CYS AT TN: N0P403<br />
ATTN: NOP 50<br />
U S AIR FORCE ACADEMY<br />
ATTN: NOP60<br />
ATTN: LIBRARY<br />
ATTN: NOP60D<br />
ATTN: USAFA/SP<br />
ATTN: NOP 603<br />
ATTN: NOP 91<br />
USAF SCHOOL OF AEROSPACE MEDICINE<br />
ATTN: OP 654<br />
ATTN: RADIATION SCIENCES DIV<br />
OPERATIONAL TEST & EVALUATION FORCE<br />
ATTN: COMMANDER<br />
PLANS, POLICY & OPERATIONS<br />
ATTN: CODE-P<br />
ATTN: CODE-POC-30<br />
TACTICAL TRAINING GROUP, PACIFIC<br />
ATTN: COMMANDER<br />
DEPARTMENT OF THE AIR FORCE<br />
AFIA/INIS<br />
AFIS/INT<br />
ATTN:<br />
ATTN: AFIA/INKD<br />
INT<br />
AIR UNIVERSITY<br />
ATTN: STRATEGIC STUDIES<br />
WEAPONS LABORATORY<br />
ATTN: WL/SUL<br />
DEPARTMENT OF ENERGY<br />
LAWRENCE LIVERMORE NATIONAL LAB<br />
ATTN: PAUL THERMAN<br />
ATTN: Z DIVISION LIBRARY<br />
LOS ALAMOS NATIONAL LABORATORY<br />
ATTN: REPORT LIBRARY<br />
ATTN: T DOWLER<br />
SANDIA NATIONAL LABORATORIES<br />
ATTN: DIV 9013, W LING<br />
ATTN: TECH LIB<br />
Dist-3
DNA-TR-89-45 (DL CONTINUED)<br />
OTHER GOVERNMENT<br />
KAMAN SCIENCES CORPORATION<br />
ATTN: DASIAC<br />
CENTRAL INTELLIGENCE AGENCY<br />
ATTN: COUNTER-TERRORIST GP<br />
PACIFIC-SIERRA RESEARCH CORP<br />
ATTN: DIRECTOR OF SECURITY<br />
2 CYS ATTN: A LAUPA<br />
ATTN: MEDICAL SERVICES<br />
2 CYS ATTN: G ANNO<br />
ATTN: NIO-T<br />
ATTN: H BRODE<br />
ATTN: N10 R & D - STRATEGIC SUBCOMMITTEE SYS<br />
PACIFIC-SIERRA RESEARCH CORP<br />
ATTN: RC &LDISBCOM E ATTN: D GORMLEY<br />
ATTN: TECH LIBRARY 2 CYS ATTN: G MCCLELLAN<br />
FEDERAL BUREAU OF INVEST ACADEMY<br />
ATTN: BEHAVIORAL RSCH UNIT<br />
ATTN: LIBRARY<br />
FEDERAL EMERGENCY MANAGEMENT AGENCY<br />
ATTN: ASST ASSOC DIR FOR RSCHTN<br />
ATTN: CIVIL SECURITY DIVISION<br />
ATTN: G ORRELL NP-CP<br />
ATTN: OFC OF CIVIL DEFENSE<br />
R & D ASSOCIATES<br />
ATTN: C MCDONALD<br />
2 CYS ATTN: DOCUMENT CONTROL<br />
J THOMPATN<br />
ATTN: J THOMPSON<br />
SCIENCE APPLICATIONS INTL CORP<br />
ATTN: DOCUMENT CONTROL<br />
U S DEPARTMENT OF STATE ATTN: E SWICK, MS 33<br />
ATTN: PM/STM<br />
ATTN: J MARTIN<br />
ATTN: J WARNER<br />
U S N CLEAR REGULATORY COMMISSION<br />
ATTN: L HUNT<br />
ATTN: DIR DIV OF SAFEGUARDS<br />
ATTN: R J BEYSTER<br />
ATTN: SYANIV<br />
SCIENCE APPLICATIONS INTL CORP<br />
DEPARTMENT OF DEFENSE CONTRACTORS<br />
ADVANCED RESEARCH & APPLICATIONS CORP<br />
ATTN: SECURITY OFFICE<br />
AEROSPACE CORP<br />
ATTN: LIBRARY ACQUISITION<br />
BDM INTERNATIONAL INC<br />
ATTN: C SOMERS<br />
ATTN: J BRADDOCK<br />
DATA MEMORY SYSTEMS, INC<br />
ATTN: T DUPUY<br />
KAMAN SCIENCES CORP<br />
ATTN: DASIAC<br />
ATTN: E CONRAD<br />
ATTN: B BENNETT<br />
ATTN: DOCUMENT CONTROL<br />
ATTN: J FOSTER<br />
ATTN: J PETERS<br />
ATTN: LGOURE<br />
ATTN: M FINEBURG<br />
ATTN: W LAYSON<br />
SCIENCE APPLICATIONS INTL CORP<br />
ATTN: R CRAVER<br />
SCIENCE APPLICATIONS INTL CORP<br />
ATTN: JOHN A SHANNON<br />
SYSTEMS RESEARCH & APPLICATION CORP<br />
ATTN: R STEELE<br />
ATTN: S GREENSTEIN<br />
Dist-4
""n" ""e*""UMENTATION PAGE Form Approved<br />
1<br />
filltiae onr reprint)e o vrgeIhu inludg tmefor eewing ristucilons srn~l.g 741""g<br />
0 MB No. 0704 .0188<br />
7s<br />
o,'o<br />
6. AUTHOR(S) EPR TP AD AESCVEE<br />
Worm Uniten No.<br />
Browne, D.<br />
N/A<br />
7. PERFORMING ORGANIZATIC NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION<br />
Armed Forces Radiobiology Research Institute<br />
REPORT NUMBER<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency S9-<br />
Bethesda, ?4D 20889-5145<br />
9. SPONSORING/ MONITORING AGENCY NAME(S) AND ADORE!P(E5 10. SPONSORING/ MONITORING<br />
AGENCY REPORT NUMBER<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency -<br />
Washington, DC 20305<br />
11. SUPPLEMENTARY NOTES<br />
1 2a. DISTRIBUTION/AVAILABILITY STATEMENT 1 2b. DISTRIB3UTION CODE<br />
Approved for public release; distribution unlimited.<br />
13. ABSTRACT (Maximum 200 words) ~~S<br />
fewl<br />
JUitiwOW<br />
6v<br />
!A By.<br />
A, 14 SUBJECT TERMS I15 NUMBER OF PAGES<br />
5<br />
________________________________________________________IS._<br />
16 PRICE CODE<br />
17. SECURITY CLASSIFICATION I S. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF<br />
OF REPORT OF THIS PAGE OF AB3STRACT ABSTRACT<br />
UNCLASSIFIED<br />
UNCLASSIFIED<br />
StnadFrI9 Rv 9<br />
NSN 7S40-01 -2604500 StanardO Form 298 e -9<br />
2,14102
Biomedical Lessons From the<br />
<strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant <strong>Accident</strong><br />
A4PIPID ;i)RLE$ AAOIOLOGY<br />
0 AESEARCM .%STITUT5<br />
SCIAiNTIPIC ReSpO"<br />
SR91-2<br />
Lt Col Doris Browne, MC<br />
The <strong>Chernobyl</strong> nuclear accident afforded the treating physicians a chance to observe clinical ARS in man, defining the degree<br />
of severity according to average radiation dose exposure, and to mahe prognoses for the individual patients on the course<br />
of the ARS, based on biological criteria. The author gives a detailed account of the clinical cause of the disease including<br />
all available laboratory values. She provides valuable data that can be utilized in handling similar accidents in the future.<br />
The world's worst radiation accident posited on the skin and mucous mem<br />
occurred at the <strong>Chernobyl</strong> nuclear branes from the molten steam and Ā Lt Operations Col Dori Division, Browne, Military MC, is Chief,<br />
Requirements Medical<br />
power plant in the USSR during the dust. Wet clothing contaminated by and Applications <strong>Department</strong> IMRAI, Armed<br />
early hours of April 26, 1986. This ac- the steam and dust provided another Forces Radiobiology Research Institute (AFRRI).<br />
cident unleashed megacuries of radio- source of contamination. She is a licensed physician in hematologyactive<br />
contamination into the atmos- Within 15 minutes of the acident, oncology. Dr. Browne attended medical school<br />
phere, generated an explosive blast first aid was provided by middle level at Georgetown University, and completed an<br />
internship and residency in internal medicine<br />
that knocked the thousand-ton lid off medical personnel and emergency at Walter Reed Army Medical Center. She<br />
the top of the reactor, and sent burn- team members. Individuals with acute was a fellow in hematology and oncology<br />
ing graphite and heat in a plume about symptoms were transported to the<br />
three miles high. The aftermath and hospital in Pripyat, where the initial<br />
at Walter Reed Army Medical Center. She<br />
joined the staff of William Beaumont Army<br />
significance of this disaster are still screening took place; others in satis- Medical Center, El Paso, Texas as the Assistant<br />
Director of the Hematology-Oncology<br />
being realized in the USSR and neigh- factory condition were instructed to Clinic, <strong>Department</strong> of Medicine. She joined<br />
boring countries, go to the hospital for examination. The the staff of the AFRRI as Chief of the Med-<br />
Details of the accident were re- initial care consisted of antiemetics, ical Operations Division and is responsible<br />
ported at the International Atomic symptomatic medication, and stable for the Medical Effects of the <strong>Nuclear</strong> Weapons<br />
Energy Agency (IAEA) meeting held saturated potassium iodide. The spa- (MENW) Course and all technology transfer<br />
activities at the institute. She is the Officerin<br />
Vienna, Austria, in August 1986, cialized emergency team of radiation in-Charge of the Medical Radiobiology Adand<br />
summarized in IAEA Safety Series accident specialists arrived at the visory Team (MRA T that recently sponsored<br />
Technic-' Report No. 75.1.2 This report accident site within 12 hours and, the First Consensus Development Conference<br />
summarizes the basic information on with the on-site medical personnel, on the Treatment of Radiation Injuries. A sumcasualties,<br />
triage, and treatment, and screened and triaged more than 350 mary report of this conference was published<br />
in an international journal and the proceedings<br />
the radionuclides released into the persons within the first 36 hours. Dur- of the conference are in press. She isa member<br />
atmosphere. The immediate casual- ing the first 24 hours, 132 persons of the American Society of Clinical Oncology,<br />
ties included only plant personnel, were hospitalized; one individual died American College of Physicians, and the Nefiremen,<br />
and auxiliary staff present at, from severe thermal burns during the tional Medical Association.<br />
or in the vicinity of the accident site. first hour, and another worker (a re-<br />
These casualties were all subject to actor operator) was unaccounted for body, as well as with the intake of adthe<br />
combined effects of the following: and believed to be buried under the ditional radionuclides through inhalashort-term<br />
beta/gamma radiation re- collapsed debris. tion. These patients were diagnosed<br />
b leased in the emission cloud; external The triage officer, a physician, as having ARS resulting from extenbeta/gamma<br />
radiation from fragments made decisions based on the initial sive beta radiation bums to the skin<br />
* of the damaged reactor core scat- symptoms and lymphocyte counts. and significant whole-body gamma<br />
tered through the accident site; in- Persons with severe symptoms were radiation exposure.<br />
halation of gaseous and aerosolized hospitalized with clinical complaints The diagnostic criteria used to assess<br />
dust composed primarily of radio- of acute radiation sickness (ARS). the presence of ARS was the presence,<br />
isotopes of cesium, plutonium, and Three hundred of these patients were intensity, and duration of symptoms<br />
iodine; and beta/gamma particles de- sent to a specialized treatment center is, nausea, vomiting, and erythema of<br />
*Chief, Medical Operations Division, Military Re- in Moscow and another 200 were the skin and mucose); time of onset;<br />
quirements and Applications <strong>Department</strong>. Armed sent to a hospital in Kiev. The 237 and the peripheral lymphocyte count,<br />
Forces Rediobiology Research Institute BWthsda, hospitalized individuals received sig- which decreased to less than 10/L<br />
MD 20814.<br />
Supported by the Armed Forces Rodioboogy Re- nificant combined radiation effects during the first 24 hours following<br />
search Institute. <strong>Defense</strong> <strong>Nuclear</strong> Agency. Views from the extensive beta/gamma ex- radiation exposure in patients with<br />
presented in this paper are those of the author;<br />
no endorsement by the <strong>Defense</strong> <strong>Nuclear</strong> Agency posure, which was generally external ARS. During the first 36 hours after<br />
hes boen given or ahould be inferred. and relatively uniform over the whole the accident, the 237 hospitalized<br />
Pe a-90-91i0. see emtbiOctaor 990 25<br />
91 5 C7 054
persons were diagnosed as having a and prognosis of ARS. Hyperamylas- treated with intravenous administraclinical<br />
pattern consistent with first emia was used as a supplementary tion of triple broad-spectrum antidegree<br />
through fourth degree ARS. diagnostic tool. biotics, including aminoglycoside,<br />
After admission to the hospital, they Treatment consisted of supportive cephalosporin and semi-synthetic<br />
were monitored again for contamina- therapy, which included selective anti- penicillin. If this regimen did not retion<br />
and, when necessary, decon- microbial intestinal decontamination, duce the fever within 48 hours, three<br />
taminated with soap, water, and a reverse isolation, empiric systemic anti- or four doses of gamma globulin were<br />
clothing change. Routine samples of biotic administration, and transfusion administered. An intravenous antiurine<br />
and blood were drawn for anal- replacement of blood and blood prod- fungal (amphotericin B) was adminysis,<br />
and thyroid scanning was per- ucts. Definitive treatment of allogeneic istered if the neutropenic fever performed.<br />
The radiation dose received bone marrow transplantation (BMT) sisted for seven days, along with the<br />
was estimated by counting the num- and human embryonic liver cell trans- antibiotics and gamma globulin. Paber<br />
of aberrant chromosomes (dicen- plantation (LCT) was performed on tients with herpes simplex were given<br />
trics) in cultured lymphocytes (cyto- patients with irreversible myelosup- acyclovir. Approximately one third of<br />
genetic analysis). The diagnosis of pression. A sterile environment was the patients with third and fourth<br />
ARS was confirmed during the first maintained through strict observance degree ARS had the herpes virus.<br />
five days for persons admitted to of hand washing by all attending per- Viral skin lesions were treated with<br />
the Moscow Hospital. Approximately sonnel upon entering and leaving the topical acyclovir. No deaths were<br />
seven days after the accident, the room; mandatory use of disposable attributed to bacterial infection alone<br />
radiation dose was estimated and the gowns, masks, and caps; antiseptic in patients with hematopoietic synpatients<br />
were categorized into four decontamination of footwear; chang- drome. However, infectious complicagroups<br />
according to prognosis and ing of patient undergarments daily; tion was the primary cause of death<br />
severity of hematopoietic syndrome antiseptic washing of walls, floors in patients with ARS complicated by<br />
(Table I). Twenty-two injured persons and items used in the room; and indi- thermal burns, radiation-induced enwere<br />
classified as having fourth de- vidually assigned antiseptically treated teritis, or acute graft-versus-host<br />
gree (extremely severe) ARS; 23 as nursing items. Isolation rooms pro- disease from BMT. The etiology of<br />
having third degree (severe) ARS; 53 vided air sterilization with ultraviolet terminal septicemia, documented by<br />
as having second degree (moderate) lamps. The microorganism population surveillance cultures, was most often<br />
ARS; and 139 as having first degree was maintained at less than 500M 3 from Staphylococcus epidenrdts.<br />
(mild) ARS. 3 in the room air. Raw fruits and veg- The hematopoietic syndrome was<br />
Neutrophil count was used to de- etables and canned products were treated with prophylactic and theratermine<br />
finally the magnitude of radia- eliminated from the patients' diet. 4 peutic fresh random donor platelets<br />
tion dose. The parameter used was The decision was made early to when the platelet count dropped to<br />
the time required for the neutrophils perform BMT on patients with third 20 x 10 9 /L or lower, or with the first<br />
to decrease to 0.5 x 10 9 /L, based on and fourth degree ARS and possible sign of bleeding. Transfusions usudata<br />
collected over a period of up to irreversible myelosuppression. 5 These ally were required every one to three<br />
three months in cases that exhibited patients vomited within the first half- days. To inactivate the immunocomtypical<br />
postirradiation platelet and/or hour, suffered from diarrhea during petent cells from the donor, all blood<br />
neutrophil counts with distinct de- the first one to two hours, and from components were irradiated with<br />
pletion and restoration phases. Coin- swelling of the parotid gland during 1,500 cGy of gamma radiation before<br />
plete blood counts were performed the first 24 to 36 hours of exposure, transfusion. Only one person received<br />
two to three times per week for two in addition to myelosuppression. single donor platelets. While the mato<br />
three months. This data was used Infections, manifested by the on- jority of patients showed no evidence<br />
to definitively confirm the diagnosis set of fever and neutropenia, were of overt bleeding, autopsy results disclosed<br />
micro-circulatory failure and very<br />
porous capillaries in several organs.<br />
Table I. Diagnostic Categories for Acute Radiation Sickness (ARS1. Inroms crorgse<br />
In some situations, cryo-preserved<br />
Degree of ARS Dose (cGy) Seventy of ARS Prognosis autologous platelets, as well as allogeneic<br />
platelets, were used success<br />
100-200 Mild Very favorable fully. Autologous platelets were taken<br />
from patients with second and third<br />
11 200-400 Moderate Relatively favorable dreens onth st d postd<br />
degree ARS on the first day post-<br />
III 400-600 Severe Doubtful irradiation. Platelet transfusions pre-<br />
IV z 600 Extremely severe Poor vented life-threatening bleeding. Three<br />
to eight transfusions of 250cc per<br />
26 The Journal of tte US Amiv Medical Deoorye "
person were used to treat patients Table It. Transplantation Cases, Estimated Radiation Dose, and Outcome.<br />
with second and third degree ARS.<br />
No evidence of refractoriness de- Degree of ARS Dose IcGy) Treatment Day of Death Cause of Death<br />
veloped. A considerable number of<br />
packed red blood cells were trans- IV 920 BMT 15 Skin; pneumonitis<br />
fused in patients with second and IV 1200 BMT 17 Skin: GI injury<br />
third degree ARS accompanied by IV 1180 BMT 18 Skin; GI injury<br />
severe radiation burns.<br />
Allogeneic BMT taken from 113 Iv 1000 BMT 18 Skin; GI injury<br />
random related donors was performed I1 550 BMT 21 Hemorrhage 1<br />
on 13 patients with third and fourth Iv 830 BMT 24 Pneumonitis<br />
degree ARS. Additionally, six patients 660 BMT 25 ARDS; toxicity<br />
with fourth degree ARS received<br />
embryonic LCT, which contained III 440 BMT 2 34 Mixed infection; GVH<br />
*<br />
stem cells and few immunocompe- IV 640 BMT 3 48 Mixed infection; GVH<br />
tent cells to decrease<br />
tetclst the risk eraeters of<br />
fIV 750 BMT 3 86 Mixed infection; GVH<br />
developing acute graft-versus-host<br />
disease (Table II). Fifty percent (seven IV 1020 BMT 4 91 Mixed infection; GVH<br />
patients) of the BMT patients died III 560 BMT 5 Alive<br />
within 17 days of transplantation (15 IV 870 BMT 5 Alive<br />
to 25 days following radiation exposure)<br />
from acute radiation injury IV 1110 LCT 14 Skin; GI injury<br />
to lung, intestine, and/or skin. The IV > 1000 LCT 14 Skin; GI injury<br />
remaining six patients did not have<br />
severe skin burns or intestinal injuries IV 1370 LCT 15 Skin; GI injury<br />
but received a total radiation dose IV 1240 LCT 17 Skin; GI injury<br />
estimated to be between 440 and IV 1090 LCT 18 Skin; GI injury<br />
1,020 cGy. Two of the six patients<br />
survived BMT (having received 560cGy IV 830 LCT 6 30 Toxicity; ARDS<br />
and 870 cGy doses, respectively) from<br />
N o t a BMT = bone marrow transplantation; LCT = liver cell transplantation; GI = gastrointestinal ,nurv<br />
haplo-identical female (sisters) donors. GVH = graft-versus-host: ARDS = acute respiratory distress syndrome irespiratory insufficiency).<br />
Both experienced transient partial en- 'Hemorrhage from mechanical trauma during catheterization.<br />
23 graftment of the transplanted mar- BMT M from rmHAietcidnr haplo - 1 identical donor but own myelopoiesis restored.<br />
38MT from HLA-identical donor.<br />
4 BMT from haplo-identical donor but own myelopoiesis restored.<br />
row before rejection 32 and 35 days<br />
after BMT, respectively, with res-<br />
5 SMT from haplo-identical donor rejected, own myelopoiesis restored.<br />
toration of their own myelopoiesis sLCT from male, postmortem evidence of own myelopoisais being restored.<br />
after 28 days. These two patients<br />
are still alive at more than three planted marrow engrafted. While re- ity, duration, and recurrence, contriband<br />
one half years after the acci- covery of autologous myelopoiesis uted significantly to the overall pathodent.<br />
5 A 62-year-old female patient may occur following large doses of physiology and outcome of the patient.<br />
who received LCT lived for 30 days; radiation exposures, such as experi- Severe skin injuries were manifested<br />
postmortem findings showed evidence enced by those patients with third by diffuse hyperemia; secondary eryof<br />
regeneration of her own myelo- and fourth degree ARS, it is unknown thema; dry and wet desquamation<br />
poiesis, indicated by female cell karyo- if this occurred due to transient en- with blistering, ulceration, and necrotype.<br />
She had received a male donor graftment of transplanted stem cells. tic dermatitis; recurrent waves of erytransplant.<br />
4 6 Radiation-induced skin injuries (beta thema; and after evidence of healing<br />
The effectiveness of BMT in an radiation burns) were seen only in of the primary lesions, edema, fever,<br />
emergency situation may be limited to combination with hematopoietic syn- and a worsening of the patient's clinpatients<br />
receiving less than 900cGy drome radiation injury. Skin doses of ical picture. 4 Topical treatment was<br />
of gamma radiation with at least 1 % radiation were estimated to be 10 to necessary, with glucocorticoids and<br />
of marrow stem cells remaining, no 20 times greater than bone marrow analgesia in the more severe cases.<br />
skin or intestinal radiation injuries, or whole-body doses, confirming the Pain control was relatively ineffecand<br />
no combined injury. 7 Seven of uncontrolled, nonuniform nature of tive, especially topical anesthesia,<br />
the 13 BMT patients died of skin and radiation accident exposure. These which seems to be typical for radiaintestinal<br />
injuries before the trans- skin injuries, according to their sever- tion injuries.<br />
Pil S90-9110. 20terb,,octobev 1990 27
Bums were fatal in the 19 of 56 pa- suiting hyperamylasemia. 8 No treat- Periodic follow-up examinations were<br />
tients with radiation burns on > 40% ment was indicated for the parotitis, made during the first year after the<br />
to 100% of body surface area.6 If early which gradually resolved; salivation, accident. Patients usually had dyssecondary<br />
erythema over > 40% body however, recurred very slowly. trophic and ulcerative skin lesions,<br />
surface area was present, a clinical Rapidly intense dyspnea with acute some with subcutaneous edema pripicture<br />
of febrile-toxemia, followed by respiratory insufficiency (adult respi- marily over the knees and feet. Skin<br />
hepatorenal insufficiency, encephal- ratory distress-like syndrome) was lesions were treated with agents that<br />
opathy with cerebral edema, coma, seen in seven patients with third and improved local blood circulation and<br />
and death resulted 14 to 48 days post- fourth degree ARS. This condition tissue trophism. Five patients sufirradiation.<br />
Plasmapheresis was used to rapidly progressed for two to three fered deep ulcers which required recontrol<br />
the hepatorenal insufficiency. 1 4 days leading to death. Postmortem peated plastic surgery.<br />
This treatment prolonged survival examination revealed enlarged blue The immunologic status of patients<br />
slightly but did not prevent death lungs with interstitial edema but no with second, third, and fourth degree<br />
from encephalopathic coma. The destruction of mucous membranes of ARS, tested 1 to 1.5 years after the acburns<br />
may have been the primary the trachea and bronchi. These pa- cident showed a persistent decrease<br />
cause of death in some cases; how- tients also had severe skin and in- in T-helper lymphocytes with an inever,<br />
in most cases, the burns were testinal radiation injuries, crease in T-suppressor lymphocyte<br />
associated with severe hematopoietic Beta radiation caused early damage activity and a significant decrease in<br />
syndrome and severe acute gastro- to the eye tissues; erythema of the the helper-suppressor ratio. However,<br />
intestinal syndrome (enteritis). eyelid, with increased vasculature of there was no evidence of a decrease<br />
In ten patients, the gastrointestinal the lid, and conjunctiva. Cutaneous in the absolute lymphocyte level or in<br />
syndrome was the life-threatening changes were manifested by waves the T- and B-subpopulations. These<br />
manifestation of ARS, with severe of erythema, hyperpigmentation, and changes in lymphocyte helper and<br />
diarrhea suggesting a radiation dose scaling. Partial epilation of the eye- suppressor populations were not seen<br />
greater than 1,000 cGy. All of these brows was transient, and all patients in patients with first degree ARS. Durpatients<br />
died within three weeks of ir- retained their eyelashes. (Scalp hair ing the follow-up period, no severe or<br />
radiation. When the enteritis persisted growth recovered fully.) Other eye life-threatening infections were noted.<br />
in spite of supportive fluid and elec- changes noted were decreased cor- Immunocorrective therapy was attrolyte<br />
therapy, death may have been neal sensitivity and superficial radia- tempted using T- and B-activin in<br />
caused solely by the gastrointestinal tion-induced keratitis, which regressed several cases. Respiratory infections<br />
syndrome. over one to two months without corneal occurred in three of eight patients<br />
Large amounts of thick rubber-like opacification. Treatment for the eye with third and fourth degree ARS and<br />
mucous formed in the oropharyngeal changes included topical ointments to only one of 22 patients with second<br />
area of about 82 patients and in some the eyelid skin and eyedrops of 20% degree ARS. A competent immune<br />
cases resulted in respiratory difficulty. albucid, sophradex, and vitamin solu- system remains critical in enhancing<br />
Initially, some patients showed be- tions into the conjunctival cavity. 4 microbial and viral resistance during<br />
nign acute radiation-induced inflam- One severely ill patient with fourth convalescence of the irradiated pamation<br />
of cheeks, tongue, and gums. degree ARS, who survived the acute tient. A plan of long-term follow-up<br />
Those having third and fourth degree phase, developed angioretinopathy observation remains in effect.<br />
ARS had, in addition to the rubbery with hemorrhage and plasma dismucous<br />
plugs, painful erosions and charges about five months post- Conclusion<br />
ulcers of oral mucosa, which required irradiation. He also had persistent The consequences ot the <strong>Chernobyl</strong><br />
sterile saline irrigation and frequent low diastolic pressure in the central radiation accident provide data on a<br />
debridement. In a significant number retinal artery. He is one of the two large group of critically ill patients<br />
of patients this radiation-induced in- surviving BMT patients. No radiation- who received uniform whole-body irflammation<br />
was complicated by sec- induced lens changes were observed radiation and required treatment of<br />
ondary bacterical and viral infections, one year postirradiation. ARS in a massive casualty situation.<br />
In one third of the patients with Convalescence of three to four The event afforded the opportunity to<br />
severe hematopoietic syndrome, her- months was required for those pa- learn many lessons regarding the<br />
petic lesions formed massive crusts tients with first and second degree biomedical effects of ionizing radiaon<br />
the lips and face about three to ARS; a much longer period was tion and to clarify many aspects about<br />
four days postirradiation. Patients necessary for those having third and the early radiobiclogical effects in<br />
with fourth degree ARS and herpetic fourth degree ARS. The majority of humans. The accident also provided<br />
lesions also developed radiation-induced the patients have resumed work but data on severe and extensive beta<br />
parotitis, inability to salivate, and re- cannot work with radiation sources. radiation skin injuries, which com-<br />
28 The Journlsl of the US Arrmy Moica DeDa1" ,e"
plicated the course of illness and REFERENCES oflomzng Radiaton, United Nations<br />
played a significant role in the death 1. USSR State Commission on the Scientific Committee on the Effects<br />
of 19 of the 31 patients. Combined Utilization of Atomic Energy. The of Atomic Radiation. New York,<br />
injuries consisting of trauma, thermal accident at the <strong>Chernobyl</strong> nuclear 1988, pp 613-647.<br />
burns, and radiation were the cause power plant and its consequences. In- 5. Baranov AE, Gale RP, Guskova AK,<br />
of death in two patients very early in formation complied for the Post- et at: Bone marrow transplantation<br />
the course of ARS. Clinical ARS was <strong>Accident</strong> Review Meeting, part II, after the Chemobyl nuclear accident.<br />
observed in man, the degrees of se- Annex 7, Vienna, Austria, August NEnglJMed321(4):205-212, 1989.<br />
verity defined according to average 25-29, 1986. 6. Young RW: <strong>Chernobyl</strong> in retrospect.<br />
radiation dose exposure, and prog- 2. International <strong>Nuclear</strong> Safety Ad- Pharmac Ther 39:27-32, 1988.<br />
nosis made for the course of ARS visory Group. Summary report on the 7. Browne D, Weiss JF, MacVittie TJ,<br />
based on biological criteria. Informa- post-accident review meeting on the et at: Conference report: The first<br />
tion for biological dosimetry was <strong>Chernobyl</strong> accident. Vienna, Austria: consensus development conference<br />
obtained from karyotypical analysis, International Atomic Energy Agency on the treatment of radiation injuries.<br />
lymphocyte counts, and symptoms (IAEA), 1986. (Safety Series No. lntJ]jdwtBio/57(2):437-442, 1990.<br />
during the early stages of illness; later 75-INSAG-1). 8. Guskova AK, Nadezhina NM, Barathe<br />
granulocyte count proved to be 3. Bair WJ: Radiological impacts of the banova AV, et al: Acute effects of<br />
a useful dosimetric tool. While con- <strong>Chernobyl</strong> accident. Health Physics radiation exposure following the<br />
tinuous data and follow-up assess- Society Newsletter. February, 1987. <strong>Chernobyl</strong> accident: Immediate rement<br />
is necessary, this information 4. Guskova AK, Barabanova AV, suits of radiation sickness and outshould<br />
prove useful in responding Baranov AE, et al: Acute radiation come of treatment. In Treatment ,.r<br />
to radiation accidents and providing effects in victims of the <strong>Chernobyl</strong> Radwhon nrunes. Browne D, Weiss JF,<br />
effective medical carg to the result- nuclear power plant accident. Ap- Mac Vittie TJ, et al (eds). Plenum,<br />
ing casualties. pendix to Sources, Effects and Risks New York, 1990. 0<br />
29
OATE:
$A t#4- 44", ONIb004lt-# Ne Oft0-*.0wM o I<br />
,*Stl eb,4 W ,11 0' 0Mg . '1 A'f00 I, IP4 1 ** $ et*14d4 *. oa 41A<br />
-----------.... PAGE Ou OAp .e<br />
AD A254 669 , e11e,,( sf'1 I144 e f I6m 4OI I .E4<br />
, AGENC 3Ili REPORT TYPE AND DATES COVERED<br />
DNA-TR-<br />
4 ItTLE ANO SUBTITLE<br />
<strong>Nuclear</strong> <strong>Accident</strong>s in the Former Soviet Union:<br />
5. FUNDING NUMBERS<br />
Kyshtym, Chelyabinsk and <strong>Chernobyl</strong> DNA/AFRRI 4020<br />
*. AUTHORIS)<br />
Daniel L. Collins, Ph.D.<br />
Lt Col, USAF<br />
lE L E T E<br />
1. PERFORMING ORGANIZATION NAME(S)AND AOORESSIES) GANIZATION<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency R U<br />
Armed Forces Radiobiology Research Institute 24 j 1 S12 "<br />
Bethesda, HD 20889-5145<br />
9. SPONSORING/MONITORING AGENCY NAMEIS) AN() ADORE SS(E S) 10. SPONSORING/MONTORING<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency<br />
AGENCY REPORT NUMBER<br />
Armed Forces Radlobiology Research Institute<br />
Bethesda, MD 20889-5145<br />
SUPPLEMENTARY NOTES Many of the data contained in this report comes from discussion<br />
nd brefings. In addition, on 2 December 1991, several scientists from the<br />
Institute of Biophysics of the USSR Ministry of Health, Chelyabinsk Branch Office,<br />
ho have studied the Kyshtym and Chelyabinsk nuclear accidents for decades,<br />
resented human data at AFRRI that have never before been released to the West.<br />
12e. DISTRIBUTION/AVAILABIUITY STATEMENT 12b. DISTRIBUTION CODE<br />
Approved for public release: Distribution is unlimited. A<br />
13. ABSTRACT(Msuimum200 w0 sJ Three nuclear accidents besides <strong>Chernobyl</strong> have occurred in<br />
the Former Soviet Union (FSU). The accidents occurred over the geographic area<br />
around Kyshtym and Chelyabinsk in the Urals between 1949 and 1967 and cntaminated<br />
over half a million people. The first accident occurred in 1949-1951, the stcond<br />
on 29 September 1957, and the third in 1967, and involved the air transfer of<br />
irradiated sand particles. Although these accidents occurred between 25 and 43<br />
years ago, the first official admission by the FSU was made in June 1989, and it<br />
was only during late November 1991 that the FSU declared a national disaster<br />
emergency concerning the affected area. The health ministries are now interested<br />
in data previously collected from these irradiated populations to examine health<br />
effects, including cancer, and genetic damage in humans. Data collected from these<br />
large populations and occupationally exposed workers offer a unique opportunity to<br />
quantify the adverse health effects of chronic exposure to ission products,<br />
reactor neutrons and enviromental chemicals.<br />
94,1 126<br />
14. SUBJECT TERMS 16. NUMBER OF PAGES<br />
Kyshtym, Chelyabinsk, <strong>Chernobyl</strong>, REM, human, psychological 0<br />
radiation 9 0 Sr, Curies, 13 7 Cs, Neutrons, Isotopes, radiation<br />
injury, cancer, genetics, enviromental chemicals<br />
To. PRICECOOt<br />
17. SECURITY CLASSIFICATION 15. SECURITY CLASSIFICATION 19. SECURITY CLASSIICATION 20. LIMITATION OF<br />
OF REPORT OF THIS PAGE OF ABSTRACT ABSTRACT<br />
Unclassified Unclassified Unclassified None
ABSTRACT<br />
Three nuclear accidents besides <strong>Chernobyl</strong> have occurred in the Former<br />
Soviet Union (FSU).<br />
The accidents occurred over the geographic area around<br />
Kyshtym and Chelyabinsk in the Urals between 1949 and 1967 and contaminated<br />
over half a million people. The first accident occurred in 1949-1951, the<br />
second accident on 29 September 1957, and the third in 1967, and involved<br />
the air transfer of irradiated sand particles.<br />
Although these accidents<br />
occurred between 25 and 43 years ago, the first official admission by the<br />
FSU was made in June 1989, and it was only during late November 1991 that<br />
the FSU declared a national disaster emergency concerning the affected area.<br />
The health ministries are now interested in obtaining data previously<br />
collected on this irradiated population to examine the health and heredity<br />
(genome damage, etc.) implications associated with these victims. A<br />
collaborative involvement by DNA/AFRRI with the Health Ministries would be a<br />
unique opportunity to obtain previously unavailable human data.<br />
-il1<br />
92-23453
<strong>Nuclear</strong> <strong>Accident</strong>s in the Former Soviet Unions :<br />
Kyshtym, Chelyabinsk, and <strong>Chernobyl</strong><br />
Avftf1atj1±t 7 Codeq<br />
Daniel L. Collins, Ph.D. . .'A/l ad/-or<br />
Dltt<br />
Spe3cial<br />
Lt Col, USAF<br />
I<br />
DTMIC QUALIY U11TPECTED 8<br />
INTRODUCTION<br />
In September 1991, three weeks after the attempted putsch, I represented<br />
the <strong>Defense</strong> <strong>Nuclear</strong> Agency/Armed Forces Radiobiology Research Institute as<br />
part of a United States Information Agency (USIA) contingent in Moscow and<br />
St. Petersburg, Russia.<br />
I met numerous scientists that were and are<br />
currently involved in all the nuclear accidents that occurred in their<br />
country. Much of the data contained in this report comes from discussions<br />
and briefings.<br />
In addition, on 2 December 1991, several scientists from the<br />
Institute of Biophysics of the USSR Ministry of Health, Chelyabinsk Branch<br />
Office, who have studied the Kyshtym and Chelyabinsk nuclear accidents for<br />
decades, presented human data that have never before been released to the<br />
West.<br />
The radiation situations in the area of Kyshtym and Chelyabinsk are<br />
unique because over the last 40 years masses of people have been exposed to<br />
9 0 Sr in the food and water chain. Basic dosimetry investigations focused on<br />
2
the doses of 9 0 Sr.<br />
The first estimates of exposure were from nuclide<br />
measurements taken from the river sediment. These analyses began in 1951,<br />
one and a half years after the accident occurred.<br />
ACCIDENTS<br />
In 1948, the U.S.S.R. began operating a plutonium production plant<br />
called Mayak in the Kyshtym/Chelyabinsk region. In 1949-1951, an accident<br />
released 3 million Ci of radiation into the Techa River (Degteva, 1991, see<br />
Figure 1).<br />
A second accident occurred in 1957, southeast of Kyshtym, when<br />
improperly ventilated storage tanks exploded, and 20 million Ci of<br />
radioactive waste were released into the atmosphere (Akleev, 1991).<br />
The<br />
storage complex was located 1.5 Km from the reprocessing plant and consisted<br />
of 60 underground storage tanks. The radionuclide composition of the<br />
fallout indicated it was comparatively fresh nuclear waste from<br />
reprocessing, stored for not more than 6-7 months, from which not only<br />
uranium and plutonium but also cesium had been removed (Medvedev, 1991).<br />
Due to the confined nature of the blast, the majority (90%) of the<br />
nuclear waste dispersed near the tanks in the form of a liquid pulp<br />
(Burnazyan, 1990).<br />
However, a plume cloud with an activity of 2 million Ci<br />
dispersed its contamination over the area shown on the map in Figure 1. The<br />
contaminates from the plume cloud were confined to the Chelyabinsk and the<br />
Sverdlovsk provinces. Before the accident, more than 28,000 people lived in<br />
the 38 villages along the Techa River. Dosages of 90Sr exceeded 0.01 Ci/km 2<br />
3
and were distributed over 23,000 km 2 (Kosenko, 1991).<br />
The highest<br />
concentration of 90Sr was located in an area known as Metlino, located near<br />
Kyshtym (see Figure 1). The doses in the Metlino area averaged 3 Sv/km 2 .<br />
The dispersion of radioactivity from the plume cloud is shown by the map in<br />
Figure 1.<br />
In the spring of 1967, a third accident occurred resulting in an air<br />
transfer of radioactive sand particles from the beach of Lake Karachay<br />
(Akleev, 1991). This lake is located due west of Argayash. A total of 600<br />
Ci 137Cs and 9 0 Sr was released by the air transfer of sand particles from<br />
the Lake Karachay beach (Akleev, 1991).<br />
Radiation from this accident was<br />
primarily gamma radiation along the Techa River and reached 5R/hour<br />
(Kosenko, 1991). The average activity along the Techa River was 10 - 5<br />
Ci/liter of water. A massive three month cleanup between the summer and<br />
autumn of 1967 (Degteva, 1991), resulted in a ten fold decrease in radiation<br />
levels. This occurred because the Soviets removed all the water from the<br />
contaminated reservoir located near the Metlino area (Degteva, 1991).<br />
(NOTE: We were not told where the water was moved to nor how it was<br />
disposed of.)<br />
DISCUSSION<br />
In the aftermath of the Kyshtym accident, if a village was found to be<br />
radioactive it was scheduled for evacuation as political circumstances<br />
dictated (Soyfer, 1991).<br />
In addition to the area dosimetry, which measured<br />
4
iver sediment for contamination, human dosimetry began in the summer of<br />
1951 measuring body excrement and contaminated clothes (Degteva, 1991). It<br />
should be noted, that during this 1.5 year hiatus between the accident and<br />
the beginning of dosimetry measurements, the people were neither informed of<br />
the accident nor of their (internal or external) exposure to any form of<br />
ionizing radiation (Soyfer, 1991).<br />
Consequently, all the food and water<br />
consumed during this time period was contaminated with radiation (Akleev,<br />
1991).<br />
In an attempt to further quantify dosimetry from the Kyshtym and<br />
Chelyabinsk accidents, the bones of deceased persons were exhumed during the<br />
1960's and resulted in the creation of a data base that reflected their<br />
*lifetime* exposures to 90Sr (Kosenko, 1991).<br />
In 1974, the Soviets created<br />
a data base using live subjects to determine their whole-body doses of 90Sr.<br />
The methodology used to make these determinations included dosimetric<br />
examinations of urine samples and frontal lobes (Douplenski, 1991).<br />
The<br />
urine samples were obtained from 1,500 residents living beside the Techa<br />
River (Kosenko, 1991); the 12,500 subjects involved in the frontal lobe<br />
study also lived along the Techa River (Kosenko, 1991). The teeth of 15,000<br />
living subjects in this area were also examined for 90Sr doses (Kosenko,<br />
1991). The location from which these residents were evacuated is known<br />
today as the "Post Box Chelyabinsk-400 area (Medvedev, 1991).<br />
The results show that 1,000 people living by the Techa River had greater<br />
than 1 mCi of 9 0 Sr in their bones (Kosenko, 1991).<br />
Further analyses showed<br />
5
that these people born in 1932-1933, and were teenagers during the first<br />
accident, accumulated three to five times more 90Sr than did those who were<br />
adults at the time of the first accident (Akleev, 1991).<br />
The maximum<br />
reading of 6 mCi per person occurred in those individuals that were<br />
teenagers during the time of the accidents (Akleev, 1991).<br />
Measuring the<br />
metabolism measurements of 9 0 Sr and overlaying them with all three previous<br />
accidents provided the following results: The exposure levels averaged 0.42<br />
Sv along the Techa River to the southwest, 0.52 Sv along the middle portion<br />
of the Techa River, and 2 Sv near Chelyabinsk (Degteva, 1991).<br />
The three<br />
accidents affected 437,000 people (Kosenko, 1991).<br />
Of these, 1,200 people<br />
obtained 200 rems over a 2 year period (Kosenko, 1991).<br />
In addition, some<br />
people received doses of up to 400 rems to their bone cells (Kosenko, 1991).<br />
RELOCATION<br />
Of the 39 villages along the Techa River before the accident, only four<br />
villages are safe to inhabit today (Akleev, 1991).<br />
An analysis of people<br />
along the Techa River show decreased leukocyte immune system functioning<br />
(Akleev, 1991).<br />
This is attributed to the 3- year period when people daily<br />
drank the radioactive water and ate contaminated food before their<br />
relocation (Douplenski, 1991).<br />
The average time of relocation took 8 years<br />
to complete (Degteva, 1991). The time range for relocation spanned 3-11<br />
years for evacuating all residents from the 35 contaminated villages along<br />
the Techa River (Douplenski, 1991; Medvedev, 1991).<br />
The total number of<br />
people moved during this time exceeded 10,500 (Akleev, 1991).<br />
After the<br />
6
people were relocated the villages were incinerated to ensure that no human<br />
habitation would occur in this highly radioactive are?.<br />
However, the heat<br />
and smoke created by the incineration process spread the contamination over<br />
the streams, rivers and lakes further contaminating the food chain for<br />
animals and humans.<br />
One exception to the protracted relocation of residents living in the<br />
contaminated areas occurred out of necessity. Following the Kyshtym<br />
accident, 1,154 residents of Kasli, near Kyshtym, were removed 7-10 days<br />
after the accident, due to extremely high 9 0 Sr levels (Akleev, 1991).<br />
People that were removed during this time period, now have twice the accute<br />
myeloid leukemias that the control groups have exhibited (Akleev, 1991).<br />
An historical note, the construction of the radioactive facilities was<br />
conducted between 1945-1948 by approximately 70,000 inmates from 12 labor<br />
camps (Douplenski, 1991). The Khyshtym location is N 55-44, E 60-35. The<br />
Khyshtym restricted area covers 2700 sq/km and contains eight lakes with<br />
interconnecting watercourses. The Khyshtym atomic plant is situated in a<br />
tunnel, which extends beneath a river, with only a smoke stack visible from<br />
the air or ground. During the construction process, one lake was drained, a<br />
building was built on its lakebed with cement, rubber and lead, then the<br />
lake was refilled (Medvedev, 1979).<br />
During the cold war several high<br />
altitude, reconnaissance aircraft routinely photographed this area.<br />
In<br />
1960, Major Francis Gary Powers was shot down by a surface-to-air missile<br />
while flying over the Khyshtym, Chelyabinsk atomic facilities.<br />
7
CONTROL GROUPS<br />
Two control groups were selected for comparison purposes for this<br />
longitudinal field study of irradiated humans (Degteva, 1991).<br />
The control<br />
groups were located just south of this area and were not involved in the<br />
radiation. The first control group consisted of 34,000 persons of the same<br />
socioeconomic status as the victims and had no access to the contaminated<br />
Techa River and were not contaminated by the plume cloud or the other<br />
accidents (Degteva, 1991).<br />
The second control group consisted of all people<br />
in the Chelyabinsk rural province but not living in the contaminated city of<br />
Chelyabinsk or the irradiated area surrounding Chelyabinsk (Kosenko, 1991).<br />
The control groups within the nonradioactive portions of the Chelyabinsk<br />
province consisted of 1.5 million people (Kosenko, 1991).<br />
This data base<br />
has been a part of the ongoing research for 40 years (Degteva, 1991).<br />
The<br />
information we obtained is a summary compiled from 33 years of data (Akleev,<br />
1991). [NOTE: There are still 7 years of untapped data because the<br />
analysis is 7 years behind due to a lack of computer equipment.]<br />
IMPACT OF IONIZING RADIATIONi ACCIDENTS<br />
Statistical results showed significantly increased death rates along the<br />
Techa River over the last 33 years.<br />
Coefficients were described in terms of<br />
excessive risk per Gy.<br />
Findings showed stomach cancers to be two to three<br />
times greater than in survivors of Nagasaki and Hiroshima, breast cancer two<br />
8
times greater, and lung and esophageal cancers two to three times greater<br />
(Degteva, 1991).<br />
Research is in progress that examines the progeny of<br />
irradiated mothers for stillborn children, abortions, and miscarriages.<br />
Analyses thus far indicate a significantly greater number of birthing<br />
complications in the irradiated mothers than existed among the control<br />
groups (Degteva, 1991).<br />
Today, near the Kyshtym reservation, where the town<br />
of Kasli used to be, the soil still contains from 1000 to 2000 Ci/km 2 of<br />
9 0 Sr (Medvedev, 1991). It was from this area that 1,154 previous residents<br />
of Kasli were rapidly evacuated in a 7-10 day period following the accident.<br />
Several types of military training maneuvers are conducted in these<br />
contaminated areas today (Douplenski, 1991).<br />
In addition to the ionizing radiation doses the victims of these three<br />
radiation accidents received, no humanitarian support existed, the people<br />
lived on a less than a well-balanced diet, and there was a pervasive lack of<br />
medical attention and equipment (Degteva, 1991).<br />
There were and still are<br />
only 50 beds to care for the 500,000 irradiated people from the three<br />
accidents (Kosenko, 1991).<br />
Russian Ministry of Health officials in the Chelyabinsk Branch Office<br />
wou I like AFRRI as a research partner to perform some collaborative<br />
research (Kosenko, 1991).<br />
They perceive AFRRI as being able to assist them<br />
in increasing the accuracy of their dosimetry and in the creation of an<br />
interactive computer register. An increased knowledge regarding the late<br />
effects of ionizing radiation would result from this collaborative effort.<br />
9
Specifically, this would assist in determining the:<br />
1) long-term, 2) lowdose,<br />
3) immunologic, and 4) genetic alterations associated with exposure to<br />
ionizing radiations (Akleev, 1991).<br />
It would also provide insight regarding<br />
the effects that exposure to ionizing radiation has on subsequent<br />
generations (child birth, miscarriages, abortions, etc.).<br />
PSYCHOLOGICAL ASPECTS OF NUCLEAR ACCIDENTS<br />
It is noteworthy that people living in the Kyshtym/Chelyabinsk area are<br />
very antinuclear as evidenced by the voluntary shutdown of two nuclear<br />
electrical generation reactors during June-July 1989 (Akleev, 1991;<br />
Medvedev, 1991).<br />
Also, although billions of rubles (equivalent to millions<br />
of U.S. dollars) had been spent on tht<br />
construction of a breeder reactor, it<br />
too was shut down approximately 3 years ago, an aftermath of the Che.nobyl<br />
accident in 1986 (Douplenski, 1991; Medvedev, 1991).<br />
The reason given for<br />
shutting down both of these "nuclear" facilities after years of use and<br />
construction was the psychological animosity that existed in the people<br />
living in this contaminated region (Douplenski, 1991).<br />
A similar<br />
psychologically induced result occurred in Mcscow, where a new, ready to be<br />
used nuclear power plant was prevented from opening due to public outcry<br />
(Douplenski, 1991).<br />
In addition, a nuclear power plant approximately 40 km<br />
south of St. Petersburg was closed to appease the psycholo_, tally upset<br />
populace (Karpov, 1991).<br />
The impact that individual perceptions of ionizing<br />
radiation have on society is now being realized (IAEA Report, 1991).<br />
10
In other parts of the FSU, specifically around the contaminated<br />
<strong>Chernobyl</strong> region, 35,000 people in the Belorus city of Gomel went on strike<br />
on 26 April 1991 to protest the fourth anniversary of <strong>Chernobyl</strong>.<br />
Similarly,<br />
60,000 people waving "nationalist" flags packed the square in front of the<br />
Sofia Cathedral in Kiev, the capital of the Ukraine, and demanued punishment<br />
for those responsible for the world's worst nuclear accident (Bohlen, 1987,<br />
Reuters, 1990).<br />
The Rukh press agency reported similar antinuclear<br />
demonstrations in Kiev, the western Ukranian city of Lvov, and in the<br />
Belorus city of Minsk, as well as demonstrations elsewhere in the two<br />
republics that were the worst hit by fallout from the accident (Reuters,<br />
1990).<br />
Reasons for the psychological outcry among the <strong>Chernobyl</strong> ,ictims are<br />
numerous.<br />
The delays caused by scientific and political discussions finally<br />
resulted in the evacutation of 40,000 residents from an area where the<br />
contamination was 40 Ci/km 2 . Furthermore, 10,000 people receiving 15-20<br />
Ci/km 2 and 60,000 people receiving 5-15 Ci/km 2 were not allowed to relocate<br />
(Kedrovsky, 1991).<br />
In addition, as part of the <strong>Chernobyl</strong> cleanup effort,<br />
numerous people are employed in 2 week (on/off) shifts inside the 6 km "Hot<br />
Zone" (Shishchits, 1991).<br />
It is expected that the rate of thyroid cancer is<br />
5 to 10 times the rate expected for 1.5 million Soviet citizens, leukemia<br />
rates among children in some areas of the Ukraine are 2 to 4 times normal<br />
levels, and the death rate for peopie working in the <strong>Chernobyl</strong> plant since<br />
the accident is 10 times what it was before the accident (Barringer, 1990).<br />
Furthermore, the scientific director of the zone surrounding the damaged<br />
11
<strong>Chernobyl</strong> power station estimated that the disaster has currently claimed<br />
over 7,000 lives (Wise, 1991).<br />
Perhaps the worst aspect of these nuclear accidents is the omnipresent<br />
invisible threat and the continuing fear that the future is marred by<br />
irreversible cancer or genetic defects. This may have increased since the<br />
accidents, when radioactive fallout contaminated the environment, animals,<br />
and people. An undeniable and continual reminder for the residents is that<br />
all the timber in the affected areas is radioactively contaminated and<br />
cannot be used for furniture, for construction, or even for firewood<br />
(Matukousky, 1990).<br />
In addition to the forests, the waters of the Pripyat, Sozh, Nevsich,<br />
Iput, Besyoad, Braginka, Kolpita, and Pokot Rivers are carrying radioactive<br />
silt into the Dnepr River. The entire grid of power stations on the Dnepr<br />
River down to the Black Sea is threatened with 60 million tons of<br />
radioactive silt, as are the 60 million people in these regions.<br />
The anticipatory stress of these people is readily apparent and can be<br />
easily understood. Interestingly, according to an assessment by 200<br />
scientists from 25 countries and 7 multinational organizations done for the<br />
United Nations International Atomic Energy Agency; stress-related illnesses<br />
are caused by lack of public information about the disaster and the mass<br />
evacuations that follow (I.A.E.A., 1991).<br />
In sum, the psychological stress<br />
that followed in the surrounding areas outside the radioactive hot zones was<br />
12
*wholly disproportionate to the biological significance of the radioactive<br />
contamination" (I.A.E.A., 1991).<br />
Even when no radiation is released from a nuclear accident, but only<br />
threatens, as was the case at Three Mile Island, the Kemeny Commission and<br />
other documents (Collins, 1984; Collins, 1991; Davidson, 1982, I.A.E.A.,<br />
1991) concluded that mental stress would be the main effect. The<br />
psychological findings could be criticized if used alone, for their<br />
potential self-serving function. To avoid this perception, neurochemical<br />
analyses that measured individual stress values were employed (Collins,<br />
1983; Collins, 1991; Davidson, 1983). By using this multidisciplinary<br />
approach, we further clarified the adverse effects that exposure, or<br />
potential exposure to ionizing radiation has on humans. Consequently, since<br />
increased stress and associated behavioral alterations occur from<br />
anticipation as occurred at TMI; when actual exposure and deaths occur from<br />
radiation, the psychological and behavioral actions of the FSU residents are<br />
easily understood.<br />
Consequently, the situation in the FSU is likely to become of major<br />
concern in future years as the future of 60 million people is adversely<br />
affected by the <strong>Chernobyl</strong> accident, and another million people (+/-) are<br />
adversely affected by the three accidents in the Kyshtym and Chelyabinsk<br />
areas. The psychological, physiological and epidemiological implications of<br />
these disasters require further study.<br />
13
2:149-166, 1983.<br />
Collins, D. L. Persistence differences between Three Mile Island residents and<br />
a control group. AD-A145567, National Technical Information Service,<br />
Springfield, VA, 1984.<br />
Davidson, L. M., Baum, A., and Collins, D. L. Stress and control related<br />
problems sat Three Mile Island. Journal of Applied Psychology<br />
12:349-359, 1982.<br />
Pegteva, M. 0. Vice Director, Candidate of Technical Sciences, Institute<br />
of Biophysics of the USSR, Ministry of Health, Chelyabinsk Branch Office.<br />
Personal Conversation, November 1991.<br />
Douplenski, N. First Secretary, Commission of the USSR for UNESCO,<br />
Ministry of Foreign Affairs of USSR. Personal Conversation, September 1991.<br />
(He grew up in this region and still has family living in the area.)<br />
International <strong>Chernobyl</strong> Project, The, Technical Report.<br />
ISBN 92-0-129191-4, Int<br />
ernational Atomic Energy Agency, pp 277-413, 1991.<br />
Karpov, V. I. Director of Research, Academy of Sciences of the USSR, Center for<br />
International, Environmental Cooperation (INENCO), St. Petersburg, Russia.<br />
Personal Conversation, September 1991.<br />
15
Kedrovsky, 0. L. Professor, Director of the Institute, All-Union designing<br />
and research institute of production engineering.<br />
Personal Conversation,<br />
September 1991.<br />
Kosenko, M. M. Head of Clinical <strong>Department</strong>, Candidate of Medical Sciences,<br />
Institute of Biophysics of the USSR, Ministry of Health, Chelyabinsk Branch<br />
Office. Personal Conversation, November 1991.<br />
Lee, G. A. <strong>Chernobyl</strong> trial opens; accused officials blame design faults.<br />
The Washington Post, July 8, 1987, p. Al.<br />
Matukovsky, N. The Lessons of <strong>Chernobyl</strong>. Izvestia, Moscow, March 26, 1990,<br />
p. 3.<br />
Medvedev, Z. Bringing the skeleton out of the closet.<br />
<strong>Nuclear</strong> Engineering<br />
International 35(436):26-32, 1991.<br />
<strong>Nuclear</strong> Regulatory Commission:<br />
Investigation into the March 1979 Three Mile<br />
Island-2 <strong>Accident</strong> by the Office of Inspection and Enforcement, NUREG-0600.<br />
<strong>Nuclear</strong> Regulatory Commission, Washington, DC, 1979.<br />
Reuters News Service. <strong>Chernobyl</strong> rally attended by thousands. The New York<br />
Times, New York, April 27, 1990, p. A6.<br />
Shishchits, I. Leading Researcher, Scientific Secretary of Section of<br />
16
Scientific-Technical Council, USSR <strong>Nuclear</strong> Society, All-Union Design and<br />
Research Institute of Industrial Technology.<br />
Personal Conversation,<br />
September 1991<br />
Soyfer, V. N. Chairman Laboratory of Molecular Genetics, George Mason<br />
University, Fairfax, VA, Personal Conversation, November 1991,<br />
Tsaturor, Y. S. Duputy chairman, USSR State committee for Hydrometerology.<br />
Personal Conversation, September 1991.<br />
Wise, M. Z. U.N. report blames stress, not radiation, for <strong>Chernobyl</strong> illnesses.<br />
The Washington Post, Washington, DC, May 22, 1991, p. A25.<br />
17
ADDENUM 1<br />
During the course of these conversations at AFRRI, Professor Valery N.<br />
Soyfer, Chairman of the Laboratory of Molecular Genetics at George Mason<br />
University, Fairfax, VA, and a recent defector from the Soviet Union, said,<br />
"The Soviet Union was the first nation to develop the hydrogen bomb."<br />
then looked at me and said, "I bet you didn't know that, but we were."<br />
He<br />
So I<br />
said, "Well can you tell me about that?"<br />
and he said he couldn't say any<br />
more on that topic.<br />
ADDENDUM 2<br />
Dr. Alexander V. Akleev, Institute of Biophysics, U.S.S.R. Ministry of<br />
Health, Chelyabinsk Branch Office, summarized their research areas.<br />
They<br />
have been looking for a radioprotective drug for the last 15 years, but have<br />
not found anything that has worked. This unsuccessful effort involved<br />
looking at spleen extracts, compounds in herbs with low molecular weights,<br />
and peptides to defuse through the cell membrane. Reactor operators have<br />
been irradiated since they began their research back in the 1948-1950. The<br />
Institute is interested in late effects of low and medium doses; low doses<br />
are defined as 20 rads or less, and medium doses are 20 rads or greater.<br />
18
NORTH<br />
FIGURE<br />
Kfis/dn A'ccadett<br />
*Szire6&sk A 19z<br />
LEGENDri<br />
inMie
AD-A284 746<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency<br />
Alexandria, VA 22310-3398<br />
DNA-TR-92-37-V1<br />
<strong>Chernobyl</strong> Doses<br />
Volume 1-Analysis of Forest Canopy Radiation Response<br />
from Multispectral Imagery and the Relationship to Doses<br />
DTIC<br />
Gene E. McClellan<br />
George H. Anno<br />
A ELECTE<br />
F. Ward Whicker<br />
Pacific-Sierra Research Corp. SEP 2 01994<br />
2901 28th Street<br />
Santa Monica, CA 90405-2938 F:<br />
September 1994<br />
Technical Report<br />
VOriginal oontains oolor<br />
plates; All DTIC reproduot,.<br />
ions wuil be in bleaok an/<br />
whites<br />
CONTRACT No. DNA 001-87-C-0104<br />
94-30186 Approved for public release;<br />
~II ~<br />
unlimited.<br />
II<br />
!111 l! t Iill Idistribution Is<br />
94 W'%= D<br />
I2
Destroy this report when it is no longer needed. Do not<br />
return to sender.<br />
PLEASE NOTIFY THE DEFENSE NUCLEAR AGENCY,<br />
ATTN: CSTI, 6801 TELEGRAPH ROAD, ALEXANDRIA, VA<br />
22310-3398, IF YOUR ADDRESS IS INCORRECT, IF YOU<br />
WISH IT DELETED FROM THE DISTRIBUTION LIST, OR<br />
IF THE ADDRESSEE IS NO LONGER EMPLOYED BY YOUR<br />
ORGANIZATION.<br />
.0
DISTRIBUTION LIST UPDATE<br />
This mailer is provided to enable DNA to maintain current distribution lists for reports• (We would<br />
appreciate your providing the requested information.)<br />
NOTE:<br />
1l Add the individual listed to your distribution list. Please return the mailing label from the<br />
document so that any additions, changes,<br />
F1 Delete the cited organization /individual,<br />
corrections or deletions can be made easily.<br />
For distribution cancellation or more<br />
El Change of address. information call DNA/IMAS (703) 325-1036.<br />
NAME:<br />
ORGANIZATION:<br />
OLD ADDRESS<br />
CURRENT ADDRESS<br />
z<br />
I-<br />
Z<br />
DEFENSE NUCLEAR AGENCY<br />
ATTN: IMAS<br />
6801 TELEGRAPH ROAD<br />
ALEXANDRIA, VA 22310-3398<br />
DEFENSE NUCLEAR AGENCY<br />
ATTN: IMAS<br />
6801 TELEGRAPH ROAD<br />
ALEXANDRIA, VA 22310-3398
DISCLAIMER NOTICE<br />
THIS DOCUMENT IS BEST<br />
QUALITY AVAILABLE. THE COPY<br />
FURNISHED TO DTIC CONTAINED<br />
A SIGNIFICANT NUMBER OF<br />
COLOR PAGES WHICH DO NOT<br />
REPRODUCE LFGIBLY ON BLACK<br />
AND WHITE MICROFICHE,.
REPOT DCUMNTATON AGEFor mApproved<br />
REPO UME T DO TATIN PA E OMB Noý 0704-0188<br />
Public reporting burden for this collection of information is estimated to average 1 hour per response including the time for reite.irng instructon*, seatcnrng exsting data soufces,<br />
gathering and maintaining the data needed, and completing and reviewing the collection of information Send comments regarding this burden estimate or any other aspect ot mst<br />
collection of information, including suggestions for reducing this burden, to Washington headquarters Services Directorate for informaton Operations and <strong>Reports</strong>. 1215 Jefferson<br />
Davis Highway. Suite 1204, Arlington, VA 22202-4302. and to the Office of Management and Budget, Pal.orvrrk Reduction Project (0734 -0l88), Washitngton, DC 2060<br />
1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED<br />
940901 Technical 870929- 930930<br />
4. TITLE AND SUBTITLE 5. FUNDING NUMBERS<br />
<strong>Chernobyl</strong> Doses<br />
Volume 1-Analysis of Forest Canopy Radiation Response from Multispectral C - DNA 001-87-C-0104<br />
Imagery and the Relationship to Doses PE -62715H<br />
"6. AUTHOR(S)<br />
PR - RM<br />
TA - RH<br />
Gene E. McClellan, George H. Anno, and F. Ward Whicker<br />
WU - DH026130<br />
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION<br />
REPORT NUMBER<br />
Pacific-Sierra Research Corp.<br />
2901 28th Street PSR Report 2251<br />
Santa Monica, CA 90405-2938<br />
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING<br />
AGENCY REPORT NUMBER<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency<br />
6801 Telegraph Road DNA-TR-92-37-Vl<br />
Alexandria, VA 22310-3398<br />
RAEM/Kehlet<br />
11. SUPPLEMENTARY NOTES<br />
This work was sponsored by the <strong>Defense</strong> <strong>Nuclear</strong> Agency under RDT&E RMC Code B4662D RM RH 00038<br />
STRP 3500A 25904D.<br />
12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE<br />
Approved for public release; distribution is unlimited.<br />
13. ABSTRACT (Maximum 200 words)<br />
This volume of the report <strong>Chernobyl</strong> Doses presents details of a new, quantitative method for remotely sensing<br />
ionizing radiation dose to vegetation. Analysis of Landsat imagery of the area within a few kilometers of the<br />
<strong>Chernobyl</strong> nuclear reactor station provides maps of radiation dose to pine forest canopy resulting from the accident<br />
of April 26, 1986. Detection of the first date of significant, persistent deviation from normal of the spectral<br />
reflectance signature of pine foliage produces contours of radiation dose in the 20 to 80 Gy range extending up to<br />
4 km from the site of the reactor explosion. The effective duration of exposure for the pine foliage is about 3<br />
weeks. For this exposure time, the LD 50 of Pinus sylvestris (Scotch pine) is about 23 Gy. The practical lower<br />
dose limit for the remote detection of radiation dose to pine foliage with the Landsat Thematic Mapper is about 5<br />
Gy or 1/4 of the LD 5 0.<br />
'DTic QU~ALI77 ',,:-L: - ,.,<br />
14. SUBJECT TERMS 15. NUMBER OF PAGES<br />
<strong>Chernobyl</strong> Dose 226<br />
Forest Damage Conifer Stress Fallout 16. PRICE CODE<br />
Change Detection Ionizing Radiation Multispectral Imagery<br />
17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19, SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT<br />
OF REPORT OF THIS PAGE OF ABSTRACT<br />
UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED SAR<br />
NSN 7540-280-5500<br />
Standard Form 298 (Rev.2-89)<br />
Prescribed by ANSI Sta 239-18<br />
290-102
UNCLASSIFIED<br />
SECURi'Y CLASSIFICAnION OF THiS PAGE<br />
CLASSIFIED BY:<br />
N/A since Unclassified.<br />
DECLASSIFY ON:<br />
N/A since Unclassified.<br />
Accesion For<br />
NTIS CRA&i<br />
D TIG TAj3<br />
Unan n-ouliced 0<br />
JUSltifca~ioin<br />
...............<br />
*<br />
By<br />
Ava 1 Idbiiily Cod'es<br />
Dist Aviat & ,ndl/or<br />
Special<br />
MII<br />
SECURITY CLA SSIVICATIO'N OF THIS PACG<br />
UNCLASSIFIED
EXECUTIVE SUMMARY<br />
This volume of the report <strong>Chernobyl</strong> Doses presents details of a new, quantitative method for<br />
remotely sensing ionizing radiation dose to vegetation.<br />
The method uses a time series of<br />
multispectral images taken from an orbital or airborne platform to reveal changes in spectral<br />
reflectance of foliage that has been exposed to radiation. The threshold of detection is about 1/4 of<br />
the median lethal dose (LD 5 0 ) of the dominant plant species.<br />
Figure S-1 illustrates the sequence of events leading to detection of radiation dose. The<br />
effective duration of the vegetation exposure is determined by the radioactive decay rate of the<br />
fallout deposited on the foliage and the rate at which fallout is removed from the foliage by<br />
weathering. The accumulated dose during this exposure time causes biological damage at the<br />
cellular level in plant tissues. At high doses (several times the LD 5 0 ), radiation damage to multiple<br />
tissues kills vegetation in a short time, a matter of one or two weeks for pine trees. At lower<br />
doses, radiation damage to foliage is significant only for growth tissue at the tips of branches.<br />
Such damage takes more time to change the appearance of foliage. The dose dependence of the<br />
delay until the onset of observable foliage response provides the basis for remote detection of<br />
radiation dose.<br />
Figure S-2 shows a map of the foliage doses received by the pine forest canopy near the site of<br />
the <strong>Chernobyl</strong> nuclear power station as derived from analysis of a time series of eleven Landsat<br />
Thematic Mapper images spanning a period from one year before to two years after the explosion<br />
of the Unit 4 reactor on April 26, 1986. Table S- 1 shows the doses for the three contours drawn<br />
on the map as well as the dose range between contours and the total area within each contour.<br />
Individual pixels, representing 25 m squares of pine forest, are color-coded according to the<br />
legend on the map, which shows the image number of first-observed, persistent deviation from<br />
normal of the spectral signature of the pixel. Table S-2 provides a conversion of this image<br />
number to a radiation dose range. The dose is quoted as a range since the first observable response<br />
may have occurred at any time during the interval between the date of the image showing first<br />
response and the date of the previous image. Black areas on the map are either not pine forest or<br />
were cleared of pine forest before showing a radiation response.<br />
Gray pixels on the map<br />
correspond to pine forest that appears normal at the end of the two year observation period.<br />
Although not indicated on the map, some of these pixels showed a transient radiation response<br />
corresponding to doses less than about 20 Gy.<br />
The methodology developed during this effort and the resulting data contribute to an improved<br />
understanding of the effects of high levels of fallout radioactivity on vegetation and, especially, on<br />
the remote observation of radiation-induced foliage response and the extraction of dose estimates<br />
from those ohservations. The results aid in the understanding of the consequences to personnel of<br />
ini
Falu<br />
Deposition<br />
ProtractedFalu<br />
_ Radiation<br />
Exposure<br />
Removal by<br />
of Foliage<br />
Efet at the<br />
Early, Systemic Continuum Late,<br />
Effects at .,K- -_ -- > Growth-Related<br />
Higher Doses of Responses Effects at Lower<br />
Doses<br />
Remote Sensing<br />
Figure S-1.<br />
Sequence of events that enables remote detection of the exposure of<br />
vegetation to ionizing radiation and calculation of the dose from the<br />
time of resulting foliage changes relative to the start of exposure.<br />
iv
Figure S-2. <strong>Chernobyl</strong> radiation dose contours derived from foliage changes in pine forest canopy<br />
observed with the Landsat Thematic Mapper multispectrai imaging device; 1 kan grid lines<br />
originating at the Unit 4 reactor site. See Table S- I for dose values.<br />
V
in an area contaminated by radioactive fallout and the impact of vegetation on that<br />
Table S-1. Description of contours drawn in Figure S-2.<br />
Pine foliage response Dose range Cumulative<br />
time at contour Dose at contour within contour enclosed area<br />
ontour (days) (Gy) (Gy) (kin 2 )<br />
inner (blue) 35 54 54 - 80 0.8<br />
Middle (green) 172 30 30- 54 2.9<br />
Outer (yellow) 499 20 20 30 14.5<br />
The final section of this report lists recommendations for improving and extending the results<br />
presented herein. The recommendation of primary importance is the establishment of a cooperative<br />
effort with scientists of the former Soviet Union to compare the satellite data with ground studies<br />
made at specific locations within a few kilometers of the <strong>Chernobyl</strong> power plant. The comparison<br />
of the image analysis with other data in Section 8 is encouraging but only qualitative because of the<br />
unavailability of data with the temporal and spatial resolution of the satellite images. We believe<br />
better data exists (Gamache, 1993); furthermore, some of the affected forest is probably still<br />
standing. Histological examination of exposed pine trees observed in the imagery would be of<br />
great value both for interpreting the imagery and interpreting the histological data.<br />
Table S-2.<br />
Pine foliage doses corresponding to first detected response at the times of<br />
the 9 postaccident images (Images 1 and 2 are preaccident).<br />
Time postaccident Dose<br />
Image number (days) (Gy)<br />
3 3 >133<br />
4 12 80- 133<br />
5 28 59-80<br />
6 35 54-59<br />
7 172 30-54<br />
8 220 28-30<br />
9 380 23-28<br />
10 499 21-23<br />
11 763 18-21<br />
vi
PREFACE<br />
This volume is the first of three volumes composing the final report to the <strong>Defense</strong> <strong>Nuclear</strong><br />
Agency (DNA) for contract DNAOOI-87-C-0104, <strong>Chernobyl</strong> Doses. In addition to summarizing<br />
investigations carried out by Pacif~c-Sierra Research Corporation (PSR) under that contract, this<br />
volume presents the analytical work that connects satellite multispectral observations of pine forests<br />
around <strong>Chernobyl</strong> to the nuclear radiation dose received by the trees as a censequence of the<br />
reactor accident of 26 April 1986. Volume 2, Conifer Stress Near Cherno'?yl Derived from<br />
Landsat Imagery, describes the acquisition and processing of Landsat imagery of the area<br />
containing the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station and presents the exploratory analysis of the<br />
imagery using PSR's proprietary HyperscoutTl change detection algorithm. Volume 3, Habitat<br />
and Vegetation Near the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station presents a detailed exposition on the<br />
soils, climate, and vegetation of the Poles'ye region of the Ukraine and Belorussia with emphasis<br />
on the area around <strong>Chernobyl</strong>. This data provides background for interpretation of the satellite<br />
imagery.<br />
The authors wish to acknowledge Frank Thomas of PSR who pointed out the possibility of<br />
observing radiation effects on vegetation near <strong>Chernobyl</strong> with multispectral imagery. They also<br />
wish to recognize the considerable computational expertise of Leigh Matheson of PSR who<br />
implemented the image processing and analysis described in this report and the skillful manuscript<br />
preparation by Kathy Howell. Dr. Gerald Gamache was helpful in providing data he obtained in<br />
Ukraine.<br />
The authors wish to acknowledge the technical monitor of this project, Robert W. Young of<br />
DNA's Radiation Policy Division, for his support and encouragement during this work.<br />
Dr. Young was assisted first by Major Bruce West and then by Major Robert Kehlet. The authors<br />
also wish to acknowledge Dr. Marvin Atkins and Dr. David Auton of DNA whose interest made<br />
this work possible.<br />
rMHyperscout is a trademark of Pacific-Sierra Research Corporation.<br />
vii
CONVERSION TABLE<br />
Conversion factors for U.S. customary to metric (SI) units of measurement<br />
To Convert From To Multiply<br />
angstrom meters (m) 1.000 000 X E-I0<br />
atmosphere (normal) kilo pascal (kPa) 1.013 25 X E.2<br />
bar kilo pascal (kPa) 1.000 000 X E÷2<br />
barn meter 2 (M 2 ) 1 000 000 X E-28<br />
British Thermal unit (thermochemical) joule (J) 1.054 350 X E+3<br />
calorie (thermochemical) joule (J) 4.184 000<br />
cal (thermochemical)/cm 2 mega joule/m 2 (MJ/m 2 ) 4.184 000 X E-2<br />
curie giga becquerel (GBq)l 3.700 000 X E+I<br />
degree (angle) radian (rad) 1.745 329 X E-2<br />
degree rahrenheit degree kelvin (K) tK=(t0f + 459.67)/1 .8<br />
electron volt joule (j) 1 .02 19 X L-19<br />
erg joule (i) 1.000 000 X E-7<br />
erg/second watt (W) 1.000 000 X E-"'<br />
foot meter (m) 3.048 000 X E-I<br />
foot-pound-force joule (J) 1.355 818<br />
gallon (U.S. liquid) meter 3 (, 3 ) 3.785 412 X E-3<br />
inch meter (m) 2.540 000 X E-2<br />
jerk joule (J) 1.000 000 X E+9<br />
joule/kilogram (J/Kg) (radiation dose<br />
absorbed) Gray (Gy) 1.000 000<br />
kilotons terajoules 4.183<br />
kip (1000 lbf) newton (N) 4.448 222 X E+3<br />
kip/inch 2 (ksi) kilo pascal (kPa) 6.894 757 X E+3<br />
ktap newton-second/mr 2 (N-s/m 2 ) 1.000 000 X E+2<br />
micron meter (m) 1.000 000 X E-6<br />
rail meter (M) 2.540 000 X E-5<br />
mile (international) meter (m) 1.609 344 X E+3<br />
ounce kilogram (kg) 2.834 952 X E-2<br />
pound-force (lbf avoirdupois) newton (N) 4.448 222<br />
pound-force inch o'ton-meter (N'm) 1.129 848 X E-I<br />
pound-force/inch newton/meter (N/m) 1.751 268 X E+2<br />
pound-force/foot 2 kilo pascal (kPa) 4.788 026 X E-2<br />
pound-force/inch 2 (psi) kilo pascal (kPa) 6.894 757<br />
pound-mass (Ibm avoirdupois) kilogram (kg) 4.535 924 X E-I<br />
pound-mass-foot 2 (moment of inertia) kilogram-meter 2 (kg-m 2 ) 4.214 011 X E-2<br />
pound-mass/foot kilogram/metr3 (kg/rn 3 ) 1.601 846 X E+1<br />
rad (radiation dose absorbed) Gray (Gy)°" 1 .000 000 X E-2<br />
roentgen coulomb/kilogram (C/kg) 2.579 760 X E-A<br />
shake second (s) 1.000 000 X E-S<br />
slug kilogram (kg) 1.459 390 X E+1<br />
torr (mm Hg, 0 0 C) kilo pascal (kPa) 1.333 22 X E-1<br />
*The becquerel (Bq) is the SI unit of radioactivity; Bp<br />
"*The Gray (Gy) is the SI unit of absorbed radiation.<br />
1 event/s.<br />
viii
CONTENTS<br />
Section<br />
SUM M A R Y ............................................................ ........<br />
PREFACE ...............................................<br />
CONVERSION TABLE ........................................................<br />
F IG U R E S ........................................................................<br />
TAB LES .................................................................. ... xvi<br />
Page<br />
vii<br />
viii<br />
ix<br />
I INTRODUCTION ............................................................... I<br />
2 REMOTE FALLOUT DETECTION THROUGH IMAGERY ANALYSIS 5<br />
3 HABITAT AND VEGETATION NEAR THE CHERNOBYL<br />
NUCLEAR POWER PLANT ............................................... 11<br />
3.1 Regional Description ................................................ . 11<br />
3.1.1 G eography ................................................... . 11<br />
3.1.2 Soils ........................................................... 14<br />
3.1.3 C lim ate .................................................. ..... 15<br />
3.1.4 Native Vegetation ........................................... 16<br />
3.1.5 C rops ......................................................... 19<br />
3.2 Tasseled Cap Transformation of Landsat Imagery ................ 19<br />
3.2.1 Description of the Tasseled Cap Transformation ........ 19<br />
3.2.2 Tasseled Cap Images of the Analysis Area ............... 23<br />
3.3.2 Unsupervised Clustering of Evergreen Pixels ........... 38<br />
3.3 Preliminary Classification of Evergreens .......................... 35<br />
3.3. 1 Selection of Evergreen Pixels .............................. 35<br />
3.3.2 Unsupervised Clustering of Evergreen Pixels ........... 38<br />
3.3.3 Identification of Evergreen Classes ............. 41<br />
3.4 Reference Sites for Pine Forest Classes ............................. 43<br />
3.4.1<br />
3.4.2<br />
Avoidance of Clouds and Haze ...........................<br />
Representative Spectral Signatures .......................<br />
45<br />
58<br />
3,. Sites C hosen ................................................. 58<br />
3.4.4 Other Impacts of Clouds and Haze ....................... 58<br />
3.5 Preaccident Pine Forest Classification Map ................... 59<br />
4 RADIONUCLIDE FALLOUT FROM THE CHERNOBYL ACCIDENT 65<br />
4.1 Release And Deposition ............................................... 65<br />
4.2 Radionuclide Composition ........................................... 65<br />
5 RADIOBOTANICAL AND DOS IMETRIC CONSIDERATIONS ......... 67<br />
5.1<br />
5.2<br />
Tree Characteristics Relevant To Radiation Damage ..............<br />
Beta Radiation Exposure Of Pine Meristems .......................<br />
67<br />
69<br />
5.3 Beta Versus Gamma Exposure ....................................... 70<br />
5.3. i Values from the Literature .................................. 70<br />
ix
CONTENTS (CONTINUED)<br />
Section<br />
Page<br />
5.3.2 Calculations .................................................. 71<br />
5.5<br />
Dose Rate Effects .....................................................<br />
Dose Rate Scenario for <strong>Chernobyl</strong> ..................................<br />
78<br />
79<br />
5.6 Sum m ary ............................................................... . 84<br />
6 DOSE-RESPONSE RELATIONSHIPS FOR PINE TREES ............... 87<br />
6.1 Literature Review ...................................................... 87<br />
6.2 Analysis Of Dose Rate Data .......................................... 91<br />
6.3 Temporal Progression of Damage .................................. 93<br />
6.4 Dose Versus Time-to-Response for Multispectral Detection ...... 94<br />
7 DOSE DETERMINATION FOR CHERNOBYL FORESTS ............... 103<br />
7.1 Spectral Deviation from Class ........................................ 103<br />
7.1.1 Mahalanobis Distance and the Mahalanobis Spectral<br />
Deviation Vector ............................................. 104<br />
7.1.2 Scaled Mahalanobis Distances and Vectors ............... 104<br />
7.1.3. Spectral Deviations for the <strong>Chernobyl</strong> Images ........... 108<br />
7.2 Deviations Caused By Forest Clearing .............................. 113<br />
7.3 Time-to-response from Imagery ..................................... 115<br />
7.4 Foliage Dose M aps .................................................... 120<br />
7.4.1 Dose Map with Hand-Drawn Contours ................... 120<br />
7.4.2 Iterative Smoothing of the Dose Map ..................... 122<br />
7.5 S um m ary ................................................................ 129<br />
8 D ISC U S SIO N ..................................................................... 131<br />
8.1 Comparison with Measurements of Forest Damage .............. 131<br />
8.2 Comparison with Aerographic Surveys of Dose Rate ............. 132<br />
8.2. 1 1-Meter Gamma Dose Rates for the Close-in Area ...... 132<br />
8.2.2<br />
8.2.3<br />
Close-in Foliage and 1-Meter Gamma Doses ............<br />
Calculated Ratio of Foliage to 1-Meter Gamma Dose...<br />
135<br />
137<br />
8.3 Comparison with Human Exposure Data ........................... 139<br />
8.4 Uncertainties in the Image Analysis ................................. 141<br />
9 CONCLUSIONS AND RECOMMENDATIONS ............................ 143<br />
10 LITERATURE CITED .................................... 145<br />
Appendix<br />
A FALLOUT DOSE CALCULATIONS FOR PINE FOREST CANOPY... A-1<br />
B SPECTRAL DEVIATIONS FROM CLASS .................................. B-1<br />
C CHRONOLOGY OF SELECTED EVENTS .................................. C-1<br />
D EVERGREEN SPECTRAL SIGNATURES BY CLASS AND DATE .... D-1<br />
x
FIGURES<br />
Figure<br />
S-i<br />
Sequence of events which enables remote detection of the exposure<br />
of vegetation to ionizing radiation and calculation of the dose from<br />
the time of resulting foliage changes relative to the start of exposure ........<br />
Page<br />
iv<br />
S-2 <strong>Chernobyl</strong> radiation dose contours derived from pine forest canopy<br />
responses as observed with the Landsat Thematic Mapper multispectral<br />
imaging device; i km grid lines originating at the Unit 4 reactor site ........<br />
V<br />
1-1 Immediate vicinity of the <strong>Chernobyl</strong> nuclear power station with Unit 4<br />
reactor (circled) still burning .................................................... 3<br />
2-1 Typical coloration of dying pine foliage. Landscape setting ................. 6<br />
2-2 Forest stress map for 8 May 1986 as presented in Volume 2.<br />
<strong>Accident</strong>-affected area circled. White cross is Unit 4 reactor site ............ 8<br />
2-3 Method of radiation dose determination from remote multispectral<br />
sensing of pine forest ................................................................ 9<br />
3-1 Regional map encompassing the 30 kilometer danger zone around the<br />
<strong>Chernobyl</strong> nuclear accident. (Reference ..... ) ................................ 12<br />
3-2 Landsat image of the confluence of the Dnieper and Pripyat Rivers<br />
at the upper end of the Kiev Reservoir. The sharply outlined black<br />
footprint shape is the reactor station cooling pond along the Pripyat<br />
River. [Thematic Mapper false color presentation (R,G,B) = (7,4,1),<br />
6 June 1985, geocoded, 1 degree longitude by 0.5 degree latitude.] ....... 13<br />
3-3 Illustration of the Tasseled Cap spectral transformation for Landsat<br />
Thematic Mapper images; transformation coefficients are listed<br />
in T able 3-2 ......................................................................... 2 1<br />
3-4 Tasseled Cap false color images, (R,G,B) = (Br,Gr,Wt) of 38.4 km<br />
square area around Chemobyl: a) early summer image one year<br />
preaccident and b) late winter image 5 weeks preaccident .................... 24<br />
3-5 Tasseled Cap false color images, (R,G,B) = (Br,Gr,Wt) of 38.4 km<br />
square area around <strong>Chernobyl</strong>: a) 3 days postaccident and b) 12 days<br />
postaccident ........................................................................ . 25<br />
3-6 Tasseled Cap false color images, (R,G,B) = (Br,Gr,Wt) of 38.4 km<br />
square area around Chemobyl: a) 4 weeks postaccident and b) 5 weeks<br />
postaccident ......................................................................... 27<br />
xi
FIGURES (CONTINUED)<br />
Figui-e<br />
Page<br />
3-7 Tasseled Cap false color images, (R,G,B) = (Br,Gr,Wt) of 38.4 km<br />
square area around <strong>Chernobyl</strong>: a) 5.6 months postaccident<br />
and b) 7.2 months postaccident .................................................. 29<br />
3-8 Tasseled Cap false color images, (R,G,B) = (Br,Gr,Wt) of 38.4 km<br />
square area around <strong>Chernobyl</strong>: a) I year postaccident and b) 1.4 years<br />
postaccident ......................................................................... 31<br />
3-9 Tasseled Cap (TC) false color images of 38.4 km square area around<br />
<strong>Chernobyl</strong> 2.1 years postaccident: a) (R,G,B) = (Br,Gr,Wt) as in<br />
Figures 3-4 through 3-8 and b) (R,G,B) = (Hz,TC5,TC6), the last<br />
3 components of the TC transformation ......................................... 33<br />
3-10 Evergreen vegetation appears bright green and deciduous and<br />
annual vegetation appears magenta in this false color presentation<br />
of COMP21 with (R,G,B) = (5,2,5) ............................................ 37<br />
3-11 ?rocedure for classifying evergreen vegetation ................................ 39<br />
3-12 a) Like Figure 3-10 except color reversal (R,G,B) = (2,5,2) displays<br />
deciduous or annual vegetation in bright green and evergreens in<br />
magenta, and b) areas passing the evergreen criteria of Figure 3-11<br />
show n in green ..................................................................... 40<br />
3-13 Spectral signatures of selected classes from the unsupervised clustering<br />
of evergreen pixels ................................................................. 42<br />
3-14 Procedure for selecting one good reference site for each<br />
coniferous/evergreen class ....................................................... 44<br />
3-15 Cloud/haze enhancement for Date 1 for the 38.4 km square analysis<br />
area, 46 (R,G,B) = (Th,Th,Hz). Clouds appear blue, warm areas<br />
yellow. Thermal gradient appearing as variation in yellow shade<br />
of the cooling pond shows counterclockwise flow of water around<br />
central barrier ...................................................................... 46<br />
3-16 a) Reference (control) site locations for Classes 3, 4, 5, 6, and 8 in<br />
the 38.4 km analysis area and b) cloud/haze enhancement for Date I<br />
showing three polygons used for classification merger described in<br />
Section 3.5. Red dots are reference site locations ............................. 47<br />
xii
FIGURES (CONTINUED)<br />
Figure<br />
Page<br />
3-17 Cloud/haze enhancements for a) Date 2 and b) Date 3. Only Date 3 has<br />
clouds. Red dots are reference site locations ................................... 49<br />
3-18 Cloud/haze enhancements for a) Date 4 and b) Date 5. Only Date 5 has<br />
clouds. Red dots are reference site locations .................................. 51<br />
3-19 Cloud/haze enhancements for a) Date 6 and b) Date 7. Only Date 6 has<br />
clouds. Red dots are reference site locations ................................... 53<br />
3-20 Cloud/haze enhancements for a) Date 8 and b) Date 9. Only Date 9<br />
has clouds. Reference Sites 5, 6 and 8 have been moved on Date 9<br />
to avoid a faint jet contrail ......................................................... 55<br />
3-21 Cloud/haze enhancements for a) Date 10 and b) Date 11. Neither date<br />
has clouds, but Date 10 has a jet contrail. Red dots are reference<br />
site locations ........................................................................ 57<br />
3-22 Procedure for generating final preaccident classification of pine forest ...... 60<br />
3-23 Final preaccident classification of pine forest for the 12.8 km square<br />
area analyzed in Volume 2. Class numbers according to Table 3-6 ......... 62<br />
3-24 Final preaccident classification of pine forest for the full 38.4 km<br />
square analysis area. Class color code same as in Figure 3-23 .............. 63<br />
5-1 An example of Pinus sylvestris. A six-foot specimen cut and<br />
photographed in December in Maryland ......................................... 68<br />
5-2 Illustration of fallout distribution described by a foliar interception<br />
fraction with neglect of winds ................................................... 72<br />
5-3 Beta to gamma dose ratio versus contact source density relative<br />
to unit areal density of incident fallout. Curves are for various<br />
radii of foliage elements located at the top of the canopy ...................... 76<br />
5-4 The beta to gamma dose (dose rate) ratio at the center of cylindrical<br />
foliage elements of various radii at a) the top of the canopy and<br />
b) the middle of the canopy ....................................................... 77<br />
xiii
FIGURES (CONTINUED)<br />
Figure<br />
Page<br />
5-5 Calculated dose rates (normalized to 1.0 at t = I day) from gamma<br />
radiation emanating from fallout on the soil and from beta activity<br />
on the foliage ............................................................ ......... 82<br />
5-6 Integral of the beta dose rate curve in Figure 5-4 and a constant rate<br />
curve that should produce a similar biological response ...................... 85<br />
6-1 Relationship of LD 5 0 doses at various constant rate (CR) exposure<br />
times in hours to those for 16-hour CR exposures ............................ 88<br />
6-2 The relationship of total dose to duration of constant rate exposure<br />
for various endpoints of damage to pines ....................................... 92<br />
6-3 The time required to reach the LD100 or GR50 versus short-term<br />
(8-30 day) dose. Circles represent data for the LD 100 ....................... 95<br />
6-4 Detectability of the radiation response of pine trees using multispectral<br />
im agery ............................................................................. . 98<br />
6-5 Regression line with 68% confidence band for the relationship between<br />
dose and time of earliest detection relative to the start of exposure for<br />
two- to four-week, springtime exposures ....................................... 101<br />
7-1 Two-dimensional illustration of the cluster of points formed by the<br />
pixel intensity vectors of a forest reference site. See Appendix B for<br />
the normalization procedure used to express deviations of pixels from<br />
the cluster center in standard units ............................................... 105<br />
7-2 The cluster of scaled Mahalanobis vectors for a pine forest class<br />
reference site using the first three features of the Tasseled Cap<br />
spectral transform ation ............................................................ 107<br />
7-3 Color graphic presentation of the scaled Mahalanobis deviation vectors<br />
for pine forest on Dates I to 5 (194 by 177 pixel area) ........................ 109<br />
7-4 Color graphic presentation of the scaled Mahalanobis deviation<br />
vectors for pine forest on Dates 6 to I 1 (194 by 177 pixel area) .............<br />
IIl<br />
7-5 Forest cleared (bulldozed) after the Chemobyl accident is color coded<br />
by the date number of the first observed clearing .............................. 114<br />
xiv
FIGURES (CONTINUED)<br />
Figure<br />
Page<br />
7-6 Algorithm for determination of time-to-response for radiation-damaged<br />
foliage on a pixel-by-pixel basis .................................................. 117<br />
7-7 Date of onset of persistent deviation from normal of pine forest<br />
pixels according to algorithm of Figure 7-6 ..................................... 119<br />
7-8 Time to first observed response with 1 km grid and three hand-drawn<br />
dose contours. See Table 7-2 for dose ranges .................................. 121<br />
7-9 Series of intermediate stages leading to a smoothed map of date<br />
of first observed response by numerical relaxation as described<br />
in the text ............................................................................ 123<br />
7-10 Smoothed dose contours from pine foliage response using numerical<br />
relaxation of dose values from individual pine forest pixels inside the<br />
w hite border ........................................................................ 125<br />
7-11 Smoothed dose contours from Figure 7-10 displayed with a gray-scale<br />
background of the western end of the <strong>Chernobyl</strong> nuclear power station .... 127<br />
8-1 Dose rate measurements for close-in fallout contaminated areas for two<br />
different dates ....................................................................... 133<br />
8-2 Change in gamma dose rate from radioactive materials in the close-in<br />
zone as a function of time based on aerographic surveys (data from<br />
A sm olov et al., 1987) ............................................................. 135<br />
8-3 Accumulated close in fallout dose (1-m above ground) in I and<br />
4 km 2 areas ......................................................................... 136<br />
8-4 Ratio versus time of the accumulated dose (b and g) at the center of<br />
cylindrical foliage elements at the top of pine forest canopy to the<br />
accumulated 1-m gamma dose a) under the canopy and b) in an open<br />
field for the same total fallout deposition ........................................ 138<br />
xv
TABLES<br />
Table<br />
Page<br />
3-1 Original Landsat-5 TM Tasseled Cap coefficients (Christ et al., 1986) ..... 20<br />
3-2 Modified Tasseled Cap coefficients used in the present work to place<br />
all feature intensities in the range 0 to 25......................... 22<br />
3-3 Landsat scenes analyzed and time of scene relative to reactor explosion .... 23<br />
3-4 Coordinates (UTM, Zone 36) of the upper left comers of the comer<br />
pixels of the 1536 x 1536 pixel area analyzed in this report .................. 34<br />
3-5 Band structure for the winter/summer composite image, COMP21,<br />
generated from Dates 2 and 1 ..................................................... 36<br />
3-6 Pine forest classes ................................................................. 61<br />
5-1 Ratio of beta to gamma doses to foliage calculated from the results<br />
of Appendix A for fallout retained in the canopy ............................... 73<br />
5-2 Dose rate per unit source for various sources and dose points.<br />
Reference canopy volume source density is taken equal to the unit<br />
ground source density spread over the assumed .............................. 75<br />
5-3 Calculation of normalized beta dose rate (BDR) versus time assuming<br />
all dose from beta particles on foliage, all fallout on t = I day ................ 80<br />
5-4 Calculation of normalized gamma dose rate (GDR) versus time<br />
assuming all dose from gamma emitters on the ground and all<br />
fallout on t = I day ................................................................ 83<br />
6-1 Estimated total doses to produce three endpoints of damage to pines<br />
for an exposure period of three weeks (assumed equivalent to the<br />
effective exposure period at <strong>Chernobyl</strong>) ........................................ 93<br />
6-2 Detectability by multispectral remote sensing of radiation damage to<br />
pine trees for spring exposures as a function of time since start of<br />
exposure and dose ................................................................. 96<br />
6-3 Dose estimates according to Equation 6.1 for a first detected response<br />
corresponding to the times of the 9 postaccident images presented<br />
in this report ........................................................................ 100<br />
xvi
TABLES (CONTINUED)<br />
Table<br />
Page<br />
7-1 Area of pine forest near the <strong>Chernobyl</strong> nuclear reactor station cleared<br />
by the listed image date. Clearing of other vegetation not included ......... 115<br />
7-2 Description of contours drawn in Figure 7-8 ................................... 120<br />
7-3 Dose bands and affected areas for the smoothed radiation contour maps<br />
of Figures 7-10 and 7-11 .......................................................... 128<br />
8-1 Results of forest damage by radiation for an area with edge about 1 km<br />
from the site of the Unit 4 reactor explosion (Sobdovych et al., 1992,<br />
courtesy of Gam ache, 1993) ..................................................... 131<br />
8-2 Estimation of initial dose rates for contours enclosing specified areas ....... 134<br />
8-3 The ratio of the pine canopy foliage dose to the gamma dose I m off the<br />
ground using satellite image and aerographic survey data ................... 137<br />
8-4 Summary of typical b/g dose ratios for four groups of people who<br />
suffered radiation lesions in the skin at <strong>Chernobyl</strong> (Barabanova and<br />
O sanov, 1990) ...................................................................... 140<br />
xvii
SECTION 1<br />
INTRODUCTION<br />
During an extreme radiation accident such as occurred at the <strong>Chernobyl</strong> nuclear power station<br />
on April 26, 1986, explosion and fire may cause lofting of radioactive aerosols and vapors from a<br />
few hundred meters to more than a kilometer into the air. Winds then transport the material away<br />
from the accident site. The aerosols settle to the ground, larger, heavier particles nearby and<br />
smaller, lighter ones farther away. The resulting terrestrial deposition of fallout particles occurs in<br />
a more or less continuous but irregular fashion, following local and regional weather patterns. The<br />
finest aerosols enter the global atmospheric circulation.<br />
Much attention has been given to the regional and global patterns of fallout from the <strong>Chernobyl</strong><br />
accident because of worldwide concerns for human radiation exposure and the entry of radio<br />
nuclides into the food chain. The primary health concern is the induction of cancers in the human<br />
population from the resulting low level radiation exposures. The doses at regional and global<br />
distances are far below the levels required to induce acute radiation sickness or cause visible<br />
changes in vegetation.<br />
Locally, however, at the site of the reactor explosion and more than a kilometer downwind<br />
from the site, radiation doses from radio nuclide fallout were above lethal levels for humans.<br />
Symptoms of radiation sickness occurred in some accident victims within the first hour of<br />
exposure (Young, 1988). Early fatalities and causes of death have been reviewed as part of this<br />
project and described in an earlier report (Laupa and Anno, 1989).<br />
The circumstance of one victim is of particular interest in the context of the analysis in this<br />
report. According to Barabanova and Osanov (1990), this person was 1.0 km downwind from the<br />
reactor at the time of the explosion, remained there for about an hour, and was exposed to both the<br />
radioactive plume and particles of fallout. According to Barabanova and Osanov, the individual<br />
was covered with black dust and received an estimated total gamma radiation dose of 12.7 Gy.<br />
The estimated beta radiation dose to the skin of his scalp, neck and upper body was 250-360 Gy at<br />
a depth of 7 mg/cm 2 and about 30 Gy at a depth of 150 mg/cm 2 . He died on the 17th day after<br />
exposure.<br />
As detailed in Section 6 of this report, the gamma dose alone for this victim is comparable to<br />
the median lethal dose (LD 50 ) for pine trees, which are only slightly less radiosensitive than<br />
humans. Beta dose will contribute further to vegetation damage depending on the depth of sensitive<br />
plant tissues. In addition, vegetation remains in the contaminated area and accumulates dose over a<br />
much longer exposure time than would mobile human beings who leave the area and are<br />
decontaminated.
Figure 1-1 shows the <strong>Chernobyl</strong> nuclear power station and immediate vicinity. The burning<br />
Unit 4 reactor (dark red dot) is circled and the approximate wind direction at the time of the reactor<br />
explosion is indicated by the arrow. The image was produced by merging data from a Landsatl<br />
Thematic Mapper image from April 29, 1986 with data from a SPOT 2 panchromatic image from<br />
May 1, 1986. The merged image combines the 10 m spatial resolution of the SPOT image and the<br />
wider spectral coverage of the 25 rn resolution Landsat image. The area shown in the image is a<br />
square with 5 km sides oriented along the Landsat orbital path. The edge of the forest at the tail<br />
end of the arrow in Figure 1-1 is only a little over 1 km from the burning reactor, so the victim<br />
described by Barabanova and Osanov was near the forest. Exposure of the victim is presumably a<br />
lower limit to the exposure of the nearby forest.<br />
Volume 3 of this report (Painter and Whicker, 1993) presents a description of the geography<br />
and vegetation of the Poles'ye region of Ukraine and Belarus where the <strong>Chernobyl</strong> nuclear power<br />
station is located. As is true in many other regions of the world, pines trees are the most<br />
radiosensitive (see, for example, Whicker and Fraley, 1974) of the large, widely occurring species<br />
of vegetation in this region. Thus, substantial effects on pine forest near the <strong>Chernobyl</strong> accident<br />
site are to be expected. Because the radiosensitivity of pines is comparable to that of humans, the<br />
study of such pine tree responses in a fallout radiation field, especially through remote sensing, is<br />
relevant to human operations in a radiation environment.<br />
The area around the <strong>Chernobyl</strong> nuclear power station after the accident of 26 April 1986<br />
provides a unique opportunity to observe the response of large scale plant communities to fallout<br />
radiation. In a review article, Whicker and Fraley (1974) quote only two studies involving realistic<br />
fallout exposures of plants. One (Murphy and McCormick, 1971) involved spreading feldspar<br />
particles coated with 90 Y on small plant communities on granite outcrops in the southeastern<br />
United States. The other (Rhoads and Platt, 1971) observed damage to desert vegetation receiving<br />
fallout from two small cratering explosions at the Nevada Test Site. Neither of these studies<br />
encompassed the size and variety of plant communities present at <strong>Chernobyl</strong>. No other studies<br />
involve the appropriate mix of beta and gamma exposure to foliage.<br />
1Landsat is the United States' civil land remote sensing satellite system. Data is obtained from the Earth<br />
Observation Satellite (EOSAT) Company of Lanham, Maryland.<br />
2 The Syst•me Probatoire d'Observation de la Terre (SPOT) satellite is operated by the French space agency<br />
Centre National d'Etudes Spatiales. Data is obtained from the SPOT Image Corporation, a U.S. subsidiary of the<br />
French company SPOT IMAGE.<br />
2
Figure 1-1. Immediate vicinity of the <strong>Chernobyl</strong> nuclear power station with Unit 4 reactor (circled)<br />
still burning; arrow shows approximate wind direction when the reactor exploded.<br />
3
Time plays several essential roles in the radiation response of a plant community (Whicker and<br />
Fraley, 1974), including:<br />
1) duration of exposure (or dose rate),<br />
2) season of exposure,<br />
3) time for manifestation of injury,<br />
4) repair and recovery times, and<br />
5) time for development of secondary effects such as community succession.<br />
Control areas are of primary importance in analyzing these time factors and in separating radiationinduced<br />
changes from naturally occurring ones. There should be multiple observations over at<br />
least one annual cycle of both the control and exposed vegetation. In addition, preirradiation<br />
observations of the experimental area are required to establish the initial condition of the exposed<br />
vegetation.<br />
Fortunately, satellite multispectral imagery provides observations of both control areas and<br />
preirradiation observations of the area affected by the <strong>Chernobyl</strong> reactor explosion. Control areas<br />
are available within individual Landsat scenes outside the highly irradiated areas and are an integral<br />
feature of our analysis. Furthermore, both summer and winter preaccident scenes of the irradiated<br />
area are included. Finally, the 16-day revisit time for acquisition of Landsat scenes from the same<br />
satellite on the same path offers ample opportunity to obtain multiple cloud-free images over an<br />
annual cycle. The analysis reported here uses 7 images during the first year postaccident and 2<br />
more during the following year. This time series of remote observations is a unique record of the<br />
spectral response of pine foliage to fallout radiation exposure. Section 2 outlines the procedure<br />
used in the remainder of the report to estimate dose to pine foliage from these images.<br />
4
SECTION 2<br />
REMOTE FALLOUT DETECTION THROUGH IMAGERY<br />
ANALYSIS<br />
A substantial amount of forest within a few kilometers of the <strong>Chernobyl</strong> nuclear power station<br />
was heavily contaminated with radionuclides by the April 26, 1986 explosion of the Unit 4 reactor<br />
and the ensuing fire. Radiation doses to conifers in some areas were sufficient to cause<br />
discoloration of needles within four to five weeks. Other areas, receiving smaller doses, showed<br />
foliage changes six months to a year later. Although we do not have color photographs of the trees<br />
affected by the accident, Figure 2-1 shows typical dying pine foliage from landscape plantings in<br />
northern Virginia. One photo shows branches with needles dying only at the tips. The other<br />
shows part of a tree whose needles are completely dead except for a few isolated bunches of green<br />
needles. Healthy trees are located in the background. Progression from green to yellow-green to<br />
reddish orange to dry brown is typical of the death of pine foliage from many causes including<br />
exposure to ionizing radiation.<br />
Multispectral imagery available from satellite sensors is especially suited for remote monitoring<br />
of such changes in vegetation since the changes affect both the visible and infrared reflectivity of<br />
foliage. A series of Landsat Thematic Mapper images spanning two years after the <strong>Chernobyl</strong><br />
accident are analyzed for significant, accident-induced change in a companion document, Volume 2<br />
(McClellan et al., 1992). <strong>Accident</strong>-related changes in foliage are apparent three days after the<br />
accident. Changes due to cleanup activities and the progression of foliage deterioration are present<br />
in images taken in the following months. Finally, delayed effects of low radiation doses on initially<br />
undamaged foliage were still appearing more than a year after the accident.<br />
Pacific-Sierra Research Corporation's Hyperscout" m algorithm provided the analytical basis in<br />
Volume 2 for demonstrating significant, radiation-induced change in the forests around <strong>Chernobyl</strong>.<br />
The Hyperscout algorithm will detect image-to-image change, providing a measure of significance<br />
of change on a pixel-by-pixel basis. The algorithm analyzes reflectivity and emission changes<br />
occurring in the detection bands of the imaging device. Spectral signature, spatial distribution, and<br />
time dependence of the observed changes help identify stresses causing the change. The algorithm<br />
is especially useful when the induced change has a low signal-to-noise ratio and when general<br />
image variations such as seasonal illumination and annual vegetation cycles cause change that is not<br />
of interest. It operates on pairs of spatially registered images and is capable of useful results even<br />
Hyperscout is a trademark of Pacific-Sierra Research Corporation.<br />
5
Figure 2-1. Typical coloration of dying pine foliage. Landscape setting; upper photo, branches<br />
almost completely brown with healthy trees in background; lower photo, only needles at tips of<br />
branches are brown.<br />
6
when comparing images from different detectors. The sensitive nature of the Hyperscout<br />
algorithm enabled delineation of both the earliest changes in heavily exposed forest and the latedeveloping<br />
deterioration of less exposed forest.<br />
Figure 2-2 shows an example of one of the Hyperscout stress maps (12 days postaccident)<br />
from Volume 2, with the area of affected forest nearest the reactor site circled. The Unit 4 reactor<br />
is marked with a white cross. The site marked is the brightest pixel of the fire visible in the<br />
Landsat image at 3 days postaccident (see Figure 1-1).<br />
With guidance from the changes detected by the Hyperscout algorithm, this volume presents<br />
quantitative estimates of radiation dose to the pine foliage near the <strong>Chernobyl</strong> nuclear power station<br />
derived from remote Landsat imagery. The physical basis for detecting foliage stress with Landsat<br />
imagery is presented in Volume 2 and will not be repeated here. Quantitative dose estimation for<br />
the exposed conifers requires extracting the time of onset of observable radiation-induced foliage<br />
damage from the images for each pine forest pixel and combining this data with the relationship<br />
between dose and time of onset (time-to-response) of conifers for ionizing radiation.<br />
Figure 2-3 outlines the remote sensing method for dose determination from multispectral<br />
images of pine forest. Similar considerations apply to other vegetation; however, pine trees are<br />
among the most radiation sensitive species of widely occurring vegetation. The first step indicated<br />
in Figure 2-3 is the determination of preaccident vegetation classes in the multispectral imagery,<br />
particularly, the identification of pixels that consist predominantly of pine trees. Generally, forested<br />
areas are readily apparent from their color, texture and location in false color images as presented in<br />
Volume 2. However, to optimize data quantity and quality it is necessary to classify individual<br />
pixels. Section 3 of this volume describes the procedure for quantitative determination of class<br />
membership on a pixel-by-pixel basis. Since a primary problem is the discrimination among<br />
coniferous, deciduous, and mixed coniferous-deciduous pixels, we use preaccident images from<br />
both summer and winter to achieve a reliable forest classification.<br />
The extent to which tree foliage excludes a view of the ground from above is called canopy<br />
closure. If closure is high then satellite images will see mostly foliage. If closure is low then rocks,<br />
soil, and understory vegetation will dominate the image. We used a preaccident, late winter image<br />
(21 March 1986) with extensive snow cover to assist selection of areas with a high canopy cover<br />
of pine trees. The snow is a good discriminate for areas with low canopy cover or areas with a<br />
substantial number of deciduous trees. By eliminating these areas, we obtain a set of pixels<br />
containing a high canopy cover of pines that increases confidence in our dose estimates for these<br />
areas. Also, in this late winter image, the Pripyat River and its tributaries are mostly frozen. The<br />
resulting bright snow and ice along the river basin eliminates substantial misclassification that<br />
7
S128 256<br />
Stress Index<br />
Figure 2-2. Forest stress map for 8 May 1986 as presented in Volume 2. <strong>Accident</strong>-affected area<br />
circled. White cross is Unit 4 reactor site; Zone 36 Universal Transverse Mercator (UTM)<br />
coordinates about X = 298,275 m and Y = 5,697,175 m.<br />
8
of PracPreaccident<br />
e C e aImages<br />
Landsat<br />
Reference Site "X Pre- and<br />
Spectral urignatures Postaccident<br />
Ve sImage Date Landsat Images<br />
Noml Pixels i.se , \ ,- .<br />
[ Map for Pine Forest V ersus radiation<br />
Figure 2-3. Method of radiation dose determination from remote multispectral sensing of<br />
pine forest.<br />
9
occurred in our preliminary analysis (Volume 2) using only the summer image (6 June 1985) for<br />
determination of pine forest classes. In that analysis, many mixed pixels of dark water and<br />
deciduous vegetation at the rivers edge mimic the spectral response of pine forest . These<br />
misidentified pixels lead to the spurious indications of stress seen along the river in Figure 2-2.<br />
The classification of pine forest is based on the location of a reference site for each class that<br />
contains on the order of one hundred contiguous pixels belonging to the class. The reference sites<br />
are defined by polygons at fixed geographic locations. These locations must be far enough from<br />
the reactor to be unaffected by radiation and must not be obscured by clouds or haze on any image.<br />
The mean spectral signature and its covariance matrix for each class on each date is determined<br />
from the pixels in these reference sites. This procedure is described in Section 3.<br />
For many biological responses to radiation exposure, there is a dose-dependent delay between<br />
exposure and onset of the response. Such is the case for foliage damage in pine trees. We use the<br />
deviation of the spectral signal of each pine forest pixel from its class mean to detect significant<br />
change in foliage and, hence, estimate a time-to-response for areas of forest affected by the<br />
accident. Section 7 presents our time-to-response results.<br />
The biological endpoint for these satellite observations is change of spectral reflectivity of the<br />
foliage. However, the endpoints most commonly reported in plant radiobiology literature are<br />
100% lethality and 50% reduction in growth rate. There are no published reports, as far as we<br />
know, that provide controlled measurements of the spectral response of foliage to radiation<br />
exposure that can be directly correlated with other reported endpoints. Fortunately, though, there<br />
are sufficient visual observations reported for pine trees and sufficient reports of other vegetation<br />
stress response- to establish a plausible connection between pine tree response and multispectral<br />
detection. Section 6 discusses the connection made between reported endpoints and their<br />
relationship to spectral appearance from remote sensing. A functional relationship between timeto-response<br />
as determined from visible and infrared imagery and foliage dose is deduced.<br />
Finally, in Section 7 the time-to-response map from the imagery and the relationship between<br />
radiation dose and time-to-response are combined to generate a map of estimated radiation dose to<br />
pine trees in the vicinity of the reactor explosion.<br />
10
SECTION 3<br />
HABITAT AND VEGETATION NEAR THE CHERNOBYL<br />
NUCLEAR POWER PLANT<br />
This section summarizes the geography, soils, climate, native vegetation, and crops of the<br />
region surrounding Cherrobyl and provides details on the analysis of Landsat imagery leading to<br />
the delineation of four classes of pine forest within a 38.4 km square area centered approximately<br />
on the <strong>Chernobyl</strong> nuclear power station.<br />
Figure 3-1 is a regional map of the border area between the former Soviet republics of Ukraine<br />
and Belorussia (Belarus) stretching from Kiev in Ukraine to Gomel in Belorussia. The city of<br />
<strong>Chernobyl</strong> lies on the Pripyat River near the northwest comer of the Kiev Reservoir. The city of<br />
Pripyat is about 15 km further up the Pripyat River adjacent to the <strong>Chernobyl</strong> nuclear power<br />
station. The power station and its cooling pond are indicated by a small black rectangle labeled<br />
"Site of <strong>Chernobyl</strong> power station." Figure 3-1 provides the geographical framework for the<br />
regional description in Section 3.1 below.<br />
Figure 3-2 is the earliest of the series of Landsat images analyzed in this report. It shows a<br />
north/south oriented rectangle about 58 km by 72 km. The upper left, or northwest, corner of the<br />
rectangle is clipped because the desired area was near the edge of the Landsat path on this date.<br />
The angle of the clipped portion corresponds to the angle of the satellite orbit relative to lines of<br />
longitude at this location. The Dnieper River and the Pripyat River both empty into the Kiev<br />
Reservoir at the lower right of Figure 3-2. Only the upper end of the reservoir is visible in the<br />
image as an indistinctly outlined dark area merging with the meandering river beds. The cooling<br />
pond of the nuclear power station appears as the large, black, footprint-like shape outlined in white<br />
along the Pripyat River. The north/south extent of the Landsat image of Figure 3-2 corresponds<br />
approximately to the north/south extent of the dashed circle on the map in Figure 3-1. This 30 km<br />
radius circle is the danger zone evacuated in the aftermath of the reactor explos'e.' A 38.4 km<br />
square subset of Figure 3-2 is selected for the ground cover analysis in Sections 3.2 through 3.5<br />
below.<br />
3.1 REGIONAL DESCRIPTION.<br />
In the following regional description, we use the term Former Soviet Union (FSU) to denote<br />
the old Union of the Soviet Socialist Republics (USSR) but continue to use references to the names<br />
of the Soviet Socialist Republics (SSRs).<br />
3.1.1 Geography.<br />
The <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant (NPP) and much of the area within 30 km around it are<br />
located in the Poles'ye region of the Ukrainian and Belorussian (Byelorussian, White Russian)<br />
11
RechittaDobrush<br />
Kilinkovichi<br />
GOMEL' OBLAST'<br />
Kopachia e ybc<br />
% %CCeembyi<br />
.qi~ esicaI~~<br />
Orodyanka.<br />
Chernbyl nclear ccient<br />
Poiesitoy %12
Figure 3-2. Landsat image of the confluence of the Dnieper and Pripyat Rivers at the upper end of<br />
the Kiev Reservoir. The sharply outlined black footprint shape is the reactor station cooling pon4<br />
along the Pripyat River. [Thematic Mapper false color presentation (RG,B) = (7,4,1), 6 June<br />
1985, geocoded, 1 degree longitude by 0.5 degree latitude.]<br />
13
Soviet Socialist Republics. The power plant itself and the cities of Pripyat and <strong>Chernobyl</strong> are in<br />
the Ukrainian SSR. It lies at about 51°12'N longitude 30 0 8E latitude.<br />
The Poles'ye (Polesye, Poles'e, Poles'ya, Polesie, Polessie) lies in the Pripyat (Pripyat',<br />
`qip'at, Pripiat, Pripet) River basin and part of the Dnepr (Dnieper) River basin and includes the<br />
area called the Pripyat marshes or bogs. It is a vast lowland on the Russian platform, extending<br />
south to the Volyno-Podolsk Plateau (Keller 1927; Berg 1950; Fridland 1976; Lysenko and<br />
Golovina 1982). The relief from the center to the edges of the basin is only 55-100 m (Berg 1950;<br />
USSR 1987). Drainage is very poor and the ground-water table is usually high (Keller 1927; Berg<br />
1950; Fridland 1976). The banks of the streams in the Poles'ye are very low (Keller 1927). In the<br />
spring and after heavy summer rains, streams overflow into areas between neighboring streams<br />
and water from one stream passes into another. The Poles'ye is often divided into three sections:<br />
the western, the central (or right bank--on the right bank of the Dnepr River), and the eastern (or<br />
left bank) (Golovina, et al., 1980; Lysenko and Golovina 1982). In the Ukraine, the central<br />
section is divided into the Kiev Poles'ye and the Zhitomir Poles'ye. The Kiev Poles'ye lies on the<br />
middle Dnepr slope, in the interfluve of the Pripyat and Teterev Rivers, extending east to the Dnepr<br />
River (Golovina, et al., 1980; Lysenko and Golovina 1982), and includes the <strong>Chernobyl</strong> NPP and<br />
the immediately area<br />
3.1.2 Soils.<br />
The soils in the Poles'ye formed in a humid climate on platform plains from a blanket of<br />
unconsolidated Quaternary fluvioglacial sand and loamy sand over and underlying a glacial<br />
moraine (Fridland 1976; Lysenko and Golovina 1982). The moraine itself is mostly loamy sand.<br />
Clay-loam lake deposits are rare and loesses even rarer (Lysenko and Golovina 1982). The parent<br />
materials are very boldery and gravely; fine particles are usually washed away (Golovina, et al.,<br />
1980). Glacial waters were active for along time during the formation of the parent material<br />
(Lysenko and Golovina 1982). The soils generally lack carbonates in the parent materials. The<br />
sandy soils of the Poles'ye are particularly vulnerable to erosion if plowed (Symons 1972). Some<br />
areas possess aeolian relief, often in the form of parabolic, west-facing dunes (Berg 1950; Fridland<br />
1976). This significantly affects soil cover composition.<br />
Most of the Poles'ye soils are poor in humus (USSR 1986). Soils in the western and central<br />
Poles'ye often have 1-1.5% humus and can have less than 1% (Krupskiy, et al., 1970). The soils<br />
generally have a low pH (Golovina, et al., 1980) and are low in available nutrients, including<br />
boron, zinc, phosphorus, potassium, and nitrate (Golovina, et al., 1980; Oleynik 1981; Lynsenko<br />
and Golovina 1982). There are large quantities of weakly podzolic sandy soils. Podzols develop<br />
under coniferous forests (Sukachev 1928; Berg 1950). The distribution of podzols is patchy in<br />
14
the Poles'ye (Fridland 1976). Medium-podzolic sod loamy sandy soils have formed on moraine<br />
outcrops in the Kiev Poles'ye (Lysenko and Golovina 1982).<br />
Much of the area is boggy. Bog soils receive excessive moisture for the greater part of the<br />
year, are sometimes covered with shallow water, and have poor drainage (Berg 1950). Much of<br />
the Poles'ye is low lying, with poor drainage, ideal for bogs (Keller 1927; Berg 1950). There are<br />
also sod-podzolic ("half-bog" or "meadow") soils (Berg 1950). Soils on flood plains may be bog,<br />
meadow, or podzols. Some parts of the Poles'ye contain "islands" of soils foreign to it (Berg<br />
1950). One such island occurs on the Ovruch ridge (about 90 km west of <strong>Chernobyl</strong>). This ridge<br />
is 320 m above sea level and 60 m above the surrounding lowland.<br />
Based on the soils map in the Ukrainian SSR Atlas (Anonymous 1962), the site of the<br />
<strong>Chernobyl</strong> nuclear power plant appears to be at or near the boundary of:<br />
Type 1:<br />
Type 27:<br />
Soddy-slightly podzolic sands and sandy-clayey soils.<br />
Sods and meadowlands, gleys, sandy loams, and loams.<br />
It is likely that the coniferous forests near the site are on type I soils.<br />
3.1.3 Climate.<br />
The climate of the Poles'ye is influenced by maritime, continental, and local factors. Moisture<br />
from the Baltic Sea and the "great valley" region of Poland influences the Poles'ye (Borisov 1965;<br />
Szafer 1966). There is unrestricted passage of marine winds, unhindered by major land relief, s-<br />
there is considerable marine influence on the climate.<br />
A small local maximum of relative humidity, caused by the intensified evaporation of water of<br />
standing water, is noticed over the marshes of the Poles'ye(Borisov 1965). Bogs influence the<br />
local microclimate (Szafer 1966). These areas have high humidity, low temperature minima,<br />
evening and morning mists Close to the soil surface, and frequent frosts. Forests also influence the<br />
microclimate (Szafer 1966). The mean air temperatures are lower than in the open, the daily<br />
temperature ranges are smaller, the snow cover is less, and winds are stilled.<br />
In the Poles'ye, the minimum relative humidity at 1 pm in May is 50-55%. The mean humidity<br />
for June, July, and August is 56-60% (Borisov 1965). Mean temperatures in the region are -6 to<br />
-7*C in January, 6 to 70C in April, 190 in July, and 7°C in October (Anonymous 1962). Absolute<br />
minimum temperatures are -30 to -35'C in January, while summer maximums range up to 40'C<br />
(Anonymous 1962). The growing season (days with mean temperatures above 5 C) is about 160-<br />
190 days (Anonymous 1962; Szafer 1966). It begins about April 11 and runs to about October 25<br />
(Anonymous 1962). Summer precipitation exceeds that in winter by a factor of two (Borisov<br />
1965). In any season, there is precipitation every 2-3 days. In the warm season (April-<br />
September), the amount of precipitation in the interior of the European FSU is much greater (350-<br />
500 mm) than on the coasts (200-300 mm) (Borisov 1966). During the cold half of the year<br />
15
(November to March) in the central belt of European FSU, the precipitation amounts are 100-300<br />
mm (Borisov 1965). A precipitation maximum is situated on the Pripyat and the upper reaches of<br />
the Dnepr and Western Dvina Rivers. The maximum precipitation in the Pripyat basin is 680-695<br />
mm/year (Berg 1950). Precipitation diminishes eastward from the Pripyat (Borisov 1965). In the<br />
specific region of interest, mean precipitation is 500-600 mm/year, with around 180 days/year<br />
recording precipitation (Anonymous 1962). The mean intensity of precipitation amounts to 8- 10<br />
mm/hr in the central belt o the country (51-59°N) (Borisov 1965).<br />
At 500N, by the second 10-day period in November snow cover is normally continuous<br />
(Borisov 1965). In western European FSU, snow cover is usually continuous by the last 10 days<br />
in October or the first 10 days in November. According to Anonymous (1962), sniow cover in the<br />
<strong>Chernobyl</strong> area usually lasts from mid-December to Mod-March. At 55'N 30'E, there are an<br />
average of 2.5 temporary snow covers before the onset of winter (Borisov 1965). In central<br />
European FSU, winter temperatures and precipitation are highly variable (Borisov 1965). Mean<br />
maximum snow depth is 10-30 cm. Continuous snow cover generally lasts about 80 days. Rivers<br />
and streams are generally frozen for about 100 days/year (Kendrew 1942).<br />
3.1.4 Native Vegetation.<br />
Most of Belorussia and Ukraine, including the Poles'ye, is in the Eastern European Vegetation<br />
Province (Takhtajan 1986). The northern, eastern and southeastern boundaries correspond to the<br />
distributions of Quercus roburl (English oak), Acer platanoides (Norway maple), and Corylus<br />
avellana (European filbert). The basic plant community types in the region are forests, woodlands<br />
or carrs, meadows, fens, and bogs. 2 . Many of these communities can grade into one another.<br />
Bog, meadow, and forest vegetation may occur on flood plains. Meadows may be transitional or<br />
ecotonal between bogs and forests. In poor sandy soils, peat bogs occupy the low areas, forests<br />
the more upland areas. Succession can progress from forest to bog or from bog to forest. Most<br />
bogs in central Europe are being drained (Walter 1978). The replacement vegetation is usually<br />
meadow grasses, birch, pine, or spruce. While the vegetation types to be discussed are specific to<br />
the Poles'ye, the species lists given are those known to occur in that type of vegetation in northern<br />
Ukraine, southern Belorussia, and/or adjacent areas in Poland. Information on both the vegetation<br />
IBoth Common and Latin names come from a number of literature sources. In order to eliminate synonyms and<br />
assure current nomenclature, the names in the Flora of the USSR and the Flora Europaea will be used for<br />
the final report.<br />
16
of the Poles'ye and the species lists was synthesized from Keller 1927; Sukachev 1928; Wulff<br />
1943; Berg 1950; Szafer 1966; Walter 1978; Oleynik 1981; and Takhtajan 1986.<br />
The sandy riverine sections along river channels receive an annual sand deposit Berg (1950).<br />
The community lying closest to the stream is often a sedge-horsetail fen. The plat", found in this<br />
type of habitat include Equisetum arvense (field horsetail), E. variegatum, E. palustre (horsetail or<br />
scouring rush), Carexfusca, C. canescens, C. stellulata, C. dioica, C. flava (sedges), Eriophorum<br />
latifolium, and E. augustifolium (cottongrass or cottonsedge).<br />
Immense tracts of the Poles'ye are occupied by bogs. There are several types, but most are<br />
Sphagnum or peat bogs. These are dominated by sphagnum mosses, including Sphagnum<br />
recurrum, S. fuscum, S. cuspitatum, S. medium, S. revellum, S. acutifolium, Hypnum schreberi,<br />
and H. crista-castrensis. Such bogs also have Eriophorum vaginatum (sheathed cottongrass),<br />
Carex pauciflora, C. limosa, and C. vaginata, and the carnivorous plant Drosera intermedia<br />
(roundleaf sundew). The vegetation of reed-bulrush bogs include Phragmites communis (common<br />
reed), S. lacuster (bulrush), Phalaris arundinacea (canarygrass), Calamagrostis neglecta<br />
(reedgrass), Typha angustifolia, T. latifolia (cattails), Carex vesicaria, C. gracilis, C.<br />
pseudocyperus, C. rostrata. Sedge bogs often include Carexfiliformis (large sedge), C. vesicaria,<br />
C. pseudocyperus, C.rostrata. C. elata, C. gracilis, C. acutiformis, C. riparia, C. lasiocarpa, C.<br />
diandra, C. aespitosa, and C. omaskiana (omskiana sedge) with Deschampsia caespitosa (tufted<br />
hairgrass) on hummocks. Sedge bogs are often found where Alnus glutinosa (European alder) has<br />
been cut. Transitional areas begin with either Sphagnum or sedge bogs and these can grade to<br />
alder, birch, birch-spruce, pine-birch, pine, and birch-aspen-conifer bogs.<br />
There are two basic types of meadows, flood-plain (or wet) meadows and upland (or fresh)<br />
meadows. Flood-plain meadows may be inundated for some time each year, especially in late<br />
winter and early spring. These are found in river valleys and at the peripheries of shallow lakes.<br />
They often consist of secondary vegetation resulting from drainage of fens or bogs. The<br />
vegetation includes Festuca ovina (sheep fescue), F. rubra (red fescue), F. pratensis (fescue),<br />
Agropyron repens (quack- or conchgrass), Poa pratensis (Kentucky bluegrass), P. annua (annual<br />
bluegrass), Phleum pratense (timothy), Agrostis stolonifera (redtop), Calamagrostis neglecta,<br />
Deschampsia caespitosa, Molinia coerulea, Carex panicea, C. caespitosa, Eriophorum neglecta,<br />
and Trifolium pratense (red clover).<br />
Fresh meadows generally are found in interstream areas, developing as secondary vegetation<br />
on cut or burned forest sites. Mowing is often used to prevent forest encroachment. They have a<br />
2 A fen is a constantly or frequently flooded area through which water flows. For simplification, the term "bog"<br />
includes marsh and swamp and is used to mean flooded areas having zero or very low rates of water flow.<br />
17
moderate moisture supply which fluctuates widely but generally does not surface. These meadows<br />
are usually used as pastures or hay meadows and some of the plants, especially the legumes, may<br />
have been sown. The vegetation includes Arrhenatherum elatius (tall or false oatgrass), Bromus<br />
mollis (smooth brnme), Agrostis stolonifera, , P. trivialis (bluegrass), P. annua, Phleumn pratense,<br />
Lolium perenne (perennial rye grass), Cynosurus cristatus (crested dogtail), Festuca pratensis, F.<br />
rubra, Agropyron repens, Phalaris arundinacea, Calamagrostis negecta, Alopecurus pratensis<br />
(meadow foxtail), Trifolium repens (white clover), T. pratense, T. hybridum (Alsike clover), T.<br />
incarnatum (crimson clover), and Medicago sativa (luceme or alfalfa).<br />
There are several types of carrs (or woodlands). The plants in willow carrs include Salix<br />
cinerea (gray willow), S. rosmarinifolia (rosemary willow), S. alba, S. fragilis, S. triandra, S.<br />
purpurea, S. rossica (willows), Alnus gutinosa, Fraxinus excelsior (ash), Betula laevis (European<br />
birch), Cornus alba (Siberian or white dogwood), Rosa (rose) sp., and Ribes nigrum (black<br />
current). The plants in elm carrs include Ulnus pedunculata (Russian elm), U. glabra (Scotch<br />
elm), Quercus robur, and Acer Platinoides. Alder-ash carrs usually include Alnus glutinosa,<br />
Fraxinus excelsior, Ulnus pedunculata, U. glabra, Acer platinoides, Carpinus betulus (European<br />
hornbeam), Festuca gigantea, and Agropyron Caninum (wheatgrass).<br />
The Kiev Poles'ye lies on the ecotone between the mixed coniferous-deciduous forest and the<br />
forest steppe. There are two basic mixed coniferous-deciduous forest types: Pine-oak forest and<br />
spruce-oak forest. The common steppe-forest of the area is oak-hornbeam. There are also some<br />
pine (bor) woodlands in the Poles'ye. It is unlikely that any virgin forests remain in eastern<br />
Europe (Walter 1978).<br />
Quercus robur is the dominant in forest-steppe in the European USSR. It grows best in the<br />
southwestern part of the European FSU and in the Poles'ye. Quercus robur will not grow on<br />
strongly podzolic soils. It is commonly found on floodplains and often grows in mixed stands<br />
with Pinus sylvestris (Scotch, Scots, or common pine), Picea abies (European spruce), or<br />
Carpinus betulus. Fagus sylvatica (European beech) is sometimes a component in oak-hornbeam<br />
forests on richer soils.<br />
Pinus sylvestris is a light-loving species, not tolerant of shade. It is not very exacting in soil<br />
and moisture requirements. It tolerates relatively poor soils and is often associated with bogs and<br />
"bor" soils. "Bor" forest is a type of sparse pine forest growing on sandy soil, dunes, etc.<br />
(Fridland 1976). The sandy forested areas, such as the parabolic, west-facing dunes, of the<br />
Poles'ye are usually covered with Pinus sylvestris, giving the region a "northern" appearance. The<br />
understory in pine (bor) woodland is often dune vegetation, including Festuca ovina, F. rubra, Poa<br />
pratensis, Deschampsia caesrvitosa, Carex arenaria (sand sedge), and Trifolium arvense (clover).<br />
When these forests are cut, the dune vegetation replaces them until there is a secondary pine<br />
growth. The understory in pine-oak forests often include Betula laevis, Tilia cordata (linden,<br />
18
lime). Populus tremula (aspen), and Corylus avellana. Pinus trees in bogs are frequently stunted.<br />
Based on the maps in Anonymous (1962), common pine (Pinus sylvestris) is the most likely forest<br />
type in the immediate vicinity of the <strong>Chernobyl</strong> nuclear power plant.<br />
Picea abies endures shade well but requires humid, relatively rich soils. Its southern limit runs<br />
through Kiev Poles'ye. Spruce-oak forests are probably occur mainly on "islands" of richer<br />
podzolic soils scattered in the Poles'ye. Near the southern limit of Picea's range, the most frequent<br />
understory species in spruce-oak forest are Tilia cordata, Acer platanoides, Fraxinus excelsior,<br />
Ulnus pedunculata, Corylus avellana, Betula laevis, and Populus tremula. Isolated Pinus sylvestis<br />
trees are also occasionally encountered in the spruce-oak type.<br />
3.1.5 Crops.<br />
Only about half of the area around <strong>Chernobyl</strong> is suitable for agriculture and much of it is in hay<br />
meadows and grazing land (Trifolium spp. and grasses) and only fodder crops, including corn<br />
(Zea mays) and fodder beets LBeta Vulgaris) (Symons 1972; USSR 1987). Winter wheat<br />
(Triticum aestivum), winter rye (Secale cereale), millet (Panicum miliaceum), winter and spring<br />
barley (Hordeum vulgare), oats (Avena sativa), flax (Linum usitatissimumn), hemp (Cannabis<br />
sativa), sugar beets (Beta vulgaris), buckwheat (Fagopyrum esculentum), and potatoes (Solanum<br />
tuberosum) are all grown from human use in the southern part of the Byelorussian SSR and<br />
northern part of the Ukrainian SSR (Sanbur and Kovalenko 1969; Fullard 1972; Symons 1972;<br />
Dewdney 1982). However, these crops are not grown in the poorly drained, boggy parts of the<br />
Poles'ye (Dewdney 1982). At any one time, 12-17% of the land is being fallowed (Symons<br />
1972).<br />
3.2 TASSELED CAP TRANSFORMATION OF LANDSAT IMAGERY.<br />
As a first step in the analysis of vegetation in Landsat imagery, it is useful to transform the<br />
spectral bands of the Thematic Mapper (TM) multispectral imaging sensor to a new spectral<br />
coordinate system that compresses the useful information into fewer bands and provides a physical<br />
interpretation of the transformed bands. We use the so-called Tasseled Cap transformation.<br />
3.2.1 Description of the Tasseled Cap Transformation.<br />
The Tasseled Cap (TC) transformation for the six reflective spectral bands of the Landsat<br />
Thematic Mapper imaging multispectral sensor was first described by Crist and Cicone (1984). It<br />
is a descendant of the Kauth-Thomas transformation developed for agricultural and vegetation<br />
scenes for the Multispectral Scanner (MSS) imaging device on earlier Landsat missions. The TC<br />
transformation is linear and preserves the Euclidean relationship of the pixel data contained in the<br />
six reflective TM bands. It is chosen so that most of the spectral-temporal variance in agricultural<br />
19
and vegetation scenes is included in the first three TC bands. Furthermore, the linear combinations<br />
of the original TM band intensities that provide the first three TC band intensities are chosen so that<br />
the magnitude of each of the three has a physical internretation with respect to vegetation and soils.<br />
Later Crist and coworkers (Crist et al., 1986), adjusted the fourth, fifth and sixth components of<br />
the tasseled cap transformation to maximize the correlation of the fourth component with<br />
atmospheric haze. Table 3-1 lists the resulting transformation coefficients. The coefficients are<br />
used according to the following example:<br />
TC1 = 0.2909*TM1 + 0.2493*TM2 + ... + 10.3695,<br />
where TCi represents the ith Tasseled Cap band intensity and TMi represents the ith Thematic<br />
Mapper band intensity.<br />
Table 3-1. Original Landsat-5 TM Tasseled Cap coefficients (Christ et al., 1986).<br />
Tasseled cap Coefficients Additive<br />
Band Feature TM1 TM2 TM3 TM4 TM5 TM7 Term<br />
TCI Brightness .2909 .2493 .4806 .5568 .4438 .1706 10.3695<br />
TC2 Greenness -.2729 -.2174 -.5508 .7221 .0733 -. 1648 -0.7310<br />
TC3 Wetness .1446 .1761 .3322 .3396 -.6210 -.4186 -3.3828<br />
TC4 Haze .8461 -.0731 -.4640 -.0032 -.0492 .0119 0.7879<br />
TC5 Fifth .0549 -.0232 .0339 -. 1937 .4162 -.7823 -2.4750<br />
TC6 Sixth .1186 -.8069 .4096 .0571 -.0228 .0220 -0.0336<br />
Figure 3-3 illustrates the transformation from the Landsat TM images presented in Volume 2 to<br />
the tasseled cap images analyzed in this volume. The first four TC features have interpretations in<br />
terms of the physical features brightness, greenness, wetness and haze:<br />
Brightness (Br).<br />
Brightness is the overall reflectivity with band-by-band weighting; red (TM3) and near infrared<br />
(TM4) bands are emphasized; soils are generally brighter than vegetation; pines forest has a<br />
relatively low brightness compared to other vegetation.<br />
Greenness (Gr).<br />
Greenness is a basic vegetation index, dominated by the difference between the near infrared<br />
(TM4) and red (TM3) band intensities; chlorophyll and other plant pigments absorb red light but<br />
strongly reflect near infrared light giving foliage a high greenness value. In principle, the difference<br />
20
Landsat Thematic Mapper Image<br />
- Geocoded and resampled to 25 m square pixels<br />
- Pixel vector = Intensity in seven (7) spectral bands<br />
" Six (6) reflective bands<br />
Band Intensity Spectral Region Wavelength (microns)<br />
TM1 Blue 0.45 - 0.52<br />
TM2 Green 0.52 - 0.60<br />
TM3 Red 0.63 - 0.69<br />
TM4 Near Infrared 0.76 - 0.90<br />
TM5 Short Wave Infrared 1.55 - 1.75<br />
TM7 Short Wave Infrared 2.08 - 2.35<br />
"• One (1) emissive band<br />
TM6 Thermal 10.4- 12.5<br />
Tasseled Cap<br />
spectral transformation<br />
I (Linear transformation<br />
on the six reflective bands)<br />
Tasseled Cap Image<br />
- Tailored for agricultural and vegetation scenes, same pixel size<br />
- Pixel vector = Intensity in seven (7) features<br />
" Six (6) reflective features<br />
Band Intensity Feature Abbreviation<br />
TC 1 Brightness Br<br />
TC2 Greenness Gr<br />
TC3 Wetness Wt<br />
TC4 Haze Hz<br />
TC5 (minimal information) -<br />
TC6 (minimal information) -<br />
"* One (1) emissive band<br />
TC7 Thermal (same as TM6)<br />
Figure 3-3. Illustration of the Tasseled Cap spectral transformation for Landsat Thematic<br />
Mapper images; transformation coeffcients are listed in Table 3-2.<br />
21
etween green light intensity and either blue or red provides a similar index. However, the<br />
contrast is higher between near infrared and red with the added benefit that atmospheric scattering<br />
introduces less background noise in the infrared than at the shorter visible wavelengths.<br />
Wetness (Wt).<br />
Wetness increases with soil and vegetation moisture content; it is predominately determined by<br />
the difference between the combined red/near infrared reflectivity (TM3 + TM4) and the combined<br />
short wave infrared reflectivity (TM5 + TM7); it is affected by absorption bands of water and<br />
structural effects related to foliage hydration; there is some contribution from shadowing which<br />
relates to vegetation height and stand density.<br />
Haze (Hz).<br />
Haze is aerosol scattering index that emphasizes the difference between blue light reflectance<br />
(TM 1) and red light reflectance (TM3); Rayleigh scattering by small aerosols strongly scatters light<br />
at shorter (blue) wavelengths providing a large haze signal from clouds, mist, smoke and aircraft<br />
contrails; on clear days the largest "haze" signals come from urban/industrial surfaces, snow and<br />
ice.<br />
Our image processing software (see Volume 2) requires band intensities in the range 0 to 255.<br />
We modified the tasseled cap transformation slightly to insure that all TC band intensities for the<br />
Landsat <strong>Chernobyl</strong> images fall in this range. Table 3-2 shows the resulting transformation<br />
coefficients. Band 1 requires a multiplicative scale factor of 1/2 applied to the corresponding<br />
coefficients in Table 3-1 to reduce its dynamic range below 2"5. Tasseled cap bands 2 through 6<br />
require an additive term to eliminate negative values. These changes do not effect the linearity or<br />
physical interpretation of the TC transformation.<br />
Table 3-2. Modified Tasseled Cap coefficients used in the present work to place all feature<br />
intensities in the range 0 to 255.<br />
Tasseled cap Coefficients Additive<br />
Band Feature TMJ TM2 TM3 TM4 TM5 TM7 Term<br />
TCI Brightness .1455 .1247 .2403 .2784 .2219 .0853 5.2<br />
TC2 Greenness -.2729 -.2174 -.5508 .7221 .0733 -.1648 160.<br />
TC3 Wetness .1446 .1761 .3322 .3396 -.6210 -.4186 128.<br />
TC4 Haze .8461 -.0731 -.4640 -.0032 -.0492 .0119 64.<br />
TC5 Fifth .0549 -.0232 .0339 -. 1937 .4162 -.7823 160.<br />
TC6 Sixth .1186 -.8069 .4096 .0571 -.0228 .0220 160.<br />
22
Finally, we note that the Tasseled Cap transformation is named for the distinctive shape in the<br />
brightness-greenness plane of a scatter plot of crop and soil pixels from an agricultural scene.<br />
3.2.2 Tasseled Cap Images of the Analysis Area.<br />
The change detection analysis in Volume 2 covers a 12.8 km square area consisting of 512 x<br />
512 pixels centered on the main westward trace of initial local radioactive deposition caused in the<br />
first hours by the reactor explosion and fire. For the analysis of radiation doses presented in this<br />
volume, we have extended the dimensions of the analysis area threefold to a 38.4 km square<br />
consisting of 1536 x 1536 pixels centered at the same spot. Thus, the area analyzed in Volume 2 is<br />
the central 1/9th of the area analyzed here.<br />
Table 3-3 lists chronologically the 11 images analyzed and the time of each image relative to the<br />
day of the reactor explosion. References to date numbers or image numbers in this report are made<br />
according to this table.<br />
Table 3-3. Landsat scenes analyzed and time of scene relative to reactor explosion.<br />
Time relative to reactor<br />
Imageldate<br />
explosion,<br />
number Date Days<br />
1 6/06/85 -324 (-10.7 months)<br />
2 3/21/86 -36 (-5.1 weeks)<br />
Day of <strong>Accident</strong> 4/26/86<br />
3 4/29/86 3<br />
4 5/08/86 12<br />
5 5/24/86 28 (4.0 weeks)<br />
6 5/31/86 35 (5.0 weeks)<br />
7 10/15/86 172 (5.6 months)<br />
8 12/02/86 220 (7.2 months)<br />
9 5/11/87 380 (1.04 years)<br />
10 9/07/87 499 (1.4 years)<br />
11 5/28/88 763 (2.1 years)<br />
Figures 3-4 through 3-9 present the 1 TC-transformed images of the 38.4 km square area in<br />
chronological order. These false color images were produced by displaying the first three tasseled<br />
cap features, brightness, greenness and wetness in red, green and blue, respectively. For<br />
shorthand notation, we use (R,G,B) = (Br,GrWt). Table 3-4 gives the geographic coordinates of<br />
the area included in the images in the Universal Transverse Mercator (UTM) coordinate system.<br />
As usual, north is up and east is the right in these images.<br />
23
a) DATE 1: 6 JUNN 85<br />
b) DATE 2: 21 MAR 66<br />
Figure 3-4. Tasseled Cap false color images, (RG.B) =(BrGrWt) of 38.4 km square area<br />
around <strong>Chernobyl</strong>: a) early summer image one year preaccident and b) late winter image 5 weeks<br />
preaccident.<br />
24
a) DATE 3: 29 APR,86<br />
b) DATE 4: 13 MAY 86<br />
Figure 3-5. Tasseled Cap false color images, (RG,B) =(BrGr,Wt) of 38.4 )an square area<br />
around <strong>Chernobyl</strong>: a) 3 days postaccident and b) 12 days postaccident.<br />
25/26
a)DATE 2 4 MIAY 86<br />
b) DATE 6: 31 MAY 86<br />
Figure 3-6. Tasseled Cap false color images, (RG,B) =(BrGr,Wt) of 38.4 kmn square area<br />
around <strong>Chernobyl</strong>: a) 4 weeks postaccident and b) 5 weeks postaccident.<br />
27/28
A)I-\E~ 15) OCT 86<br />
b) DATE 8. DEC 86<br />
Figure 3-7. Tasseled Cap false color images, (R,G,B) (Br,Gr,Wt) of 38.4 km squam, area<br />
around <strong>Chernobyl</strong>: a) 5.6 months postaccident and b) 7.2 months postaccident.<br />
29/30
i). .XTj'E 9:1 \AY 6i7<br />
b) DATE 10. SEPT 87<br />
Figure 3-8, Tasseled Cap false color images, (R,G,B) =(Br,GrWt) of 38.4 km square area<br />
around <strong>Chernobyl</strong>: a) 1 year postaccident and b) 1.4 years postaccident.<br />
3 1/32
zi) ['ATE 11: 25 MAY 58<br />
b) DATE 11: 228 MAY 88 TC 4,5.6<br />
Figure 3-9. Tasseled Cap (TC) false color images of 38.4 kmn squae area around <strong>Chernobyl</strong> 2.1<br />
years postaccident: a) (R,G,B) = (Br,Gr,Wt) as in Figures 3-4 through 3-8 and b) (RG,B)=<br />
(Hz,TC5,TC6), the last 3 components of the TC transformation.<br />
33
Table 3-4. Coordinates (UTM, Zone 36) of the upper left comers of the comer pixels of the<br />
1536 x 1536 pixel area analyzed in this report.<br />
X<br />
Corner pixel (meters) (meters)<br />
Upper left 277,100. 5,714,900.<br />
Upper right 315,475. 5,714,900.<br />
Lower left 277,100. 5,676,525.<br />
Lower right 315,475. 5,676,525.<br />
Y<br />
Date 1, shown in Figure 3-4a, is a summer image one year before the nuclear accident. This<br />
image is used as the reference image for the change detection analysis presented in Volume 2. The<br />
red patches in Figure 3-4a are mostly soil in agricultural fields. The <strong>Chernobyl</strong> nuclear power<br />
station at the upper left of the cooling pond and the city of Pripyat also show as red. These areas<br />
have high brightness, low greenness, and low wetness in the TC spectral coordinate system. The<br />
darker, mostly bluish areas to the lower left (southwest) of the cooling pond are areas of<br />
predominately coniferous forest. Portions of the forest have been cleared for farmland. The<br />
lighter, more greenish areas sometimes grading into the coniferous forest are predominately<br />
deciduous vegetation.<br />
Date 2, shown in Figure 3-4b, is a late winter image taken five weeks before the accident. Ice<br />
and snow have high brightness, low greenness, and high wetness and appear magenta in the image<br />
since (R,G,B) = (Br,Gr,Wt). The Pripyat River is frozen over and residual snow cover is<br />
apparent bordering the areas of coniferous forest, which are now the strongest green areas of the<br />
image. Clearings in the coniferous forest are snow covered even though farmland is not. This<br />
observation suggests that the snowfall is not recent. It is likely that the floor of the forest is snow<br />
covered but the canopy is snowless. The snow cover will then show through from the satellite<br />
perspective wherever pine trees are thinner. This observation from the image implies that in areas<br />
of high TC greenness, TC brightness and wetness provide a measure of coniferous canopy<br />
closure.<br />
Date 3, shown in Figure 3-5a, is three days after the accident. There are more bare fields on<br />
this date than on Date 1 since it is earlier in the growing season. Likewise, deciduous areas have<br />
greened somewhat, but there is only moderate contrast between the evergreen and deciduous areas.<br />
Dates 4, 5, and 6, shown in Figures 3-5b, 3-6a, and 3-6b, respectively, are range from 12 days to<br />
5 weeks after the accident. By the end of May (Date 6), the contrast between evergreen and<br />
deciduous vegetation has returned to the higher level of Date I a year earlier.<br />
34
Date 7, shown in Figure 3-7a, is apparently after the autumn leaf senescence since the<br />
greenness of the deciduots vegetation has dropped dramatically. Date 8, shown in Figure 3-7b, is<br />
in December after the accident. It is similar to Date 2 except that the Pripyat River is not frozen and<br />
there is no snow cover.<br />
Dates 9 and 10 are shown in Figure 3-8. They are 1.0 and 1.4 years after the accident,<br />
respectively. Date 11, shown in Figure 3-9a, is 2.1 years postaccident. The cessation of<br />
agriculture in the analysis area after the accident is apparent from the absence of any fields with<br />
bare soil after Date 6 (31 May 1986). Such fields show up as bright red areas because of the high<br />
brightness of most bare soils. The red fields in the images before Date 6 are fading as cultivated<br />
fields are overtaken by indigenous vegetation.<br />
On the other hand, the increasing size of the red area around the reactor station on Dates 9, 10,<br />
and II shows land, including forests, being cleared as part of the radiation decontamination effort.<br />
On close examination, Date 7 shows earlier forest clearing along a road that crosses the trace of<br />
highest fallout deposition about 2 km west of Reactor Unit 4. The analysis of radiation response<br />
must account for these cleared areas and avoid mistaking cleared forest for radiation damaged trees.<br />
Figure 3-9b shows the other three bands of the Tasseled Cap image for Date 11 presented as<br />
(R,G,B) = (Hz,5,6). The lack of contrast in this image relative to the first three bands of the same<br />
image presented in Figure 3-9a illustrates how the Tasseled Cap transformation includes most of<br />
the pixel variance for vegetation scenes in the first three bands. The largest contrast in Figure 3-9b<br />
is between the urban/industrial area and the cooling pond, for which the Tasseled Cap<br />
transformation was not optimized.<br />
In summary, the Tasseled Cap images for Dates I through 11 as shown in Figures 3-4 through<br />
3-9 provide the data for all analysis in this volume.<br />
3.3 PRELIMINARY CLASSIFICATION OF EVERGREENS.<br />
This subsection describes the preliminary classification of vegetation in the 38.4 km square<br />
analysis area aimed at the identification of pixels in the image dominated by pine forest. We first<br />
define areas of evergreen vegetation and then use an unsupervised clustering algorithm (see, for<br />
example, Duda and Hart, 1973) in TC space to group the pixels into spectrally-related classes.<br />
3.3.1 Selection of Evergreen Pixels.<br />
In order to provide good discrimination between coniferous and deciduous forest, we construct<br />
a composite image from the preaccident images of 21 March 1986 (late winter) and 6 June 1985<br />
(early summer). Table 3-5 shows the band structure of the composite image. We refer to this<br />
winter/summer composite image as COMP21 since it is generated from Dates 2 and 1.<br />
35
Table 3-5.<br />
Band structure for the winter/summer composite image, COMP2 1, generated from<br />
Dates 2 and 1.<br />
Tasseled cap Tasseled cap<br />
COMP21 Band band(date number) feature name<br />
1 Br(2) Winter brightness<br />
2 Gr(2) Winter greenness<br />
3 Wt(2) Winter wetness<br />
4 Br(l) Summer brightness<br />
5 Gr(1) Summer greenness<br />
6 Wt(1) Summer wetness<br />
7 Th(1) Summer thermal<br />
An obvious difference between coniferous and deciduous forest is that coniferous forest has<br />
moderate greenness in both winter and summer while deciduous forest has high greenness in<br />
summer and low greenness in the winter. Figure 3-10 exploits this contrast by displaying winter<br />
greenness and summer greenness as the complementary colors green and magenta, respectively, in<br />
the same image. Evergreen vegetation, including coniferous forest, appears in varying shades of<br />
bright green. Deciduous vegetation appears magenta.<br />
Because the presentation colors in Figure 3-10 are complementary, pixels that have equal<br />
relative greenness in summer and winter appear as a shade of gray. Very dark shades of green are<br />
associated with the cooling pond, the power station, and some cultivated fields. Other cultivated<br />
fields appear nearly white. The Pripyat River and its meandering stream bed are nicely defined<br />
across the diagonal of the image. The river flows from upper left to lower right. The smaller Uzh<br />
River flows to the right across the lower portion of the image, dipping temporarily off the lower<br />
edge before emptying into the Pripyat River at the lower right corner of the image. Although it is<br />
not readily visible in Figure 3-10, the city of <strong>Chernobyl</strong> lies on the right bank of the Pripyat River<br />
along the north side of the Uzh River.<br />
The false color presentation of COMP21 in Figure 3-10, interpreted in the context of regional<br />
information presented in Section 3. 1, clearly indicates the presence of the large contiguous areas of<br />
pine forest in the vicinity of the <strong>Chernobyl</strong> nuclear power station, especially to the west in the<br />
direction of the main trace of early fallout deposition. To maximize the amount of data, we want to<br />
find all pixels (25 m by 25 m each) whose ground cover is dominated by pine trees. The method<br />
of classifying individual pixels based on their spectral signature is discussed in Volume 2. In<br />
short, we find a representative set of pixels, called a reference site, which defines the mean vector<br />
36
JJ<br />
Abb<br />
ji<br />
Figure 3-10. Evergreen vegetation appears bright green and deciduous and annual vegetation<br />
appears magenta in this false color presentation of COMP21 with (R,G,B) = (5,2,5).<br />
37
and the covariance matrix for the spectral signature of each vegetation class of interest. Maximum<br />
likelihood is then used to assign pixels to the various classes.<br />
Figure 3-11 outlines the procedure for defining classes of evergreen vegetation occurring in the<br />
38.4 km square analysis area. The procedure described in Figure 3-11 is a preliminary<br />
classification and need not account for all pixels in the image. In fact, to limit the complexity of the<br />
task, it is best to preselect pixels of the general type needed for the final analysis. Therefore, we<br />
limit our preliminary classification to pixels of evergreen vegetation.<br />
The selection of evergreen pixels is based on ranges of brightness, greenness, and wetness<br />
from the winter scene of Image 2. The ranges are listed in Figure 3-11. These ranges were chosen<br />
after examination of histograms of values of brightness, greenness and wetness for identifiable<br />
areas in Image 2.<br />
The ranges were adjtisted to cut out areas of water, ice, snow, soil,<br />
urban/industrial areas, and deciduous vegetation while keeping all pixels with a moderate to high<br />
value of wintertime greenness.<br />
As mentioned in Section 3.2.2, because of snow cover on the forest floor, TC brightness and<br />
wetness provide a measure of canopy closure in areas of coniferous forest. By inspection, we<br />
found that the wetness intensity value increased from about 146 to about 184 as the canopy closure<br />
varied from maximum (presumably near 100%) to zero. The wetness value for snow and ice was<br />
distributed mostly between 184 and 216, with a tail extending to 255.<br />
Figure 3-12b shows the set of pixels that pass the brightness, greenness, and wetness cuts<br />
defining evergreen pixels as shown in Figure 3-11. These evergreen pixels are marked in green<br />
over a gray-scale background image. This figure should be compared with Fig :e 3-10 to judge the<br />
effectiveness of the evergreen cuts. Notice that very few pixels in the agricultural fields or in the<br />
bottom land along the river passed the evergreen criteria. Most importantly, the obvious magenta<br />
areas in Figure 3-10 corresponding to deciduous or annual vegetation do not appear in the set of<br />
evergreen pixels.<br />
For the purpose of contrast, Figure 3-12a shows the same presentation as Figure 3-10 except<br />
that the roles of winter and summer greenness are reversed. The resulting false color image<br />
displays deciduous areas in bright green and evergreen areas in magenta.<br />
3.3.2 Unsupervised Clustering of Evergreen Pixels.<br />
As a first step toward identifying pine forest pixels in the analysis area, we use the first six<br />
bands of the winter/summer composite image COMP21 to divide the evergreen pixels into a<br />
manageable number of spectrally-related classes. Manual examination of the shape and structure of<br />
the evergreen pixel distribution in the six-dimensional spectral hyperspace is difficult, so we use an<br />
unsupervised clustering algorithm to divide the structure into ten neighborhoods or clusters.<br />
38
Transformed Images<br />
Preaccident Winter/<br />
for All Dates ,_ Summer Composite<br />
(Tasseled Cap Bands) A Image - COMP2I<br />
(See Table 3-5)<br />
Hayo°n Evergreens 125 _< G(2") _< 255<br />
Image•<br />
No Haze<br />
145 5Wt(2) 5 180<br />
(See Figure 3-12b)<br />
Composite Image<br />
(COMP21)<br />
Classified According to<br />
Coniferigouf/Evergreen<br />
Ten Cluster Signatures<br />
Conpstentag<br />
Figure 3-11.<br />
Procedure for classifying evergreen vegetation.<br />
39
a)Deeidtiou-z, ill gr~een<br />
b) Evergreen areas<br />
Figure 3-12. a) Like Figure 3-10 except color reversal (RG,GB) =(2,5,2) displays deciduous or<br />
annual vegetation in bright green and evergreens in magenta, and b) areas passing the evergreen<br />
ciiteria of Figure 3- 11 shown in green.<br />
40
The avoidance of pixels affected by clouds and haze is discussed in Section 3.4 below. Since<br />
Image 1 and, hence, COMP21 contains significant clouds in the lower third of the image, we use<br />
the haze mask from Section 3.4 to avoid affected pixels in the subset used for unsupervised<br />
clustering.<br />
For unsupervised clustering, we use the ISODATA algorithm, which stands for "Iterative Self-<br />
Organizing Data Analysis Technique" (Tou and Gonzalez, 1974). The multipass algorithm is<br />
based only on the spectral signature of each pixel and not on that of its spatial neighbors in the<br />
image. The algorithm begins with a selected number (ten in this case) of class mean intensity<br />
vectors spaced evenly along a straight line through the region occupied by the evergreen pixels. On<br />
the first pass of the algorithm, all pixels are assigned to the class with the nearest mean. Each class<br />
mean is then moved to the mean value of the pixels assigned to it. On the next pass, class<br />
assignments of all pixels are reevaluated and adjusted based on distances to the new means. The<br />
process is repeated until subsequent iterations produce changes in pixel assignments less than a<br />
specified percentage. The method is reasonably nonparametric and is not biased toward any spatial<br />
location on the image. Results are also reasonably independent of the initial placement of means.<br />
The mean pixel vector and the covariance matrix of the resulting ten classes from the<br />
unsupervised clustering of evergreen pixels were used to classify all pixels of the COMP21 image.<br />
Examination of the spectral signatures of the ten classes and the spatial distribution of pixels<br />
assigned to each class resulted in the elimination of four of the ten as either only marginally<br />
cvergreen or lacking in contiguous areas of significant size. The spectral signatures of the<br />
remaining six classes, numbered 3 through 8 are displayed in Figure 3-13 along with the spectral<br />
signatures of the two forest classes used in the preliminary analysis of Volume 2. Detailed listings<br />
of the spectral signatures appear in Appendix D.<br />
3.3.3 Identification of Evergreen Classes.<br />
The spectral signatures of the six classes retained from the unsupervised clustering are best<br />
distinguished from one another in Figure 3-13 by the plot of Band 5 versus Band 4 (CHAN 5<br />
versus CHAN 4) of the COMP21 image. This is a plot of summer greenness versus summer<br />
brightness. The spectral signatures organize into two branches emanating from Class 5. The 4/3<br />
branch extends to higher brightness and lower greenness and the 6/8 branch extends to both higher<br />
brightness and greenness.<br />
Similarity to the Forest 1 and Forest2 signatures from the analysis of Volume 2 shows that the<br />
4/3 branch corresponds to the large areas of coniferous forest evident in Figure 3-10. Class 5 is<br />
pine forest with the highest canopy closure and the least component of deciduous trees since it has<br />
the lowest brightness of all the classes in both summer and winter and the highest greenness in the<br />
winter.<br />
41
U)<br />
LO<br />
Class 3 has lower greenness and higher brightness than Class 5 in both summer and winter; it<br />
apparently consists of similar pine trees with lower canopy closure (less dense foliage) but no<br />
significant component of deciduous trees which would increase the greenness feature in summer.<br />
The signature of Class 4 lies between Classes 5 and 3 in both summer and winter, indicating that it<br />
consists of the same pine trees with intermediate canopy closure. Thus, the 5/4/3 class sequence is<br />
apparently due to varying canopy closure of similar trees.<br />
Class 6 is adjacent to Class 5, the high density pine forest. In the winter it has essentially the<br />
same spectral signature as Class 4 and so may be inferred to contain about the same density of pine<br />
trees as Class 4. On the other hand, Class 6 has a higher greenness in summer than Class 6,<br />
indicating that the pine trees in Class 6 are interspersed with deciduous trees.<br />
Class 8 is most removed from Class 5 along the 6/8 branch in the summer<br />
brightness/greenness plane.<br />
The very high summer greenness value indicates a dominant<br />
component of deciduous or annual vegetation. On the other hand, the winter greenness value still<br />
indicates a significant evergreen component. Three possibilities are sparse pine trees in deciduous<br />
forest, evergreen shrubs or vines as understory in deciduous forest, and evergreen shrubs with<br />
other deciduous or annual vegetation. In exploratory calculations, we found clear examples of<br />
interspersed pixels of Class 6 and Class 8 where Class 6 showed definite radiation response and<br />
Class 8 did not. Also, by association with water patterns in the image, Class 8 apparently tends<br />
toward lower lying areas. We assume for further analysis that Class 8 contains few or no pine<br />
trees. A reference site for Class 8 is defined and Class 8 is retained in the pixel classification<br />
procedure, however, to minimize the number of Class 8 pixels that might otherwise be mistaken<br />
for Class 6.<br />
Class 7 lies near the 6/8 branch of the summer spectral signatures between Class 6 and Class 8.<br />
With respect to of its spatial distribution, Class 7 tends to occur at the edges of Class 6 and to have<br />
no significantly sized areas of its own. We assume that it consists of mixed pixels rather than a<br />
primary forest type. We do not include it in further analyses because of the lack of a spatially<br />
homogeneous reference site.<br />
3.4 REFERENCE SITES FOR PINE FOREST CLASSES.<br />
Figure 3-14 outlines the procedure for selecting a single reference site for each of the retained<br />
classes of evergreen vegetation. The starting point is the winter/summer composite image with<br />
pixels classified according to the signatures of the classes from the unsupervised clustering<br />
described in Section 3.3.2. Using an interactive mode of the image processing software, polygons<br />
are drawn around several representative areas of contiguous pixels for each of the retained<br />
evergreen classes (3,4,5,6, and 8).<br />
reference sites.<br />
These homogenous polygons are candidates for class<br />
43
Composite Image<br />
Cloud/Haze<br />
(COMP21)<br />
Enhancements<br />
Classified According to (Figures 3-16<br />
Unsupervised Clusters to 3-21)<br />
Forslcatig oer g<br />
HomogneousHaze<br />
Composite<br />
Mask<br />
6, 56and 8<br />
Si4for Classes 43,<br />
Rersnaivue 3-4 rcdRefrslcigoeodreference Raiatioe onor<br />
Spectral Signtur ~o fo ac Soiteros/eren<br />
oiferos/everrenc foclass. class. oi az/lod az/Cous<br />
Cls4nal4ae
The reference sites will be used to define the normal spectral signature of each pine forest class on<br />
each image date analyzed. As such, the reference sites must not be obscured by clouds or haze on<br />
any date and must not be significantly affected by radiation exposure on any date. Guidance from<br />
the forest response and radiation contours reported in Volume 2 is used to avoid areas of forest<br />
shcwing radiation damage. Images enhanced to show clouds and haze are used to insure a clear<br />
view of each reference site on each image.<br />
3.4.1 Avoidance of Clouds and Haze.<br />
Figure 3-15 shows a cloud and haze enhanced image for Date 1. The presentation shows<br />
clouds and hazy areas in blue and all other areas in shades of yellow and gray. The use of<br />
complementary colors (here, blue and yellow) follows the same principle used to contrast summer<br />
and winter greenness in Figure 3.10. The cloud and haze enhancement contra.sts the Tasseled Cap<br />
haze band with the thermal band from the same date. The TC haze band, as discussed in Section<br />
3.2, provides a high signal from the aerosols in clouds and haze and a generally lower signal from<br />
land and water surfaces. On the other hand, the thermal band provides a high signal from land<br />
surfaces warmed by the sun and a lower signal from clouds and haze, which are at the usually<br />
cooler temperature of the atmosphere. Display of the haze signal in blue and the thermal signal in<br />
yellow results in good definition of clouds and any nonuniform haze as shown in Figure 3-15.<br />
Figures 3-16 through 3-21 show the cloud and haze enhanced images for each date. Only<br />
Dates 1, 3, 5, 6, 9, and 10 have significant clouds or nonuniform haze. As indicated in the<br />
procedural diagram of Figure 3-14, the cloud/haze enhancements are used in two ways. First, we<br />
construct a composite haze mask for the full set of images. A threshold value for the TC haze<br />
feature is determined for each of the six images with clouds or haze such that pixels in obvious<br />
areas of clouds and haze always exceed the threshold. With these thresholds, we create a single<br />
numerical mask that marks each pixel of the analysis area that is affected by clouds or haze on one<br />
or more dates. This haze mask is used during the selection of candidate reference sites to limit<br />
consideration to areas that are clear of clouds or haze on all dates.<br />
Second, the individual cloud/haze enhancements are used in the final choice of a single<br />
reference site for each class to double check that the chosen site is always free of clouds and haze.<br />
This step is accomplished by plotting the polygons of the candidate reference sites directly on the<br />
cloud/haze-enhanced images.<br />
45
Figure 3-15. Cloud/haze enhancement for Date 1 for the 38.4 km square analysis area,<br />
(R,G,B) = (Th,Th,Hz). Clouds appear blue, warm areas yellow. Thermal gradient appearing as<br />
variation in yellow shade of the cooling pond shows counterclockwise flow of water around<br />
central barrier.<br />
46
a) Trair•ing •ite location•<br />
--- u<br />
•!1 •]"•<br />
•'<br />
Su[]<br />
-<br />
• .I]<br />
SrJm<br />
• -<br />
rJ<br />
b) Date 1: 6 JUN 85<br />
Figure 3-16. a) Reference (control) site locations for Classes 3, 4, 5, 6, and 8 in the 38.4 km<br />
analysis area and b) cloud/haze enhancement for Dam 1 showing three polygons used for<br />
classification merger described in Section 3.5. Red dots are reference site locations.<br />
47148<br />
g J • II III
t<br />
a1 Date 2.: 211 MAR 86<br />
b) Date 3: 29 APR 86<br />
Figure 3-17. Cloud/haze enhancements for a) Date 2 and b) Date 3. Only Date 3 has clouds. Red<br />
dots are reference site locations.<br />
49/50
it) Date 4:<br />
8 MAY 8U<br />
b) Date 5: 24 MAY 86<br />
Figure 3-18. Cloud/haze enhancements for a) Date 4 and b) Date 5. Only Date 5 has clouds. Red<br />
dots are reference site locations.<br />
5 1/52
a) Datet 6- .31 MAY 86<br />
41<br />
INVb)<br />
Date 7: 15 OfT 86<br />
Figure 3-19. Cloud/haze enhancements for a) Date 6 and b) Date 7. Only Date 6 has clouds. Red<br />
dots are reference site locations.<br />
53154
aD Date 8: 2 DE" 86<br />
b) Date 9: 11 MAY 87<br />
Figure 3-20. Cloud/haze enhancements for a) Date 8 and b) Date 9. Only Date 9 has clouds.<br />
Reference Sites 5, 6 and 8 have been moved on Date 9 to avoid a faint jet contrail.<br />
55/56
a) Date 10: 7 SEP 87<br />
b) Date 11: 28 MAY 88<br />
Figure 3-21. Cloud/haze enhancements for a) Date 10 and b) Date 11. Neither date has clouds,<br />
but Date 10 has a jet contrail. Red dots are reference site locations.<br />
57
3.4.2 Representative Spectral Signatures.<br />
During the final selection of reference sites, the spectral signature of each site is examined to<br />
insure that it is truly representative of its class. This step is accomplished by plotting the spectral<br />
signatures of each of the candidate sites with the signature of the overall class aý in Figure 3-13.<br />
3.4.3 Sites Chosen.<br />
Figure 3-16a shows the final choice of primary reference sites for Classes 3, 4, 5, 6, and 8<br />
overlaying a gray-scale image of the analysis area. On this image, the actual size and shape of each<br />
reference site is displayed as a patch inside a circle labeled with the class number of the reference<br />
site. Included on the cloud/haze enhancements in Figures 3-16 through 3-21 for each of the eleven<br />
image dates are five red dots marking the location of the reference sites. For better visibility, the<br />
red dots on the enhanced images are larger than the actual reference sites.<br />
The primary reference sites shown in Figure 3-16a are used on all images except Date 9. On<br />
this date, the primary Reference Sites 5 and 8 are slightly obscured by the remnants of a jet<br />
contrail. Site 5 is moved to a nearby patch in the same forested area and Site 8 is moved across the<br />
river to a similar site. Finally, primary Reference Site 6 is off the edge of the available image on<br />
Date 9. It is also moved across the river to a similar site on this date. Reference Sites 3 and 4 are<br />
unchanged. Site 3 is near, but safely outside, the contrail. Note that a very well defined,<br />
presumably newer, jet contrail appears in the image on Date 10. Date 5 shows hints of a contrail<br />
along the same flight path. Date 5 also has a broader diffuse band passing north/south over the<br />
reactor cooling pond. It may be an old contrail or a thin cloud.<br />
3.4.4 Other Impacts of Clouds and Haze.<br />
The enhanced images show that Dates 2, 4, 7, 8, and 11 are entirely free of clouds over the<br />
analysis area. These images presumably have varying degrees of atmospheric haze from one date<br />
to another, but the haze is uniform over each image and does not impact the radiation response<br />
analysis.<br />
The occurrence of occasional scattered clouds on the other image dates is tolerable in our<br />
analysis even though the presence of a cloud over a forested area in an image causes an apparent<br />
deviation from normal of the obscured pixels. We avoid difficulties with the apparent deviation<br />
due to clouds by requiring persistence from date to date in the detection of radiation-induced<br />
response. Because the percentage of cloud cover is low and because we have several cloud-free<br />
images, there is no case where coincidence of cloud cover in successive images mimics damaged<br />
forest.<br />
It is interesting to note in passing that the thermal signals from the reactor cooling pond<br />
represented by the shades of yellow in Figures 3-16 to 3-21 show that after the explosion of Unit<br />
58
4, the other three reactors were turned off on Dates 3 through 6. At least one reactor was operating<br />
on Date 7 (October 1986) and thereafter through Date 11.<br />
3.5 PREACCIDENT PINE FOREST CLASSIFICATION MAP.<br />
The classification of evergreen pixels in the winter/summer composite image COMP21 requires<br />
extracting the spectral signature of the reference sites for Classes 3, 4, 5, 6, and 8 from COMP2 1.<br />
These spectral signatures consist of the mean pixel intensity vector and its covariance matrix for<br />
each reference site. With this information, the likelihood that a pixel belongs to each class is<br />
calculated. The pixel is assigned to the class with the highest likelihood as long as a threshold<br />
value is exceeded. Pixels below the threshold value are not assigned to any class. We used the<br />
first six bands of COMP21 for the signatures and the maximum likelihood assignment to classes.<br />
These six bands are the brightness, greenness, and wetness for winter and summer as listed in<br />
Table 3-5. The maximum likelihood method is discussed further in Section 5.1 of Volume 2.<br />
Because Image I is missing one corner and is also affected by clouds as illustrated in Figure 3-<br />
15, the pixel classification from COMP21 is not satisfactory for the whole analysis area. The areas<br />
of unacceptability are outlined in white by three polygons in Figure 3-16b. Fortunately, Image 4<br />
from 12 days after the accident is cloud free and can provide a substitute summer component for a<br />
composite winter/summer image. We have constructed such a composite called COMP24<br />
according to Table 3-5 by substituting Date 4 for Date 1. Although technically not a preaccident<br />
composite, COMP24 serves as a reasonable substitute since the areas of unacceptability in<br />
COMP21 are far enough from the reactor station to show no radiation effects by Date 4. The final<br />
classification uses COMP21 where acceptable and COMP24 elsewhere.<br />
The procedure for generating the final pixel preaccident classification map is illustrated in<br />
Figure 3-22. COMP24 is classified using spectral signatures from its own reference sites in the<br />
same way as COMP21. Classifications from COMP21 and COMP24 are merged by using the<br />
COMP21 pixel assignments everywhere except within the three polygons displayed in Figure 3-<br />
16b where assignments from COMP24 are used.<br />
Figure 3-23 displays the final pine forest classification on a pixel-by-pixel basis for the central<br />
512 x 512 pixel section of the analysis area. Classes 3, 4, 5, and 6 are color-coded according to<br />
the labeled squares in the figure. Unassigned pixels are represented by a gray-scale image taken<br />
from Date 4 to provide spatial background for the classification map. Figure 3-24 provides the<br />
same display at a lower magnification for the full analysis area using the same color code. Class 8,<br />
although used in the maximum likelihood pixel assignments, is not represented in Figures 3-23 and<br />
3-24 because it is presumed to have negligible pine tree content as discussed in Section 3.3.3.<br />
59
Preaccident<br />
Winter/Summer<br />
Composite Image<br />
COMP21<br />
Substitute<br />
Winter/Summer<br />
Composite Image<br />
COMP24<br />
Maximum"• Class _ •Maximum--•<br />
SLikelihood Reference Likelihood<br />
Casfcto Sites lsicto<br />
Pixel-by-Pixel ] Pixel-by-Pixel<br />
Assignments<br />
Assignments<br />
CLASS21<br />
CLASS24<br />
Preaccident<br />
Classification Map<br />
of Pine Forest<br />
Figure 3-22. Procedure for generating final preaccident classification of pine forest.<br />
60
Table 3-6 defines the pine forest classes based on the discussion in Section 3.3.3. According<br />
to Section 3.1, the pine trees are likely to be Pinus sylvestris (ý- -')-h pine).<br />
Table 3-6. Pine forest classes.<br />
Class number<br />
Tree composition<br />
3 Predominately pines, low canopy closure<br />
4 Predominately pines, moderate canopy closure<br />
5 Predominately pines, high canopy closure<br />
6 Mixed pines and deciduous, pine density like Class 4<br />
8 Believed to contain very few pine trees<br />
Figures 3-23 and 3-24 show that the highest density pine trees (Class 5 in dark green) are in<br />
the large areas of forest southwest of the reactor site. These same areas have large sections of the<br />
moderate density pine forest (Class 4 in light green). Mixed forest (Class 6 in cyan) tends to<br />
occupy the borders of these areas. Large areas of mixed forest occur near the southern edge of the<br />
analysis area in Figure 3-24. The patches of pine forest directly west of the reactor station within a<br />
few kilometers are dominated by Classes 3, 4, and 6 with essentially no high density pine stands.<br />
Likewise, across the Pripyat River in tae northeast corner of the analysis area, there is little Class 5<br />
forest.<br />
In conclusion, Classes 3, 4, 5, and 6 are used in Section 7 for the analysis of radiation dose<br />
response ater the explosion of Unit 4 at the <strong>Chernobyl</strong> nuclear power station.<br />
61
Figure 3-23. Final preaccident classification of pine forest for the 12.8 km square area analyzed in<br />
Volume 2. Class numbers according to Table 3-6.<br />
62
63/4k
SECTION 4<br />
RADIONUCLIDE FALLOUT FROM THE CHERNOBYL ACCIDENT<br />
This brief section summarizes the release, deposition, and composition of radioactive fallout<br />
from the <strong>Chernobyl</strong> accident.<br />
,*.1 RELEASE AND DEPOSITION.<br />
The most reliable description of the daily radionuclide release from <strong>Chernobyl</strong> is provided in<br />
Table 4.13 of Annex 4 of USSR (1986). Some 24% of the total release occurred on April 26,<br />
1986. During the next five days (4/27-5/1), the release dropped considerably, declining each day.<br />
From 5/2 to 5/5, the release rate increased again, peaking on 5/5 to a level roughly 2/3 that on April<br />
26. Then, successful mitigative efforts reduced releases on 5/6 and afterward to levels<br />
insignificant in comparison to the releases over the first 10 days.<br />
During the initial 10-day release, the primary wind direction changed almost daily. On<br />
April 26, the wind was toward the west. During the ensuing nine days, it shifted in clockwise<br />
fashion toward the north, then the east, the south, and southwest (Izrael, Petrov and Severov,<br />
1987; NEA, 1987). This indicates that fallout deposition to the west of the reactor occurred<br />
primarily on April 26. The report of Izrael, Petrov and Severov (1987) strongly suggests an<br />
instantaneous spread of radioactive debris directly to the west immediately after the explosion.<br />
Continuing releases over the next 24-36 hours appeared to have been spread in directions ranging<br />
from southwest to northwest. There appears to have been no westward plume during the 10-day<br />
release after 12 noon on April 27th. This information indicates that the source term relevant to the<br />
westward-extending plume can be described as an acute (< 1 day) deposition episode occurring<br />
early in the day on April 26.<br />
4.2 RADIONUCLIDE COMPOSITION.<br />
The radionuclide composition of the close-in fallout can be approximated by the discharges<br />
reported by the Soviet experts (USSR, 1986). The primary radionuclides listed by the Soviets<br />
include 1311, 134, 137 Cs, 99 Mo, 95 Zr, 103, 1 06 Ru, 1 40 Ba, 141, 144 Ce, 89, 90 Sr, and 2 39 Np. Several<br />
of these radionuclides have short-lived radioactive daughters in equilibrium with them, which<br />
should be add to the list (DOE, 1987). These daughter products include 9 9 mTc, 9 5 Nb, 1 0 6 Rh,<br />
140 La, 1 44 Pr, and 90 Y. In addition, measurements of airborne radioactivity outside the Soviet<br />
Union indicated the presence of 1331 in large quantities (Lange, Dickerson and Gudikson, 1987),<br />
so this radionuclide was also added to the list for our dose rate analysis. It is possible, and is in<br />
fact likely, that radionuclide fractionation within the debris occurred after the material escaped to<br />
the atmosphere (DOE, 1987); thus the radionuclide composition of fallout debris could differ<br />
65
somewhat from that reported by the Soviets. A major effect is the condensation of volatile<br />
radionuclides such as 1311, 132 Te, and 137 Cs onto refractory particles (DOE, 1987). However,<br />
this is a kinetic process subject to complexities of time, space, microphysics, and meteorological<br />
conditions not specifically addressed in this effort. Therefore, in the absence of specific<br />
information relative to the dominant, westward plume, it is assumed that the relative quantities of<br />
radionuclides in the fallout that affected the vegetation are the same as those reported by the Soviet<br />
experts (USSR, 1986), with the addition of radioactive daughter products and 1331, as noted<br />
above. These radionuclides are used for dose and dose rate calculations in Section 5 and Appendix<br />
A where tables of relative quantities and radioactive properties are provided.<br />
66
SECTION 5<br />
RADIOBOTANICAL AND DOSIMETRIC CONSIDERATIONS<br />
This section discusses radiobotanical and dosimetric considerations regarding the observation<br />
and interpretation of radiation damage to pine forest. Tree characteristics such as age,<br />
morphology, and annual growth cycles play an important role. The interplay of ionizing particle<br />
range and depth of sensitive tissues in trees determines how the forest responds to the distribution<br />
of fallout. Detailed calculations are required to judge the relative importance of beta and gamma<br />
radiation for the induction of observable foliage response. As with most biological systems, the<br />
response of pine trees to radiation exposure depends not only on dose but on the rate at which the<br />
dose is delivered. Finally, specific characteristics of the radionuclide mix and how it was released<br />
during the <strong>Chernobyl</strong> accident influence the course of radiobotanical response.<br />
5.1 TREE CHARACTERISTICS RELEVANT TO RADIATION DAMAGE.<br />
Our satellite imagery analysis and reports such as that of Bohlen (1987) leave little doubt that<br />
the predominant radiobotanical effects were observed on pine trees. The dominant species of pine<br />
in the area is assumed (Painter and Whicker, 1993) to be Scotch (or common) pine,<br />
Pinus sylvestris. Figure 5-1 shows an example of a six-foot Scotch pine grown commercially in<br />
Maryland that was cut down in December. The date of the <strong>Chernobyl</strong> accident, 26 April 1986,<br />
suggests that the trees were in the spring growth phase. In this phase the trees are probably the<br />
most radiosensitive to acute or short-term irradiation (Woodwell and Sparrow, 1963). On average,<br />
the growing season begins about 11 April in the region (Anonymous, 1962).<br />
The radiobiological response of pines is also affected by the age and size of the trees, with<br />
seedlings being the most radiosensitive and large, healthy trees the least sensitive (Sparrow,<br />
Schwemrnmer and Bottino, 1971; Sparrow, Rogers and Schwemmer, 1968; McCormick, 1967).<br />
Tree dimensions and leaf density are also important in the context of understanding the fallout<br />
exposure scenario. Higher and denser tree canopies initially retain more of the fallout particles on<br />
foliage (Chamberlain, 1970), also, the lesser is the gamma radiation exposure from fallout on the<br />
ground surface (Beck and de Planque, 1968).<br />
It is most likely that irradiation of pine apical and lateral meristems at <strong>Chernobyl</strong> from fallout on<br />
the foliage was due primarily to beta particles rather than gamma photons based on a variety of<br />
dosimetric studies in the past (Aleksakhin, Tikhomirov and Kulikov, 1970; Broido and Teresi,<br />
1961; Mackin, Brown and Lane, 1971; Kantz, 1971; Rhoads et al., 1969). Even though the<br />
relative biological effect (RBE) of beta and gamma radiation is similar (~1), the distinction is<br />
important because the beta radiation dose rate from fallout on foliage is expected to differ<br />
significantly from that of gamma radiation due to fallout on the ground; and, it is well known that<br />
67
dose rate is a very important modifier of the radiobiological response (Sparrow, Schwemmer and<br />
Bottino, 1971). The dose rate versus time for beta and gamma exposures differ because 1) as time<br />
passes, fallout particles are lost by weathering from foliage (and concurrently build up on the<br />
ground) and 2) the decay schemes depend on radionuclide species and radiation energy spectra.<br />
h~20<br />
Figure 5-1. An example of Pinus sylvestris. A six-foot specimen cut and photographed in<br />
December in Maryland.<br />
No precise information was available for this study on the average height, foliar density, or age<br />
of the radiation-affected pines near <strong>Chernobyl</strong>. However, a forest is not usually referred to as<br />
"forest" unless most of the trees are relatively mature, and several published accounts do refer to<br />
"forest" (Asmolov et al., 1987; Bohlen, 1987). In addition, photo-pictorial documents of the<br />
accident's environs reveal mature trees in the background. Therefore, in this analysis we assume<br />
that most affected trees were mature and averaged 12 to 15 m in height. This assumption is<br />
consistent with a study (Painter and Whicker, 1993) of the area surrounding the <strong>Chernobyl</strong><br />
68
<strong>Nuclear</strong> Plant where the pines can be 20-30 m in height on solid soil but can be stunted to only 6 to<br />
8 m high in the bogs that exist in the vicinity.<br />
We also assume that the canopy was sufficiently dense to intercept initially 50% or more of the<br />
fallout particles. This assumption is consistent with a calculated mean green biomass (dry weight)<br />
of at least 2 kg/im 2 for boreal forests (Rodin, Bazilevich and Rozov, 1975). Using the<br />
Chamberlain (1970) filtration model and a conservative foliar interception constant of<br />
0.4 m 2 /kg (Whicker and Kirchner, 1987), the predicted foliar interception fraction is<br />
1-exp(-0.4 x 2.0) = 55%.<br />
5.2 BETA RAMIATION EXPOSURE OF PINE MERISTEMS.<br />
To address the question of whether beta exposures could be significant for pines, we consider<br />
the geometry of pine needles and meristems. If there were a sufficient thickness of nondividing<br />
tissue to protect the meristem by absorbing the beta energy, then the dominant exposure mode<br />
would be gamma radiation. This issue involves a review of the histology of pines, as well as an<br />
assessment of the beta particle energies of the radionuclides composing the fallout. According to<br />
Biatobok and Zelawski (1976), the shoot apices (within which reside the dividing cells of the<br />
meristematic tissue) vary considerably in shape and dimensions depending on the stage of<br />
ontogenetic and annual development. However, the larger apices in Pinus ponderosa (which has<br />
similar apex morphology to Pinus sylvestris), are about 500 gim in diameter and 120 gim in height.<br />
These apices are usually surrounded by scales some 200-500 gim thick. While some airborne<br />
particles might find their way underneath these scales and lie in direct contact with the apex, most<br />
particles would settle on the outer surfaces of the scales, and their beta particles would have to<br />
traverse this layer to reach the meristem.<br />
From the descriptions and photomicrographs of longitudinal sections of Pinus sylvestris apices<br />
in Biatobok and Zelawski (1976: pp. 207-209), we estimate that the mean distance through<br />
nondividing tissue that a beta particle would have to traverse in order to reach the meristem would<br />
be about 400 im, with a lower limit of 100 gm and an upper limit of about 1200 gim. If the<br />
tissues were fully hydrated (as they should be in late April), the tissue density would be about<br />
1,0 g/cm 3 .<br />
The range of a beta particle having an initial kinetic energy of 0.2 Mev is 0.04 g/cm 2 (Public<br />
Health Service, 1960) or 400 g in water. Thus, beta particles having energies exceeding 0.2 Mev<br />
should, on average, be able to penetrate into the meristem of Scotch pines if the radionuclide is in<br />
contact with the scales directly over the apex. All of the radionuclides under consideration in this<br />
report have beta transformation energies in excess of 0.2 Mev; they range from 0.41 to 4.83 Mev<br />
(Public Health Service, 1960). Indeed, many of these beta particles exceed 0.5 Mev and would be<br />
able to traverse well over 1200 jim of unit density material, which would likely correspond to the<br />
69
upper limit of protection afforded by the nondividing tissue. We conclude from the probable<br />
dimensions of the Scotch pine apices and the energies anticipated of beta particles from the<br />
predominant fallout radionuclides, that beta particle exposure must be considered in the total doses<br />
received by the pine trees around <strong>Chernobyl</strong>. Radiation transport calculations are presented in<br />
Appendix A for dose depths relevant to both needles and meristematic tissue.<br />
5.3 BETA VERSUS GAMMA EXPOSURE.<br />
Evidence from published reports is cited in this subsection to implicate beta particles as the<br />
dominant exposure mode for the apical meristem of pine trees. Estimates of the ratio of the betadose<br />
to the gamma-dose components for various parts of the upper canopy appropriate for pine<br />
forests near the <strong>Chernobyl</strong> nuclear power plant, however, require models and calculations<br />
specifically designed for that purpose. Such calculations are presented in Appendix A and<br />
discussed in Section 5.3.2 below.<br />
5.3.1 Values from the Literature.<br />
Theoretical calculations by Osanov, Tissen and Radzievsky (196% show that beta depth doses<br />
from a mixture of 239 pu fission products at 0.04 g/cm 2 range from 16 to 61 rad/day per tCi/cm 2 ,<br />
depending on time after fission (which affects the radionuclide mix and hence the beta energy<br />
distribution). These values are 0.13 to 0.38 of the beta exposure at a depth of only 0.005 g/cm 2<br />
(50 gtm).<br />
Other theoretical considerations by Broido and Teresi (1961), show that for equal fallout<br />
depositions on skin and the soil surface, "the 13-dose at the surface of a contaminated individual<br />
would be approximately forty times the y-dose measured 1 m above the contaminated surface"<br />
(assumed to be of infinite extent). Elevating the individual to a height 12 to 15 m above the surface<br />
(as is the case for Fine meristems of the upper canopy), would decrease the gamma exposure rate<br />
due to the ground surface fallout to roughly one half that at I m (Beck and de Planque, 1968). The<br />
y-dose component to the skin (or foliage) due to the fallout on the skin would be approximately the<br />
same as that I m above the ground. 13-radiation from ground surface fallout does not contribute to<br />
exposure in the upper canopy since the air between the ground and 10 m up is equivalent to about<br />
1 cm of water and further shielding is provided by the lower branches. Finally, the estimated 13-<br />
to-y exposure ratio is about 40:1.5 or 27 for equal deposition on the foliage and on the ground.<br />
Actual measured doses to vegetation from close-in fallout debris at the Nevada Test Site<br />
showed that beta doses exceeded gamma doses by more than an order of magnitude (Kantz, 1971).<br />
In this case, the vegetation comprised shrubs about 25 cm above the soil. In the same setting at the<br />
Nevada Test Site, Rhoads et al. (1969), demuistrated that mortality of desert vegetation was<br />
caused by beta particles rather than gamria radiation.<br />
70
Finally, the world-renowned Russian radioecologist, R. M. Aleksakhin (Aleksakhin,<br />
Tikhomirov, and Kulikov, 1970) has the following to say about beta radiation damage to trees of<br />
the coniferous forest:<br />
"On the basis of data obtained in a study of global fallout, it has been established that<br />
the coefficient of primary retention of the most important fission fragments in middle-aged<br />
plantings is not less than 40% - in dense conifer plantings radioactive substances may be<br />
completely retained."<br />
"-- the duration of the period of half-purification (weathering half-time) may fluctuate<br />
from two weeks in plantings purged well by the wind and washed by precipitation to three<br />
or four months in dense coniferous plantings."<br />
"--the highest radiation doses (from fallout) will be obtained by the crowns of woody<br />
plants in the topmost layer"<br />
"--the needles and buds are comparable in size with the run length of 13 particles -- and<br />
all cells of these tissues prove ko be accessible to 03 radiation --- a considerable portion of<br />
the 13 energy will be absorbed in meristematic tissue"<br />
"-- with radioactive contamination of the crowns, the main contribution to the radiation<br />
dose of meristematic tissues will be made by 13 radiation."<br />
"-- radiation on the crowns in the topmost layer -- will exceed by ten or more times the<br />
radiation dose of mammals -- under the forest canopy."<br />
5.3.2 Calculations.<br />
An estimate of the dose of radiation from fallout must account for 1) the spectrum of radiation<br />
emitted by the radionuclide mix, 2) the transport of radiation from source to exposed foliage, and<br />
3) the relative exposure rates of beta and gamma particles within the pine tissue. Appendix A<br />
presents detailed calculations for specific geometrical arrangements of fallout source and foliage<br />
accounting for these factors. Additionally, quant-tative estimation of the relative importance of beta<br />
and gamma ray exposures of pine tree tissues requires a model for the distribution of fallout in the<br />
forest. This distribution involves the initial arrangement of fallout radioactivity at the end of the<br />
deposition episode and the time dependent redistribution caused by weathering.<br />
First, we consider the distribution of fallout at the end of the deposition episode. Figure 5-2<br />
illustrates a simple model for visualizing the distribution of fallout particles in the forest. We will<br />
not treat horizontal gradients of fahlout concentratior, in our calculations so horizontal movement<br />
caused by wind is neglected. Fallout particles stick (aadiere) with some probability when they<br />
encounter a surface within the canopy. This probability varies with particle size, shape,<br />
composition and the nature of the encountered surface. The overall fraction of radioactivity that is<br />
71
etained in the canopy at the end of the deposition episode is the foliar interception fraction, f,<br />
discussed in Section 5.1.<br />
Fallout<br />
Canopy><br />
77.7•<br />
Foliar interception fraction f<br />
'.<br />
Ground deposition fraction (1 f)<br />
Figure 5-2. Illustration of fallout distribution described by a foliar interception fraction with<br />
neglect of winds.<br />
The fraction (1-f) of the incident fallout radioactivity not intercepted by foliage is assumed to<br />
fall uniformly on to the floor of the forest. Thus, the two sources of radioactivity are the fallout<br />
assumed to be uniformly distributed in the foliage and the fallout deposited on the ground. We<br />
refer to these two sources as the canopy source and the ground source, respectively. We assume<br />
that the canopy and ground sources have the same radionuclide composition so that both emit the<br />
same spectra of gamma and beta radiation.<br />
In Appendix A, the canopy source is further divided into two contributions to improve<br />
calculational accuracy. The first consists of fallout particles in direct contact (deposited on) the<br />
surface of the foliage element under consideration. The second is the fallout deposited on all other<br />
foliage throughout the canopy volume. Calculations for the first contribution, called the foliage<br />
contact source, use an estimated surface density of fallout on the foliage element. Calculations for<br />
the second contribution, referred to as the canopy volume source, use a mean volume density of<br />
72
fallout distributed uniformly throughout the canopy. For the transport of radiation from the ground<br />
source and from the canopy volume source, the canopy mass (exclusive of tree trunks) is<br />
approximated by its mean density over the volume of the horizontal layer occupied by the canopy.<br />
Table 5-1 summarizes the ratio of beta radiation doses to gamma radiation doses for the foliage<br />
contact source and the canopy volume source. The dose due to the foliage contact source is<br />
overwhelmingly due to beta radiation for all the dimensions of foliage elements chosen. The<br />
homogenized canopy volume source produces beta dose rates that are two or three times larger than<br />
the gamma dose rates at the same location.<br />
Table 5-1.<br />
Ratio of beta to gamma doses to foliage calculated from the results of Appendix A<br />
for fallout retained in the canopy.<br />
Foliage Contact Source<br />
Dimensions offoliage element<br />
Radius (p) Length (21) P3/ ydose tatio at center of element<br />
(cm)<br />
(cm)<br />
0.04 4.5 58.8<br />
0.04 9.0 59.1<br />
0.0625 9.0 46.3<br />
0.10 4.5 34.7<br />
0.15 9.0 25.<br />
0.40 2.0 8.5<br />
Canopy Volume Source<br />
Dose point at: middle of canopy 3.2<br />
top of canopy 2.0<br />
In order to estimate the beta to gamma dose ratio for the total absorbed dose to a foliage<br />
element, doses from all radiation components are summed including the ground source. A model<br />
for relative densities oi the fallout sources is specified and results from Appendix A are utilized<br />
calculate total reference dose rates or doses for foliage elements.<br />
The complications from the wind-driven, horizontal component of fallout motion and the finite<br />
size of forested areas are neglected. The calculated dose rates to the foliage are referenced to the<br />
same amounts of radioactive fallout distributed throughout the canopy mass and on the ground<br />
surface below at the end of the deposition episode. Accordingly, on the basis of a unit source<br />
density, reference dose rates to the canopy are calculated for I P3-particle or y-ray per cm--sec on<br />
the ground surface and 10-3 3-particles or y-rays per cm 3 -sec within the canopy volume, based on<br />
73
vertical thickness of 10 m for the canopy. For summing dose contributions at a given dose point,<br />
the canopy volume source density is scaled in proportion to the foliar intercept fraction f. The<br />
ground fallout source is scaled in proportion to (4-f).<br />
An appropriate model for the areal source density for fallout material in surface contact with<br />
foliar canopy components is less straightforward to formulate than for ground surface fallout<br />
underneath. Complicating factors include the geometry of canopy components, the aerodynamic<br />
behavior of particle trajectories in the canopy, the effective adherence of particles that make contact<br />
with canopy component surfaces, and weathering influences.<br />
In the absence of definitive information regarding the contact source density, we examine the<br />
effect of varying the density between the unit incident fallout density and one tenth of that value.<br />
The value of the initial foliar intercept fraction, f = 0.6, is taken from Kerr et al. (1971) to be<br />
consistent with the average foliage density used for transport calculations in Appendix A. Also,<br />
this value is close to that of 0.55 estimated in Section 5.1. Table 5-2 summarizes the dose rate per<br />
unit source for the various sources and dose points from Appendix A. Summing the doses over<br />
beta and gamma sources scaled appropriately for each dose point gives the dependence of the beta<br />
to gamma dose ratio as a function of the contact source density in the assumed range for each of the<br />
foliage elements.<br />
Figure 5-3 shows the variation of the beta to gamma dose ratio with contact source density for<br />
selected foliage elements at the top of the canopy. Since the dose from the contact source at the<br />
center of the thinner foliage elements is dominated by beta radiation, the beta to gamma ratio for<br />
these elements is strongly affected by the assumed contact source density factor. Thicker elements<br />
are less influenced by the contact source strength because the beta penetration to their centers is<br />
suppressed.<br />
It is likely that the adherence probability and, hence, the contact source density varies from one<br />
part of the foliage to another and is surely close to one for areas of the pine branches that are sticky<br />
to the touch, which are quite common. Smoother surfaces such as the needles are likely have a<br />
lower adherence probability. Also, there is some reduction in fallout flux as it filters through to<br />
lower canopy levels. We assume that a value of 0.5 for the contact source density factor is<br />
reasonably representative and use it for all further calculations.<br />
With the assumption of a contact source density factor of 0.5 and an initial foliar intercept<br />
fraction of 0.6, Figure 5-4 shows how the beta to gamma dose ratio changes with weathering.<br />
Just after deposition, the ground fraction is 0.4; it increases as weathering effects move fallout<br />
particles from the canopy to the ground. The values plotted in Figure 5-4 for foliage elements of<br />
various radii located at the top and middle of the canopy are calculated with the assumption that<br />
74
Table 5-2.<br />
Dose rate per unit source for various sources and dose points. Reference canopy<br />
volume source density is taken equal to the unit ground source density spread over<br />
the assumed 10 m canopy height.<br />
Beta dose rate Gamma dose rate<br />
Dose point<br />
(cGy/h x 10 7 )/unit<br />
source<br />
(cGy/h x 10 7 )/unit<br />
source<br />
Unit Ground Source 12 m 10.1<br />
I(f or y)/cm 2 - sec 7 m- 16.9<br />
1 m 185a 29.4<br />
Reference Canopy 12 m 33.4 16.5<br />
Volume Source 7 m 66.9 20.6<br />
10- 3 (p3 or ))/cm 3 -sec I m 16.5b<br />
Radius of foliage<br />
element (cm)<br />
Unit Contact 0.04 302. 5.13<br />
Source, Cylindrical 0.0625 236. 5.10<br />
Geometry 0.10 173. 4.98<br />
1(P or /cm 2 -sec 0.15 124. 4.95<br />
0.40 32.0 3.78<br />
aBased on J / dose ratio of 6.3 according to Barabanova and Osanov (1990).<br />
bAssumed equal to the value at the 12 m dose point.<br />
weathering lowers the canopy retention for both the foliage contact source and the canopy volume<br />
source in unison and that the ground fraction increases accordingly. Weathering of the ground<br />
fallout into the earth is neglected.<br />
As discussed in Section 5.2, the most sensitive growth tissues of the pine are located in the<br />
apical meristems at the tips of branches (and roots) and are typically less than 0. 10 cm below the<br />
surface of the foliage elements. Figure 5-4 shows that the calculated b-kta to gamma ratio for these<br />
tissues is greater than six in both the middle and upper canopy before the effects of weathering.<br />
Even when half the intercepted fallout has moved to the ground (ground fraction = 0.7), the ratio is<br />
still above four. These calculations confirm the assertions in the literature as reported above that<br />
beta doses dominate for the most sensitive growth tissues of the pine.<br />
75
14<br />
Dose point at canopy top<br />
12 Foliage element<br />
radius = 0.04 cm<br />
o 0.0625<br />
10..1<br />
8e 8<br />
0<br />
6 -<br />
4<br />
2<br />
,- - 0.40<br />
0 I I<br />
0.0 0.2 0.4 0.6 0.8 1.0<br />
Contact source density factor<br />
Figure 5-3. Beta to gamma dose ratio versus contact source density relative to unit areal density<br />
of incident fallout. Curves are for various radii of foliage elements located at the<br />
top of the canopy.<br />
76
0<br />
12 I<br />
10<br />
S8<br />
S80.0625<br />
0.04 cm<br />
a) Canopy top<br />
S6 - 0 10_O~ O_<br />
:• 4 0 .15 - -- - -- . _- .<br />
o<br />
2 ~-<br />
0<br />
0.4 0.5 0.6 0.7 0.8 0.9 1.0<br />
12 1 1 1<br />
b) Canopy middle<br />
10<br />
84<br />
80 0.041cm<br />
S 0.0625<br />
S6<br />
2 0.40<br />
0 , I<br />
0.4 0.5 0.6 0.7 0.8 0.9 1.0<br />
Ground source fraction<br />
Figure 5-4,<br />
The beta to gamma dose (dose rate) ratio at the center of cylindrical foliage elements<br />
of various radii at a) the top of the canopy and b) the middle of the canopy.<br />
77
Lateral meristematic tissue, which is located in the cambium and provides for growth in the<br />
diameter of branches, may be better protected from beta radiation than the apical meristems.<br />
However, Figure 5-4 shows that tissues as deep as 0.4 cm still have beta to gamma ratios larger<br />
than two until substantial weathering has taken place. Figure 5-3 shows that lowering the assumed<br />
contact source density factor from 0.5 to 0.1 has only a weak influence on the beta to gamma ratio<br />
for such deep tissues.<br />
We conclude that beta doses are much more important than gamma doses for the sensitive<br />
growth tissue in the apical meristems of pine trees. At doses below the LD 50 , the dominant effect<br />
of radiation is on the apical meristems. Established growth is outwardly unaffected. Thus, visible<br />
affects may not show up until the growing season when the lack of viable apical meristems inhibits<br />
new growth. We expect that late developing effects from these relatively low doses are likely to be<br />
associated with high beta to gamma ratios for the causative dose.<br />
For doses much higher than the LD 5 0 , however, trees can be completely dead in a matter of a<br />
few weeks. Radiation response in this short time requires more than just sterilization of the growth<br />
tissue. The broader systemic effects associated with acute mortality (involving irradiation and<br />
response of all canopy foliar components) may be more influenced by penetrating gamma<br />
radiation. It is likely that early death of a tree from large doses is associated with lower beta to<br />
gamma ratios (but still not lower than one or two as discussed above) than for later occurring<br />
foliage deterioration at lower doses.<br />
5.4 DOSE RATE EFFECTS.<br />
It has long been recognized that the radiobiological response to a given dose depends on the<br />
time over which the dose is delivered. This dependency is sometimes caused by subcellular repair<br />
which is more effective if the exposure is protracted. In other cases, if exposure times are short<br />
compared to the cell division cycle time, high dose rates are clearly more damaging than low dose<br />
rates because of reduced recovery time during irradiation.<br />
This is particularly true in<br />
physiologically active plants which may have greater recovery potential than dormant plants. In the<br />
case of chronic or long-term exposures however, somewhat different patterns may emerge. For<br />
example, dormant plants might show a greater effect from a long-term irradiation exposure than do<br />
actively-growing individuals; the dormant cells can accumulate a higher total dose and repair<br />
mechanisms may be less efficient (Whicker and Fraley, 1974).<br />
Most of the research on the effects of radiation on pines has involved either acute (< 1 day) or<br />
chronic (> 1 year) exposures, with constant dose rates applied over the specified period.<br />
Unfortunately, neither of these dose rate regimes match the exposure of the pines at <strong>Chernobyl</strong>.<br />
The likely exposure conditions and mix of radionuclides at <strong>Chernobyl</strong> would have produced a<br />
78
declining dose rate with time. The rate of decline was too slow to allow the exposure to be<br />
considered acute, but was too rapid to approximate a chronic exposure.<br />
We have dealt with this dilemma by estimating a credible dose rate versus time curve for the<br />
westward plume at <strong>Chernobyl</strong>. From the integral of such a curve, one can estimate the length of a<br />
comparable constant-rate exposure. Knowing the length of a constant-rate exposure that would be<br />
comparable to <strong>Chernobyl</strong>, one can examine the more relevant literature to estimate the doseresponse<br />
relationship. A few studies on short-term (8-30 day) constant rate exposures of pines<br />
have been conducted (e.g., Monk, 1966; McCormick, 1967; Pedigo, 1963; Miller, 1968; and<br />
Platt, 1963).<br />
These studies involved pine species other than P. sylvestris; however, the<br />
chromosome characteristics of P. sylvestris are similar to those of pines in general and there is not<br />
a great deal of variation among the pines in radiosensitivity (Sparrow, Rogers and Schwemmer,<br />
1968). In addition, dose rate effects per se have been studied (e.g., Sparrow, Schwemmer and<br />
Bottino, 1971; Amiro, 1986) and these data will also be considered in the development of the doseresponse<br />
algorithm.<br />
5.5 DOSE RATE SCENARIO FOR CHERNOBYL.<br />
Because the larger part of the dose to pine meristems is assumed to have been delivered by beta<br />
particles, a normalized beta dose rate function is estimated from the list of radionuclides and their<br />
decay rates, the relative quantities released, the total beta transformation energies, and an assumed<br />
rate of weathering from the foliage. The relative beta dose rate (BDR) at time t (in days) is:<br />
where:<br />
BDR = Y' Ri Ei exp(-ki t) (5.1)<br />
i<br />
Ri = Estimated abundance proportion of nuclide i, MCi (USSR, 1986)<br />
Ei = Energy of beta transformation of nuclide i (Public Health Service, 1960)<br />
ki = Effective loss rate constant of nuclide i, ki = A + Xw, where Xp = physical decay<br />
constant and Xw = weathering rate constant, 0.0495/day (Hoffman and Boes, 1979)<br />
The normalized beta dose rate (NBDR) is calculated from:<br />
=t BDR(t)<br />
NBDR(t) = BDR(t = 1) (5.2)<br />
79
C') Ný Ný ',- Cii N Ni- N<br />
J CD CD t w 0 C N al C-4<br />
1) It<br />
i-C N) N' CM LO N 0' a)<br />
0 '-0 Cl) 00 00C0 000 0 l0<br />
cU0 lr N O4 N () 1, U) 1, () 00 LD U) N U)<br />
cm 0) 00 * 0 mt 0i it (D ONC - U)0<br />
o 0 ~ \ ~ \ '<br />
-V- o I I<br />
)C<br />
ZO 000 00-0 01-00 0-0 0 In 0<br />
it O D - in U) '-T ( CO) NY N NM W) CV) 0 0<br />
o b0 00(0 CD') CO) D It N4 CO) N4 W N Uf) 0 to U)<br />
(0 CD 0 qr V) 0 C)000 U) 0)LO t CM NO-)C<br />
66666 C;ci i 6;C ci 6 c~i 6 uC;<br />
ý0 CD N - M NON vt CD 0 CD) 0 CD N C) 0 0q<br />
0 CD0 C 0) CD OD N`ý L I 0) C14 N In M -N N 0' N<br />
T- ODC) Cf(D UN C\) CD U) 0~d NO) - C1 D 0) CJ C 0) qc<br />
00 ) C ( D U) CD ON N D N -0 0' 00<br />
C ) 0-D CD N- LOOU CD ) U) ON NO CO C 0) '<br />
0) W) MC i-D W C N N W M 4 NO 10MN(0<br />
'1 o Mi Ni C M C') Mt<br />
M( 0i 66o 100 C<br />
ON~~~ 0-0))CD0)iN))0 N<br />
0 0<br />
Nu0 NC CD 010 IN N O CD M' N' U) '-N<br />
13C' 0 ON 0)- M4 C i-CD U) M '4 )q CD CD<br />
0 U) C') C' CD a 0<br />
04' 04 CD 1000 00<br />
(- 0' 0 i0 CY ')D ') 14 ' N U) OU 0 00) LO N O<br />
-), 00-ID CO'q ) N D U C DD I -o N c) 6D6* - Nr -a<br />
C~~~~~~~l~~ 0 0 0 0 C)<br />
0<br />
~<br />
0 t- ) T- 10 OCD tn0N (0(0 ) CU ON U) (D '4- M4<br />
u -<br />
o)'-N O O N 0 0 W 0 0 00 t<br />
C') COD<br />
UL U)CýC<br />
*00<br />
0<br />
N<br />
-J 0<br />
~) U) C ) + + Cu (0)i + +<br />
4) WN~~ (1 ) (00 cc '41) (D ~ + 06-n9<br />
~ ' ' e '<br />
LC) 0)1 0 wZ a ) '<br />
0-<br />
80
The values for Ri, Ei, ki and BDR(t) for each radionuclide (see Section 4) is summarized in<br />
Table 5-3. The normalized beta dose r,:! s (NBDR) are plotted through 60 days in Figure 5-5.<br />
For comparison, there are also shown a normalized gamma dose rate curve based on aerial surveys<br />
(Asmolov et al., 1987) and a normalized gamma exposure rate curve that is calculated for the list of<br />
radionuclides in Table 5-3. In the latter calculation, the relative gamma dose rate (GDR) for<br />
radionuclides on the ground is estimated from:<br />
GDR = 7, Ri Gi exp(-Xpi t) (5.3)<br />
where:<br />
R i = is the same quantity as in Equation 5.1<br />
Gi = gamma dose factor for nuclide i in tR/hr per mCi/km 2 (Beck, 1980)<br />
Xpi = physical decay constant for nuclide i<br />
As with the BDR, the GDR values were normalized to 1.0 at t = 1 day. Table 5-4 summarizes the<br />
calculations for the normalized gamma exposure rate curve.<br />
Inspection of Figure 5-5 reveals that the estimated dose rate curve for the beta component is<br />
very similar to the aerially-measured gamma component for the first two weeks, after which the<br />
beta curve declines more rapidly. The shapes of the calculated and measured gamma exposure rate<br />
curves are similar after the first two weeks. It is not clear exactly how the aerial measurements<br />
were made. If they were taken just above the forest canopy and if a large fraction of the fallout<br />
was initially retained by the canopy, then the aerially-measured curve would include a beta<br />
component and should resemble the calculated beta curve for the first couple of weeks, since both<br />
would be affected by foliar weathering (Xw) as well as the mix of rqdionuclides. The more likely<br />
situation is that the aerial measurements were taken sufficiently above the canopy to avoid turbulent<br />
propeller down wash that might disturb measurement . If this were the case, the measurements<br />
would be largely of gamma radiation. As the fallout material weathered from the foliage and<br />
accumulated on the ground, the aerial measurements would be expected to more closely approach<br />
the calculated gamma component arising from the soil, and this seems to be the case. The<br />
somewhat widening increase with time between the measured curve (dashed) and that from gamma<br />
emitters "on soil" (solid) could reflect weathering, both from the foliage to the ground and from<br />
migration of fallout into the soil over time. Based on this discussion, it appears that the relative<br />
shapes of the three curves in Figure 5-4 are compatible with the essential facts and with plausible<br />
assumptions we have made.<br />
81
1.0<br />
From gamma emitters<br />
- •" on soil (GDR)<br />
"0.1<br />
U _Aerial survey Z.<br />
Z _(Asmolov et al. 1987)<br />
From beta activity<br />
on foliage (NBDR)<br />
0.01<br />
0 10 20 30 40 50 60 70<br />
Time (days after 4/26/86)<br />
Figure 5-5.<br />
Calculated dose rates (normalized to 1.0 at t = I day) from gamma radiation<br />
emanating from fallout on the soil and from beta activity on the foliage.<br />
82
~ 3)TV ~ V C E CV) CE)ý<br />
W C W111 1 w w w w w II<br />
M~ to CE) CV) to UD CE) CO) N4 () It t. C.<br />
L WWWWWbWWWIII LI b L L L L<br />
W CD<br />
N~ 100 0 CE) N- cr -U) IN CE) 0: 0 U- N, Ný<br />
C') co mE t- coE) VDC) in c) NO 0 t q N d<br />
o NU)10ws 4 40; NýCýi-N 0)<br />
4 ~ ~ N It ~C) CM)<br />
ItN CV) NVN O CY V) It Lc<br />
b W WW WWWW W W WWWII W cr)<br />
'R 10 ME W- W- 0 ON CD l 0 'I W N<br />
U D CE) 10'U N 'It "I Pý r- c0i<br />
00 - 0) (D Cl ( 0 Ci -NO'( T- N1 N - cm U<br />
N c4 ) C') CE) CV) N NC' ) N CE) 'r q V<br />
S0 '0 M NCDN 0 t W -tCE)q N 1<br />
nE i-C CE) 00 N ) M M0) N<br />
. . . .<br />
cz<br />
0m<br />
NN<br />
-q- NO N 0- m o m N R<br />
00 aE 0 It NO 0 r- N) ' 0<br />
N C) CE CE CE N N CE) i .t 0- C) 6V<br />
> m 0 o0 N 0 0) N 0 0 m) 100 CN E)<br />
0 0 0C CO) N 0)1 (D Itt CY) 0 0 Y N D<br />
N E) C) CE) N N CE) i-CE CE P, 6<br />
00D ro 10- Nb NL C) 11-0 Ni a,1.11.1O NL )<br />
CM 0O C\ N\ 0'<br />
0 1 .-LWUJWWWWWWUWW 0)<br />
"M CON C C'l ON 00' CE ) N 10N<br />
0) Nl (p i-i '- N c Ný Ný q !<br />
00<br />
CE) CE RN CE) CE ( CE) N R C CE) N) C<br />
M M It N O<br />
N 0 M V<br />
&CE~e'Ow 00<br />
00<br />
W2<br />
u Ei 0 0<br />
o) ,- -<br />
0Y) + + + C<br />
0 L)0 a:cN c Z<br />
n cV)C) CE) CE) 0)LO000 lqT It mi E '<br />
'- -' '-)0)' i i ~ i N NV) itl 1 if<br />
83
None of the three dose rate curves in Figure 5-5 would produce an "acute" exposure. Of the<br />
three, the beta dose rate curve declines the most rapidly and hence provides the shortest effective<br />
duration of exposure. The beta curve is considered the most plausible representation of the actual<br />
dose rate curve experienced by meristems of the damaged pines; its time-integral is plotted in<br />
(Figure 5-6). An exposure following this curve produces 50% of the total exposure in nine days<br />
and some 80% of the exposure in 21 days. A subjective estimate of the exposure time at a constant<br />
rate that would produce a similar biological response is about 21 days. Note that from day 2 to day<br />
12, when nearly 50% of the dose has been delivered, the dose rates of the constant rate and<br />
calculated beta rate curves are quite similar. We assume that the lower dose rates for the constantrate<br />
exposure for the first few days would be roughly compensated for by the higher dose rates<br />
after day 12. After 20-30 days, the additional cumulative exposure from fallout is probably not<br />
very significant in terms of biological response.<br />
From this analysis, we surmise that experiments in which constant-rate exposures are delivered<br />
to mature pines for periods ranging from roughly 2 to 4 weeks during the early growing season<br />
should produce dose-response relationships comparable to those for the <strong>Chernobyl</strong> forest.<br />
5.6 Summary.<br />
Important results from Section 5 regarding the radiobotanical response of the pine forests near<br />
the <strong>Chernobyl</strong> acrdent include:<br />
1) beta radiation doses are expected to dominate gamma doses by at least a factor of six for<br />
late foliage responses (after months) induced by doses less than the LD 50 ,<br />
2) beta radiation dose contributions are expected to be at least one or two times the gamma<br />
contribution for large doses that cause tree mortality within a matter of weeks or days,<br />
3) the effective exposure time for foliage doses from beta radiation near the power station<br />
determined by the decay of radionuclides and weathering of fallout from the foliage is<br />
estimated to be about 3 weeks, and<br />
4) results of constant dose rate experiments with pines exposed during their growing seasons<br />
for periods of 2 to 4 weeks may be used to interpret the observed responses at <strong>Chernobyl</strong>.<br />
84
0<br />
z<br />
cc$<br />
_0 -0<br />
cu<br />
co<br />
S -4<br />
CLD<br />
xN<br />
o<br />
a c<br />
0o<br />
ci) 0<br />
aso<br />
013ils<br />
asopeleqAII~nwE<br />
0-.0'0<br />
485/8
SECTION 6<br />
DOSE-RESPONSE RELATIONSHIPS FOR PINE TREES<br />
Based on a review of radiobotanical literature for pine tree response to ionizing radiation<br />
exposure, this Section examines the dependence of pine tree response on dose and dose rate and<br />
the relationship between time-to-response and dose for short term exposures relevant to the<br />
<strong>Chernobyl</strong> accident. A quantitative relationship is derived relating the time of earliest detection of<br />
response in multispectral imagery to the dose received by pine trees.<br />
6.1 LITERATURE REVIEW.<br />
Stress to foliage changes its spectral reflectance in various wavelength bands. Temperature,<br />
water content and leaf pigments largely determine the observed spectrum. Reflectance changes<br />
observed in the tree canopy in response to radiation stress are probably primarily due, initially, to a<br />
reduction in photosynthesis and an increase in dark respiration (Ursino, Moss, and Stimac, 1974).<br />
These effects are caused by damage to shoot apical meristems which retards growth and function<br />
of photosynthetic tissues (Bostrack and Sparrow, 1969). Eventual death of trees likely results<br />
from starvation brought about by a critical reduction in the amount of photosynthetic tissue<br />
(Bostrack and Sparrow, 1970). Thus, dying and dead trees should reveal a progressive decline in<br />
the amount of green tissue and water content of the foliage. As the foliage dries and falls from the<br />
branches, reflectance should become progressively similar to open fields or even to bare ground if<br />
the understory vegetation is sparse. The rate of progression of these changes is expected to depend<br />
on dose and season of the year. Because of the nature of these changes, it is appropriate to focus<br />
on mortality and growth reduction as the damage endpoints likely to be revealed by satellite<br />
images.<br />
Based upon measured chromosome characteristics, Sparrow, Schwemmer, and Bottino (1971)<br />
estimated the acute LD50 exposures for 82 woody plants, including 13 species of pine. The<br />
predicted LD 50 exposure for Pinus sylvestris was 620 R. The comparable values for the other<br />
pines ranged from 410 to 770 R, with an overall mean of 615 ± 97 (1 s.d.) R. These values<br />
were all for a 16 hour constant rate exposure. Sparrow, Schwemmer, and Bottino (1971) also<br />
provide some data on the total dose required to produce an LD 50 for a given constant rate exposure<br />
time. The data were not developed from pines, but pines are expected to behave similarly. The<br />
relevant data are plotted in Figure 6-1 relative to the LD 5 0 for a 16-hour constant rate exposure.<br />
Unfortunately, the data do not reflect exposure times over 36 hours. A linear extrapolation to<br />
exposure times of 2-4 weeks, however risky, predicts LD 5 0 values for Scotch pines of<br />
1300-1500 R.<br />
87
00<br />
_ _ _ _ _ _<br />
_ _ _ _ _ _ _ _ _ _)<br />
E<br />
oo0<br />
-CI<br />
t<br />
00<br />
03 u0<br />
CL 0<br />
0)nsdx q 9L -'-10C-<br />
88L
Sparrow, Schwemmer, and Bottino (1971) also indicate that a fallout decay simulation (FDS)<br />
treatment is comparable to an 8-hour constant rate exposure.<br />
However, this is based on<br />
experiments in which the decay rate followed the t -1.2 relationship with I hour as the initial<br />
reference time. In this case, the exposure rate at 1 hour would be 16 times that at 10 hours. There<br />
is no indication, to our knowledge, that the initial dose rates at <strong>Chernobyl</strong> were that much higher<br />
than the 10-hour dose rates. This would require a very acute deposition of fresh fission products<br />
(as from a weapon explosion) with no prior inventory buildup of longer-lived fission products as<br />
would be expected with used reactor fuel.<br />
The study by Miller (1968) appears quite relevant because it involved a 29-day constant rate<br />
gamma exposure to Pinus palustris trees during the month of April. Sixty percent of the<br />
population survived an exposure of 2,100 R. Complete mortality occurred within three months at<br />
8,700 R or greater. We estimate the LD 5 0 to be about 2,600 R from a plot of survival versus total<br />
exposure. Noticeable (- 12%) mortality occurred at an exposure of 700 R. A 50% reduction in<br />
growth was observed at approximately 500 R. In Miller's study, 29-day irradiations of nearby<br />
plots were also conducted in summer, fall, and winter. The LD 5 0 for the summer and fall<br />
irradiations was only slightly higher than that of the spring irradiation. In the case of growth<br />
reduction, the spring, summer and fall irradiations gave nearly identical results. In winter, the<br />
pines were considerably more resistant.<br />
Monk (1966) also studied the response of Pinus palustris to a constant rate gamma exposure.<br />
In this study, the trees (which were 5 years old) were exposed for 16 days in mid-May. A 50%<br />
reduction in growth was observed at 400 R, a value similar to that reported by Miller (1968). The<br />
exposure required to kill all the trees by the end of the growing season was about 9,300 R, again<br />
similar to the Miller (1968) study. Data on partial mortality were not reported.<br />
A third study of Pinus palustris was carried out by McCormick (1967). In this case, the forest<br />
was exposed at a constant rate for 8.3 days during August. All these pines < 5 years of age<br />
receiving 800 R or more died within 4 months. Older trees (up to 12 year old) required 2,800 R<br />
for complete mortality. This forest also contained some Pinus ellottii. All individuals of this<br />
species receiving > 300 R died and a clear relationship between plant size and radiosensitivity was<br />
observed. In all cases, the larger trees were more resistant. Microenvironmental changes were<br />
also observed at various exposure levels. In areas receiving > 2,000 R, temperature gradients<br />
were more like those of open fields than forests.<br />
The effects of mixed neutron-gamma radiation in air emanating from the Lockheed Aircraft<br />
Corporation reactor in northern Georgia on Pinus taeda and P. rigida were studied by Platt (1963)<br />
and Pedigo (1963). Most of the exposures were delivered in a 2-week period in June 1959 and a<br />
3-week period in August 1960, so the dose rate regime should be relevant to <strong>Chernobyl</strong>. After the<br />
June irradiation, pines receiving more than 7,500 rads turned reddish-brown within a few days and<br />
89
were dead within a few weeks. Those receiving about -',000-6,000 rads took much longer (up to<br />
8-10 months) to dic. Doses of about 1,000-3,000 rads caused death of terminal buds, inhibited<br />
reproduction and growth, and reduced photosynthesis, but caused little mortality. A complication<br />
in this Lockheed reactor study was the presence of neutrons, which may have an RBE greater than<br />
1, but we have no data to assess this effect for pine mortality or growth. The gamma/neutron dose<br />
ratios reported in Platt (1963) ranged from 0.5:1 to 3.0:1, depending on location. If the neutron<br />
RBE is greater than 1 for these effects, the doses of pure gamma radiation needed to produce the<br />
same effects would be higher.<br />
Donini (1967) studied the histological response of five year old Pinus pinea and P. halepensis<br />
to various total doses and dose rates. From such data, the relationship of dose rate to total<br />
exposure time to produce plant death was estimated. The exposure times ranged from 25 to 380<br />
days for P. halepensis. Both curves plotted as a straight line on a log-log scale and were thus<br />
extrapolated back to a 20-day exposure period. This yielded a lethal total exposure of 860 R for P.<br />
halepensis and 3,000 R for P. pinea. These species have nuclear volumes of 1,000-1,100 4tm 3<br />
(Donini, 1967), from which one would predict acute (16 hr) LD 5 0 values of roughly 700 R<br />
(Sparrow, Rogers, and Schwemmer, 1968). The large discrepancy in radiosensitivity between the<br />
two species is puzzling in view of their similar nuclear volumes. Nevertheless, it is evident that to<br />
produce an LD 50 a 2-3 week irradiation requires a somewhat to much larger dose than does a<br />
16-hour exposure.<br />
A nine-day gamma exposure in autumn of a pine birch stand in the Soviet Union resulted in an<br />
LD 1 00 for pine of 5,000-6,000 rads (Karaban et al., 1978). The threshold dose for obvious<br />
damage to the buds and needles was about 800 rads.<br />
Two excellent chronic gamma irradiation studies on pine forests were carried out by Woodwell<br />
and Rebuck (1967) and by Amiro and Dugle (1985). Unfortunately, it is not possible to infer from<br />
their data lethal doses to pines subjected to 2 to 3-week exposures. Nevertheless, both of these<br />
studies support the concept that prolonged exposures are less effective than acute ones.<br />
Approximately 80% mortality to Pinus rigida was observed after 7,000 R was delivered over an<br />
11 -month period (Woodwell and Rebuck, 1967). A total dose of 8,000 rad over a period of about<br />
2 years produced significant (- 80%) mortality in Pinus banksiana (Amiro and Dugle. 1985).<br />
One year old Pinus sylvestris seedlings were exposed to various constant exposure rates for<br />
150 days in Canada (Sheppard, Thibault, and Guthrie, 1982). Needle growth at 17 rad/day<br />
essentially terminated after 60 days with a total dose of about 1,000 rad. Relationships in<br />
Klechkovskii, Polikarpov, and Aleksakhin (1973) indicate that this dose would need to be about<br />
1,700 rad to cause severe growth inhibition and result in total lethality of the population. The ratio<br />
in radiosensitivity for seedlings relative to adult trees is reported by McCormick (1967) to be about<br />
90<br />
MUM
3.5 for P. palustris, so that a dose of about 6,000 rad delivered over 60 days should kill all adult<br />
Scotch pines if P. sylvestris shows the same ratio.<br />
A study by Amiro (1986) on 2 year old Pinus banksiana seedlings examined the effects of<br />
both dose rate and total dose on the relative growth of the plants. Converting his multiple<br />
regression predictive equation to the units used in this report:<br />
RG = 1.072 - (2.303 x 10-4) D - (2.369 x 10-3) D/t<br />
where:<br />
RG = growth rate relative to controls<br />
D = total dose in raf<br />
t = time of the exposure in days<br />
The third term in the equation above is the dose rate factor. For an exposure time of 21 days and a<br />
measurable growth response of RG = 0.8, this equation predicts a dose of - 800 rad. A severe<br />
growth inhibition (RG = 0.2) is predicted for a dose of 2,500 rad. If this dose corresponds to<br />
0.6 of the LD100 (Klechkovskii, Polikarpov, and Aleksakhin, 1973), the LD 100 would be about<br />
4,200 rad for a 21-day exposure.<br />
6.2 ANALYSIS OF DOSE RATE DATA.<br />
We have plotted the relevant data from the literature review discussed above such that all<br />
estimates can be given some weight in developing a dose rate or exposure duration relationship for<br />
assessing the pine tree damage in the <strong>Chernobyl</strong> forest. The total dose required to produce three<br />
endpoints of damage is plotted against the duration of the constant-rate exposure in Figure 6-2.<br />
The LD 100 represents the minimum dose required to produce 100% mortality to the pines, while<br />
the LD 5 0 represents the lethal dose to 50% of the population. The GR 5 0 represents a 50%<br />
reduction in growth rate, relative to controls. Generally, lethal effects were scored within the<br />
current growing season for exposure periods < 60 days. In the case of the longer exposure<br />
periods, the effects were generally scored within 1-3 years after the start of irradiation. Growth<br />
rate effects were usually scored within the current growing season for exposure periods < 60 days.<br />
The lines in Figure 6-2 are drawn by eye.<br />
The point estimates of the dose required to produce the three endpoints of damage for a 21-day<br />
exposure period as read from Figure 6-2, along with uncertainty bounds, are listed in<br />
Table 6-1. The uncertainty bounds are subjective, but they include a 1-week uncertainty in the<br />
effective exposure period at <strong>Chernobyl</strong>, as well as the scatter in the data. The suggested<br />
uncertainty bounds are within a factor of 1.8 of the best point estimate. Because the data are based<br />
91
H i l l I I 1 111 1il 1 1<br />
0~<br />
ElE<br />
00<br />
CIOI<br />
00<br />
00<br />
ot<br />
G 11<br />
-l If3 I.2<br />
cc -00~<br />
0 0 orFR<br />
O~ Z<br />
-0<br />
0.0 ~ 0<br />
0 0"<br />
0<br />
C~j<br />
(E~o)asopje-o<br />
92
on short-term exposures (< 60 days) received in the early growing season, the endpoints in Table<br />
6-1 should be reached by the end of the growing season in late September or early October.<br />
Table 6-1. Estimated total doses to produce three endpoints of damage to pines for an exposure<br />
period of three weeks (assumed equivalent to the effective exposure period at<br />
<strong>Chernobyl</strong>). The upper and lower bounds consider the scatter in the data from the<br />
literature, as well as a I-week uncertainty in the effective equivalent exposure period.<br />
Required Dose (rad)<br />
Endpointa Best Estimate Lower Bound Upper Bound<br />
LD 100 (complete mortality) 6,000 3,300 10,800<br />
LD50 (50 percent mortality) 2,300 1,300 4,100<br />
GR50 (50 percent growth reduction) 800 440 1,400<br />
"aEndpoint observed by the end of the growing season.<br />
We judge that the maximum stress indicated by satellite image processing would correspond<br />
roughly to the LD 100 . Stress at the GR 50 level is probably not detectable. At the GR 5 0, the<br />
vegetation will be hydrated, green, and functional, even though growth rate and reproduction are<br />
impaired. The LD 5 0 is about a factor of 2.6 lower than the LD 10 0 and roughly a factor of 2.9<br />
higher than the GR 50 . The LD 50 will be detectable by satellite image processing certainly by the<br />
time half the trees are dead.<br />
6.3 TEMPORAL PROGRESSION OF DAMAGE.<br />
Most of the relevant studies on the effects of short-term radiation exposure on pines failed to<br />
score the vegetation frequently enough to document the rate of progression of the damage. In most<br />
cases, the damage was simply scored at the end of the growing season, and sometimes the<br />
following season. Fundamental radiobiological considerations would suggest that the rate of<br />
progression of damage should increase with total dose and dose rate, and also with the rate of<br />
mitotic activity.<br />
The study by Platt (1963) is one of the more helpful with regard to time-progression of<br />
damage. Quoting from this author:<br />
"Within one week after the June irradiation, pine trees receiving doses of 7,500 rads or<br />
more began to turn a brilliant orange-red and died within a few weeks. Those receiving<br />
about 4,000 rads took much longer to die."<br />
93
Multiple regression analyses of data from this study indicated that needle production in pines was a<br />
complex function of time after exposure, dose, and time-dose interactions. The work by Miller<br />
(1968), also one of the more relevant studies, showed that 100% mortality occurred within three<br />
months in pines receiving > 8,700 R. Unfortunately the time to death for pines receiving even<br />
higher exposures was not reported.<br />
Pedigo (1963) reported noticeable effects within a few days in pines receiving a 2-week<br />
exposure of 8,000 rads or more. Significant mortality was recorded within one week after the<br />
exposure for pines receiving > 12,000 rads, but not until about 100 days after exposure for pines<br />
receiving 9,000 rads.<br />
In the case of McCormick's (1967) study (an 8-day exposure in August), the older pines<br />
receiving 2,800 R took about four months to die. No additional trees died during the subsequent<br />
two years of the study.<br />
Using the limited database, Figure 6-3 shows the approximate time required to reach the LD 1 oo<br />
or the GR 5 0 of pines for short-term (8 to 30-day) exposure periods plotted against total dose. The<br />
lines are drawn subjectively. The GR 5 0 is offset from the LD 5 0 by a factor of 7.5, as was the case<br />
in Figure 6-2. Considering the scatter in the data and other neglected variables such as growth<br />
stage, temperature, and moisture availability, we estimate that the uncertainty in the time to reach<br />
the endpoint is roughly a factor of 2. The curve would predict that very high doses (> 12,000 rad)<br />
would be required for prompt (< I week) mortality. Doses on the order of 1500 rad or more could<br />
cause growth impairment within a few days.<br />
6.4 DOSE VERSUS TIME-TO-RESPONSE FOR MULTISPECTRAL<br />
DETECTION.<br />
The reports of Monk (1966), Pedigo (1963), Platt (1963), and Miller (1968) provide visual<br />
descriptions and other data on pine tree response to radiation that allow qualitative estimation of the<br />
detectability of radiation response using Landsat Thematic Mapper images. These four reports are<br />
applicable for trees at least five years old and for springtime exposures lasting from two to four<br />
weeks. Table 6-2 lists data extracted from these reports.<br />
For various combinations of absorbed dose and time since start of exposure, Table 6-2 repeats<br />
with minimal paraphrasing the description of damage found in thc indicated Reference. The<br />
qualitative estimate of multispectral detectability for each dose/time combination in Table 6-2 is<br />
based on the descriptions of damage and the authors' experience with vegetation analysis using<br />
multispectral imagery. Factors considered include:<br />
1) leaf senescence in deciduous trees is easily detected, so dead pine trees with brown needles<br />
will be easily detected,<br />
94
10o 3 2<br />
102 0D<br />
04-.<br />
C<br />
40<br />
16<br />
E<br />
10<br />
GR 5 o<br />
LD 1 0o<br />
1 I Iiiiul I I I ill<br />
102 103 104 40,000<br />
Short-term dose (cGy)<br />
Figure 6-3.<br />
The time required to reach the LD 1 00 or GR 50 versus short-term (8-30 day)<br />
dose. Circles represent data for the LDIoO; squares are data for the GR 50 .<br />
Numbers represent the authors: 1 = Pedigo (1963); 2 = Platt (1963); 3 = Miller<br />
(1968); 4 = McCormick (1967); 5 = Sheppard, Thibault and Guthrie (1982);<br />
and 6 = Amiro (1986).<br />
2) new growth needles and old growth needles on pine trees have a different color to the eye<br />
and have infrared spectral reflectivities that differ by 50% to 80% through the growing<br />
season (Wolfe and Zissis, 1978), and<br />
3) minor variations in needle color or number will be masked by pixel-to-pixel fluctuations in<br />
canopy cover and residual misregistration of pixels.<br />
95
Table 6-2. Detectability by multispectral remote sensing of radiation damage to pine trees for<br />
spring exposures as a function of time since start of exposure and dose. Exposure<br />
duration of two to four weeks.<br />
T.ie since<br />
start of Estimate of<br />
Dose exposure multispectral<br />
(Gy) (days) detectabilitya Description of damageb Referencec<br />
3. 330 X No visible evidence of damage, terminal and 3<br />
lateral growth are 60-65% of normal<br />
3.5 90 X All trees alive, 85% survival of terminal 4<br />
buds, terminal growth 70% of normal<br />
5. 142 X Terminal buds alive, stem elongation<br />
reduced<br />
11.6±1.7 142 X Terminal buds dead, proliferation of lateral<br />
bud formation with subsequent growth<br />
20. 20 x No unusual needle fall 2<br />
(16- 23)<br />
7. 90 0 12% survival of terminal buds, terminal 4<br />
growth 15% of normal, 50% survival of<br />
lateral buds, 10% of trees dead (may not be<br />
distinguishable from normal variations in<br />
canopy cover)<br />
25. 280 [] First effects 2<br />
30. 199 0 First effects 2<br />
45. 20 0 Trees shed needles produced during first 2<br />
(16- 23) flush<br />
80. 7 El First signs of color, trees began to turn 2<br />
reddish brown<br />
20. 390 U Pines that had received more than 2000 rads 3<br />
were markedly affected<br />
21. 90 U 0% survival of terminal buds, no terminal 4<br />
growth, 40 % of trees dead<br />
32 ± 6 142 U Trees alive but terminals dead, no lateral I<br />
development<br />
40. 270 U Dead 8 to 10 months following an exposure 3<br />
greater than 4000 rads<br />
42. 90 U 90% of trees are dead 4<br />
45. 390 U 13 months following irradiation, pines that 3<br />
had received more than 4000 to 5000 rads<br />
were dead<br />
75. 28 U Began to turn a brilliant orange-red and died 3<br />
within a few weeks<br />
96
Table 6-2. Detectability by multispectral remote sensing of radiation damage to pine trees for<br />
spring exposures as a function of time since start of exposure and dose. Exposure<br />
duration of two to four weeks. (Continued)<br />
Time since<br />
start of Estimate of<br />
Dose exposure multispectral<br />
(Gy) (days) detectabilirya Description of damnageb Referencec<br />
87. 90 U All trees dead at this dose and higher 4<br />
93. 142 U Trees dead, minimum lethal exposure (at this<br />
time) was 9261 R.<br />
120. 16 U Brilliant red-brown, coloring completed 2<br />
aDetectability with Landsat Thematic Mapper (judgment of authors):<br />
x no, unlikely.<br />
o maybe, uncertain,<br />
[ yes, very likely,<br />
b"Dead" means no green needles whatsoever; terminal and lateral refer, respectively, to the ends<br />
and sides of branches; terminal and lateral buds both contain apical meristem.<br />
cReferences:<br />
1. Monk (1966). 17 day exposure to 137 Cs y-radiation; starting 11 May 1965, about one<br />
month after pine shoot elongation had begun; five-year-old longleaf pine (Pinus palustris).<br />
2. Pedigo (1963): 14 day irregular exposure to output of Lockheed air-shielded reactor. n:7<br />
ranging from 1:1.7 to more than 3:1; dominant erposure during second week starting about<br />
June 14, 1959: loblolly pine (Pinus taeda) forest containing trees with trunks up to at least<br />
12" in diameter.<br />
3. Platt (1963): Report based on same experiment as Reference 2 (Pedigo, 1961).<br />
4. Miller (1968): 29 day exposure to 137 Cs y-radiation; starting April 6, 1966, "at the very<br />
beginning of the growing season;" 8-year-old longleaf pine (Pinus palustris).<br />
Figure 6-4 provides a graphical presentation of the detectability estimates from Table 6-2 with<br />
different symbols representing the three categories of detectability. Both dose and time since stadt<br />
of exposure should be considered as independent variables in Figure 6-4. The detectability data<br />
divides the dose/time plane into three regions corresponding to the three categories of detectability.<br />
97
000 -- f- -*t<br />
- i~rzz<br />
+-4<br />
TI<br />
----- j~ ~<br />
00<br />
Foes Dee __ -r<br />
---<br />
(98
Although the data is somewhat sparse, the general outlines of the three regions are apparent in<br />
Figure 6-4.<br />
Figure 6-4 includes the LD 10 0 line from Figure 6-3 for reference. Because dead in the<br />
literature is defined as having no green needles, it is clear that any patch of forest lying along or to<br />
the right of the LD 1 00 line will be detectable with the multispectral images of the Thematic Mapper.<br />
The other straight line in Figure 6-4 is hand-selected to approximate the boundary between the<br />
likely and uncertain estimates of detectability from Table 6-2. We assume that this line is a<br />
reasonable estimate of the relationship between dose an(.' the time of earliest detectable response.<br />
This choice is conservative on the high side.<br />
A controlled experiment might reveal that<br />
multispectral analysis will detect responses well into the uncertain region of the dose/time plane. It<br />
seems unlikely that the line of earliest detectability would be much closer to the LD 100 line.<br />
We assume for the present analysis that the relationship between dose and time of first<br />
detectable response is linear on a log-log plot, that is, the relationship follows a power law. The<br />
factor likely to cause the biggest deviation from the assumed linearity is the seasonal variation in<br />
detectability arising from the spectral contrast between new and old needles. Consider a patch of<br />
pine trees that has received a dose at the beginning of the growing season large enough to kill most<br />
buds but not enough to significantly affect old growth. According to data presented by Miller<br />
(1968) such a dose would lie in the range of 7 to 15 Gy for a 4 week exposure in April. During<br />
the growing season, the appearance of unexposed trees is dominated by new growth, which would<br />
be absent from the exposed trees. Consequently, in multispectral imagery, unexposed and<br />
exposed trees would differ according to the spectral contrast between new and old needles. This<br />
contrast would fade after the growing season with the onset of cold weather as new needles on the<br />
unexposed trees assume the spectral reflectivity of mature growth. The contrast presumably<br />
returns the following growing season as radiation damage again is manifested by inhibited growtl<br />
and mortality of some trees. Thus, in this intermediate dose range, the contrast between exposed<br />
and unexposed trees and, hence, the detectability of response, probably goes through a maximum<br />
toward the end of the first growing season and a minimum during the following winter.<br />
This potential fluctuation in the detectability of pine tree radiation response in the intermediate<br />
dose range is represented by the s-shaped, dotted curve in Figure 6-4. Detectability occurs at a<br />
dose minimum 3 to 5 months after the start of exposure (during August or September) and then<br />
moves back to higher doses 7 to 9 months after the start of exposure (December to March).<br />
.•',though this fluctuation in detectability may influence the interpretation of our results, we<br />
nevertheless approximate the earliest detectability with a straight line in Figure 6-4 since the<br />
detectability data is sparse and we do not have late summer satellite observations to interpret.<br />
The times of the 9 postaccident images relative to the day of the <strong>Chernobyl</strong> accident are<br />
indicated by the circles along the left edge of Figure 6-4.<br />
99
Figure 6-5 presents a regression analysis intended to quantitatively estimate the boundary<br />
between the likely and uncertain estimates of detectability from Table 6-2. The data for the<br />
regression is selected from Figure 6-4 by taking the left-most occurrences (lowest dose for a given<br />
time interval) of the likely category and all of the points in the uncertain category. The data at 90<br />
days has been excluded in both cases since we do not have an image of <strong>Chernobyl</strong> at that time and<br />
because of the probable temporary increase in detectability at the end of the growing season<br />
discussed above. Finally, the data at 142 days does not have a point in the uncertain category, so<br />
one has been interpolated halfway between the likely and unlikely points. Time of observation is<br />
taken as the independent variable and the dose as the dependent variable for the regression<br />
analysis. The regression line with 68% confidence band is plotted in Figure 6-5. The regression<br />
line is given by<br />
D = 198 t- 0 . 365 (6.1)<br />
where t is the time of first detectable response in days since start of exposure and D is the estimated<br />
dose in Gray. This relationship is our best estimate for detectability with remotely sensed<br />
multispectral imagery.<br />
It does not apply to close visual inspection of trees nor to growth<br />
measurement, both of which would reveal response at lower doses and earlier times.<br />
Table 6-3 provides evaluations of Equation 6.1 for the postaccident image dates analyzed in<br />
this report. Equation 6.1 is based on data ranging from only 1 week to 1 year. Values for times<br />
outside this range are extrapolations as noted in Table 6-3 and are subject to additional uncertainty.<br />
Table 6-3. Dose estimates according to Equation 6.1 for a first detected response corresponding to<br />
the times of the 9 postaccident images presented in this report.<br />
Time postaccident Estimated dose<br />
Image number (Days) (Gy)<br />
3 3 133. (extrapolation)<br />
4 12 80.<br />
5 28 59.<br />
6 35 54.<br />
7 172 30.2<br />
8 220 27.6<br />
9 380 22.6<br />
10 499 20.5 (extrapolation)<br />
11 763 17.6 (extrapolation)<br />
100
1000 __<br />
-- --<br />
2 Symbols as<br />
if<br />
ii -___ ___<br />
nKgure 6 / 4-<br />
__________ - - t<br />
_____ I Jl<br />
100<br />
CD<br />
(/3<br />
0<br />
cz<br />
10<br />
100 Li<br />
first detection, d<br />
Figure 6-5.<br />
Regression line with 68% confidence band for the relationship between dose and<br />
time of earliest detection relative to ihe start of exposure for two- to four-week,<br />
springtime exposures.<br />
101
In conclusion, Equation 6.1 represents the relationship between total absorbed dose and the<br />
time of first detected response relative to the start of exposure. This relationship is used for the<br />
dose estimates presented in this report. It is based mostly on gamma dose.<br />
102
SECTION 7<br />
DOSE DETERMINATION FOR CHERNOBYL FORESTS<br />
This section maps estimated radiation doses to pine forest near the <strong>Chernobyl</strong> nuclear power<br />
station through an analysis of a time series of Landsat Thematic Mapper (TM) images for two years<br />
following the <strong>Chernobyl</strong> accidental nuclear reactor explosion. The first subsection describes a<br />
method for defining spectral deviation from normal for a pixel known to belong to a given class<br />
and presents such deviations based on the 11 Tasseled Cap images of <strong>Chernobyl</strong> and the 4 pine<br />
forest classes with reference (or control) sites for each class as given in Section 3. The second<br />
subsection discusses the progression of forest clearing during postaccident cleanup operations.<br />
The third subsection describes the calculation of time-to-response on a pixel-by-pixel basis for pine<br />
forest showing radiation response and discusses the correlation with radiation dose to foliage. The<br />
fourth subsection presents two alternative methods for generating dose contours from the satellite<br />
data. The final subsection discusses the resulting dose maps and summarizes results from this<br />
section.<br />
7.1 SPECTRAL DEVIATION FROM CLASS.<br />
Section 3 describes the procedures used to map four preaccident classes of pine forest based on<br />
pixel-by-pixel classification of spectral reflectivity. Each class of pine forest is represented by a<br />
reference site containing on the order of 100 contiguous pixels. Each pixel encompasses a 25 m by<br />
25 m square of forest and is represented by a vector of spectral intensity values transformed to the<br />
Tasseled Cap coordinate system. The vector components measure the average spectral features of<br />
all trees and other surfaces within the pixel that are visible from above. In addition, each pixel<br />
vector has noise components due mainly to atmospheric scattering of light into the line of sight of<br />
the pixel.<br />
In this subsection, we establish the bagis for detecting the average radiation response of the<br />
pine trees in a pixel by quantifying the deviation of each pixel vector from normal, defined as the<br />
mean vector of the reference (or control) site for the class to which the pixel belongs. The<br />
deviation for each pixel is expressed as a Mahalanobis vector. The deviations for the pine forest<br />
within a few kilometers of the <strong>Chernobyl</strong> nuclear reactor station are presented in a color format<br />
designed to show deviations in Tasseled Cap brightness, greenness, and wetness for each date<br />
analyzed.<br />
103
7.1.1 Mahalanobis Distance and the Mahalanobis Spectral Deviation Vector.<br />
Figure 7-1 illustrates a cluster of pixels from a class reference site using two axes of the<br />
Tasseled Cap coordinate system. Each pixel is represented by a point plotted according to the<br />
components of its spectral intensity vector x. For calculational purposes, we assume that each<br />
cluster of pixels may be reasonably approximated by a multivariate normal distribution. The<br />
resulting hyperelliptical distribution in general has its principle coordinate axes tilted with respect to<br />
the Tasseled Cap axes and has unequal variance along the principle coordinate axes.<br />
The spectral deviation of a pixel x from its class is expressed by the deviation vector d = x - pt,<br />
where p. is the mean vector of the class reference site. In order to have a consist meaning for the<br />
spectral deviation from class to class, the deviation vector must be expressed in standard units.<br />
The method that we use is explained in Appendix B. For each class of pine forest, a normalizing<br />
matrix N is used to convert the deviation vector d to a Mahalanobis vector m by the transformation<br />
m = Nd. (7.1)<br />
The matrix N is calculated from the covariance matrix of the class reference site as described in<br />
Appendix B.<br />
The normalization of each pixel deviation according to Equation 7. 1, transforms the set of<br />
deviations for the reference site to a distribution with unit variance in all directions. By definition,<br />
the deviations have a mean of zero. The magnitude of the Mahalanobis vector, called the<br />
Mahalanobis distance (Duda and Hart, 1973), measures the distance of a pixel from its class mean<br />
in standard units. It is the multivariate equivalent of the normal deviate z customarily used with a<br />
univariate standard normal distribution.<br />
The Mahalanobis distance is used to determine the significance of the deviation of a pixel from<br />
its class mean. To define significant deviation, a threshold is set such that any pixel with<br />
Mahalanobis distance greater than the threshold is judged to have moved significantly away from it<br />
class through the action of some factor not affecting the class in general. Since the Mahalanobis<br />
distance is expressed in standard units, we can use the same threshold for deviations with respect<br />
to all classes.<br />
7.1.2 Scaled Mahalanobis Distances and Vectors.<br />
The image processing software used for the present analysis stores only 8 bit intensity values,<br />
that is, integers ranging from 0 to 255. Calculations are done with floating point arithmetic, but 8<br />
bit numbers are used to store results in image format and for color display of images. For<br />
104<br />
0i
V)<br />
Q) CSt Reference Deviation<br />
Vector d<br />
Site<br />
Cluster<br />
Q-<br />
-o.... - .. .-<br />
S~PrincipOe<br />
0 Coordinate<br />
Axes of<br />
Cluster<br />
Tasseled Cap Greenness<br />
Figure 7-1.<br />
Two-dimensional illustration of the cluster of points formed by the pixel<br />
intensity vectors of a forest reference site. See Appendix B for the<br />
normalization procedure used to express deviations of pixels from the cluster<br />
center in standard units.<br />
105
convenience, a scaled Mahalanobis vector v and a scaled Mahalanobis distance s are defined<br />
according to<br />
v = 4m+128i (7.2)<br />
and<br />
s = 4 Iml, (7.3)<br />
where i is a vector with all components are equal to one. The dimensionality of the scaled<br />
Mahalanobis vector v is equal to the number of spectral bands being utilized in the analysis and can<br />
range from one up to the original number of bands in the multispectral image. The scale factor of 4<br />
in both equations scales the values such that in one dimension, a standard deviation becomes four<br />
units. In other words, the scaled Mahalanobis distance s is expressed in integer multiples of 0.25<br />
standard deviations. This resolution is sufficient since deviations of one or two standard<br />
deviations, expressed here as 4 or 8 units, are below the level of significance. A full scale<br />
deviation of 255 corresponds to 255/4 = 63.75 standard deviations, a dynamic range that is quite<br />
sufficient for present purposes.<br />
The scaled Mahalanobis vector v has an additive term of 128 on each of its components to shift<br />
the mean value from zero to 128, the center of the 8 bit scale. In this way, the cluster of scaled<br />
deviations for a reference site forms a spherical distribution at the center of the hypercube of<br />
possible deviation vectors with 8 bit components. The cluster has a standard deviation of 4 along<br />
each axis. Again, the only purpose of the scaling defined by Equations 7.2 and 7.3 is to<br />
accommodate the 8 bit data scale.<br />
The analysis of vector deviations in this report uses only three components from the Tasseled<br />
Cap transformed images, namely, brightness, greenness, and wetness. Figure 7-2 illustrates a<br />
scaled cluster for a class reference site using these three dimensions. The value of the scaled<br />
Mahalanobis distance s from Equation 7.3 corresponds to the radial displacement of a pixel from<br />
the center of the cluster in Figure 7-2.<br />
It is used to judge the significance of a deviation<br />
independently of the combination of feature changes causing the deviation. A threshold is chosen<br />
such that any deviation greater than the threshold is flagged as significant. Following the usual<br />
procedures of statistical analysis, the threshold must be set high enough that the random occurrence<br />
of deviations above threshold on the tail of the multivariate distribution for normal forest as<br />
illustrated in Figure 7-2 does not interfere with the detection of the spatial pattern of deviations<br />
caused by radiation exposure.<br />
106
0 50<br />
L0n<br />
1001<br />
o v fn<br />
o<br />
o• Ln,<br />
Figure 7-2.<br />
The cluster of scaled Mahalanobis vectors for a pine forest class reference site using<br />
the first three features of the Tasseled Cap spectral transformation. (Cluster<br />
enlarged for clarity.)<br />
107
The scaled vector deviation v in Equation 7.2 specifies the direction of the deviation in the<br />
three-dimensional space, that is, it tells how much each of the Tasseled Cap features deviates from<br />
normal. It may be used to distinguish deviations with specific spectral signatures. It particular, the<br />
Mahalanobis vector is used in this report to detect pine forest that was cleared as part of the<br />
decontamination effort after the <strong>Chernobyl</strong> accident.<br />
7.1.3. Spectral Deviations for the <strong>Chernobyl</strong> Images.<br />
We have calculated both vector and scalar spectral deviations according to Equations 7.1-3 and<br />
Appendix B for all pixels in the 38.4 km square analysis area presented in Section 3 belonging to<br />
the four pine forest classes defined in Figure 3-24. Figures 7-3 and 7-4 provide a color graphic<br />
presentation of the vector deviations for a 4.4 km by 4.8 km area near the reactor station. The<br />
upper left panel in Figure 7-3 shows the four pine forest classes color-coded as in Figure 3-23 and<br />
superimposed on a gray-scale background image of the area for spatial orientation. The other five<br />
panels of the same size show the vector deviations for the pine forest pixels on the first five inmage<br />
dates; for clarity, no background image is include. The five panels are numbered according to the<br />
image/date numbering scheme of Table 3-3. Figure 7-4 shows the vector deviations in the same<br />
area for the last six image dates.<br />
The color graphic presentation of vector deviations in Figures 7-3 and 7-4 is generated by<br />
displaying the components of the scaled Mahalanobis vector (Equation 6.2) for Tasseled Cap<br />
brightness, greenness, and wetness in the red, green, and blue channels, respectively, of the color<br />
images. The process can be visualized in the color space of the display images with the aid of<br />
Figure 7-2 by replacing each Tasseled Cap label on the axes with the corresponding red, green, or<br />
blue display intensity. The scale of gray levels (equal intensities of red, green, and blue) from<br />
black to white stretch along the diagonal of the color space from the origin (0,0,0) to<br />
(255,255,255).<br />
The cluster corresponding to a class reference site is centered at midscale (128) on each of the<br />
color channels. The resulting display is gray at the middle of the intensity scale, halfway between<br />
black and white. A patch of forest with negligible deviations from the class mean would appear as<br />
nearly uniform gray in Figure 7-3. Since the pixels of normal forest have random deviations from<br />
the class mean as illustrated :n Figure 7-2, normal forest will appear to be speckled with pastel<br />
colors lying near the midlevel gray point.<br />
The three reference bars along the right edge of Figures 7-3 and 7-4 show the appearance of<br />
normal distributions like that of Figure 7-2 with standard deviations of 4, 6, and 8 units along each<br />
axis as indicated by the number in each bar. Note that only the intensity of color fluctuations and<br />
not their spatial scale is different in the three reference bars. By definition, reference sites for the<br />
108
Figure 7-3. Color graphic presentation of the scaled Mahalanobis deviation vectors for pine forest<br />
on Dates I to 5 (194 by 177 pixel area). Upper left panel shows forest classes (see Figure 3-23).<br />
Three reference bars along right edge illustrate appearance of normal forest<br />
109/110
Figure 7-4. Cclor graphic presentation of the scaled Mahalanobis deviation vectors for pine forest<br />
on Dates 6 to 11 (194 by 177 pixel area). Three reference bars along right edge llustrate<br />
appearance of normal forest.<br />
III
pine forest classes will appear like the upper bar with a stmtidlrd d-iiaticn of 4 units along each<br />
color axis. The reference bars with standard deviations cf 6 , 8 are included since<br />
geographically separated patches of norrnal forest may havc aciditicnal random variance not present<br />
in the corresponding reference site.<br />
Examination of the Date 1 spectral deviattons as represented by the speckled gray patterns of<br />
the color graphic display in Figure 7-3 shows co signifikant diffeerence from the reference bars as<br />
expected since Date 1 was before the reactor accidtnt and is used to define the forest classes. Date<br />
2, the preaccident winter image, is also reasonably close to the reference bars as it should be.<br />
The strongest deviations in Date 3 of Figure 7-3 are the magenta patches caused by clouds<br />
which have high brightness and wetness with low greenness. The resulting mixture of red and<br />
blue with negligible green gives the magenta color.<br />
Figure 7-4 clearly shows the progression of increasingly significant disturbance of the pine<br />
forest as a result of the reactor accident. By Date 9, strong red patches show forest that has been<br />
cleared for operational purposes and for radioactive decontamination These areas are expanded by<br />
Date 10. Some of the ground cover on these cleared areas has changed by Date 11 as evidenced by<br />
the shift from red to magenta. The red line through the large forest patch in the lower third of this<br />
image area is probably the disturbance caused by the construction of a ground water barrier as<br />
discussed in Appendix C. It first appears in the images on Date 7 (October 1986) with an<br />
additional segment on Date 9 (May 1987).<br />
Not much forest clearing had been done by Date 7. There were no clouds on this date. The<br />
spectral deviations of the pine forest show a strong radiation response as the brown and orange<br />
swath across the middle of the left half of the image area. This swath corresponds to the westward<br />
trace of initial fallout deposition from the reactor explosion. Additionally, tbc. irregular patch of<br />
forest at the top of the image area on Date 7 shows a radiation response with a somewhat different<br />
spectral signature as evidenced by the magenta hue in Figure 7-4. This response of the irregular<br />
patch has diminished on the winter image of Date 8 but returns the following year on Dates 9 and<br />
10.<br />
According to the data and discussion in Section 6.4, this fluctuation in detcctability of the<br />
radiation response of the irregular patch suggests a short term total dose to tome canopy in the 10 to<br />
20 Gy range. The magenta hue on Date 7 indicates a reduction in greenness relative to normal<br />
pines, corresponding to a lack of new growth during the sumnmer after the accident. As the new<br />
growth on unexposed pines matures, the winter image of Date 8 shows almost no difference<br />
between the exposed patch and normal forest. The shift to an orange hue by Date 10 indicates that<br />
many pines have died by that time.<br />
112<br />
i m Bm *. =now
Although it may not be apparent because of the resolution and fidelity of color reproduction of<br />
this report, the forest along the westward trace of fallout shows a widening response on Dates 4<br />
through 6. This widening is demonstrated by the change detection analysis in Volume 2 of this<br />
report (McClellan et al., 1992). It is also apparent when the deviation images of Figures 7-3 and<br />
7-4 are displayed on a high resolution color monitor of an image processing system. Because of<br />
limitations of color reproduction, further analysis of radiation response in this report is based on<br />
numerical analysis of the scaled Mahalanobis deviation vectors and distances.<br />
7.2 DEVIATIONS CAUSED BY FOREST CLEARING.<br />
In the detection of radiation-induced spectral deviations, clearing of forest must not be<br />
interpreted as severe radiation response. Figure 7-5 shows a map of pine forest that was present<br />
postaccident through May 1986 but appears to have been cleared (bulldozed) by the date of the<br />
image number indicated in the figure. The figure includes. I km grid lines originating at the reactor<br />
(Unit 4). Clearing activity west of the reactor station seems to be mainly for decontamination. The<br />
earliest clearing is near the elbow in the main road about 1.5 km west and 0.5 km south of the<br />
reactor site where the road crosses the main trace of fallout deposition. Appendix C includes an<br />
eyewitness description by a man who walked along this road on the morning of the accident. He<br />
encountered a band of graphite flakes crossing the road and reported that the forest later "came<br />
under the ax."<br />
The large area in green in Figure 7-5 located 1.5 krn east and 2 km south of the reactor site was<br />
cleared between December 1986 and May 1987. This area shows no apparent radiation response<br />
before being cleared and so is presumed to have been cleared for operational reasons. The spectral<br />
signature of this area on Date 9 is used to define cleared forest on all dates. For this purpose, the<br />
scaled Mahalanobis vector is used as described in Section 7.1.2.<br />
A representative scaled<br />
Mahalanobis vector for cleared forest is calculated on Date 9 from a sample of pixels from the large<br />
cleared area. The vector is then normalized to provide a unit vector pointing in the direction of<br />
spectral deviation of cleared forest relative to normal forest. A forest pixel on any date is<br />
designated as cleared if the dot product of its scaled Mahalanobis deviation vector with this unit<br />
vector exceeds a threshold. The threshold is set high enough to eliminate spurious indications of<br />
clearing for isolated pixels and to avoid classifying radiation-damaged forest as being cleared.<br />
Table 7-1 lists the areas of newly cleared pine forest on the various dates in Figure 7-5. These<br />
areas are lower limits to the amount of vegetation actually cleared since Figure 7-5 includes only<br />
pixels belonging to one of the four classes of pine forest defined in Section 3. The listed areas<br />
include the pixels along the path cleared through the forest 3 km south of the reactor for the<br />
construction of a ground water barrier as described in Appendix C.<br />
113
Figure 7-5. Forest cleared (bulldozed) after the <strong>Chernobyl</strong> acident is color coded by the date<br />
number of the first observed clearing. Grid lines are spaced by I kn,, with an intersection at the<br />
Unit 4 reactor.<br />
114
Table 7-1. Area of pine forest near the <strong>Chernobyl</strong> nuclear reactor station cleared by the listed image<br />
date. Clearing of other vegetation not included.<br />
Image Area Cumulative Area<br />
Number Date Number of Pixels Cleared (ha) (ha)<br />
7 15 Oct 86 5 0.3 0.3<br />
8 2 Dec 87 0 0.0 0.3<br />
9 11 May 87 1115 69.7 70.0<br />
10 7 Sep 87 493 30.8 100.8<br />
11 28 May 88 410 25.6 126.4<br />
7.3 TIME-TO-RESPONSE FROM IMAGERY.<br />
The time-to-response for the foliage comprising a single pixel is derived from the onset of<br />
significant, persistent deviation of the spectral signature of the pixel from normal the spectral<br />
signature of the class to which the pixel belonged before the accident. The time-to-response for the<br />
foliage in the pixel is the time difference between this response and the time of the accident, that is,<br />
time-to-response is measured relative to the start of radiation exposure.<br />
Significant deviation from normal on a given date is judged by the scaled Mahalanobis distance<br />
of a pixel relative to its class mean as described in Section 7.1.2. A threshold distance is set for<br />
each date so that any pine forest pixel with scaled Mahalanobis distance exceeding threshold is<br />
flagged as having a significant deviation on that date. The procedure is equivalent to using a<br />
multivariate z score to judge the significance of an observed deviation. The Mahalanobis distance<br />
is calculated in 3 dimensions using the intensity values for Tasseled Cap brightness, greenness,<br />
and wetness. Thresholds of 20 to 24 are low enough to give good detection of the radiationinduced<br />
spatial patterns of change described in Volume 2 of this report and high enough to<br />
minimize the random occurrence of isolated pixels exceeding threshold at large distances from the<br />
reactor site.<br />
In addition, a neighborhood-dependent algorithm for threshold comparison is used to improve<br />
the signal-to-noise ratio for detecting radiation response. In this algorithm, the Mahalanobis<br />
distance for each pixel is replaced by the magnitude of the average Mahalanobis vector of the pixel<br />
and any of its eight neighboring pixels that are also pine forest. The vector average reduces the<br />
influence of slight registration errors from image to image as well as noise from other sources. To<br />
take advantage of this noise reduction, the threshold for each pixel is scaled by the inverse of the<br />
square root of the number of pixels being averaged. That is, when a pine pixel at the center of a<br />
115
3 by 3 square patch of pixels has N - I neighbors in the patch that are also pine forest, the pixel is<br />
flagged as having a significant deviation if<br />
N<br />
Xvi<br />
T<br />
> N 1/2 (7.4)<br />
N<br />
where vi is the scaled Mahalanobis distance of the i th pine forest pixel in the 3 by 3 patch and T is<br />
the threshold value for a single pixel. A form of Equation 7.4 that is more efficient for calculations<br />
is<br />
N 122(75<br />
vi > NT 2 (7.5)<br />
i= I<br />
Figure 7-6 illustrates the algorithm for the determination of time-to-response given that<br />
significant deviations have been flagged in each image and lists the threshold T used for each<br />
image. The algorithm is based on finding the first significant deviation that persists for the same<br />
pixel on all subsequent image dates. The deviation must precede any indication that the forest in<br />
the pixel has been cleared. Any pixel that does not contain pine forest, or is cleared of forest before<br />
any indication of radiation response, yields no data regarding radiation exposure. Such pixels are<br />
assigned a code of zero and will become black background in response maps presented as images<br />
in this report. Any pixel that appears normal (does not exceed threshold) on Date 11, the last<br />
image of the series, has no persistent response by definition and is coded as normal forest with no<br />
radiation response. Such forest pixels yield a null response and are coded to appear gray in<br />
response or dose images. These areas show where fallout contamination was too low to produce<br />
lasting visible damage to the foliage of pine trees. A pixel that shows a first persistent deviation<br />
from normal on Date n is coded with the value n to indicate the date of first observed response. Of<br />
course, the first "observable" response will have actually occurred sometime between this date and<br />
the previous image date.<br />
The requirement of a persistent response over two years has the disadvantage of eliminating<br />
responses to low doses where the trees recover normal appearance within two years. On the other<br />
hand, the requirement has two major advantages. The first, and of most practical importance, is<br />
the elimination of spurious responses caused by clouds, jet contrails, and local haze patterns<br />
occurring on a single date. The second advantage is further reduction of noise signals that are<br />
uncorrelated from date to date. These advantages derive from the fact that several of the images in<br />
116
Start Processing<br />
Thresholds for significant deviation<br />
Image: 1 2 3 45 67 8 91011<br />
T: 40 55 40 40 60 40 40 40 40 40 45<br />
(Scaled Mahalanobis Distance)<br />
belong to a pine<br />
forest class?<br />
Yes<br />
NDoes the pixel show<br />
•"Yes<br />
a persistent<br />
deviation from<br />
normal? N<br />
No Data<br />
(Coded black)<br />
Did an indication of<br />
forest clearing occur<br />
persistent as early deviation as the<br />
Ye from normal? No<br />
Null<br />
Response,<br />
Forest normal<br />
(Coded gray)<br />
No Data<br />
(Coded black)<br />
Response Time<br />
Detected<br />
(Color-coded by<br />
date)<br />
Figure 7-6. Algorithm for determination of time-to-response for radiation-damaged foliage on a<br />
pixel-by-pixel basis. Parenthetical remarks refer to presentations in response time<br />
figures.<br />
117
the series are cloud free and none of the images have extensive cloud cover, so the probability of a<br />
persistent response at a given location due to a combination of clouds and random noise is nil.<br />
Since the last date (Date 11) has no subsequent image for establishing persistence, the threshold<br />
used for Date 11 is set higher than that of the other images to reduce spurious responses on that<br />
date.<br />
Figure 7-7 presents results for the date of first observed radiation response according to the<br />
algorithm of Figure 7-6. The date number of response for each pixel is color-coded according to<br />
the legend in the image. Table 6-3 lists the date numbers and corresponding estimated foliage<br />
doses according to the analysis of published data presented in Section 6. No persistent responses<br />
exceed threshold on the preaccident Dates 1 and 2 or on postaccident Date 3. Only two pixels in<br />
the central blue area of Figure 7-7 have responses on Date 4. The main early response in<br />
Figure 7-7 is for Dates 5 and 6, four to five weeks after the accident date. Since the responses<br />
could have started immediately after Date 4 at 12 days, the dose for the blue area is estimated to be<br />
between 54 and 80 Gy according to Table 6-3. Surrounding the blue area and extending further to<br />
the west is a dark green area showing first observed response on Date 7 (October 1986). The<br />
estimated dose for this area is 30 to 54 Gy. The yellow-green and yellow areas have first observed<br />
response between Dates 8 and 10 with estimated doses in the 20 to 30 Gy range.<br />
Note that the irregular patch of forest in the upper middle of Figure 7-7 shows only a few<br />
pixels with persistent responses, beginning on Dates 9 and 10 with estimated dose in the 20 to 28<br />
Gy range. The irregular patch consists mostly of gray pixels, indicating a dose less than about 18<br />
Gy. As pointed out in Section 7.1.3, this area shows a transient response maximizing on Date 7<br />
(October 1986) and disappearing over the winter. As discussed in conjunction with Figure 6-4,<br />
this transient response is likely to be associated with doses in the range of 7 to 15 Gy. The<br />
response of this irregular patch shows that the multispectral detection technique for pine tree<br />
radiation response could be extended to doses below 20 Gy if the analysis accounted for transient<br />
as well as persistent responses.<br />
A small, vertically oriented, rectangular group of about 20 pixels just to the right of the<br />
irregular patch shows a solid response on Date 7 for an estimated dose of 30 to 54 Gy. Figure 7-5<br />
shows that this same group of pixels was cleared between Dates 9 and 10. The group is 1.2 km<br />
west and 1.3 km north of the reactor site. Figure 7-5 shows that the irregular patch of forest<br />
discussed above is located 2 km west and 1.5 km north of the reactor site and had not been cleared<br />
as of Date 11 (May 1988). The patch is at the eastern edge of Pripyat. If these trees are still<br />
standing, they would provide a good retrospective sample on which to perform histological studies<br />
to correlate with the responses and estimated doses derived from satellite images.<br />
The path cut through the largest patch of gray, normal forest in the lower part of Figure 7-7 is<br />
associated with a ground water barrier several kilometers long and 30 - 35 m deep constructed to<br />
118
Figure 7-7. Date of onset of persistent deviation from normal of pine forest pixels according to<br />
algorithm of Figure 7-6; 6.4 km square area. See Table 6-3 for corresponding dose estimates. No<br />
responses occur for the preaccident Dates I and 2 or for Date 3 three days postaccident.<br />
119
impede the migration of radionuclides in the !round. The presence of the ground wdter barrier<br />
apparently affects the pine trees adjacent to the path, especially on the southwest side. The trees<br />
show significant deviations from normal extend:-rg further from the barrier as time progresses<br />
between 1 and 2 years after the accident. These indications of stress are probably not due to<br />
radiation damage to the foliage. They may indicate a ground water effect on the health of the root<br />
,s stems of the trees not involving any radiation damage.<br />
7.4 FOLIAGE DOSE MAPS.<br />
Two procedures have been used to construct dose contours from the data on time to first<br />
response for pine trees shown in Figure 7-7. The first method uses shape of gamma dose rate<br />
contours measured aerographically as guidance for the hand-drawn contours wherever satel iite<br />
image data is sparse or nonexistent. The second method uses a numerical relaxation techniqte, to<br />
generate smoothed contours from the satellite image data alone in an area with sufficient data for<br />
analysis.<br />
7.4.1 Dose Map with Hand-Drawn Contours.<br />
Figure 7-8 shows the time-to-response data from the satellite image analysis with three handdrawn<br />
radiation dose contours. Table 7-2 lists the total area enclosed by each contour and the dose<br />
range of the area inside each contour.<br />
Table 7-2. Description of contours drawn in Figure 7-8.<br />
Time of observed Dose within Cumulative<br />
response contour enclosed area<br />
Contour (image number) (Gy) (ha) (ki2)<br />
Inner (blue) 5 and 6 54 to 80 a 82. 0.8<br />
.liddle (green) 7 30 to 54 287. 2.9<br />
:ter (yellow) 8, 9 and 10 20 to 30 1450. 14.5<br />
'r limit set by the sparsity of response on Date 4.<br />
tours in Figure 7-8 are hand-drawn starting in the westward trace of highest fallout<br />
'here there is sufficient data from pine forest response to draw the contours as<br />
;tween areas with different dates of first observed response. The intent is to draw<br />
.ooth contours along the average boundary between different dates of response.<br />
-)urse, scattered pixels with response dates that do not match the contours. These<br />
120
Figure 7-8. Time to first observed response with 1 km grid and three hand-drawn dose contours.<br />
See Table 7-2 for dose ranges.<br />
121
variations presumably are caused by nonuniformity in fallout deposition and by noise in the<br />
detection process. The contour segments drawn where there is image data are extended around the<br />
reactor where there is no image data following the shape of published gamma dose rate contours<br />
measured a year after the reactor explosion (Asmolov et al., 1987). These Soviet-supplied<br />
contours are described in Volume 2 of this report (McClellan, 1992). Note the contours in Volume<br />
2 and the outer contour in Figure 7-8 have been extended across the reactor cooling pond to<br />
approximate the situation in the absence of the body of water.<br />
Since the contours in Figure 7-8 are derived from observations of pine tree response, they<br />
represent an estimate of the dose at any location that would be received by a patch of pine forest at<br />
that location. The relationship of this dose to the dose that would be accumulated by some other<br />
detector can be estimated from the results of the radiation transport calculations of Appendix A. In<br />
particular, Section 8 discusses the relation between the pine tree dose and the gamma dose 1 m<br />
above the ground under the canopy.<br />
The extrapolation of the contours around the reactor in Figure 7-8, based on gamma dose rate<br />
contours, is uncertain because of the difference between in situ pine forest response and that of the<br />
gamma detectors one year after the accident. As discussed in Section 5, the pine forest canopy<br />
responds mainly to the dose accumulated during the first few weeks after the reactor explosion<br />
with a beta contribution at least as large as the gamma contribution. On the other hand, the curves<br />
of Asmolov et al. (1987) show the gamma activity a year after the explosion affected by weathering<br />
effects and cleanup activities. Because of this difference, it is of interest to consider a method of<br />
generating contours from the pine tree response alone.<br />
7.4.2 Iterative Smoothing of the Dose Map.<br />
Figure 7-9 illustrates an iterative numerical relaxation method for smoothing of the dose map.<br />
About 3 km 2 is cut from the area of highest radiation response in Figure 7-8. The upper left panel<br />
in Figure 7-9 shows the date of first observed response for this area following the color code of<br />
Figure 7-8. Gray pixels indicating forest that is normal on Date 11 are assigned date value 12.<br />
The lower right panel shows the result of the relaxation method with date numbers of first<br />
observed response used as the numerical value assigned to each pixel.<br />
On each iteration, the response date number for each pixel not containing pine forest is replaced<br />
by the average date number (taken as a continuous variable) of its four nearest neighbor pixels.<br />
Values for pine forest pixels in the initial data set are held fixed at their initial values to provide<br />
boundary conditions for the relaxation solution. Relaxation of values at the border of the cut out<br />
area use a neighborhood average excluding pixels outside the border. This choice lets the value at<br />
122
Figure 7-9. Series of intermediate stages leading to a smoothed map of date of first observed<br />
response by numerical relaxation as described in the text.<br />
123
the border float according to the influence of the nearest pine forest pixels in the initial data set<br />
without requiring a definite value at the border. Pixels with no value (displayed as black) in the<br />
initial data set are initialized with date value zero. Iteration of the map until the value in every pixel<br />
changes by less than 0.0001 from the previous iteration gives the lower right panel in Figure 7-9.<br />
Stopping the iteration when all changes are below 0.5 and other intermediate values indicated in the<br />
figure gives the full series shown in Figure 7-9. For practical purposes, stopping the iteration<br />
when changes are less than 0.001 is good enough.<br />
This numerical relaxation technique converges to a solution of Laplace's equation,<br />
V 2 n = 0 (7.6)<br />
where n is the date of first observed response treated as a continuous variable. In electrostatics,<br />
Laplace's equation describes the spatial variation of the electric potential in a region of uniform<br />
electrical conductivity. The pixels in the initial data set used as fixed boundary conditions<br />
correspond to areas held at a fixed voltage by an external source of electrical potential. There is no<br />
physical reason why the radioactivity level in a wind-driven fallout field should satisfy Laplace's<br />
equation. The numerical relaxation technique is simply a convenient tool for smoothing the dose<br />
map.<br />
Figure 7-9 shows substantial small scale variation caused by isolated pixels in the initial data<br />
set. Furthermore, Figure 7-9 gives a map of date of first response rather than dose. Figure 7-10<br />
addresses both of these issues. Based on the success of the relaxation solution in Figure 7-9,<br />
Figure 7-10 uses a larger area of data. To reduce small scale variations, isolated pixels in the initial<br />
data set are ignored as boundary conditions and allowed to vary in the relaxation calculation. Next,<br />
before relaxation, the date numbers in the initial data set are replaced by radiation doses according<br />
to Table 6-3. Iteration proceeds until fractional changes in all dose values drop below 0.001, then<br />
the pixels are assigned to dose bands and colored according to Table 7-3.<br />
Figure 7-11 displays the smoothed contours from Figure 7-10 with a gray-scale image of the<br />
area. Both Figures 7-10 and 7-11 include one kilometer grid lines originating from the site of the<br />
Unit 4 reactor. The site of the reactor is chosen as the brightest pixel from the reactor fire visible<br />
on 26 April 1986 (Date 3), three days after the explosion. The Zone 36 UTM coordinates of the<br />
upper left hand comer of this pixel are X = 298,275 m and Y = 5,697,175 m as determined by the<br />
geocoding of our Landsat image data (see Volume 2).<br />
124
Figure 7-10. Smoothed dose contours from pine foliage response using numerical relaxation of<br />
dose values from individual pine forest pixels inside the white border; I km grid lines.<br />
See Table 7-3 for estimated dose bands.<br />
125/126
Figure 7-11. Smoothed dose contours from Figure 7-10 displayed with a gray-scale background<br />
of the western end of the <strong>Chernobyl</strong> nuclear power station; I km grid lines. See Table 7-3 for<br />
estimated dose bands.<br />
127
Table 7-3. Dose bands and affected areas for the smoothed radiation contour maps of Figures 7-10<br />
and 7-11. Areas are within the relaxation area (white boundary of Figure 7-10) only.<br />
Dose band Band area Cumulative area<br />
Color (Gy) (ha) (ha)<br />
Red >94 0.1 0.1<br />
Yellow 54 to 94 9.1 9.2<br />
Green 35 to 54 16.8 26.0<br />
Cyan 23 to 35 86.1 112.1<br />
Blue 13 to 23 353.8 465.9<br />
The smoothed contours produced by the relaxation method appear too irregular and sometimes<br />
discontinuous in the central area. For example, the yellow areas in Figure 7-11 with a dose range<br />
from 54 to 94 Gy originate mainly from the observed responses on Dates 5 and 6. The appearance<br />
of this data in Figure 7-8 (within the inner contour), suggests that all of the area between the<br />
yellow patches in Figure 7-11 probably received at least as much dose as the yellow patches<br />
themselves. However, because of the wide spacing of the three main areas of response on Dates 5<br />
and 6, the relaxation method allows the influence of more distant pixels to lower the interpolated<br />
dose between the yellow patches. Although hot spots do occur in fallout fields, the spacing of the<br />
yellow patches in Figures 7-10 and 7-11 would not be expected in this situation and is clearly an<br />
artifact of the spacing of the patches of pine forest in the initial data set.<br />
Except for the exaggerated irregularities induced by the patchiness of pine forest in the initial<br />
data set, the smoothed contours of Figure 7-11 seem reasonable. For example, the western outline<br />
of the outer most dose band (blue, 14 -26 Gy) in Figure 7-11 has the same general shape as the<br />
outer most, hand-drawn contour (yellow, 20 - 30 Gy) in Figure 7-8.<br />
The relaxation method has the advantage of providing smoothed radiation dose contours<br />
without reference to other data and free of human bias. It has the disadvantage that it does not<br />
account for the physical transport properties of wind blown fallout patterns and tends to generate<br />
dose contours having the same spottiness as the pine forest spatial distribution. With further<br />
study, it might be possible to modify the relaxation method to better approximate true fallout<br />
patterns.<br />
128
7.5 SUMMARY.<br />
Both methods presented in Section 7.4 for generating maps with dose contours or bands from<br />
the pixel-by-pixel pine tree response data have advantages and disadvantages. Both provide a<br />
qualitative description of the likely initial fallout deposition pattern, but both must be used with care<br />
if quantitative extrapolations of dose outside the existing areas of pine forest are required. The most<br />
reliable results are the actual pixel-by-pixel dose estimates obtained through interpretation of the<br />
date of first observed response (Figure 7-7 ur 7-8) with the dose and time-to-response relationship<br />
from Figure 6-5 or Table 6-3, keeping in mind that response observed on a certain date means that<br />
the actual first observable response may have occurred any time in the interval bracketed by the<br />
date of first observed response and the previous image date. The dose estimate for the pixel lies in<br />
the interval defined by the doses for the two dates from Table 6-3. Accordingly, the dose bands in<br />
Table 7-3 were obtained using the geometric mean of the dose for the first observed response and<br />
the dose for the preceding date.<br />
The following list summarizes findings from this section:<br />
1) Pixel spectral deviations for pine forest canopy relative to the average spectral signature of a<br />
pine forest class may be used on a pixel-by-pixel basis to detect foliage radiation response<br />
in multispectral imagery (Figures 7-3 and 7-4).<br />
2) Time to first observed response for persistent spectral deviations over an annual cycle may<br />
be used to map pine canopy doses (Figure 7-7) near and above the LD 50 , about 23 Gy for<br />
a three week exposure.<br />
3) Mapping of dose responses significantly below the LD 5 0 requires interpretation of<br />
temporal variation in detectability of spectral deviations caused by the annual growth cycle.<br />
It is likely that transient responses may be detected down to 10 or even 5 Gy.<br />
4) Dose estimates at distances of 1 to 2 km downwind of the reactor site and from 30 to 54 Gy<br />
at 2 to 3 km downwind (Figure 7-8) from the Landsat multispectral imagery for pine forest<br />
canopy along the main westward trace of initial fallout deposition from the <strong>Chernobyl</strong><br />
reactor explosion range from 54 Gy to 80 Gy.<br />
These doses estimates are based on the<br />
equivalent gamma dose that produces the same foliage spectral response as the actual mixed<br />
beta and gamma doses received by the foliage.<br />
129/130
SECTION 8<br />
DISCUSSION<br />
This section discusses the pine forest foliage doses derived from the satellite imagery in relation<br />
to data from the former Soviet Union regarding three types of measurements: 1) doses and<br />
responses of trees near the <strong>Chernobyl</strong> nuclear power plant, 2) aerographic and ground surveys of<br />
gamma dose rates following the accident, and 3) exposures of accident victims. A final subsection<br />
discusses uncertainties in the derivation of foliage doses from the satellite imagery.<br />
8.1 COMPARISON WITH MEASUREMENTS OF FOREST DAMAGE.<br />
The only quantitative exposure and response data we have for trees within a few kilometers of<br />
the <strong>Chernobyl</strong> nuclear power station has been obtained by Gamache (1993) through visits with<br />
Ukrainian scientists (Sobdovych et al. 1992). Table 8-1 lists this data, which refers to forest with<br />
edge about 1 km from the Unit 4 reactor site. This distance corresponds to the boundary of the<br />
nuclear power station on the west side. We are fortunate to have this information although some<br />
uncertainty remains regarding the data. For example, it is tempting to assume that the subject<br />
forest is west of the reactor site along the main trace of fallout deposition, but there are also forest<br />
patches northwest, southwest, and south of the reactor site that are not much more than 1 km<br />
away. The forest reportedly consisted of pine trees but Table 8-1 implies an LD 50 of 49 Gy, above<br />
the upper bound in Table 6-1 for pine trees suffering a few week exposure. Figure 6-2 indicates<br />
that an effective exposure time of about four months would be required to increase the LD 5 0 for<br />
pines to 49 Gy; however, our image analysis shows significant response along the main trace after<br />
only one month.<br />
Table 8-1. Results of forest damage by radiation for an area with edge about 1 km from the site of<br />
the Unit 4 reactor explosion (Sobdovych et al., 1992, courtesy of Gamache, 1993).<br />
Distance from<br />
edge offorest<br />
Calculated<br />
absorbed dose<br />
Tree crown<br />
damage<br />
Recovery<br />
of trees<br />
(M) (Gy) (%) Degree of harm (%)<br />
0 100 100 Completely dry 0<br />
(Edge of forest<br />
wood<br />
nearest reactor)<br />
35 65 50 Very strong damage 10- 15<br />
90 49 20-30 Medium 50<br />
350 5
In addition, Table 8-1 lacks an indication of the time at which the listed endpoints were<br />
observed. Presumably, the percent of crown damage must have been observed well before the<br />
percent recovery of trees since 50% lethality of trees would be inconsistent with only 20 - 30 %<br />
crown damage if both endpoints were observed at the same time. Also, the nature of the calculated<br />
absorbed dose is not known to us. In particular, does it include dose from beta radiation and to<br />
what tissue and depth does it correspond?<br />
The precipitous drop reported for the dose 350 m from the edge of the forest also needs<br />
explanation. Given that the plume from the initial explosion and fire rose more than a kilometer<br />
(Appendix C) and that the forest location is only 1 to 1.5 km from the reactor site, it would be<br />
surprising to have such a sharp gradient in dose within 250 to 350 meters along the main trace of<br />
initial deposition. Such a drop would be more likely, however, in moving off the trace laterally.<br />
Another possibility is that the reported forest is not on the main trace but was contaminated later in<br />
the 10 day release sequence by a plume that remained near the ground and was rapidly absorbed as<br />
it moved through the forest canopy. In the absence of accurate position information, the data in<br />
Table 8-1 cannot be directly compared with the satellite data.<br />
In spite of the unknown factors, it is encouraging that the doses listed in Table 8-1 are of the<br />
same order of magnitude as the doses from the analysis of the satellite image data listed in Table 7-<br />
2 and 7-3. The attempt at comparison strongly emphasizes the need for well documented ground<br />
measurements at specific times and with locational accuracy and resolution comparable to the 25 m<br />
spatial resolution of the imagery.<br />
8.2 COMPARISON WITH AEROGRAPHIC SURVEYS OF DOSE RATE.<br />
Following the <strong>Chernobyl</strong> accident, frequent aerographic and surface measurements were made<br />
by the Soviets to determine and monitor the gamma field dose rates surrounding the nuclear reactor<br />
station. These measurements provide spatial contours and decay rates of the gamma dose rate<br />
within the area of our satellite observations. Utilizing the Soviet data, this subsection calculates<br />
integrated gamma ray doses for the three week period following the accident and compares these to<br />
the pine forest doses extracted from the satellite image data. The ratio of the pine foliage dose to<br />
the gamma dose one meter off the ground is compared to the same ratio derived from the<br />
calculations of Appendix A.<br />
8.2.1 1-Meter Gamma Dose Rates for the Close-in Area.<br />
Figure 8-1 shows plots of gamma dose rate versus contaminated area based on measurements<br />
referenced to two dates, May 29, 1986 and May 1, 1987. The area on the abscissa is that enclosed<br />
by a given isodose-rate contour. Accordingly, dose rates within a given area are larger than that at<br />
the contour bounding it. The Soviet dose rate data are given only for close-in areas greater than<br />
132
about 2 km 2 . Therefore, estimates down to 0.8 km 2 are based on extrapolation indicated by the<br />
dashed part of the curves in Figure 8-1.<br />
104<br />
103<br />
E6<br />
May 29,1986<br />
D102%<br />
(D<br />
0<br />
May 1, 1987<br />
10<br />
1 10 102 103<br />
Contaminated fallout area (kin 2 )<br />
Figure 8-1.<br />
Dose rate measurements for close-in fallout contaminated areas for two different<br />
dates: May 29, 1986 (Izrael, Petrov and Severov, 1987) and May 1, 1987<br />
(Asmolov et al., 1987).<br />
133
Table 8-2 provides, in the second and fourth columns, values for selected contour dose rates<br />
read from Figure 8-1 for the two reference dates.<br />
Table 8-2. Estimation of initial dose rates for contours enclosing specified areas.<br />
Measurements of: May 29, 19 8 6 a May 1, 19 8 7 b<br />
Contaminated area Contour Dose ratec Contour Dose ratec<br />
within contour dose rate on 4126/86 dose rate on 4126/86<br />
(kin 2 ) (mR/h) (Rid) (mR/h) (Rid)<br />
14.5 450 68<br />
4.0 1150 173 38 158<br />
2.9 1450 218<br />
1.0 2500 376 100 416<br />
0.8 3000 450<br />
aMeasurements from Izrael, Petrov and Severov (1987).<br />
bMeaurements from Asmolov et al. (1987).<br />
CExtrapolated back to date of accident according to Figure 8-2.<br />
In order to obtain doses accumulated over the first three weeks after the accident, it is necessary<br />
know the time dependence of the dose rate. Figure 8-2 gives a plot of Soviet data for the change<br />
with time of the gamma dose rate in the close-in zone due to the decay and weathering of<br />
radioactive material. We have approximated the data with a smooth curve by fitting it with an<br />
empirical relationship,<br />
F (t) = exp{- [in (1 + t)] 2 / a), (8.1)<br />
where, a = 6.7784 and postaccident time t is measured in days. Figure 8-2 plots the gamma dose<br />
rate in relative units normalized to the day of the accident (t = 0 on April 26, 1986); note that the<br />
abscissa is (1 + t). The dose rate values on the day of the accident are estimated with Equation 8.1<br />
and shown in columns 3 and 5 of Table 8-2. The cumulative gamma dose for a postaccident<br />
exposure time t is determined for any initial dose rate R 0 by integrating Equation 8.1 over time,<br />
i.e.,<br />
D (t) = Ro F (t')dt'. (8.2)<br />
134
2<br />
0.1<br />
a:<br />
0.01<br />
1 10 100 500<br />
Time (days)<br />
Figure 8-2. Change in gamma dose rate from radioactive materials in the close-in zone as a<br />
function of time based on aerographic surveys (data from Asmolov et al., 1987).<br />
Figure 8-3 compares the 1-m gamma dose accumulated at the 1 km 2 and 4 km 2 contours as a<br />
function of postaccident time. The a and b curves reflect the Soviet measurements reported on<br />
May 1, 1987 and May 29, 1986, respectively. The middle curve, labeled 1 - 4 km 2 , of Figure 8-3<br />
plots the geometric mean dose for the 1 and 4 km 2 curves. Both the a and b curve pairs give<br />
essentially the same geometric mean value for the area between the 1 and 4 km 2 contours. The a<br />
and b curves are in good agreement considering the 11-month time difference of the data. We will<br />
use the earlier set for comparison with the satellite data since it is taken closer to the time of<br />
important exposure for the pine forest.<br />
8.2.2 Close-in Foliage and 1-Meter Gamma Doses.<br />
The analysis in Section 5 indicates an effective exposure time of about 3 weeks for the pine<br />
foliage around <strong>Chernobyl</strong>. This duration is based on the assumption that fallout initially retained in<br />
135
the forest canopy makes an important dose contribution, especially the beta component. The<br />
effective duration of exposure is determined by the decay of radionuclides and the weathering of<br />
fallout particles from the foliage to the ground. Accordingly, radiobotanical data for exposure<br />
times of two to four weeks is used to derive radiation doses from the satellite imagery. With this<br />
point of view, the pine forest doses extracted from the imagery represent total doses received in<br />
about the first 21 days after the initial deposition.<br />
104 a 2<br />
10b 1 km 2<br />
1 -4 km 2<br />
4 km 2<br />
0<br />
102<br />
1 10 102 103<br />
Post-accident time (day)<br />
Figure 8-3.<br />
Accumulated close in fallout dose (1-m above ground) in 1 and 4 km 2 areas; a and b<br />
based on gamma field measurements (USSR, 1987) on 5/1/87 and 5/29/86<br />
respectively; the middle curve is the geometric mean between 1 and 4 km 2 .<br />
Table 8-3 lists the pine forest foliage doses represented by the hand-drawn contours of Figure<br />
7-8. For comparison, Table 8-3 lists the accumulated three week gamma dose one meter off the<br />
ground according to the survey data of May 29, 1986 (Izrael et al. 1987). These gamma doses are<br />
calculated with Equation 8.2 and the initial dose rates listed in Table 8-2. For each of the three<br />
136
contours, the foliage dose exceeds the 1-m gamma dose. The ratios, also listed in Table 8-3,<br />
decrease from 2.9 at the lowest dose to 1.2 at the highest dose.<br />
Table 8-3. The ratio of the pine canopy foliage dose to the gamma dose 1 m off the ground<br />
using satellite image and aerographic survey data.<br />
Foliage dose at 1-m r-dose (21 days)<br />
Dose rato<br />
Contour area contour (Table 7-2) at contour<br />
(km 2 ) (Gy) (Gy) (foliage/I-m )I<br />
14.5 20. 6.9 2.9<br />
2.9 30. 22. 1.4<br />
0.8 54. 38. 1.2<br />
8.2.3 Calculated Ratio of Foliage to 1-Meter Gamma Dose.<br />
Section 5.3.2 presents calculations of the expected beta to gamma dose ratio for cylindr;.cal pine<br />
foliage elements of various radii using the radiation transport calculations of Appendix A. The<br />
same assumptions and transport calculations provide an estimate of the ratio of the accumulated<br />
dose to foliage from beta and gamma rays to the accumulated 1-m gamma dose either under the<br />
canopy or in an open field. Figure 8-4 shows the resulting dose ratios as a function of<br />
accumulation time for foliage elements in the upper canopy. The physical parameters assumed for<br />
the canopy are given in Appendix A. We assume that a calculation of the gamma dose in an open<br />
field is the best analog for the Soviet survey data.<br />
The dose ratios in Figure 8-4 are based on an assumed initial foliar intercept fraction of 0.60<br />
and a weathering rate of 0.0495 per day as discussed in Section 5.5. Fallout decay rates are<br />
neglected. The measured ratios listed in Table 8-3 may be compared with the calculated ratios at 21<br />
days postaccident from Figure 8-4b. The comparison shows that the two higher dose contours<br />
have measured dose ratios corresponding to foliage elements of about 0.4 cm in radius. The ratio<br />
for the lower dose contour corresponds to a foliage element of 0.10 cm radius.<br />
The measured ratios are similar to the calculated ratios, although the indicated radii of foliage<br />
elements are somewhat larger than expected from the morphological discussion in Section 5. The<br />
apical meristematic tissue is estimated to be at a depth ranging from 0.01 to 0.12 cm in depth with a<br />
typical value of 0.04 cm. It is possible that other sensitive tissues at greater depths may contribute<br />
to the radiation response. For example, the new growth extension behind the apical meristem has a<br />
radius of about 0.15 cm and may be relatively sensitive to exposure early in the growing season.<br />
137
10 " 1I<br />
o 9 a) 1-my dose<br />
under canopy<br />
8<br />
I 70.047 - cm = foliage element radius<br />
)D"-0.10<br />
C 4<br />
0.0625<br />
S6<br />
b4 0. 15<br />
2 0.40 - - SZ4<br />
' - .. . .. .<br />
0<br />
O 1 I<br />
I I . . . I i , I<br />
0 10 20 30 40 50 60<br />
7<br />
M i•b) 1-mydose<br />
S6 04cmin open field<br />
0.04 cm<br />
5<br />
4<br />
-4<br />
" 0.0625<br />
~4 '-o.io<br />
.0.15<br />
blj -l<br />
---<br />
0<br />
44 0.40 -<br />
40O.<br />
0 I I I I I I ,<br />
0 10 20 30 40 50 60<br />
Time postaccident, d<br />
Figure 8-4. Ratio versus time of the accumulated dose (P and y) at the center of cylindrical foliage<br />
elements at the top of pine forest canopy to the accumulated 1 -m gamma dose a)<br />
under the canopy and b) in an open field for the same total fallout deposition.<br />
138
It is noteworthy that the trend of the data in Table 8-3 is toward a higher ratio and an implied<br />
smaller depth of sensitive tissue for lower doses. This trend supports the conclusion in Section 5<br />
that responses of pine trees to lower doses of radiation are likely to be more sensitive to beta<br />
exposure of apical meristems than are the responses at higher doses, which may be dominated by<br />
the systemic effects of more penetrating gamma rays.<br />
8.3 COMPARISON WITH HUMAN EXPOSURE DATA.<br />
Skin lesions from P-irradiation were an integral feature of the acute radiation syndrome<br />
suffered by the victims of the accident at the <strong>Chernobyl</strong> nuclear power plant (Barabanova and<br />
Osanov, 1990). About half of 115 patients who were exposed in and around the plant had<br />
radiation-induced lesions in addition to injury to the hematopoietic system. This significant<br />
contribution of .B-dose to human injury in a fallout field parallels the importance of the 0-dose to<br />
vegetation in a fallout field as emphasized in Section 5.<br />
Table 8-4 lists typical 5/y dose ratios for four groups of accident victims at the <strong>Chernobyl</strong>. The<br />
patients are grouped according to mode of exposure as described in the footnotes to Table 8-4.<br />
Barabanova and Osanov (1990) describe the variations in occurrence and severity of skin lesions<br />
within these groups. The two depths at which dose ratios are presented in Table 8-4 correspond to<br />
0.007 cm and 0.15 cm of unit density material. The smaller depth is essentially at the skin surface<br />
with little attenuation of beta dose. At 0.15 cm depth, however, beta dose is substantially<br />
attenuated as evidenced by the third column of Table 8-4. Even so, the P/t dose ratio at 0.15 cm<br />
depth is still well above 1, especially when contact sources contribute or when only a few feet of<br />
air shields the skin from the beta source.<br />
According to calculations presented in Section 5, the 0/y dose ratio at the center of a cylindrical<br />
element of pine foliage with radius 0.15 cm varies from about 5 to about I as the ground source<br />
fraction increases from 40% to 90% and the contact source fraction decreases correspondingly.<br />
This range agrees qualitatively with that in Table 8-4 at 150 mg/cm 2 for the more nearly planar<br />
geometry of human skin. Quantitative comparison would require accounting for any differences in<br />
assumed radionuclide mix and source distribution and the difference between cylindrical and planar<br />
geometry.<br />
Cylindrical geometry increases the 0/1y ratio relative to the same depth in planar<br />
geometry.<br />
The last column of Table 8-4 lists the range of gamma doses for the individuals in each patient<br />
group. Generally, these doses cannot be compared to the foliage doses derived from the satellite<br />
imagery because the victims were located in or very close to the Unit 4 reactor building rather than<br />
near the forests. However, the most severe case in Group II was an individual located 1.0 km<br />
downwind from the reactor site for about 1 hour following the reactor explosion. As reported by<br />
Barabanova and Osanov, his y dose was 12.7 Gy and his 0 dose was about 30 Gy at a depth of<br />
139
150 mg/cm 2 . These doses should be somewhat comparable to those received by any nearby trees<br />
over the same exposure period.<br />
Table 8-4. Summary of typical P/y dose ratios for four groups of people who suffered<br />
radiation lesions in the skin at <strong>Chernobyl</strong> (Barabanova and Osanov, 1990).<br />
13/r dose ratio #-dose fl'ydose ratio Range of 7 doses<br />
Groupa (at 7 mg/cm 2 ) attenuationb (at 150 mg/cm 2 ) (GY)<br />
1 3 head(face) 3 1 2- 5.8<br />
5 shoulder/chest 2<br />
20 feet 7<br />
11 20-30 10 2-3 4-12.7<br />
III 20 4 5 9-14<br />
IV (> 13 )c 3 (>4) 2-11.5<br />
aGrouped according to characteristic patterns of irradiation:<br />
Group I.<br />
Group II.<br />
Distant f-exposure (high energy); 15 patients exposed in and around the plant<br />
commencing 3 to 5 hours after explosion, little contact dose.<br />
Deposition of thin source (relatively low energy); 6 patients exposed downwind by<br />
plume and by contact with fallout immediately after the explosion.<br />
Group 111. Exposure in cloud (fireman); 6 patients exposed on roof of Unit 4 for 30 to 40<br />
minutes, clothing protected somewhat from 1-rays.<br />
Group IV. Deposition of thick source (various energies); 29 patients who were plant operators<br />
working in Unit 4 at the time of the explosion, wet clothing impregnated with<br />
radionuclides, various exposure modes.<br />
bAttenuation factor from 7 mg/cm 2 to 150 mg/cm 2 skin tissue depth.<br />
clnferred from Table 1 of Barabanova and Osanov.<br />
The y dose of 12.7 Gy in about one hour implies a dose rate of about 30,000 R/d. This rate is<br />
one hundred times larger than the rates calculated from the aerographic surveys and listed in<br />
Table 8-2. If applied to a stationary pine tree for three weeks according to Equation 8.2, this<br />
initial dose rate would result in a dose about one hundred times higher than actually observed either<br />
from the satellite data or from the aerographic surveys. It seems likely then that the dose received<br />
by this individual was dominated by the airborne cloud dose for both beta and gamma rays as the<br />
140
adioactive plume passed over and around him and not by fallout deposition. Since the cloud dose<br />
would affect the forest for only a limited time as well, this assumption avoids inconsistency in the<br />
dose calculations for the pine forest. It indicates, however, that the cloud dose may be a<br />
substantial contributor to the dose received by the pine forest since the doses received by this<br />
individual are as much as 20% or 30% of the doses estimated for the nearby pine forest. Occurring<br />
over a shorter exposure time, the cloud dose would also be weighted more strongly than the fallout<br />
dose in the induction of biological responses in the foliage.<br />
8.4 UNCERTAINTIES IN THE IMAGE ANALYSIS.<br />
The following uncertainties provide important caveats for the interpretation of the results<br />
presented in this report and equally important guides for future research:<br />
1) Detectability offoliage radiation response.<br />
There is little quantitative information available on radiation-induced spectral changes in pine<br />
foliage. Our assumptions regarding the detectability from orbit of morphological and biological<br />
changes based on visual descriptions in the literature are subject to uncertainty.<br />
2) Seasonal variation offoliage response.<br />
The seasonal variation in the detectability of radiation-induced damage to foliage needs better<br />
definition, especially for intermediate doses in the sublethal to mid-lethal range. For these<br />
intermediate doses, meristem damage may be obvious during the growing season but not cause<br />
observable spectral changes outside the growing season.<br />
3) Relative biological effect (RBE).<br />
Most of the published data for the time dependence of spectral changes of foliage are for<br />
gamma rays or mixed exposure to gamma rays and fission neutrons. Fallout involves mixed<br />
exposures of gamma and beta rays. Although beta and gamma rays have similar microdosimetry,<br />
differences in depth distribution of dose may influence observable spectral changes. Neutron<br />
exposure causes ionization tracks at higher average linear energy transfer (LET) than either beta or<br />
gamma rays. In mammals, the RBE of neutrons is sometimes unity but may vary upwards or<br />
downwards by a factor of two or more depending on the biological endpoint under consideration.<br />
We are not aware of published data on RBE for spectral changes in foliage.<br />
141
4) Radiation transport in foliage.<br />
We have not accounted for the full three dimensional, heterogeneous nature of the fallout<br />
deposition and radiation transport problem. This neglect will influence the relative importance of<br />
the exposure contribution from beta and gamma exposure.<br />
5) Radionuclide release and fractionation.<br />
Uncertainty in the radionuclide mix released from the reactor and fractionation during transport<br />
and deposition contributes to the uncertainty in the beta to gamma dose ratio for the sensitive<br />
tissues of the foliage at different distances from the reactor.<br />
6) Cloud dose to the foliage.<br />
We have little information regarding the contribution of the radioactive cloud to direct exposure<br />
of the foliage as it drifted over and about the trees within a few kilometers of the reactor.<br />
7) Weathering of the fallout.<br />
Weathering of fallout particles from the foliage to the ground and subsequent migration into the<br />
ground are important factors in the time dependence of the beta to gamma dose ratio and the<br />
effective exposure time for the foliage.<br />
8) Physiological basis of pine tree response.<br />
Analysis of the spectral response of pine tree foliage to irradiation, especially the time<br />
dependence of that response, requires a careful study of the physiology of pines, including the<br />
behavior of important radiosensitive tissues and their depth distribution. The change in the<br />
sequence of physiological events leading to mortality as the dose is increased above the LD 5 0 is of<br />
particular importance and needs further elucidation.<br />
9) Influence of decontamination and containment activities.<br />
We know that helicopters sprayed polymers around the reactor site to immobilize fallout<br />
particles, but we do not know the timing, or location of such spraying and whether it influenced<br />
the satellite spectral observations for pine forest.<br />
142
SECTION 9<br />
CONCLUSION AND RECOMMENDATIONS<br />
Our analysis of Landsat imagery of the area within a few kilometers of the <strong>Chernobyl</strong> nuclear<br />
reactor station provides maps of radiation dose to pine forest canopy resulting from the accident of<br />
April 26, 1986. Detection of the first date of significant, persistent deviation from normal of the<br />
spectral reflectance signature of pine foliage produces contours of radiation dose in the 20 to 80 Gy<br />
range extending up to 4 km from the site of the reactor explosion.<br />
According to arguments presented in Section 5, the effective duration of the dominant exposure<br />
of the pine foliage is about 3 weeks. For this exposure time, the L 5 0 of Pinus sylvestris (Scotch<br />
pine) is about 23 Gy. At this dose level, the onset of persistent deviation from normal for the<br />
spectral signature is delayed until about one year after exposure. Around twice the L 50 , persistent<br />
deviation begins as soon as 4 to 5 weeks after the start of exposure. In a limited area, response of<br />
pine foliage was observed as soon as 12 days after the start of exposure, corresponding to a dose<br />
above 80 Gy.<br />
A patch of forest about 2 km west and 1.5 km north of the Unit 4 reactor site showed a<br />
transient deviation from normal during the late growing season of 1986. This deviation is likely<br />
the result of a dose in the 7 to 15 Gy range, about 1/3 to 1/2 of the L 5 0 . Accounting for such<br />
transient deviations, the practical threshold for remote detection of radiation dose to pine foliage<br />
with the Landsat Thematic Mapper is probably about 1/4 of the L 50 .<br />
These conclusions, stated relative to the L 50 , should remain valid even if the effective exposure<br />
time for the pine forest at <strong>Chernobyl</strong> is found to be different than three weeks and our dose<br />
estimates are adjusted accordingly. The threshold of detectability by remote observation of foliage<br />
response relative to the L 50 is likely to apply to other evergreen plant species, as well.<br />
We beliel/e that the results reported here contribute to an improved understanding of the effects<br />
of high levels of fallout radioactivity on vegetation and, especially, on the remote observation of<br />
radiation-induced foliage response and the extraction of dose estimates from those observations.<br />
The results may be used to gain an improved understanding of the radiation exposure<br />
consequences to personnel operating in an area contaminated by radioactive fallout and the impact<br />
of vegetation on that exposure.<br />
The following recommendations build on the results of the present research:<br />
1) Establish a cooperative effort with scientists of the former Soviet Union to better compare<br />
satellite data with ground studies made at specific locations within a few kilometers of the<br />
<strong>Chernobyl</strong> power plant.<br />
143
2) Reexamine the question of the relative importance of specific pine tree tissues in causing<br />
outward manifestations of radiation injury, especially considering variations during the<br />
annual cycle.<br />
3) Determine the contribution and impact of the airborne cloud exposure to the total dose<br />
received by the pine forest.<br />
4) Perform calculations of the beta and gamma exposures of sensitive tissues in the pine tree<br />
using improved geometry and improved data or tCie appropriate radionuclide mix.<br />
5) Analyze the satellite imagery to account for variations in seasonal observability of sublethal<br />
to midlethal damage.<br />
6) Extract characteristic, time-dependent spectral signatures for the pine forest as a function of<br />
dose to the foliage as a guide for future observations.<br />
7) Extend the methodology for dose estimation to include exposures during seasons other than<br />
spring.<br />
144
SECTION 10<br />
LITERATURE CITED<br />
Aleksakhin, R.M., F.A. Tikhomirov and N.Y. Kulikov, 1970, Status and Problems of Forest<br />
Radioecology, Soviet J. Ecology, 1:19-27.<br />
Amiro, B.D. and J.R. Dugle, 1985, Temporal Change in Boreal Forest Tree Canopy Cover Along<br />
A Gradient of Gamma Radiation, Canadian J. Botany, 63:15-20.<br />
Anno, G. H. and A. Laupa, 1989, <strong>Chernobyl</strong> <strong>Accident</strong> Fatalities and Causes, Technical Report<br />
DNA-TR-89-275, <strong>Defense</strong> <strong>Nuclear</strong> Agency, Washington, DC, February 1989.<br />
Anonymous, 1962, Atlas of the Ukrainian SSR and Maldavian SSR, Central Administration of<br />
Geodesy and Cartography, Moscow (In Russian).<br />
Asmolov, V.G., et. al., 1987, The <strong>Accident</strong> at the <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant: One Year<br />
After, International Conference on <strong>Nuclear</strong> Power Performance and Safety, Vienna,<br />
28 September-2 October 1987, IAEA-CN-48/63, International Atomic Energy Agency, Vienna<br />
(Translated from Russian).<br />
Barabanova, A. and D. P. Osanov, 1990, The Dependence Of Skin Lesions On The Depth-Dose<br />
Distribution From Beta-irradiation Of People In The <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant <strong>Accident</strong>,<br />
Int. J. Radiation. Biol., Vol. 57, No. 4, 775-782.<br />
Beck, H.L., 1980, Exposure Rate Conversion Factors for Radionuclides Deposited on the<br />
Ground, EML-378, Environmental Measurements Laboratory, U.S. <strong>Department</strong> of Energy.<br />
New York, NY.<br />
Beck, H. and G. De Planque, 1968, The Radiation Field in Air Due to Distributed Gamma-Ray<br />
Sources in the Ground, HASL-195, Environmental Measurements Laboratory, U.S.<br />
<strong>Department</strong> of Energy, New York, NY.<br />
Berg, L.S. 1950, Natural Regions of the USSR (O.A. Titelbaum, Translator), MacMillan Co.,<br />
New York.<br />
Biatobok, S. and W. Zelawski (eds.), 1976, Outline of the Physiology of Scots (Scotch) Pine,<br />
Foreign Scientific Publications, <strong>Department</strong> of the National Center for Scientific, Technical and<br />
Economic Information, Warsaw.<br />
Borisov, A.A., 1965, Climates of the USSR (R. A. Ledward, translator), Aldine Publishing Co.,<br />
Chicago, 225pp<br />
Bohlen, C., 1987, <strong>Chernobyl</strong>'s Slow Recovery, Washington Post, July 21, 1987:A 1, A25.<br />
Bostrack, J.M. and A.H. Sparrow, 1969, Effects of Chronic Gamma Irradiation on the Anatomy<br />
of Vegetative Tissues of Pinus Rigida Mill, Radiation Botany 9:367-374.<br />
Bostrack, J.M. and A.H. Sparrow, 1970, The Radiosensitivity of Gymnosperms, i1, One the<br />
Nature of Radiation Injury and Cause of Death of Pinus Rigida and P. Strobus after Chronic<br />
Gamma Irradiation, Radiation Botany 10:131-143.<br />
145
Broido, A. and J.D. Teresi, 1961, Analysis of the Hazards Associated with the Radioactive Fallout<br />
Material, I, Estimation of A- and f3-doses, Health Physics 5:63-69.<br />
Chamberlain, A.C., 1970, Interception and Retention of Radioactive Aerosols by Vegetation,<br />
Atmos. Env. 4:57-78.<br />
Crist, E. P. and R. C. Cicone, 1984, A Physically-Based Transformation of Thematic Mapper<br />
Data -- The TM Tasseled Cap, IEEE Transactions on Geoscience and Remote Sensing, Vol.<br />
GE-22, No. 3, May 1984.<br />
Crist, E. P., R. Laurin, and R. C. Cicone, 1986, Vegetation and Soils Information Contained in<br />
Transformed Thematic Mapper Data, Proceedings of IGARSS 1986 Symposium, Zurich, 8-11<br />
September 1986, pp. 1465-1470.<br />
DOE, 1987, Health and Environmental Consequences of the <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant<br />
<strong>Accident</strong>, DOE/ER-0332, UC-41 and 48, National Technical Information Service, Springfield,<br />
VA 22161.<br />
Dewdney, J.C., 1982, USSR in Maps, Holmes & Meier Publishers, Inc., New York.<br />
Donini, B., 1967, Effects of Chronic Gamma-Irradiation on Pinus Pinea and Pinus Halepensis,<br />
Radiation Botany, 7:183-192.<br />
Duda, R. 0., and Hart, P. E., 1973, Pattern Classification and Scene Analysis, John Wiley &<br />
Sons.<br />
Fridland, V.M., 1976, Patterns of Soil Cover, Israel Program for Scientific Translation,<br />
Jerusalem, 291 pp.<br />
Fullard, H., 1972, Soviet Union in Maps, George Philip & Son Ltd., London.<br />
Gamache, G., 1993, Private communication. Report in preparation for the <strong>Defense</strong> <strong>Nuclear</strong><br />
Agency.<br />
Golovina, L. P., M. N. Lysenko and T. I. Kisel., 1980, Content and Distribution of Zinc in the<br />
Soil of the Ukrainian Poles'ye, Soviet Soil Science 12(1):73-80<br />
Hoffman, F.O. and C.F. Baes, III, 1979, A Statistical Analysis of Selected Parameters for<br />
Predicting Food Chain Transport and Internal Dose of Radionuclides, NUREG/CR-<br />
1004/ORNL/NUREGHM-282, Oak Ridge National Laboratory, Oak Ridge, TN 37830.<br />
International Advisory Committee, 1991, The International <strong>Chernobyl</strong> Project: An Overview,<br />
International Atomic Energy Agency, Vienna.<br />
Izrael, Y.A., V.N. Petrov and D.A. Severov, 1987, Modeling Radioactive Fallout Near the<br />
<strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant <strong>Accident</strong>, Meteorologiya i Gidrologiya 7:5-12.<br />
IZVESTIYA, 1989, Confession of a Veteran of <strong>Chernobyl</strong>. Written From a Hospital Bed,<br />
18001257 Moscow IZVESTIYA in Russian, 19 June 1989, Morning Edition, page 4.<br />
Reprinted in English in National Affairs, Suffering From <strong>Chernobyl</strong>-Related Illness Continues,<br />
FBIS-SOV-89-124, 29 June 1989, page 77-79.<br />
146
Kantz, A.D., 1971, Measurement of Beta Dose to Vegetation From Close-in Fallout, pp. 56-70, In<br />
D.W. Benson and A.H. Sparrow (eds.), Survival of Food Crops and Livestock in the Event of<br />
<strong>Nuclear</strong> War, CONF-700909, National Technical Information Service, Springfield, VA.<br />
Karaban, R.T., N.N. Mishenkov, B.S. Prister, R.M. Aleksakhin, F.N. Tikhomirov, G.N.<br />
Romanov, and M.A. Naryshkin, 1978, Radiation Effects on Arboreal Plants During the First<br />
Year After Acute Gamma-Irradiation of a Forest, Lesovedenii, 1:39-44.<br />
Keller, B. A., 1927, Distribution of Vegetation on the Plains of European Russia, Journal of<br />
Ecology 15:189-233.<br />
Kendrew, M. A., 1942, The Climates of the Continents, Oxford University Press, New York,<br />
473 pp.<br />
Klechkovskii, V.M., G.G. Polikarpov, and R.M. Aleksakhin, 1973, Radioecology (Translated<br />
from Russian by H. Kaner and H. Mills; translation edited by D. Greenberg, Israel Program<br />
for Scientific Translations, London), John Wiley and Sons, NY.<br />
Krupskiy, N. K., V. P. Kuz'michev, and R. G. Derevyanko., 1970, Humus Content in<br />
Ukrainian Soils, Soviet Soil Science 2:278-288.<br />
Lange, R., M.H. Dickerson and P.H. Gudiksen, 1987, Dose Estimates from the <strong>Chernobyl</strong><br />
<strong>Accident</strong>, UCRL-96934 (preprint), Lawrence Livermore National Laboratory, Livermore, CA<br />
94550.<br />
Lysenko, M. N. and L. P. Golovina., 1982, Boron Content and Distribution in the Soils of the<br />
Ukrainian Poles'ye, Soviet Soil Science 14(1):89-97.<br />
Mackin, J., S. Brown and W. Lane, 1971, Measurement and Computational Techniques in Beta<br />
Dosimetry, pp. 51-55. In D.W. Benson and A.H. Sparrow (eds.), Survival of Food Crops<br />
and Livestock in the Event of <strong>Nuclear</strong> War. CONF-700909. National Technical Information<br />
Service, Springfield, VA.<br />
McClellan et al., 1992, <strong>Chernobyl</strong> Doses: Volume 2 -- Conifer Stress Near <strong>Chernobyl</strong> Derived<br />
from Landsat Imagery, Technical Report DNA-TR-92-37-V2, <strong>Defense</strong> <strong>Nuclear</strong> Agency,<br />
Washington, DC, December 1992.<br />
McCormick, J.F., 1967, Effects of Ionizing Radiation on a Pine Forest, pp. 78-87. In D.J.<br />
Nelson and F.C. Evans, (eds.), Symposium of Radioecology, CONF-670503, U.S. Atomic<br />
Energy Comm., Div. Tech. Inform. Ext., Oak Ridge, TN.<br />
Miller, G. L., 1968, The Influence of Season on the Radiation Sensitivity of an Old Field<br />
Community, Ph.D. Thesis, University of North Carolina, Chapel Hill, NC.<br />
Monk, C.D., 1966, Effects of Short-Term Gamma Irradiation on an Old Field, Radiation Botany<br />
6:329-335.<br />
NEA, 1987, The Radiological Impact of the <strong>Chernobyl</strong> <strong>Accident</strong> in OECD Countries, <strong>Nuclear</strong><br />
Energy Agency, Organisation for Economic Co-Operation and Development, Paris Cedex,<br />
France.<br />
Oleynik, V. S., 1981, Genetic Characteristics of Peat Soils in the Ukrainian Western Poles'ye,<br />
Soviet Soil Science 13(4):32-37.<br />
147
Osanov, D.P., J.J. Tissen, and G.B. Radzievsky, 1969, Dose Distribution of 13 Radiation of<br />
Fission Products in the Tissue-Equivalent Material, Health Physics 17:489-495.<br />
Painter. E.L., F.W. Whicker, 1993, <strong>Chernobyl</strong> Doses: Habitat and Vegetation Near the<br />
<strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station, Vol. 3, <strong>Defense</strong> <strong>Nuclear</strong> Agency, DNA-TR-92-37-V3.<br />
Pedigo, R.A., 1963, Effects of Ionizing Radiation on Pinus Taeda L, pp. 295-299. In V. Schultz<br />
and A.W. Klement, Jr., (eds.), Radioecology, Van Nostrand-Reinhold, Princeton, NJ.<br />
Platt, R.B., 1963, Ionizing Radiation and Homeostasis of Ecosystems, pp. 39-60. In<br />
G.M. Woodwell (ed.), Ecological Effects of <strong>Nuclear</strong> War, BNL-917 (C-43), Brookhaven<br />
national Laboratory, Upton, NY.<br />
Public Health Service, 1960, Radiological Health Handbook, Division of Radiological Health,<br />
Public Health Service, U.S. Dept. of Health, Education, and Welfare, Washington, DC.<br />
Rhoads, W.A., R.B. Platt, R.A. Harvey, and E.M. Romney, 1969, Ecological and Environmental<br />
Effects from Local Fallout from Cabriolet, 1, Radiation Doses and Short-Term Effects on the<br />
Vegetation from close-in Fallout, USAEC Rept, PNE-956, EG&G, Inc., Goleta, CA.<br />
Rodin, L.E., N.I. Bazilevich, and N.N. Rozov, 1975, Productivity of the World's Main<br />
Ecosystems, pp. 13-26. In Productivity of World Ecosystems, National Academy of<br />
Sciences, Washington, DC.<br />
Sanbur, G. N. and I. I. Kovalenko., 1969, Improvement and Rational Utilization of Lowland<br />
Saline Soils of the Southern Poles'ye and Northern Forest-Steppe of the Ukrainian SSR,<br />
Soviet Soil Science 1:1401-1408.<br />
Sheppard, M.I., D.H. Thibault, and J.E. Guthre, 1982, Radiosensitivity and the Role of Pinus<br />
Sylvestris L. In the Revegetation of Radioactive Disposal Areas, Environmental and<br />
Experimental Botany 22(2):193-198.<br />
Sobdovych, E. V., V.P. Dudkin, S.N. Kireyev, N.N. Kaletnik, I.N. Los, V.V. Gudzenko, D.A.<br />
Bobyleva, and L.Y. Tabachny, 1992, <strong>Chernobyl</strong> Catastrophe: Reasons and Consequences for<br />
Kiev, Minsk, and Moscow. Ukrainian Academy of Sciences: Kiev.<br />
Sparrow, A.H., A.F. Rogers, and S.S. Schwemmer, 1968, Radiosensitivity Studies with Woody<br />
Plants - I. Acute Gamma Irradiation Survival for 28 Species and Predictions for 190 species,<br />
Radiation Botany 8:149-186.<br />
Sparrow, A.H., S.S. Schwemmer, and P.J. Bottino, 1971, The Effects of External Gamma<br />
Radiation from Radioactive Fallout on Plants with Special Reference to Crop Production,<br />
Radiation Botany 11:85-118.<br />
Sukachev, V. N., 1928, Principles of Classification of the Spruce Communities of European<br />
Russia, Journal of Ecology, 16:1-18.<br />
Symons, L., 1972, Russian Agriculture: A Geographic Survey, John Wiley and Sons,<br />
New York, 348 pp.<br />
Szafer, W., 1966, The Vegetation of Poland, Pergamon Press, Oxford, 728 pp.<br />
148
Takhtajan, A., 1986, Floristic Regions of the World, (T. J. Crovello, translator), University of<br />
California Press, Berkeley.<br />
Tou, J. T. and R, C. Gonzalez, 1974, Pattern Recognition Principles, Addison-Wesley Publishing<br />
Company, Reading, Massachusetts, 1974<br />
Ursino, D.J., A. Moss, and J. Stimac, 1974, Changes in the Rates of Apparent Photosynthesis in<br />
21% and 1% Oxygen and of Dark Respiration Following a Single Exposure of Three-year-old<br />
Pinus Strobus L. Plants to Gamma Radiation, Radiation Botany 14:117-125.<br />
USSR, 1986, USSR State Committee on the Utilization of Atomic Energy: The <strong>Accident</strong> at the<br />
<strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant and its Consequences, Information compiled for the IAEA<br />
experts' meeting, 25-29 August 1986, Vienna.<br />
USSR, 1987, The <strong>Accident</strong> at the <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant: One Year After, International<br />
Conference on <strong>Nuclear</strong> Power Performance and Safety, Vienna, Austria,<br />
28 September-2 October 1987, IAEA-CN-48/63, International Atomic Energy Agency,<br />
Vienna.<br />
Walter, H., 1978, Vegetation of the Earth and Ecological Systems of the Geobiospher, (J. Wieser,<br />
translator), Springer-Verlag, New York.<br />
Whicker, F.W. and L. Fraley, Jr., 1974, Effects of Ionizing Radiation on Terrestrial Plant<br />
Communities, pp. 317-366. In Advances in Radiation Biology, Vol. 4, Academic Press, Inc.<br />
Whicker, F.W. and T.B. Kirchner, 1987, PATHWAY. A Dynamic Food Chain Model to Predict<br />
Radionuclide Ingestion After Fallout Deposition, Health Physics 52(6):717-737.<br />
Wolfe, W. L. and G. J. Zissis, eds., 1978, The Infrared Handbook (Environmental Research<br />
Institute of Michigan), p. 3-140.<br />
Woodwell, G.M. and A.L. Rebuck, 1967, Effects of Chronic Gamma Radiation on the Structure<br />
and Diversity of an Oak-Pine Forest, Ecological Monographs, 37:53-69.<br />
Woodwell, G.M. and A.H. Sparrow, 1963, Effects of Ionizing Radiation on Ecological Systems,<br />
pp. 20-38. In G.M. Woodwell (ed.), Ecological Effects of <strong>Nuclear</strong> War, BNL-917 (C-43).<br />
Brookhaven National Laboratory, Upton, NY.<br />
Wulff, E. V., 1943, An Introduction to Historical Plant Geography, (E. Brissenden, translator),<br />
Chronica Botanica Co., Walham, Massachusetts, 223 pp.<br />
Young, R. W., 1988, <strong>Chernobyl</strong> in Retrospect, in Pharmac. Ther. Vol. 00, Eds. J. F. Weiss and<br />
M. G. Simic, Pergamon Press, Great Britain, 1988<br />
149
APPENDIX A<br />
FALLOUT DOSE CALCULATIONS FOR PINE FOREST CANOPY<br />
This appendix describes fallout beta and gamma radiation dose calculations in pine<br />
forest canopy for three dose points located 1, 7, and 12 meters above the ground, as<br />
indicated in Figure A-1. The 1-meter dose point is in air; the 7-meter point is in the<br />
middle of the canopy; and the 12-meter point is at the top of the canopy. Although we<br />
refer to dose calculations, the results are given in terms of dose rate, in cGy/h per unit<br />
source intensity.<br />
Fallout radiation sources are assumed to be homogeneously distributed both in the<br />
canopy mass (leaves and twigs) and on a flat surface of the ground below.<br />
homogenized canopy mass density of 0.0054 gm/cm 3 for medium density pine forest in<br />
mid-European latitudes, based on Kerr et al. (1971), is assumed to be of infinite extent<br />
laterally and finite vertically as indicated in Figure A-1. The dose calculations are all<br />
given in terms of the dose rate cGy/h per unit area or volume intensity, depending upon<br />
An<br />
Top canopy dose pt.<br />
Zu77<br />
Tree,. 32f Mid-canopy dose pt.<br />
• !•i:liiii!!i!!~ i !iiiii!~~iiii• . S.. T re e 32 ft 3• •<br />
CaCanopy<br />
.:+il!:iiii+ • i~~iii~:::+. height 40 ftIJ (10 m)*<br />
]<br />
....... ii~ (12 I s ,,<br />
.: !!N .~iii<br />
i'....<br />
~ii!iii~ii..........iil S....<br />
Air 8ft (2 m)<br />
"Dimensions assumed<br />
777777771'/77/7777•7//////<br />
7m<br />
Figure A-1. Pine forest canopy model for dose calculations based on Kerr et al., 1971.<br />
Dose points are 1, 7, and 12 meters ground surface.<br />
A-1
whether the fallout source is distributed over a surface (such as the ground) or within a<br />
space (such as the canopy), respectively.<br />
The beta particle dose calculations are not explicitly performed for each beta<br />
decay spectral component, but rather are based on the mean and maximum beta energies<br />
according to a semiempirical beta dose relationship (discussed later in this Appendix).<br />
Accordingly, the convention we utilize to designate the unit source intensity for beta<br />
particles is fp/cm 2 -sec or P/cm 3 -sec for radionuclide decay where "Pr" effectively<br />
represents all the beta decay spectra components from the excited radionuclide states.<br />
For the gamma dose calculations, the explicit gamma ray energies and<br />
corresponding fractional yields associated v. ith radionuclide decay<br />
are utilized.<br />
Accordingly, the convention we utilize to designate the unit source intensity for gamma<br />
rays is y/cm 2 -sec or y/cm 3 -sec where "y" represents the frequency weighted effective<br />
gamma ray decay components.<br />
GAMMA DOSES.<br />
Gamma dose calculations for foliage (pine needles or budding meristems) at the<br />
canopy dose points indicated in Figure A-1 are done in three parts. The ground fallout<br />
source is treated as one part and the fallout retained in the canopy is divided into two<br />
parts, that due to fallout in direct contact with (deposited on) the foliage element whose<br />
dose is being calculated and that due to fallout deposited on the rest of the canopy.<br />
Parameters for the gamma dose calculations are listed by gamma energy in Table A-1.<br />
Calculations were first performed for each gamma energy and then the specific<br />
radionuclide source parameters listed in Table A-2 were applied to calculate the dose for<br />
the radionuclide source mix in Table A-2 based on weighting by the individual released<br />
source strength, Q(MCi).<br />
The doses determined for each gamma energy in Table A-1 provide enough data to<br />
easily interpolate to find dose values for the radionuclide disintegration energies in<br />
Table A-2. Accordingly, if Rij(x) represents the gamma dose rate component for the ith<br />
energy of the jth radionuclide, the total dose rate at a dose point position x is<br />
Dy(x) = Y fiX Rij(x)Yij (A.1)<br />
j<br />
where, fj = Qj / XQj and Yij is disintegration yield in Table A-2.<br />
J<br />
The gamma dose is determined by integrating a point source dose function G(r) over<br />
a surface or volume source region, 9t,<br />
A-2
-)<br />
M<br />
A T-- (N CN 0 W) ot M<br />
w<br />
ChW<br />
l<br />
N 0 - O<br />
W)4 (iel 4C (2r<br />
t- r - It 0 n r- A t t<br />
(A 0<br />
U)<br />
0~<br />
to<br />
E<br />
0u<br />
r<br />
a.tt<br />
011<br />
Cuo<br />
3 7o<br />
Cu > e<br />
COz<br />
rig<br />
w<br />
n<br />
'-4 co<br />
I~~~<br />
W E<br />
g~ oo o0oo o 0 c<br />
A-3
00 1,-<br />
o<br />
00t<br />
0 000<br />
C6<br />
C6<br />
N %000<br />
00<br />
0<br />
en Rt00<br />
a e0%<br />
en 0<br />
cc 00 e<br />
0 0 C)N<br />
ui 00 N C 0<br />
A? 65 66 c; 6 cc,<br />
Q6 z C) C4<br />
00 n 00 )e<br />
0n 0 0 000<br />
0 - 00 W)C n000ýN<br />
>4 6 65dC 6 6 i6 6 6 66 1U0<br />
0)0<br />
E %C - 0% q C, *0' '0 ON -1('<br />
00~ 00 q ~ n 0 e<br />
'0 00 CPO t 0 4N N0 0r'<br />
6 CC6 6 C5 66 d d 6 d 6<br />
N N% Q '0 r- -N n<br />
- ?- M I)<br />
00000~0 l-4:0'0002<br />
00<br />
ff 0~ei Cý 4 ( C4 " %f 46c<br />
E<br />
~~-4N.O00<br />
'00%<br />
0%cc<br />
E ~ 4<br />
o 66666 66<br />
NN''0y4~ '0 ~U<br />
0%b<br />
A-4a
L= KEy(9en/p)SJG(r)d9C<br />
where, Ey is the gamma energy in MeV, (pen/P) is the energy absorption coefficient in<br />
cm 2 /gm, S= 1, is the unit gamma source strength per unit area or volume, and K is a dose<br />
conversion constant,<br />
1.6xl 10( ergs )X3600(s-e)<br />
K 1=, MeV 7 =5.76x 10- 5 ( gm -cGy -sec<br />
10(ergs)<br />
MeV-h }<br />
Then an energy dependent dose constant, P.f(E), is<br />
Py(E) =5.76x 1O 5 E(-ten / p)<br />
and the dose rate, DL, will have units of (cGy/h)/(y/sec-cm 2 or cm 3 ).<br />
Canopy Fallout Volume Source/12- and 7-Meter Dose Point. The calculational<br />
geometry for the canopy fallout source and dose point at the top of the canopy is shown<br />
below in Figure A-2.<br />
Co 12 m dose point, Dp<br />
Canopy top<br />
_________<br />
Ar $- -<br />
1= 10 m<br />
;~dA pd4odp<br />
Figuie A-2. Calculational geometry--canopy fallout volume source, dose point at 12 meters, side<br />
view upper panel, aerial view lower.<br />
A-5
As shown in Figure A-2, a flat disk source of radioactive fallout material of radius a<br />
is imbedded in the homogenized canopy mass a distance z from the top of the canopy.<br />
The differential element of area is dA=pd4dp. The dose at the top of the canopy from<br />
disc source having unit gamma source intensity, Sa=1 y/cm 2 -sec, is given by,<br />
2n a -U<br />
D= (E)Sa 2 f d B(4r)e" pdp<br />
o<br />
o<br />
- Py(E)Sa 2 Br)2 dp (cGy/h)/(y/cm 2 -sec). (A.2)<br />
0<br />
Since, r 2 =p 2 +z 2 , and rdr=pdp,<br />
DpPy(E)Sa s B(jr)e- d<br />
2 (A.3)<br />
z<br />
Implementing the Berger's form of the dose buildup factor for gamma radiation given by,<br />
B(p) = 1 + agreb<br />
,(A.4)<br />
where a and b are fit constants given in Table A-1, the gamma dose integral becomes,<br />
D Dp= Py(E)Sa[je 2 dr+aj e-(-b)Prdr . (A.5)<br />
The first term in the bracket, we call I(z,s) can be written as<br />
1 1 (z,s) = -du -J--du<br />
u u<br />
These integrals are a specific form of the general exponential integral function,<br />
b<br />
m -t<br />
Then I(Ipz, ps) = EI(Mtz) - Ej(ps)<br />
Integrating the second term in the bracket of Equation A.5 above, the dose, Dp1<br />
becomes,<br />
A-6
DP= Y(E)Sa Ejjz IP) ,(-~u - e -(1-b)gs] (A.6)<br />
Assuming the radioactive fallout to be distributed over a homogenized volume of the<br />
canopy mass, a unit finite volume source element at a depth z in the canopy is dSv =<br />
Sadz. Then, dDp<br />
dDp= 2){ }dz (A.7)<br />
For purposes of integrating the disc source over the canopy mass volume, the source disc<br />
may be approximated by a radially infinite source plane were s-'o and therefore, both<br />
EI(jts) and the exponential term, exp[-(1-b)ps], in Equations (A.6) and (A.7) are zero.<br />
Accordingly, the canopy mass is approximated as a homogeneous, vertically finite,<br />
laterally infinite slab source of radioactive material. Integrating Equation (A.7),<br />
D<br />
= Py(E)Sv 2 {tzz-z(ab)pzdz} Ej(pz)dz + 1--a e-1b)fd<br />
Py(E)Sv { a [1_ e-(l-b)W (A.8)<br />
0 E(d+(1- b)2•<br />
Utilizing a property of exponential integral functions,<br />
En(y)<br />
d-n+1 d E (y)<br />
the firbt integral in the bracket of Equation (A.8) is,<br />
t<br />
•<br />
fJE 1 (pz)dz= dE 2 (Y) [1- E 2 (W)]<br />
0 90 I<br />
Employing the parameters given in Table A-I, the dose at the 12-meter point at the top<br />
of the canopy, for f = 1000 cm, is then given by,<br />
2<br />
DP yE) { 1 - E 2 (Ige) + (a-) e--(I-b)WI<br />
=2.88x10-Ey(Y(ten/p){ } (
Similarly, the dose at the 7-meter point, in the middle of the canopy (see Figure A-1), is<br />
given by Equation (A-9) for f = 500 cm, and then doubled. Figure A-3 gives the<br />
calculated gamma dose rates per unit source, (cGy/h)/(y/cm3-sec) at the top (12 rn dose<br />
point) and middle (7 rn dose point) of the canopy for gamma energies from 0.1 to<br />
3.0 MeV. The results are interpolated to obtain the dose rates, Rij(x), for the gamma-ray<br />
source, disintegration energies of the radionuclides in Table A-2.<br />
Then the source<br />
intensity-weighted gamma dose rate due to the homogenized canopy source at the two<br />
canopy dose points are given below employing Equation (A. 1).<br />
10-2<br />
EC.)<br />
10-3 -<br />
7i<br />
CD<br />
10-4<br />
0.1 1.0 3.0<br />
Gamma energy, MeV<br />
Figure A-3. Gamma dose rate in the canopy from canopy fallout volume source versus gamma energy.<br />
A-8
Dose Point<br />
Dose Rate, (cGy/h)/(Y/cm 3 -sec)<br />
Top of canopy 1.651 x 10-3<br />
12-meters<br />
Middle of canopy 2.064 x 10- 3<br />
7-meters<br />
Ground Fallout Source/12- and 7-Meter Dose Points.<br />
The calculational geometry for the ground fallout source for the dose point at the<br />
middle and top of the canopy is shown below in Figure A-4. In Figure A-4, a flat disc of<br />
radioactive material represents the ground fallout source a distance f = 7 m below the<br />
mid canopy dose point. The vertical air distance between the bottom of the canopy and<br />
the source disc is e 1 = 2 m; this region has an air density, Pi = 1.226 x 10- 3 g/cm 3 . The<br />
canopy thickness between the dose point and bottom of the canopy is f = 5 m; this<br />
region has a homogenized canopy density P2 = 0.0054 g/cm 3 .<br />
Canopy top<br />
_ _ _ _ __ OA__ _ _ _ _ _ _ __ _ _ _ _ _ _ _<br />
12 m dose point<br />
''S r t-7m<br />
Canopy<br />
Air =2m<br />
77-r~d'~7,7777L1<br />
p<br />
dA PdOp<br />
Figure A-4. Calculational geometry--ground fallout source/dose point at 7 meters.<br />
A-9
The dose at the middle of the canopy from the disc source of unit source intensity<br />
(Sa = 1 y/cm 2 -sec) is given by,<br />
2x<br />
a<br />
Dp Py(E)Sa fd*f B(ri) exp(-ET.tri)pdp (A.10)<br />
0 0<br />
Since r 2 = p 2 + e 2 , and rdr = pdp, changing the variable of integration,<br />
P<br />
f(E)Sa B( ri)-exp(-4.r)rdr . (A.11)<br />
2 = r22<br />
Using Berger's form of the buildup factor, given by B(gtiri)<br />
combining with the exponential term in Equation (A.10),<br />
1 +alg, rii exp(bip.giri) and<br />
B(.•liri)exp(-•,lgiri) =exp(-Igiri)+ .•ailairi exp[-,Y.,(-bi)•tiri ]<br />
i i i i i<br />
Then since ri = eir/e, some of the summation terms can be simplified for purposes of<br />
integration, i.e.,<br />
y~ =-gX.ri r =(r , a=19 1<br />
lagiri =r aigitir = Or, P3=-Yaigite ,and<br />
i i i<br />
r y<br />
F,(l-bj)gjtri=-<br />
i<br />
.(1-bjifi =gi= Y= •-Y(1-b)gjtji<br />
i<br />
where, p&i = (t/p)ipi and (p/p) 1 is the gamma mass attenuation coefficient for air (Pt =<br />
1.226 x 10-3, g/cm 3 ); (p,/P) 2 is that for the canopy (p 2 = 0.0054, g/cm 3 ).<br />
Equation (A. 11) can be rewritten as,<br />
(ESaj Oc+Jre idr<br />
D 2 f(e OeT)r (A. 12)<br />
t<br />
Integrating of first term in brackets results in the difference between two exponential<br />
integral functions: EI(ac) - E 1 (cts). Integrating the second term in the brackets, the dose<br />
is given as,<br />
Dp= PyP(E)S, FEI(f) --EI(as) + P.(e-Yf -e-15)<br />
A2 O1<br />
A-10
Then assuming an infinitely extended ground source plane (s--oo), the dose in the canopy<br />
from the fallout source on the ground is,<br />
Dp = 2.88x10-SE,(l.te /P)[Ej(oe)+ke-'ie] (cGy /h) /(/cm 2 -sec) . (A.13)<br />
Employing Equation (A.13) with the parameters given in Table A-1, doses were<br />
determined for various gamma energies at the middle of the canopy for t = 700 cm, f, =<br />
200 cm, and f2 = 500 cm; and at the top of the canopy for t = 1200 cm, tj = 200 cm,.<br />
and e2 = 1000 cm.<br />
Figure A-5 gives the calculated gamma dose rates per unit fallout ground source<br />
(cGy/h)/(y/cm 2 -sec) at the 7-meter and 12-meter dose points as function of gamma<br />
energy ranging from 0.1 to 3.0 MeV.<br />
Ground Fallout Dose at 1-Meter Height.<br />
The same form as given by Equation (A.13) was employed to calculate the ground<br />
fallout source dose at 1 meter in air above the surface for t = 100 m, a = (p/p) 1 pl,<br />
P = a(g/p) 1 and y = (1 - b)(gVp) 1 . Figure A-5 also gives the calculated gamma dose rates<br />
for the 1-meter dose as a function of gamma-ray energy.<br />
The results of the ground fallout source dose rates at all three dose points, 1 meter,<br />
7 meters, and 12 meters, were interpolated to obtain the dose rates, Rij(x), for the<br />
gamma-ray source, disintegration energies of the radionuclides in Table A-2. These<br />
values were then weighted and summed by the disintegration energy yields and<br />
radionuclide source intensity according to Equation (A. 1), and given below.<br />
Dose Point<br />
Dose Rate (cGy/h)/(y/cm 2 -sec)<br />
Top of canopy 1.011 X 10-6<br />
12 meters<br />
Middle of canopy 1.690 x 10-6<br />
7 meters<br />
1-meter height in air 2.936 x 10-6<br />
Canopy Fallout Contact Source/Dose to Foliage Elements.<br />
Calculations were performed to estimate the fallout gamma dose at the center of<br />
cylindrical foliage elements such as a pine needle or a budding meristem due to fallout<br />
deposited directly on the foliage element in question. It is assumed that fallout material<br />
A-I1
is homogeneously deposited on the surface of a cylinder of length 21 and radius p as<br />
shown in Figure A-6.<br />
10"5 - " .... __ ___<br />
10-5<br />
(A<br />
E<br />
10-6<br />
107<br />
o / ,<br />
0.1 1.0 3.0<br />
Gamma energy, MeV<br />
2t<br />
Figure A-5. Gamma dose rate from ground fallout source versus gamma energy.<br />
Dose point -<br />
I<br />
SdA<br />
= pddz<br />
Figure A-6. Calculational geometry for canopy fallout contact source.<br />
A-12
A point source gamma dose function at a small element of source area dA is<br />
integrated over the cylindrical surface to give the gamma dose at the dose point due to<br />
radionuclide fallout material on the surface.<br />
2M<br />
te -Ur<br />
Dy =Py(E)Sa fdfe 'dodz . (A.14)<br />
0 0<br />
Changing the variable of integration z--r, r2 = z 2 + p 2 , dz = rdr/z and z 2 = r 2 - p2,<br />
D = = Py(E)PSa 2Y 2 O r -"7r dr--<br />
2 P r ,qr- -p"<br />
Py'(E)PSa e Per r (A.15)<br />
- 2.__''- '-____<br />
22-<br />
Since wr
Calculations were based on an average length of 4.5 cm. For the budding meristem, the<br />
calculations were based on cylinders 0.04 and 0.0625 cm in diameter to simulate the dose<br />
depths discussed in Section 5.2. For horizontal orientation of the needle or meristem, the<br />
results obtained with Equation (A.17) should be doubled since the integration was over<br />
the half length f. On the other hand, if fallout was only on the top surface of the cylinder<br />
the results would in turn be halved due to symmetry considerations. We assume that<br />
these two factors cancel. However, assuming a random directional orientation of the axis<br />
of the pine needle (or meristem) between --t/2 from the horizontal, the values obtained<br />
from Equation (A.17) were multiplied by the average value of the cosine, Z (2/R) =<br />
0.63662 to account approximately for the effective area seen by incident fallout particles.<br />
Accordingly, the gamma dose values for various foliage element dimensions are given<br />
below.<br />
Dose Point<br />
Dose Rate (cGy/h)y/cm 2 -sec<br />
Radius (p) cm Length (2f) cm Random Orientation<br />
0.04 4.5 5.13 x 10-7<br />
0.04 9.0 5.14 x 10-7<br />
0.0625 9.0 5.10 x 10-7<br />
0.10 4.5 4.983 x 10-7<br />
0.15 9.0 4.95 x 10-7<br />
0.40 2.0 3.783 x 10-7<br />
BETA DOSES.<br />
Beta dose calculations were performed for the two dose points in the canopy<br />
indicated in Figure A-1 from fallout source radionuclide material dispersed in the canopy<br />
biomass. Beta dose calculations were also performed to estimate the dose in the foliage<br />
elements from radioactive fallout deposited on their surfaces. Because of the limited<br />
range of beta particles, the beta dose to the canopy dose points from fallout material on<br />
the ground surface was neglected.<br />
Fallout radionuclide parameters for the beta dose calculations are listed in Table A-3.<br />
Beta doses were calculated for each radionuclide component and then weighted<br />
according to each radionuclide source intensity, Q(MCi), and summed to obtain the dose<br />
for the source mix. If Rj(x) represents the beta dose rate component for the jth<br />
radionuclide, the dose rate at position x is,<br />
A-14
00<br />
* 72<br />
.00<br />
en enz 0r ce 0' 4C, t<br />
a, tn M Cu<br />
oZ<br />
aal<br />
-tr 0<br />
C<br />
2~~~ 0o0-u~<br />
CA<br />
6.-<br />
AU-
D6(x)= fjRj(x) , (A.18)<br />
where, f 1 = Qj /I Qj .<br />
j<br />
The beta dose calculations are based on integrating an energy dependent point source<br />
dose function over the source geometries of interest. The point source dose function J(r)<br />
is a semi-empirically derived relationship for beta particles given by Loevinger, Japha,<br />
and Brownell (1956),<br />
where<br />
1.6 x 10-<br />
k<br />
Jr=k 2 {c[1 -<br />
(vr)<br />
(yr / c -e1vr/c)]+ vrel-vr IA1<br />
[]f0, r2c/v<br />
= 1.273 x 10-9p 2 v 3 E[t ,<br />
a =-[3c2(c2-1)eC<br />
-0.55E.<br />
c = 3(e") 1<br />
V = v(E) 18.2 1. 37 (cm 2 ,g)an<br />
(E 0 - 0.036)<br />
r =source<br />
2<br />
-to -detector distance, g / cm<br />
p = density, g / cm 3 ,<br />
Ep= average beta energy, MeV,<br />
Eo= maximum beta energy, MeV,<br />
v(Eo) = apparent absorption coefficient, cm 2 / g.<br />
(A.19)<br />
Canopy Fallout Volume Source/12- and 7-Meter Dose Point.<br />
The calculation procedure for canopy fallout source and canopy dose points is the<br />
same as that given for integrating the point source gamma dose function over the canopy<br />
volume of finite vertical width, and infinite lateral extent. Referring to the calculational<br />
geometry given in Figure A-2, the dose at the top of the canopy a vertical distance z from<br />
disc source of infinite radius imbedded in the canopy is,<br />
A-16
2n<br />
Dp = fJ(r)dA = f d# J(r)rdr<br />
V 0 z<br />
=_2xk ['c/v c[1- (vr /c)e1-(vr/c)] rd vrelw rdr (A.20)<br />
1 f r 2 fd+ dr.(.0<br />
V z rzr<br />
[ ]=O, rZc/v.<br />
Carrying out the integration of Equation (A.20), we obtain,<br />
D = pl(E){c[l + n(c / vz) - el-(vz/c) ]+ e-Vz}.<br />
(A.21)<br />
Expressing Dp above as a dose rate per unit source, (cGy/h)/(I/cm 2 -sec), Pp(E) = 2.879 x<br />
10" 5 vfa(Sa - 1).<br />
In order to obtain the beta dose at the top of the canopy, the infinite disc source<br />
(Equation A.21) is integrated over the canopy source region of thickness t = 1000 cm<br />
and of density Pc = 0.0054 g/cm. Furthermore, at that point, the beta dose is calculated<br />
at the center, x, of a pine needle of density p = 1 g/cm 3 , and thickness 0.2 cm, where<br />
then, x = 0.1 gm/cm 2 . The surface source of unit intensity S. extends to a unit volume<br />
source Sv, i.e., dSV = Sadz. The volume integral is then,<br />
x+1<br />
DP = fSDp(z)dz<br />
SC+(A .22)<br />
= PO(E)S,{fc41 +en(c/vz)-e1-(Pz/c)]+ 9lvzdz}<br />
For both dose points, 12 m or 7 m, , t = 1000 cm (5.4 g/cm 2 ) and f = 500 cm<br />
(2.7 g/cm 2 ) respectively, the canopy slab source is of infinite thickness since the<br />
distances exceed the maximum range, r., of the radionuclide disintegration betas given in<br />
Table A-3. Then given, I k ro, and integrating Equation (A.22),<br />
DP(x,*) = 0.5Dp c 2 [3 -e'(vx/c) - ix-( 2 + n c + el-VX }<br />
(.3<br />
]=0-x>c/v1<br />
[<br />
A-17
where, Dp is the beta particle dose in the interior of a large (infinite) source as is the case<br />
of the canopy slab source. The beta dose at the surface of the top of the canopy is Dp/2,<br />
as indicated in Equation (A.23), assuming negligible back scatter from the air above.<br />
Expressing Dp as the dose rate per unit source, (cGy/h)/(I,/cm 3 -sec), Dp is,<br />
= 1.6x10-'6(ergs/g)xfEp(MeV/j3)<br />
100( ergs ) 0 X .0054( g__<br />
[,g -cGy<br />
c<br />
= 1.0667 x 10- 2 E19 (cGy / h)/ (D/cm3 -sec) . (A.24)<br />
The dose point given by Equation (A.23) is actually 0.1 cm above the canopy slab<br />
surface to account for penetration into the center of a pine needle. Locating the dose<br />
point 0.1 cm below the canopy slab surface just inside the canopy, the beta dose is<br />
D , (x inside, o) = Do - Dp(x,o-) (cGy / h)/(P /cm' -sec). (A.25)<br />
Equation (A.25) is used to determine the beta dose rate at the top of the canopy (12-<br />
meter dose point). Then because of infinite slab thickness conditions of the canopy mass,<br />
the beta dose rate in the middle of the canopy (7-meter dose point) is obtained by<br />
doubling the value obtained from Equation (A.25) for the 12-m dose point. Calculations<br />
were performed for each radionuclide and weighted according to the source intensity and<br />
summed to obtain the beta dose rate for the radionuclide fallout mix; they are given<br />
below.<br />
Dose Point Beta Dose Rate (cGy/h)/(P/cm 3 -sec)<br />
Top of canopy<br />
3345 x<br />
12 meters<br />
Middle of canopy<br />
6.69 x<br />
7 meters 6.69_×_10-3<br />
Canopy Fallout Contact Source/Dose to Foliage Elements.<br />
Calculations were performed to estimate the fallout beta dose at the center of foliage<br />
elements such as a pine needle or a budding meristem. It is assumed that fallout material<br />
is homogeneously deposited on the surface of a cylinder of length 2V and radius p as<br />
shown in Figure A-6.<br />
A-18
The point source dose function, J(r), at the small element of source area, dA, is<br />
integrated over the cylindrical surface to give the beta dose at the dose point,<br />
Dp = SaJ J(r) 2npdz.<br />
(A.26)<br />
A<br />
Changing the variable of integration to r, r 2 = z2 + p 2 dz = rdr/z, and z = 2 , the<br />
dose becomes<br />
27pk 2l• f 2 J(r)rdr r2 V r2-p2<br />
•1 c[i_(vr/c)e - (vr/c)] dr + el-vrdr (A.27)<br />
Since in Equation (A.27) [ 0, sJt -+p2 > c / v, the integral of the first term in the [I<br />
brackets is (1/p)cos"1 (p/(c/v)). Then the dose becomes,<br />
P P [L L(c/S. v ,-.Lpdrl+v j e dr (A.28)<br />
.i Vr 2 -P 2 P 1 7 7 J<br />
The two integrals in Equation (A.28) were evaluated numerically.<br />
Because of the<br />
conditions indicated above for the [] bracket, the upper limit, UL is subject to the<br />
conditions given as follows.<br />
Condition<br />
c/v>Jp-2+t2 >P<br />
Fpv +t2 >c/V>p<br />
UL<br />
I/p 2 -+e<br />
c/V<br />
Jfp-7+ -t2 >p>c/v 0<br />
Equation (A.28) gives the beta dose rate per unit source intensity (cGy/h)/(p/cm 2 -sec) in<br />
the center of the foliage element where P3(E) = 2.879 x 10-5 pvE'a.<br />
A-19
The assumptions made for the calculations regarding dimensions, fallout source,<br />
surface symmetry, orientation, etc., are all discussed above under the subsection "Canopy<br />
Fallout Contact Source/Dose to Foliage Elements," which describes fallout gamma ray<br />
calculations parallel to those described here for beta radiation. Utilizing Equation (A.18)<br />
with the radionuclide fallout source intensity values in Table A-3, the beta dose rates<br />
obtained are as follows.<br />
Dose Point<br />
Dose Rate (cGy/h)y/cm 2 -sec<br />
Radius (p) cm Length (21) cm Random Orientation<br />
0.04 4.5 3.02x 10-5<br />
0.04 9.0 3.04 x 10-5<br />
0.0625 9.0 2.36 x 10-5<br />
0.10 4.5 1.73 x 10-5<br />
0.15 9.0 1.24 x 10-5<br />
0.40 2.0 0.32 x 10-5<br />
SUMMARY.<br />
The various calculated beta and gamma cose components from fallout radiation are<br />
summarized in Table A-4. The dose rates are all expressed in terms of cGy/h per unit<br />
source either in the canopy or on the ground surface and reflect equal radionuclide fallout<br />
deposition in the canopy mass and on the ground surface below. Dose rates are given at<br />
three locations (see Figure A-i), two in the canopy and a third at 1 meter above the<br />
ground surface for reference. In the canopy mass, one dose point is located at the top, 12<br />
meters above the ground, and another in the middle, 7 meters above the ground.<br />
The results given for gamma and beta dose rates per unit volume source in the<br />
canopy reflect the adjustment made for the reference volume density in the canopy to be<br />
consistent with equal fallout deposition in the canopy mass and on the ground surface.<br />
That is, the fallout deposited in the canopy mass is assumed to be homogeneously<br />
distributed over the 1000 cm thickness of the canopy rather than on the plane surface of<br />
the ground, i.e., Sv(A or y/cm 3 -sec) = Sa(O or y/cm 2 -sec) x10- 3 .<br />
In addition to the dose rate contributions in the canopy from fallout radionuclide<br />
sources distributed in canopy mass and ground surface below, Table A-4 gives dose rates<br />
at these locations in the middle of cylindrical foliage elements from fallout radionuclides<br />
assumed to be deposited on their surfaces.<br />
A-20
caa<br />
a. 4.a<br />
.4,r4<br />
2- - -<br />
2 x x x x * 2<br />
0%q ON r<br />
oo<br />
0 A-<br />
c<br />
a-<br />
m co U. co U<br />
Q<br />
U.:<br />
-<br />
- - - - 2<br />
A-21 X
Inspection of Table A-4 indicates that the dominant dose component in canopy mass<br />
is beta radiation due to fallout deposition on the surface of foliage elements assuming<br />
that the fallout deposits equally in the canopy mass and on the ground surface below and<br />
that the surface density of fallout on the foliage elements is the same as that on the<br />
ground.<br />
A-22
APPENDIX B<br />
SPECTRAL DEVIATIONS FROM CLASS<br />
This appendix presents our method for constructing a deviation vector for a pixel in standard<br />
units relative to a reference class of pixels.<br />
Each pixel in a multispectral image is represented by a vector of its band intensities. The<br />
pixels belonging to a single class form a cluster in the hyperspace of band intensities. We<br />
assume that the cluster of pixels belonging to a reference site for the class may be approximated<br />
by a multivariate normal distribution. Let u be the mean vector of the pixels in ihe reference site<br />
and I be the covariance matrix for the reference site.<br />
In Bayes decision theory, the distance of a pixel x from the mean of its class is called the<br />
Mahalanobis distance (Duda and Hart, 1973) when it is scaled relative to the covariance matrix<br />
Y. The square of the Mahalanobis distance r is given by<br />
r2 = (x -)T Z-1 (x - A). (B-1)<br />
For a multivariate distribution, the Mahalanobis distance r is the equivalent of the normal<br />
deviate z used for a univariate normal distribution. It is a measure of the deviation of a pixel<br />
from its class mean in standard units. Since r is a scalar quantity, it provides no information<br />
regarding the direction of the deviation in hyperspace.<br />
We have generalized the Mahalanobis distance to a Mahalanobis vector m that points in the<br />
direction of deviation and whose magnitude is the Mahalanobis distance r. For a pixel that<br />
deviates from its class, the magnitude r indicates the significance of the deviation. For a<br />
statistically significant deviation, the direction of the vector m carries information regarding the<br />
likely cause of the deviation.<br />
Assume that a multispectral image has been transformed to Tasseled Cap space (TC space).<br />
In general, the cluster for a reference site will be hyperelliptical with its principle axes tilted with<br />
respect to the TC axes and with unequal variance along the principle axes. The upper panel of<br />
Figure B-I illustrates a reference site cluster in two dimensions. The eigenvectors of the<br />
covariance matrix of the cluster lie along the principle component coordinate axes of the cluster.<br />
The eigenvalues are the variances of the cluster in the direction of these axes.<br />
B-1
U)<br />
ou Reference Deviation<br />
SCluster<br />
-~Site<br />
Vco<br />
0 A •'<br />
(U<br />
0)<br />
Principle<br />
Coordinote<br />
Axes of<br />
Cluster<br />
Tu,,;:eled Cap Greenness<br />
"1 Normalization"<br />
Wetness<br />
Deviation<br />
•. t. ;" .. ;= p<br />
S..-Maholonobis Vector m<br />
* ..- *- .-<br />
"- Greenness<br />
.. "'" ... . Deviation<br />
Figure B-I.<br />
The normalization procedure defined in Appendix B transforms the reference site<br />
cluster into a cluster with unit covariance matrix.<br />
B-2
Let U be the matrix whose column vectors are the unit eigenvectors of the covariance matrix<br />
Y- Let x be a pixel vector in TC space and d = x - g be the deviation of the pixel from the<br />
reference site mean in TC space. The deviation d can be expressed in the principle component<br />
coordinate system of the reference site cluster by the transformation<br />
d = UT d. (B-2)<br />
Likewise, the covariance matrix .p of the cluster in the principle component coordinate system is<br />
: p = UT yU. (B-3)<br />
Since U is constructed from the eigenvectors of 1, the matrix I<br />
is diagonal with its diagonal<br />
components equal to the variance of the cluster along each of its principle axes. In fact, the usual<br />
procedure for finding the matrix U is to diagonalize Y-<br />
The reference site cluster may be scaled to a spherical distribution with unit variance along<br />
each of its axes in the principle component coordinate system by dividing each component of the<br />
pixel deviation vectors dp by the standard deviations of the cluster along the corresponding<br />
principle axis. This scaling is accomplished by multiplication of each pixel deviation vector by a<br />
normalizing matrix Np:<br />
mp =- Np dp,<br />
(B-4)<br />
The matrix Np is a diagonal matrix given by<br />
N Np = Ip-1/2 (B-5)<br />
where the root is taken term-by-term on the right hand side of the equation. In other words, the<br />
diagonal terms of N are the inverses of the standard deviations of the reference site cluster along<br />
the principle axes.<br />
The vector mp given by Equation 4 for a pixel is the desired Mahalanobis vector expressed in<br />
the principle component coordinate system of the cluster. Since U is a unitary matrix,<br />
U- 1 = UT. So mnp may be expressed in the Tasseled Cap coordinate system by the<br />
transformation<br />
which is the inverse transformation of that in Equation 2.<br />
m= Ump. (B-6)<br />
B-3
Combining Equations 6, 4, 2, and the definition of d, we have<br />
m= UNPUT (x-) (B-7)<br />
Equation 7 shows that the normalizing matrix N for calculating Mahalanobis vectors in the<br />
TC coordinate system is<br />
N = U [UT L U] -1/2 UT. (B-8)<br />
where the root on the square brackets is taken term-by-term. Finally, the desired Mahalanobis<br />
vector expressing the deviation vector of a pixel x from its class in standard units in Tasseled<br />
Cap space is<br />
m = N (x - R). (B-9)<br />
This normalization, when applied to each pixel of the reference site cluster, results in a cluster<br />
with unit covariance matrix as illustrated in the lower panel of Figure B- 1.<br />
If the cluster is well approximated by a multivariate normal distribution, then the significance<br />
of the deviation of any given pixel from the mean of the cluster may be judged with a chisquared<br />
test on the value of r 2 for the pixel. In practice, the usual procedure is to adjust a<br />
threshold of significance on r 2 by inspection of the resulting spatial and temporal patterns of<br />
significant deviations. The threshold is raised just enough to eliminated random spatial patterns<br />
of deviation or to eliminate deviations that occur before the time of a known stimulus.<br />
B-4
APPENDIX C<br />
CHRONOLOGY OF SELECTED EVENTS<br />
SATURDAY, 26 APRIL 1986,01:23<br />
" two explosions of Unit 4; concrete, graphite, and debris escaped through roof; hole<br />
exposed graphite core (1)<br />
"* smoke and fumes with radioactive material rose in a hot plume about 1800 m high (1)<br />
"* heavier debris and particles fell near site (1)<br />
"• lighter particles to west and north (1)<br />
"• winds at 1500 m were 8-10 km/s from SE (1)<br />
"* plant firemen arrived within minutes (1)<br />
"* burning graphite on roof of Unit 3 (1)<br />
"• At < I h first case of acute radiation syndrome (1)<br />
"° At = 1.5h Unit 3 shutdown (1)<br />
"• At = 24h<br />
Units 1 &2shutdown(1)<br />
26 APRIL 1986,0500<br />
First person report of Valeriy Fedorovich Zosimov (2)<br />
"Early in the morning, about 0500, I slipped into Kiev to meet my<br />
family. We returned in a private car. We reached Kopachy, not far<br />
from Pripyat, near the station. A captain with a portable radio gave us<br />
permission to leave the vehicle in Kopachy and from there walk home<br />
to Pripyat. So, we went.... My 10-year-old daughter, my wife, and me.<br />
Ahead and behind people were also walking, and from Kopachy the<br />
destuoyed fourth unit could already be seen.<br />
Where the power<br />
transmission line crosses the road there was a long band of graphite<br />
smoke.<br />
I shook the light black flakes from my coat, and they<br />
immediately dispersed. We arrived home at night (later a large part of<br />
the forest around that path was declared hot and came under the ax-it<br />
was dangerous there! Just as it was when I walked along it with my<br />
family?)"<br />
26 APRIL 1986, DAWN<br />
° all fires extinguished except burning graphite in core (1)<br />
C-1
27 APRIL 1986<br />
1200 announcement of evacuation broadcast in Pripyat (1)<br />
1400 - 1700 Pripyat evacuated with 1200 buses that had assembled in <strong>Chernobyl</strong> (1)<br />
Line several kilometers long (1)<br />
Some of population of Pripyat had already left, so the number transported was<br />
less than the 44,600 projected (1)<br />
Population of Pripyat moved initially to surrounding towns and villages<br />
2 MAY 1986<br />
Evacuation of 30 km danger zone begun.<br />
4 MAY 1986<br />
High radiation levels force government headquarters from Pripyat to <strong>Chernobyl</strong>.<br />
6 MAY 1986<br />
End of atmospheric release of radioactivity from core (1)<br />
6 MAY 1986<br />
Evacuation of the danger zone (30 km radius) completed (1)<br />
JUNE 1986<br />
Start construction of hydraulic emergency structures (3)<br />
Water Protection (1)<br />
As part of the protection of rivers and the Kiev Reservoir, an effort was made to slow the<br />
movement of long-lived radionuclides through ground or surface water. Three major<br />
undertakings were:<br />
• 140 dams and dikes to limit runoff from the site area into the codirn pond and the Pripyat<br />
river.<br />
* existing silt traps at the bottoms of the rivers, the cooling pond, and the Kiev Reservoir<br />
were scoured.<br />
* ground water barrier was built around the plant to prevent the flow of radioactive water<br />
towards the River Dnepr. The barrier was 8 km long and 30-35 m deep, down to the<br />
impxrmeable clay layer.<br />
C-2
Hydraulic Emergency Structures (3)<br />
a filtration-proof wall in the soil along part of the perimeter of the site of the power plant<br />
and wells to lower the water table<br />
* a drainage barrier for the cooling pond<br />
* a drainage cutoff barrier on the right bank of the Pripyat river<br />
• a drainage interception barrier in the south-west section of the plant<br />
* drainage water purification facilities<br />
MID-NOVEMBER 1986<br />
Completion of sarcophagus (1)<br />
APRIL 1987<br />
Completion of work begun in May 1986 for protecting the water system. (1)<br />
SOURCES<br />
(1) International Advisory Committee, 1991.<br />
(2) IZVESTIYA, 1989.<br />
(3)<br />
C-3/C-4
APPENDIX D<br />
EVERGREEN SPECTRAL SIGNATURES BY CLASS AND DATE<br />
Z1. Spectral Signature of Classes on the Winter/Summer Composite Image CONP21<br />
Name of class = FOREST1 Number of points in sample = 481<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 45.00 124.00 150.00 57.00 138.00 111.00 147.00<br />
Mean 53.53 134.61 160.67 65.30 147.03 128.89 151.16<br />
Max 66.00 144.00 180.00 81.00 154.00 143.00 157.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 21.86 -19.39 24.98 17.52 -11.79 -19.44 6.87<br />
Band 2 -19.39 19.16 -23.20 -16.96 12.02 19.46 -6.31<br />
Band 3 24.98 -23.20 31.25 21.48 -14.66 -23.80 7.58<br />
Band 4 17.52 -16.96 21.48 21.89 -13.55 -25.50 6.35<br />
Band 5 -11.79 12.02 -14.66 -13.55 10.49 16.15 -4.39<br />
Band 6 -19.44 19.46 -23.80 -25.50 16.15 35.48 -7.12<br />
Band 7 6.87 -6.31 7.58 6.35 -4.39 -7.12 5.72<br />
Name of class = FOREST2 Number of points in sample = 1945<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 43.00 122.00 142.00 54.00 140.00 98.00 137.00<br />
Mean 46.39 145.35 149.59 56.57 155.10 142.73 138.33<br />
Max 74.00 150.00 179.00 93.00 160.00 149.00 142.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band F Band 6 Band 7<br />
Band 1 6.73 -6.66 7.96 4.67 -1.30 -5.27 1i15<br />
Band 2 -6.66 8.56 -9.37 -5.18 2.73 6.66 -1.33<br />
Band 3 7.96 -9.37 12.41 6.27 -2.98 -7.82 1.60<br />
Band 4 4.67 -5.18 6.27 6.25 -2.68 -7.18 1.03<br />
Band 5 -1.90 2.73 -2.98 -2.68 3.63 4.37 -. 66<br />
Band 6 -5.27 6.66 -7.82 -7.18 4.37 11.86 -1.18<br />
Band 7 1.15 -1.33 1.60 1.03 -. 66 -1.18 1.06<br />
D-I
Name of class = UNSUP CLASS 3<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 42.00 126.00 146.00 58.00 123.00 89.00 125.00<br />
Mean 55.32 132.61 158.74 71.12 147.73 118.98 152.09<br />
Max 59.00 149.00 175.00 97.00 174.00 133.00 180.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 11.33 -5.84 9.35 -1.41 -. 32 4.63 -. 18<br />
Band 2 -5.84 13.56 -6.40 1.01 1.20 1.15 .59<br />
Band 3 9.35 -6.40 36.26 -9.20 -3.18 10.78 -1.46<br />
Band 4 -1.41 1.01 -9.20 28.99 3.12 -15.72 -9.25<br />
Band 5 -. 32 1.20 -3.18 3.12 29.52 6.53 -6.79<br />
Band 6 4.63 1.15 10.78 -15.72 6.53 34.63 -7.10<br />
Band 7 -. 18 .59 -1.46 -9.25 -6.79 -7.10 43.78<br />
Name of class = UNSUP CLASS 4<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 42.00 126.00 146.00 18.00 137.00 115.00 125.00<br />
Mean 51.91 136.68 157.51 62.03 150.74 130.99 146.72<br />
Max 59.00 151.00 174.00 78.00 163.00 147.00 174.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 10.17 -7.66 9.11 2.67 -. 35 -1.66 -. 85<br />
Band 2 -7.66 11.74 -9.06 -1.64 1.08 2.31 2.28<br />
Band 3 9.11 -9.06 18.57 .77 -. 43 -1.17 -2.34<br />
Band 4 2.67 -1.64 .77 10.25 -. 39 -5.85 -1.02<br />
Band 5 -. 35 1.08 -. 43 -. 39 8.53 1.75 .19<br />
Band 6 -1.66 2.31 -1.17 -5.85 1.75 12.17 -3.08<br />
Band 7 -. 85 2.28 -2.34 -1.02 .19 -3.08 23.89<br />
Name of class = UNSUP CLASS 5<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
---- ............... ----- --------.--........................<br />
Min 42.00 126.00 146.00 5.00 113.00 125.00 34.00<br />
Mean 46.82 142.92 151.21 56.30 154.12 140.47 139.67<br />
Max 59.00 153.00 167.00 69.00 169.00 159.00 161.00<br />
Covar Band I Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 4.06 -2.71 2.70 .54 -1.26 -. 32 -1.13<br />
Band 2 -2.71 6.73 -4.05 -1.05 2.93 3.20 -. 16<br />
Band 3 2.70 -4.05 5.62 .61 -2.36 -2.30 .03<br />
Band 4 .54 -1.05 .61 8.11 .89 -3.98 3.05<br />
Band 5 -1.26 2.93 -2.36 .89 8.21 3.03 .50<br />
Band 6 -. 32 3.20 -2.30 -3.98 3.03 10.68 -6.14<br />
Band 7 -1.13 -. 16 .03 3.05 .50 -6.14 17.15<br />
D-2
Name of class = UNSUP CLASS 6<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 42.00 126.00 146.00 37.00 137.00 107.00 104.00<br />
Mean 49.97 138.04 154.41 61.46 159.59 136.74 139.40<br />
Max 59.00 151.00 168.00 77.00 176.00 156.00 156.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 7.19 -3.56 3.73 .41 -3.19 1.09 -1.35<br />
Band 2 -3.56 8.94 -3.20 -. 13 1.46 .59 3.31<br />
Band 3 3.73 -3.20 10.78 -1.84 -5.09 .05 -. 12<br />
Band 4 .41 -. 13 -1.84 8.30 2.24 -2.16 -3.38<br />
Band 5 -3.19 1.46 -5.09 2.24 24.94 4.07 1.72<br />
Band 6 1.09 .59 .05 -2.16 4.07 10.97 -3.73<br />
Band 7 -1.35 3.31 -. 12 -3.38 1.72 -3.73 20.67<br />
Name of class UNSUP CLASS 7<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 42.00 126.00 146.00 57.00 147.00 109.00 114.00<br />
Mean 55.18 131.70 159.84 67.61 163.87 130.24 142.73<br />
Max 59.00 146.00 175.00 88.00 175.00 143.00 166.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 7.75 -3.97 4.85 1.18 -1.81 -1.14 .08<br />
Band 2 -3.97 9.72 -4.55 -1.12 .80 1.79 2.22<br />
Band 3 4.85 -4.55 24.13 -5.80 -1.34 1.59 .43<br />
Band 4 1.18 -1.12 -5.80 18.08 1.04 -10.21 -. 90<br />
Band 5 -1.81 .80 -1.34 1.04 21.94 3.48 4.36<br />
Band 6 -1.14 1.79 1.59 -10.21 3.48 17.77 -6.67<br />
Band 7 .08 2.22 .43 -. 90 4.36 -6.67 27.50<br />
Name of class = UNSUP CLASS 8<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 42.00 126.00 146.00 61.00 160.00 123.00 85.00<br />
Mean 54.21 131.43 157.61 70.10 177.29 136.55 137.41<br />
Max 59.00 146.00 173.00 82.00 193.00 157.00 155.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 11.00 -6.93 7.83 2.89 2.30 -2.29 -. 01<br />
Band 2 -6.93 12.81 -6.84 -4.95 -2.17 2.12 3.75<br />
Band 3 7.83 -6.84 24.59 -. 39 6.41 -3.94 4.62<br />
Band 4 2.89 -4.95 -. 39 11.97 4.68 -2.58 -5.08<br />
Band 5 2.30 -2.17 6.41 4.68 27.57 3.32 4.10<br />
Band 6 -2.29 2.12 -3.94 -2.58 3.32 13.36 -7.05<br />
Band 7 -. 01 3.75 4.62 -5.08 4.10 -7.05 24.94<br />
D-3
Name of class = CLASS 3 REFERENCE SITE Number of points in sample 70<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 50.00 128.00 155.00 64.00 137.00 112.00 151.00<br />
Mean 55.31 132.07 161.81 69.44 144.97 123.59 154.19<br />
Max 59.00 137.00 169.00 77.00 153.00 131.00 158.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
---- --------........<br />
----- --------........................-<br />
Band 1 3.96 -3.29 3.97 2.35 -. 79 -2.21 1.26<br />
Band 2 -3.29 3.82 -4.13 -2.76 1.92 3.27 -1.45<br />
Band 3 3.97 -4.13 7.35 .88 -3.35 -. 92 1.02<br />
Band 4 2.35 -2.76 .88 9.79 -2.40 -9.56 2.70<br />
Band 5 -. 79 1.92 -3.35 -2.40 8.00 2.05 -. 37<br />
Band 6 -2.21 3.27 -. 92 -9.56 2.05 13.70 -3.74<br />
Band 7 1.26 -1.45 1.02 2.70 -. 37 -3.74 2.72<br />
Name of class CLASS 4 REFERENCE SITE Number of points in sample = 253<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 45.00 131.00 151.00 56.00 145.00 126.00 146.00<br />
Mean 50.11 137.43 156.95 59.35 150.43 131.92 150.58<br />
Max 56.00 143.00 164.00 63.00 156.00 138.00 157.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 4.94 -4.09 5.64 1.99 -. 13 -2.06 -. 36<br />
Band 2 -4.09 6.30 -6.03 -2.09 1.09 3.31 -. 51<br />
Band 3 5.64 -6.03 9.20 2.77 -. 80 -3.59 -. 34<br />
Band 4 1.99 -2.09 2.77 2.57 -. 94 -3.21 .14<br />
Band 5 -. 13 1.09 -. 80 -. 94 4.22 2.96 -. 39<br />
Band 6 -2.06 3.31 -3.59 -3.21 2.96 7.45 -1.46<br />
Band 7 -. 36 -. 51 -. 34 .14 -. 39 -1.46 6.83<br />
Name of class = CLASS 5 REFERENCE SITE Number of points in qample = A4<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 43.00 141.00 146.00 54.00 152.00 134.00 137.00<br />
Mean 45.12 146.38 148.48 55.53 155.51 143.44 137.69<br />
Max 50.00 150.00 155.00 59.00 160.00 147.00 140.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .66 -. 32 .40 .19 .11 -. 15 .12<br />
Band 2 -. 32 1.48 -. 80 -. 27 .47 .61 -. 20<br />
Band 3 .40 -. 80 1.58 .27 -. 37 -. 60 .19<br />
Band 4 .19 -. 27 .27 .61 -. 13 -. 47 .00<br />
Band 5 .11 .47 -. 37 -. 13 2.19 1.20 -. 40<br />
Band 6 -. 15 .61 -. 60 -. 47 1.20 2.57 -. 41<br />
Band 7 .12 -. 20 .19 .00 -. 40 -. 41 .48<br />
D-4
Name of class = CLASS 6 REFERENCE SITE Number of points in sample = 467<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 46.00 131.00 146.00 57.00 149.00 130.00 132.00<br />
Mean 48.29 138.08 154.12 61.14 157.37 137.04 136.99<br />
Max 55.00 143.00 160.00 66.00 169.00 143.00 141.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.65 -. 71 1.25 .70 -1.18 -1.55 .81<br />
Band 2 -. 71 2.89 -2.22 -. 47 -. 43 1.43 -. 51<br />
Band 3 1.25 -2.22 4.60 .06 -2.18 -2.50 .98<br />
Band 4 .70 -. 47 .06 2.47 2.56 -. 32 -. 71<br />
Band 5 -1.18 -. 43 -2.18 2.56 14.12 5.21 -3.63<br />
Band 6 -1.55 1.43 -2.50 -. 32 5.21 6.40 -2.39<br />
Band 7 .81 -. 51 .98 -. 71 -3.63 -2.39 3.29<br />
Name of class = CLASS 8 REFERENCE SITE Number of points in sample = 133<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 48.00 126.00 147.00 63.00 169.00 134.00 135.00<br />
Mean 53.65 131.73 158.92 67.95 177.39 138.28 137.26<br />
Max 59.00 137.00 169.00 73.00 186.00 142.00 139.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 7.33 -5.52 8.04 3.90 6.26 -. 40 -.01<br />
Band 2 -5.52 6.96 -7.57 -3.83 -6.27 .29 -.14<br />
Band 3 8.04 -7.57 12.63 4.63 7.68 -. 53 -.04<br />
Band 4 3.90 -3.83 4.63 4.42 6.75 -. 32 -. 13<br />
Band 5 6.26 -6.27 7.68 6.75 14.60 1.02 -.72<br />
Band 6 -. 40 .29 -. 53 -. 32 1.02 2.27 .03<br />
Band 7 -. 01 -. 14 -. 04 -. 13 -. 72 .03 .40<br />
D-5
E2. Spectral Signatures for Five Zvergreen Classes on each Date<br />
Class 3, Date 1; 6 JUN 85 Number of points in sample = 70<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 64.00 137.00 112.00 119.00 152.00 159.00 151.00<br />
Mean 69.44 144.97 123.59 121.96 156.47 160.17 154.19<br />
Max 77.00 153.00 131.00 125.00 160.00 162.00 158.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 9.79 -2.40 -9.56 .28 -3.07 .17 2.70<br />
Band 2 -2.40 8.00 2.05 -. 66 1.66 -. 29 -. 37<br />
Band 3 -9.56 2.05 13.70 .14 3.17 -. 01 -3.74<br />
Band 4 .28 -. 66 .14 1.24 .19 .17 -. 10<br />
Band 5 -3.07 1.66 3.17 .19 2.73 -. 21 -. 90<br />
Band 6 .17 -. 29 -. 01 .17 -. 21 .49 -. 09<br />
Band 7 2.70 -. 37 -3.74 -. 10 -. 90 -. 09 2.72<br />
Class 3, Date 2; 21 MAR 86 Number of points in sample = 72<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 50.00 128.00 155.00 121.00 158.00 159.00 89.00<br />
Mean 55.31 132.07 161.83 127.69 159.75 160.62 89.47<br />
Max 59.00 137.00 169.00 134.00 162.00 162.00 90.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 3.88 -3.23 3.90 3.66 -. 37 .12 -. 14<br />
Band 2 -3.23 3.73 -4.08 -3.73 .25 -. 05 -. 03<br />
Band 3 3.90 -4.08 7.21 4.84 -. 55 .09 .04<br />
Band 4 3.66 -3.73 4.84 5.97 .03 .06 -. 08<br />
Band 5 -. 37 .25 -. 55 .03 .79 -. 13 -. 08<br />
Band 6 .12 -. 05 .09 .06 -. 13 .50 .00<br />
Band 7 -. 14 -.03 .04 -. 08 -. 08 .00 .25<br />
Class 3, Date 3; 29 APR 86 Number of points in sample = 72<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 58.00 137.00 124.00 120.00 156.00 158.00 131.00<br />
Mean 61.18 139.53 131.51 123.36 158.92 159.65 132.78<br />
Max 66.00 143.00 136.00 126.00 161.00 162.00 136.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 3.23 -. 68 -4.04 -. 29 -. 47 .23 .73<br />
Band 2 -. 68 1.94 .38 -. 16 .23 -. 26 -. 42<br />
Band 3 -4.04 .38 7.92 .53 .05 -. 28 -1.18<br />
Band 4 -. 29 -. 16 .53 1.43 -. 01 -. 08 .07<br />
Band 5 -. 47 .23 .05 -. 01 1.28 -. 16 .07<br />
Band 6 .23 -. 26 -. 28 -. 08 -. 16 .49 .20<br />
Band 7 .73 -. 42 -1.18 .07 .07 .20 1.03<br />
D-6
Class 3, Date 4; 8 MAY 86 Number of points in sample = 72<br />
Band I Band 2 Band 3 Band 4 Band 5 Eand 6 Band 7<br />
Min 57.00 139.00 119.00 117.00 155.00 159.00 140.00<br />
Mean 62.51 143.28 127.58 120.69 158.74 160.43 143.69<br />
Max 69.00 148.00 134.00 123.00 162.00 162.00 148.,"<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 4.81 -1.96 -5.09 .43 -1.65 -. 15 .53<br />
Band 2 -1.96 4.01 1.87 -1.24 1.16 .29 -. 39<br />
Band 3 -5.09 1.87 8.80 .48 1.89 .50 -2.30<br />
Band 4 .43 -1.24 .48 2.26 .00 .10 - .42<br />
Band 5 -1.65 1.16 1.89 .00 1.86 .02 -. 34<br />
Band 6 -. 15 .29 .50 .10 .02 .58 -. 24<br />
Band 7 .53 -. 39 -2.30 -. 42 -. 34 -. 24 3.81<br />
Class 3, Date 5; 24 MAY 86 Number of points in sample = 72<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 63.00 137.00 121.00 118.00 155.00 159.00 147.00<br />
Mean 66.93 143.11 128.10 123.19 158.24 160.28 150.40<br />
Max 73.00 150.00 134.00 126.00 161.00 162.00 154.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 4.56 -1.78 -4.53 .64 -1.55 .26 .89<br />
Band 2 -1.78 5.02 2.63 -1.41 1.43 .02 -. 86<br />
Band 3 -4.53 2.63 8.51 -. 24 2.03 .07 -1.97<br />
Band 4 .64 -1.41 -. 24 1.85 -. 44 .14 .22<br />
Band 5 -1.55 1.43 2.03 -. 44 1.45 .06 -. 31<br />
Band 6 .26 .02 .07 .14 .06 .59 -. 02<br />
Band 7 .89 -. 86 -1.97 .22 -. 31 -. 02 2.73<br />
Class 3, Date 6; 31 MAY 86 Number of points in sample = 70<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 63.00 136.00 124.00 124.00 154.00 159.00 151.00<br />
Mean 67.59 140.59 131.67 126.63 158.56 160.57 154.63<br />
Max 74.00 147.00 139.00 129.00 162.00 162.00 159.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 5.13 -1.37 -6.03 .62 -1.34 .66 1.03<br />
Band 2 -1.37 3.74 2.94 -. 32 .48 -. 07 -. 13<br />
Band 3 -6.03 2.94 9.49 -. 74 1.67 - .44 -1.84<br />
Band 4 .62 -. 32 -. 74 1.21 -. 02 .03 .07<br />
Band 5 -1.34 .48 1.67 -. 02 1.59 - .13 -. 36<br />
Band 6 .66 -. 07 -. 44 .03 -. 13 .51 -. 09<br />
Band 7 1.03 -. 13 -1.84 .07 -. 36 -. 09 3.32<br />
D-7
Class 3, Date 7; 15 OCT 86 Number of points in sample 72<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 30.00 150.00 133.00 97.00 159.00 158.00 101.00<br />
Mean 33.32 152.24 137.10 99.65 160.65 159.03 101.54<br />
Max 36.00 155.00 141.00 101.00 163.00 161.00 102.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.55 .54 -. 86 -. 10 -. 07 .09 .36<br />
Band 2 .54 1.40 .25 -. 17 - .07 .05 .19<br />
Band 3 -. 86 .25 2.54 .37 -. 24 -. 07 -. 12<br />
Band 4 -.10 -. 17 .37 .76 -. 03 11 -. 02<br />
Band 5 -. 07 -. 07 -. 24 -. 03 .74 .10 -. 08<br />
Band 6 .09 .05 -. 07 .11 .10 .53 -. 01<br />
Band 7 .36 .19 -. 12 -. 02 -. 08 -. 01 .25<br />
Class 3, Date 8; 2 DEC 86 Number of points in sample = 72<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 22.00 146.00 137.00 90.00 158.00 159.00 78.00<br />
Mean 23.86 147.79 138.87 92.65 160.43 159.63 78.85<br />
Max 26.00 149.00 141.00 95.00 162.00 161.00 79.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .59 -. 03 -. 04 .08 .07 .20 .05<br />
Band 2 -.03 .61 .03 .00 .07 -. 03 -. 03<br />
Band 3 -. 04 .03 .59 .22 .01 .08 .00<br />
Band 4 .08 .00 .22 1.09 .06 .09 .02<br />
Band 5 .07 .07 .01 .06 .76 -. 04 -. 01<br />
Band 6 .20 -. 03 .08 .09 -. 04 .30 .05<br />
Band 7 .05 -. 03 .00 .02 -. 01 .05 .13<br />
Class 3, Date 9; 11 MAY 87 Number of points in sample = 72<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 57.00 140.00 106.00 115.00 153.00 159.00 127.00<br />
Mean 61.22 144.14 124.44 117.43 158.15 160.18 129.64<br />
Max 74.00 147.00 131.00 120.00 162.00 162.00 132.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 9.06 -2.49 -10.47 .60 -2.95 .31 1.36<br />
Band 2 -2.49 3.02 2.77 -. 10 .72 .09 .16<br />
Band 3 -10.47 2.77 16.02 .12 3.73 -. 38 -2.24<br />
Band 4 .60 -. 10 .12 1.25 -. 20 -. 07 .38<br />
Band 5 -2.95 .72 3.73 -. 20 1.97 -. 17 -. 64<br />
Band 6 .31 .09 -. 38 -. 07 -. 17 .74 .23<br />
Band 7 1.36 .16 -2.24 .38 -. 64 .23 1.87<br />
D-8
Class 3, Date 10; 7 SEP 87 Number of points in sample 72<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 50.00 141.00 134.00 114.00 158.00 158.00 124.00<br />
Mean 52.53 144.90 140.28 116.81 159.93 159.42 126.14<br />
Max 56.00 150.00 145.00 119.00 163.00 161.00 128.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.75 -. 03 -1.84 -. 09 -. 20 .17 .48<br />
Band 2 -. 03 3.61 1.34 .15 .46 -. 34 -. 39<br />
Band 3 -1.84 1.34 5.74 .87 .62 -. 06 -.99<br />
Band 4 -. 09 .15 .87 1.47 .21 -. 03 -. 33<br />
Band 5 -. 20 .46 .62 .21 1.03 -. 07 -. 09<br />
Band 6 .17 -. 34 -. 06 -. 03 -. 07 .35 -. 06<br />
Band 7 .48 -. 39 -. 99 -. 33 -. 09 -. 06 .96<br />
Class 3, Date 11; 28 MAY 88 Number of points in sample = 72<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 52.00 146.00 103.00 108.00 150.00 158.00 145.00<br />
Mean 56.49 149.96 123.54 112.11 156.17 159.93 148.24<br />
Max 65.00 159.00 132.00 115.00 159.00 161.00 153.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 5.33 -. 70 -7.86 .28 -2.02 .34 1.83<br />
Band 2 -. 70 7.13 1.28 -. 84 -. 33 -. 86 -. 23<br />
Band 3 -7.86 1.28 16.69 1.04 3.29 -. 40 -4.47<br />
Band 4 .28 -. 84 1.04 1.57 .21 .15 -. 47<br />
Band 5 -2.02 -. 33 3.29 .21 2.42 -. 07 -. 42<br />
Band 6 .34 -. 86 -. 40 .15 -. 07 .52 .26<br />
Band 7 1.83 -. 23 -4.47 -. 47 -. 42 .26 4.22<br />
Class 4, Date 1; 6 JUN 85 Number of points in sample = 253<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 56.00 145.00 126.00 116.00 156.00 158.00 146.00<br />
Mean 59.35 150.43 131.92 118.71 159.34 160.18 L50.58<br />
Max 63.00 156.00 138.00 122.00 162.00 162.00 157.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Sand 7<br />
Band 1 2.57 -. 94 -3.21 -. 02 -. 21 -. 15 .14<br />
Band 2 -. 94 4.22 2.96 -. 10 .14 -. C4 -. 39<br />
Band 3 -3.21 2.96 7.45 .25 .41 .21 -1.46<br />
Band 4 -. 02 -. 10 .25 1.23 .21 .09 -. 29<br />
Band 5 -. 21 .14 .41 .21 1.10 .02 - .07<br />
Band 6 -. 15 -. 04 .21 .09 .02 .54 .28<br />
Band 7 .14 -. 39 -1.46 -. 29 -. 07 .28 6.83<br />
D-9
Class 4, Date 2; 21 MAR 86 Number of points in sample = 257<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 45.00 131.00 151.00 118.00 157.00 158.00 88.00<br />
Mean 50.12 137.42 156.96 123.72 159.64 160.58 89.24<br />
Max 56.00 143.00 164.00 129.00 162.00 163.00 91.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 4.92 -4.07 5.62 3.72 -. 25 .36 -. 47<br />
Band 2 -4.07 6.27 -6.03 -4.48 .27 -. 28 .60<br />
Band 3 5.62 -6.03 9.14 5.18 -. 43 .41 -. 81<br />
Band 4 3.72 -4.48 5.18 5.15 -. 12 .41 -. 53<br />
Band 5 -. 25 .27 -. 43 -. 12 .73 .01 .07<br />
Band 6 .36 -. 28 .41 .41 .01 .70 -. 04<br />
Band 7 -. 47 .60 -. 81 -. 53 .07 -. 04 .68<br />
Class 4, Date 3; 29 APR 86 Number of points in sample = 257<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 51.00 139.00 132.00 119.00 159.00 158.00 127.00<br />
Mean 53.91 143.06 137.56 121.47 160.78 160.02 129.60<br />
Max 57.00 147.00 142.00 124.00 164.00 162.00 135.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.29 .09 -1.46 -. 04 .02 .10 .29<br />
Band 2 .09 1.84 .49 -. 39 .03 -. 18 -. 22<br />
Band 3 -1.46 .49 4.37 .33 .15 -. 14 -. 54<br />
Band 4 -. 04 -.39 .33 1.32 .03 .17 -. 13<br />
Band 5 .02 .03 .15 .03 .82 .04 -. 10<br />
Band 6 .10 -.18 -. 14 .17 .04 .51 .03<br />
Band 7 .29 -. 22 -. 54 -. 13 -. 10 .03 2.78<br />
Class 4, Date 4; 8 MAY 86 Number of points in sample = 257<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 51.00 144.00 126.00 115.00 157.00 159.00 135.00<br />
Mean 55.02 148.54 133.40 118.34 160.17 160.18 140.12<br />
Max 59.00 155.00 140.00 122.00 163.00 162.00 148.00<br />
Covar Band I Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 2.12 -. 36 -2.71 -. 13 -. 18 .18 .61<br />
Band 2 -. 36 3.08 1.72 -. 28 .03 -. 29 -. 64<br />
Band 3 -2.71 1.72 6.50 .59 .24 -. 23 -1.02<br />
Band 4 -. 13 -. 28 .59 1.52 .18 .06 - .16<br />
Band 5 -. 18 .03 .24 .18 1.08 -.11 -. 16<br />
Band 6 .18 -. 29 -. 23 .06 -.11 .47 .20<br />
Band 7 .61 -. 64 -1.02 -. 16 -. 16 .20 5.98<br />
D-10
Class 4, Date 5; 24 MAY 86 Number of points in sample = 257<br />
Band 1 Band 2 Band 3 Band 4 Bind 5 Band 6 Band 7<br />
Min 55.00 144.00 126.00 116.00 158.00 158.00 141.00<br />
Mean 58.52 148.84 134.08 119.60 160.03 160.31 146.03<br />
Max 63.00 155.00 ±40.00 122.00 163.00 162.00 153.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 2.17 -. 28 -2.41 .02 -. 13 .13 .36<br />
Band 2 -. 28 4.28 1.73 -. 90 .01 .12 -. 37<br />
Band 3 -2.41 1.73 5.73 .21 .19 -. 04 -. 88<br />
Band 4 .02 -. 90 .21 1.44 .00 .10 - .51<br />
Band 5 -. 13 .01 .19 .00 1.01 -. 02 -. 38<br />
Band 6 .13 .12 -. 04 .10 -. 02 .60 -.15<br />
Band 7 .36 -. 37 -. 88 -. 51 -. 38 -. 15 5.63<br />
Class 4, Date 6; 31 MAY 86 Number of points in sample = 253<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 57.00 141.00 131.00 122.00 156.00 158.00 143.00<br />
Mean 60.10 145.17 138.00 125.03 160.02 160.02 147.70<br />
Max 64.00 152.00 143.00 128.00 163.00 162.00 153.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.91 -. 36 -2.14 -. 23 -.11 .00 .59<br />
Band 2 -. 36 3.50 1.47 .05 .01 -. 22 -. 35<br />
Band 3 -2.14 1.47 5.31 1.03 .11 .02 -1.54<br />
Band 4 -. 23 .05 1.03 1.33 .03 .01 -. 45<br />
Band 5 -.11 .01 .11 .03 1.17 -. 06 -. 27<br />
Band 6 .00 -.22 .02 .01 -. 06 .75 .09<br />
Band 7 .59 -. 35 -1.54 -. 45 -. 27 .09 3.19<br />
Class 4, Date 7; 15 OCT 86 Number of points in sample = 257<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 30.00 148.00 131.00 97.00 158.00 157.00 100.00<br />
Mean 32.05 152.93 138.25 99.37 160.67 159.48 101.35<br />
Max 36.00 156.00 142.00 102.00 164.00 162.00 103.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.28 .39 -. 52 -. 08 -.11 -. 16 .04<br />
Band 2 .39 2.20 1.26 -. 20 -. 01 -. 07 -. 01<br />
Band 3 -. 52 1.26 3.01 .15 -. 03 .09 .03<br />
Band 4 -. 08 -. 20 .15 .90 .02 .08 -. 02<br />
Band 5 -.11 -. 01 -. 03 .02 1.04 .10 .02<br />
Band 6 -. 16 -. 07 .09 .08 .10 .69 .00<br />
Band 7 .04 -. 01 .03 -. 02 .02 .00 .28<br />
D-11
Class 4, Date 8; 2 DEC 86 Number of points in sample = 257<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 22.00 146.00 136.00 89.00 159.00 158.00 79.00<br />
Mean 23.41 148.33 138.48 91.72 160.47 159.38 79.39<br />
Max 25.00 150.00 141.00 94.00 162.00 161.00 81.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .44 A08 -. 07 .11 .04 .09 -. 01<br />
Band 2 .08 .75 -. 25 -. 19 -. 03 -. 10 .01<br />
Band 3 -. 07 -. 25 .74 .03 .05 .11 -. 04<br />
Band 4 .11 -. 19 .03 1.09 .00 .07 .06<br />
Band 5 .04 -. 03 .05 .00 .54 .08 00<br />
Band 6 .09 -. 10 .11 .07 .08 .38 .00<br />
Band 7 -. 01 .01 -. 04 .06 .00 .00 .27<br />
Class 4, Date 9; 11 MAY 87 Number of points in sample = 257<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 50.00 143.00 126.00 113.00 157.00 158.00 125.00<br />
Mean 54.30 148.23 132.39 115.82 159.75 159.70 127.62<br />
Max 59.00 153.00 138.00 119.00 162.00 161.00 131.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 2.54 -. 51 -2.76 -. 05 -. 21 .32 .13<br />
Band 2 -. 51 2.07 1.59 -. o2 .02 -. 24 -. 21<br />
Band 3 -2.76 1.59 6.08 .22 .18 -. 19 -. 28<br />
Band 4 -. 05 -. 02 .22 1.02 .08 -. 01 .05<br />
Band 5 -. 21 .02 .18 .08 .93 -. 02 -.11<br />
Band 6 .32 -. 24 -. 19 -. 01 -. 02 .57 .00<br />
Band 7 .13 -. 21 -. 28 .05 -.11 .00 1.89<br />
Class 4, Date 10; 7 SEP 87 Number of points in sample = 257<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 47.00 142.00 138.00 113.00 158.00 157.00 123.00<br />
Mean 49.09 145.98 142.65 116.34 160.18 159.33 124.18<br />
Max 51.00 153.00 147.00 119.00 163.00 161.00 128.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .81 -. 07 -. 86 .07 .16 .14 .14<br />
Band 2 -.07 2.33 .86 -. 07 -. 16 -. 18 .12<br />
Band 3 -. 86 .86 2.76 .34 -. 33 .05 .05<br />
Band 4 .07 -. 07 .34 1.29 .15 .03 -. 09<br />
Band 5 .16 -. 16 -. 33 .15 1.00 .08 .01<br />
Band 6 .14 -. 18 .05 .03 .08 .60 .09<br />
Band 7 .14 .12 .05 -. 09 .01 .09 1.24<br />
D-12
Class 4, Date 11; 28 MAY 88 Number of points in sample 257<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 47.00 149.00 120.00 108.00 153.00 157.00 139.00<br />
Mean 50.36 153.25 130.41 110.87 157.78 159.98 142.70<br />
Max 56.00 161.00 136.00 114.00 162.00 162.00 151.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.78 -. 29 -2.01 .54 -. 39 -. 03 .40<br />
Band 2 -. 29 3.41 1.79 -. 58 .07 -. 07 -. 61<br />
Band 3 -2.01 1.79 5.96 .00 .61 .19 -1.55<br />
Band 4 .54 -. 58 .00 1.39 -. 06 .08 .24<br />
Band 5 -. 39 .07 .61 -. 06 1.34 -. 02 -. 09<br />
Band 6 -. 03 -. 07 .19 .08 -. 02 .58 .02<br />
Band 7 .40 -. 61 -1.55 .24 -. 09 .02 4.43<br />
Class 5, Date 1; 6 JUN 85 Number of points in sample 664<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 54.00 152.00 134.00 117.00 156.00 157.00 137.00<br />
Mean 55.53 155.51 143.44 119.85 159.20 159.55 137.69<br />
Max 59.00 160.00 147.00 123.00 161.00 162.00 140.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .61 -. 13 -. 47 .06 .00 .03 .00<br />
Band 2 -. 13 2.19 1.20 -. 03 -.11 -. 06 -. 40<br />
Band 3 -. 47 1.20 2.57 .25 -. 09 .08 -. 41<br />
Band 4 .06 -. 03 .25 1.33 A04 .04 -. 07<br />
Band 5 .00 -. 11 -. 09 .04 .78 .02 .06<br />
Band 6 .03 -. 06 .08 .04 .02 .61 -. 03<br />
Band 7 .00 -. 40 -. 41 -. 07 .06 -. 03 .48<br />
Class 5, Date 2; 21 MAR 86 Number of points in sample = 665<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Mil 43.00 141.00 146.00 114.00 158.00 158.00 87.00<br />
Mean 45.12 146.38 148.43 117.71 159.95 159.97 89.56<br />
Max 50.00 150.00 155.00 122.00 163.00 162.00 91.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .66 -. 32 .40 .48 .05 .08 .07<br />
Band 2 -. 32 1.47 -.80 -. 72 -. 05 -. 17 -. 08<br />
Band 3 .40 -. 80 1.59 .65 -. 03 .18 -. 03<br />
Band 4 .48 -. 72 .65 1.59 .08 .10 -. 01<br />
Band 5 .05 -. 05 -. 03 .08 .66 -. 02 .02<br />
Band 6 .08 -. 17 .18 .10 -. 02 .47 -. 03<br />
Band 7 .07 -. 08 -.03 -. 01 .02 -. 03 .51<br />
D-13
Class 5, Date 3; 29 APR 86 Number of points in sample 665<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 52.00 143.00 138.00 119.00 158.00 158.00 123.00<br />
Mean 54.00 145.96 144.03 122.73 160.86 159.72 123.80<br />
Max 56.00 149.00 148.00 126.00 164.00 161.00 126.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .40 -. 01 -. 25 .09 .01 .07 .08<br />
Band 2 -. 01 1.02 .23 -. 29 .00 -. 12 -. 08<br />
Band 3 -. 25 .23 1.38 .23 .04 .00 -. 15<br />
Band 4 .09 -. 29 .23 1.30 .06 .02 -. 02<br />
Band 5 .01 .00 .04 .06 .88 .04 -. 02<br />
Band 6 .07 -. 12 .00 .02 .04 .41 .02<br />
Band 7 .08 -. 08 -. 15 -. 02 -. 02 .02 .43<br />
Class 5, Date 4; 8 MAY 86 Number of points in sample = 665<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 52.00 147.00 136.00 116.00 157.00 158.00 128.00<br />
Mean 5:,.75 151.02 141.14 119.68 160.28 159.86 129.90<br />
Max 57.00 155.00 144.00 123.00 163.00 162.00 133.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .52 -. 15 -. 36 .07 -. 06 .12 .11<br />
Band 2 -. 15 1.29 .46 -. 18 .07 -. 18 -. 25<br />
Band 3 -. 36 .46 1.67 .09 .21 .01 -. 31<br />
Band 4 .07 -. 18 .09 1.17 .04 .03 .05<br />
Band 5 -. 06 .07 .21 .04 .81 -. 03 -. 05<br />
Band 6 .12 -. 18 .01 .03 -. 03 .49 .03<br />
Band 7 .11 -. 25 -. 31 .05 -. 05 .03 .58<br />
Class 5, Date 5; 24 MAY 86 Number of points in sample = 665<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 54.00 149.00 136.00 116.00 157.00 158.00 134.00<br />
Mean 55.49 152.51 142.02 119.99 160.08 159.60 135.21<br />
Max 59.00 155.00 145.00 124.00 163.00 162.00 138.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .57 -.11 -. 22 .07 .01 .03 -. 02<br />
Band 2 -.11 1.43 .57 -. 16 .06 -. 09 - .31<br />
Band 3 -. 22 .57 1.64 .20 .04 .06 -. 16<br />
Band 4 .07 -. 16 .20 1.40 .05 .11 -. 02<br />
Band 5 .01 .06 .04 .05 1.01 .03 -. 07<br />
Band 6 .03 -. 09 .06 .11 .03 .50 .08<br />
Band 7 -. 02 -. 31 -. 16 -. 02 -. 07 .08 .95<br />
D-14
Class 5, Date 6; 31 MAY 86 Number of points in sample = 665<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 55.00 144.00 138.00 120.00 157.00 158.00 138.00<br />
Mean 57.00 148.73 145.70 125.28 160.19 159.87 139.94<br />
Max 60.00 152.00 149.00 129.00 164.00 162.00 142.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .62 -. 15 -. 04 .37 .00 -. 02 -. 09<br />
Band 2 -. 15 1.46 .41 -. 48 -. 02 -. 01 -. 02<br />
Band 3 -. 04 .41 1.78 .46 .01 -. 04 -. 28<br />
Band 4 .37 -. 48 .46 1.96 .07 .01 -. 19<br />
Band 5 .00 -. 02 .01 .07 .85 -. 05 -. 02<br />
Band 6 -. 02 -. 01 -. 04 .01 -. 05 .60 .01<br />
Band 7 -. 09 -. 02 -. 28 -. 19 -. 02 .01 .72<br />
Class 5, Date 7; 15 OCT 86 Number of points in sample = 665<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 30.00 154.00 138.00 96.00 156.00 157.00 101.00<br />
Mean 31.98 156.68 142.43 98.86 160.25 159.21 102.04<br />
Max 34.00 160.00 146.00 103.00 163.00 161.00 103.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .49 .26 -. 02 .04 -. 01 .01 -. 01<br />
Band 2 .26 1.23 .22 -. 15 -. 10 -. 12 -. 15<br />
Band 3 -. 02 .22 1.14 .13 -. 10 .07 -. 09<br />
Band 4 .04 -. 15 .13 .99 -. 03 -. 03 -. 01<br />
Band 5 -. 01 -. 10 -. 10 -. 03 .91 .00 .06<br />
Band 6 .01 -. 12 .07 -. 03 .00 .72 .05<br />
Band 7 -. 01 -. 15 -. 09 -. 01 .06 .05 .50<br />
Class 5, Date 8; 2 DEC 86 Number of points in sample = 665<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 21.00 147.00 137.00 89.00 157.00 158.00 78.00<br />
Mean 23.39 149.57 139.48 91.70 160.22 159.31 78.58<br />
Max 25.00 153.00 145.00 95.00 162.00 161.00 80.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .33 .05 -. 05 .10 .04 .04 .00<br />
Band 2 .05 .79 -. 06 -. 17 -. 07 -. 05 .00<br />
Band 3 -. 05 -. 06 .76 .07 -. 05 .05 .03<br />
Band 4 .10 -. 17 .07 .99 .08 -. 03 .01<br />
Band 5 .04 -. 07 -. 05 .08 .59 -. 01 -. 02<br />
Band 6 .04 -. 05 .05 -. 03 -. 01 .32 -.01<br />
Band 7 .00 .00 .03 .01 -. 02 -. 01 .29<br />
D-15
Class 5, Date 9; 11 MAY 87 Number of points in sample 673<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 49.00 150.00 135.00 112.00 157.00 157.00 118.00<br />
Mean 51.55 153.28 139.53 115.86 159.91 159.23 119.29<br />
Max 55.00 157.00 143.00 120.00 163.00 161.00 122.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .58 .01 -. 30 .23 -. 02 .08 .07<br />
Band 2 .01 1.54 .82 -. 25 -. 06 -. 10 -.29<br />
Band 3 -. 30 .82 2.23 .18 -. 05 .02 -. 40<br />
Band 4 .23 -. 25 .18 1.26 .12 .11 .03<br />
Band 5 -. 02 -. 06 -. 05 .12 .83 .02 .09<br />
Band 6 .08 -. 10 .02 .11 .02 .42 .02<br />
Band 7 .07 -. 29 -. 40 .03 .09 .02 .60<br />
Class 5, Date 10; 7 SEP 87 Number of points in sample = 665<br />
Band I Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 46.00 148.00 145,00 114.00 148.00 157.00 118.00<br />
Mean 47.87 151.43 149.96 116.19 159.69 159.18 119.03<br />
Max 64.00 193.00 169.00 119.00 164.00 162.00 121.00<br />
Covar Rand 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .99 1.69 .72 .18 -. 42 .10 -. 02<br />
Band 2 1.69 5.11 2.15 -. 25 -1.19 .26 -. 07<br />
Band 3 .72 2.15 1.78 .13 -. 67 .15 -. 05<br />
Band 4 .18 -. 25 .13 1.29 .04 .03 -. 05<br />
Band 5 -. 42 -1.19 -. 67 .04 .94 -.10 -. 01<br />
Band 6 .10 .26 .15 .03 -. 10 .47 -. 01<br />
Band 7 -. 02 -. 07 -. 05 -. 05 -. 01 -.01 .24<br />
Class 5, Date 11; 28 MAY 88 Number of points in sample = 665<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 46.00 150.00 122.00 108.00 153.00 158.00 131.00<br />
Mean 48.19 154.80 135.24 110.55 158.38 160.17 132.67<br />
Max 52.00 i58.00 139.10 115.00 168.00 162.00 135.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .49 -. 02 -. 41 .16 .10 .07 .02<br />
Band 2 -. 02 1.19 .44 -. 30 709 -. 14 -. 22<br />
Band 3 -. 41 .44 2.07 .27 -. 15 .03 -. 20<br />
Band 4 .16 -. 30 .27 1.66 .06 .17 -. 09<br />
Band 5 .13 .09 -. 15 .06 1.37 -. 06 -. 03<br />
Band 6 07 -. 14 .03 .17 -. 06 .35 -. 07<br />
Band 7 .02 -. 22 -. 20 -. 09 -. 03 -. 07 .41<br />
D-16
Class 6, Date 1; 6 JUN 85 Number of points in sample 467<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
---- ................ ----- --------.-.......................-<br />
Min 57.00 149.00 130.00 118.00 157.00 158.00 13.-00<br />
Mean 61.14 157.37 137.04 121.26 159.55 159.66 136.9-'<br />
Max 66.00 169.00 143.00 125.00 162.00 162.00 141.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 2.47 2.56 -. 32 .62 -. 14 .19 - .71<br />
Band 2 2.56 14.12 5.21 -. 55 -. 34 .31 -3.63<br />
Band 3 -. 32 5.21 6.40 .61 -. 54 .25 -2.39<br />
Band 4 .62 -. 55 .61 1.97 .00 .14 -. 49<br />
Band 5 -. 14 -. 34 -. 54 .00 .99 -. 06 .17<br />
Band 6 .19 .31 .25 .14 -. 06 .54 - .17<br />
Band 7 -. 71 -3.63 -2.39 -. 49 .17 -. 17 3.29<br />
Class 6, Date 2; 21 MAR 86 Number of points in sample = 475<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
---- ........................--- .............................<br />
Min 46.00 131.00 146.00 120.00 158.00 159.00 86.00<br />
Mean 48.29 138.10 154.11 123.77 160.13 160.48 87.53<br />
Max 55.00 143.00 160.00 129.00 163.00 162.00 89.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.64 -. 73 1.26 1.02 -. 12 .08 .04<br />
Band 2 -. 73 2.92 -2.29 -1.60 .16 -. 04 .17<br />
Band 3 1.26 -2.29 4.63 1.95 -. 32 .02 -. 29<br />
Band 4 1.02 -1.60 1.95 2.42 -. 06 .12 -. 07<br />
Band 5 -. 12 .16 -. 32 -. 06 .66 .01 .03<br />
Band 6 .08 -. 04 .02 .J2 .01 .46 .01<br />
Band 7 .04 .17 -. 29 -. 07 .03 .01 .39<br />
Class 6, Date 3; 29 APR 86 Number of points in sample = 475<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
---- ........................--- .............................<br />
Min 53.00 139.00 131.00 121.00 159.00 158.00 124.00<br />
Mean 55.68 142.57 140.01 124.55 161.25 159.84 126.26<br />
Max 60.00 146.00 144.00 128.00 164.00 162.00 130.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.02 .12 -. 84 .19 -. 10 -. 03 .39<br />
Band 2 .12 1.33 .45 -. 26 -. 08 -.13 -. 19<br />
Band 3 -. 84 .45 3.19 .31 .04 .02 -. 47<br />
Band 4 .19 -. 26 .31 1.46 .01 .14 .06<br />
Band 5 -. 10 -. 08 .04 .01 .89 .02 -. 03<br />
Band 6 -. 03 -. 13 .02 .14 .02 .56 .01<br />
Band 7 .39 -. 19 -. 47 .06 -.03 .01 1.26<br />
D-17
Class 6, Date 4; 8 MAY 86 Number of points in sample 475<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 52.00 146.00 128.00 11'.00 158.00 158.00 130.00<br />
Mean 55.11 1J0.67 137.15 120.04 160.66 159.91 133.41<br />
Max 60.00 161.00 142.00 123.00 164.00 162.00 139.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.47 .83 -. 91 -. 13 -. 10 .08 .27<br />
Band 2 .83 5.07 2.04 -. 16 -. 21 -. 22 -1.50<br />
Band 3 -. 91 2.04 4.06 .41 -. 22 -. 07 -1.48<br />
Band 4 -. 13 -. 16 .41 1.30 -. 02 -. 08 -.10<br />
Band 5 -.10 -. 21 -. 22 -. 02 1.17 -. 07 .02<br />
Band 6 .08 -. 22 -. 07 -. 08 -. 07 .51 .22<br />
Band 7 .27 -1.50 -1.48 -. 10 .02 .22 2.18<br />
Class 6, Date 5; 24 MAY 86 Number of points in sample = 475<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 56.00 148.00 131.00 117.00 158.00 158.00 134.00<br />
Mean 58.56 154.63 138.10 121.14 160.41 159.92 138.30<br />
Max 66.00 169.00 144.00 125.00 163.00 162.00 143.00<br />
Covar Band I Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 2.32 2.18 -. 12 .30 -. 28 .14 -. 06<br />
Band 2 2.18 11.45 4.34 -. 07 -. 97 -. 22 -3.25<br />
Band 3 -. 12 4.34 5.40 .44 -. 50 .05 -2.14<br />
Band 4 .30 -. 07 .44 1.47 -. 04 .15 -. 14<br />
Band 5 -. 28 -. 97 -. 50 -. 04 .89 -. 03 .26<br />
Band 6 .14 -. 22 .05 .15 -. 03 .64 -. 02<br />
Band 7 -. 06 -3.25 -2.14 -. 14 .26 -. 02 3.52<br />
Class 6, Date 6; 31 MAY 86 Number of points in sample = 467<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 58.00 143.00 135.00 122.00 157.00 157.00 137.00<br />
Mean 60.39 149.61 141.88 125.93 160.63 159.46 141.68<br />
Max 65.00 160.00 147.00 130.00 163.00 162.00 147.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.51 1.51 -. 62 -. 03 -. 17 .02 .05<br />
Band 2 1.51 11.08 3.94 -. 20 -. 36 .23 -4.18<br />
Band 3 -. 62 3.94 4.46 .30 -. 09 .31 -2.47<br />
Band 4 -. 03 -. 20 .30 1.62 .18 .02 -. 14<br />
Band 1 -. 17 -. 36 -. 09 .18 1.12 -. 01 .09<br />
Bend 6 .02 .23 .31 .02 -. 01 .67 -. 24<br />
Band 7 .05 -4.18 -2.47 -. 14 .09 -. 24 3.57<br />
D-18
Class 6, Date 7; 15 OCT 86 Number of points in sample 475<br />
Band 1 Band 2 Band 3 Bind 4 Band 5 Band 6 Band 7<br />
Min 29.00 151.00 130.00 95.00 158.00 157.00 97.00<br />
Mean 32.05 154.14 138.69 99.49 160.63 159.66 99.03<br />
Max 42.00 159.00 142.00 103.00 164.00 162.00 101.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 2.54 1.06 -1.49 -. 31 -. 16 -. 02 .53<br />
Band 2 1.06 1.69 -.04 -. 24 - .10 -. 07 .20<br />
Band 3 -1.49 -. 04 2.57 .33 09 -. 03 -. 44<br />
Band 4 -. 31 -. 24 .33 1.22 -. 02 .06 -.13<br />
Band 5 -. 16 - .10 .09 -. 02 1.01 -. 04 -. 02<br />
Band 6 -. 02 -. 07 -.03 .06 -. 04 .69 -. 04<br />
Band 7 .53 .20 -. 44 -. 13 -. 02 -. 04 .62<br />
Class 6, Date 8; 2 DEC 86 Number of points in sample = 475<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 21.00 143.00 136.00 88.00 158.00 141.00 78.00<br />
Mean 23.51 148.73 138.-2 91.85 160.43 159.20 78,95<br />
Max 27.00 152.00 143.00 95.00 163.00 162.00 81.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .67 .18 -. 19 -. 04 .0: -. 16 .03<br />
Band 2 .18 1.00 -,32 -. 08 .09 .38 02<br />
Band 3 -. 19 -. 32 .88 .05 -.03 -. 37 -.05<br />
Band 4 -. 04 -. 08 .05 1.16 -.11 .22 -.06<br />
Band 5 .02 .09 -. 03 -.11 .62 .01 .01<br />
Band 6 -. 16 .38 -. 37 .22 .01 2.38 .00<br />
Band 7 .03 .02 -. 05 -. 06 .01 .00 .24<br />
Class 6, Date 9; 11 MAY 87 Number of points in sample = 103<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 51.00 147.00 123.00 113.00 157.00 158.00 124.00<br />
Mean 54.75 151.17 130.56 115.37 159.68 159.65 125.47<br />
Max 61.00 154.00 136.00 118.00 161.00 161.00 131.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 3.00 .09 -2.87 .22 .05 .18 1.16<br />
Band 2 .09 1.99 .37 .01 .17 -. 43 .21<br />
Band 3 -2.87 .37 5.95 .12 -. 21 -. 21 - .70<br />
Band 4 .22 .01 .12 1.23 .06 -. 05 -. 01<br />
Band 5 .05 .17 -. 21 .06 .75 - .09 - .11<br />
Band 6 .18 -. 43 -. 21 -. 05 -. 09 .47 .19<br />
Band 7 1.16 .21 -. 70 -. 01 -.11 .19 3.14<br />
D- 19
Class 6, Date 10; 7 SEP 87 Number of points in sample = 475<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 47.00 143.00 143.00 114.00 158.00 157.00 117.00<br />
Mean 49.52 147.73 146.81 117.94 160.93 159.17 119.07<br />
Max 54.00 153.00 151.00 121.00 163.00 161.00 121.00<br />
Covar P.4*d 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .89 .66 -. 36 -. 07 .04 .16 .14<br />
Band 2 .66 2.71 .43 -. 29 -. 08 -. 03 -. 34<br />
Band 3 -.36 .43 1.82 .12 -. 14 .01 -. 15<br />
Band 4 -.07 -. 29 .12 1.03 .00 .00 -. 03<br />
Band 5 .04 -. 08 -. 14 .00 .90 -. 05 .03<br />
Band 6 .16 -. 03 .01 .00 -. 05 .52 .10<br />
Band 7 .14 -. 34 -. 15 -. 03 .03 .10 .90<br />
Class 6, Date 11; 28 MAY 88 Number of points in sample = 475<br />
Band I Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 48.00 153.00 126.00 107.00 154.00 158.00 129.00<br />
Mean 50.45 159.47 133.48 110.87 158.31 159.78 134.19<br />
Max 56.00 171.00 141.00 116.00 161.00 162.00 140.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 1.54 1.50 -. 49 .20 -. 08 .10 .13<br />
Band 2 1.50 11.28 4.48 -. 48 -. 18 -. 15 -3.76<br />
Band 3 -. 49 4.48 4.76 .20 .15 -. 04 -2.16<br />
Band 4 .20 -. 48 .20 1.73 .08 .05 .19<br />
Band 5 -. 08 -. 18 .15 .08 1.27 -.01 -.11<br />
Band 6 .10 -. 15 -. 04 .05 -. 01 .48 .09<br />
Band 7 .13 -3.76 -2.16 .19 -.11 .09 3.86<br />
Class 8, Date 1; 6 JUN 85 Number of points in sample = 133<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 63.00 169.00 134.00 117.00 155.00 158.00 135.00<br />
Mean 67.95 177.39 138.28 119.17 158.59 159.74 137.26<br />
Max 73.00 186.00 142.00 122.00 161.00 162.00 139.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 4.42 6.75 -. 32 -. 20 -. 54 .19 -. 13<br />
Band 2 6.75 14.60 1.02 -. 88 -1.48 .24 -. 72<br />
Band 3 -- 32 1.02 2.27 .15 -. 32 .15 03<br />
Band 4 -. 20 -. 88 .15 .93 .09 -. 04 .16<br />
Band 5 -. 54 -1.48 -. 32 .09 1.27 .00 .18<br />
Band 6 .19 .24 .15 -. 04 .00 .68 .02<br />
Band 7 -. 13 -. 72 .03 .16 .18 .02 .40<br />
D-20
Class 8, Date 2; 21 MAR 86 Number of points in sample = 135<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
-----------------.--------.--------.--------- --------------<br />
Min 49.00 126.00 147.00 120.00 157.00 159.00 88.00<br />
Mean 53.65 131.71 158.93 126.55 159,76 161.13 89,21<br />
Max 59.00 137.00 169.00 132.00 162.00 163.00 91.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
------------.--------.----------------.--------.--------<br />
Band 1 7.35 -5.55 8.06 5,49 -. 86 .35 -. 22<br />
Band 2 -5.55 6.98 -7.57 -5.49 .37 -. 46 .26<br />
Band 3 8.06 -7.57 12.58 7.25 -1.13 .62 -. 26<br />
Band 4 5.49 -5.49 7.25 6.79 -. 56 .47 -. 22<br />
Band 5 -. 86 .37 -1.13 -. 56 .76 -. 03 .04<br />
Band 6 .35 -. 46 .62 .47 -. 03 .52 -. 03<br />
Band 7 -. 22 ,26 -. 26 -. 22 .04 -,03 .27<br />
Class 8, Date 3; 29 APR 86 Number of points in sample = 135<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 56.00 143.00 119.00 117.00 158.00 158.00 126.00<br />
Mean 61.39 147.63 130.09 120.01 160.72 160.01 127.68<br />
Max 67.00 152.00 139.00 124.00 163.00 161.00 129.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
------------- -------- -------- -------- -------- -------- --------<br />
Band 1 4.82 1.16 -5.01 -. 65 .09 -. 06 .14<br />
Band 2 1.16 5.25 1.71 -. 89 .41 -. 31 .42<br />
Band 3 -5.01 1.71 10.11 .77 -. 15 -. 16 .26<br />
Band 4 -. 65 -. 89 .77 1.32 .02 .18 -. 14<br />
Band 5 .09 .41 -. 15 .02 1.01 .05 -. 09<br />
Band 6 -. 06 -. 31 -. 16 .18 .05 .41 -. 14<br />
Band 7 .14 .42 .26 - .14 -. 09 -. 14 .76<br />
Class 8, Date 4; 8 MAY 86 Number of points in sample = 135<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
------------.--------.----------------.--------.--------<br />
Min 57.00 151.00 121.00 116.00 156.00 158.00 131.00<br />
Mean 62.92 160.24 131.91 118.81 159.69 159.92 132.87<br />
Max 69.00 172.00 141.00 122.00 162,00 162.00 136.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 5.59 7.11 -1.56 -. 51 -. 63 -. 50 .30<br />
Band 2 7.11 22.16 9.40 -. 30 -. 81 -1.68 -. 69<br />
Band 3 -1.56 9.40 13.43 .51 -. 18 -. 76 -,96<br />
Band 4 -. 51 -. 30 .51 1.01 .12 .14 -. 22<br />
Band 5 -. 63 -. 81 -. 18 .12 1.26 -. 09 -.11<br />
Band 6 -. 50 -1.68 -. 76 .14 -. 09 .56 .01<br />
Band 7 .30 -. 69 -. 96 -. 22 -.11 .01 .96<br />
D-21
Class 8, Date 5; 24 MAY 86 Numbe:- of points in sample 135<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 60.00 162.00 135.00 117.00 157.00 158.00 134.00<br />
Mean 67.19 172.80 139.22 120.01 159.32 159.76 135,54<br />
Max 73.00 183.00 143.00 123.00 162.00 161.00 137.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 5.46 9.44 .73 .82 -. 65 .37 .18<br />
Band 2 9.44 20.93 3.54 1.45 -1.74 .62 .59<br />
Band 3 .73 3.54 2.87 .53 -. 52 .17 .21<br />
Band 4 .82 1.45 .53 1.61 -. 24 .17 .08<br />
Band 5 -. 65 -1.74 -. 52 -. 24 .94 -. 03 -.11<br />
Band 6 .37 .62 .17 .17 -. 03 .50 .04<br />
Band 7 .18 .59 .21 .08 -.11 .04 .54<br />
Class 8, Date 6; 31 MAY 86 Number of points in sample = 133<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 63.00 156.00 139.00 122.00 157.00 158.00 136.00<br />
Mean 67.46 167.14 142.46 125.16 159.56 160.36 137.97<br />
Max 72.00 174.00 147.00 129.00 162.00 162.00 143.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 3.83 5.87 -. 22 .49 -. 48 .04 .14<br />
Band 2 5.87 11.65 .54 .45 -. 90 .26 -.11<br />
Band 3 -. 22 .54 2.36 .16 -. 30 .20 -. 03<br />
Band 4 .49 .45 .16 1.63 -. 15 .24 .06<br />
Band 5 -. 48 -. 90 -. 30 -. 15 3.04 -. 03 .04<br />
Band 6 .04 .26 .20 .24 -. 33 .68 -. 01<br />
Band 7 .14 -.11 -. 03 .06 .04 -. 01 .95<br />
Class 8, Date 7; 15 OCT 86 Number of points in sample = 135<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 33.00 151.00 126.00 94.00 157.00 159.00 102.00<br />
Mean 38.20 154.40 134.71 97.36 161.30 160.08 102.42<br />
Max 45.00 158.00 138.00 101.00 163.00 162.00 103.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 4.72 1.40 -2.28 -1.05 .35 .29 -. 12<br />
Band 2 1.40 2.07 .29 -. 44 -. 27 -,02 .06<br />
Band 3 -2.28 .29 4.18 .68 -. 62 -. 21 .25<br />
Band 4 -1.05 -. 44 .68 1.28 -.19 -. 10 .07<br />
Band 5 .35 -. 27 -. 62 -. 19 .97 -. 02 -. 08<br />
Band 6 .29 -. 02 -. 21 -. 10 -.02 .39 -. 02<br />
Band 7 -. 12 .06 .25 .07 -.08 -. 02 .24<br />
D-22
Class 8, Date 8; 2 DEC 86 Number of points in sample = 135<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 23.00 145.00 134.00 90.00 158.00 158.00 77.00<br />
Mean 24.68 147.41 137.64 92.24 160.64 160.04 77.73<br />
Max 27.00 150.00 140.00 95.00 162.00 161.00 78.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 .51 .02 -. 37 .06 .02 .06 -. 01<br />
Band 2 .02 .74 -. 09 -. 06 - .01 -. 08 .09<br />
Band 3 -. 37 -. 09 1.38 -. 03 .17 .08 .04<br />
Band 4 .06 -. 06 -. J3 1.20 .03 -. 01 .00<br />
Band 5 .02 -. 01 A17 .03 .68 .01 -. 05<br />
Band 6 .06 -. 08 .08 -. 01 .01 .35 .01<br />
Band 7 -. 01 .09 .04 .00 -. 05 .01 .20<br />
Class 8, Date 9; 11 MAY 87 Number of points in sample 94<br />
Band I Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 53.00 145.00 116.00 112.00 156.00 158.00 128.00<br />
Mean 58.74 150.18 124.34 114.59 159.23 159.90 130.07<br />
Max 65.00 157.00 132.00 118.00 162.00 161.00 133.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 6.32 2.59 -6.56 .02 -. 71 .20 -. 29<br />
Band 2 2.59 7.78 2.23 -. 03 .00 -. 19 -. 73<br />
Band 3 -6.56 2.23 13.48 .22 1.05 -. 09 .25<br />
Band 4 .02 -. 03 .22 1.05 .01 .02 -. 18<br />
Band 5 -. 71 .00 1.05 .01 1.06 .00 .12<br />
Band 6 .20 -. 19 -. 09 .02 .00 .58 .03<br />
Band 7 -. 29 -. 73 .25 -. 18 .12 .03 1.22<br />
Class 8, Date 10; 7 SEP 87 Number of points in sample = 135<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 51.00 151.00 139.00 114.00 158.00 158.00 120.00<br />
Mean 54.83 156.87 143.07 117.13 160.37 159.62 120.24<br />
Max 58.00 162.00 145.00 119.00 163.00 161.00 121.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 2.14 2.37 -1.12 -. 25 .11 -.13 .07<br />
Band 2 2.3- 4.11 -. 84 -. 49 -. 01 -. 19 .01<br />
Band 3 -1.12 -. 84 2.38 .29 -. 35 .13 -. 01<br />
Band 4 -. 25 -. 49 .29 1.03 -. 02 .06 -. 06<br />
Band 5 .11 -. 01 -. 35 -. 02 .75 .05 .00<br />
Band 6 -. 13 -. 19 .13 .06 .05 .47 .00<br />
Band 7 .07 .01 -. 01 -.06 .00 .00 .19<br />
D-23
Class 8, Date 11; 28 MAY 88 Number of points in sample = 135<br />
Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Min 56.00 170.00 129.00 107.00 148.00 158.00 130.00<br />
Mean 60.39 177.44 134.47 110.21 156.41 159.75 131.67<br />
Max 65.00 187.00 149.00 114.00 160.00 161.00 135.00<br />
Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7<br />
Band 1 5.39 8.15 .61 .24 -1.17 .13 .48<br />
Band 2 8.15 17.00 4.27 -. 64 -2.58 .20 1.20<br />
Band 3 .61 4.27 7,07 .15 -2.22 .27 .86<br />
Band 4 .24 -. 64 ,15 1.95 -. 12 .10 -. 13<br />
Band 6 -1.17 -2.58 -2.22 -. 12 2.95 -. 15 -. 26<br />
Band 6 .13 .20 .27 .10 -. 15 .49 .19<br />
Band 7 .48 1.20 .86 -. 13 -. 26 .19 1.49<br />
D-24
DEPARTMENT OF DEFENSE<br />
ARMED FORCES RADIOBIOLOGY RSCH INST<br />
ATTN: DEPT OF RADIATION BIOCHEMISTRY<br />
ATTN: BHS<br />
ATTN: DIRECTOR<br />
ATTN: EXH<br />
ATTN: MRA<br />
ATTN: PHY M WHITNALL<br />
ATTN: RSD<br />
ATTN: SCIENTIFIC DIRECTOR<br />
ATTN: TECHNICAL LIBRARY<br />
ASSISTANT SECRETARY OF DEFENSE<br />
INTERNATIONAL SECURITY POLICY<br />
ATTN: NUC FORCES & ARMS CONTROL PLCY<br />
ASSISTANT TO THE SECRETARY OF DEFENSE<br />
ATTN: EXECUTIVE ASSISTANT<br />
ATTN: MIL APPL C FIELD<br />
DISTRIBUTION LIST<br />
DNA-TR-92-37-V1<br />
OASD<br />
ATTN: DUSP/P<br />
ATTN: USD/P<br />
OFC OF MILITARY PERFORMANCE<br />
ASSESSMENT TECHNOLOGY<br />
2 CY ATTN: F HEGGE<br />
STRATEGIC & SPACE SYSTEMS<br />
ATTN: DR SCHNEITER<br />
U S EUROPEAN COMMAND/ECJ-6-DT<br />
ATTN: ECJ-6<br />
U S EUROPEAN COMMAND/ECJ2-T<br />
ATTN: ECJ-3<br />
ATTN: ECJ2-T<br />
ATTN: ECJ5-N<br />
ATTN: ECJ5N<br />
USSSTRATCOMIJ531T<br />
DEFENSE INTELLIGENCE AGENCY ATTN: J-521<br />
ATTN: DB<br />
ATTN: JPEP<br />
5 CY ATTN: DB-4 RSCH RESOURCES DIV 3416TH TTS INTERSERVICE NUC WPNS SCHOOL<br />
ATTN: DB-5C<br />
ATTN: TTV<br />
ATTN: DB-6B<br />
2 CY ATTN: TTV 3416TH TTSO<br />
ATTN: DB-6E<br />
ATTN: DiW-4<br />
DEPARTMENT OF THE ARMY<br />
ATTN: DN<br />
ATTN: DT<br />
ARMY RESEARCH LABORATORIES<br />
ATTN: OFFICE OF SECURITY<br />
ATTN: TECH LIB<br />
ATTN: OS<br />
COMBAT MATERIAL EVAL ELEMENT<br />
DEFENSE INTELLIGENCE COLLEGE<br />
ATTN: SECURITY ANALYST<br />
ATTN: DIC/RTS-2<br />
ATTN: DIC/2C<br />
DEP CH OF STAFF FOR OPS & PLANS<br />
ATTN: DAMO-SWN<br />
DEFENSE NUCLEAR AGENCY<br />
ATTN: DAMO-ZXA<br />
ATTN: DFRA JOAN MA PIERRE<br />
ATTN: NANF<br />
NUCLEAR EFFECTS DIVISION<br />
ATTN: NASF<br />
ATTN: STEWS-NE-T<br />
10CYATTN: RAEM<br />
2 CY ATTN: TITL TEXCOM<br />
ATTN: D PACE<br />
DEFENSE SYSTEMS SUPPORT ORGANIZATION<br />
ATTN: JNGO<br />
U S ARMY AIR DEFENSE ARTILLERY SCHOOL<br />
ATTN: COMMANDANT<br />
DEFENSE TECHNICAL INFORMATION CENTER<br />
2 CY ATTN: DTIC/OC U S ARMY ARMOR SCHOOL<br />
ATTN: ATSB-CTD<br />
FIELD COMMAND DEFENSE NUCLEAR AGENCY<br />
ATTN: FCPR<br />
ATTN: FCPRT<br />
FIELD COMMAND DEFENSE NUCLEAR AGENCY<br />
ATTN: FCTO<br />
NATIONAL DEFENSE UNIVERSITY<br />
ATTN: ICAF TECH LIB<br />
ATTN: NWCLB-CR<br />
ATTN: LIBRARY<br />
ATTN: STRAT CONCEPTS DIV CTR<br />
NET ASSESSMENT<br />
ATTN: DOCUMENT CONTROL<br />
Dist-1<br />
ATTN: TECH LIBRARY<br />
U S ARMY BALLISTIC RESEARCH LAB<br />
ATTN: AMSRL-SL-BS DR KLOPCIC<br />
ATTN: AMSRL-SL-BS E DAVIS<br />
ATTN: SLCBR-D<br />
ATTN: SLCBR-DD-T<br />
ATTN: SLCBR-TB<br />
U S ARMY CHEMICAL SCHOOL<br />
ATTN: CRAL FRAKER<br />
U S ARMY COMBAT SYSTEMS TEST ACTIVITY<br />
ATTN: JOHN GERDES<br />
ATTN: MIKE STANKA
DNA-TR-92-37-V1 (DL CONTINUED)<br />
U S ARMY COMD & GENERAL STAFF COLLEGE<br />
ATTN: ATZL-SWJ-CA<br />
ATTN: ATZL-SWT-A<br />
NAVAL PERSONNEL RES & DEV CENTER<br />
ATTN: CODE P302<br />
NAVAL POSTGRADUATE SCHOOL<br />
ATTN: CODE 52 LIBRARY<br />
U S ARMY CONCEPTS ANALYSIS AGENCY<br />
ATTN: TECHNICAL LIBRARY<br />
NAVAL RESEARCH LABORATORY<br />
U S ARMY FIELD ARTILLERY SCHOOL ATTN: CODE 1240<br />
ATTN: ATSF-CD ATTN: CODE 5227<br />
U S ARMY FORCES COMMAND<br />
ATTN: AF-OPTS<br />
U S ARMY FOREIGN SCIENCE & TECH CTR<br />
ATTN: C WARD<br />
NAVAL SEA SYSTEMS COMMAND<br />
ATTN: PMS-423<br />
ATTN: SEA-07R<br />
NAVAL SURFACE WARFARE CENTER<br />
ATTN: CODE F-31<br />
U S ARMY INFANTRY CENTER<br />
ATTN: G RIEL<br />
ATTN: ATSH-CD-CSO<br />
NAVAL TECHNICAL INTELLIGENCE CTR<br />
U S ARMY ITAC<br />
ATTN: NTIC-DA30<br />
ATTN: IAX-Z<br />
NAVAL WAR COLLEGE<br />
U S ARMY LABORATORY COMMAND ATTN: CODE E-1 1<br />
ATTN: DIRECTOR<br />
ATTN: CTR FOR NAV WARFARE STUDIES<br />
ATTN: DR D HODGE<br />
ATTN: DOCUMENT CONTROL<br />
ATTN: LIBRARY<br />
U S ARMY NUCLEAR & CHEMICAL AGENCY<br />
ATTN: STRATEGY DEPT<br />
4 CY ATTN: MONA-NU DR D BASH<br />
ATTN: MONA-NU MAJ BLISS<br />
NAWCWPNSDIV DETACHMENT<br />
U S ARMY TEST & EVALUATION COMMAND<br />
ATTN: STECS-NE<br />
U S ARMY VULNERABILITY/LETHALITY<br />
ATTN: AMSLC-VL-NE DR J FEENEY<br />
U S ARMY WAR COLLEGE<br />
ATTN: LIBRARY<br />
ATTN: STRATEGIC STUDIES INSTITUTE<br />
U S MILITARY ACADEMY<br />
ATTN: DEPT BEHAVORIAL SCI & LEADERSHIP<br />
ATTN: COL J G CAMPBELL<br />
ATTN: SCIENCE RESEARCH LAB<br />
ATTN: CLASSIFIED LIBRARY<br />
NUCLEAR WEAPONS TNG GROUP ATLANTIC<br />
ATTN: CODE 222<br />
ATN: DOCUMENT CONTROL<br />
NUCLEAR WEAPONS TNG GROUP PACIFIC<br />
ATTN: CODE 32<br />
OPERATIONAL TEST & EVALUATION FORCE<br />
ATTN: COMMANDER<br />
PLANS. POLICY & OPERATIONS<br />
ATTN: CODE-P<br />
ATTN: CODE-POC-30<br />
US ARMY MATERIEL SYS ANALYSIS ACTVY<br />
STRATEGY AND POLICY DIVISION<br />
ATTN: DRXSY-DS ATTN: N 51<br />
US ARMY MEDICAL RESEARCH &<br />
ATTN: NIS-22<br />
ATTN: NOP 50<br />
DEVELOPMENT COMMAND ATTN; NOP 603<br />
ATTN: SGRD-PLE ATTN. NOP 91<br />
US ARMY MODEL IMPROVEMENT STUDY<br />
MANAGEMENT AGENCY<br />
ATTN: E VISCO<br />
ATTN: NUC AFFAIRS & IN5L NEGOT BR<br />
ATTN: N455<br />
DEPARTMENT OF THE AIR FORCE<br />
WALTER REED ARMY MEDICAL CENTER<br />
ATTN: G RIPPLE<br />
DEPARTMENT OF THE NAVY<br />
MARINE CORPS<br />
ATTN: CODE PPO<br />
ATTN: PSI G RASP<br />
NAVAL OCEAN SYSTEMS CENTER<br />
ATTN: CODE 9642-B<br />
AFIS/INT<br />
ATTN: INT<br />
AIR UNIVERSITY<br />
ATTN: STRATEGIC STUDIES<br />
AIR UNIVERSITY LIBRARY<br />
ATTN: AUL-LSE<br />
ATTN: LIBRARY<br />
ARMSTRONG LABORATORY<br />
ATTN: AL/CFBE DR F S KNOX<br />
Dist-2
DNA-TR-92-37-V1 (DL CONTINUED)<br />
ASSISTANT CHIEF OF STAFF<br />
ATTN: AFSAAISAMI<br />
ARI<br />
ATTN: JOHN LOCKHART<br />
DEPUTY CHIEF OF STAFF FOR PLANS & OPERS<br />
ATTN: AFXOOSS<br />
HO ACC/XP-JSG<br />
ACrTN: ACCXP-JSG<br />
EAI CORPORATION<br />
ATTN: MILLARD MERSHON<br />
ATTN: DENNIS METZ<br />
ATTN: KRISTIN GAVLINSKI<br />
HQ USAFA/DFSELD<br />
FULTON FINDINGS<br />
ATTN: LIBRARY<br />
ATTN: WES FULTON<br />
HORIZONS TECHNOLOGY INC<br />
NATIONAL AIR INTELLIGENCE CENTER ATTNS F GREY<br />
ATTN:<br />
CCN<br />
ATTN: SDA<br />
HORIZONS TECHNOLOGY, INC<br />
ATT'IN: J MARSHALL-MIES<br />
USAF NUCLEAR CRITERIA GROUP SERETARIAT<br />
ATTN: AFWL/NTN R SUDER<br />
INSTITUTE FOR DEFENSE ANALYSES<br />
ATTN: DOUGLAS<br />
DEPARTMENT<br />
SCHULTZ<br />
OF ENERGY<br />
ATTN: J ORLANSKY<br />
DEPARTMENT OF ENERGY<br />
ATTN: DR T JONES<br />
DPEARTMENT OF ENERGY<br />
ATTN: B SANTORO<br />
ATTN: G KERR<br />
ATTN: J WHITE<br />
ATTN: W RHOADES<br />
LAWRENCE LIVERMORE NATIONAL LAB<br />
ATTN: A WARSHAWSKY<br />
ATTN: Z DIVISION LIBRARY<br />
LOS ALAMOS NATIONAL LABORATORY<br />
ATTN: D STROTTMAN<br />
ATTN: REPORT LIBRARY<br />
SANDIA NATIONAL LABORATORIES<br />
ATTN: TECH LIB 3141<br />
OTHER GOVERNMENT<br />
CENTRAL INTELLIGENCE AGENCY<br />
ATTN: COUNTER-TERRORIST GROUP<br />
ATTN: DIRECTOR OF SECURITY<br />
ATTN: MEDICAL SERVICES<br />
ATTN: NIO-T<br />
ATTN: N10 STRATEGIC SYS<br />
FEDERAL EMERGENCY MANAGEMENT AGENCY<br />
ATTN: CIVIL SECURITY DIVISION<br />
ATTN: NP-CP<br />
ATTN: OFC OF CIVIL DEFENSE<br />
U S DEPARTMENT OF STATE<br />
ATTN: PM/STM<br />
U S NUCLEAR REGULATORY COMMISSION<br />
ATTN: DIR DIV OF SAFEGUARDS<br />
ATTN: S YANIV<br />
DEPARTMENT OF DEFENSE CONTRACTORS<br />
ARES CORP<br />
2 CY ATTN: A DEVERILL<br />
JAYCOR<br />
ATTN: M FINEBERG<br />
ATTN: CYRUS P KNOWLES<br />
KAMAN SCIENCES CORP<br />
ATTN: DASIAC<br />
KAMAN SCIENCES CORPORATION<br />
AT-IN: R STOHLER<br />
KAMAN SCIENCES CORPORATION<br />
ATTN: DASIAC<br />
LOCKHEED MISSILES & SPACE CO, INC<br />
ATTN: WE-YOUNG WOO<br />
LOGICON R & D ASSOCIATES<br />
ATTN: DOCUMENT CONTROL<br />
ATTN: DOUGLAS C YOON<br />
LOGICON R & D ASSOCIATES<br />
ATTN: S WOODFORD<br />
MASSACHUSETTS INSTITUTE OF TECHNOLOGY<br />
ATTN: DUNCAN C MILLTER<br />
MICRO ANALYSIS AND DESIGN<br />
ATTN: N LAVINE<br />
ATTN: R LAUGHERY<br />
ATTN: TROTH<br />
PACIFIC-SIERRA RESEARCH CORP<br />
2 CY ATTN: F W WHICKER<br />
2 CY ATTN: G ANNO<br />
20 CY ATTN: G E MCCLELLAN<br />
ATTN: H BRODE<br />
PACIFIC-SIERRA RESEARCH CORP<br />
ATTN: D GORMLEY<br />
2 CY ATTN: G MCCLELLAN<br />
PSYCHOLOGICAL RESEARCH CENTER<br />
ATTN: L GAMACHE<br />
Dist-3
DNA-TR-92-37-VI (DL CONTINUED)<br />
SCIENCE APPLICATIONS INTL CORP<br />
ATTN:D KAUL<br />
ATTN: E<br />
L<br />
SWiCK<br />
HUNT<br />
ATTN: R J BEYSTER<br />
AT'ITN: W WOOLSON<br />
SCIENCE APPLICATIONS INTL CORP<br />
ATTN: JMCGAHAN<br />
ATTN: J PETERS<br />
ATTN: P VERSTEEGEN<br />
ATTN: W LAYSON<br />
SCIENCE APPLICATIONS INTL CORP<br />
ATTN: R CRAVER<br />
TECHNICO SOUTHWEST INC<br />
ATTN: S LEVIN<br />
TRW INC<br />
ATTN: TIC<br />
TRW OGDEN ENGINEERING OPERATIONS<br />
ATTN: D C RICH<br />
TRWSIG<br />
ATTN- DR BRUCE WILSON<br />
UNIVERSITY OF CINCINNATI MEDICAL CENTER<br />
ATTN: E SILBERSTEIN<br />
DIRECTORY OF OTHER<br />
UNIV OF ILLINOIS-WILLIARD AIRPORT<br />
ATTN: H L TAYLOR<br />
Dist-4
<strong>Defense</strong> <strong>Nuclear</strong> Agency<br />
Alexandria, VA 22310-3398<br />
AD-A259 085<br />
S.... IiilII|IlH~l D.A-TR-92-3,,.v2<br />
<strong>Chernobyl</strong> Doses<br />
Volume 2-Conifer Stress Near <strong>Chernobyl</strong> Derived<br />
from Landsat Imagery<br />
Gene E. McClellan<br />
Terrence H. Hemmer<br />
Ronald N. DeWitt<br />
Pacific-Sierra Research Corporation<br />
12340 Santa Monica Boulevard<br />
Los Angeles, CA 90025-2587<br />
December 1992 ST EI<br />
Technical Report<br />
SELEC f<br />
S DF'C2 2 1992 II<br />
E<br />
CONTRACT No. DNA 001-87-C-0104<br />
Approved for public release;<br />
distribution ti unlimited.<br />
92-32436<br />
I!IIII4lU 9 9 .0 f.j, .• - !
Destroy this report when it is no longer needed. Do not<br />
return to sender.<br />
PLEASE NOTIFY THE DEFENSE NUCLEAR AGENCY,<br />
ATTN: CSTI, 6801 TELEGRAPH ROAD, ALEXANDRIA, VA<br />
22310-3398, IF YOUR ADDRESS IS INCORRECT, IF YOU<br />
WISH IT DELETED FROM THE DISTRIBUTION LIST, OR<br />
IF THE ADDRESSEE IS NO LONGER EMPLOYED BY YOUR<br />
ORGANIZATION.<br />
-4a Zi
-DISCLA]IIR NOTICE<br />
THIS DOCUMENT IS BEST<br />
QUALITY AVAILABLE. THE COPY<br />
FURNISHED TO DTIC CONTAINED<br />
A SIGNIFICANT NUMBER OF<br />
COLOR PAGES WHICH DO NOT<br />
"REPRODUCE<br />
LEOIBLY ON BLACK<br />
AND WHITE MICROFICHE.
DISTRIBUTION LIST UPDATE<br />
This mailer is provided to enable DNA to maintain current distribution lists for reports. (We would<br />
appreciate your providing the requested information.)<br />
NOTE:<br />
o- Add the individual listed to your distribution list. Please return the mailing label from<br />
the document so that any additions,<br />
0 Delete the cited organization/individual. changes, corrections or deletions can<br />
be made easily.<br />
O3 Change of address.<br />
NAME-<br />
ORGANIZATION:<br />
OLD ADDRESS<br />
CURRENT ADDRESS<br />
Z<br />
TELEPHONE NUMBER: ( )<br />
I-I<br />
WU<br />
w DNA PUBUCATION NUMBER/TITLE CHANGES/DELETIONSIADDITIONS, etc.)<br />
as<br />
Z<br />
,<<br />
uI<br />
UJ<br />
I-I<br />
(Atafh ShtM if mmr. Space is Required)<br />
* DNA OR OTHER GOVERNMENT CONTRACT NUMBER:<br />
CERTIFICATION OF NEED-TO-KNOW BY GOVERNMENT SPONSOR (if other than DNA):<br />
SPONSORING ORGANIZATION:<br />
CONTRACTING OFFICER OR REPRESENTATIVE:<br />
SIGNATURE:
DEFENSE NUCLEAR AGENCY<br />
ATTN: TITL<br />
6801 TELEGRAPH ROAD<br />
ALEXANDRIA, VA 22310-3398<br />
DEFENSE NUCLEAR AGENCY<br />
ATTtN- TITL<br />
6801 TELEGRAPH ROAD<br />
ALEXANDRIA, VA 22310-3398
REPORT DOCUMENTATION PAGE<br />
OBN.00-I<br />
Public reporting burden tor Vthe colection of informaton is estmatead to average I hour per response Including the tuime fr reviewing instructions. searching existing dao sources.<br />
gatlherng and maintaining the data needed, and completing and reviw-ng the collection of information Send comments regarding this burden estimate or any other aspect of thIs<br />
colectlion of formabon. inctuding suggestions for reducing this burden. to Washington HeedQuarters Services Directorate for intormation O oeration5 and <strong>Reports</strong>. 1215 Jefferson<br />
Davis Highway, Suite 1204. Arlington. VA 22202-4302. and to the Office at Management and Budget. Paperwort Reduction Protect (0704-0188}. Washington. DC 20503<br />
1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED<br />
1 921201 Technical 870929 - 920228<br />
4. TITLE AND SUBTITLE 5. FUNDING NUMBERS<br />
<strong>Chernobyl</strong> Doses C - DNA 001-87-C-0104<br />
Volume 2- Conifer Stress Near <strong>Chernobyl</strong> Derived from Landsat Imagery PE - 62715H<br />
6. AUTHOR(S) PR - RM<br />
Gene E. McClellan, Terrence H. Hemmer, and Ronald N. DeWitt<br />
TA -RH<br />
WU - DH026130<br />
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION<br />
REPORT NUMBER<br />
Pacific-Sierra Research Corporation<br />
12340 Santa Monica Boulevard PSR Report 2250<br />
Los Angeles, CA 90025-2587<br />
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING<br />
AGENCY REPORT NUMBER<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency<br />
6801 Telegraph Road DNA-TR-92-37-V2<br />
Alexandria, VA 22310-3398<br />
RARP/Young<br />
11. SUPPLEMENTARY NOTES<br />
This work was sponsored by the <strong>Defense</strong> <strong>Nuclear</strong> Agency under RDT&E RMC Code B4662D RM RH 00038<br />
STRP 3500A 25904D.<br />
12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE<br />
Approved for public release; distribution is unlimited.<br />
13. ABSTRACT (Maximum 200 words)<br />
This volume presents Landsat Thematic Mapper imagery of the area surrounding the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor<br />
Station and derives quantitative estimates of the spatial extent and time progression of stress on coniferous forests<br />
resulting from the 26 April 1986 reactor explosion and release of radioactive material. Change detection between<br />
pre- and postaccident images demonstrates convincingly that remote sensing of the spectral reflectance of<br />
coniferous forests in visible and infrared wavelengths at moderate spatial resolution (30 meters) will detect the<br />
effects of large radiation doses to the forest canopy.<br />
This work was initiated at a time when the expectation for direct data from the Soviet Union on local, accidentinduced<br />
radiation levels was limited and the satellite data provided an alternative source. Although information<br />
exchange with the former Soviet Union has improved dramatically, the results of this report are important, since<br />
they prove the feasibility of large-scale, spectral response measurements on radiation-exposed pine trees in a natural<br />
environment. Volume I presents the derivation of radiation doses from the imagery reviewed in this volume,<br />
describes changes in spectral reflectivity of the affected trees as a function of dose and time, and discusses the military<br />
operational implications of these results.<br />
14. SUBJECT TERMS 15. NUMBER OF PAGES<br />
<strong>Chernobyl</strong> Forest Damage Landsat 80<br />
Change Detection Conifer Stress Fallout 16. PRICE CODE<br />
Ionizing Radiation Multispectral Imagery<br />
17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19, SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT<br />
OF REPORT OF THIS PAGE OF ABSTRACT<br />
UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED SAR<br />
NSN 7540-20-5=<br />
Standrcd Form 296 (Rev.2-9)<br />
i Preecnbed by, ANSI SM. 239-18<br />
255- 02
UNCLASSIFIED<br />
SECURITY CLASSIFICATION OF THIS PA(rE<br />
CLASSIFIED BY:<br />
N/A since Unclassified.<br />
DECLASSIFY ON:<br />
N/A since Unclassified.<br />
UN•cLASSIFIED
PREFACE<br />
This volume is the second in a series of three volumes composing the final report to the<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency (DNA) for contract DNAOO1-87-C-0104, <strong>Chernobyl</strong> Doses. This<br />
document was prepared by investigators at Pacific-Sierra Research Corporation (PSR) as a topical<br />
report for that contract but is being published as a volume of the final report. It decribes the<br />
acquisition and processing of Landsat imagery of the area containing the <strong>Chernobyl</strong> <strong>Nuclear</strong><br />
Reactor Station and presents the exploratory analysis of the imagery using PSR's proprietary<br />
Hyperscoutr' change detection algorithm. Volume 1, Analysis of Forest Canopy Radiation<br />
Response from Multispectral Imagery and the Relationship to Doses, presents the analytical work<br />
that connects these multispectral observations of pine forests in the images to the nuclear radiation<br />
dose received by the treas as a consequence of the reactor accident of 26 April 1986. Volume 3,<br />
Habitat and Vegetation Near the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station, presents a detailed exposition<br />
on the soil, climate, and vegetation of the Poles'ye region of the Ukraine and Belorussia with<br />
emphasis on the area around <strong>Chernobyl</strong>.<br />
The authors wish to acknowledge Frank Thomas and George Anno of PSR, who recognized<br />
the potential for remote sensing of radiation-damaged foliage around <strong>Chernobyl</strong>; Wayne Hallada<br />
and the late Quentin Wilkes of PSR, who arranged the necessary equipment and image<br />
acquisitions; and finally, the skillful manuscript preparation by Kathy Howell and Sunny Wiard.<br />
The authors wish to acknowledge the technical monitor of this project, Robert W. Young of<br />
DNA's Radiation Policy Division, for his support and encouragement during this work.<br />
Dr. Young was assisted first by MAJ Bruce West and then by MAJ Robert Kehlet. The authors<br />
also wish to acknowledge Dr. Marvin Atkins and Dr. David Auton of DNA whose interest made<br />
this work possible.<br />
DTIC QUALITY 1 i-Y'YD 2<br />
Accesion For<br />
NTIS CRA&I<br />
DTIC TAB<br />
Unannounced [<br />
Justification<br />
By .........................................<br />
Distribution I<br />
Dist<br />
Availability Codes<br />
IAvail and/or<br />
Special<br />
ThEHypersou is a regisweid trdemark of Pacific-Sierra Reserc Caupmuon.<br />
SI . ,
CONVERSION TABLE<br />
Conversion factors for U.S. customary to metric (SI) units of measurement<br />
To Convert From To MulUply<br />
angstom meters (m) 1.000 000 X K-10<br />
atmosphere (normal) kilo pascal (kPa) 1.013 25 X E+2<br />
bar kilo pascal (kPa) 1.000 000 X E+2<br />
barn meter 2 (M 2 ) 1.000 000 X E-28<br />
British Thermal unit ( emicl) joule (JW 1.054 350 X E+3<br />
calorie (thermochemicsl) joule (JW 4.184 000<br />
cal (thermoWhe cal)/,ml mep joule//m2(MJ/m2) 4.184 000 X E-2<br />
curie g"ig becquerel (GBqr 3.700 000 X E+I<br />
degree (angle) radian (rad) 1.745 329 X E-2<br />
degmre Fahrenheit degree kelvin (K) tK=(tof + 459.67)11.8<br />
electron volt joule (J) 1.602 19 X E-19<br />
erg Joule (J) 1.000 000 X E-7<br />
erg/second watt MW) 1.000 000 X E-7<br />
foot meter (m) 3.048 000 X E-I<br />
foot-pound-force joule (J) 1.355 818<br />
gallon (U.S. liquid) meter 3 (m3) 3.785 412 X E-3<br />
inch meter (m) 2.540 000 X E-2<br />
jerk joule (J) 1.000 000 X E+9<br />
joule/kilogram tJ/Kg) (radiation dose<br />
absorbed) Gray (Gy) 1.000 000<br />
kilotons terajoules 4.163<br />
kip (1000 Ibf) newton (N) 4.448 222 X E+3<br />
kip/tnch2 (ksl) kilo pascal (kPa) 6.894 757 X E+3<br />
ktap newton-second/mr (N-aim2) 1.000 000 X E+2<br />
micron meter (mW 1.000 000 X E-6<br />
mil meter (mW 2.540 000 X E-5<br />
mile (international) meter (m) 1.609 344 X E+3<br />
ounce kilogram (kg) 2.834 952 X E-2<br />
pound-force (lbf avoirdupois) newton (N) 4.448 222<br />
pound-force inch newton-meter (N-m) 1.129 848 X E-I<br />
pound-force/Inch newton/meter (N/m) 1.751 268 X E+2<br />
pound-force/foot' kilo pascal (kPa) 4.768 026 X E-2<br />
pound-force/Inch2 (psi) kilo pascal (kPa) 6.894 757<br />
pound-mass (Ibm avoirdupois) kilogram (kg) 4.535 924 X E-I<br />
pound-mass-foot2 (moment of inertia) kilogram-meter' (1g.m') 4.214 01! X E-2<br />
pound-mass/foot3 kilogram/meter' (kg/lnm) 1.601 846 X E+ I<br />
red (radiation dose absorbed) Gray (Gy)" 1.000 000 X E-2<br />
roentgen coulomb/kilogram (C/kg) 2.579 760 X E-4<br />
shake second Is) 1.000 000 X E-8<br />
slug kilogram (kg) 1.459 390 X E+1<br />
torn Imm 14ft OC) kilo pascal (kPa| 1.333 22 X E-I<br />
"The becquerel (Dq) Is the St unit of radioactivity: Bp a I event/s.<br />
"•The Gray (Gy) ti the S8 unit of absorbed radiation.<br />
iv
TABLE OF CONTENTS<br />
Section<br />
Page<br />
PREFACE ................................................................................ iii<br />
CONVERSION TABLE ............................................................... iv<br />
FIGURES ............................................................................... vi<br />
TABLES .................................................................................. vii<br />
1 INTRODUCTION ......................................................................<br />
1.1 Background ....................................................................<br />
1<br />
1<br />
1.2 Organization of report ....................................................... 2<br />
2 THE PHYSICAL BASIS FOR THE SPECTRAL DETECTION OF<br />
VEGETATION STRESS ............................................................. 3<br />
3 QUANTITATIVE ANALYSIS OF MULTIDATE IMAGERY .................... 6<br />
3.1 Factors affecting the utility of multidate imagery .......................... 6<br />
3.2 Selection of an algorithm ..................................................... 7<br />
4 THEMATIC MAPPER IMAGERY ................................................. 9<br />
4.1 Introduction to the thematic mapper ..................................... 9<br />
4.2 Motivation for use of themaic mapper imagery ........................... 11<br />
4.3 Image selection and procurement ........................................ 11<br />
4.4 Image preparation ............................................................ 16<br />
5 IMAGE ANALYSIS ........................................ 19<br />
5.1 Image partitioning ............................................................ 19<br />
5.2 Mapping stress ................................................................ 25<br />
6 GROUND TRUTH ................................................................... 53<br />
7 DISCUSSION .......................................................................... 58<br />
8 CONCLUSION ........................................................................ 64<br />
9 REFERENCES ........................................................................ 66
FIGURES<br />
Figure<br />
Page<br />
4-1 Vicinity of the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station on 8 May 1986 .......... 18<br />
5-1 Partitioning of 6 June 1985 image into classes derived from training<br />
sites 3 (yellow) and 4 (green) .................................................... 24<br />
5-2 Date: 6 June 1985, 1 year preaccident ....................................... 27<br />
5-3 Date: 21 March 1986, 2 months preaccident ............................... 28<br />
5-4 Date: 29 April 1986, 3 days postaccident .................................. 29<br />
5-5 Date: 29 April 1986, 3 days postaccident, forest stress map, colored<br />
according to indicated scale. Gray-scale background consists of<br />
nonforested areas ................................................................. 31<br />
5-6 Date: 8 May 1986, 12 days postaccident ................................... 32<br />
5-7 Date: 8 May 1986, 12 days postaccident, forest stress map .............. 33<br />
5-8 Date: 24 May 1986, 4 weeks postaccident ................................. 35<br />
5-9 Date: 24 May 1986, 4 weeks postaccident, forest stress map ............ 36<br />
5-10 Date: 31 May 1986, 5 weeks postaccident ................................... 37<br />
5-11 Date: 31 May 1986, 5 weeks postaccident, forest stress map ............ 38<br />
5-12 Date: 15 October 1986, 5.6 months postaccident ......................... 39<br />
5-13 Date: 15 October 1986, 5.6 months postaccident, forest stress map ...... 40<br />
5-14 Date: 2 December 1986, 7.2 months postaccident ........................ 41<br />
5-15 Date: 2 December 1986, 7.2 months postaccident, forest stress map..... 42<br />
5-16 Date: 11 May 1987, 1 year postaccident .................................... 44<br />
5-17 Soviet-supplied gamma dose rate contours of 1 May 1987 overlaid<br />
on the I IM ay 1987 ................................................................ 45<br />
5-18a Date: 11 May 1987, 1 year postaccident, forest stress map .............. 46<br />
5-18b Date: 11 May 1987, forest stress map with gamma dose rate contours<br />
of 1 M ay 1987 ..................................................................... 47<br />
5-19 Date: 7 September 1987, 16 months postaccident ........................... 48<br />
5-20 Date: 7 September 1986, 16 months postaccident, forest stress map. 49<br />
5-21 Date: 28 May 1988, 25 months postaccident ............................. 51<br />
5-22 Date: 28 May 1988, 25 months postaccident, forest stress map ......... 52<br />
7-1 Date: 11 May 1987, enhanced image ....................................... 61<br />
7-2 Boundaries of areas classified as similar to training site 4 overlaid on the<br />
11 May 1987 enhanced image ................................................... 62<br />
7-3 Date: 6 June 1985, preaccident image with the same image<br />
enhancement transformation shown in Figure 7-1 ............................. 63<br />
vi
TABLES<br />
Table<br />
Page<br />
2-1 Estimated short-term radiation exposures required to damage<br />
various plant communities .......................................................... 3<br />
4-1 Landsat TM spectral bands ........................................................ 9<br />
4-2 Landsat scene quality and cloud contamination analysis for scenes<br />
with less than 30 percent cloud cover ........................................ 12<br />
4-3 Scene procurement recommendations ........................................... 14<br />
4-4 Landsat scene acquisitions ...................................................... 15<br />
4-5 Corner points of the acquired images ........................................ 15<br />
4-6 Corner points of the geocoded images after resampling ...................... 16<br />
4-7 Corner points (Zone 36 UTM coordinates) of the 512 x 512 pixel area<br />
analyzed in this report ............................................................ 17<br />
5-1 Training site modified transformed divergencies .............................. 23<br />
6-1 Second order warp coefficients for the gamma dose contours ............... 54<br />
6-2 Accuracy of the registration of GCPs ........................................... 55<br />
7-1 Comparison of present work with that of Goldman and coworkers ........... 59<br />
vii
SECTION 1<br />
INTRODUCTION<br />
On 26 April 1986, at 1:24 a.m. local time, the <strong>Chernobyl</strong> nuclear reactor number four blew up.<br />
This report describes the research of investigators at Pacific-Sierra Research Corporation (PSR) in<br />
studying the effects of radioactivity deposited in the immediate area by the accident. Specifically, it<br />
describes the quantitative analysis of the multispectral (MS) imagery of the area surrounding the<br />
<strong>Chernobyl</strong> nuclear reactor.<br />
1.1 BACKGROUND.<br />
Before the explosion only theoretical models suggested our ability to monitor remotely the<br />
effects caused by widely distributed large doses of radiation. The sensitivity of vegetation to large<br />
doses of radiation had been measured only in the lab and in small-scale field experiments. The<br />
<strong>Chernobyl</strong> accident provides an opportunity to observe both short- and long-term effects of high<br />
radiation dose on plants. The large spatial scale of the affected area allows us to evaluate the utility<br />
of remotely sensed multispectral imagery in quantifying the extent of damage to foliage and in<br />
estimating the radiation dose that was deposited by the accident.<br />
The accident occurred before the changes of Glasnost could be taken for granted. Indeed,<br />
because the initial tendency of Soviet authorities was to deny the accident, there was little basis for<br />
anticipating sufficient and accurate information from Soviet sources. Therefore, any independent<br />
source of information that could augment or verify the information released by Soviet authorities<br />
seemed desirable. In particular, some method for obtaining an independent assessment of dose<br />
was regarded to be of primary interest<br />
The most obvious method was to use remote sensors. Because overflights of Soviet territory<br />
were deemed impossible, the remote sensor would have to be in a low-altitude orbit. The remote<br />
measurement of radiation at this distance is prevented by atmospheric attenuation. On the other<br />
hand, the remote measurement of the effects of the radiation on vegetation was possible with the<br />
existing satellite sensor mix. Two of these sensors were the Landsat Thematic Mapper (TM) and<br />
SPOT maintained by the United States and France, respectively. Because both of these are<br />
commercial systems, data is readily available. In effect, vegetation provides an on-site biological<br />
dosimeter that can be read remotely with multispectral imagery.<br />
Such remote sensing of vegetation stress is an important scientific endeavor contributing to<br />
early detection of the effects of major disasters such as <strong>Chernobyl</strong> as well as the chronic effects of<br />
major pollution sources. Analysis of imagery of the <strong>Chernobyl</strong> accident provides a benchmark for<br />
this capability.
1.2 ORGANIZATION OF REPORT.<br />
Following this introduction, Section 2 presents the physical basis for the spectral detection of<br />
vegetation stress. Section 3 describes the usual approaches to the detection of stress. Section 4<br />
follows with a description of the sensor, imagery selection criteria, and preprocessing<br />
requirements. Section 5 presents the imagery, including detected indications of vegetation stress.<br />
Since limited ground truth (radiation measurements) was published by the Soviets, this ground<br />
truth can be compared to the detected stress. Section 6 presents this comparison. Section 7<br />
reiterates salient features and compares this work with other similar work. Section 8 contains<br />
concluding remarks and makes recommendations for continuation and improvements.<br />
2
SECTION 2<br />
THE PHYSICAL BASIS FOR THE SPECTRAL DETECTION OF<br />
VEGETATION STRESS<br />
We have asserted that dose estimates for the <strong>Chernobyl</strong> area can be obtained by monitoring the<br />
effects of the radiation on indigenous vegetation. In making this assertion, we assume that the<br />
accident sufficiently dosed the vegetation to invoke a response detectable in the spectral regions<br />
monitored by satellite sensors. This section provides the justification for these assumptions.<br />
We simplify the analysis by concentrating on only one type of plant community. Three criteria<br />
governed the selection of the type of plant community to be monitored. First, our desire to<br />
determine the radiation dose dictated a plant type relatively sensitive to radiation. Second, the<br />
selection of a plant species pervading the area of interest allows estimates of the areal extent of the<br />
radiation effects. Third, because the time scale of the manifestation of stress could not be<br />
anticipated, we desired a plant type subject to little seasonal variation.<br />
Coniferous forests meet all three requirements. Table 2-1 shows that coniferous forests are<br />
very sensitive to radiation. The extensive forests located throughout the area around the <strong>Chernobyl</strong><br />
reactor complex comprise predominantly conifers, with common pine, Pinus sylvestris (see<br />
Volume Ill of this reportl) the primary species. Finally, conifers show relatively little seasonal<br />
spectral variation compared to deciduous trees. However, seasonal changes in illumination and<br />
variations in the canopy closure of intermixed deciduous species, as well as changes caused by the<br />
maturing of the predominant foliage, vary the observed reflectance spectrum somewhat from<br />
season to season.<br />
Table 2-1. Estimated short-term radiation exposures required to damage various plant communities<br />
(Whicker and Fraley, 1974).<br />
Exposures to produce (kR)<br />
Community type Minor effects Intermediate Severe effects<br />
effects<br />
Coniferous forest 0.1-1 1-2 >2<br />
Deciduous forest 1-5 5-10 >10<br />
Shrub 1-5 5-10 >10<br />
Tropical rain forest 4-10 10-40 >40<br />
Grassland 8-10 10-100 >100<br />
Moss-lichen 10-50 50-500 >500<br />
3
Before presenting the spectral manifestations of stress, we discuss the spectral characteristics<br />
of healthy vegetation, with emphasis on those areas of the spectrum that can be monitored by either<br />
Landsat or SPOT. In the visible region of the electromagnetic spectrum extending from<br />
wavelengths of 0.4 to 0.7 micrometers (Wim), chlorophyll absorption (Hoffer, 1978) dominates the<br />
reflectance spectra of healthy vegetation. Chlorophyll absorbs throughout this region, but it<br />
absorbs less in the green than in the blue or red. This preferential absorption causes the green<br />
color of healthy vegetation.<br />
The infrared (IR) region of the spectrum comprises several subregions: the near infrared (NIR);<br />
short-wave infrared (SWIR); mid-wave IR (MWIR); and thermal or long-wave IR (LWIR). The<br />
wavelength interval defining each of these spectral regions varies from author to author. This<br />
report defines the NIR region to extend from about 0.7 jim to 1.3 Aim. In this subregion, the<br />
spongy mesophyll tissue in the interior of the leaves causes the foliage to reflect strongly, typically<br />
as much as 45 to 50 percent of incident illumination. The SWIR spectral region, from<br />
approximately 1.5 jim through 2.5 jim, includes two absorption features of the leaf water content,<br />
reducing the foliage spectral reflectance. Neither Landsat nor SPOT are sensitive in the MWIR,<br />
which extends from about 3 to 5 jim. The LWIR includes electromagnetic radiation from 8 jim to<br />
12 am and beyond. Because at terrestrial temperatures most materials have spectral characteristics<br />
in the LWIR that tend to be emissive rather than reflective, this band is referred to as the thermal<br />
MR. Thus, in this region the temperature and the emissivity of the vegetation determines its spectral<br />
response. Under normal conditions, transpiration and efficient heat exchange keep the<br />
temperature of foliage close to air temperature (Estes, 1983; Weibelt and Henderson, 1976).<br />
The spectral manifestations of stress depend strongly on the cause of the stress (Estes, 1983).<br />
For example, moisture stress initially results in increased reflectance in the SWIR, because there is<br />
less water to absorb the infrared radiation (Hoffer, 1978). Desiccation of pine needles to<br />
approximately 48 percent of fresh weight reduces NIR reflectance dramatically, while SWIR<br />
reflectance continues to rise (Westman and Price, 1988). An expected lack of cooling by<br />
transpiration would allow stressed foliage to be warmed by sunlight to temperatures higher than<br />
those of healthy vegetation and would result in relatively higher emissions in the thermal HR. If the<br />
moisture stress impedes chlorophyll production, the reflectivity in the visible can also be expected<br />
to increase. The spectral properties of other pigments in the leaf may now become dominant,<br />
resulting in a yellowish or reddish color.<br />
Cellular damage results in a marked decrease in the reflectance in the NIR. Little or no change<br />
in the visible is expected unless the damage reduces chlorophyll content or causes the production or<br />
destruction of other pigments.<br />
Unfortunately, the specific spectral manifestations of ionizing radiation damage to foliage are<br />
not well known. However, one form of radiation-induced stress may be at least partially<br />
4
predictable. The spectral reflectance of new pine needle growth is significantly higher than that of<br />
old pine needle growth at all wavelengths (Wolfe and Zissis, 1978). If the radiation stress results<br />
in either reduced or accelerated growth of new needles (an effect expected to be highly dependent<br />
on dose), then this deviation may be detectable.
SECTION 3<br />
QUANTITATIVE ANALYSIS OF MULTIDATE IMAGERY<br />
The problem of identifying stress in imagery can be approached in two ways. The first<br />
approach assumes that stressed vegetation has unique spectral signatures, and that these spectral<br />
signatures are sufficiently different from the spectral signatures of healthy vegetation that a suitable<br />
spectral transformation can be applied to the image to render stressed vegetation readily apparent to<br />
the image analyst. For example, Johnson (Johnson, 1989) developed a phenomenologically based<br />
image-enhancement transformation designed to detect a specific spectral manifestation of stress.<br />
Section 7 discusses this transformation in more detail.<br />
The other approach to identifying stress in imagery attempts to exploit the additional<br />
information content of multidate imagery. Multidate algorithms generally fall into two major<br />
categories: change detection algorithms and dynamic-system algorithms. Change detection<br />
algorithms tend to ignore explicit spectral properties and concentrate only on differences from date<br />
to date, except insofar as the spectral properties are associated with particular kinds of ground<br />
covers that manifest the change. Spectral properties are occasionally used, but only at a later stage<br />
to characterize the detected change. Dynamic-system algorithms tend to retain most of the spectral<br />
information.<br />
3.1 FACTORS AFFECTING THE UTILITY OF MULTIDATE IMAGERY.<br />
An accurate assessment of the fallout from the <strong>Chernobyl</strong> nuclear reactor accident through its<br />
effect on coniferous forests can be made only if a number of rather stringent requirements are met.<br />
First, there must be a sensor capable of detecting and tracking the changes in the spectral properties<br />
of the vegetation. Second, all unrelated factors that affect the spectral properties of vegetation on<br />
an image-to-image basis must be normalized out so that direct multidate comparisons can be made.<br />
Third, the data must be very accurately registered. Fourth, data must be available at frequent<br />
enough intervals.<br />
Numerous extraneous factors can affect the comparison of two or more images. Some of these<br />
factors are environmental (e.g., haze, humidity, sun angle, and cloud cover). Others are<br />
instrumental (e.g., view angle, pointing accuracy, spectral sensitivities, and resolution). Even if<br />
the same sensor at the same relative position at the same time of day is used, the differences<br />
between images can be quite large. Either the images must be corrected for these differences, or an<br />
algorithm that is insensitive to these extraneous differences must be used.<br />
Because multidate algorithms use the spectral information for the same ground point on two or<br />
more dates, the images must be registered (geocoded). Geocoding processes the images from all<br />
6
dates so that the picture element (pixel) at the same position in each date's image corresponds to a<br />
common area on the ground. Registration accuracy is of fundamental importance in the analysis of<br />
multidate imagery. Variations in the satellite orbit, pointing accuracy, jitter, and numerous other<br />
reasons make exact registration impossible, but registration to within a fraction of a pixel width is<br />
often possible.<br />
Finally, multidate image analysis requires that images be collected at time intervals that are<br />
small compared to the time scale of the process being monitored. In the case of radiation stress,<br />
the local radiation dose determines that time scale and causes it to vary from one place to another<br />
according to the distribution of the radiation fallout. This variation imposes a requirement for<br />
frequent images to monitor the development of stress in higlidy contaminated areas, but less<br />
frequent imaging will suffice in the regions of lesser contamination. Because cloud cover often<br />
renders satellite data useless, data collections at time intervals smaller than the time scale of the<br />
stress development process must be attempted.<br />
3.2 SELECTION OF AN ALGORITHM.<br />
The approach first envisioned for this project involved tracking the spectral properties of large<br />
areas of forest as a function of time. This approach falls in the dynamic-system category. Large<br />
areas of the forest would be delineated by polygons for the analysis. We can fit multivariate<br />
normal distributions to the measurements taken on each date within each polygon in the following<br />
manner. For a given ground pixel, there is an intensity value measured in each band, and these<br />
intensities can be formed into an ordered set called the pixel vector. A polygon's mean ii tensity<br />
vector can be comlited by averaging the pixel vectors over all pixels within the polygon. Once the<br />
polygon's mean intensity vector has been determined, there can be computed for each pixel a<br />
deviation vector (representing a pixel intensity vector's departure from the mean vector) by simply<br />
subtracting the polygon's mean intensity vector from the pixel vector. The polygon's covariance<br />
matrix can then be computed by forming a matrix comprising all pairings of the intensities for each<br />
pixel deviation vector and then averaging these matrices over all pixels in the polygon (Swain,<br />
1978). The resulting mean intensity vector and intensity covariance matrix suffice to characterize a<br />
multivariate normal distribution for that particular polygon for that particular day. This process<br />
would be performed for all polygons for all images.<br />
Next, a reference image (date) would be chosen, and quantitative methods would be used to<br />
identify the spectral deviations of each forested area from the spectral properties of the same<br />
forested area on the reference image. This approach requires the removal from the multidate<br />
images of any variations that are not directly related to stress. Many of these variations can be<br />
removed by rescaling the date-of-interest image (the image for which the stress is to be calculated)<br />
to the reference image. This rescaling can significantly reduce the effects of seasonal variations in<br />
7
the incidence angle of solar illumination. If this rescaling is done on a band-by-band basis,<br />
rescaling can also reduce the effects of those variations in the atmospheric condition that are<br />
uniform over the entire area of interest. Rescaling cannot reduce the effects of variation in cirrus<br />
cloud density or other local atmospheric conditions.<br />
To perform the rescaling, we selected several forest "training sites." A forest "training site" is<br />
a forested area that is assumed to be spectrally identical, before solar and atmospheric variations,<br />
on the two dates. Statistics were computed for each of the training sites. Unfortunately, the scale<br />
factors that were computed from the mean vectors were different for different training sites. This<br />
suggests that there were significant spectral variations across the image that were not related to<br />
stress and could not be normalized out. Further, rescaling by a multiplicative constant<br />
concomitantly requires the rescaling of the covariance matrix elements by the square of the<br />
corresponding multiplicative constant for diagonal elements and by the products of multiplicative<br />
constants for the off-diagonal elements. The resulting rescaled covariance matrices did not match<br />
the reference date covariance matrices. Because these difficulties would adversely affect the<br />
sensitivity, and hence the usefulness of the ensuing analysis, this approach was abandoned.<br />
Fortunately, in the interim PSR developed in an independent research and development (IR&D)<br />
project, an extremely sensitive method of stress detection, the HyperscoutT' algorithim. Because<br />
of its sensitivity, this algorithm was used in the quantitative portion of the study. In addition,<br />
several procedures for identifying the spectral signatures that correspond to stressed vegetation<br />
were developed.<br />
TmHypwcm is a rgisred tradeark of Pacific-Siena Reanwh Cpocradon.
SECTION 4<br />
THEMATIC MAPPER IMAGERY<br />
4.1. INTRODUCTION TO THE THEMATIC MAPPER.<br />
The Landsat Thematic Mapper (TM) sensor collects data in seven bands in 4 regions of the<br />
electromagnetic spectrum (see Table 4-1): three visible; one near infrared (NIR); two short-wave<br />
infrared (SWIR); and one thermal infrared (Engel, 1984). That is, seven images were obtained:<br />
one through each of the seven spectral band filters. Each region of the spectrum supplies<br />
information on different manifestations and levels of stress as discussed in Section 2.<br />
A comment on the numbering of the TM bands may help prevent confusion. As discussed<br />
above and displayed in Table 4-1, the numbering of the bands is not strictly in order of increasing<br />
wavelength. 1 Specifically, band 6 is out of sequence. In order of increasing wavelength, the<br />
bands are: 1 through 5, 7, and then 6.<br />
Table 4-1. Landsat TM spectral bands.<br />
Band<br />
number<br />
Spectral<br />
region<br />
Wavelength band IFOV*<br />
(microns) (meters) Stress sensitivities<br />
1 Visible 0.45-0.52 30 Pigmentation changes<br />
(Especially chlorophyll)<br />
2 Visible 0.52-0.60 30<br />
3 Visible 0.63-0.69 30<br />
4 NIR 0.76-0.90 30 Plant structure damage<br />
5 SWIR 1.55-1.75 30 Moisture stress<br />
7 SWIR 2.08-2.35 30<br />
6 Thermal 10.40-12.50 120 Plant heat stress<br />
*Instantaneous field of view<br />
1 Thematic Mapper Simulator (TMS) data, on the other band, is sequential (ie., TMS bands 6 and 7 are reversed<br />
from TM bands 6 and 7). No further reference to TMS data is made in this study. Band numbers will always refer to<br />
"TIM band natnbers and, theefore, will be out of wavelength sequence.<br />
9
A TM image is built up by scanning a sensor that has an instantaneous field of view (IFOV)<br />
that when projected on the ground measures 30 meters (m) in bands 1 through 5 and 7, and 120 m<br />
in band 6. During processing, all bands are usually resampled to 28.5 m. Geocoded products,<br />
such as those used in this study, are resampled to 25 m.<br />
A frequently encountered misconception is that because the smallest resolvable object can be no<br />
smaller than a pixel, no information on a scale smaller than a pixel can be extracted. In fact,<br />
information from objects much smaller than the pixel dimensions can sometimes be extracted. One<br />
dramatic example of this was the identification of natural gas flarings in the NOAA-6 Advanced<br />
Very High Resolution Radiometer (AVHRR) data (Matson and Dozier, 1981). The pixel size of<br />
the NOAA-6 AVHRR sensor was 1.1 kilometer (kin) on a side at nadir, while the dimensions of<br />
natural gas flarings are clearly only a very small fraction of that size. Further, for AVHRR pixels<br />
only partially contaminated by cloud, it was possible to determine both the fraction of pixel<br />
occupied by the cloud and the brightness temperature of the cloud. How is this possible? In the<br />
case of the natural gas flarings, the flaring was many orders of magnitude brighter than the<br />
background. Thus, the overall brightness of a pixel was affected although the pixel was orders of<br />
magnitude larger than the flarings. In the case of cloud contamination, the key is that the data are<br />
multispectral. Because there were two "thermal" bands (one MWIR and one thermal) on the<br />
AVHRR instrument, two intensity levels were recorded for each pixel. In some cases these<br />
intensity levels may be used to extract information on a subpixel scale. The more bands collected,<br />
the more subpixel information that can be extracted.<br />
The Landsat 4 and 5 satellites are in 705-kmn sun-synchronous orbits (Irons, 1985). The<br />
descending node equatorial crossing time (local time) is 9:45 a.m. Thus, all Landsat images are<br />
acquired in the morning at about the same local time. This minimizes the scene-to-scene image<br />
variations caused by the solar illumination angle.<br />
The TM sensor collects a 185-kmi wide swath centered at nadir. To completely cover the<br />
Earth's surface, 233 orbits are required. Because the orbit period is 98.9 minutes, 16 days are<br />
required to complete the 233 orbits. Thus, neglecting overlap, each area on the ground can be<br />
imaged no more frequently than once every 16 days. This 16-day repeat cycle of the TM is not<br />
frequent enough to monitor the areas receiving the highest dose, especially when allowance was<br />
made for cloud cover. Fortunately, because the image swaths overlap, a TM image of the<br />
<strong>Chernobyl</strong> area could be acquired every 7 or 9 days. The very high probability of cloud cover is<br />
still a problem for which there is no solution. For areas receiving lower doses, the 7- or 9-day<br />
repeat cycle is probably frequent enough to monitor radiation stress.<br />
10
4.2 MOTIVATION FOR USE OF THEMATIC MAPPER IMAGERY.<br />
Early detection and continuous monitoring of the coniferous tree stress was critical to the<br />
accuracy of the exposure estimates. Thus, the sensor needs to be sensitive to those regions of the<br />
spectrum that most clearly exhibit the effects of radiation stress. Unfortunately, the spectral<br />
manifestations of radiation stress were uncertain. This uncertainty was the primary motivation for<br />
the use of TM data; TM collects data in four regions of the spectrum, while SPOT collects data in<br />
only two regions of the spectrum. Thus, even though SPOT has a much higher spatial resolution<br />
(10-m panchromatic, 20-m multispectral), it does not have the spectral range of TM.<br />
PSR's Hyperscout change detection algorithm is theoretically independent of the sensor,<br />
provided that any sensor that is used collects data in the appropriate spectral regions.<br />
4.3 IMAGE SELECTION AND PROCUREMENT.<br />
To select images for analysis, we first obtained a list of all TM acquisitions from EOSAT<br />
Corporation. Second, to avoid purchasing a large number of images that were not usable because<br />
of cloud cover, we identified those images collected in the area around <strong>Chernobyl</strong> that were cloud<br />
free. The simplest approach to identifying cloud free images would have been to use the automatic<br />
cloud cover assessment on the Landsat acquisition listing. Unfortunately, this number represents<br />
the cloud cover percentage over the entire scene (an area which is 100 nautical miles (nmi) on a<br />
side) and not the cloud cover over the immediate area of interest. Even if the cloud cover rating for<br />
the scene is 80 percent or more, it is still possible that the area of interest is cloud free.<br />
Conversely, a cloud cover rating of 10 percent did not guarantee that the area of <strong>Chernobyl</strong> was<br />
cloud free. Thus, a more localized analysis was desirable. For this study, cloud contamination<br />
was assessed in two ways. First, Multispectral Scanner data, which are recorded simultaneously<br />
with TM data, were reviewed at the (MSS) microfilm library located at EOSAT Corporation's<br />
headquarters in Lanham, Maryland, and observations on image quality and cloud cover<br />
contamination were recorded (see Table 4-2). Next, black and while (B&W) prints were obtained<br />
as indicated in Table 4-2 for the more interesting dates and studied in detail. Unfortunately, MSS<br />
data were not recorded in the spring of 1988, and there is a long waiting time for TM prints. For<br />
these reasons the procurement decisions for the 1988 data were based solely on the automatic cloud<br />
cover assessments.<br />
11
Table 4-2.<br />
Landsat scene quality and cloud contamination analysis for scenes with less<br />
than 30 percent cloud cover (see notes following table).<br />
Date Landat PathWrow Scene cloud B&W print Comment<br />
sensors cover (%) ordered<br />
06/29/88 TM-4 182/24 20 No Not processed to film<br />
* 05/28/88 TM-4 182/24 0 No Not processed to film<br />
01105/88 TM-4 182/24 20 No Not processed to film<br />
10/18/87 TM & MSS 181/24 0 No Not reviewed<br />
10/02/87 TM 181/24 10 No Not processed to film<br />
09/23/87 TM & MS 182/24 30 Yes Not processed to film<br />
* 09107/87 TM & MSS 182/24 20 Yes Clouds in left quarter,<br />
good full scene<br />
08/22/87 IM 182/24 30 No Not processed to fldm<br />
07/22/87 TM-4 181/24 10 Yes 6.7-in. film available;<br />
cloudy<br />
07/21/87 MSS 182/24 20 Yes Cloud firee near nuclear<br />
plant<br />
06/28/87 TM & MSS 181/24 10 Yes Clouds to northwest<br />
* 05/11/87 TM & MSS 181/23 10 Yes Very good; cloud free<br />
02/04/87 TM & MSS 181/24 0 Yes Snow cover<br />
01/10/87 MSS 182/24 0 Yes Snow cover; ice free<br />
cooling pond<br />
01/03/87 TM 181/24 0 Yes Cirrus wisps and<br />
shadows; poor<br />
12/25/86 TM 182/24 0 No No film<br />
12/18/86 TM 181/24 20 No No film<br />
12/09/86 TM&MSS 182/24 10 Yes Some cinrus and<br />
popcorn; poor<br />
12/02/86 TM & MSS 181/24 10 Yes Good<br />
10/22I86 TM 182/24 20 No No film<br />
10/15/86 TM & MSS 181/24 0 Yes Good<br />
09/28/86 MSS 182/24 10 Yes Popcorn clouds to<br />
north<br />
08/111/86 MSS 182/24 20 Yes Some cirrus<br />
08/03/86 TM & MSS 182/24 30 Yes Clouds to the<br />
northwest<br />
07/18/86 TM 182/24 30 No No film<br />
07/03/86 MSS 181/24 10 Yes Some popcorn clouds<br />
12
Table 4-2.<br />
Landsat scene quality and cloud contamination analysis for scenes with less than<br />
30 percent cloud cover (see notes following table). (Continued)<br />
Date Landsat Path/row Scene cloud B&W print Comment<br />
sensors cover (%) ordered<br />
06/17/86 MSS 181/24 10 Yes Very good<br />
06/16/86 IM 182/24 20 Yes Heavy popcorn<br />
* 05/31/86 TM & MSS 182/24 10 Yes Very good full scene<br />
* 05/24/86 TM & MSS 181/24 10 Yes Ok, but some popcorn<br />
clouds<br />
05/16/86 MSS-4 181/24 10 Yes Good<br />
* 05/08/86 TM & MSS 181/24 0 Yes Very good<br />
05/07/86 MSS-4 182/24 0 Yes Very good<br />
04/30/86 MSS-4 181/24 10 Yes Good<br />
* 04/29/86 TM & MSS 182/24 10 Yes Some popcorn clouds<br />
* 03/21/86 TM & MSS 181/24 0 Yes Frozen river, ice-free<br />
cooling pond<br />
09/01/85 MSS 182/24 20 No Cloud to the south<br />
08/25/85 MSS 181/24 10 No Cloud free<br />
* 06/06/85 TM & MSS 181124 0 Yes Cirrus to the south<br />
10/25/84 TM&MSS 181/23 10 Yes Thin cirrus over reactor<br />
area<br />
07/13/84 MSS 181/24 10 No Very good<br />
05/01/84 MSS 182/24 20 Yes Some cloud but good<br />
06/25/83 MSS 181/24 10 No Some haze but good<br />
Notes:<br />
1. An * to the left of the date indicates that the data were purchased for that date.<br />
2. Landsat license restrictions apply only to data acquired after 25 September 1985.<br />
3. Only scenes with a cloud cover rating of
From these observations, we produced a prioritized list of TM imagery (see Table 4-3), and<br />
ordered an initial set of imagery (see Table 4-4). This initial set included nine images that spanned<br />
the period from 6 June 1985 through II May 1987. The two more recent images were ordered<br />
later (see Table 4-4) after preliminary results (discussed below) indicated that the area showing<br />
stress was still expanding.<br />
Table 4-3. Scene procurement recommendations.<br />
Date Priority Sensor Medium Area<br />
05/31/86 1 TIM Tape Full Scene<br />
08/03/86 2 TM Tape Quad 4<br />
10/15/86 3 TM Tape Quad 3<br />
10/25/84 4 TM Tape Quad 3<br />
03/21/86 5 7IM Tape Quad 3<br />
12/02/86 6 TM Tape Quad 3<br />
05/11/87 7 TM Tape Quad 3<br />
06/17/86 8 MSS Tape Full Scene<br />
02/04/87 9 TIM Tape Quad 3<br />
05/08/86 10 TM Tape Quad 3<br />
06/06/85 11 TIM Tape Quad 3<br />
04/29/86 12 TM Tape Quad 4<br />
08/25/85 13 MSS Tape Full Scene<br />
07/13/84 14 MSS Tape Full Scene<br />
06/28/87 15 TIM Tape Quad 3<br />
09/07/87 16 TIM Tape Full Scene<br />
01/10/87 17 MSS Tape Full Scene<br />
05/08/86 18 MSS Film Full Scene<br />
14
Table 4-4. Landsat-scene acquisitions.<br />
Image number Date Scene ID Padh/row/quad<br />
1 6/06/85 Y5046208185 181/024/3<br />
2 3121/86 Y5075008144 181/024/3<br />
3 4/29/86 Y5078908200 182/024/4<br />
4 5/08/86 Y5079808133 181/024/3<br />
5 5/24/F6 Y5081408131 181/024/3<br />
6 5/31/86 Y5082108191 182/024/4<br />
7 10/15/86 Y5095808082 181/024/3<br />
8 12/02/86 Y5100608071 181/024/3<br />
9 5/11/87 Y5116608123 181/024/3<br />
10 9/07/87 Y5128508213 182/024/4<br />
11 5/28/88 Y4214308224 182/024/4<br />
Because only a small area was believed to have received a sufficient radiation dose to result in<br />
visible stress, only a 1 degree of longitude by 0.5 degree of latitude area was ordered. Table 4-5<br />
shows the comer points of this area area.<br />
Table 4-5. Comer points of the acquired images.<br />
Corner Longitude Latitude<br />
(:±ddd:mm.ss.s)<br />
(±dd:nmm'ss.s)<br />
Upper left 29:45:00.0 51:40:00.0<br />
Upper right 30:45:00.0 51:40:00.0<br />
Lower left 29:45:00.0 51:10:00.0<br />
Lower right 30:45:00.0 51:10:00.0<br />
15
4.4. IMAGE PREPARATION.<br />
As discussed above, precise registration of images to one another is crucial for change<br />
detection. Registration involves three basic processes: ground control point selection, image<br />
warping, and resampling. Ground control points (GCPs) are spatial features whose location is<br />
known in both images or in one image and a map. The locations of these GCPs are then used to<br />
calculate a mathematical mapping (called a warp) from an arbitrary position in one image to the<br />
corresponding ground position in the other image or map. This mapping can be envisioned as the<br />
stretching and twisting (warping) of one image so that it can be overlaid on the reference image<br />
(map). Once overlaid, any position on the ground occurs at the same location in both images.<br />
After successful warping, pixels in the warped image will not, in general, be coincident with pixels<br />
in the reference image. Instead, they will be off center and skewed relative to those in the reference<br />
image. To correct for this problem, the warped image is resampled to a set of pixels congruent<br />
with those at the reference image.<br />
The TM data were delivered directly to STX Corporation of Lanham, Maryland, for<br />
registration (geocoding). Only the image dated 31 May 1986 was directly registered to maps. The<br />
maps used were 1:250,000 scale Joint Operations Graphics (JOGs) identified as Series 1501,<br />
Sheet NM 35-3, Edition 2-GSGS and Sheet NM 36-1, Edition 1. The remaining eight images<br />
were then registered to the map-registered 31 May 1986 image. This allowed the selection of a<br />
greater number of control points, resulting in a more precise registration among the images than<br />
could have been attained by registering all of the images to the maps. Two more recently procured<br />
images (numbers 10 and 11) were registered in like fashion to the 6 June 1985 image.<br />
Because the maps used to register the first image were Universal Transverse Mercator (UTM)<br />
projections, Grid Zone 36, all of the subsequent images were registered to this projection. That is,<br />
any of the images can be overlaid on a UTM map (or visa versa). The comer points of the area<br />
covered by these images are shown in Table 4-6.<br />
Table 4-6. Comer points of the geocoded images after resampling.<br />
Longitude Latitude UTM X UTM Y<br />
Corner (degrees) (degrees) (meters) (meters)<br />
Upper left 294246.9574E 513956.7881N 272700. 5729000.<br />
Upper right 304459.6821E 514124.8298N 344475. 5729000.<br />
Lower right 304631.7845E 510959.8775N 344475. 5670725.<br />
Lower left 294501.3475E 510833.4597N 272700. 5670725.<br />
16
A 12.8-km sub-area (512 x 512 pixels) was identified for intensive analysis. The corner points<br />
of this area are shown in Table 4-7. Unless otherwise specified, these coordinates apply to all<br />
color figures shown in this report.<br />
Table 4-7. Comer points (Zone 36 UTM coordinates) of the<br />
512 x 512 pixel area analyzed in this report.<br />
Comer X Y<br />
(meters)<br />
(meters)<br />
Upper left 289900. 5702100.<br />
Upper right 302675. 5702100.<br />
Lower left 289900. 5689325.<br />
Lower right 302675. 5689325.<br />
Figure 4-1 shows a sample image of this 512 x 512 pixel area. The Pripyat river flows<br />
through the upper right hand comer of the image. The dark area adjacent to the river is in the<br />
cooling pond of the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station. The industrial area comprising the reactor<br />
station itself is at the upper left corner of the cooling pond. The cooling water intake channel (right<br />
angle bend) and the outlet channel (obtuse angle) are evident. Water circulates counterclockwise in<br />
the pond. The reactor buildings are located in an east-west row just above the east-west portion of<br />
the intake channel. Reactor four is leftmost in the row.<br />
The city of Pripyat is at the upper center of the image, just south of the river and northwest of<br />
the reactor site. The city of <strong>Chernobyl</strong>, an old river port, is about 10 km to the southeast, not on<br />
this image.<br />
The bright green areas in Figure 4-1 are mostly agricultural lands; the bright pinkish or reddish<br />
areas are bare fields. The dark green areas are predominantly conifer forests as evidenced by<br />
seasonal progressions in subsequent figures.<br />
The standard geometric-correction algorithm applied to TM data (Irons, 1985) results in an<br />
image that appears to be somewhat blurred. But, unlike Goldman and coworkers (Goldman,<br />
1987), we made no attempt to sharpen the images. If the images are sharpened, even small<br />
(subpixel) registration errors may result in large apparent image-to-image changes for<br />
corresponding pixels. Further, because these errors are systematic, real stress at the single pixel<br />
level may be entirely obscured by this misregistration-induced apparent stress.<br />
17
Figure 4-1.<br />
Vicinity of the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station on 8 May 1986; 7,4,1 false-color<br />
presentation (12.8 km by 12.8 km, north is up). Sixteen of the indicated polygons<br />
are areas selected as candidate training sets for the forest classification process.<br />
18
SECTION 5<br />
IMAGE ANALYSIS<br />
This section discusses the techniques used to identify forest pixels in the Landsat imagery and<br />
reviews the false-color images and forest stress results for each date.<br />
5.1. IMAGE PARTITIONING.<br />
Because only forested areas are of interest for our analysis, we applied methods based on<br />
statistical decision theory to eliminate other areas from consideration. Maximum-likelihood<br />
classification (Swain, 1978), a parametric method for statistical classification, requires estimates of<br />
the parameters of the statistical distributions that are assumed to characterize the variation of<br />
spectral intensities over statistically homogeneous regions of the image. The pixels within selected<br />
areas of the image (referred to as training areas or training sites) are appropriately analyzed to<br />
estimate statistical parameters of an assumed statistical distribution for the pixels belonging to each<br />
class. Each training area yields a set of statistical parameters and a corresponding class distribution<br />
in the form of an analytic function involving the class parameters.<br />
The distributions are usually assumed to be multivariate normal in their analytic form. The<br />
parameters of the multivariate normal distribution law include the mean vector and the covariance<br />
matrix. The mean vector is a band-by-band average of the intensity taken over every pixel in the<br />
training site. Thus, if seven bands are used, the mean vector is seven dimensional (i.e., it has<br />
seven components, one for each band average). The covariance matrix comprises the covariances<br />
estimated from all pairings of the band intensities and will be a square array of dimension equal to<br />
the number of bands.<br />
An analytic distribution derived in this way provides a means for estimating the probability that<br />
a pixel belonging to the corresponding class will be found to lie at any given point in the<br />
multispectral feature space. Thus, probabilities that a pixel belongs to each of the classes under<br />
consideration can be estimated at any given point in the multispectral feature space. Maximumlikelihood<br />
classification assigns a pixel to the class for which its position in the feature space has<br />
the greatest probability. In this manner the pixel classification process assigns each pixel of the<br />
image to one or the other of the classes, resulting in a partitioning of the image into the predefined<br />
classes.<br />
Training sites are usually determined from ground truth. Unfortunately, because ground truth<br />
of the <strong>Chernobyl</strong> area was not readily available to us at the time of this analysis, we identified<br />
training sites visually by spectral characteristics and characteristic texture. Potential training sites<br />
identified in this way are displayed as polygons in Figure 4-1.<br />
19
Ideally, assignment of a pixel to a given class depends on the joint probability that it would be<br />
characterized by a pixel vector X (its vector of intensity values) while belonging to class Coi. This<br />
joint probability is given by the following equation:<br />
p( Op(G) . exp- .(X - M). T'1: (X - M.)]<br />
(2xj(1)<br />
where,<br />
n = the number of bands (also the dimensionality of X and U, and the order<br />
of S),<br />
p(co) = the probability of class wi,<br />
Mi = the mean vector for class coi,<br />
X = the pixel vector (n intensity values), and<br />
Mi = the covariance matrix for class mi.<br />
Note that this is not a conditional probability. However, it is related to the conditional probabilities<br />
by<br />
p(X C, O)= p(Xji) p(Ci) =p(OiIX) p(X) (2)<br />
The pixel is then assigned to class wji if and only if<br />
pXoI) po)>pXcj) p( j) (3)<br />
for all j = 1,2,..., m, where m is the number of classes. Because in this study the class<br />
probabilities p(coi) were unknown, they were assumed to equal one another, causing the class<br />
assignments to be determined by the conditioned probability of the pixel vector's occurrence, given<br />
the hypothetical class.<br />
For this study we required the training sites to be minimal in number and near the reactor, but<br />
not subject to radiation stress. We chose the smallest possible number of training sites for the<br />
analysis because maximum-likelihood classification is computationally expensive. Ideally, one<br />
training site would be selected for each distinguishable forest type. The distinctions among forest<br />
types might be based on ground truth data such as differences in the mixture of coniferous and<br />
deciduous trees, age of the stand, densities or crown closures, or species in the area. Because little<br />
ground truth was available at the time of the study, the selection of training sites had to be based<br />
entirely on their remotely sensed multispectral characteristics. It was found that two training sites<br />
20
provide a satisfactory classification of the pixels in the image, a classification that includes nearly<br />
all of the pixels in regions believed to be forested while including few in regions that were clearly<br />
not forested. In the process of selecting these two training sites a number of potential training sites<br />
were selected in the area near the <strong>Chernobyl</strong> nuclear reactor, these are shown on Figure 4-1. Note<br />
that only 16 of these are forest. Five were chosen to include only water, and one was believed to<br />
be sand. These nonforest sites were chosen in anticipation of a need to normalize the various<br />
images using features whose spectral response could be assumed stable over time. Because the<br />
method finally chosen to detect stress did not require such normalization, the nonforest training<br />
sites were not used.<br />
Partitioning the image assigned every pixel in the image to one or the other of the predefined<br />
classes. Because we were interested only in forests, we selected only forest training sites and<br />
included only forest classes in the set of predefined classes. After the image was partitioned in this<br />
way, pixels that obviously contained water, bare soil, and all manner of other things were<br />
assigned, along with the forest pixels, to the closest of the forest classes. As yet no consideration<br />
had been given to the possibility that a pixel's spectral features might have been far more likely to<br />
have arisen among pixels belonging to nonforest classes such as water, bare soil, etc; we had yet to<br />
define what was and was not a forest.<br />
Nonforested areas could have been removed from consideration by maximum-likelihood<br />
classification in the same manner as was used for the forest classes. However, this would have<br />
entailed developing training areas or sites for every class of landcover that occurred in the image,<br />
and it would have required that the a probability be calculated for every pixel for every such class.<br />
A computationally more efficient process was to apply a threshold to the probability computed for a<br />
pixel's membership in the forest classes. If the pixel's multispectral features occurred in a<br />
particular forest class with a probability less than a certain threshold value, the pixel was assumed<br />
to belong to a class other than that forest class. In practice, we adjusted the probability threshold<br />
while monitoring the outcome of the classification process, raising the threshold if too many stray<br />
pixels in the image seemed to be assigned to the desired classes and lowering the threshold if too<br />
many of the pixels thought to belong to the desired classes were being rejected. The resulting<br />
statistical decision process may be looked upon as a casual implementation of the Neyman-Pearson<br />
decision rules; the threshold for pixels belonging to the forest class was made as low as possible<br />
without accepting too many pixels that lay outside the apparent areas of forest.<br />
When the image was partitioned using only one training site (e.g., training site 4 on<br />
Figure 4-1), many of the forested areas were misclassified'as nonforested areas (errors of<br />
omission) when a reasonable threshold value was selected. This threshold value was determined<br />
qualitatively by viewing the effect of a given threshold value on the classification. Lowering the<br />
21
threshold sufficiently would include all suspected forested areas, but at the cost of including too<br />
many other areas that were perceived to be nonforested (errors of commission).<br />
Finding that no adjustment of the threshold would lead to an entirely satisfactory result, we<br />
concluded that the use of only one training site was not sufficient and considered using two. The<br />
question then became "which two?" As discussed above, the selection should be based on<br />
statistical arguments. To this end, PSR developed a modified form of the transformed divergence<br />
(Swain, 1978) specifically for use in this study. The divergence between two training sites was<br />
defined by<br />
4I _ ) _ I<br />
•)= (<br />
l1]+l<br />
The transformed divergence is given by<br />
DT=<br />
ij I 8(5)<br />
Unfortunately, the transformed divergence saturated (yielded 100) for even small differences in<br />
cluster statistics if a large number (>4) of bands was used. In addition, the results were strongly<br />
dependent on the number of bands used. To reduce the effects of these problems, we defined the<br />
modified transformed divergence by<br />
D MT = 100 1 xp(-<br />
ii L\ 2n/ (6)<br />
The modified transformed divergence was relatively independent of the number of bands used<br />
in the analysis provided that the bands used were not highly correlated. Table 5-1 presents the<br />
modified transformed divergence between every pair of potential forest training sites shown in<br />
Figure 4-1. These results led us to choose sites 3 and 4 as training sites to partition the 6 June<br />
1985 image. Figure 5-1 shows the results of this classification. Areas colored green were<br />
assigned to the class developed from training site 4, and areas colored yellow, to the class<br />
developed from training site 3. This classification map (recall that the image is registered, so this<br />
really is a map) was then overlaid on a gray-scale TM band 4 image of the area. It is not important<br />
for the present analysis to know the difference between the types of forest in these training sites.<br />
We need only to account for their statistically distinct multispectral characteristics.<br />
22
cn eno<br />
00k0t- t Oo %0 roo~~ M C 4e<br />
oo0% a 00 00-c m0% a m<br />
4000 Q 00 N*AC 0 4% 1<br />
Ml V_ t*- r- n 0 t- N 00 Aý t~-00O<br />
C.) eq 0%~ N~~'O - (nl* 4 in~ 00*4 - t 00<br />
00e cC C : *M ýQ oc* o ioc-t<br />
oo a6f~ 0 %Vcý0 ~ 4 % %0O~<br />
0 0%Ot-00 w' w we4 m0 M<br />
So -00 C -%n W 0o0 in t- cc cc O<br />
%Q~ 0000 00 ) ) o ý - Q<br />
co % 000<br />
0%* n0<br />
cn n o r- @000 r- V % 00 0% %C<br />
w 00<br />
23 %t
Ar<br />
Figure 5-1.<br />
Partitioning of 6 June 1985 image into classes derived from training sites 3 (yellow)<br />
and 4 (green). Gray-scale background consists of nonforested areas.<br />
24
To minimize the effect of local atmospheric variations, we wished to choose training sites as<br />
close to the reactor as possible. On the other hand, if training sites too close to the reactor were<br />
chosen, they could have been subject to radiation stress, especially if the image used to identify<br />
forested areas was acquired after the accident. This study used the image dated 6 June 1985<br />
(before the accident) to partition the image. Nevertheless, the training sites were selected to be<br />
outside the area in which radiation stress was expected to appear in postaccident images.<br />
5.2. MAPPING STRESS.<br />
In this section, each of the selected images is reviewed in chronological order with important<br />
features identified. These require some explanation. Standard false-color prints of TM data<br />
display band 4 as red (R), band 3 as green (G), and band 2 as blue (B).<br />
This presentation is referred to as a 4,3,2 RGB false color. In this report, the color ordering<br />
will always be RGB unless otherwise stated). As discussed in Section 2 above, vegetation should<br />
appear red in this standard presentation. The shadowing of trees by one another causes forests to<br />
appear highly textured and darker.<br />
There is another standard false-color display gaining acceptance in the community (Johnson):<br />
the 7,4,2 false color. This results in a more appealing and less arcane image because it appears<br />
much like a normal (3,2,1 RGB), true-color image but has better contrast information content; the<br />
7,4,2 false-color image contains information from three quite different regions of the spectrum,<br />
rather than being restricted to the visible region as is a 3,2,1 image. In the 7,4,2 presentation bare<br />
soils tend to appear red, with drier soils appearing a brighter red color. Vegetation appears green,<br />
and shallow or turbid water tends to appear blue.<br />
The color prints presented here use a nonstandard 7,4,1 false-color display. This combination<br />
captures most of the advantages of the standard 7,4,2 false-color display, but with the added<br />
advantage that it makes cirrus clouds more obvious. Because cirrus clouds can (and invariably do)<br />
interfere with analysis, knowledge of the location of even light cirrus clouds can be important<br />
For images acquired on or after 29 April 1986, stress maps have been produced using PSR's<br />
Hyperscout change detection algorithm with the 6 June 1985 image as reference. The stress<br />
algorithm assigns a single number, called the stress index, to each forested pixel in the image, and<br />
these are displayed in the stress map. Again, the resulting stress map, like the images from which<br />
it was derived, is registered to the UTM projection, Grid Zone 36. The stress map could be<br />
displayed as a gray scale, but the resulting image would be difficult to discern in areas manifesting<br />
low stress. In this report, the stress gray scale is converted to a color scale. A legend showing<br />
this color scale appears with the stress maps shown below. Values of the highest stress index are<br />
colored red. As the stress index decreases, the color shifts continuously through orange, yellow,<br />
green, and blue to magenta. Thus, magenta-colored areas are the least stressed. Nonforested<br />
25
areas, which could not be analyzed for stress, are filled in with a gray-scale TM band 4 image to<br />
establish a spatial context for the interpretation. The accompanying text discusses significant<br />
features of these stress maps, as well as of the 7,4,1 false-color images.<br />
The stress maps derive from an analysis of only three of the 7 bands, namely, 3, 4, and 7.<br />
Band 6 (the thermal band) was not used because the images frequently included clouds. The low<br />
temperature of clouds, coupled with the 120-m resolution of band 6, affects an area larger than that<br />
where clouds are evident. Even light cirrus clouds would mislead a stress calculation that includes<br />
band 6 because of its high sensitivity to the temperature of the cloud cover. Bands 1 and 2 (blue<br />
and green) were not used because they also are relatively sensitive to cirrus clouds and atmospheric<br />
scattering as compared to the red or infrared. This left only bands 3, 4, 5, and 7. Bands 5 and 7<br />
are very highly correlated, so either may be used. We arbitrarily choose band 7 over band 5.<br />
Figure 5-2 shows the reference image from 6 June 1985, 1 year before the accident. Recall<br />
that this is also the image used to classify forest types. A significant feature of this image is that<br />
the forested areas due west of the reactor appear darker and may even have a brownish tint. This is<br />
also the type of coloring (spectral reflectance) that might be expected from radiation-stressed<br />
coniferous forests. Because it is unlikely that the coloring present on 6 June 1985 was caused by<br />
radiation stress (certainly not that caused by the release on 26 April 1986), this image indiates that<br />
qualitative (visual) identification of stress from the false-color images can be misleading.<br />
Figure 5-3 shows the 21 March 1986 image also taken before the accident. Some snow cover<br />
was evident (blue color), and the rivers (but not the warm cooling pond) were ice covered. At<br />
training site 3 very little snow cover was visible through the canopy. On the other hand. some<br />
snow appeared at training site 4. This observation can be interpreted in at least two ways. First,<br />
training site 4 may have more deciduous trees, or, second, the forest may not be as dense in<br />
training site 4 as it is in training site 3. Which, if either, of these interpretations is correct has not<br />
been determined.<br />
A study, similar in intent to this study, was recently preformed by Goldman and coworkers<br />
(Goldman, 1987). They used enhanced Landsat TM data to identify stress visually. They<br />
concluded that stress could not be detected visually on images taken before 16 June 1986 (i.e.,<br />
until more dian 7 weeks after the accident).<br />
Three days after the accident, Landsat acquired an image of the reactor area. This image, taken<br />
29 April 1986, shows some cloud cover as seen in Figure 5-4. The reactor, still extremely hot,<br />
was readily apparent as a deep red area due north of the reactor water inlet pooL Another reddish<br />
white feature appeared 1.1 km to the west of the reactor and was mistakenly reported to be another<br />
burning reactor. There was no visibly stressed vegetation.<br />
26
Figure 5-2.<br />
Date: 6 June 1985, 1 year preaccident; 7,4,1 false color. This date was used as<br />
a reference for change detection.<br />
27
Figure 5-3. Date: 21 March 1986, 2 months preaccident; 7,4,1 false color. Snow cover is<br />
evident (bright blue) around the forest patches.<br />
28
Figure 5-4.<br />
Date: 29 April 1986, 3 days postaccident; 7,4,1 false color. The deep red pixel is<br />
the thermal emission from the hot reactor core.<br />
29
Figure 5-5 presents the stress map for this date. The area of high stress index that appears 4<br />
km due west of the reactor was cleared before the accident. Nearly all of the rest of the apparent<br />
stress was caused by clouds. The centers of cumulus clouds and their shadows usually appear<br />
deep red in the stress index maps, while at their fringes there appears a sharp, rainbowlike<br />
progression of colors from deep red to magenta. Cirrus clouds are much more insidious; their<br />
effects vary in color and extent. In some cases cirrus clouds show no effect at all, while in others,<br />
they produce a pattern of apparent stress similar to that expected of real stress. In Figure 5-5, there<br />
is a small area approximately 0.9 km west by southwest of the reactor that has a sufficiently high<br />
stress index to be displayed in the green to yellow color range. On the basis of this single image, it<br />
is debatable whether this is real stress, cloud effects, smoke, debris covering the foliage, or<br />
something else. A qualitative measure of the statistical nature of the noise can be obtained by<br />
looking at areas that were not stressed (e.g., training site 3). These areas are colored mostly blue<br />
and magenta except near the edges of the forests. From this observation, we conclude that areas in<br />
the stress map speckled blue and magenta are probably not really stressed but only the effect of<br />
random noise-level detections of change. In addition, any apparent stress that appears only at the<br />
edges of forests is suspect, because of possible registration errors.<br />
The 8 May 1986 image, presented in Figure 5-6, was collected 12 days after the accident.<br />
There were no clouds in the image, nor was there stressed foliage visible in the 7,4,1 false-color<br />
presentation.<br />
Figure 5-7 presents the stress map made from the 8 May 1986 image. In it appears a very well<br />
defined 0.9-km long strip of forest, beginning 2 km west by southwest of the reactor and running<br />
to the west, where stress was apparent. This area also had a higher than normal stress index on<br />
29 April 1986, but the area was not as clearly deft-ed. The shape of this stressed area was<br />
consistent with that of a directed explosion or wind-deposited fallout. Because by this time the fire<br />
at the reactor was reported to have been extinguished, the detected change seems unlikely to be<br />
attributable to smoke. Thus, in this stress map we can very clearly see some indications of<br />
accident-related stress or change that appeared within 12 days after the accident. Located 6 km<br />
south of the reactor there appears an apparent stress or change feature that we have interpreted to be<br />
an artifact of the change analysis algorithm. The exact cause is unknown. This artifact appears on<br />
almost all of the images that follow; its stress index changes very little in time.<br />
30
0 128 256<br />
St•es Index<br />
Figure 5-5.<br />
Date: 29 April 1986, 3 days postaccident, forest stress map. colored according to<br />
indicated scale. Gray-scale background consists of nonforested areas.<br />
31
Figure 5-6. Date: 8 May 1986, 12 days postaccident; 7,4,1 false color.<br />
32
0 128 256<br />
Sbss Index<br />
Figure 5-7.<br />
Date: 8 May 1986, 12 days postaccident, forest stress map.<br />
33
Figure 5-8 shows the 24 May 1986 image collected 4 weeks after the accident. Although it is<br />
difficult to see on the 7,4,1 false-color image, there was a large cirrus cloud in the area of the<br />
reactor. There were also some "popcorn" clouds in the area. The obscuration of the cirrus cloud<br />
and the natural color of the area make positive visual detection of stress uncertain in the false-color<br />
image. On the other hand, an area of change becomes very obvious and spatially well defined<br />
when presented on the corresponding stress map (Figure 5-9). Unfortunately, so does the large<br />
cirrus cloud. The cirrus cloud appears as a blue to green swath on the stress map, affecting the<br />
bulk of the eastern side of the image. The "popcorn" clouds and their shadows show up as deep<br />
red. By 24 May 1986 aerial spraying to prevent the wind from redistributing radioactive dust,<br />
along with the dust itself, may have contributed to the changes that were detected.<br />
The image taken on 31 May 1986 (Figure 5-10) also included some clouds. Positive visual<br />
identification of stress remains uncertain, but the stress map shows the stressed area very clearly<br />
(Figure 5-11). Little change appears in the spatial distribution of stress between 24 and 31 May,<br />
except that a small new area just south of the previously stressed area began to indicate a very high<br />
stress index. Because this new area was adjacent to a road, it probably indicates changes caused<br />
by human activities other than radiation release.<br />
After 31 May 1986, the next image selected for analysis was that collected 15 October 1986.<br />
An interim image collected 16 June 1986 was not purchased because of cloud cover exceeding our<br />
standards.<br />
By 15 October 1986 a band of stressed forests that appeared to the west of the reactor can be<br />
readily discerned in the false-color image (Figure 5-12). Also visible is evidence of extensive<br />
human activity. The Soviets had started to clear some of the contaminated forests near the road that<br />
passes to the west of the reactor. There was evidence of extensive diking operations in the<br />
northwest and northeast comers of the image. Indications appeared of a cleared swath passing<br />
through the coniferous forest to the southwest of the reactor. Because this image was collected in<br />
the fall and there were deciduous trees in the area, the spectral definition of stress becomes affected<br />
by the natural changes of the deciduous foliage and the seasonal sun angle changes. As a result,<br />
the apparent stress is lower than before. Examination of the stress map in Figure 5-13 confirms<br />
this hypothesis. The stressed area remains obvious and fairly well defined, but the stress index is<br />
lower.<br />
The situation regarding deciduous foliage is similar on 2 December 1986 (Figure 5-14). In<br />
addition, the low sun elevation causes a further loss of sensitivity. Although the stresses<br />
(Figure 5-15) appear lower than before, the areal extent of the stress does not appear to have<br />
changed significantly.<br />
34
Figure 5-8. Date: 24 May 1986,4 weeks postaccident; 7,4,1 false color.<br />
35
I I I<br />
0 128 256<br />
Stress Index<br />
Figure 5-9. Date: 24 May 1986,4 weeks postaccident, forest stress map.<br />
36
Figure 5-10. Date: 31 May 1986, 5 weeks postaccident; 7,4,1 false color.<br />
37
I<br />
II<br />
0 128 256<br />
Stess Index<br />
Figure 5-11. Date: 31 May 1986, 5 weeks postaccident, forest stress map.<br />
38
Figure 5-12. Date: 15 October 1986, 5.6 months postaccident; 7,4,1 false color.<br />
39
0 128 256<br />
Stress Index<br />
Figure 5-13. Date: 15 October 1986, 5.6 months postaccident, forest stress map.<br />
40
Figure 5-14. Date: 2 December 1986, 7.2 months postaccident; 7,4,1 false color.<br />
41
..........<br />
0 128 256<br />
Stress Index<br />
Figure 5-15. Date: 2 December 1986, 7.2 months postaccident. forest stress map.<br />
42
The Soviet mitigation efforts were much more extensive by 11 May 1987 (Figure 5-16). A<br />
large area is cleared to the west of the reactor complex, as is an even larger area south by southeast<br />
of the reactor. Much of the diking appears to have been completed, and the areas behind the dikes<br />
appear flooded. Again, there is a rather large cirrus cloud partially obscuring most of the forests in<br />
the western part of the image.<br />
Goldman and coworkers (Goldman, 1987) state that no new stressed areas had appeared by 11<br />
May 1987. Indeed, a visual inspection of the false-color image provides no evidence to the<br />
contrary. However, the stress map indicates significant new areas beginning to show stress.<br />
Before looking at the stress map, the reader may find it enlightening to attempt to identify these<br />
areas visually in the false-color image. As an aid to the reader, Figure 5-17 shows Sovietsupplied<br />
gamma dose rate contours dated 1 May 1987 overlaid on the TM image. This ground<br />
truth will be discussed in detail in Section 6 below. The innermost contour is for 100 mR/hr, with<br />
contours for 50, 10, 5, 2, 1, 0.5, and 0.1 mR/hr appearing progressively outward. If these<br />
contours are correct, the bulk of the coniferous forest to the south was still receiving a dose of at<br />
least 5 mR/hr, and all areas between the coniferous forest and the cooling pond to the east and the<br />
reactor complex to the north were receiving a dose of at least IOmR/hr. Because the areas<br />
receiving doses higher than this had already shown an elevated stress index, these areas were the<br />
most likely next candidates.<br />
Turning to the stress map shown in Figure 5-18a, we see that these areas were indeed<br />
beginning to show symptoms of stress. These newly stressed areas included much of the<br />
coniferous forest to the south of the reactor and nearly all of the forested areas between the<br />
coniferous forest and the cooling pond to the east and the reactor complex to the north. For ease of<br />
comparison, Figure 5-18b shows the dose rate contours overlaid on the stress map from<br />
Figure 5-18a.<br />
No images that were free of clouds and haze (see Table 4-2) were collected in 1987 after<br />
22 July 1987. However, an image taken 7 September 1987 (Figure 5-19) was procured to help<br />
confirm the indications of continued stress implied by the 11 May 1987 stress map. Unfortunately<br />
for the comparison, the Soviets had cleared much of the suspect area lying within the 10 mR/hr<br />
dose contour. The northeastern edge of the coniferous forest continued to appear heavily stressed<br />
(Figure 5-20) but not as heavily stressed as before. The apparent reduction in stress may have<br />
been due to the heavy haze. Indications of new stress appear in the 7 September 1987 stress map<br />
in an area just north of the area first showing stress.<br />
43
Figure 5-16. Date: 11 May 1987, 1 year postaccident; 7,4,1 false color.<br />
44
Figure 5-17.<br />
Soviet-supplied gamma dose rate contours of 1 May 1987 overlaid on the<br />
11 May 1987 7,4,1 false-color image. The innermost contour is 100 mR/h, the<br />
next innermost 50 mR/h, then 10, 5, 2, 1, 0.5 and 0.1 mR/h, respectively.<br />
45
I I I<br />
0 128 256<br />
Stress Index<br />
Figure 5-18a. Date: 11 May 1987, 1 year postaccident, forest stress map.<br />
46
I<br />
II<br />
0 128 256<br />
Stress Index<br />
Figure 5-18b.<br />
Date: 11 May 1987, forest stress map with gamma dose rate contours of<br />
1 May 1987 (see Figure 5-17 for contour values).<br />
47
Figure 5-19. Date: 7 September 1987, 16 months postaccident; 7,4,1 false color.<br />
48
I I I<br />
0 128<br />
Sb'ess Index<br />
256<br />
Figure 5-20. Date: 7 September 1986, 16 months postaccident, forest stress map.<br />
49
Because newly stressed areas were still being found and the 7 September 1987 image was of<br />
low quality, a more recent image was procured. This image, dated 28 May 1988, is presented on<br />
Figure 5-21. Again, Soviet mitigation efforts had cleared some of the newly suspect arrias, so<br />
continued monitoring of those areas was not possible. The clearing of the newly suspect areas<br />
suggests that these armas had been affected enough that they needed to be decontaminated, although<br />
no mitigation efforts were applied to the coniferous forest to the south of the reactor, at which<br />
signs of stress persisted (Figure 5-22).<br />
5o
Figure 5-2 1. Date: 28 May 1988, 25 months postaccident; 7,4,1 false color.<br />
51
I I I<br />
0 128 256<br />
S&ess Index<br />
Figure 5-22. Date: 28 May 1988, 25 months postaccident, forest stress map.<br />
52
SECTION 6<br />
GROUND TRUTH<br />
Some ground truth became available (Asmolov, 1987) in the form of gamma radiation dose<br />
contours for 1 May 1987 (i.e. approximately I year after the accident). Unfortunately, there are a<br />
number of difficulties associated with the use of these data. First, the copy of the dose contours<br />
that was available to us is very poor; the contours appear to be hand drawn, and in some places the<br />
contours are not closed, making digitization difficult. Also, the contour map has no grid lines and<br />
no accurate ground control points (GCPs). In fact, the only usable GCPs are river bends.<br />
Further, the indicated courses of these rivers on the contour map do not match the maps that were<br />
used to register the images. Because they match the images fairly well, registration directly to the<br />
images is a viable alternative. Two cities are shown as circles, but the location of one of them,<br />
<strong>Chernobyl</strong>, is apparently in error by several kilometers. The course of the river Uzh also appears<br />
to be in error near this city. The location of the reactor itself is not marked on the map.<br />
In spite of these difficulties, 36 GCPs were identified. Using these GCPs, the coefficients of a<br />
second-order warp were calculated by means of a least squares algorithm. When this warp was<br />
applied to the digitized contours and the results overlaid on an image, the 100 mR/h contour was<br />
centered on reactor 3 instead of reactor 4. Use of a third-order warp resulted in unacceptable (i.e.,<br />
unlikely) distortions of the contours near the edges of the map. Attempts to generate a fourth-order<br />
warp were unsuccessful. Because the ability to weigh GCPs preferentially is inherent to the<br />
algorithm used, another GCP, the best guess of the location of the reactor based on the shape of<br />
the contours, was added to the set and given a relative weight of 100. The coefficients for a<br />
second-order warp were then recalculated using the least squares algorithm. The equations of this<br />
warp are shown in equations 7 and 8:<br />
and<br />
xref=al+a 2 x+a 3 -y+a 4 - x2+a 5 * x- y+a 6 *y 2 (7)<br />
yrf= bi +b 2 . x+b 3 . y+b 4 . x2+b • x- y+b 6 .y 2 (8)<br />
53
where,<br />
ai = the coefficients of the warp for the x-coordinate of the image,<br />
bi = the coefficients of the warp for the y-coordinate of the image,<br />
x - the x-coordinate of a point on the dose contour map,<br />
xref = the corresponding column (x-coordinate) on the image, which is related to UTM<br />
map coordinates,<br />
y - the y-coordinate of a point on the dose contour map, and<br />
yref - the corresponding row (y-coordinate) on the image, which is related to UTM<br />
map coordinates.<br />
The x- and y-axes of the contour plots are arbitrary because no grid lines or map projection was<br />
given for the dose contour map. Still, the relative magnitudes of these coefficients are of some<br />
interest; they are presented in Table 6-1. A measure of the accuracy of the registration can be<br />
obtained by warping the location of the GCPs on the contour map through Equations (7) and (8);<br />
then comparing these warped coordinates to the true location of those GCPs in the image. These<br />
results are presented in Table 6-2. The contour lines were warped and overlaid on an image<br />
(Figure 5-17). In order starting with the innermost, the contours are for doses of 100, 50, 10, 5,<br />
2, 1, 0.5 and 0.1 mR/h.<br />
Table 6-1. Second order warp coefficients for the<br />
gamma dose contours.<br />
i ai bi<br />
1 -0.2760500E+04 0.4388116E+04<br />
2 0.1866979E+00 0.1391407E+00<br />
3 -0.4149658E+00 -0.4166605E+00<br />
4 0.8310569E-05 -0.3711088E-05<br />
5 0.1337978E-04 -0.1319325E-05<br />
6 0.1379850E-04 -0.5189849E-05<br />
54
Table 6-2. Accuracy of the registration of GCPs.<br />
x Xref (pixels) xfit (pixels) Error (pixels)<br />
14864.79 1024.00 1024.99 0.99<br />
13347.76 249.00 271.18 22.18<br />
13376.74 271.00 282.24 11.24<br />
13525.32 365.00 357.72 -7.28<br />
13434.05 297.00 301.57 4.57<br />
13533.28 364.00 352.89 -11.11<br />
13620.08 397.00 387.50 -9.50<br />
13750.20 470.00 459.64 -10.36<br />
13847.89 505.00 504.52 -0.48<br />
13866.87 509.00 511.20 2.20<br />
14116.60 653.00 648.12 -4.88<br />
14083.66 635.00 619.49 -15.51<br />
14224.67 712.00 700.87 -11.13<br />
14381.22 794.00 775.78 -18.22<br />
14688.25 939.00 936.75 -2.25<br />
15065.93 1148.00 1140.12 -7.88<br />
15145.92 1204.00 1184.85 -19.15<br />
15427.47 1334.00 1320.82 -13.18<br />
15801.45 1519.00 1518.99 -0.01<br />
15829.66 1530.00 1532.57 2.57<br />
15960.41 1608.00 1606.15 -1.85<br />
16016.58 1637.00 1633.56 -3.44<br />
16197.20 1732.00 1734.57 2.57<br />
15469.94 1337.00 1334.33 -2.67<br />
15438.92 1319.00 1316.72 -2.28<br />
15338.93 1262.00 1262.07 0.07<br />
15215.50 1196.00 1194.90 -1.10<br />
15246.26 1218.00 1211.36 -6.64<br />
15326.76 1251.00 1254.99 3.99<br />
14944.11 1049.00 1049.89 0.89<br />
14803.87 976.00 975.37 -0.63<br />
14782.98 971.00 965.25 -5.75<br />
14692.60 917.00 917.88 0.88<br />
14563.63 849.00 848.41 -0.59<br />
55
Table 6-2. Accuracy of the registration of GCPs (Continued).<br />
X Xref (pixels) xfu (pixels) Error (pixels)<br />
14174.82 636.00 642.65 6.65<br />
14255.32 685.00 684.05 -0.95<br />
13806.38 449.00 448.57 -0.43<br />
Y Yref(pixels) yfu (pixels) Error (pixels)<br />
9029.81 1274.00 1273.78 -0.22<br />
10839.32 260.00 267.20 7.20<br />
10758.94 331.00 311.86 -19.14<br />
10647.04 365.00 376.65 11.65<br />
10548.19 434.00 428.17 -5.83<br />
10486.92 475.00 463.98 -11.02<br />
10235.79 607.00 602.26 -4.74<br />
10244.13 592.00 600.88 8.88<br />
10062.87 689.00 701.09 12.09<br />
9982.62 745.00 744.78 -0.22<br />
9959.43 756.00 762.79 6.79<br />
9729.84 888.00 885.48 -2.52<br />
9808.04 834.00 846.50 12.50<br />
9536.02 993.00 995.44 2.44<br />
9302.09 1116.00 1126.06 10.06<br />
9117.25 1237.00 1230.63 -6.37<br />
9116.23 1239.00 1232.37 -6.63<br />
8452.58 1591.00 1586.75 -4.25<br />
7977.77 1843.00 1839.49 -3.51<br />
7837.40 1915.00 1912.75 -2.25<br />
7895.74 1868.00 1883.86 15.86<br />
7595.00 2046.00 2040.27 -5.73<br />
7642.69 2014.00 2017.33 3.33<br />
7861.99 1892.00 1895.45 3.45<br />
7782.38 1938.00 1936.27 -1.73<br />
7783.66 1936.00 1934.14 -1.86<br />
7515.21 2069.00 2070.78 1.78<br />
7574.82 2037.00 2040.58 3.58<br />
7613.80 2019.00 2021.74 2.74<br />
7408.67 2132.00 2120.83 -11.17<br />
56
Table 6-2. Accuracy of the registration of GCPs (Continued).<br />
y yref(pixels) Yfit (pixels) Error (pixels)<br />
7390.46 2123.00 2127.52 4.52<br />
7320.72 2162.00 2162.85 0.85<br />
7291.87 2183.00 2175.80 -7.20<br />
7373.52 2124.00 2131.29 7.29<br />
7468.49 2079.00 2073.78 -5.22<br />
7507.46 2050.00 2055.69 5.69<br />
7593.20 2000.00 2000.42 0.42<br />
57
SECTION 7<br />
DISCUSSION<br />
Because only very limited ground truth is available for comparison with the results of this<br />
study, it is difficult to verify the accuracy of the stress maps that were produced. The 1987 gamma<br />
dose rate contours seem to correlate well with the later stress maps (Figure 5-18b), but the<br />
correlation is not perfect.<br />
There are several important points to be made concerning these contours. First, the contours<br />
correspond to the gamma dose rate at a time (1 May 1987) approximately 1 year after the accident<br />
and not to the accumulated dose. Second, these contours show only the gamma dose rate, not the<br />
alpha or beta particle dose rates. Third, as discussed above, the position or these contours is<br />
probably not very accurate. In fact, for these reasons, the forest response can be considered to<br />
provide a more accurate indication of the integrated dose to foliage than do the 1 May 1987 dose<br />
contours.<br />
There is one final noteworthy point concerning the position of the contours. It is possible that<br />
Soviet mitigation efforts were more effective over the damaged reactor than over the neighboring<br />
undamaged reactor. In that case, the center of the 100 mR/hr dose contour might really belong<br />
over the undamaged reactor. Mitigation efforts might also help to explain the southerly<br />
displacement of the dose contours relative to the areas of forest .showing the highest stress<br />
(Figure 5-18). Much of the area that first showed stress, and that was later removed, lies outside<br />
the 100 mR/hr dose contour.<br />
In lieu of ground truth, other methods for checking the results of this study can be applied to<br />
increase confidence in the results. Are the stress maps consistent with Soviet mitigation efforts in<br />
the area? Are the results self-consistent over time? Do the results of this study agree with those of<br />
other similar studies?<br />
The first of these questions was addressed in considerable detail in Section 5. Basically, most<br />
Soviet mitigation efforts in forested areas appear to have been consistent with the stress map. The<br />
only notable exception was the northern edge of the coniferous forest located to the south of the<br />
reactor. This area consistently showed stress in the stress maps, but no signs of mitigation were<br />
evident. Some possible explanations are (1) that the Soviets may believe that this area will recover,<br />
(2) the area may have low priority for current operations, or (3) the afflicted area is too large for<br />
mitigation to be practicable.<br />
With the exception of the winter months (when stress seems less detectable in the present<br />
analysis), the stress maps appear to be consistent with one another. The discolored area visible on<br />
the 29 April 1986 stress map is the same area that appeared stressed on the 8 May 1986 stress map.<br />
58
Similarly, the stressed area on 24 May 1986 includes all the stressed areas of 8 May 1986. The<br />
24 May 1986 and the 31 May 1986 stress maps show little difference. The stressed area of<br />
11 May 1987 subsumes that in all the previous images. In addition, apart from that for the winter<br />
months, the effects measured by the stress index increased fairly uniformly with time. Thus, the<br />
stress maps are consistent with one another.<br />
The last question is somewhat more difficult to address because of the lack of similar studies.<br />
Although the scope and even the data used for the study by Goldman and coworkers (Goldman,<br />
et al., 1987) were similar to that of the present work, the approaches and the results obtained were<br />
quite different. A comparison of the results obtained by these two studies is shown in Table 7-1.<br />
Most of the aspects of this comparison were discussed either above or in Section 5. In their study,<br />
a group of analysts (albeit experienced) attempted to identify stress visually by looking at<br />
photographic prints. Such a process is obviously subjective. Photographic prints are far from an<br />
ideal media, for the very production of the prints tends to involve variables usually controlled by<br />
subjective judgments (e.g., exposure and color balance). In addition, the dynamic range that can<br />
be achieved with the photographic process is not very large, resulting in a loss of information. In<br />
contrast, the results of this work were obtained largely by quantitative means. In this respect, we<br />
believe the methods and results of this study speak for themselves.<br />
Johnson (1989) has developed a phenomena-based image-enhancement transformation that is<br />
gaining popularity in the intelligence community. Designed to detect a specific spectral<br />
manifestation of stress (the decrease in reflectance in TM band 4 and concomitant increase in<br />
reflectance in band 5), the transformation squares each pixel's intensity in band 4 and divides that<br />
by the pixel's intensity in band 5 (i.e., 42/5). The transformed data are displayed as a color<br />
composite image (42/5 as red, band 4 as green, and band 5 as blue).<br />
Table 7-1. Comparison of present work with that of Goldman and coworkers (1987).<br />
Technique/result Present work Goldman and coworkers<br />
Algorithm type Change detection Normalized difference vegetation<br />
index<br />
Data preparation Image registration Spatially filtering image sharpening<br />
Method of injury assessment Quantitative (algorithmic) Qualitative (visual)<br />
First detectable change 8 May 1986 (12 days) 16 June 1986 (51 days)<br />
Latest significant change 11 May 1987 (1 year) 15 October 1986 (5.6 months)<br />
59
Care must be exercised in the interpretation of such single-date composite images. For<br />
example, one interpretation of a 42/5,4,5 composite of the 11 May 1987 image of the <strong>Chernobyl</strong><br />
area has vast new areas of forest beginning to show stress. Figure 7-1 shows this composite<br />
image. The area of stressed foliage indicated by our analysis (Figure 5-18) shows as a dark gray<br />
band in Figure 7-1. Johnson interprets the forest patches along the middle and upper left edge of<br />
the image as also being stressed. These purportedly newly stressed forested areas appear darker in<br />
the composite image than do other "unstressed" forest areas.<br />
This interpretation is based in part on the assumption that this pattern did not appear on any<br />
imagery dated before 11 May 1987 and that the darkened forest should appear the same as the<br />
undarkened. However, comparing the forest classification map (Figure 5-1), which was derived<br />
from the 6 June 1985 image, with the color composite in Figure 7-1 suggests that the pattern<br />
actually did exist a year before the accident. To make this comparison easier, the boundaries of<br />
forested areas classified as being similar to training site 4 are colored white; all other forested<br />
regions are classified as being similar to training site 3. Figure 7-2, showing this outline overlaid<br />
on the composite, indicates that these supposedly newly stressed areas seem to correspond to areas<br />
covered by a different forest type rather than to the radiation dose contours published by the<br />
Soviets. In other words, the apparent stress is likely an artifact of the analysis rather than an effect<br />
of the radiation release. Indeed, a similar composite image shown in Figure 7-3 derived from the<br />
preaccident image of 6 June 1985 shows the same darkening pattern as the 11 May 1987<br />
composite; therefore, the slightly dark appearing forest cannot be a new manifestation of radiation<br />
induced stress.<br />
Dose estimates based on our analysis are not presented in this interim report. Such estimates<br />
will depend on being able to relate the radiation dose either to the spectral manifestations of stress<br />
or to the times at which these manifestations first became detectable. Dose rate may complicate the<br />
relationship by affecting the spectral characteristics of injury as well as of the dose-response time.<br />
Also, the question of the relative proportion of beta dose versus gamma dose requires careful<br />
examination. These issues will be addressed in the final report for this project.<br />
60
Figure 7-1. Date: 11 May 1987, enhanced image; 42/5,4,5 false color.<br />
61
Figure 7-2.<br />
Boundaries of areas classified as similar to training site 4 overlaid on the<br />
11 May 1987 enhanced image; 42/5,4,5 false color.<br />
62
Figure 7-3. Date: 6 June 1985, preaccident image with the same image enhancement<br />
transformation shown in Figure 7-1. Darker forested areas in this enhancement<br />
correspond to a different forest type rather than to radiation damaged areas.<br />
63
SECTION 8<br />
CONCLUSION<br />
Perhaps the most significant conclusion that can be drawn from this work is that remotely<br />
collected, low-resolution multispectral data can be used to identify radiation stressed foliage and to<br />
monitor it quantitatively. Furthermore, this capability is aided by the availability of a sensitive<br />
change detection algorithm, Hyperscout, that can detect this stress very early and reliably. Early<br />
detection is critical for accurate dose estimates.<br />
When the data are in final form, a number of important results are anticipated. The first of<br />
these are estimates of the doses received by indigenous plant life. Secondly, the time history of the<br />
multispectral manifestations of various levels of radiation stress can be obtained, thus providing the<br />
first large-scale spectral measurements of the effects of a full range of radiation doses on conifers.<br />
This database would facilitate monitoring future accidents or even determining whether an accident<br />
had occurred. Such a database is not readily available from any other source.<br />
Much work remains to be done. The most important task remaining is to estimate radiation<br />
dose from the foliage response. Once the stress maps are finalized, the spectral changes<br />
corresponding to various levels of stress will be determined. These spectral histories, in turn, will<br />
be used to estimate the health of the conifers. Finally, these health histories will be used to infer<br />
dose estimates.<br />
Once dose estimates have been obtained, the process can be inverted. That is, the temporal<br />
aspects of the spectral manifestations of various levels of radiation stress can be extracted from the<br />
data.<br />
Continued monitoring of the area is recommended for at least two reasons. First, the<br />
identification of new areas showing stress would expand the area for which dose estimates are<br />
available. This new dose information might be incorporated to refine the estimates of dose levels<br />
received by human populations. Second, but not unrelated, lightly stressed forests should be<br />
monitored for evidence of recovery. Again, the primary goal is the refinement of dose estimates.<br />
Of equal importance is the collection of spectral information not available from other sources.<br />
Obtaining accurate dose estimates is the primary goal, and accurate determination of the dates<br />
of various stages of injury to the conifer community is required to obtain these estimates. There<br />
are available no high quality TM data for the period extending from 8 May 1986 to 24 May 1986,<br />
nor from 31 May 1986 to 15 October 1986. Yet, most of the spectral manifestations of the stress<br />
appear to have occurred during these two time intervals, while there seems to have been little<br />
change between 24 May 1986 and 31 May 1986. Because of the lack of suitable TM images, the<br />
question arises as to whether any other sources of multispectral data can be used to fill in these two<br />
64
critically important gaps. The algorithm used in this work can detect and map stress using data<br />
from any sensor, provided the data are sufficiently complete. Thus, both SPOT and MSS data<br />
should be considered potentially useful for filling in these gaps, provided that the analysis requires<br />
neither SWIR nor LWIR data.<br />
Another potentially useful approach to determining stress is to use TM data that was acquired at<br />
night. Night data should be much more sensitive to thermal stress. Thermal stress may be<br />
detectable much earlier than other forms of stress. If so, it could prove extremely valuable for<br />
detecting and monitoring the early stages of stress and, thus, for estimating dose.<br />
There are also two modifications in processing that should be considered. First, instead of<br />
using the same reference image for all analyses, as was done in this study, the reference date<br />
should be chosen to correspond to the season of the image being analyzed. This may significantly<br />
improve the accuracy of the resulting stress map. For example, the 15 October and 2 December<br />
1986 dates would be referenced to the 21 March 1986 date instead of the 6 June 1985 date.<br />
Unfortunately, snow cover present in the 21 March 1986 image may have a negative impact.<br />
The second processing modification involves using the first three components of the Tasselled<br />
Cap transformation (Crist, Laurin, and Cicone, 1986) instead of bands 7, 4, and 3. The fourth<br />
component, usually referred to as "Haze," might also allow the automatic elimination of distracting<br />
areas of apparent change attributable to cloud cover or haze.<br />
Extensions or alternatives to current algorithms should be investigated in the hope that even<br />
more sensitive algorithms may be found. One alternative is the use of neural networks to identify<br />
stress areas. Another promising area to explore is the simultaneous investigation of several<br />
(instead of just two) dates of imagery. This multitemporal processing might significantly improve<br />
the accuracy of the stress maps by removing the effects of noise and less than perfect registration.<br />
The final kinds of potential improvement are aesthetic. Some such kinds of improvement<br />
include the removal of areas contaminated by cloud, cloud-shadow, or haze from consideration,<br />
because these areas spuriously produce high stress values and confuse the presentation. In like<br />
manner, measures could be developed to the remove edge effects and to reduce noise by mode or<br />
other filtering.<br />
65
SECTION 9<br />
REFERENCES<br />
Asmolov, V.G., et. al., 1987<br />
"The <strong>Accident</strong> at the Chemobyl <strong>Nuclear</strong> Power Plant: One Year After," International Atomic<br />
Energy Agency, IAEA-CN-48/63.<br />
Crist, E.P., R. Laurin, and R.C. Cicone, 1986<br />
"Vegetation and Soils Information Contained in Transformed Thematic Mapper Data,"<br />
Proceedings of IGARSS 1986 Symposium, Zurich, 8-11 September 1986, pp.1465-1470.<br />
Engel, J., 1984<br />
"Thematic Mapper (TM) Instrument Description," LANDSAT-4 Science Investigations<br />
Summary, Vol. 1, NASA Conference Publication 2326, pp. 41-61.<br />
Estes, J. E., Ed., 1983<br />
Manual of Remote Sensing, Vol. II (American Society of Photogrammetry), pp. 1511-1513<br />
and 2136-2187.<br />
Goldman, M., et aL, 1987<br />
"Radiation Exposure Near <strong>Chernobyl</strong> Based on Analysis of Conifer Injury Using Thematic<br />
Mapper Satellite Images," University of California, Davis, California 95616 (unpublished).<br />
Hoffer, R. M., 1978<br />
"Biological and Physical Considerations in Applying Computer-Aided Analysis Techniques to<br />
Remote Sensor Data, Remote Sensing: The Quantitative Approach, edited by P. H. Swain and<br />
S. M. Davis (McGraw-Hill,.New York) pp. 227-289.<br />
Irons, J., 1985<br />
"An Overview of LANDSAT-4 and the Thematic Mapper," LANDSAT-4 Science<br />
Characterization Early Results, Vol. II, NASA Conference Publication 2355, pp. 15-46.<br />
Johnson, A. J., 1989<br />
Central Intelligence Agency, Washington, D.C. 20505, private communication.<br />
Matson, M., and J. Dozier, 1981<br />
"Identification of Subresolution High Temperature Sources Using a Thermal IR Sensor,"<br />
Photogrammetric Engineering and Remote Sensing, 47(9):1311-1318.<br />
Swain, P.H., 1978<br />
"Fundamentals of Pattern Recognition in Remote Sensing," Remote Sensing: The Quantitative<br />
Approach, edited by P. H. Swain and S. M. Davis (McGraw-Hill, New York) pp. 136-187.<br />
Westman, W.E. and C.V. Price, 1988<br />
"Spectral Changes in Conifers Subjected to Air Pollution and Water Stress: Experimental<br />
Studies," IEEE Transactions on Geoscience and Remote Sensing, 26(1): 11-21.<br />
Whicker, F.W., and L. Fraley, Jr., 1974<br />
Effects of Ionizing Radiation on Terrestrial Plant Communities, Vol. 4 of Advances in<br />
Radiation Biology, edited by J.T. Lett, H. Adler, and M. Zelle (Academic Press, New<br />
York).<br />
66
Wiebelt, J.A., and J.B. Henderson, 1976<br />
Techniques and Analysis of Thermal Infrared Camouflage in Foliated Backgrounds (U.S.<br />
Army Mobility Equipment Research and Development Command), AD A038186.<br />
Wolfe, W. L., and G. J. Zissis, eds., 1978<br />
The Infrared Handbook (Environmental Research Institute of Michigan), p. 3-140.<br />
67
DISTRIBUTION LIST<br />
DNA-TR-92-37-V2<br />
DEPARTMENT OF DEFENSE<br />
INTERSERVICE NUCLEAR WEAPONS SCHOOL<br />
ATTN: TTV<br />
ARMED FORCES RADIOBIOLOGY RSCH INST<br />
2 CYS ATTN: TTV 3416TH TTSO<br />
ATTN: DEPT OF RADIATION BIOCHEMISTRY<br />
ATTN: BHS<br />
JOINT DATA SYSTEM SUPPORT CTR<br />
ATTN: DIRECTOR ATTN: C-332<br />
ATTN: EXH<br />
ATTN: JNSV<br />
ATTN: MRA<br />
ATTN: PHY<br />
NATIONAL DEFENSE UNIVERSITY<br />
ATTN: RSD<br />
ATTN: ICAF TECH LIB<br />
ATTN: SCIENTIFIC DIRECTOR<br />
ATTN: NWCLB-CR<br />
ATTN: TECHNICAL LIBRARY<br />
ATTN: LIBRARY<br />
ATTN: STRAT CONCEPTS DIV CTR<br />
ASSISTANT SECRETARY OF DEFENSE<br />
INTERNATIONAL SECURITY POLICY<br />
NET ASSESSMENT<br />
ATTN: NUC FORCES & ARMS CONTROL<br />
ATTN: DOCUMENT CONTROL<br />
ASSISTANT TO THE SECRETARY OF DEFENSE<br />
ATTN: EXECUTIVE ASSISTANT<br />
ATTN: MIL APPL C FIELD<br />
OFC OF MILITARY PERFORMANCE<br />
ASSESSMENT TECHNOLOGY<br />
ATTN: F HEGGE<br />
DEFENSE INTELLIGENCE AGENCY<br />
PROGRAM ANALYSIS & EVALUATION<br />
ATTN: DB<br />
ATTN: NAVAL FORCES<br />
5 CYS ATTN: DB-4 RSCH RESOURCES DIV ATTN: STRATEGIC PROGRAMS & TNF<br />
ATTN: DB-5C<br />
STRATEGIC AND THEATER NUCLEAR FORCES<br />
ATTN: DB-6B<br />
ATTN: DR E SEVIN<br />
ATTN: DB-6E<br />
ATTN: DR SCHNEITER<br />
ATTN: DIAs'VPA-2<br />
ATTN: DIW-4<br />
THE JOINT STAFF<br />
ATTN: DN<br />
ATTN: JKAC<br />
ATTN: DT<br />
ATTN:JLT<br />
ATTN: OFFICE OF SECURITY<br />
ATTN: JPEP<br />
ATTN: OS<br />
DEFENSE INTELLIGENCE COLLEGE<br />
THE JOINT STAFF<br />
DEFENSE INEIGCE CATTN:<br />
ED30 J-3 STRATEGIC OPS<br />
ATTN:<br />
DIV<br />
DIC/RTS-2<br />
ATTN: J-3 SPECIAL OPERATIONS<br />
ATTN: DIC/2C<br />
ATTN: J-8 S CONLIN<br />
DEFENSE LOGISTICS AGENCY<br />
ATTN: JAD/SFD<br />
ATTN: COMMAND SECURITY OFC<br />
ATTN: JSOA<br />
DEFENSE NUCLEAR AGENCY<br />
U S EUROPEAN COMMAND/ECJ-6-DT<br />
ATTN: CID<br />
ATTN: ECJ-6<br />
ATTN: DFRA JOAN MA PIERRE<br />
U S EUROPEAN COMMAND/ECJ2-T<br />
ATTN: NANF<br />
ATTN: ECJ2-T<br />
ATTN: NASF<br />
ATTN: OPNA<br />
U S EUROPEAN COMMAND/ECJ3-CCD<br />
ATTN: OPNS<br />
ATTN: RAEE<br />
ATrN: ECJ-3<br />
20 CYS ATTN: RARP U S EUROPEAN COMMAND/ECJ4-LW<br />
2 CYS ATTN: TITL ATTN: ECJ4-LW<br />
DEFENSE TECHNICAL INFORMATION CENTER<br />
U S EUROPEAN COMMAND/ECJ5-N<br />
2 CYS ATTN: DTICIFDAS ATTN: ECJ5-N<br />
FIELD COMMAND DEFENSE NUCLEAR AGENCY<br />
ATTN: FCPR<br />
ATTN: FCPRT<br />
ATTN: NUC SECURITY<br />
FIELD COMMAND DEFENSE NUCLEAR AGENCY<br />
ATrN: FCNM<br />
U S EUROPEAN COMMAND/ECJ7/LW<br />
ATiN: ECJ-7 LW<br />
UNDER SEC OF DEFENSE FOR POLICY<br />
ATTN: DUSP/P<br />
ATTN: USD/P<br />
Dist-1
DNA-TR-92-37-V2 (DL CONTINUED)<br />
DEPARTMENT OF THE ARMY<br />
COMBAT MATERIAL EVAL ELEMENT<br />
ATTN: SECURITY ANALYST<br />
DEP CH OF STAFF FOR OPS & PLANS<br />
ATTN: DAMO-SWN<br />
ATTN: DAMO-ZXA<br />
HARRY DIAMOND LABORATORIES<br />
ATrN: SLCH-NW-RS JOSIP SOLN<br />
ATTN: SLCIS-IM-TL<br />
JOINT SPECIAL OPERATIONS COMMAND<br />
ATTN: J-2<br />
ATTN: J-5<br />
NUCLEAR EFFECTS DIVISION<br />
ATTN: STEWS-NE-T<br />
U S ARMY AIR DEFENSE ARTILLERY SCHOOL<br />
ATTN: COMMANDANT<br />
U S ARMY ARMAMENT RSCH DEV & ENGR CTR<br />
ATTN: DRDAR-LCN-F<br />
U S ARMY ARMOR SCHOOL<br />
ATTN: ATSB-CTD<br />
ATTN: TECH LIBRARY<br />
U S ARMY MATERIAL COMMAND<br />
ATTN: DRCDE-D<br />
U S ARMY NUCLEAR & CHEMICAL AGENCY<br />
ATTN: MONA-NU<br />
4 CYS ATTN: MONA-NU DR D BASH<br />
U S ARMY TEST & EVALUATION COMMAND<br />
ATTN: STECS-NE<br />
U S ARMY WAR COLLEGE<br />
ATTN: LIBRARY<br />
ATTN: STRATEGIC STUDIES INSTITUTE<br />
U S MILITARY ACADEMY<br />
ATTN: BEHAVORIAL SCI & LEADERSHIP<br />
ATTN: COL J G CAMPBELL<br />
ATTN: SCIENCE RESEARCH LAB<br />
US ARMY MATERIEL SYS ANALYSIS ACTVY<br />
ATTN: DRXSY-DS<br />
USA SURVIVABILITY MANAGMENT OFFICE<br />
ATTN: SLCSM-SE J BRAND<br />
USACACDA<br />
ATTN: ATZL-CAD-N<br />
DEPARTMENT OF THE NAVY<br />
U S ARMY BALLISTIC RESEARCH LAB<br />
DEPARTMENT OF THE NAVY<br />
ATTN: SLCBR-D<br />
ATTN: PMS-423<br />
ATTN: SLCBR-DD-T<br />
ATTN: SEA-06GN<br />
ATTN: SLCBR-TB<br />
ATTN: SLCBR-VL-I DR KLOPCIC<br />
MARINE CORPS<br />
ATTN: SLCBR-VL-V<br />
ATTN: CODE PPO<br />
ATTN: PSIG RASP<br />
U S ARMY COMBAT SYSTEMS TEST ACTIVITY<br />
ATTN: JCOHN GERDES<br />
NAVAL OCEAN SYSTEMS CENTER<br />
ATTN: MIKE STANKA ATTN: CODE 9642-B<br />
U S ARMY COMD & GENERAL STAFF COLLEGE<br />
ATTN: ATZL-SWJ-CA<br />
ATTN: ATZL-SWT-A<br />
U S ARMY CONCEPTS ANALYSIS AGENCY<br />
ATTN: TECHNICAL LIBRARY<br />
NAVAL PERSONNEL RES & DEV CENTER<br />
ATTN: CODE P302<br />
NAVAL POSTGRADUATE SCHOOL<br />
ATTN: CODE 1424 LIBRARY<br />
NAVAL RESEARCH LABORATORY<br />
U S ARMY FIELD ARTILLERY SCHOOL ATTN: CODE 1240<br />
ATTN: ATSF-CD ATIN: CODE 2627<br />
U S ARMY FORCES COMMAND<br />
NAVAL SURFACE WARFARE CENTER<br />
ATTN: AF-OPTS ATTN: CODE F-31<br />
ATTN: G RIEL<br />
U S ARMY FOREIGN SCIENCE & TECH CTR<br />
ATTN: C WARD<br />
U S ARMY INFANTRY CENTER<br />
ATTN: ATSH-CD-CSO<br />
U S ARMY ITAC<br />
ATTN: IAX-Z<br />
U S ARMY LABORATORY COMMAND<br />
ATTN: DIRECTOR<br />
ATTN: DR D HODGE<br />
NAVAL TECHNICAL INTELLIGENCE CTR<br />
ATTN: NTIC-DA30<br />
NAVAL WAR COLLEGE<br />
ATTN: CODE E-1 1<br />
ATTN: CTR FOR NAV WARFARE STUDIES<br />
ATTN: DOCUMENT CONTROL<br />
ATTN: LIBRARY<br />
ATTN: STRATEGY DEPT<br />
NAVAL WEAPONS EVALUATION FACILITY<br />
ATTN: CLASSIFIED LIBRARY<br />
Dist-2
DNA-TR-92-37-V2 (DL CONTINUED)<br />
NUCLEAR WEAPONS TNG GROUP, ATLANTIC<br />
ATTN: CODE 222<br />
ATTN: DOCUMENT CONTROL<br />
NUCLEAR WEAPONS TNG GROUP, PACIFIC<br />
ATTN: CODE 32<br />
LAWRENCE LIVERMORE NATIONAL LAB<br />
ATTN: Z DIVISION LIBRARY<br />
LOS ALAMOS NATIONAL LABORATORY<br />
ATTN: D STROTTMAN<br />
ATTN: REPORT LIBRARY<br />
OFFICE OF CHIEF OF NAVAL OPERATIONS<br />
MARTIN MARIETTA ENERGY SYSTEMS INC<br />
ATTN: NIS-22<br />
ATTN: B SANTORO<br />
ATTN: NOP 06D<br />
ATTN: G KERR<br />
ATTN: NOP 403<br />
ATITN: J WHITE<br />
ATTN: NOP 50<br />
ATTN: W RHOADES<br />
ATTN: NOP 60<br />
ATTN: NOP 60D<br />
SANDIA NATIONAL LABORATORIES<br />
ATTN: NOP 603 ATTN: TECH LIB 3141<br />
ATTN: NOP 91<br />
ATTN: OP 654<br />
OTHER GOVERNMENT<br />
ATTN: PMS/PMA-423<br />
CENTRAL INTELLIGENCE AGENCY<br />
OPERATIONAL TEiT & EVALUATION FORCE<br />
ATTN: COMMANDER<br />
PLANS, POLICY & OPERATIONS<br />
ATTN: CODE-P<br />
ATTN: CODE-POC-30<br />
TACTICAL TRAINING GROUP, PACIFIC<br />
ATTN: COMMANDER<br />
DEPARTMENT OF THE AIR FORCE<br />
ACADEMY LIBRARY DFSELD<br />
ATTN: LIBRARY<br />
AFIS/INT<br />
ATTN: INT<br />
AIR UNIVERSITY<br />
ATTN: STRATEGIC STUDIES<br />
AIR UNIVERSITY LIBRARY<br />
ATTN: AUL-LSE<br />
ATTN: LIBRARY<br />
ATTN: COUNTER TERRORIST GROUP<br />
ATTN: DIRECTOR OF SECURITY<br />
ATITN: MEDICAL SERVICES<br />
ATTN: NIO-T<br />
ATTN: N10 STRATEGIC SYS<br />
FEDERAL EMERGENCY MANAGEMENT AGENCY<br />
ATTN: CIVIL SECURITY DIVISION<br />
ATTN: G ORRELL NP-CP<br />
ATTN: R SANDS<br />
U S DEPARTMENT OF STATE<br />
ATTN: PM/STM<br />
U S NUCLEAR REGULATORY COMMISSION<br />
ATTN: DIR DIV OF SAFEGUARDS<br />
ATTN: S YANIV<br />
DEPARTMENT OF DEFENSE CONTRACTORS<br />
ARES CORP<br />
ATTN: A DEVERILL<br />
HORIZONS TECHNOLOGY, INC<br />
ATTN: F GREY<br />
ASSISTANT CHIEF OF STAFF<br />
2 CYS ATTN: AF/SAMI HORIZONS TECHNOLOGY, INC<br />
ATTN: J MARSHALL-MIES<br />
ASSISTANT CHIEF OF THE AIR FORCE<br />
ATTN: SAF/AQA<br />
DEPUTY CHIEF OF STAFF PLANS & OPERS<br />
ATTN: AFXOOSS<br />
FOREIGN TECHNOLOGY DIVISION<br />
ATTN: CCN<br />
ATTN: SDA<br />
KAMAN SCIENCES CORP<br />
ATTN: DASIAC<br />
KAMAN SCIENCES CORPORATION<br />
ATTN: R STOHLER<br />
KAMAN SCIENCES CORPORATION<br />
ATTN: DASIAC<br />
TACTICAL AIR COMMAND/XPSC<br />
ATTN: TAC/DOA<br />
USAF SCHOOL OF AEROSPACE MEDICINE<br />
ATTN: RADIATION SCIENCES DIV<br />
DEPARTMENT OF ENERGY<br />
DEPARTMENT OF ENERGY<br />
ATTN: DR T JONES<br />
LOCKHEED MISSILES & SPACE CO, INC<br />
ATTN: WE-YOUNG WOO<br />
LOGICON R & D ASSOCIATES<br />
ATTN: DOCUMENT CONTROL<br />
ATTN: DOUGLAS C YOON<br />
LOGICON R & D ASSOCIATES<br />
ATTN: S WOODFORD<br />
Dist-3
DNA-TR-92-37-V2 (DL CONTINUED)<br />
MICRO ANALYSIS AND DESIGN<br />
ATTN: R LAUGHERY<br />
MISSION RESEARCH CORP<br />
ATTN: DR NEIL GOLDMAN<br />
SCIENCE APPLICATIONS INTL CORP<br />
ATTN: B BENNETT<br />
ATTN: D BAREIS<br />
ATTN: FOSTER<br />
ATTN: J MCGAHAN<br />
PACIFIC-SIERRA RESEARCH CORP<br />
ATTN: J PETERS<br />
2 CYS ATTN: G ANNO<br />
12 CYS ATTN: G E MCCLELLAN SCIENCE APPLICATIONS INTL CORP<br />
ATTN: H BRODE<br />
ATTN: R CRAVER<br />
2 CYS ATTN: R N DEWITT<br />
2 CYS ATTN: T H HEMMER SCIENCE APPLICATIONS INTL CORP<br />
ATTN: JOHN A SHANNON<br />
PACIFIC-SIERRA RESEARCH CORP<br />
ATTN: D GORMLEY<br />
TECHNICO SOUTHWEST INC<br />
2 CYS ATTN: G MCCLELLAN ATTN: S LEVIN<br />
SCIENCE APPLICATIONS INTL CORP<br />
ATTN: D KAUL<br />
ATTN: E SWICK<br />
ATTN: L HUNT<br />
ATTN: R J BEYSTER<br />
ATTN: W WOOLSON<br />
TRW OGDEN ENGINEERING OPERATIONS<br />
ATTN: D C RICH<br />
TRW SPACE & DEFENSE SECTOR<br />
ATTN: DR BRUCE WILSON<br />
UNIVERSITY OF CINCINNATI MEDICAL CENTER<br />
ATTN: E SILBERSTEIN<br />
Dist-4
AD-A260 167<br />
A lexandria, VA 22310-3398 l,*<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency<br />
Alexandria, VA 22310-.3398<br />
DNA-TR-92-37-V3<br />
<strong>Chernobyl</strong> Doses,<br />
Volume 3-Habitat and Vegetation Near the <strong>Chernobyl</strong><br />
<strong>Nuclear</strong> Reactor Station<br />
DTIC<br />
~ ELECTF.<br />
Elizabeth L. Painter<br />
i IN•9 199EIF<br />
F. Ward Whicker JAN % 93f<br />
Pacific-Sierra Research Corporation I U<br />
12340 Santa Monica Boulevard C<br />
Los Angeles, CA 90025-2587<br />
January 1993<br />
Technical Report<br />
CONTRACT No. DNA 001-87-C-0104<br />
I<br />
Approved for public release;<br />
distribution Is unlimited.<br />
93-01259
Destroy this report when it is no longer needed. Do not<br />
return to sender.<br />
PLEASE NOTIFY THE DEFENSE NUCLEAR AGENCY,<br />
ATTN: CSTI, 6801 TELEGRAPH ROAD, ALEXANDRIA, VA<br />
22310-3398, IF YOUR ADDRESS IS INCORRECT, IF YOU<br />
WISH IT DELETED FROM THE DISTRIBUTION LIST, OR<br />
IF THE ADDRESSEE IS NO LONGER EMPLOYED BY YOUR<br />
ORGANIZATION.<br />
*N4o
DISTRIBUTION LIST UPDATE<br />
This mailer is provided to enable DNA to maintain current distribution lists for reports. (We would<br />
appreciate your providing the requested information.)<br />
o Add the individual listed to your distribution list<br />
o Delete the cited organization/individual,<br />
o Change of address.<br />
NOTE:<br />
Please return the mailing label from<br />
Fthe document so that any additions,<br />
changes, corrections or deletions can<br />
be made easily.<br />
NAME:<br />
ORGANIZATION:<br />
OLD ADDRESS<br />
CURRENT ADDRESS<br />
Z<br />
TELEPHONE NUMBER: ( )<br />
cc DNA PUBLICATION NUMBER/TITLE CHANGES/DELETIONS/ADDITIONS, etc.)<br />
0' (Attach Sheot f more Space is fequftd)<br />
Z<br />
UJ<br />
WI<br />
I'-<br />
• : DNA OR OTHER GOVERNMENT CONTRACT NUMBER:<br />
CERTIFICATION OF NEED-TO-KNOW BY GOVERNMENT SPONSOR (if other than DNA):<br />
SPONSORING ORGANIZATION:<br />
CONTRACTING OFFICER OR REPRESENTATIVE:<br />
SIGNATURE:
DEFENSE NUCLEAR AGENCY<br />
ATTN: TITL<br />
6801 TELEGRAPH ROAD<br />
ALEXANDRIA, VA 22310-3398<br />
DEFENSE NUCLEAR AGENCY<br />
ATTN: TITL<br />
6801 TELEGRAPH ROAD<br />
ALEXANDRIA, VA 22310-3398
REPORT DOCUMENTATION PAGE<br />
Form ppoved•<br />
I OMB No. 0704-0188<br />
PUli tO ri In. Vim Of oou•o on ofnto i a li' s•l ooa d io wvrageo I hour per respaonee riclding tio o orie lr rov@iwig Wn ucn. oeawcwg eom ng dala sources.<br />
gw O arni Wi m V ftain date neessided, arid coimpil arnd reviewing Vi coelcrti of iiff n n, Sand comen * regard•ig V wrdon eeanite or may o~e aepect of ow<br />
l Cbibimaion. .c.ud.lig t sggeanime r• g V1 lulden W i hlin Hedqua•rer Soei Diectorate fr mnborrmabon Operalns arid <strong>Reports</strong>. 1215 Jefterson<br />
Datlighway. Suite 1204. Arbingtorn. VA 220-40. gird W Vie Office of Managementl<br />
and Budget. Paperwor Reduction Proiect (0704-oiU). Washingtion. DC 20603<br />
1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED<br />
1 930101 Technical 870929- 920228<br />
4. TITLE AND SUBTITLE 5. FUNDING NUMBERS<br />
<strong>Chernobyl</strong> Doses C - DNA 001-87-C-0104<br />
Volume 3-Habitat and Vegetation Near the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station PE - 62715H<br />
6. AUTHOR(S) PR -RM<br />
TA -RH<br />
Elizabeth L Painter and F. Ward Whicker<br />
WU - DH026130<br />
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION<br />
REPORT NUMBER<br />
Pacific-Sierra Research Corporation<br />
12340 Santa Monica Boulevard PSR Report 2251<br />
Los Angeles, CA 90025-2587<br />
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING<br />
AGENCY REPORT NUMBER<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency<br />
6801 Telegraph Road DNA-TR-92-37-V3<br />
Alexandria, VA 22310-3398<br />
RARP/Young<br />
11. SUPPLEMENTARY NOTES<br />
This work was sponsored by the <strong>Defense</strong> <strong>Nuclear</strong> Agency under RDT&E RMC Code B4662D RM RH 00038<br />
STRP 3500A 25904D.<br />
12a. DISTRIBUTION/AVAILABIUTY STATEMENT 12b. DISTRIBUTION CODE<br />
Approved for public release; distribution is unlimited.<br />
13. ABSTRACT (Maximum 200 words)<br />
This volume presents a detailed exposition on the soils, climate, and vegetation of the Poles'ye region of<br />
Ukraine and Belorussia with emphasis on the area around the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station. This data<br />
provides background for interpretation of multispectral satellite imagery of the area. Volume I uses these<br />
images and the information of this report to analyze the radiation response of the canopy of the coniferous forests<br />
in the immediate vicinity of the reactor station after the accident of 26 April 1986.<br />
14. SUBJECT TERMS 15. NUMBER OF PAGES<br />
Chemobyl Forest Damage Landsat 90<br />
Change Detection Conifer Stress Fallout 16. PRICE CODE<br />
Ionizing Radiation Multispectral Imagery<br />
17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. UMITATION OF ABSTRACT<br />
OF REPORT OF THIS PAGE OF ABSTRACT<br />
UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED SAR<br />
NSN 7S0-220-6600<br />
Form 296 (Rv.2-B19)<br />
U NISN byI A Va.<br />
290-102
UNCLASSIFIED<br />
SECVWV CA3S W•TE *% OF ThIS MOE<br />
CLASSIFIED BY:<br />
N/A since Unclassified.<br />
DECLASSIFY ON:<br />
N/A since Unclassified.<br />
i fl•, OAWIFUAl OF TM MW<br />
UNCLASSIFIED
PREFACE<br />
This volume is the third in a series of three volumes composing the final report to the <strong>Defense</strong><br />
<strong>Nuclear</strong> Agency (DNA) for contract DNA001-87-C-0104, <strong>Chernobyl</strong> Doses. It was prepared at<br />
Colorado State University from locally available literature under a consulting agreement with<br />
Pacific-Sierra Research Corporation (PSR). It was generated as a topical report for that contract,<br />
but is being published as a volume of the final report. It provides a detailed exposition on the<br />
soils, climate, and vegetation of the Poles'ye region of the Ukraine and Belorussia with emphasis<br />
on the area around the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station. This data provides background for the<br />
interpretation of satellite imagery of the area. Volume 2, Conifer Stress Near <strong>Chernobyl</strong> Derived<br />
from Landsat Imagery, describes the acquisition and processing of satellite multispectral imagery<br />
of the area containing the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station and presents the exploratory analysis<br />
of the imagery using PSR's proprietary Hyperscout m change detection algorithm. Volume 1,<br />
Analysis of Forest Canopy Radiation Response from Multispectral Imagery and the Relationship<br />
to Doses, presents the analytical work that connects these multispectral observations of pine forests<br />
in the images to the nuclear radiation dose received by the trees as a consequence of the reactor<br />
accident of 26 April 1986 and summarizes other work conducted for the contract.<br />
ST ! S<br />
6t&Ltt<br />
Wt4 flit 0<br />
Uzstuaoji* JUD__Strtdit on/ jaw/<br />
Availability Codes<br />
[Avail sad/or<br />
rDist<br />
Special
CONVERSION TABLE<br />
Conversion factors for U.S. customary to metric (SI) units of measurement<br />
TO Convert Proma TO multiply<br />
angstroin meters 1m) 1.000 000 X E-10<br />
atmosphere (normal) kilo pascal (kPa) 1.013 25 X E+2<br />
bar kilo pascal (kPa) 1.000 000 X E+2<br />
barn meter' (in2) 1.000 000 X E-28<br />
British Thermal unit (tbrobnia) joule WJ 1.054 350 X E+3<br />
calori (thermochemical) joule (J) 4.184 000<br />
cal (tbermowmchemcl)/cm2 mega jDUle/M2(Mj/m2) 4.184 000 X E-2<br />
curse gi4p becquerel (G~qr- 3.700 000 X E+I<br />
degree (angle) radian (rad) 1.745 329 X E-2<br />
degree Fahrenheit degree kelvin (K) t 3 1 u(tf.+ 459.67)/ 1.8<br />
electron volt joule(JW 1.602 19 XE-19<br />
asg joule (J) 1.0O00O00xE-"<br />
erg/second watt MW 1.000 000 X "-<br />
foot meter (M) 3.046000 X E-1<br />
foot-pound-force joule (J) 1.3558616<br />
gallon (U.S. liquid) meter 3 EM3) 3.765 412 X E-3<br />
inch meter 1m) 2.540 000 X E-2<br />
jerk joule (J) 1.000 000 X E+9<br />
joule/kilogrm (J/Kg) (radiation dose<br />
absorbed) Gray 10y) 1.000 000<br />
kilotons terajoules 4.183<br />
kip (10001b1) newton (N) 4.448 222XEZ+3<br />
hip/Inch 2 (kal) kilo pascal (kPa) 6.694 757 X E.3<br />
ktap newton-second/rn' (N-s/rn') 1.000 000 X E+2<br />
Micron meter (m) 1.000 000 X E-6<br />
mil meter (m) 2.540 000 X "-<br />
mile (International) mete 4m) 1.609 344 X +3.<br />
ounce kilogram (kg) 2.634 952 X 3-2<br />
pound-force (bf avoirdupois) newton (N) 4.448 222<br />
pound-force inch newton-meter Min) 1.129 648 XE-1<br />
pound-frc/inch newton/meter Motr) 1.751 266 X E.2<br />
pound-force/foot' kilo pascal (kPa) 4.766 026 X E-2<br />
pound-forc/inch' (psi) kilo paacal (kPa) 6.694 757<br />
pound-mass MIm avoirdupois) kilogram (kg) 4.535 924 X E-1<br />
pound-mass-lbot2 (xmomet of inertial kilogram-eter' (kgrn'3 4.214011l X E-2<br />
pound-rnaslhst kilogram/meter' (hg/rn') 1.6018464 X E+1<br />
rdaflt " w don absorbed) Gray EGyr' 1.000 000 X E-2<br />
roenie.coulomb/kiogram (C/hg) 2.579 760 X E-4<br />
sakeh second (s) 1.000 000 X E-8<br />
48kilogram (kg) 1.459 390 X E3.1<br />
eory (mmo HS, 0C) kilo pascal (kPal 1.333 22 X E- I<br />
*Ybe b soquere (Dq) Is tde 91 unit of radloac~tvty Sp a 1 event/s.<br />
I-Tbe Grey (Qy) in the 81 unit of absorbed radiation.<br />
iv
TABLE OF CONTENTS<br />
Section<br />
Page<br />
PREFACE ................................................................................. iim<br />
CONVERSION TABLE ............................................................... iv<br />
1 INTRODUCTION ....................................................................... 1<br />
2 GEOGRAPHY ........................................................................... 2<br />
3 SOILS ..................................................................................... 4<br />
4 CLIMATE ................................................................................. 7<br />
5 VEGErATION ........................................................................... 9<br />
6 SUCCESSION ......................................................................... 19<br />
7 BOG DRAINAGE AND RECLAMATION ....................................... 21<br />
8 CROPS .................................................................................. 22<br />
9 REFERENCES ........................................................................... 23<br />
Appendices<br />
A TREES AND SHRUBS ............................................................... A-1<br />
B<br />
HERBS .................................................................................... B-i<br />
C CROPPLANTS .......................................................................... C-1<br />
v
SECTION 1<br />
INTRODUCTION<br />
This report was prepared to support the use of satellite multispectral imagery to detect changes<br />
in the vegetation near the <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station that were caused by the release of<br />
radioactive material during the accident of 26 April 26 1986. It is based on a review of literature<br />
available locally at Colorado State University.<br />
The review covered information on the Poles'ye region of the Ukraine and Belorussia,<br />
republics of the former Soviet Union. This area contains the 30 kilometer danger zone that was<br />
evacuated in the aftermath of the accident. The information is intended to aid in the interpretation<br />
of features and vegetation types in the imagery of the area within the danger zone.<br />
Section briefly describes the geography, Section 3 the soils, and Section 4 the climate of the<br />
Poles'ye, with emphasis on the area near the reactor station. Section 5 presents an extensive<br />
discussion of vegetation of the region. Sections 6, 7, and 8 describe succession, bog drainage,<br />
and typical crops, respectively. Appendixes A. B, and C provide descriptions of the more<br />
important species of trees and shrubs, herbs, and crop plants, respectively.
SECTION 2<br />
GEOGRAPHY 1<br />
The <strong>Chernobyl</strong> <strong>Nuclear</strong> Reactor Station (NRS) and much of the area within 30 kilometers (kin)<br />
around it are located in the Poles'ye 2 region of the Ukrainian and Belorussian 3 Soviet Socialist<br />
Republics. The power plant itself and the cities of Pripyat 4 and <strong>Chernobyl</strong> are in northeastern<br />
Kievskaya Oblast, in the interfluve of the Pripyat and Teterev Rivers, in the Ukrainian SSR. It lies<br />
at about 51 0 12'N longitude 30 0 8'E latitude.<br />
The Poles'ye lies in the Pripyat River basin and in part of the Dnepr 5 River basin and includes<br />
the area called the Pripyat marshes or bogs. It is a vast lowland (more than 13 million hectares) on<br />
a large, morainal outwash and outwash-alluvial plain on the Russian platform. It is bounded on<br />
the north, east, and south by greater rises in elevation than can be found within the region itself.<br />
The area has a very nonhomogeneous geological structure. From west to east, it occupies the<br />
various geostructural regions of the Russian platform: the Galician-Volyn lowland, the Ukrainian<br />
crystalline shield, the Dnepr-Donestsk depression, with the east approaching the Voronezh<br />
crystalline shield.<br />
The terrain is almost uniformly low, broken only by low moraine hills, sandy knolls or<br />
mounds (usually old dunes), and slightly elevated plains. The relief from the center to the edges of<br />
the basin is only 50 to 150 meters (m). Some areas possess eolian relief, often in the form of<br />
parabolic, west-facing dunes, but maximum heights within the Poles'ye are only 200 m. The<br />
interfluve of the Pripyat and Teterev Rivers has an unusual morainal-hilly relief with some<br />
relatively deep erosion. Poorly drained sandy plains are located between the moraine hills.<br />
The Poles'ye is often divided into three sections: the western, the central (or right bank-on<br />
the right bank of the Dnepr River), and the eastern (or left bank). In the Ukraine, the central<br />
section is divided into the Kiev Poles'ye and the Zhitomir Poles'ye. The Kiev Poles'ye lies on the<br />
1 Geological information has been synthesized from Keller 1927, Berg 1950, Akademiya Nauk SSSR 1963, Lydolph<br />
1964, Fridland 1976, Golovina et al. 1980, Lysenko and Golovina 1982, USSR 1987.<br />
2 polesye, Poles'e, Poles'ya, Polesie, Polessie, Polesiye, Pollessic lowlands, Polyesye (Where there are multiple<br />
spellings of a Russian name in English. the one most frequently encountered in Soviet literature has been used. All<br />
others are listed in a foomote.)<br />
3 Byelorussian or White Russian.<br />
4 pripyan. Prip'az. Pripiat, Pripet.<br />
2
middle Dnepr slope, in the interfluve of the Pripyat and Teterev Rivers, extending east to the Dnepr<br />
River, and includes the <strong>Chernobyl</strong> NRS and the area immediately around it. The Zhitomir is west<br />
of the Kiev Poles'ye.<br />
3
SECTION 3<br />
SOILS 6<br />
Glaciofluvial plains are formed by deposition from thaw waters of glaciers. They consist of<br />
outwash and broken-up moraine materials. The soils in the Poles'ye formed in a humid climate on<br />
glaciofluvial plains from a blanket of unconsolidated Quaternary outwash deposits of sand and<br />
loamy sand over- and underlying a terminal glacial moraine. Glaciofluvial sediments are sandy or<br />
gravelly-pebbly stratified sediments of flowing glacial waters. Outwash plains are relatively<br />
smooth sandy surfaces deposited near the edges of continental ice. The thickness of these deposits<br />
range from 1.5 m in the southeast to 150 m in the northwest Moraines are deposits of continental<br />
ice, consisting mostly of unstratified, unsorted, nonuniform materials of varying mechanical<br />
composition. The Poles'ye terminal moraine is mostly loamy sand. Medium and clay-loam lake<br />
deposits are rare in the Poles'ye and loesses, even rarer.<br />
The parent materials are course-textured, very bouldery, and gravelly; fine particles are usually<br />
leached or washed away. In the Kievskaya Oblast, glacial waters were active for a long time<br />
during the formation of the parent material. The soils generally lack carbonates in the parent<br />
materials. Most of the soils are at least periodically waterlogged (with spring and autumn floods,<br />
and sometimes after prolonged summer rains).<br />
Most of the Poles'ye soils are poor in humus; soils in the western and central Poles'ye often<br />
have I to 1.5 percent humus and can have less than 1 perceat. The soils are mostly noncalcareous.<br />
They generally are acid, with a pH of 4.5 to 5.5, and are low in available nutrients, including<br />
boron, zinc, phosphorus, potassium, and nitrate.<br />
The soils in the Poles'ye are podzolic, 7 peaty bog, and peaty-gley meadow soils. More than<br />
75 percent of the soils are meadow or bog soils. Soils on floodplains may be bog, meadow, or<br />
podzols. Parent materials around the <strong>Chernobyl</strong> NRS is from the Ukrainian crystal shield and is<br />
mainly sands or sandy-loams. Many of the soils in the northeastern part of the Kiev Poles'ye, in<br />
the interfluve of the Pripyat and Teterev Rivers, are weakly podzolic (frequently gleyed) sods.<br />
6Descriptions, distributions, and definitions of soils have been synthesized from Sukachev 1928, Berg 1950,<br />
Vilenskii 1960, Lydolph 196%, Fridland 1965, Symons 1972, Brady 1974, Krupskiy et al. 1970, Walter 1978,<br />
Smith 1980, Golovina et al. 1980, Oleynik 1981, Lysenko and Golovina 1982, USSR 1986, USSR 1987.<br />
7 The soil classification system found in the Soviet literature surveyed for this report is not the system currently<br />
being used in much of the western world. Podzols have been placed in a group called "spodosols," most bog soils in<br />
"histosols," and most gley soils in "inceptisols" (Brady 1974, Smith 1980). This report continues to use the<br />
classification used in the USSR. Many of the old soil names come originally from Soviet soil literature. 'Podzol"<br />
is a Russian word, and "gley" is Ukrainian.<br />
4
The soils of the area frequently are flooded or contain excess soil water, at least part of the<br />
year. There are three sources of excess soil water in the Poles'ye: (1) in the spring and often in<br />
the autumn from stream overflow; (2) from a groundwater table that is very close to the soil<br />
surface; and (3) from high amounts of atmospheric precipitation. The banks of the streams and<br />
floodplain terraces are both very low. In the spring and autumn, and after heavy summer rains,<br />
streams overflow onto floodplain terraces and into interfluvial moraine plains between neighboring<br />
streams, and water from one stream sometimes passes into another.<br />
Much of the area around the <strong>Chernobyl</strong> NRS is poorly drained and contains a large block of<br />
low-lying wetlands with bog soils. Because bog soils receive excessive moisture for the greater<br />
part of the year, are sometimes covered with shallow water, and have poor drainage, they become<br />
peaty.<br />
Podzols have a very thin organic layer on top of a gray, leached eluvial layer, which in turn<br />
overlies a dark brown horizon. Podzols develop in cool, moist climates under coniferous forest.<br />
Podzolization occurs when rainfall in forests is sufficient to carry away many elements, especially<br />
calcium, magnesium, potassium, iron, and aluminum. Becuase the needles of coniferous trees are<br />
acid, conifers generally return insufficient bases to the surface soil, which becomes acid. The<br />
distribution of podzols is patchy in the Poles'ye.<br />
There are large quantities of weakly podzolic sandy soils ("bor sands"). ("Bor" is a pine<br />
woodland on poor, sandy soil.) Dry bor sands are found on eolian and moraine hills in the<br />
Poles'ye.<br />
Meadow soils are derived from bog or podzolic soils and are also called "halfbog," "turfpodzol,"<br />
or "sod-podzol" soils. "Half-bog" soils develop when herbaceous plants and/or<br />
deciduous trees grow on drained bogs. These are often found under wet meadows. Deciduous<br />
trees growing on podzols add nitrogen and other minerals. An herbaceous understory can<br />
develop, transforming the upper horizons of podzols into "sod-" or "turf-podzols." Such soils are<br />
found under fresh meadows. Soddy, moderately podzolic soils have formed on moraine outcrops<br />
in the Kiev Poles'ye.<br />
When forests are cleared but not adequately drained, the podzols deteriorate. Deterioration is<br />
marked by formation of gley horizons. Gleyed soils include gley, gley-podzol, and sod-gley soils.<br />
Gleization occurs when the iron in soils where water is at or near the surface is usually reduced to<br />
ferrous compounds, having a gray or bluish color. These soils therefore have a more or less dense<br />
but not unusually viscous layer of loamy or clayey material of a gray color. Gley are also found in<br />
seasonally waterlogged sites. Sod-gley soils are found in meadows that have become boggy; areas<br />
with gleyed soils often become sphagnum bogs.<br />
Some parts of the Poles'ye contain "islands" of soils foreign to it. The are small loess islands<br />
in both the Zhitomir Poles'ye to the west of the Kiev Poles'ye and the left bank to the east. One<br />
5
such island occurs on the Ovruch ridge (about 90 km west of <strong>Chernobyl</strong>). This ridge is 320 m<br />
above sea level and 60 m above the surrounding lowland.<br />
6
SECTION 4<br />
CLIMATE 8<br />
The Poles'ye has a humid temperate-continental climate, influenced by maritime, continental,<br />
and local factors.<br />
Bogs, which are common around the <strong>Chernobyl</strong> NRS, influence the microclimate. These areas<br />
have high humidity, low temperature minima, evening and morning mists close to the soil surface,<br />
and frequent frosts.<br />
Forests, which are common, also influence the microclimate. The air temperatures are lower<br />
than in the open; the daily temperature ranges are smaller, the snow cover is less thick; and winds<br />
are stilled.<br />
The moisture from the Baltic Sea and the "great valley" region of Poland influences the<br />
Poles'ye. Because there is unrestricted passage of marine winds, unhindered by major land relief,<br />
there is considerable marine influence on the climate.<br />
A small local maximum of relative humidity, caused by the intensified evaporation of water<br />
from wetlands, is noticed over the Poles'ye wetlands. The minimum relative humidity at 1 p.m. in<br />
May is 50 to 55 percent. The mean humidity for June, July, and August is 56 to 60 percent.<br />
The Poles'ye has hot summers and relatively mild winters. Mean temperatures in the region<br />
are -6 to -7 0 C in January, 6 to 7 °C in April, 19 °C in July, and 7 0 C in October. Absolute<br />
minimum temperatures are -30 to -35 °C in January, while summer maximums range up to 40 0C.<br />
The growing season (days with mean temperatures above 5 °C) is about 160 days. The<br />
growing season in bogs begins 8 to 10 days later than in the surrounding nonbog habitat and ends<br />
10 to 12 days earlier.<br />
Summer precipitation exceeds winter by a factor of 2. In any season, there is precipitation<br />
every 2 to 3 days. In the warm season (April to September), the amount of precipitation in the<br />
interior of the European USSR is much greater (350 to 500 millimeters [mm]) than on the coasts<br />
(200 to 300 mm). During the cold half of the year (November to March) in the central belt of<br />
European USSR, the precipitation amounts are 100 to 300 mm.<br />
A precipitation maximum is situated on the Pripyat and the upper reaches of the Dnepr and<br />
Western Dvina Rivers. The maximum precipitation in the Pripyat basin is 680 to 695 mm/year.<br />
Precipitation diminishes eastward from the Pripyat. In the specific region of interest, mean<br />
precipitation is 500 to 600 mm/yr, with around 180 days/year recording precipitation. Two-thirds<br />
81nfmrmation about the climate has been synthesized from Kendrew 1942, Berg 1950, Anonymous 1962, Borisov<br />
1965, Szafer 1966, Skoropmnov 1968, Lysenko and Golsvina 1982, USSR 1987.<br />
7
of the precipitation is received in the warm season (April to September). The mean intensity of<br />
precipitation amounts to 8 to 10 mm/hr in the central belt of the country (51 to 59 0 N).<br />
At 50°N by the second 10-day period in November snow cover is continuous. In western<br />
European USSR, snow cover is usually continuous by the last 10 days in October or the first 10<br />
days in November. At 55ON 300E, there is an average of 2.5 temporary snow covers before the<br />
onset of winter. In central European USSR, winter temperatures and precipitation are highly<br />
variable. Mean maximum snow depth is 10 to 30 centimeters (cm). Continuous snow cover<br />
generally lasts about 80 days. Rivers and streams are generally frozen for about 100 days/year.<br />
8
.. SECTION 5<br />
VEGETATION<br />
On a global scale, the Poles'ye is in the Eastern European vegetation province, whose<br />
northern, eastern, and southeastern boundaries correspond to the distributions of Quercus robur,<br />
Acerplatanoides, and Corylus avelkana (Takhtajan 1986). In the Flora of the USSR, the Poles'ye<br />
is in the Upper Dnepr floral region (Komarov 1968). According to Lydolph (1964), Poles'ye<br />
means "woodland" and much of the natural vegetation was once pine-dominated woodlands.<br />
About 30 percent is still forested.<br />
Habitats in the region are classified by how wet they are and whether the vegetation includes<br />
trees. The basic plant community types in the region are wetlands, meadows, carts, woodlands,<br />
and forests. 9 ,1 0 , 11 These may grade into one another. In the <strong>Chernobyl</strong> district, natural<br />
vegetation (mostly forest, woodlands, and wetlands) covers nearly 50 percent of the terrain (USSR<br />
1987). This does not include the meadows that have been created from draining mires or felling<br />
forests, in which the natural vegetation is often augmented with pasture grasses and legumes. In<br />
all, only one-third of the landscape is croplands.<br />
Much of the Kiev Poles'ye is floodplain terraces. Floodplains are usually divided into the area<br />
adjacent to the streambed (the streamside zone), a central zone, and an "upper" zone. The<br />
streamside zone has the thickest silt deposit and relatively high mesorelief (depressions and<br />
ridges1 2 ). Ridges in this zone are usually 1 to 10 m high and depressions awe quite shallow. The<br />
central zone is usually undulating, with low mesorelief and shallow groundwater. Ridges in the<br />
central zone usually are not flooded and may even experience some drought. Depressions are<br />
9 While the vegetation types are specific to the Poles'ye, the species lists are those known to occur in that type of<br />
vegetation in northen Ukraine, southern Belorussia, andlor adjacent areas in Poland.<br />
10 [)finitions and descriptions of habitat, vegetation types, succession, bog drainage, and reclamation have been<br />
synthesized from Keller 1927, Berg 1950, Vilenskii 1960, Akademiya Nauk SSSR 1963, Viktorov et al. 1964,<br />
P'yavchemko 1965, Szafer 1966, Skmpanov 1968, Stankevich and Rubin 1968, Remezov and Pogrebnyak 1969,<br />
Ivanov 1972, Fridland 1976, Walter 1978, Komamv 1980, Smith 1980, Moore 1984, USSR 1987.<br />
1 llnformation on the flora of vegetation types and descriptions on the species lists was synthesized from Keller<br />
1927. Sukachev 1928, Wulff 1943, Berg 1950, Vilenskii 1960, Larin 1962a"b, 1964, Akademiya Nauk SSSR 1963,<br />
Komarov 1963, 1964, 1967, 1968, 1970, 1971a,b, Tutin 1964, 1968, 1972, P'yavachenko 1965. Szafer 1966,<br />
Reme and Pogrebnyak 1969, Ivanov 1972, Bialobok and Zelawski 1976, Leathart 1977, Phillips 1978, Walter<br />
1978, Hora 1981, Oieynik 1981, Moore 1984, Takhtajan 1986. USSR 1987.<br />
12 MorW e idgsa anl depressions ue gomocphic in origin and ae not synonymous with hummocks and hollows<br />
which ae formed by the vegetation and are much smaller in scale than ridges and depressin. There are some<br />
differnces in flood meakow species disribution elated to meorelief. Teme cam be found with the individual species<br />
in the species •ist that follow.<br />
9
flood-prone. The central zone can contain areas in which flooding is prolonged (greater than 40<br />
days), moderate (20 to 40 days), and short (fewer than 20 days), and areas that do not experience<br />
annual flooding; the vegetation in them varies accordingly. The "upper" zone is the lowest in<br />
absolute elevation and is usually flat and excessively wet. Fen, bog, meadow, and woodland<br />
vegetation may all occur on floodplains. Fens and carrs are usually found close to the stream.<br />
Transitional bogs are usually found in the central zone. Annually flooded meadows are found both<br />
close to the stream and in the central zone. Fresh meadows and moist forests, habitats that do not<br />
experience annual are usually in the central zone. Floodplain bogs, boggy wet forests, and wet<br />
meadows are found in the upper zone.<br />
There are also extra- (or supra-) floodplain areas. The interfluve of the Pripyat and Teterev<br />
Rivers is unusually morainal-hilly. There are poorly drained sandy plains are located between the<br />
moraine hills. Dry woodlands and forests are found on the moraine hills. Sphagnum bogs and<br />
wet forests occur on the poorly drained moraine plains and fresh meadows in better drained areas.<br />
Immense tracts of the Poles'ye are occupied by wetlands. The Russian word "botolo" is a<br />
general term for wetlands. It is translated in a number of ways, but the best may be mire: a<br />
peatland, 13 a wetland with considerable water retained by an accumulation of partially decayed<br />
organic matter, including wetlands with or without flowing water. There are a number of English<br />
terms used to describe different types of mires. Marshes are wetlands with emergent vegetation<br />
(i.e., plants with their leaves above water but their roots in soil that is covered [part or all of the<br />
time] with water). Marshes include both fens, swamps, and bogs. Fens are mires fed by moving<br />
water, either surface or ground water. Swamps are fens with only surface water influence. If<br />
woody vegetation is added to a fen, these communities are called carrs. Fens are sometimes<br />
called low moors or low moor bogs. The community closest to a stream is usually a fen or carr.<br />
Fens are also frequently found at the edges of lakes. Bogs are mires where water is supplied<br />
exclusively by atmospheric precipitation. The terms "high moor" bog and "raised" bog are<br />
sometimes used to distinguish true bogs from fens. Transitional bogs are intermediate between<br />
fens and bogs. The term "marsh" is sometimes reserved for transitional bogs. Most bogs are not<br />
flat but consist of hummocks, hollows, and intermediate sites.1 4<br />
131B geologcal terms, a mire must have a layer of peat at least 20 So 30 an deep (Walter 1978). "Moom" means<br />
mir in Gamm; a Ein SU it is used for tranutional bogs<br />
14 1ouauom Abou which Vpeda grow on hummocfs hollows, and intmeediaw sites can be found with the plant<br />
4in<br />
dt species IeL<br />
10
The dominant vegetation in sedge fens (swamp) is usually sedges, horsetails, and/or<br />
cottongrass. Plants 15 include Equisetuu arvenSe, 16 E. pahistre, E pratense, E. variegatum, Carex<br />
acutiformis, C. appropin quata, C. aquatilis, C. caespitosa, C. canescens, C. chordorrhiza,<br />
C. diandra, C. dioica, C. gracilis, C. heleonastes, C. lasiocarpa, C. limosa, C. panicea,<br />
C. paniculata, C. psuedocyperus, C. ripania. C. stellulata, C. vesicaria, C. vulpina, Cyperus<br />
Jlavescens, C. pal wnicus, Eriophorum angustifolium, E~ gracile, E. latifolium, Scirpus Lacustris,<br />
silvauicus, Junciss compressus, J. inflexus, J. tenageia, Alopecurus arwulinaceus, A. geniculatus,<br />
Glyceria aquasica, Leersia oryzoides, Phragmites commwuns, Poa palustris, Setaria glauca, and<br />
S. viridis. In colored aerial photographs, 17 fens appear smooth dark green (the darker, the wetter<br />
the mire). In black and white photos, they appear dark, either without outlines or with an<br />
elongated outline. Soils beneath sedge fens become peaty.<br />
Reed-bulrush fens (swamps) also have peat soil. The vegetation includes Phragmites<br />
communis, Scirpus acicudaris, S. eupalustris, S. holoschoenus, S. lacuster, S. mamillatus,<br />
S. ovatus, S. pauciflorus, S. sylvaticus, Beckmannia eruciformis, Calamagrostis neglecta,<br />
Glyceria aquauica, G. fruitans, G. hemoralis, Leersia 0' yzides, Phalaris arwzdinacea, Typha<br />
angustifolia, T. latifilia, Carex ekuta, C. omskiana, C. panicea, C. pseudo-cyperus, C. vesicaria,<br />
Juncus compressus, and J. tenageia. These usually sit in 50 to 100 cm of water, with plants<br />
anchored in a silty bottom.<br />
Soils in transitional bogs are also peaty. Vegetation includes Carex acutiformis,<br />
C. appropin quata, C. aquatilis, C. caespitosa, C. canescens, C. chordorrhiza, C. diandra,<br />
C. dioica, C. disticta, C. disperma, C. elata, C. fihiformis, C. gracilis, C. heleonastes,<br />
C. junceila, C. lauiocarpa, C. limosa, C. muricata, C. omskiana, C. pauciflora, C. pseudocyperus,<br />
C. riparia, C. vesicaria, C. vulpina, Eriopho rum angustifoblium, E. gracile,<br />
E. latifoblium, K. vaginatum, Scirpus lacustris, Juncus compressus, JI effusus, JI lampocarpus,<br />
I. eeruji, J. squarrosus, Alopecurus arundinaceus, A. geniculatus, Beckmannia eruciformis,<br />
Calamagrostis neglecta, Deschampsia caespitosa, Molirsia coerulea, Equisetum heleocharis,<br />
E. palustre, and L. variegatum. Sphagnum apiculatum, Sph. compactum, Sph. cuspidatum,<br />
Sph. fuscum, Sph. papillosum, Sph. recurvum, Sph. rubellum, and Sph. squarrosum are<br />
common mosses. Transitional bogs are often found in shallowwater (less than 1 in), where Alnus<br />
1 5 Commuumty type we gmueraL They are often subdvided int two to may mate detailed communities. Not all<br />
pheat musd far a comumity wfi neoeussay be found together, rahe they an found in one or more of the<br />
eommities of the type. Mit nondcuakum species order d=e no nesammily reflect umportace.<br />
1C mm<br />
for indvidua ueciet cm be found in the specie li&<br />
1 7 pbrdw iForo About the upince of vegetatio in saera photographs con be found with the descriptions in<br />
he Species lsL
glutinosa has been cut. Transitional bogs often contain shrubs, including Betulda humilus, Salix<br />
aurita, S. cinerea, S. lapponum, S. myrtilloides, S. phylicifolia, S. rosmarinifolia, Ledum<br />
palutre, and Chamaedaphne calyculata. Trees sometimes found in or at the margin of transitional<br />
bogs include A/nus incana, Betula pubescens, Pinus sylvestris, and Populus tremula. From the<br />
air, transitional bogs appear as a patchwork of dark green (wet peat), green (hollow), and yellow<br />
(hummock) patches, without sharp boundaries. Water appears as black pools. Tree crowns are<br />
visible in aerial photographs.<br />
Most of the bogs in the Chemobyl region are Sphagnum peat bogs, with or without trees. The<br />
common shrubs in sphagnum bogs include Andromeda polifolia, Betula humiUs, Calluna vulgaris,<br />
Chamaedaphne calyculata, Empetrum nigrum, Ledum palhssre, Rubus chamaemus, Vaccinium<br />
oxycoccus, V. uliginosum, and V. vitis-idaea. Common herbs include Carex limosa,<br />
C. chordorrhiza, C. heleonastes, Eriophorum vaginatum, Juncus bulbosus, J. squarrosus,<br />
Equisetum variegatum, and the carnivorous plant Drosera intermedia. The mosses include<br />
Sphagnum cuspidatum, Sph. fuscum, Sph.medium, Sph. recurvum, Sph. rubellum, Sph.<br />
russovii, Pohlia nutans, Polytrichum commune, and P. stricum. In aerial photographs, sphagnum<br />
bogs have parallelnarrow, dark zigzag lines on a lighter, brownish (sphagnum) background. In<br />
black and white photos, they are light to dark gray (with light dominant), or they appear to have<br />
sinuous light belts with dark intervals (hummocks and hollows).<br />
Sphagnum-pine bogs are the most common type of wooded sphagnum bog. Trees are not a<br />
principal vegetation component of the community. The species are a mixture of those found in<br />
sphagnum bogs and pine-sphagnum forests. Pines in bogs only reach 8 m in height. From the<br />
air, sphagnum-pine bogs have the same appearance as raised bogs; however, the crowns of trees<br />
are visible.<br />
Meadows are usually (1) wet meadows, (2) flood meadows, or (3) fresh meadows.<br />
Meadows may be transitional or ecotonal between wetlands and moist forest. In black and white<br />
aerial photographs, meadows appear uniformly light colored and usually have very definite<br />
boundaries.<br />
(1) Wet meadows are inundated for some time each year, especially late winter and early<br />
spring. The groundwater table fluctuates frequently but never drops very low. They often have<br />
peaty soils. Wet meadows often are secondary vegetation resulting from drainage of fens or bogs<br />
or when moist forest is removed. The vegetation includes Agrosuis alba, A. canina, A. stolonifera,<br />
Alopecurus arundinaceus, A. geniculatus, A. pratensis, Beckmannia eruciformis, Calamagrostis<br />
neglecma, C. phragmitoides, Catabrosa aquatica, Deschampsia caespitosa, Festuca arundinacea,<br />
F. ovina, F. pratensis, F. rubra, Glyceria aquatica, G. fruitans, G. plicata, Hierochloe odorata,<br />
Molinia oedruea, Nardw mnictv , Phajaris armudinacea, Poa palustris, Triseium sibiricum, Carex<br />
aquatiuis, C. caespimosa, C. canescens, C diandra, C dioica, C. disticta, C. gracilis, C. juncella,<br />
12
C. lasiocarpa, C. limosa, C. muricata, C. panicea, C. riparia, C. vesicaria, C. vulpina,<br />
Eriophorum angustifolium, Scirpus compressus, S. eupalustris, S. holoschoenus, S. ovatus,<br />
S. pauciforus, S. silvaticus, Juncus ambiguus, J. compressus, J. effusus, J. tenageia, Equisetum<br />
arvense, E. hekocharis, and E. palustre.<br />
(2) Flood meadows are usually inundated for at least short periods but are reasonably well<br />
drained, have a lower water table than wet meadows during most the year, and have moist soils for<br />
the rest of the growing season. Plants include Agropyron repens, Agrostis alba, A. canina, A.<br />
stolonifera, Alopecurus arundinaceus, A. pratensis, A. tenuis, Anthoxanthum odoratum,<br />
Arrhenatherum elatius, Beckmannia eruciformis, Bromus inermis, Calamagrostis epigeios,<br />
Deschampsia caespitosa, Festuca arundinacea, F. pratensis, Glyceria aquatica, G. fruitans, G.<br />
plicata, Hierochloe odorata, Koeleria delavignii, Phalaris arundinacea, Phleum pratense, Poa<br />
palustris, P. pratensis, P. trivialis, Trifolium hybridum, T. pratense, T. repens, Carex aquatilis,<br />
C. gracilis, C. riparia, C. vesicaria, Scirpus eupalustris, Juncus ambiguus, J. compressus, J.<br />
tenageia, Equisetum arvense, E. heleocharis, and E palustre. Species restricted to flood meadows<br />
with short periods of inundation include Festuca ovina, Arrhenatherum elatius, Medicagofalcata,<br />
Trifolium pratense, and T. repens.<br />
(3) Fresh ("true" or "dry") meadows are often secondary vegetation on cut or burned forest<br />
sites. Mowing is often used to prevent forest encroachment. They have a moderate moisture<br />
supply that fluctuates widely but generally does not surface. They sometimes become quite dry in<br />
summer. These meadows are usually used as pastures or haymeadows and some of the plants,<br />
especially the legumes, may have been sown. The vegetation includes Agropyron repens,<br />
Alopecurus pratensis, A. tenuis, Anthoxanthum odoratum, Arrhenatherum elaius, Briza media,<br />
Bromus inermis, Cynosurus cristatus, Dacrylis glomerata, Festuca rubra, F. pratensis, Holcus<br />
lanatus, H. mollis, Lolium perenne, Nardus stricta, Phleum pratense, Poa pratensis, P. trivialis,<br />
P. annua, Medicago falcata, M. lupulina, M. sativa, Melilomus alba, M. officinalis, Trifolium<br />
repens, T. hybridum, T. pratense, Carex leporina, and Equisetum arvense. Beckmannia<br />
erucibrmis, Phak/ris arwndinacea, and Poa palustrisare are occasionally found there.<br />
Herbaceous plants also commonly dominate forest and woodland margins and openings<br />
(glades). Many openings are man-made, the result of burning, felling, or thinning forests or<br />
woodlands. Herbs found there include Agropyron repens, Beckmannia pinnatum, Briza media,<br />
Bronu inermis, Calamagrostiu epigeios, Cynosurus criuratus, Dactylis glomerata, Melica nutans,<br />
Carex leporina, and Equisesum sylvaticum. Prunus spinosa, Rosa acicularis, R. mollis,<br />
R. tomentosa, Rubhts chamaemus, Salix caprea, S. livida, S. nigricans, S. phylicifolia, and<br />
Populus tremala are often found at the margin of forest openings. Mixed forest-meadow<br />
vegetation is divided into three types: (1) scattered, with herbaceous vegetation evenly spread<br />
throughout a thinned tree stand (in oak and linden groves there are about 300 to 800 trees<br />
13
hectare (ha) and in birch woodlands there are about 200 to 600 trees/ha); (2) flower-bed, with<br />
alternating openings and dense forests; and (3) windbreak, with forest strips about 25 to 30 m<br />
wide alternating with rectangular openings about 70 to 80 m wide.<br />
Relict Rhododendron luteum thickets are found at the margins of pine- and spruce-deciduous<br />
forests on peaty soils in the Poles'ye. Ledum palustre, Vaccinium uliginosum, and V. viuis-idaea<br />
are common dwarf shrubs there.<br />
Heaths (or heathlands) are ericoid shrub fens in poor, acid, podzolized soils, in meadow-type<br />
sites, often where pine forest has been cut. Calluna vulgaris is the dominant dwarf shrub. Other<br />
dwarf shrubs include Arctostaphylos uva-ursi, Empetrum nigrum, Ledum palustre, Vaccinium<br />
vids-daea, and V. uliginosur. The common herbs include Nardus stricta and Uncus squarrosus.<br />
Sphagnum fuscum is a common moss.<br />
More than 60 percent of the forests are dominated by Pinus sylvestris. The deciduous forests<br />
are dominated by oak, hornbeam, birch, and alder. It is unlikely that any virgin forests remain in<br />
eastern Europe (Walter 1979).<br />
There are a number of Russian terms used to describe forests and woodlands. (Woodlands<br />
are low density forests.) "Bor" is a pine woodland on poor, sandy soil. Bor woodlands are<br />
common on the poor glaciofluvial sands of the Poles'ye. Pine forests can be found in sites that<br />
climatically should be deciduous or coniferous-deciduous forest. "Bor oak" is occasionally used<br />
for oak growing in sandy soils, and "bor birch" for birch in similar soils. Pine-birch woodlands<br />
on poor soils are frequently also called "bor." "Subor" is a mixed deciduous-pine forest,<br />
dominated by pine, with a strong lower stratum of deciduous trees, usially oak but also birch or<br />
beech. Pine-oak forests are "subor," as are pine-birch forests on better soils. "Ramen" is a more<br />
closed stand, usually dominated by spruce. "Suramen" is a mixed "ramen." "Dubrava" is an oak<br />
forest or woodland (grove) and includes oak hornbeam forests, and "sudubrava" is a woodland<br />
dominated by deciduous trees usually considered subordinant in oak forests, such as birch<br />
woodlands and linden and aspen groves.<br />
Carr is an English term for a wooded wetland. There are several types of carrs in the<br />
Poles'ye.<br />
Willow (or sallow) camrs occur on either reed or sedge peat. They are usually found in the<br />
stream zone of floodplains. Trees may include Salix alba, S. caprea, S. rossica, Alnus glutinosa,<br />
A. incana, and Betula pubescens. The shrubs include Sa/ix acuufolia, S. cinerea, S. dasyclados,<br />
S. nigricans, S. pentandra, S. purpurea, S. riandra, S. xerophila, and Cornus a/ba. Dwarf shrubs<br />
include Salix aurita, S. livida, S. myrtlloides, S. rosmarinifolia, Rosa mollis, R. omenaosa, and<br />
Ribes nigrnm<br />
Osier thickets, with osier willows such as Salix purpurea and S. viminalis, are often found on<br />
shore dunes.<br />
14
Willow-poplar carts occur on recent alluvial soils in the stream and central zones of<br />
floodplains. Soils are s~turated during floods but waters flow through, and they can dry out<br />
during nonflood periods. The trees include Populus nigra, P. alba, Salix alba, S. caprea, and<br />
sometimes Alnus glutinosa and Fraxinus excelsior. Shrubs include Salix cinerea, S. livida,<br />
S. pentandra, S. purpurea, S. rosmarinifolia, S. triandra, and S. viminalis.<br />
Wet alderwoods occur on reed fen peats. The are usually located in the excessively wet areas<br />
in the stream and central zones of floodplains. Trees in alderwoods include Alnus glutinosa,<br />
A. incana, Fraxinus excelsior, and Betula pubescens. Shrub willows include Salix cinerea and<br />
dwarf shrub willows include Salix myrtilloides and S. rosmarinifolia and may also include others<br />
found in willow carts (above). Ribes nigrum, Betula humilus, Vaccinium myrtillus, and V. vitisidaea<br />
are also common dwarf shrubs. Herbs include Agropyron caninum, Deschampsia<br />
caespitosa, Festuca gigantea, Glyceria lithuanica, Milium effusum, Molinia coerulea, Phragmites<br />
communis, Carex argyroglochin, C. caespitosa, C. elongata, and C. inumbrata.<br />
Alder-ash and elm carts are small communities with damp to wet, weakly acid to neutral soils.<br />
The plants in elm cams include Ulmus iaevis, U. glabra, Quercus robur, and Acer platanoides.<br />
Elm carms are found on sites that are innundated only during major floods.<br />
The dominant trees in alder-ash carms include Alnus glutinosa, A. incana, and Fraxinus<br />
excelsior, mixed with a few Acer plantinoides and Carpinus betulus. Ribes nigrum is a common<br />
shrub. Carex remota and Equisetum silvaticum are important herbs. Alder-ash cams occur on<br />
humic-gley soils.<br />
Aspen groves (dominated by Populus tremula), linden groves (dominated by T7Iia cordata),<br />
and birch woodlands (dominated by Betula pendula in slightly moist soils and Betula pubescens in<br />
wetter soils, with a mixture in moderately moist soils) are all "sudubrava." They are all early<br />
successional woodlands, and they all have all have strongly herbaceous understories. Occasional<br />
Quercus trees may be intermixed. Alopecurus tenuis, Calamagrostis arundinacea, C. epigeios,<br />
Melica nwtans, and Eriophorum vaginatum are important herbs. Birch woodlands often develop in<br />
fire clearings in moist soils, with Arctostaphylos uva-ursi as an understory shrub.<br />
The Kiev Poles'ye lies on the ecotone between the mixed coniferous-deciduous forest and the<br />
forest steppe. There are two basic mixed coniferous-deciduous forest types: pine-oak forest and<br />
spruce-oak forest. The common steppe-forest of the area is oak-hornbeam.<br />
Quercus robur is a codominant in each of them and is the dominant tree in forest-steppe in the<br />
European USSR. It grows best in the southwestern part of the European USSR and in the<br />
Poles'ye. Quercus robur will not grow on strongly podzolic soils. It is usually found on<br />
floodplains. It usually grows in mixed stands with Pinbs sylvestris, Picea abies, or Carpinus<br />
beudus.<br />
15
In addition to Q. robur and C. betulus, the trees found in oak-hornbeam forests ("dubrava")<br />
include Acer platanoides, A. pseudoplatanus, Betula pendula, Fraxinus excelsior, Tilia cordata,<br />
Ulmus laevis, and U. glabra. On richer soils, Fagus sylvatica is sometimes a component on<br />
moderately moist soils. Cornus alba and Corylus avellana are common shrubs. Agropyron<br />
caninum, Koeleria delavignii, Melica nutans, and Millium effusum are important herbs. Carex<br />
pilosa is important on drier soils, while Festuca gigantea is important on wetter ones. Oakhornbeam<br />
forests are not found in stagnant water. The groundwater depth is usually greater than<br />
1 m. They are most common on older alluvial soils and on frontal and ground moraines. They<br />
are often found between wet alderwoods and poorer pine-oak oak forests. From the air, oakhornbeam<br />
forests appear bright light gray (in black and white photographs). The surface of the<br />
canopy appears uneven, and trees appear bunched in groups of three to five.<br />
Acidiphilous pine-oak forests C'subor") are found in poor habitats with weakly to moderately<br />
podzolized, sandy or sandy-loam soils. In addition to Pinus sylvestris and Q. robur, the trees<br />
include Betula pendula and/or Betula pubescens, Populus tremula, and Tilia cordata. Corylus<br />
avellana is the dominant shrub, and Vaccinium myrtillus is the most important dwarf shrub.<br />
Acidiphilous beechwoods (also "subor") are found on moist, moderately fertile soils. The<br />
dominant trees are Fagus sylvestris and Pinus sylvestris. The nondominant tree species are those<br />
found in acidiphilous pine-oak forests.<br />
Most natural pine stands ("bor") are in poor habitats. There are many "bor" woodlands in the<br />
Poles'ye, dominated by Pinus sylvestris. Pine-dominated stands types include (1) dry to freshon<br />
well-drained soils with marked podzolization, (2) damp or moist--on poorly drained soils with<br />
attenuated podzolization (semi-bog soils), (3) wet-on poorly drained bog soils.<br />
Dry "bor" woodlands are found on the old dunes and morainal hills. The parabolic, westfacing<br />
dunes of the Poles'ye are usually covered with Pinus sylvestris giving the region a<br />
"northern" appearance. These woodlands rarely have understory shrubs. The herbaceous<br />
understory vegetation can include Festuca ovina, F. polesica, Koeleria glauca, and Poa pratensis.<br />
These sites have light, dry sandy soils, and the lichens are common. Lichens include Cladonia<br />
alectoria, C. alpestris, C rangiferina, C. silvatica, C. uncialis, and Cetraria islandica. C alpestris<br />
often dominates in late succession. About 100 years is required for pine on dry soil to mature<br />
enough to make good timber. Trees are unequal in height, crowns are wide, and branches<br />
pendant Betuda is not found in dry "bot" woodlands.<br />
Moist "bor" forests are dominated by Pinus with either Polytrichum mosses or evergreen<br />
ericoid shrubs (e.g., Andromeda polifolia, Arctostaphylos uva-ursi, Chamaedaphne calyculata,<br />
Calluna vulgaris, Vaccinium myrtillus, and V. vitis-daea) dominating the understory.<br />
Brachypodium silvancum, Festuca ovina, F. rubra, Molinia coerulea, and Nardus stricta are<br />
16
important herbs. Trees are well proportioned, with narrow crowns. Pines in moist habitats are<br />
longer lived than in dry or wet habitats and are the best quality timber.<br />
Wet "bor" woodlands can have either flowing or stagnant water. In the former Phragmites<br />
communis (growing in hollows) is the dominant understory species, with Nardus stricta as an<br />
important herb. In the latter, Sphagnum mosses dominate the understory. Both are dominated by<br />
Pinus. In "subor" woodlands, Betula pubescens is strongly subdominant, but woodlands in<br />
which Bzga is a lesser component are often referred to as "bor." Wet pine-sphagnum woodlands<br />
are found in poorly drained depressions on terraces above floodplains. After 100 years, pines on<br />
wet soils may be only 6 m tall and reach a total heighz of only 8 m.<br />
There are several subtypes of pine-sphagnum "bor" woodlands. (1) Betula pubescens is an<br />
important tree component. Carex lasiocarpa, Equisetum heleocharis, and E. vaginatum are<br />
important herbaceous components. The mosses include Sph. centrale and Sph. warnstorfii,<br />
Sph. girgensohnii, Pleurosium schreberi, and Aulacomium palustre. This type is often found in<br />
depressions with gentle slopes, with peaty-gley or peaty, intensely boggy soils. (2) Salix caprea is<br />
an important tree, Salix pendrandra is a common shrub, and Vaccinium vitis-idaea, V. myrtillus,<br />
V. oxycoccus, Ledum palustre, Empetrum nigrum, Chamaedaphne calyculata are common dwarf<br />
shrubs. Equisetum silvaticum, Calamagrostis epigeios, Carex lasiocarpa, Eriophorum<br />
angustifolium, E. gracile, and E vaginatum are important herbaceous species. The mosses include<br />
Sph. magellanicum, Sph. compactum, Sph. fuscum, Sph. russovii. Sph. apiculatum, and<br />
Pleurosium schreberi. This type occurs in slight depressions with impeded runoff. (3) Betula is<br />
an important tree component; Pinus decreases as Betula increases in this type. Salix aurita,<br />
Vaccinium myrtillus, V. vitis-idaea, and Ledum palustre are the important dwarf shrubs; V.<br />
uliginosum is unusual. The mosses include Sph. acutifolium, Sph. warnstorfii, Polytrichum<br />
commune, Hylocomium proliferum, Pleurozium schreberi, and Aulacomium palustre. This type is<br />
transitional between slightly flowing and stagnant wet woodlands and is found in slightly hillocky<br />
shallow depressions. (4) Betula pubescens is an important tree component. There are also<br />
infrequent Alnus. Vaccinium vitis-idaea, V. myrtillus, Andromeda polifolia, Calluna vulgaris,<br />
Chamaedaphne calyculata, Ledum palustre, Empetrum nigrum, and Betula humilus are important<br />
dwarf shrubs. Carex lasiocarpa and Calamagrostis lanceolata are important herbs. The moss is<br />
mostly Polytrichum commune with some Sphagnum mosses, including Sph. fuscum. This type is<br />
found in shallow, usually closed depressions between drier pine stands. (5) Pinus is sparse. The<br />
common subshrubs are Rubus chamaemus, Vaccinium viis-idaea, and V. uliginosum; V. myrtillus<br />
and V. oxycoccus are infrequent. Equisetum variegatum is a common herb. The mosses are<br />
Sph. angust'folium, Sph. acutifolium, Sph. fuscum, Sph. magellanicum, and Sph. russovii, with<br />
Sph pulchrum and Sph. pahwstre in hollows. This type is transitional between woodland and bog.<br />
17
Soils are highly acidic. (6) Pinus is sparse. Shrub and moss species are the same as subtype 5.<br />
Eriophorwn vaginatum is the common herb.<br />
Picea abies endures shade well but requires humid, relatively rich soils. Its southern limit runs<br />
through Kiev Poles'ye. Spruce-oak forests probably occur mainly on "islands" of richer podzolic<br />
soils scattered in the Poles'ye. Understory trees include Acer platanoides, Betulda pendula,<br />
Carpinus betulus, Fraxinus excelsior, Populus tremula, Tilia cordata and Ulmus laevis. Shrubs<br />
include Corylus avellana and Rubus chamaemus. Herbs include Carex argyroglochin, C. loliacea,<br />
Scirpus sylvaticus, Agropyron caninun, Calamagrostis phragmitoides, Deschampsia caespitosa,<br />
Glyceria lithuanica, Melica nutans, and Millium effusum. Mosses include Sphagnum girgensohnii<br />
and Sph. squarrosum.<br />
18
SECTION 6<br />
SUCCESSION<br />
In poor sandy soils, peat bogs occupy the low areas, forests the upper ones. Succession can<br />
progress from forest to bog or bog to forest.<br />
With increased moisture, succession progresses from Pinus-Polytrichum moist forests to<br />
Pinus-Sphagnum wet forests, then to sphagnum-pine bog.<br />
The first sere in secondary succession after an moderately dry oak-hornbeam "dubrava" has<br />
been felled is a pine-birch "bor," followed by a pine-oak-birch "subor," and then finally a return to<br />
"dubrava."<br />
When moist pine forests are overly thinned or cleared for meadows, without adequate drainage<br />
and with grazing, boggy conditions develop and podzols deteriorate. With time these meadows<br />
become sphagnum bogs. Without grazing, meadows that do not become bogged are colonized by<br />
birch, aspen, alder. Pine eventually returns under these, and, with time, they return to pine<br />
forests.<br />
If birches are left when pines are felled in pine-birch woodlands, birch woodlands with grassy<br />
understories often develop within 3 to 5 years. Because Calamagrostis epigeios inhibits pine<br />
regeneration, it can retard secondary succession.<br />
In deforested dry pine woodlands on sandy hills or dunes, trees and lichens are often replaced<br />
by herbaceous dune vegetation, much of which is already found in the understory, including<br />
Festuca ovina, F. polesica, Koekria glauca, K. gracilis, and Poa pratensis.<br />
Aspen groves, linden groves, and birch woodlands are all early successional woodlands on<br />
moderately moist soils. These can become oak-hornbeam forest.<br />
Willow-poplar carrs usually begin as osier thickets. These are replaced with alder-ash carts,<br />
which can become oak-hornbeam forest.<br />
As succession proceeds in a fen, it first becomes a willow carr or wet alderwood, which is<br />
replaced by alder-ash can'. If the site dries, this is replaced by an oak-hornbeam forest. Further<br />
drying leads to a pine-oak woodland.<br />
If a fen is grazed, the first step is a wet meadow. If the groundwater table drops, it will<br />
become a fresh meadow. If grazing is abandoned, the site will become an oak-hornbeam forest.<br />
Further drying again results in a pine-oak woodland.<br />
Flood meadows are created when fens are drained. In time these will become forest.<br />
If flowing water decreases in a fen but the site remains wet, peat decomposition will decrease,<br />
and pines and sphagnum will increase.<br />
19
With time, a wet pine-sphagnum woodland will become increasingly bog and eventually will<br />
become a sphagnum-pine bog. As the pines become less and less robust, the tree component can<br />
be lost, or nearly so.<br />
20
SECTION 7<br />
BOG DRAINAGE AND RECLAMATION<br />
Most bogs in central Europe are being drained. The replacement vegetation is usually meadow<br />
grasses, birch, pine, or spruce.<br />
Many of the wetlands in the Poles'ye have been drained. Between 1872 to 1898, 400,000 ha<br />
of wetlands were converted to meadows, 130,000 ha to croplands, and 1,300,000 ha to accessible<br />
forests. In 1950, it was estimated that 1.3 million ha in Belorussia and 1 million ha in the Ukraine<br />
"needed" reclamation. Between 1950 to 1970, an extensive drainage program was caried out in<br />
the Poles'ye in both republics. By 1972, the Poles'ye had ceased to be a primarily waterlogged.<br />
Continuous extensive drainage of the Pripyat wetlands may disturb the groundwater balance<br />
both in the Poles'ye and in the drier regions to the south. Drainage can lead to excessive drying of<br />
rivers and to destruction of hydrophylic shrubs on berry grounds.<br />
Because the sandy soils of this region are particularly vulnerable to wind erosion when<br />
plowed, conversion to croplands has not always been successful. Drainage has made some old<br />
farmlands worthless. In some areas, the sandy soils have dried out and become blown sand; some<br />
peat soils need irrigation to grow wheat or beets. Since 1968, the formerly excessively wet<br />
Poles'ye has experienced dust storms. There is a joke that, in the Poles'ye, those attempting to<br />
"reclaim" wetlands "turn all swamps into deserts" (Komarov 1980). Because of the soils beneath<br />
the bogs, it might have been better if they had been converted to meadows instead of croplands;<br />
however the plans of the agricultural bureaucrats specified plowed croplands.<br />
Pines in some tree farms have died from viral diseases; they needed the acid soils formed from<br />
stagnant water to do well. These diseased nursery trees also infected nearby stressed older trees.<br />
21
SECTION 8<br />
CROPS 1 s<br />
Only about half of the area around <strong>Chernobyl</strong> is suitable for agriculture. Much of it is in hay<br />
meadows and grazing land (Trifolium spp. and grasses), 19 and only about 25 percent of the area is<br />
actually in croplands.<br />
Triticum aestivum, Secale cereale, Panicum miliaceum, Hordeum vulgare, Avena sativa,<br />
Fagopyrum esculentum, Beta vulgaris, Solanum tuberosum, Linum usitatissimum, Cannabis<br />
satii, Humulus lupulus, Daucus carota, Brassica oleracea var. capitata, B. oleracea var. acephala,<br />
B. rapa, and Cucumis sativa are all grown in the southern part of the Belornssian SSR and<br />
northern part of the Ukrainian SSR. However, most can not be grown on poorly drained, boggy<br />
sites of the Poles'ye. As a result, wheat and other mesic soil crops are less important here than in<br />
other parts of Belorussia and the Ukraine. Cereal grains are grown on about half the crop lands,<br />
fodder crops on 35 to 40 percent, potatoes on about 8 percent, and flax on up to 5 percent. At any<br />
one time, 12 to 17 percent of the land is being fallowed.<br />
On drained bogs with moderate peat beds, crops are often replaced, after a few years, with<br />
perennial grasses and legumes, creating an artificial meadow for hay and grazing. Many of the<br />
dwarf shrubs growing in bogs produce edible fruits and these are harvested as "crops." In some<br />
places, they are cultivated in the bogs.<br />
18 Information regvding cops has been synmeuized from Lydolph 1964, Skoropanov 1968, Sambur and Kovalenko<br />
1969, Fuflard 1972, Synau 1972, Komamov 1980, Dewdocy 1982, USSR 1987.<br />
19 See meadows, under Sectin 5, Vegeaio.<br />
22
SECTION 9<br />
REFERENCES<br />
Akademiya Nauk SSSR. 1963. The Increase of Productivity of Swamped Forests. Israel<br />
Program for Scientific Translation, Jerusalem. 291 pp.<br />
Anonymous. 1962. Atlas of the Ukrainian SSR and Moldavian SSR. Central Administration of<br />
Geodesy and Cartography, Moscow. (In Russian)<br />
Berg, L. S. 1950. Natural Regions of the U.S.S.R. Translated by 0. A. Titelbaum. MacMillan<br />
Co., New York.<br />
Bialobok, S., and W. Zelawski, eds. 1976. Outline of Physiology of Scots Pine. Foreign<br />
Scientific Publications, <strong>Department</strong> of the National Center for Scientific, Technical and<br />
Economic Information, Warsaw.<br />
Borisov, A. A. 1965. Climates of the U.S.S.R.<br />
Publishing Co., Chicago.<br />
Translated by R. A. Ledward. Aldine<br />
Brady, N. C. 1974. The Nature and Properties of Soils. MacMillan Publishing Co., Inc., New<br />
Yor0.<br />
Dewdney, J. C. 1982. U.S.S.R. in Maps. Holmes & Meier Publishers, Inc., New York.<br />
Fridland, V. M. 1976. Patterns of Soil Cover. Israel Program for Scientific Translation,<br />
Jerusalem.<br />
Fullard, H. 1972. Soviet Union in Maps. George Philip & Son Ltd., London.<br />
Golovina, L. P., M. N. Lysenko, and T. I. Kisel. 1980. "Content and Distribution of Zinc in the<br />
Soil of the Ukrainian Poles'ye." Soviet Soil Science 12(l):73-80.<br />
Hora, B., ed. 1981. Oxford Encyclopedia of Trees of the World. Oxford University Press,<br />
Oxford, England.<br />
Ivanov, K. E., ed. 1972. The Hydrology of Marshlands. Israel Program for Scientific<br />
Translations, Jerusalem.<br />
Keller, B. A. 1927. Distribution of Vegetation on the Plains of European Russia. Journal of<br />
Ecology 15:189-233.<br />
Kendrew, M. A. 1942. The Climates of the Continents. Oxford University Press, New York.<br />
Komarov, B. 1980. The Destruction of Nature in the Soviet Union.. Translated by M. Vale &<br />
J. Hollander. M. E. Sharpe, Inc., New York. 150 pp.<br />
Komarov, V. L. 1963. Flora of the USSR. Volume 2. Israel Program for Scientific Translations,<br />
Jerusalem.<br />
23
-. 1964. Flora of the USSR.. Volume 3. Israel Program for Scientific Translations,<br />
Jerusalem.<br />
• 1967. Flora of the USSR. Volume 18. Israel Program for Scientific Translations,<br />
Jerusalem.<br />
- 1968. Flora of the USSR. Volume 1. Israel Program for Scientific Translations,<br />
Jerusalem.<br />
- 1970. Flora of the USSR. Volume 5. Israel Program for Scientific Translations,<br />
Jerusalem.<br />
. 1971La. Flora of the USSR. Volume 10. Israel Program for Scientific Translations,<br />
Jerusalem.<br />
1971b. Flora of the USSR. Volume 11. Israel Program for Scientific Translations,<br />
Jerusalem.<br />
Krupskiy, N. K., V. P. Kuz'michev, and R. G. Derevyanko. 1070. "Humus Content in<br />
Ukrainian Soils." Soviet Soil Science 2:278-288.<br />
Larin, I. V. 1962. Pasture Economy and Meadow Cultivation. Israel Program for Scientific<br />
Translations, Jerusalem.<br />
1962. Pasture Rotation. Israel Program for Scientific Translations, Jerusalem.<br />
1., ed. 1965. Advances in Pasture and Hay-meadow Management. Israel Program for<br />
Scientific Translations, Jerusalem.<br />
Leathart, S. 1977. Trees of the world. Hamilyn, London.<br />
Lydolph, P. E. 1964. Geography of the USSR. John Wiley & Sons, Inc., New York.<br />
Lysenko, M. N., and L. P. Golovina. 1982. "Boron Content and Distribution in the Soils of the<br />
Ukrainian Poles'ye." Soviet Soil Science 14(1):89-97.<br />
Moore, P. D., ed. 1984. European Mires. Academic Press, London.<br />
Oleynik, V. S. 1981. "Genetic characteristics of peat soils in the Ukrainian western Poles'ye."<br />
Soviet Soil Science 13(4):32-37.<br />
Phillips, R. 1978. Trees of North American and Europe. Random House, New York.<br />
P'yavchenko, N.I., ed. 1965. Improvement in Forest Growth on Peat-Bog Soils of the USSR<br />
Forest Zone and Tundra. Israel Program for Scientific Translations, Jerusalem.<br />
Remezov, N. P., and P. S. Pogrebnyak.<br />
Scientific Translations, Jerusalem.<br />
1969. Forest Soil Science. Israel Program for<br />
24
Sambur, G. N., and I. L Kovalenko. .1969. "Improvement and Rational Utilization of Lowland<br />
Saline Soils of the Southern Poles'ye and Northern Forest-Steppe of the Ukrainian SSR."<br />
Soviet Soil Science 1:1401-1408.<br />
Skoropanov, S. G. 1968. Reclamation and Cultivation of Peat-bog Soils. Israel Program for<br />
Scientific Translations, Jerusalem.<br />
Smith, R. L 1980. Ecology and Field Biology. Harper & Row, Publishers, New York.<br />
Stankevich, V. S., and P. R. Rubin. 1968. Drainage and Reclamation of Bogs and Bogged<br />
Areas. Israel Program for Scientific Translations, Jerusalem.<br />
Sukachev, V. N. 1928. "Principles of Classification of the Spruce Communities of European<br />
Russia." Journal of Ecology 16:1-18.<br />
Symons, L. 1972. Russian Agriculture: A Geographic Survey. John Wiley and Sons, New<br />
York.<br />
Szafer, W. 1966. The Vegetation of Poland. Pergamon Press, Oxford.<br />
Takhtajan, A. 1986. Floristic Regions of the World. Translated by T. J. Crovello. University of<br />
California Press, Berkeley.<br />
Tutin, T. G., V. KL Heywood, N. A. Burges, D. H. Valentine, S. M. Walters, and D. Webb.<br />
1964. Flora Europaea. Volume 1. Cambridge University Press, London.<br />
Tutin, T. G., V. H. Heywood, N. A. Burges, D. M. Moore, D. H. Valentine, S. M. Walters,<br />
and D. A. Webb. 1968. Flora Europaea. Volume 2. Cambridge University Press, London.<br />
1972. Flora Europaea. Volume 3. Cambridge University Press, London.<br />
USSR. 1986. The <strong>Accident</strong> at the <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Pant and its Consequences.<br />
Information compiled by the USSR State Committee on the Utilization of Atomic Energy for<br />
the IAEA Experts' Meeting, 25 to 29 August 1986, Vienna (IAEA Translation).<br />
USSR. 1987. The <strong>Accident</strong> at the <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant: One Year After. International<br />
Conference on <strong>Nuclear</strong> Power Performance and Safety, 28 September to 2 October 1987,<br />
Vienna (IAEA wanslation).<br />
Viktorov, S. V. Ye. A. Vostokova, and D. D. Vyshivkin. 1964. Short Guide to Geobotanical<br />
Surveying. MacMillan Co., New York.<br />
Vilenskii, D. C. 1960. Soil Science. Israel Program for Scientific Translations, Jerusalem.<br />
Walter, H. 1978. Vegetation of the Earth and Ecological Systems of the Geobiosphere.<br />
Translated by J. Wieser translator. Springer-Verlag, New York.<br />
Wulff, E. V. 1943. An Introduction to Historical Plant Geography. Translated by E. Brissenden.<br />
Chronia Botanica Co., Waltham, Massachusetts.<br />
25
APPENDIX A<br />
TREES AND SHRUBS<br />
List of Species 2 0 ' 2 1<br />
Noncultivated 2 2<br />
"Mmeciuse more than one valid Latin nam was sometimes found for species, I chose the conservatie name<br />
(choosig thernam which stmasse simibiaites or jlumper's name). Lotn names of dicoyledons follow the Flora<br />
Ewvaerop those of mnuo olqledoms are those used in the curren liateature. The authorof Flora of she USSR tended<br />
to emphasize differences (i~e., was what is called a splitter elevating subgeneric names to genera, subspecific<br />
qilthas to specific), his names often do nom reflect those most frequently encountered tn the literature.<br />
2 1Reggreaces nchide those used for specie and habitats in Section 3. Vegetation<br />
Z 2 lchuifg species seeded int meadows, exclding species growing sa crop in plowed fields.<br />
A-1
GYMNOSPERMS<br />
PINUS SYLVESTRIS 23<br />
Scots or Scotch pine24<br />
common pine, wild pine<br />
forest pine<br />
P. sylvestrs is the dominant tree in the Poles'ye and is the most abundant in much of Europe,<br />
where it is often called simply "common pine" (also its common name in the USSR). It can live<br />
for up to 400 years. It is an erect evergreen tree and can be 20 to 30 m tall (occasionally up to<br />
40 m). In bogs, trees are quite stunted; after 100 years trees often are only 6 m tail, and they<br />
rarely get to be more than 8 m tall. Its crown is round, high, broad, and (in aerial views) appears<br />
smooth and dark (although lighter than spruce canopies). Intervals between crowns appear<br />
uniform and are not deeply shaded. The trunk is often crooked, with regularly whorled branches,<br />
diverging at acute angles. The bark is deeply fissured, thick, and dark brown on the lower trunk<br />
but smooth, thin, and red or red-brown on the upper trunk.<br />
There are two types of vegetative apices. Long shoot are elongated stems with scaled leaves in<br />
the axils. These produce dwarf shoots, lateral long shoots, and strobili (cones). Short shoot are<br />
deciduous shoots, which do not elongate and which produce the leaves. Buds are acute and more<br />
or less resinous. Twigs are glabrous and are initially yellow green, becoming gray-brown with<br />
age.<br />
Primary leaves are always single and are found only on seedlings. Within 2 to 3 years, they<br />
are completely replaced by secondary (permanent) leaves. Secondary leaves (needles) are paired,<br />
glaucous, and blue-green in color. Needles are elongate, semicylindrical, tapering toward the<br />
apex. The internal surface is flat and either slightly concave or slightly convex. Stomata are found<br />
on both sides of the needle distributed in rows along its entire length. The external surface is<br />
semicircular. They are twisted, 2 to 7 cm long, and about 2 mm in wide. Needle length varies<br />
with year of growth; they may continue to grow for more than one season. Needles on trees from<br />
eastern Europe are 5 to 7 cm long; the longest needles found have been on pines from the Poles'ye<br />
and Volhynia. Shoots with female cones usually have longer needles than shoots with male cones.<br />
Individual leaves can survive 5 or more years.<br />
Degree of stomatal opening decreases from a morning maximum to almost complete closure by<br />
evening. Regardless of soil moisture, there are considerably fewer stomata open in the afternoon.<br />
23 Someimes spelled sve .<br />
24Most raulasmrs give the English common me far a species. However, Ameaic and other European as weln<br />
as lcal, common ames wer sometimes encountezed. I have listed all common ames I found for each species.<br />
A-2
Reproduction from pollination to mature seed takes 3 years. Seeds form 1 year after<br />
pollination. Cones remain small during the first year after pollination. There is rapid growth and a<br />
change in color (from gray-brown to rich green) in the second summer. Mature seed cones are 3 to<br />
7 cm long and 2 to 3.5 cm wide. They are yellow-brown when they are mature. Free-standing<br />
trees can bear seeds when they are about 15 years old; trees in a closed canopy do not begin to bear<br />
seeds until they are about 20 years old. Abundant seed crops are produced only every 3 to 4 years,<br />
but at least a few trees yield seeds every year. Self-pollination produces poorer quality seed than<br />
does cross pollination.<br />
P. sylestris usually forms mycorrhizal associations in all soil types, although there are<br />
qualitative and quantitative differences among soils. Pines in bog soils are very mycorrhizal,<br />
probably because of the low nutrient status of the soil. More than 40 species of fungi have been<br />
found associated with P. sylvestris in Europe. Annual radial growth in P. sylvestris is dependent<br />
on temperature and nutrient depletion. There can be as many as seven flushes, with more than one<br />
growth ring per year. Annual apical growth can be completed as early as mid-June. Maximum<br />
needle elongation may occur as -arly as early May.<br />
There is quite a bit of difference among trees in different habitats. In dry sites, trees are<br />
unequal in height, crowns are wide, branches pendant, and needles are quite blue-gray. Growth<br />
begins earlier, and the growth increment is more clearly marked than in moister habitats. The size<br />
of the increment depends on the amount of summer rainfall. In moist habitats, trees are well<br />
proportioned, crowns are narrower, and leaves are greener. Trees live longer (usually more than<br />
200 years). The size of the growth increment depends on the amount of summer rainfall. The best<br />
quality timber comes from moist forests. In wet habitats, trees have shorter lives, usually 100 to<br />
200 years. Growth begins late, after the groundwater table subsides. Sphagnum mosses inhibit<br />
regeneration from seed. Trees growing in very wet habitats are susceptible to windthrow (or<br />
blowdown).<br />
It is codominant with Quercus robur in pine-oak mixed forests and with Fagus silvatica in<br />
acidiphilous beechwoods. It is dominant in "bor" forests. It is found in sphagnum-pine bogs and<br />
wooded transitional bogs.<br />
P. sylvestris is a light-loving species, not tolerant of shade. It is xerophytic, 25 not very<br />
exacting in its moisture requirements and can be found growing under almost any moisture regime<br />
from dry to swampy. On wet soils, it is susceptible to windthrow. It is an oligotroihic 26<br />
25C=win<br />
dry sad.<br />
A-3
acidophyte. 27 It can tolerate relatively poor soils and is often associated with bogs and sandy<br />
soils. It has moderate transpiration rates.<br />
PICFA ABIES2<br />
Norway spruce<br />
European spruce<br />
Common spruce<br />
P. abies is an important evergreen tree in much of eastern Europe. Its southern limit runs<br />
through Kiev Poles'ye, but it apparently is not common there. It is usually 30 to 50 m tall but<br />
occasionally reaches 60 m. It is full grown in 30 to 50 years, but it can live for more than 300<br />
years.<br />
Its crown is pointed, pyramidal, and nearly rests on the ground. In aerial views it appears very<br />
uneven and darker than that of P. sylvestris.<br />
Its branches are whorled, and it lacks short shoots. Needles are spirally arranged, 1 to 2 cm<br />
long, and green. Mature cones are 10 to 14 cm long.<br />
P. abies endures shade well but requires humid, relatively rich, loamy soils. Because poor<br />
soils predominate in the Poles'ye, P. abies is probably limited to widely scattered "islands" of<br />
richer podzolic soils.<br />
It is mesotrophic. 29 It grows best on moderately moist soils. On wet soils, it is susceptible to<br />
windthrow.<br />
27 Toume acidic wiLt<br />
2 9P. mete (fiqpemm, foud in ok" fkeme)- P. abies.<br />
29 kRqoiu at lo• uodme W&,l of mnuaims.<br />
A-4
DICOTYLEDONS<br />
QUERCUS ROBUR<br />
English oak<br />
British oak<br />
European oak<br />
pedunculate oak<br />
Q. robur is the most important deciduous tree in the mixed forest of the region. It is found<br />
with Pinus sylvestris or Picea abies in mixed forests and with Carpinus betulus in oak-hornbeam<br />
forests. It is found in elm cam and is occasional in linden and aspen groves and birch woodlands.<br />
Its crown appears blunt and is very large. Oak-hornbeam forests appear bright light-gray in<br />
aerial photographs (especially in summer). Because of the mixture of other species, the surface of<br />
the canopy appears somewhat uneven and trees may appear bunched.<br />
It is mesotrophic, has high transpiration rates, and is a mesoxerophyte. 30 It grows best on<br />
moderately moist soils. On wet soils, it is susceptible to windthrow.<br />
It grows to be 30 to 50 m tall and lives up to 800 years.<br />
ACER PLATANOIDES<br />
Norway maple<br />
sharp-leafed maple<br />
It is a deciduous tree, usually 18 to 20 m (occasionally 30 m) tall. The crown is densely<br />
leafed, and it has a spreading canopy. It is eutrophic, 3 1 has high transpiration rates, and is a<br />
xeromesophyte. It grows best in slightly moist soils. It is found in alder-ash carrs, oak-hornbeam<br />
forests, and spruce-oak forests.<br />
ACER PSEUDOPLA7TNUS<br />
sycamore<br />
sycamore maple<br />
great maple<br />
It is a deciduous tree, up to 30 m tall. It is eutrophic. It is found in oak-hornbeam forests.<br />
Dense, domed crown; spreading canopy.<br />
30 rowfum moduuely dry d~L<br />
3 IRelhu~ vrbvely high nutrent levels.<br />
A-5
ALNUS GLUTINOSA<br />
European alder, black alder<br />
common alder<br />
It is a deciduous tree, usually up to 20 m tall (rarely up to 35 m). The crown is pyramidal. Its<br />
bark is dark brown and fissured. It is mesotrophic, has moderate transpiration rates, and is a<br />
hydrophyte. 32 It always grows in moist soils, especially cleared forest sites and drained bogs. It<br />
is unable to grow in dry ones. In boggy soils, grows well only on the peaty layer or on peat,<br />
which is at times dry and whose upper layers are strongly decomposed; it is usually found on<br />
hummocks. It is codominant in alder-ash carts and wet alderwoods. It is found in sallow<br />
(willow) cans. It is sometimes found in willow-poplar carrs.<br />
ALNUS INCANA<br />
gray alder, white alder<br />
speckled alder<br />
European alder<br />
It is a small deciduous tree (sometimes a shrub), 5 to 20 m tall. It grows on moderately welldrained<br />
soils. Its bark is smooth. It is mesotrophic and is a mesohydrophyte. 33 In boggy soils, it<br />
usually grows on hummocks. It is found in alder-ash camrs, sallow (willow) carts, wet<br />
alderwoods, wooded transitional bogs, and at forest margins. It is an early successional species in<br />
cleared moist and wet forest sites. ALNUS GLUTINOSA x A. INCANA hybrids are not<br />
uncommon where parents grow together.<br />
BETULA PENIDULA3<br />
European white birch<br />
silver birch, common birch<br />
It is a deciduous tree, 10 to 30 m tall. The bark is smooth and white. Its crown is lower, more<br />
elongate, lighter colored, and grows closer to the ground than pine. Spaces between crowns are<br />
small and uniform. (In a pine-birch mixture, it is difficult to distinguish the two, and the mixture<br />
can be difficult to distinguish from pure stands of either.)<br />
32 Toklrm exaemive flooding (more than 40 days), grows only in damp arean, often in standing water.<br />
33 .rokrms mod flooding (20.40 days).<br />
34Bau pendal . ILwrucosa of ealy iteram.<br />
A-6
It is oligotrophic, an acidophyte, and a mesophyte. 35 It grows best in slightly moist soils,<br />
especially on burns and other clearings. In wet soils, it is susceptible to windthrow. Sphagnum<br />
mosses inhibit regeneration from seed. It does not grow well on dry soils. In bog soils, it grows<br />
well only on the peaty layer or on peat, which is at times dry and whose upper layers are strongly<br />
decomposed; it is usually found on hummocks.<br />
It can be dominant in early successional birch woodlands on moist grounds. It is found in<br />
acidiphilous beechwoods, pine-oak woodlands, and spruce-oak forests. It often grows with<br />
B. pubescens.<br />
BETULA PUBESCENS<br />
downy birch, red birch<br />
pubescent birch<br />
It is a low to medium-height deciduous tree, 10 to 25 m tall. Its crown is lower, more<br />
elongate, lighter colored, and grows closer to the ground than pine. Spaces between crowns are<br />
small and uniform. (In a pine-birch mixture, it is difficult to distinguish the two, and the mixture<br />
can be difficult to distinguish from pure stands of either.)<br />
It is oligotrophic, an acidophyte, and a mesohydrophyte. It grows best on moderately moist<br />
soils, especially on burns and other clearings. On wet soils, it is susceptible to windthrow.<br />
Sphagnum mosses inhibit regeneration from seed. It does not grow well on dry soils. In bog<br />
soils, it grows well only on the peaty layer or on peat, which is at times dry and whose upper<br />
layers are strongly decomposed; it is usually found on hummocks.<br />
It is often locally dominant in poor, acid (peaty) woodlands. It can be dominant in early<br />
successional birch woodlands on moist grounds. It is found in wet pine-sphagnum forests<br />
(subtype 4), sallow (willow) carrs, wet alderwoods, pine-oak woodlands, and wooded transitional<br />
bogs. It is often found with B. pendula.<br />
CARPINUS BETULUS<br />
common hornbeam<br />
European hornbeam<br />
It is a deciduous tree that is usually 15 to 25 m tall (occasionally up to 30 m). The trunk is<br />
fluted. Its bark is smooth and gray. The crown is rounded. It is mesotrophic, an acidophyte, and<br />
a mesophyte. It grows best in slightly moist soils and cannot tolerate flooding. It is codominant in<br />
oak-hornbeam forests. It is found in alder-ash carts.<br />
35<br />
equim mods aly mot sl.<br />
A-7
FAGUS SIL VA TICA<br />
European beech<br />
common beech<br />
It is a deciduous tree that is usually 30 m tall (occasionally up to 50 m in dense beechwoods).<br />
The bark is smooth and gray. It is frequently cultivated on well-drained soils. It grows best on<br />
slightly moist soils and cannot tolerate flooding. It is codominant in acidiphilous beechwoods<br />
(with Pinus sylvestris). It is sometimes found in oak-hornbeam forests.<br />
FRAXINUS EXCELSIOR<br />
European ash<br />
common ash<br />
It is a deciduous tree, usually up to 30 m tall (occasionally up to 45 m). The bark is gray. It is<br />
eutrophic, has high transpiration rates, and is a mesophyte. It grows best in slightly moist soils.<br />
In boggy soils, it usually grows on hummocks or intermediate sites. 3 6 It is found in wet<br />
alderwoods, alder-ash carrs, oak-hornbeam, and spruce forests. It is sometimes found in willowpoplar<br />
carts.<br />
POPULUS ALBA<br />
white poplar<br />
It is a deciduous tree. It grows in spring-inundated soils. It is found in willow-poplar carts.<br />
POPULUS NIGRA<br />
black-barked poplar<br />
common black poplar<br />
water poplar<br />
It a deciduous tree, up to 30 m tall. The bark is dark gray. Leaves are deltoid. It grows in<br />
spring-inundated soils. It is found in willow-poplar cams. It usually grows mixed with Salix and<br />
Alnus.<br />
36 &Mae Wpm between hummoc md Wolows in bop.<br />
A-8
POPULUS TREMULA<br />
aspen, asp<br />
European aspen<br />
trembling aspen<br />
trembling poplar<br />
It is a deciduous tree that is usually 15 to 20 m tall (occasionally up to 50 m). It is a clonal tree.<br />
Individual shoots are shortlived. Bark of young shoots is smooth, greenish-white but is brown<br />
and ridged at base of older shoots.<br />
It is mesotrophic. It grows in spring-inundated soils and is a mesohydrophyte. In very wet<br />
habitats, it is susceptible to windthrow. In boggy soils, it is usually found on hummocks. Aspen<br />
canopies are much lighter than pine or birch and appear flat. When aspen grows mixed with pine<br />
or birch, the woodland may have a "flower-bed" appearance because of the clonal growth form of<br />
aspen.<br />
It is dominant in early successional aspen groves on moist soils. It is found in acidiphilous<br />
beechwoods, pine-oak woodlands, spruce-oak forests, and at the margins of transitional bogs and<br />
forest openings.<br />
SALIX ALBA<br />
white willow<br />
European white willow<br />
It is a tree willow, 10 to 25 m tall. The bark is gray. It is a mesohydrophyte and grows best<br />
on moderately moist soils. It is found in willow-poplar carrs.<br />
SALIX CAPREA<br />
goat willow, pussy willow<br />
goat sallow, great sallow<br />
It is a tree willow, 6 to 10 m tall. It is mesotrophic and a mesohydrophyte. It grows best on<br />
moderately moist soils. In boggy soils, it is usually found on hummocks. It is found in willowpoplar<br />
and sallow (willow) carrs, at forest margins, and in wet pine-sphagnum forests (subtype 2).<br />
SALIX PENTANDRA<br />
bay willow<br />
bay-leaf willow<br />
laurel willow<br />
laurel-leaf willow<br />
It is a deciduous tree in damp woods, 7 to 10 m tall. Twigs are glabrous and shiny. It is<br />
mesotrophic. It is found in sallow (willow) and willow-poplar carms.<br />
A-9
SALIX ROSSICA<br />
It is a deciduous tree, 8 to 20 m tall. It is found in sallow (willow) carrs.<br />
T7LJA CORDATA<br />
small-leafed lime<br />
small-leafed linden<br />
It is a deciduous tree, up to 35 m tall. The crown appears pale, large, and spreading. It<br />
appears almost white in late-spring (flowering) photos. It is mesotrophic and a mesophyte. It is<br />
dominant in early successional linden woodlands on moist soils. It is found in oak-hornbeam<br />
forests, acidiphilous beechwoods, pine-oak woodlands, and spruce-oak forests.<br />
ULMUS GLABRA<br />
Scotch elm, Wych elm<br />
scabrus elm, mountain elm<br />
It is a deciduous tree, up to 40 m tall. It is eutrophic, has high transpiration rates, and is a<br />
mesophyte. It grows best in moderately moist soils. It is codominant in elm carrs. It is found in<br />
oak-hornbeam forests on flood-plains and flat interfluvial plains.<br />
ULMU$ LAEVIS37<br />
European white elm<br />
European elm, Russian elm<br />
fluttering elm<br />
It is a deciduous tree, 20 to 35 m tall. It is eutrophic, has high transpiration rates, and is a<br />
mesohydrophyte. It grows best in moderately moist soils. It is codominant in elm camf. It is<br />
found in oak-hornbcam forests and spruce-oak forests.<br />
37 Uw L/aevis - U. pedwucwata of early iitumwe.<br />
A-1O
SHRUBS (OVER 2 M)<br />
CORNUS ALBA<br />
Siberian dogwood<br />
It is a deciduous shrub, up to 3 m tall. Twigs are dark red. It is found in sallow (willow) carrs<br />
and oak-hornbeam forests.<br />
CORYLUS AVELLANA<br />
European filbert, cobnut<br />
European hazelnut<br />
common hazelnut<br />
wild hazelnut<br />
It is 4 to 6 m tall. The bark is smooth and brown. It is eutrophic and a mesophyte. It is found<br />
in oak-hornbeam forests and acidiphilous beechwoods, pine-oak woodlands, and spruce-oak<br />
forests.<br />
PRUNUS SPINOSA<br />
blackthorn, sloe berry<br />
It is a deciduous shrub, 4 to 8 m tall. It is found in at the margins of forests.<br />
RHODODENDRON LUTEUM<br />
pontic azalea<br />
It is an evergreen ericoid shrub, 1 to 6 m tall. It is a mesotrophic acidophyte. In the USSR,<br />
this species is principally found in western TransCaucasus. In the Poles'ye, it is assumed to be a<br />
Tertiary relict. It is found at the margins of pine- and spruce-deciduous forests on peaty soils.<br />
SAM1X ACUTIFOLJA<br />
sharp-leafed willow<br />
It is a tall deciduous shrub (rarely a tree), 10 to 12 m tall. It is a xerophyte. It is found in<br />
sallow (willow) cars.<br />
A-i1
SALIX CINEREA<br />
gray willow<br />
gray sallow<br />
It is a deciduous shrub, 5 to 10 m tall. It is a hydrophyte. It is found in sallow (willow) and<br />
willow-poplar carts, wet alderwoods, and shrubby transitional bogs.<br />
SALIX DASYCLADOS<br />
It is a deciduous shrub, 5 to 8 m tall. It is found in sallow (willow) carts.<br />
SALIX NIGRICANS<br />
dark-leafed willow<br />
black-leafed willow<br />
It is a deciduous shrub, 1 to 8 m tall. It is found in sallow (willow) and willow-poplar camrs<br />
and at forest margins.<br />
SALIX PENTANDRA<br />
bay willow<br />
bay-leaf willow<br />
laurel willow<br />
laurel-leaf willow<br />
It is deciduous and is a shrub (3 to 5 m tall) in peat soils. In boggy soils, it is usually found on<br />
intermediate sites. Twigs are glabrous and shiny. It is found in wet pine-sphagnum forests<br />
(subtype 2), transitional bogs, and in wet alderwoods.<br />
SALIX PHYLUCIFOLIA<br />
tea-leafed willow<br />
It is a deciduous shrub, up to 3.5 cm tall. It is found in sallow (willow) camts, transitional<br />
bogs, forest openings and at margins.<br />
SAMIX PURPUREA<br />
purple osier<br />
purple willow<br />
It is a deciduous shrub, usually 2 to 5 m tall (rarely up to 10 m). It is found in sallow (willow)<br />
and winow-poplar cans.<br />
A-12
SALIX T7JANDRA<br />
almond-leafed willow<br />
French willow<br />
It is deciduous and is usually a shrub (sometimes small tree), 4 to 10 m tall. It is found in<br />
sallow (willow) and willow-poplar carts.<br />
SALIX VIMINAUS<br />
common osier<br />
osier willow<br />
It is a deciduous shrub, 5 to 10 m tall. It is found in willow-poplar carts (earlier succession).<br />
SALMX XEROPHILA<br />
It is a deciduous shrub, up to 6 m tall. It grows in sallow (willow) carrs.<br />
A-13
DWARF SHRUBS (UNDER 2 METERS)<br />
ANDROMEDA POLUFOLUA<br />
andromeda, bog rosemary<br />
bog-rosemary andromeda<br />
It is an evergreen ericoid dwarf shrub, 15 to 40 cm tall. It is an oligotrophic acidophyte. In<br />
boggy soils, it is usually found on hummocks or tree boles. It is found in wet pine-sphagnum<br />
forests (subtype 4), sphagnum bogs, and moist pine woodlands.<br />
ARCTOSTAPHYLOS UVA-URSI<br />
bearberry<br />
It is an evergreen, mat-forming, ericoid dwarf shrub, 25 to 130 cm tall, with long, prostrate<br />
branches. It is an oligotrophic acidophyte. In boggy soils, it is usually found on hummocks. It is<br />
found in heaths, thin birch woodlands, and moist pine woodlands.<br />
BETULA NANA<br />
dwarf birch<br />
It is a deciduous dwarf shrub, 20 to 70 cm tall (rarely more than 2 m), with spreading or<br />
procumbant branches. It is oligotrophic. In boggy soils, it is commonly found on hummocks.<br />
While this is primarily a tundra-taiga species, it is sometimes found in the forest zone. Its habitats<br />
are the same as those of B. humilus.<br />
BETULA HUMILUS<br />
low birch<br />
dwarf birch<br />
It is a small, much-branched deciduous shrub, 1 to 2 m tall. It is oligotrophic.<br />
characteristic of continuously wet peat. In boggy soils, it is commonly found on hummocks or<br />
tree boles. It is found in wet pine-sphagnum forests (subtype 4), shrubby transitional bogs,<br />
alderwoods, and sphagnum bogs.<br />
It is<br />
A-14
CHAMAEDAPHNE CALYCULATA<br />
leatherleaf<br />
ground laurel<br />
It is an evergreen ericoid dwarf shrub, 15 to 50 cm tall. It is an oligotrophic acidophyte. In<br />
boggy soils, it is usually found on hummocks or tree boles. It is found in moist pine forests, in<br />
wet pine-sphagnum forests (subtype 1, 2,4), in shrubby transitional bogs, and in sphagnum bogs.<br />
CALLUNA VULGARIS heather<br />
It is an evergreen ericoid dwarf shrub, 30 to 70 cm tall. It is an oligotrophic acidophyte. In<br />
boggy soils, it is usually found on hummocks or tree boles. It is the dominant shrub in heaths. It<br />
is found in moist pine-sphagnum forests, in wet pine-sphagnum forests (subtype 4), and (in late<br />
successional stages) in sphagnum bogs.<br />
EMPETRUM NIGRUM<br />
black crowberry<br />
It is an evergreen ericoid dwarf shrub, up to 120 cm tall. It is an oligotrophic acidophyte. In<br />
boggy soils, it is usually found on hummocks or tree boles. It is found in pine-sphagnum<br />
woodlands (subtypes 2, 4, 5, 6), heaths, and sphagnum bogs.<br />
LF_•UM PALUSTRE<br />
ledum, marshtea<br />
crystaltea ledum<br />
wild rosemary<br />
It is an evergreen ericoid dwarf shrub, 12 to 125 cm tall. It is an oligotrophic acidophyte. In<br />
boggy soils, it is usually found on hummocks or tree boles. It is found in moist pine forests, wet<br />
pine-sphagnum forests (subtypes 1, 2, 3, 4), heaths, rhododendron thickets, sphagnum bogs, and<br />
shrubby transitional bogs.<br />
RIDES NIGRUM<br />
black current<br />
It is a deciduous dwarf shrub, 1 to 2 m tall. In boggy soils, it is usually found on hummocks.<br />
It is found in wet alderwoods.<br />
A-15
ROSA ACICULARIS<br />
dog rose<br />
It is a deciduous dwarf shrub, up to 1 m tall. It is found at forest margins.<br />
ROSA MOLWS<br />
hair rose<br />
It is a deciduous dwarf shrub, 50 to 200 cm tall. It is found in sallow (willow) carts.<br />
ROSA TOMENTOSA<br />
downy rose<br />
It is a deciduous dwarf shrub, 50 to 150 cm tall. It is found in sallow (willow) carts and at<br />
forest margins.<br />
RUBUS CHAMAEMUS<br />
raspberry, cloudberry<br />
It is a deciduous dwarf shrub. In boggy soils, it is usually found on hummocks. It is found in<br />
wet pine-sphagnum forests (subtypes 5, 6), shrubby sphagnum bogs, and spruce-oak forests.<br />
SALIX A URITA<br />
roundear willow<br />
roundear sallow<br />
eared sallow<br />
It is a deciduous dwarf shrub, 1 to 2 m (rarely 3 m) tall, with creeping stems. It is oligotrophic<br />
and is a hydrophyte. In boggy soils, it is usually found on hummocks or intermediate sites. It is<br />
found in wet pine-sphagnum forests (subtype 3), sallow (willow) carrs, wet alderwoods, and<br />
shrubby transitional bogs.<br />
SAMX CINEREA<br />
gray willow, gray sallow<br />
It is a deciduous shrub, dwarfed in peat-bogs, with creeping stem. It is found in shrubby<br />
sphagnum and transitional bogs and in sallow (willow) carrs.<br />
A-16
SALIX LAPPONUM<br />
burr willow<br />
It is a deciduous dwarf shrub, 1 to 1.5 m tall. It is found in shrubby transitional bogs.<br />
SALIX LIVIDA<br />
livid willow<br />
It is a deciduous dwarf shrub, about 50 cm tall. It grows at woodland margins and in sallow<br />
(willow) willow-poplar car.<br />
SALIX MYRTILLOIDES<br />
myrtle willow<br />
It is a deciduous dwarf shrub, 30 to 80 cm tall (rarely up to 2 m), with subterranean creeping<br />
stem. It is eutrophic. In boggy soils, it is usually found on intermediate sites. It is found sallow<br />
(willow) camrs, wet birch woodlands, and shrubby transitional bogs.<br />
SAMIX ROSMARINIFOL!A<br />
rosemary willow<br />
It is a deciduous dwarf shrub, 75 to 100 cm tall, with subterranean creeping stem. It is an<br />
oligotrophic acidophyte. In boggy soils, it is usually found on intermediate sites. It is found in<br />
sallow (willow) and willow-poplar carts and in transitional bogs.<br />
VACCINIUM MYRTILLUS<br />
myrtle whortleberry<br />
myrtle bilberry<br />
It is a deciduous ericoid dwarf shrub, 15 to 40 cm tall. It is oligotrophic acidophyte. In boggy<br />
soils, it is usually found on hummocks or tree boles. It is found in wet pine-sphagnum forests<br />
(subtypes 2, 3, 4), wet alderwoods, acidiphilous beechwoods, and pine-oak woodlands. It is<br />
infrequent in wet pine-sphagnum forests (subtypes 5, 6).<br />
A-17
VACCINIUM OXYCOCCUS<br />
small cranberry<br />
cranberry<br />
It is a evergreen ericoid dwarf shrub, up to 80 cm tall. It is an oligotrophic acidophyte. It is<br />
found in wet pine-sphagnum forests (subtypes 1, 2, 5, 6) and sphagnum bogs. Fruit harvested as<br />
a "crop."<br />
VACCINIUM ULIGINOSUM<br />
bog whortleberry<br />
bog bilberry<br />
It is a deciduous ericoid dwarf shrub, 50 to 100 cm tall. It is an oligotrophic acidophyte. In<br />
boggy soils, it is usually found on hummocks. It is found in moist pine forests, heaths,<br />
rhododendron thickets, and (in late successional stages) in sphagnum bogs. It is occasionally<br />
found in wet pine-sphagnum forests (subtypes 3, 4). Fruit harvested as a "crop."<br />
VACCINIUM VITJS-IDAEA<br />
cowberry<br />
cranberry<br />
mountain cranberry<br />
foxberry<br />
It is a dwarf ericoid shrub, up to 30 cm tall, with persistent leaves. It spreads by creeping<br />
rhizomes. It is an oligotrophic acidophyte. In boggy soils, it is usually found on hummocks and<br />
tree boles. It is found in wet pine-sphagnum forest (subtypes 1, 2, 3, 5, 6), wet alderwoods,<br />
heaths, rhododendron thickets, and (in late successional stages) in sphagnum bogs. Fruit<br />
harvested as a "crop."<br />
A-18
APPENDIX B<br />
HERBS<br />
Nonvascular and Lower Vascular Plants<br />
B-1
MOSSES<br />
AULACOMIUM PALUSTRE<br />
green moss<br />
In boggy soils, it is usually found in hollows. It is found in wet pine-sphagnum forests<br />
(subtype 1, 3, 4).<br />
HYLOCOMIUM PROUFORMIS<br />
green moss<br />
It is found in wet pine-sphagnum forest (subtype 3).<br />
PLEUROZIUM SCHREBERI<br />
In boggy soils, it is found on hummocks.<br />
(subtypes 1, 2, 3).<br />
green moss<br />
It is found in wet pine-sphagnum forests<br />
POHLIA NUTANS<br />
green moss<br />
It is oligotrophic. In boggy soils, it is usually found in intermediate sites. It is found in<br />
sphagnum bogs and spruce forests.<br />
POLYIRICHUM COMMUNE<br />
haircap moss<br />
It is oligotrophic. In boggy soils, it is usually found on hummocks and intermediate sites. It is<br />
found in wet pine-sphagnum forests (subtypes 3, 4), sphagnum bogs, and heaths.<br />
POLYTRICHUM STRICTUM<br />
haircap moss<br />
It is found in sphagnum bogs and heaths.<br />
B-2
SPHAGNUM ACUTIFOLUUM<br />
It is oligotrophic and an acidophyte.<br />
(subtypes 3, 5, 6).<br />
It is found in wet pine-sphagnum forests<br />
SPHAGNUM APICULATUM<br />
It is oligotrophic and an acidophyte. In boggy soils, it is usually found in hollows and<br />
intermediate sites. It is found in wet pine-sphagnum forests (subtype 2) and transitional bogs.<br />
SPHAGNUM CENTRALE<br />
It is oligotrophic and an acidophyte. It is found in wet pine-sphagnum forests (subtype 1).<br />
SPHAGNUM COMPACTUM<br />
It is oligotrophic and an acidophyte. It is found in wet pine-sphagnum forests (subtype 2) and<br />
transitional bogs.<br />
SPHAGNUM CUSPIDATUM<br />
It is oligotrophic and an acidophyte. In boggy soils, it is usually found in hollows. It is found<br />
in sphagnum bogs and transitional bogs.<br />
SPHAGNUM FUSCUM<br />
It is oligotrophic and an acidophyte. In boggy soils, it is usually found on hummocks. It is<br />
found in sphagnum transitional bogs, heath, and wet pine-sphagnum forests (subtype 2, 4, 5, 6).<br />
B-3
SPHAGNUM GIRENSOHNII<br />
It is oligotrophic and an acidophyte. In boggy soils, it is usually found on hummocks and<br />
intermediate sites. It is found in wet pine-sphagnum forests (subtype 1).<br />
SPHAGNUM MAGELLANICUM<br />
It is oligotrophic and an acidophyte. In boggy soils it is usually found on hummocks. It is<br />
found in wet pine-sphagnum forests (subtypes 2, 4, 5, 6).<br />
SPHAGNUM MEDIUM<br />
It is oligotrophic and an acidophyte. In boggy soils, it is usually found on hummocks. It is<br />
found in sphagnum bogs.<br />
SPHAGNUM PAPILLOSUM<br />
It is found in transitional bogs.<br />
SPHAGNUM PULCHRUM<br />
It is oligotrophic and an acidophyte. In boggy soils, it is usually found in hollows. It is found<br />
in wet pine-sphagnum forests (subtype 5, 6).<br />
SPHAGNUM RECURVUM<br />
It is oligotrophic and an acidophyte. In boggy soils, it is usually found in hollows. It is found<br />
in sphagnum and transitional bogs.<br />
B-4
SPHAGNUM RUBELLUM<br />
It is oligotrophic and an acidophyte. In boggy soils, it is usually found on hummocks and<br />
intermediate sites. It is found in sphagnum and transitional bogs.<br />
SPHAGNUM RUSSOVII (RUSSOWII)<br />
It is oligotrophic and an acidophyte. In boggy soils, it is usually found on hummocks. It is<br />
found in wet pine-sphagnum forests (subtypes 2, 5, 6) and sphagnum bogs.<br />
SPHAGNUM SQUARROSUM<br />
It is mesotrophic. In boggy soils, it is usually found in hollows and intermediate sites. It is<br />
found in transitional bogs and wet spruce forests.<br />
SPHAGNUM WARNSTORFI!<br />
It is oligotrophic and an acidophyte. It is found in wet pine-sphagnum forests (subtypes 1, 3).<br />
B-5
LICHENS<br />
CLADONIA ALECTORIA<br />
shrubby lichen<br />
It is found in dry pine woodlands.<br />
CLADONIA ALPESTRIS<br />
shrubby lichen<br />
It is found in dry pine woodlands and often dominates in late succession.<br />
CLADONIA RANGIFERINA<br />
shrubby lichen<br />
It is found in dry pine woodlands.<br />
CLADONIA SILVA TICA<br />
forest lichen<br />
It is found in dry pine woodlands.<br />
CLADONIA UNCIALIS<br />
It is found in dry pine woodlands.<br />
CETRARIA ISLANDICA<br />
It is found in dry pine woodlands.<br />
B-6
EQUISETACEAE<br />
EQUISETUM ARVENSE<br />
field horsetail38<br />
fens.<br />
It can grow in deep silt deposits. It is found in wet, flood, and fresh meadows and in sedge<br />
EQUISETUM HELEOCHARIS<br />
swamp horsetail<br />
It can endure prolonged 39 surface flooding and deep deposits of silt. It is found in wet<br />
pine-sphagnum forest (subtype 1), transitional bogs, and wet and flood meadows.<br />
EQUISETUM PALUSTRE<br />
marsh horsetail<br />
It can grow in deep silt deposits. It is found in sedge fens, transitional bogs, and wet and<br />
flood meadows.<br />
EQUISETUM PRA TENSE<br />
meadow horsetail<br />
fens.<br />
In flood meadows, it is found on ridges. It is found in flood and fresh meadows and sedge<br />
EQUISETUM SILVATICUM<br />
forest horsetail<br />
silvan horsetail<br />
It is found in wet pine-sphagnum forests (subtype 2), alder-ash carts, and forest openings.<br />
38 Eqiaetium is ais b known<br />
scouring nbh.<br />
39 M=e don 40 dayL<br />
B-7
EQUISETUM VARIEGATUM . variegated horsetail<br />
It can grow in deep silt deposits. It is found in wet pine-sphagnum forests (subtype 5),<br />
sphagnum bogs, transitional bogs, and sedge fens.<br />
B-8
MONOCOTYLEDONS<br />
Typhaceae<br />
TYPHA ANGUSTIFOLJA<br />
narrow-leafed cattail4o<br />
It is a hydrophyte. It is found in reed-bulrush fens.<br />
TYPHA LATIFOLIA<br />
broad-leafed cattail<br />
It is a hydrophyte. It is found in reed-bulush fens.<br />
Juncaceae<br />
JUNCUS AMBIGUUS<br />
It is found in wet and flood meadows.<br />
JUNCUS BULBOSA<br />
bulbous rush<br />
tuberous rush<br />
It is found in sphagnum and sphagnum-pine bogs.<br />
JUNCUS COMPRESSUS<br />
round-fruited rush<br />
It is found in reed-bulrush fens, and sedge fens, transitional bogs, and wet and flood<br />
meadows.<br />
4&pm ais kam ow ars<br />
mnaa<br />
B-9
JUNCUS EFFUSUS<br />
loose-spreading rush<br />
It is found in transitional bogs and wet meadows.<br />
JUNCUS INFLEXUS<br />
It is found in sedge fens.<br />
JUNCUS LAMPOCARPUS<br />
It is found in transitional bogs.<br />
JUNCUS LEERSII<br />
It is found in transitional bogs.<br />
JUNCUS SQUARROSUS<br />
squarrose rush<br />
It is found in heathlands, sphagnum bogs, and transitional bogs.<br />
JUNCUS TENAGEJA<br />
It is found in reed-bulrush fens, sedge fens, and wet and flood meadows.<br />
B-10
Cyperaceae<br />
CAREX ACU77FORMIS<br />
sharp sedge<br />
It is mesotrophic. In boggy soils, it is commonly found in hollows and intermediate sites. It is<br />
found in sedge fens and transitional bogs.<br />
CAREX APPROPINQUATA<br />
It is eutrophic. In boggy soils, it is commonly found on hummocks and intermediate sites. It<br />
is found in sedge fens and transitional bogs.<br />
CAREX AQUA T7LIS<br />
water sedge<br />
straight-leafed water sedge<br />
It is oligotrophic. In boggy soils, it is found in hollows and intermediate sites. It is found in<br />
wet and flood meadows, sedge fens, and transitional bogs.<br />
CAREX ARGYROGLOCHIN<br />
silver-arrow sedge<br />
It is found in alderwoods and spruce forest.<br />
CAREX CAESPITOSA<br />
tufted sedge, turfy sedge<br />
It is mesotrophic and a hydrophyte. In boggy soils, it is usually found in hollows and<br />
intermediate sites. It is found in wet and flood meadows, sedge fens, transitional bogs, and wet<br />
alderwoods.<br />
CAREX CANESCENS<br />
gray sedge<br />
hoary sedge<br />
It is found in wet meadows and transitional bogs.<br />
B-II
CAREX CHORDORRHIZA<br />
It is oligotrophic. In boggy soils, it is usually found in hollows and intermediate sites. It is<br />
found in sphagnum bogs, transitional bogs, and sedge fens.<br />
CAREX DIANDRA<br />
diandrous sedge<br />
It is rhizomatous. It is mesotrophic. In boggy soils, it is usually found in hollows and intermediate<br />
sites. It is found in sedge fens, transitional bogs, and wet meadows.<br />
CAREX DIOICA<br />
dioecious sedge<br />
creeping separate-headed sedge<br />
It is mesotrophic. In boggy soils, it is usually found on hummocks and in intermediate sites.<br />
It is found in sedge fens, transitional bogs, and wet meadows.<br />
CAREX DISPERMA<br />
dispermous sedge<br />
It is mesotrophic. In boggy soils, it is usually found on hummocks and in intermediate sites.<br />
It is found in transitional bogs.<br />
CAREX DISTJCTA<br />
distictous sedge<br />
It is rhizomatous. It is found in transitional bogs and wet meadows.<br />
CAREX FLAVELLA<br />
little yellow sedge<br />
It is found in wet meadows and willow-poplar cars.<br />
B-12
CAREX GRACILIS<br />
sheep sedge, marsi, sedge<br />
slender sedge<br />
It is mesotrophic and a hydrophyte. It is nonresistant to subsurface flooding but can endure<br />
prolonged surface flooding. It can grow in moderate silt deposits. In boggy soils, it is usually<br />
found in hollows and intermediate sites. It is found in flood and wet meadows, sedge fens, and<br />
transitional bogs.<br />
CAREX HELEONASTES<br />
It is mesotrophic. In boggy soils, it is usually found in hollows. It is found in sedge fens,<br />
transitional bogs, and sphagnum bogs.<br />
CAREX INUMBRATA<br />
It is found in wet alderwoods.<br />
CAREX JUNCELLA<br />
little-rush sedge<br />
It is oligotrophic. In boggy soils, it is found in hollows and intermediate sites. It is found in<br />
transitional bogs and wet meadows.<br />
CAREX LASIOCARPA<br />
hairy fruited sedge<br />
It is oligotrophic. In boggy soils, it is found in hollows and intermediate sites. It is found in<br />
wet pine-sphagnum forests (subtypes 1, 2, 4), transitional bogs, sedge fens, and wet meadows.<br />
CAREX LEPORINA<br />
hare's-foot sedge<br />
It is found in fresh meadows and at forest margins.<br />
B-13
CAREX LIMOSA<br />
bog sedge<br />
mud sedge<br />
It is oligotrophic. In boggy soils, it is usually found in hollows. It is found in sphagnum<br />
bogs, transitional bogs, sedge fens, and wet meadows.<br />
CAREX LOLJACEA<br />
It is mesotrophic. In boggy soils, it is found on hummocks and intermediate sites. It is found<br />
in spruce forests.<br />
CAREX MURICATA<br />
muricate sedge<br />
greater prickly sedge<br />
It is found in transitional bogs and wet meadows.<br />
CAREX OMSKIANA<br />
Omskian sedge<br />
It is found in transitional bogs and reed-bulrush fens.<br />
CAREX PANICEA<br />
pink-leafed sedge<br />
It is eutrophic. In boggy soils, it is found in hollows and intermediate sites. It is found in wet<br />
meadows, and sedge fens.<br />
CAREX PANICULATA<br />
paniculate sedge<br />
great panicled sedge<br />
It is eutrophic. In boggy soils, it is found in hollows. It is found in sedge fens, reed-bulrush<br />
fens, and wet meadows.<br />
B-14
CAREX PAUCIFLORA<br />
few-flowered sedge<br />
It is oligotrophic. In boggy soils, it is found on hummocks and intermediate sites. It is found<br />
in transitional bogs and sphagnum bogs.<br />
CAREX PILOSA<br />
hairy sedge<br />
It is eutrophic. It is found in drier oak-hornbeam forests.<br />
CAREX PSUEDO-CYPERUS<br />
false flatsedge<br />
It is mesotrophic. In boggy soils, it is found in hollows and intermediate sites. It is found in<br />
sedge fens and transitional bogs.<br />
CAREX REMOTA<br />
distant-spiked sedge<br />
It is found in alder-ash carrs.<br />
CAREX RIPARIA<br />
riparian sedge<br />
great common sedge<br />
riverine sedge<br />
It is mesotrophic. In boggy soils, it is usually found in hollows. It is found in wet and flood<br />
meadows, sedge fens, and transitional bogs.<br />
CAREX VESICARIA<br />
bladder sedge<br />
short-beaked bladder sedge<br />
It is found in transitional bogs, sedge fens, and wet and flood meadows.<br />
CAREX VULPINA<br />
fox sedge<br />
great sedge<br />
It is found in wet meadows, sedge fens, and transitional bogs.<br />
B-15
CYPERUS FLAVESCENS<br />
yellow flatsedge yellow galingale<br />
It is found in sedge fens.<br />
CYPERUS PANONICUS<br />
It is found in sedge fens.<br />
ERIOPHORUM ANGUSTJFOLIUM<br />
narrow-leafed<br />
cottongrass 4 1<br />
It is found in sedge fens, transitional bogs, wet meadows, and wet pine-sphagnum forest<br />
(subtype 2).<br />
ERIOPHORUM GRACILE<br />
slender cottongrass<br />
It is found in sedge fens, transitional bogs, and wet pine-sphagnum forest (subtype 2).<br />
ERIOPHORUM LATIFOLJUM<br />
broad-leafed cottongrass<br />
It is found in sedge fens, transitional bogs, and wet meadows.<br />
ERIOPHORUM VAGINA TUM<br />
sheathed cottongrass<br />
haretail cottongrass<br />
In boggy soils, it is usually found on hummocks. It is found in wet pine-sphagnum forest<br />
(subtype 2), wet birch woodlands, sphagnum bogs, and transitional bogs.<br />
4 1 Erphmm is sao caled axomede and boVwOOL<br />
B-16
SCIRPUS ACICULARIS needle bulrush 42<br />
It is found in reed-bulrush fens.<br />
SCIRPUS COMPRESSUS<br />
compressed bulrush<br />
It is found in wet meadows.<br />
SCIRPUS EUPALUSTRUS<br />
marsh bulrush, marsh clubrush<br />
creeping scirpus<br />
It is found in wet and flood meadows and in reed-bulrush fens.<br />
SCIRPUS HOLOSCHOENUS<br />
It is found in reed-bulrush fens and wet meadows.<br />
SCIRPUS LACUSTRIS<br />
lake bulrush<br />
lake clubrush<br />
It is mesotrophic and a hydrophyte. In boggy soils, it is found in hollows. It is dominant in<br />
reed-bulrush fens. It is found in wet meadows and sedge fens.<br />
SCIRPUS MAMILLATUS<br />
It is found in reed-bulrush fens.<br />
SCIRPUS OVA TUS<br />
ovate bulrush<br />
It is found in reed-bulrush fens and wet meadows.<br />
42 Bulpjh is also spelled bullrus.<br />
B-17
SCIRPUS PA UCIFLOR US<br />
few-flowered bulrush<br />
chocolate-headed clubrush<br />
It is found in reed-bulrush fens and wet meadows.<br />
SCIRPUS SILVATICUS<br />
wood scirpus, wood bulrush<br />
It is mesotrophic. In boggy soils, it is found in hollows and intermediate sites. It is found in<br />
wet meadows, reed-bulrush fens, and spruce forests.<br />
B-18
POACEAE- 3<br />
AGROPYRON CANINUM<br />
dog wheatgrass<br />
dog couchgrass<br />
doggrass<br />
It is mesotrophic. It is found in oak-hornbeam and spruce forests.<br />
AGROPYRON REPENS<br />
quackgrass<br />
creeping couchgrass<br />
creeping wheatgrass<br />
creeping doggrass<br />
twitchgrass, quitchgrass<br />
quickgrass, quakegrass<br />
squitchgrass<br />
Is a mesohydrophyte. It is nonresistant to subsurface flooding but can endure prolonged<br />
surface floods and deep deposits of silt. It is often dominant in flood meadows. It is also found in<br />
fresh meadows and at forest margins and in openings.<br />
AGROSTIS ALBA<br />
redtop bentgrass<br />
white bentgrass<br />
marsh bentgrass<br />
redtop florin<br />
bonnetgrass<br />
It is mesotrophic. It is a mesophyte and can endure prolonged surface flooding. It can adjust<br />
to acid soils. It is found in wet and flood meadows. It is recommended for cultivation in flood<br />
meadows and drained bogs.<br />
AGROSTIS CANINA<br />
velvet bentgrass<br />
dog bentgrass<br />
brown bentgrass<br />
finetop bentgrass<br />
mountain redtop<br />
It is mesotrophic. It is found in wet and flood meadows.<br />
B-19
AGROSTIS STOLONIFERA<br />
carpet bentgrass<br />
creeping bentgrass<br />
black twitch<br />
It is mesotrophic and a mesophyte. It can endure moderate"4 subsoil flooding. It is found in<br />
wet and flood meadows and on the peripheries of transitional bogs.<br />
ALOPECURUS ARUNDINA CEUS<br />
creeping foxtail<br />
In flood meadows, it is found in interridge depressions.<br />
meadows, sedge fens, and transitional bogs.<br />
It is found in wet and flood<br />
ALOPECURUS GENICULATUS<br />
marsh foxtail, water foxtail<br />
floating foxtail<br />
elbowit-grass<br />
It is found in wet meadows, sedge fens, and transitional bogs.<br />
ALOPECURUS PRA TENSIS<br />
meadow foxtail<br />
It is a mesohydrophyte. It can endure moderate subsoil flooding and prolonged surface<br />
flooding. It is found in wet, flood, and fresh meadows. It is recommended for cultivation in fresh<br />
and flood meadows and drained bogs.<br />
ALOPECURUS TENUIS<br />
slender foxtail<br />
It is found in flood and fresh meadows and birch woodlands.<br />
ANTHOXANTHUM ORDORATUM<br />
sweet vernalgrass<br />
sweet-scented vernalgrass<br />
It is mesotrophic and a mesophyte. It is found in fresh and flood meadows.<br />
4420 to 40 days.<br />
B-20
ARRHENATHERUM ELATIUS<br />
tall oatgrass<br />
meadow oatgrass<br />
false oatgrass<br />
French ryegrass<br />
It is a mesophyte. It is nonresistant to subsurface flooding but can endure short4 5 surface<br />
floods. It is found in fresh meadows and flood meadows with short inundation periods.<br />
BECKMANNIA ERUCIFORMIS<br />
common sloughgrass<br />
It is a mesohydrophyte. It is able to endure prolonged surface flooding. It is found in wet,<br />
flood, fresh meadows, and reed-bulrush fens. It is recommended for cultivation in drained bogs.<br />
BRA CHYPODIUM PINNA7TUM<br />
heath falsebrome<br />
It is often dominant at forest and woodland margins and in forest openings.<br />
BRA CHYPODIUM SILVATICUM<br />
forest falsebrome<br />
slender flasebrome<br />
It is found in moist pine woodlands.<br />
BRIZ4 MEDIA<br />
common quaking grass<br />
doddering jockies<br />
doddering dillies<br />
earthquakes<br />
It is found in fresh meadows and forest openings.<br />
45Ln thma 20 days.<br />
B-21
BROMUS INERMIS<br />
awnless bromegrass<br />
awnless brome<br />
smooth brome<br />
Hungarian brome<br />
It is a mesophyte. It cannot resist subsoil flooding but can endure moderate46 surface flooding<br />
and deep deposits of silt. In flood meadows, it is found in interridge depressions. It is found in<br />
flood and fresh meadows and at forest margins and in openings. It is recommended for cultivation<br />
in fresh meadows and drained bogs.<br />
CALAMAGROSTIS ARUNDINACEA<br />
forest reedgrass<br />
sand reedgrass<br />
beachgrass<br />
It is a mesophyte. It inhibits pine regeneration in successional woodlands. Found in linden<br />
and aspen groves and birch woodlands.<br />
CALAMAGROSTIS EPIGEIOS<br />
chee reedgrass<br />
small reedgrass<br />
wood smallreed<br />
It is mesotrophic and a mesophyte. It can grow in deep silt and can endure moderate surface<br />
flooding. It inhibits pine regeneration in successional woodlands. It is found in flood meadows,<br />
wet pine-sphagnum forest (subtype 2, 4), linden and aspen groves, and birch woodlands.<br />
CAL4MAGROSTIS NEGLECTA<br />
slim-stem reedgrass<br />
It is mesotrophic. It is found in wet meadows and at the peripheries of transitional bogs.<br />
CALAMAGROSTIS PHRAGMITOIDES<br />
tall reedgrass<br />
It is oligotrophic. It is found in wet meadows and damp spruce forests.<br />
4620 so 40 dap.<br />
B-22
CA TABROSA AQUA TI CA<br />
brookgrass<br />
water whorigrass<br />
It is found in wet meadows.<br />
CYNOSURUS CRISTA TUS<br />
crested dogtail<br />
crested dog's-tail grass<br />
It is a mesophyte. It is found in fresh meadows and moist forest openings.<br />
DACT"YJS GLOMERATA<br />
common owhardgrass<br />
cocksfoot<br />
rough cocksfoot<br />
It is a mesophyte and can endure moderate surface flooding. It is not able to grow in deep or<br />
moderate silt deposits but can grow in thin deposits. It is found in fresh and flood meadows and in<br />
forest openings. It is recommended for cultivation in fresh meadows.<br />
DESCHAMPSIA CAESPITOSA<br />
tufted hairgrass<br />
It is mesotrophic and a mesohydrophyte. It can endure moderate subsoil flooding. It grows in<br />
moist to very wet soils and in thin deposits of silt. It is found in wet and flood meadows, in<br />
alderwoods, and spruce forests.<br />
FESTUCA ARUNDINACEA<br />
tall fescue, reed fescue<br />
alta fescue<br />
It is found in wet and flood meadows.<br />
FESTUCA GIGANTEA<br />
giant fescue<br />
tall-bearded fescue<br />
It is found in wetter oak-hornbeam forests.<br />
B-23
FESTUCA OVINA<br />
sheep fescue<br />
sheep's fescue<br />
It is mesotrophic. It can grow in thin deposits of silt and can e;. lure short periods of flooding.<br />
In flood meadows, it is found on ridges. It is found in fresh meadows and flood meadows with<br />
only short floods, as well as dry pine woodlands and on dry sandy hills and dunes.<br />
FES7TUCA POLESICA<br />
Polish fescue<br />
It is found in dry pine woodlands and on dry sandy hills and dunes.<br />
FESTUCA PRATENSIS<br />
meadow fescue<br />
It is mesotrophic. It is a mesophyte but can endure moderate surface flooding. It is not able to<br />
grow in deep silt deposits but can grow in moderate deposits. It is found in wet, flood, and fresh<br />
meadows and in moist pine-sphagnum forests. It is recommended for cultivation in flood and<br />
fresh meadows and drained bogs.<br />
FESTUCA RUBRA<br />
red fescue, creeping fescue<br />
It is mesotrophic and a mesohydrophyte. It can endure moderate subsoil and surfacc flooding.<br />
It can grow in poor soils and in thin deposits of silt. It is found in wet, flood, and fresh meadows,<br />
as well as moist pine woodlands.<br />
GLYCERIA AQUATICA<br />
reed mannagrass<br />
reed sweetgrass<br />
reed meadowgrass<br />
reed whitegrass<br />
It is mesotrophic and a hydrophyte. It is found in reed-bulrush fens, sedge fens, and wet and<br />
flood meadows.<br />
B-24
GLYCERIA FRUITANS<br />
floating mannagrass<br />
It is mesotrophic and a hydrophyte. It is found in reed-bulrush fens and wet and flood<br />
meadows.<br />
GLYCERIA HEMORALIS<br />
It is found in wet alderwoods and reed-bulrush fens.<br />
GLYCERIA LITHUANICA<br />
Lithuanian mannagrass<br />
It is eutrophic. It is found in wet alderwoods and spruce forests.<br />
GLYCERIA PLJCATA<br />
plicate mannagrass<br />
It is found in wet and flood meadows.<br />
HIEROCHLOE ODORATA<br />
common sweetgrass<br />
sweet holygrass<br />
sweet vanillagrass<br />
sweet senecagrass<br />
It is mesotrophic. It is found in wet and flood meadows.<br />
HOLCUS LANATUS<br />
common velvetgrass<br />
meadow softgrass<br />
Yorkshire fog<br />
It is found in fresh meadows.<br />
HOLCUS MOLI.S<br />
downy velvetgrass<br />
creeping softgrass<br />
It is found in fresh meadows.<br />
B-25
KOELERIA DELAVIGNII<br />
Delavign junegrass<br />
It is found in flood meadows and oak forests.<br />
KOELERIA GLAUCA<br />
glaucus junegrass<br />
It is found in dry pine woodlands and on dry sandy hills and dunes.<br />
LEERSIA ORYZOIDES<br />
rice cutgrass<br />
rice whitegrass<br />
false rice<br />
It is found in sedge and reed-bulrush fens.<br />
LOLIUM PERENNE<br />
perennial ryegrass<br />
English ryegrass<br />
perennial raygrass<br />
It is a mesophyte and is nonresistant to subsurface flooding. It has only weak resistance to<br />
drought It is found in fresh meadows and is recommended for cultivation there.<br />
MELICA NUTANS<br />
mountain melic<br />
It is found in aspen groves, oak-hornbeam forests, and forest openings.<br />
MILIUM EFFUSSUM<br />
wood millet<br />
spreading milletgrass<br />
It is mesotrophic. It is found in alderwoods, oak-hornbeam forests, birch woodlands, aspen<br />
and linden groves, and spruce forests.<br />
B-26
MOLJNIA COERULUA<br />
purple heathgrass<br />
purple moorgrass<br />
purple molinia<br />
It is mesotrophic and a hydrophyte. In boggy soils, it is usually found on hummocks and<br />
intermediate sites. It is found in wet meadows, moist pine forests, wet alderwoods, at the<br />
peripheries of transitional bogs, and in spruce forests.<br />
NARDUS STRICTA<br />
matgrass<br />
It is oligotrophic and mesophytic. It can grow in poor soils and in thin deposits of silt. In<br />
boggy soils, it is usually found in intermediate sites. It is found in wet, flood, and fresh<br />
meadows, heathlands, moist pine forests, and wet pine woodlands with flowing water. It grows<br />
in moist and very moist "bor" and "subor" sites.<br />
PHALARIS ARUNDINACEA<br />
reed canarygrass<br />
reed swordgrass<br />
It is mesotrophic and a mesohydrophyte. It is able to endure prolonged surface flooding. It<br />
can grow in deep silt deposits. In boggy soils, it is usually found in intermediate sites. It is found<br />
in reed-bulrush fens and in wet, flood, and fresh meadows. It is recommended for cultivation in<br />
fresh and flood meadows and drained bogs.<br />
PHLEUM PRATENSE<br />
timothy, common timothy<br />
meadow timothy<br />
common cat's-tafilgrass<br />
meadow cat's-tailgrass<br />
herd's grass<br />
It is mesotrophic and a mesophyte and can endure moderate, but not prolonged, surface<br />
flooding. It is found in flood and fresh meadows. It is recommended for cultivation in flood and<br />
fresh meadows and drained bogs and is often supplementally sown for forage.<br />
B-27
PHRAGMITES COMMUNIS<br />
common reed<br />
giant reed<br />
ditch reed<br />
pool reed<br />
It is mesotrophic and a hydrophyte. In boggy soils, it is usually found in hollows and<br />
intermediate sites. It is dominant in reed-bulrush fens. It is found in wet pine forests with flowing<br />
water, in wet alderwoods, and wet meadows.<br />
POA ANNUA<br />
annual meadowgrass<br />
dwarf meadowgrass<br />
annual bluegrass<br />
causeway grass<br />
Suffolkgrass<br />
low speargrass<br />
It is found in fresh meadows.<br />
POA PALUSTRIS<br />
fowl meadowgrass<br />
fowl bluegrass<br />
swamp meadowgrass<br />
false redtop<br />
It is mesotrophic and a mesohydrophyte. It can endure prolonged surface flooding and can<br />
grow in moderate silt deposits. In flood meadows, it is found in interridge depressions. It is<br />
found in wet, flood, and fresh meadows and sedge fens.<br />
POA PRATENSIS<br />
smooth meadowgrass<br />
smooth-stalked meadowgrass<br />
Kentucky bluegrass<br />
junegrass<br />
speargrass<br />
greengrass<br />
It is a mesophyte and can endure moderate surface flooding. It is not able to grow in deep or<br />
moderate silt deposits but can grow in thin deposits. It is found in flood and fresh meadows. It is<br />
recommended for cultivation in flood and fresh meadows and drained bogs and is often<br />
supplementally sown for forage. It is also found in dry pine woodlands and on sandy hills and<br />
dunes.<br />
B-28
PAO TRIVIALIS<br />
rough meadowgrass<br />
rough-stalked meadowgrass<br />
rough bluegrass<br />
rough-stalked bluegrass<br />
evergreen<br />
It is mesotrophic and a mesophyte. It can endure moderate subsoil flooding. It is found in<br />
flood and fresh meadows and is often supplementally sown for forage.<br />
SETARIA GLA UCA<br />
glaucus bristlegrass<br />
glaucus foxtail<br />
pigeongrass<br />
bottlegrass<br />
It is found in sedge fens.<br />
SETARIA WRIDIS<br />
green bristlegrass<br />
green foxtail<br />
green pigeongrass<br />
It is found in sedge fens.<br />
TRISETUM SIBIRICUM<br />
Siberian trisetum<br />
Siberian falseoat<br />
Siberian oatgrass<br />
It is found in wet meadows.<br />
B-29
DICOTYLEDONS<br />
DROSERA ROTUNDIFOUA<br />
round-leaf sundew<br />
It is carnivorous, oligotrophic, and an acidophyte. In boggy soils, it is usually found on<br />
hummocks. It is found in sphagnum bogs and transitional bogs.<br />
MEDICAGO FALCATA<br />
yellow lucerne<br />
yellow alfalfa<br />
It is a mesophyte and can endure short surface floods. It is not able to grow in deep silt<br />
deposits. In flood meadows, it is found in interridge depressions. It grows in fresh meadows,<br />
flood meadows with only short inundations periods, and drained bogs. It is often supplementally<br />
sown for forage.<br />
MEDICAGO LUPULINA<br />
black medic<br />
hop medic<br />
nonesuch<br />
It is a mesophyte, It grows in fresh meadows and is often supplementally sown for forage.<br />
MEDICAGO SA TIVA<br />
common lucerne<br />
common alfalfa<br />
purple medic<br />
Spanish trefoil<br />
It is a mesophyte and can endure short surface floods. It will not grow in excessively moist<br />
soils or soil with even moderately long flooding. It grows in fresh meadows, flood meadows with<br />
short inundation periods, and drained bogs and is often supplementally sown for forage.<br />
MELILOTUS ALBA<br />
white sweetclover<br />
white melilot<br />
Bokara clover<br />
It is a mesophyte. It is found in fresh meadows and is often supplementally sown for forage.<br />
B-30
MELULOTUS OFFICINAUS<br />
yellow sweetclover<br />
yellow melilot<br />
It is a mesophyte. It grows in fresh meadows and is often supplementally sown for forage.<br />
TPJFOLIUM HYBRIDUM<br />
alsike clover, pink clover<br />
Swedish clover<br />
It is a mesophyte. It is not able to grow in deep silt deposits but can grow in moderate<br />
deposits. It can endure moderately acid soils. It grows in flood and fresh meadows and pastures.<br />
It is recommended for cultivation in fresh and flood meadows and drained bogs and is often<br />
supplementally sown for forage.<br />
TRIFOLJUM PRA TENSE<br />
red clover, meadow clover<br />
common clover<br />
broad-leaf clover<br />
cowgrass clover<br />
meadow trefoil<br />
It is mesotrophic and a mesophyte. It is not able to grow in deep silt deposits but can grow in<br />
moderate deposits. It can withstand short surface floods and soil acidity. It is found in fresh<br />
meadows, flood meadows with short inundation periods, and pastures on relatively fertile, moist<br />
(but well-drained) soils. It is recommended for cultivation in fresh meadows and drained bogs and<br />
is often supplementally sown for forage (see T. sativum).<br />
TRIFOLUUM REPENS<br />
white clover<br />
Dutch clover<br />
It is mesotrophic and a mesophyte. It grows best on moderately moist soils and can endure<br />
short surface floods. It is not able to grow in deep silt deposits but can grow in moderate deposits.<br />
It is found in fresh meadows and flood meadows with short inundation periods, usually on welldrained<br />
soils It is recommended for cultivation in fresh and flood meadows and is often<br />
supplementally sown for forage.<br />
TRIFOLJUM SATIVUM<br />
field clover<br />
This is a name for the cultivated form of T pratense.<br />
B-31
..APPENDIX C<br />
CROP PLANTS<br />
C-I
A VENA SA TIVA<br />
oats (spring)<br />
In the Poles'ye, it is cultivated on moderate podzols.<br />
BETA VULGARIS<br />
sugar beet, table beet<br />
forage or fodder beet<br />
It is grown for forage and silage as well as human consumption. In the Poles'ye, it is<br />
cultivated on moderate podzols and, with wheat, "dominates" the best soils.<br />
BRASSICA OLERACEA var. A CEPHALA<br />
kale<br />
BRASSICA OLERACEA var. CAPITATA<br />
cabbage<br />
BRASSICA RAPA<br />
turnips<br />
CANNABIS SATIVA<br />
hemp<br />
It is grown for fiber.<br />
CUCUMIS SA TIVA<br />
cucumbers<br />
DAUCUS CAROTA<br />
table carrots<br />
fodder carrots<br />
In the Poles'ye it is grown in drained bogs.<br />
FAGOPYRUM ESCULENTUM<br />
buckwheat<br />
In the Poles'ye, it is cultivated on poor and moderate podzols.<br />
C-2
HORDEUM VULGARE<br />
barley (winter and spring)<br />
In the Poles'ye, it is cultivated on poor and moderate podzols.<br />
HUMULUS LUPULUS<br />
hops<br />
In the Poles'ye, it is cultivated on moderate podzols.<br />
UNUM USITATISSIMUM<br />
flax<br />
It is grown for fiber. Fine-quality fiber can be obtained from plants grown on podzolic and<br />
gley soils with considerable fertilizing. In the Poles'ye, it is cultivated on poor and moderate<br />
podzols.<br />
PANICUM MILIACEUM<br />
millet<br />
In the Poles'ye, it is cultivated on poor and moderate podzols.<br />
SECALE CEREALE<br />
rye (winter)<br />
In the Poles'ye, it is cultivated on poor and moderate podzols.<br />
SOLANUM TUBEROSUM<br />
potato<br />
In the Poles'ye, it is cultivated on poor and moderate podzols.<br />
TRITICUM AESTIVUM<br />
wheat (winter and spring)<br />
In the Poles'ye, it is grown only on the best soils and, with beets, "dominates" them.<br />
C-3
ZEA MAYS<br />
corn<br />
It is usually grown as a forage or silage plant. In the Poles'ye, it is cultivated on moderate<br />
podzols.<br />
C-4
DEPARTMENT OF DEFENSE<br />
ARMED FORCES RADIOBIOLOGY RSCH INST<br />
ATTN: DEPT OF RADIATION BIOCHEMISTRY<br />
ATTN: BHS<br />
ATTN: DIRECTOR<br />
ATTN: EXH<br />
ATTN: MRA<br />
ATTN: PHY<br />
ATTN: RSD<br />
ATTN: SCIENTIFIC DIRECTOR<br />
ATTN: TECHNICAL LIBRARY<br />
ASSISTANT SECRETARY OF DEFENSE<br />
INTERNATIONAL SECURITY POLICY<br />
ATrN: NUC FORCES & ARMS CONTROL PLCY<br />
ASSISTANT TO THE SECRETARY OF DEFENSE<br />
ATTN: EXECUTIVE ASSISTANT<br />
ATTN: MIL APPL C FIELD<br />
DISTRIBUTION LIST<br />
DNA-TR-92-37-V3<br />
JOINT DATA SYSTEM SUPPORT CTR<br />
ATTN: C-332<br />
ATTN: JNSV<br />
NATIONAL DEFENSE UNIVERSITY<br />
ATrN: ICAF TECH LIB<br />
ATTN: NWCLB-CR<br />
ATrN: LIBRARY<br />
ATTN: STRAT CONCEPTS DIV CTR<br />
NET ASSESSMENT<br />
ATTN: DOCUMENT CONTROL<br />
OFC OF MILITARY PERFORMANCE<br />
ASSESSMENT TECHNOLOGY<br />
ATTN: F HEGGE<br />
STRATEGIC AND THEATER NUCLEAR FORCES<br />
ATTN: DR E SEVIN<br />
ATrN: DR SCHNEITER<br />
DEFENSE INTELLIGENCE AGENCY<br />
THE JOINT STAFF<br />
ATTN: DB<br />
ATTN: JLT<br />
5 CYS ATTN: DB-4 RSCH RESOURCES DIV ATTN: JPEP<br />
ATTN: DB-SC<br />
ATTN: DB-6B DB-6E<br />
U S EUROPEAN ATTN: ECJ-6 COMMAND/ECJ-6-DT<br />
ATTN: DIANPA-2<br />
ATTN: DIW-4<br />
ATTN: DN<br />
ATrN: DT<br />
ATTN: OFFICE OF SECURITY<br />
ATTN: OS<br />
DEFENSE INTELLIGENCE COLLEGE<br />
ATTN: DIC/RTS-2<br />
ATTN: DIC/2C<br />
DEFENSE LOGISTICS AGENCY<br />
ATTN: COMMAND SECURITY OFC<br />
U S EUROPEAN COMMAND/ECJ2-T<br />
ATTN: ECJ2-T<br />
U S EUROPEAN COMMAND/ECJ3-CCD<br />
ATTN: ECJ-3<br />
U S EUROPEAN COMMAND/ECJ4-LW<br />
ATTN: ECJ4-LW<br />
U S EUROPEAN COMMAND/ECJ5-N<br />
ATTN: ECJ5-N<br />
U S EUROPEAN COMMANDiECJ7/LW<br />
ATrN: ECJ-7-LW<br />
DEFENSE NUCLEAR AGENCY<br />
ATTN: DFRA JOAN MA PIERRE<br />
UNDER SEC OF DEFENSE FOR POLICY<br />
ATIN: NANF<br />
ATTN: DUSP/P<br />
ATTN: NASF<br />
ATTN: USD/P<br />
ATTN: OPNA<br />
ATTN: OPNS<br />
DEPARTMENT OF THE ARMY<br />
ATTN: RAEE<br />
20 CYS ATTN: RARP COMBAT MATERIAL EVAL ELEMENT<br />
2 CYS ATTN: TITL ATTN: SECURITY ANALYST<br />
DEFENSE TECHNICAL INFORMATION CENTER<br />
DEP CH OF STAFF FOR OPS & PLANS<br />
2 CYS ATTN: DTIC/FDAB ATTN: DAMO-SWN<br />
ATTN: DAMO-ZXA<br />
FIELD COMMAND DEFENSE NUCLEAR AGENCY<br />
ATTN: FCPR<br />
HARRY DIAMOND LABORATORIES<br />
ATTN: FCPRT<br />
ATTN: SLCH-NW-RS JOSIP SOLN<br />
ATTN: NUC SECURITY<br />
ATTN: SLCIS-M-TL<br />
FIELD COMMAND DEFENSE NUCLEAR AGENCY<br />
JOINT SPECIAL OPERATIONS COMMAND<br />
ATTN: FCNM ATIN: J-2<br />
ATTN: J-5<br />
INTERSERVICE NUCLEAR WEAPONS SCHOOL<br />
ATTN: ITV NUCLEAR EFFECTS DIVISION<br />
2CYS ATTN: T'TV3416THTTSO<br />
ATTN: STEWS-NE-T<br />
Dist-1
DNA-TFI-92-37-V3 (DL CONTINUED)<br />
U S ARMY AIR DEFENSE ARTILLERY SCHOOL<br />
ATFN: COMMANDANT<br />
U S ARMY ARMAMENT RSCH DEV & ENGR CTR<br />
ATTN: DRDAR-LCN-F<br />
USA SURVIVABILITY MANAGMENT OFFICE<br />
ATTN: SLCSM-SE J BRAND<br />
USACACDA<br />
ATTN: ATZL-CAD-N<br />
U S ARMY ARMOR SCHOOL<br />
DEPARTMENT OF THE NAVY<br />
ATTN: ATSB.CTD<br />
ATTN: TECH LIBRARY<br />
MARINE CORPS<br />
ATTN: CODE PPO<br />
U S ARMY BALLISTIC RESEARCH LAB<br />
ATTN: PSI GIRASP<br />
ATTN: SLCBR-D<br />
ATTN: SLCBR-DD-T<br />
NAVAL OCEAN SYSTEMS CENTER<br />
ATfN: SLCBR-TB ATTN: CODE 9642-B<br />
ATTN:<br />
ATTN:<br />
SLCBR-VL-I<br />
SLCBR-VL-V<br />
DR KLOPCIC<br />
NAVAL PERSONNEL RES & DEV CENTER<br />
ATTN: CODE P302<br />
U S ARMY COMBAT SYSTEMS TEST ACTIVITY<br />
NAVAL POSTGRADUATE SCHOOL<br />
ATTN: JOHN GERDES ATGRCDE 1424 OL<br />
ATTN: MIKE STANKA<br />
ATTN: CODE 1424 LIBRARY<br />
U S ARMY COMD & GENERAL STAFF COLLEGE<br />
NAVAL<br />
ATTESACO<br />
RESEARCH LABORATORY<br />
1240<br />
ATTN: ATZL-SWJ-CA ATTN: CODE 1240<br />
ATIN: ATZL-SWT-A ATTN: CODE 2627<br />
U S ARMY CONCEPTS ANALYSIS AGENCY<br />
ATIN: TECHNICAL LIBRARY<br />
U S ARMY FIELD ARTILLERY SCHOOL<br />
ATTN: ATSF-C:D<br />
U S ARMY FORCES COMMAND<br />
ATTN: AF-OPTS<br />
U S ARMY FOREIGN SCIENCE & TECH CTR<br />
ATTN: C WARD AIFRTC<br />
NAVAL SEA SYSTEMS COMMAND<br />
ATTN: PMS-423<br />
ATTfN: SEA-06GN<br />
NAVAL SURFACE WARFARE CENTER<br />
ATTN: CODE F-31<br />
ATTN: G RIEL<br />
NAVAL TECHNICAL INTELLIGENCE CTR<br />
ATTN: NTIC-DA30<br />
NAVAL WAR COLLEGE<br />
U S ARMY INFANTRY CENTER ATfTN: CODE E-1 1<br />
ATTN: ATSH-CD-CSO<br />
ATTN: CTR FOR NAV WARFARE STUDIES<br />
ATTN: DOCUMENT CONTROL<br />
U S ARMY ITAC<br />
ATTN: LIBRARY<br />
ATrN: IAX-Z<br />
ATTN: STRATEGY DEPT<br />
U S ARMY LABORATORY COMMAND<br />
ATTN: DIRECTOR<br />
ATTN: DR D HODGE<br />
NAVAL WEAPONS EVALUATION FACILITY<br />
ATTN: CLASSIFIED LIBRARY<br />
NUCLEAR WEAPONS TNG GROUP, ATLANTIC<br />
U S ARMY MATERIAL COMMAND ATTN: CODE 222<br />
ATTN: DRCDE-D<br />
ATTN: DOCUMENT CONTROL<br />
U S ARMY NUCLEAR & CHEMICAL AGENCY<br />
NUCLEAR WEAPONS TNG GROUP. PACIFIC<br />
ATTN: MONA-NU ATTN: CODE 32<br />
4 CYS ATTN: MONA-NU DR D BASH<br />
OFFICE OF CHIEF OF NAVAL OPERATIONS<br />
U S ARMY TEST & EVALUATION COMMAND<br />
ATTN: NIS-22<br />
ATTN: STEC$-NE(70)<br />
ATTN: NOP 06D<br />
ATTN: NOP 403<br />
U S ARMY WAR COLLEGE ATTN: NOP 50<br />
ATTN: LIBRARY ATTN: NOP 60<br />
ATTN: STRATEGIC STUDIES INSTITUTE<br />
ATTN: NOP 60D<br />
ATTN: MOP 603<br />
ATTN: NOP 91<br />
U S MILITARY ACADEMY<br />
AITN: BEHAVORIAL SCI & LEADERSHIP ATTN: OP 654<br />
ATTN: COL J G CAMPBELL<br />
ATTN: PMSOPMA523<br />
ATTN: SCIENCE RESEARCH LAB<br />
US ARMY MATERIEL SYS ANLYSIS ACTVY<br />
ATTN: DRXSY-DS<br />
OPERATIONAL ATrIN: TEST COMMANDER & EVALUATION FORCE<br />
Dist-2
DNA-TR-92-37-V3 (DL CONTINUED)<br />
PLANS. POLICY & OPERATIONS<br />
ATTN: CODE-P<br />
ATTN: CODE-POC-30<br />
TACTICAL TRAINING GROUP, PACIFIC<br />
ATTN: COMMANDER<br />
DEPARTMENT OF THE AIR FORCE<br />
ACADEMY LIBRARY DFSELD<br />
ATTN: LIBRARY<br />
AFIS/INT<br />
ATTN: INT<br />
ATTN: NIO-T<br />
ATTN: N10 STRATEGIC SYS<br />
FEDERAL EMERGENCY MANAGEMENT AGENCY<br />
ATTN: CIVIL SECURITY DIVISION<br />
ATfN: G ORRELL NP-CP<br />
ATTN: R SANDS<br />
U S DEPARTMENT OF STATE<br />
ATTN: PM/STM<br />
U S NUCLEAR REGULATORY COMMISSION<br />
ATTN: DIR DIV OF SAFEGUARDS<br />
ATTN: S YANIV<br />
AIR UNIVERSITY<br />
ATTN: STRATEGIC STUDIES<br />
AIR UNIVERSITY LIBRARY<br />
ATTN: AUL-LSE<br />
ATTN: LIBRARY<br />
DEPARTMENT OF DEFENSE CONTRACTORS<br />
ARES CORP<br />
ATTN: A DEVERILL<br />
HORIZONS TECHNOLOGY, INC<br />
ATTN: F GREY<br />
ASSISTANT CHIEF OF STAFF<br />
2 CYS ATTN: AF/SAMI HORIZONS TECHNOLOGY, INC<br />
ATTN: J MARSHALL-MIES<br />
ASSISTANT CHIEF OF THE AIR FORCE<br />
ATrN: SAF/AOA<br />
DEPUTY CHIEF OF STAFF FOR PLANS & OPERS<br />
ATTN: AFXOOSS<br />
FOREIGN TECHNOLOGY DIVISION<br />
ATTN: CCN<br />
ATTN: SDA<br />
KAMAN SCIENCES CORP<br />
ATTN: DASIAC<br />
KAMAN SCIENCES CORPORATION<br />
ATTN: R STOHLER<br />
KAMAN SCIENCES CORPORATION<br />
ATrN: DASIAC<br />
TACTICAL AIR COMMANDIXPSC<br />
ATTN: ACC/XP-JSG<br />
LOCKHEED MISSILES & SPACE CO, INC<br />
ATTN: WE-YOUNG WOO<br />
USAF SCHOOL OF AEROSPACE MEDICINE<br />
LOGICON R & D ASSOCIATES<br />
ATrN: RADIATION SCIENCES DIV<br />
ATrN: DOCUMENT CONTROL<br />
ATrN: DOUGLAS C YOON<br />
DEPARTMENT OF ENERGY<br />
DEPARTMENT OF ENERGY<br />
LOGICON R & D ASSOCIATES<br />
ATTN: S WOODFORD<br />
ATrN: DR T JONES<br />
MICRO ANALYSIS AND DESIGN<br />
LAWRENCE LVERMORE NATIONAL LAB<br />
ArTN: R LAUGHERY<br />
ATrN: Z DIVISION LIBRARY<br />
MISSION RESEARCH CORP<br />
LOS ALAMOS NATIONAL LABORATORY<br />
ATTN: DR NEIL GOLDMAN<br />
ArTN: D STROTTMAN<br />
ATTN: REPORT LIBRARY<br />
PACIFIC-SIERRA RESEARCH CORP<br />
2 CYS ATrN: E L PAINTER<br />
MARTIN MARIETTA ENERGY SYSTEMS INC<br />
2 CYS ATrN: F W WHICKER<br />
ATTN: SANTORO 2 CYS ArmN: G ANNO<br />
ATTN: G KERR<br />
ATrN: H BRODE<br />
ATTN: J WHITE<br />
ATTN: W RHOADES<br />
PACIFIC-SIERRA RESEARCH CORP<br />
ATrN: D GORMLEY<br />
SANDIA NATIONAL LABORATORIES<br />
2 CYS ATTN: G MCCLELLAN<br />
ATTN: TECH LIB 3141<br />
SCIENCE APPLICATIONS INTL CORP<br />
OlHER GOVERNMENT<br />
ATrN: D KAUL<br />
ArmN: E SWICK<br />
CENTRAL INTELLIGENCE AGENCY<br />
ATTN: L HUNT<br />
ATTN: COUNTER-TERRORIST GROUP<br />
ATTN: R J BEYSTER<br />
ArmN: DIRECTOR OF SECURITY<br />
ATTN: W WOOLSON<br />
ATTN: MEDICAL SERVICES<br />
Dist-3
DNA-TR42-37-V3 (DL CONTINUED)<br />
SCIENCE APPLICATIONS INTL CORP<br />
ATTN: J MCGAHAN<br />
ATTN: J PETERS<br />
ATTN: W LAYSON<br />
SCIENCE APPLICATIONS INTL CORP<br />
ATN•: R CRAVER<br />
SCIENCE APPLICATIONS INTL CORP<br />
ATTN: JOHN A SHANNON<br />
TECHNICO SOUTHWEST INC<br />
ATTN: S LEVIN<br />
TRW OGDEN ENGINEERING OPERATIONS<br />
ATTN: D C RICH<br />
TRW SPACE & DEFENSE SECTOR<br />
ATTN: DR BRUCE WILSON<br />
UNIVERSITY OF CINCINNATI MEDICAL CENTER<br />
ATTN: E SILBERSTEIN<br />
Dist-4
Armed Forces Rad I Research Institute<br />
Retrospective Reconstruction of<br />
Radiation Doses of<br />
<strong>Chernobyl</strong> Liquidators by<br />
Electron Paramagnetic Resonance<br />
A ove,<br />
reieosU<br />
Scientific Center of Radiation Medicine<br />
Academy of Medical Sciences, Ukraine<br />
19980223 032
! 35b PL • .•,s -<br />
Retrospective Reconstruction of<br />
Radiation Doses of<br />
<strong>Chernobyl</strong> Liquidators by<br />
Electron Paramagnetic Resonance<br />
Authored by<br />
Scientific Center of Radiation Medicine<br />
Academy of Medical Sciences, Ukraine<br />
254050, Kiev-5O, Melnikova 53<br />
Vadim V. Chumak, Ilia A. Likhtarev,<br />
Sergey S. Sholom, Larisa F. Pasalskaya,<br />
and Yuri V. Pavienko<br />
Published by<br />
Armed Forces Radiobiology Research Institute<br />
Bethesda, Maryland, USA<br />
Editor and NIS Initiatives Coordinator<br />
Glen I. Reeves, M.D.
Cleared for public release: distribution unlimited.<br />
AFRRI Contract Report 97-2<br />
Printed December 1997<br />
<strong>Defense</strong> <strong>Nuclear</strong> Agency Contract DNA001 -95-C-001 7<br />
For information about this publication, write Armed Forces Radiobiology Research<br />
Institute, 8901 Wisconsin Avenue, Bethesda, MD 20889-5603, USA, or telephone<br />
011-301-295-0377, or send electronic mail to reeves@mx.afrri.usuhs.mil. Find<br />
more information about AFRRI on the Internet's World Wide Web at http://www.<br />
afrri.usuhs.mil.<br />
This and other AFRRI publications are available to qualified users from the <strong>Defense</strong> Technical<br />
Information Center, Attention: OCP, 8725 John J. Kingman Road, Suite 0944, Fort Belvoir,<br />
VA 22060-6218; telephone 703-767-8274. Others may contact the National Technical<br />
Information Service, 5285 Port Royal Road, Springfield, VA 22161; telephone 703-487-<br />
4650. AFRRI publications are also available from university libraries and other libraries<br />
associated with the U.S. Government's Depository Library System.
Preface<br />
On April 26, 1986, Reactor #4 at the <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant<br />
near Pripyat, Ukraine, exploded, releasing millions of curies of radioactive<br />
materials into the environment. The reaction was swift, with<br />
firefighters and medics being mobilized within hours of the accident.<br />
After initial care of the injured and instituting measures to prevent<br />
further exposure of the general population in the vicinity, attention<br />
was turned to cleanup of the damaged reactor and the radioactive<br />
debris. Hundreds of thousands of workers (called "liquidators") were<br />
employed in the cleanup. Authorities were aware of the risks of<br />
immediate and long-term health effects to these people and took<br />
measures to limit the dose received.<br />
With the passage of time, the liquidators have developed leukemia,<br />
solid tissue neoplasms, cardiovascular disease, and other illnesses.<br />
The question of what relationship these illnesses, which also occur in<br />
unexposed populations, bear to the radiation exposure received at<br />
<strong>Chernobyl</strong> naturally arises. This question is vitally important, not only<br />
for compensation purposes, but also for advancing our knowledge of<br />
the effects of protracted radiation exposure on human health and for<br />
setting or reevaluating safety standards. But the critical first step in<br />
finding the answer is accurately ascertaining what dose was actually<br />
received. Physical dosimeters were not always used, and were not<br />
always used reliably, during the several operations involved in the<br />
<strong>Chernobyl</strong> cleanup. It is necessary to employ accurate, reliable biological<br />
indicators of radiation effects to reconstruct exposure received.<br />
The implications of using electron paramagnetic resonance (EPR)<br />
analysis as one such state-of-the-art technique in performing dose<br />
reconstruction clearly go beyond <strong>Chernobyl</strong>. There are other areas of<br />
the world with widespread environmental contamination at dose<br />
levels sufficient to cause adverse health effects, such as along the Techa<br />
River in the Southern Urals region of Russia and in the areas surrounding<br />
the former nuclear weapons test site near Semipalatinsk,<br />
Kazakstan. There have also been accidents involving small numbers<br />
111
Retrospective Reconstruction of Radiation Doses by EPR<br />
of individuals where the actual dose received is, for one reason or<br />
another, not accurately known.<br />
Because of the importance of a means of accurately and precisely<br />
estimating cumulative radiation exposure for epidemiologic studies,<br />
the Armed Forces Radiobiology Research Institute (AFRRI) elected to<br />
fund this study. The authors are highly competent investigators who<br />
also have connections with similarly skilled independent scientists<br />
who could refine their techniques and improve their results. In addition,<br />
they have access to the data from the <strong>Chernobyl</strong> liquidators<br />
accessible to followup, most of whom are now in Ukraine. Although<br />
the scope of this study was limited, its results should provide a significant<br />
step in improvingthe utility of EPR in dose reconstruction as well<br />
as in getting a clearer picture of the magnitude of the radiation<br />
exposure actually received at the world's most tragic reactor accident.<br />
Glen I. Reeves, M.D.<br />
Editor and NIS Initiatives Coordinator<br />
AFRRI<br />
iv
Contents<br />
Preface .....................................................<br />
A b stract .................................................... 1<br />
Introduction ................................................ 3<br />
Task I. Development of a Routine High-Performance EPR-<br />
Dosim etric Technique ................................. 5<br />
EPR Measurements, Spectra Analysis, and Dose Reconstruction 5<br />
Sam ple Collection ....................................... 8<br />
Sam ple Preparation ...................................... 8<br />
Task 2. Quality Assurance Program .......................... 11<br />
Stage 1. Intercalibration With Homogenized Samples ....... 11<br />
Intercomparison Design ......................... 12<br />
Methods and Results ............................. 12<br />
Stage 2. Intercalibration With Whole Teeth Irradiated Under<br />
Laboratory Conditions ......................... 14<br />
Intercomparison Design ......................... 15<br />
Methods and Results ............................. 15<br />
Stage 3. Intercomparison of Teeth From Liquidators ........ 17<br />
Intercomparison Design ......................... 17<br />
M ethods and Results ............................. 18<br />
D iscussion ............................................. 19<br />
Task 3. Test of Practical Dose Determination .................. 21<br />
Dose Reconstruction .................................... 21<br />
System Developm ent ................................... 27<br />
D iscussion ................................................. 29<br />
Effects of M edical X Rays ................................ 29<br />
Effects of UV Light ...................................... 30<br />
Nonlinearity of Dose Response Curves .................... 31<br />
Sum m ary ................................................. 33<br />
References ................................................ 35<br />
Appendix-Identification Form for Tooth Sampling ............. 37<br />
iii<br />
V
Abstract<br />
Accurate, reliable dose reconstruction is a critical component in the<br />
epidemiological followup of liquidators. Dosimetry of teeth by electron<br />
paramagnetic resonance (EPR) is a state-of-the-art laboratory technique<br />
that is key to this effort. The Scientific Center of Radiation Medicine<br />
(SCRM) has developed and refined this technique in order to meet the<br />
practical demands of large-scale epidemiologic followup of the <strong>Chernobyl</strong><br />
liquidators. Independent analysis using similar technology was<br />
performed by investigators at the University of Utah and showed good<br />
correlation with the SCRM results. The lower limit of detection for<br />
reliable dose reconstruction was 100 mGy. Techniques were applied to<br />
samples from approximately 135 liquidators involved in cleanup activities<br />
within the first 2 years after the <strong>Chernobyl</strong> accident in 1986.<br />
Mean dose was 287 mGy, geometric mean was 205 mGy, and median<br />
dose value was 200 mGy. The reconstructed dose values range from 30<br />
to 2220 mGy. Correlation of results between the two institutions was<br />
generally within 17%. This report also addresses some confounding<br />
factors (previous medical x-ray exposures, ultraviolet light effects on<br />
anterior teeth, nonlinearity of dose response curves below 100 mGy)<br />
and how to deal with them.<br />
Keywords: dosimetry, retrospective, EPR, technique, doses, liquidators,<br />
<strong>Chernobyl</strong>
Introduction<br />
Electron paramagnetic resonance (EPR) dosimetry using teeth is generally<br />
accepted as a highly attractive method for reconstructing individual<br />
radiation doses long after exposure [1]. However, until recently,<br />
EPR dosimetry was generally considered a unique state-of-the-art laboratory<br />
procedure unsuitable for practical dose reconstruction. Moreover,<br />
the accuracy and consistency of results produced by this technique<br />
were not proven.<br />
A tool for dose reconstruction is acutely needed, particularly in the<br />
<strong>Chernobyl</strong> situation. Although these dose records are incomplete and<br />
not all liquidators know their doses, an official dose record is included<br />
in most of the identification forms of cleanup workers. The objective<br />
of the effort funded by DNA (contract number DNA001-95-C-0017)<br />
was to develop reliable retrospective estimates of radiation doses<br />
received by the <strong>Chernobyl</strong> liquidators. This goal was approached in<br />
three stages.<br />
First, the Scientific Center of Radiation Medicine (SCRM) developed<br />
the EPR dosimetry technique to provide reliable and efficient reconstruction<br />
of doses. Each of the basic steps of EPR-dosimetric methodology<br />
was subjected to rigorous analysis and optimization. Methodological<br />
research and development of the technique are far beyond the<br />
scope of this contract, but the technique has been explicitly presented<br />
elsewhere [2,3] and is widely accepted worldwide. Figure 1 illustrates<br />
that every step incorporated a number of innovations and specific<br />
features. This improved version of EPR dosimetric technique was<br />
employed for both routine dose reconstruction and interlaboratory<br />
cross-calibration.<br />
Second, a sophisticated cross-calibration effort was undertaken in<br />
order to assure the quality of results. This effort included a series of<br />
internal tests as well as intercomparisons with an experienced counterpart<br />
in the USA.<br />
3
Retrospective Reconstruction of Radiation Doses by EPR<br />
Third, after completion of the above tasks, routine reconstruction of<br />
doses to liquidators began. The doses to 135 individuals were assessed<br />
in accordance with the technique designed in Task One. These results<br />
were entered into a database and are available for researchers.<br />
Sampl<br />
Acquisition<br />
Collection of teeth along with extended ID data<br />
Separate processing of inner and outer parts of tooth to account for possible x-ray<br />
exposure<br />
P State-of-the-art purification of enamel in several progressive steps under EPR<br />
quality control<br />
ck Spetr rdincof p Subtraction mixture of nonirradiated of the instrument teeth background from young people signal<br />
-b Use of Mn:MgO standard for adjustment of magnetic field<br />
"- approim of e spectra after every o s pr a e m t<br />
stage of purification Deovlto Manipulation of background and measured signals by<br />
"F Use of Mn:MgO spectrometric of EPR S etra varying amplitude under fixed g-factor and width of signals<br />
standard<br />
Radiation-induced signal intensity determined as peak-to-<br />
". Recording of the signal of empty• peak amplitude<br />
resonator for use as instrument!<br />
background "- Monte Carlo procedure for<br />
x-a Medca dosear iskron determined: of<br />
of Cumulative<br />
, Empirical relationship deerindro for rough<br />
"• Use of additive dose method with 4-6 irradiations Dose assessment of uncertainty<br />
"• Additive dose values are at the expected level of the<br />
cumulative dose. %<br />
"• Laboratory irradiation of up to 20 samples in parallel at<br />
Evlution<br />
137Cs source with TILD control<br />
"• Least squares of thefEPPRtechniqu best fit<br />
"• Account for natural background and medical exposure components of cumulative dose<br />
"• Dose from the natural background is a function of the age of patient and type of tooth<br />
•Medical x-ray dose is determined:<br />
-precisely with differential method by analysis of inner and outer parts of tooth<br />
-approximately using empirical values of dose per x-ray examination<br />
of the<br />
"<strong>Accident</strong>al"<br />
Assessment of<br />
Uncertainty<br />
•.<br />
CUK.D<br />
Figure 1. Principal steps of EPR dosimetry and innovative features of the SCRM version<br />
of the EPR technique
Task 1<br />
Development of a Routine High-<br />
Performance EPR-Dosimetric Technique<br />
Extensive scientific and technological investigations were conducted<br />
in order to make EPR dosimetry usable as a routine tool for dose<br />
reconstruction. Special attention was paid to ensuring reliable results.<br />
With respect to the demands of epidemiological followup, optimal EPR<br />
technique must meet the following requirements:<br />
"* Sensitivity of the technique and accuracy of the results must be<br />
adequate to meet the practical needs of post-<strong>Chernobyl</strong> followup.<br />
"* The results produced by the technique must be consistent with<br />
other (independent) dosimetric methods and internal standards.<br />
"* The technique must be reproducible at different times and in other<br />
laboratories.<br />
"* Performance of the technique must be high enough to meet practical<br />
demands.<br />
EPR Measurements, Spectra Analysis, and Dose<br />
Reconstruction<br />
A brief profile of the EPR measurement procedure is presented in table<br />
1. The use of the Mn:MgO spectrometric standard allows for objective<br />
control of stability of the system and ease in accounting for possible<br />
drifts and deviations. The standard is used for calibration of the<br />
instrument in terms of both sensitivity and g factor. The empty resonator<br />
signal is recorded daily and subtracted as instrument background<br />
noise in each series of measurements.<br />
5
Retrospective Reconstruction of Radiation Doses by EPR<br />
Table 1. Brief profile of the EPR measurement procedure used at<br />
SCRM<br />
Feature<br />
Characteristics<br />
1. Instrument BRUKER ECS-106<br />
2. Laboratory irradiator<br />
13 7 Cs<br />
3. Buildup for secondary electrons +<br />
4. Number of additional irradiations 5<br />
5. Dose increment/cumulative dose AD = 174 for Dx < 500 mGy<br />
values for calibration curve<br />
AD = 348 for Dx -> 500 mGy<br />
6. Method of best fit of the Linear regression<br />
calibration curve<br />
7. Sample preparation +<br />
8. Parameters of EPR registration:<br />
Microwave frequency<br />
9.81 GHz<br />
Microwave power<br />
10 mW<br />
Center field<br />
348 mT<br />
Sweep width<br />
10.0 mT<br />
Modulation frequency<br />
100 kHz<br />
Modulation amplitude<br />
0.23 mT<br />
Conversion time<br />
164 ms<br />
Time constant<br />
328 ms<br />
Number of scans 15<br />
Measurement time<br />
60 and 45 minutes<br />
See item 5<br />
9. Processing of measured Subtraction of standard backspectra<br />
and detection of ground signal by selectively<br />
radiation-induced (RI) varying its intensity and fixing<br />
signals<br />
g factor and width<br />
10. Type of standard Spectrum of the mixture of<br />
background signal<br />
nonirradiated teeth from students<br />
18-25 years old<br />
11. Use of presumably nonirradiated<br />
samples as reference<br />
12. Approach to the error For doses less than 500 mGy,<br />
propagation<br />
the error is determined by uncontrolled<br />
impurities of enamel<br />
with maximum intensity<br />
equivalent to RI signal from 50<br />
to 80 mGy dose; for higher<br />
doses, error is determined by<br />
statistical error of RI signal<br />
determination.<br />
6
Development of a Routine High-Performance EPR-Dosimetric Technique<br />
The procedure for mathematical processing was based on investigation<br />
of the principal background signal variability. We discovered that<br />
background signals of different teeth differ only in terms of intensity,<br />
while g factors and width of lines for background signals are constant<br />
for all teeth. Therefore, we use a standard background signal, which<br />
was obtained from a mixture of several dozen nonirradiated teeth that<br />
were extracted from young people (18-25 years old), in order to minimize<br />
the natural background dose.<br />
The optimal procedure for subtraction of the principal background<br />
signal using the BRUKER ECS-106 or similar instrument is as follows:<br />
1. The standard background spectrum is shifted by the constant<br />
magnetic field relative to the sample spectrum, using an Mn:MgO<br />
marker until the g factors of both spectra coincide.<br />
2. The amplitude of the standard background spectrum is adjusted to<br />
coincide with the sample spectrum.<br />
3. The two spectra are subtracted while keeping the Mn:MgO signal<br />
constant.<br />
4. The resultant subtracted g factors of the maximum and minimum<br />
components of the suspected radiation-induced (RI) signal are compared<br />
to the positions of the relevant points determined for highdose<br />
signals.<br />
5. If points coincide, the intensity of the original radiation signal is<br />
measured and the value is used as the first experimental point in<br />
the individual calibration curve. If no coincidence occurs, no confident<br />
dose reconstruction is possible, and the tooth is considered<br />
to be exposed to a dose of less than 0.1 Gy.<br />
This rather conservative approach to spectrum interpretation ensures<br />
against the measurement of artifacts and misleading readouts.<br />
To account for individual radiosensitivity of teeth, an internal standard<br />
was used. Each sample was exposed to additional doses under<br />
controlled laboratory conditions. The 1 3 7 Cs secondary standard irradiator,<br />
calibrated in terms of absorbed dose in air using an 8 mm plastic<br />
screen for buildup of secondary electrons, was used for this purpose.<br />
The additional doses and the results of subsequent measurement<br />
produce the calibration curve of the individual tooth. The intersection<br />
of this curve with the abscissa corresponds to the amount of exposure.<br />
7
Retrospective Reconstruction of Radiation Doses by EPR<br />
Sample Collection<br />
The following samples were collected for use in both practical dose<br />
reconstruction and interlaboratory cross-calibration:<br />
* 69 teeth from liquidators (40 were multiple teeth from 15 individuals)<br />
* 19 teeth from unexposed young people in Ukraine<br />
* 6 teeth from unexposed people in the United States<br />
Teeth from cleanup workers were collected in the course of routine<br />
dental treatment in the Liquidators' Clinic in Kiev, which is dedicated<br />
to the medical and dental care of liquidators. All teeth were extracted<br />
for clinical reasons only; none was collected solely for dosimetric<br />
purposes. The extraction procedure was performed by a skilled dental<br />
surgeon who had been appropriately instructed. Every sample was<br />
accompanied by a Tooth ID Form (see appendix), which contains ID<br />
information on the patient, a history of occupational and medical<br />
exposure, activities and length of stay in the restricted zone, and the<br />
officially recorded dose value. After extraction, the teeth were washed<br />
with water and dried at room temperature. Each sample was transferred<br />
to our laboratory in a clearly labeled, individual container.<br />
Sample Preparation<br />
A new, elaborate method of sample preparation was one of the most<br />
efficient innovations introduced into the EPR-dosimetric technique.<br />
Application of this procedure leads to the substantial reduction of<br />
background EPR signals that normally are superimposed on the radiation-induced<br />
signal in tooth enamel, making detection of doses below<br />
0.5 Gy difficult.<br />
When samples arrive at the laboratory, they are registered in the log<br />
and computer database. Then sample preparation begins. Since the<br />
properties of the original material vary, the degree of purification<br />
needed to obtain an optimal specimen may vary. Accordingly, purification<br />
is normally performed in several progressive steps.<br />
1. Removal of the tooth root.<br />
2. Splitting of the tooth into its inner and outer parts with a diamond<br />
saw (this step is taken in order to take possible medical x-ray<br />
exposures into account).<br />
8<br />
3. Fragmentation of the tooth into 1- to 2-mm particles.
Development of a Routine High- Performance EPR-Dosimetric Technique<br />
4. Chemical treatment of the tooth with KOH alkaline in an ultrasonic<br />
bath for 4 to 7 days to remove dentine and organic components of<br />
tooth enamel.<br />
5. Removal of the remainder of the dentine (especially in the tooth<br />
parts with the highest curvature) with a hard-alloy dental drill.<br />
6. Crushing of samples to 0.1 to 0.25 mm-sized grains.<br />
7. Additional purification of tooth enamel using a heavy liquid, so-<br />
3<br />
dium polytungstate, with a specific weight of 2.92-2.94 g/cm<br />
8. Several washings of samples with distilled water; the last washing<br />
takes several hours under ultrasonic processing.<br />
9. If needed, visual control of samples using a binocular microscope<br />
to remove nonenamel inclusions.<br />
The failure of any given procedure leads to application of further<br />
treatment. This algorithm is graphically presented in figure 2.<br />
STEP 1.<br />
" Crush bulk tooth into grains of 1-2 mm size.<br />
" Treat with NaOH under ultrasound and temperature 60 0C over the 24-hour period.<br />
P, jRecord EPR spectrum.<br />
NO Is sample ýYES<br />
STEP 2.<br />
6 Crush to smaller grains (0.1-0.25<br />
mm).<br />
. Treat with NaOH (as in STEP 1) NO"-YES<br />
over 8- to 12-hour period. Was STEP 2<br />
- Wash in distilled water under done?<br />
ultrasound for 2 hours.<br />
pure enough?<br />
STEP 3.<br />
* Separate heavy liquid (sodium NO YES<br />
polytungstate; liquids of 2.65 g/cm 3 Wa•s STP 3<br />
and 2.85 g/cm 3 densities are used<br />
for removal of light and heavy<br />
inclusions, respectively.<br />
done?<br />
STEP 4.i NO Was STEP 4<br />
. Manually remove nonenamel paramagnetic particles under visual control. done?<br />
Figure 2. Flow chart of the sample-preparation process<br />
9
Retrospective Reconstruction of Radiation Doses by EPR<br />
150<br />
100<br />
.50<br />
S0-<br />
- -50<br />
A<br />
-100<br />
-150 ,..,,,,,,, ...... i .. .i .. •,a.... a.... i......<br />
3425 3450 3475 3500 3525<br />
Gauss (G)<br />
150-<br />
100-<br />
- 50-<br />
S0-<br />
B<br />
-100<br />
- 5 . . . , . . , . . ..,. .. . . , .....<br />
., ,... . . . . .<br />
3425 3450 3475 3500 3525<br />
Gauss (G)<br />
Figure 3. Effect of sample purification. EPR spectra (A)<br />
before and (B) after purification of the low-dose sample<br />
Figure 3 shows the EPR spectra of the same low-dose sample recorded<br />
before and after application of the purification procedure. Clearly, the<br />
spectrum was dramatically improved, making dose reconstruction<br />
with the sample possible.<br />
Another important issue is evaluating the contribution of medical x<br />
rays to the measured cumulative dose. Oversensitivity of tooth material<br />
to low-energy photons is well known, causing serious difficulties<br />
in determining <strong>Chernobyl</strong>-related doses. As much as a 7 to 1 difference<br />
in deposited dose in enamel versus that in soft tissue represents a<br />
threat to the utility of EPR as a dosimeter. The results of a study of this<br />
problem are presented elsewhere [4], giving a clue to a solution to this<br />
potentially severe problem.<br />
10
Task 2<br />
Quality Assurance Program<br />
Dr. Ed Haskell of the Division of Radiobiology, College of Medicine,<br />
University of Utah (UU), agreed to take part in the cross-calibration<br />
studies. These were performed in three stages, namely:<br />
Stage 1: Intercalibration using samples of tooth enamel with uniform<br />
properties that were exposed to known radiation doses under<br />
laboratory conditions<br />
Stage 2: Intercalibration using whole teeth exposed in vitro<br />
Stage 3: Intercomparison using liquidator's teeth accidentally exposed<br />
in vivo<br />
Each stage was designed to more closely approximate reality. Thus, the<br />
first stage dealt with rather ideal samples while the third intercomparison<br />
involved full-scale dose reconstruction using teeth specifically<br />
from exposed individuals. Interpretation of the results became increasingly<br />
difficult with each stage, as the number of uncertainty factors<br />
increased. However, the overall results appear to have engendered<br />
confidence in the adequacy of the dose assessments by EPR dosimetry<br />
with teeth.<br />
Stage 1. Intercalibration With Homogenized Samples<br />
For the first stage, the simplest intercalibration was performed with<br />
unrealistic but extremely uniform samples. This stage could be considered<br />
as a check for the reproducibility of the technique for practical<br />
dose reconstruction. The significant advantage of the intercomparison<br />
design was the possibility of objectively judging the results-no uncer-<br />
11
Retrospective Reconstruction of Radiation Doses by EPR<br />
tainty factors could lead to a fuzzy interpretation of the dose<br />
determinations.<br />
Intercomparison Design<br />
The main purpose of this intercalibration 1 was to test the results<br />
produced by different techniques using samples with well-known and<br />
uniform properties, thus allowing for objective evaluation of results.<br />
Since sample preparation techniques in different laboratories vary<br />
significantly, only the minimum necessary treatment was applied to<br />
tooth samples.<br />
A large number of nonirradiated human teeth with minimum background<br />
doses were treated mechanically in order to extract tooth<br />
enamel in the form of 0.1- to 0.25-mm grains. These were mixed<br />
together to form homogeneous material. This material was then divided<br />
into 100 mg portions and forwarded for irradiation to the IAEA<br />
Laboratory in Siebersdorf, Austria. The samples were irradiated with<br />
dose levels of about 100, 250, 500, and 1000 mGy. Those receiving doses<br />
up to the 500 mGy level were irradiated with a 13 7 Cs source at a dose<br />
rate of 800 Gy/min. Those at the 1000 mGy level were exposed to a 6 0 Co<br />
source at a dose rate of 200 mGy/min. Sets of samples with five<br />
different dose levels (unknown to the participants) were shipped to<br />
SCRM and UU. The participants were invited to use their own EPR<br />
dosimetric routines (including chemical treatment of samples, EPR<br />
measurement, and interpretation of spectra) to determine the exposures<br />
received by the samples.<br />
Methods and Results<br />
The intercomparison revealed significant variations in the experimental<br />
techniques used for dose reconstruction with EPR of tooth enamel.<br />
The SCRM technique is shown in table 1. The major differences in the<br />
UU technique were as follows.<br />
" The background signal was not subtracted. The intensity of the RI<br />
signal was determined as the difference of intensity at the points<br />
corresponding to the g factor of the first maximum and first minimum<br />
of the RI signal.<br />
"* The samples were measured 10 times each, shaking the tube after<br />
every measurement.<br />
12<br />
This cross-calibration was performed as part of the First International<br />
Intercalibration of EPR dosimetry with teeth, which was sponsored in part<br />
by CEC contract COSU-CT93-0051.
Quality Assurance Program<br />
Laboratory irradiation was performed for those samples that demonstrated<br />
the most significant variations. For these samples (8 of<br />
14), the value of the initial RI signal was determined as the intersection<br />
of the calibration curve with the ordinate axis. For the<br />
remainder, the initial intensity of the RI signal was assessed as the<br />
mean of 10 measurements. At this point, three samples with minimum<br />
RI signal values were considered as empirical zero. Doses of<br />
other samples were determined by subtracting the average intensity<br />
of three low-dose signals from the RI signal intensity. The<br />
results were divided by the value of the average calibration factor<br />
derived from the analyses of eight samples.<br />
* The residual signal, which corresponded to the average dose of<br />
three nonirradiated samples, was estimated to be 68 mGy.<br />
The results of the intercalibration are presented in figure 4 and table<br />
2. Clearly, the precision of dose determination depends greatly on the<br />
dose value and the technique used for reconstruction. The results<br />
obtained at SCRM demonstrated an excellent agreement with the<br />
preset dose values [5]. On the other hand, the UU technique showed<br />
variability over the whole range of doses, from 0 to 1000 mGy.<br />
These results were discussed in May 1995 duringthe 4th International<br />
Symposium on EPR Dosimetry and Applications. The constructive<br />
discussion of the peculiarities, possible advantages, and shortcomings<br />
of the techniques used at SCRM and UU led to the harmonization of<br />
certain approaches used in the two laboratories for subsequent stages<br />
of the cross-calibration. The UU approach was changed to conform<br />
more closely with some of the particular procedures under the SCRM<br />
approach.<br />
cc ~ 1200- A 1200 . B<br />
1000 1000-<br />
S800 800<br />
">, 600 600<br />
can) 00 400<br />
0 400-<br />
" " 200 200<br />
cc 0 0<br />
a)<br />
-2 0 0 0 ' -2 0 0 L ' 8 8 L a -<br />
0 200 400 600 800 1000 0 200 400 600 800 1000<br />
CHU..... CD,<br />
Nominal dose value, mGy<br />
Figure 4. Results of intercalibration with homogenized samples. Measured<br />
doses versus nominal dose values. (A) UU, (B) SCRM<br />
13
Retrospective Reconstruction of Radiation Doses by EPR<br />
Table 2. Results of intercalibration using homogenized samples<br />
irradiated in vitro<br />
Participant Sample Measured dose Nominal dose<br />
number value, mGy value, mGy<br />
University 87 20.7±17 0<br />
of Utah 88 49.8±28 0<br />
90 -93±9 0<br />
21 132±15 100<br />
4 209±39 100<br />
6 237±26 100<br />
35 137±58 250<br />
43 138±68 250<br />
58 312±31 250<br />
75 413±67 500<br />
66 448±34 500<br />
56 479±49 500<br />
106 940±60 1000<br />
100 1128±21 1000<br />
SCRM 80 0±50 0<br />
81 10±50 0<br />
8 80±50 100<br />
16 140±50 100<br />
23 150±50 100<br />
36 240±50 250<br />
37 260±50 250<br />
42 290±50 250<br />
60 440±50 500<br />
68 460±50 500<br />
69 520±50 500<br />
76 970±100 1000<br />
108 970±100 1000<br />
Stage 2. Intercalibration With Whole Teeth<br />
Irradiated Under Laboratory Conditions<br />
14<br />
The main purpose of stage 2 was to address the possible influence of<br />
sample preparation procedures on the results obtained using different<br />
techniques.
Quality Assurance Program<br />
Intercomparison Design<br />
The sample set consisted of six teeth, collected by a local dentist in the<br />
United States. The prior dose of x rays was unknown. Each tooth was<br />
split in two at the plane perpendicular to the jaw contour. One half of<br />
each tooth was irradiated using a 6 °Co source with a dose rate of about<br />
10 Gy/h. Measurements were considered to be dependent only upon<br />
the total dose, not the dose rate. Three dose levels between 100 and 500<br />
mGy were used; dose was unknown to the participants. One half of<br />
each tooth remained unirradiated in order to provide a postmeasurement<br />
check for detectable dose due to dental x rays as well as to provide<br />
a check should anomalies appear in the spectra or the results.<br />
The participants knew that the pairs of teeth numbered 1 and 4, 2 and<br />
5, and 3 and 6 were irradiated with equal doses in order to provide<br />
direct comparison of the results with the equivalent laboratory-added<br />
dose. 2 Teeth 4, 5, and 6 were shipped to the SCRM; the others remained<br />
at UU. Shipping was done by express mail specially labeled to avoid<br />
x-ray inspection and thus minimize transportation dose. Participants<br />
were instructed to apply their customary EPR dosimetric technique.<br />
Methods and Results<br />
The samples of tooth enamel were prepared as described in Task 1<br />
above. The tooth size was sufficient to perform separate analyses of the<br />
outer and inner parts in order to control, and if possible, account for<br />
unreported x-ray examinations.<br />
The EPR spectrum of tooth 6 differed significantly from the expected<br />
shape of the signal of irradiated enamel. Chemical processing with<br />
NaOH alkali (8 hours in an ultrasonic bath at 60 °C) was used to purify<br />
the enamel (see figure 2). Although the subsequent EPR analysis revealed<br />
a noticeable improvement, the shape of the signal was still<br />
distorted. The specimen was examined under the microscope, and one<br />
paramagnetic nonenamel particle was located and removed. That<br />
significantly improved the signal, making the sample appropriate for<br />
dose reconstruction.<br />
A special effort was made to control for possible x-ray exposure prior<br />
to intercalibration. Although the results of this examination do not<br />
allow for the reliable quantitative assessment of lifetime dose, there<br />
are strong indications of both a dose gradient (i.e., the surface of the<br />
tooth nearest the x-ray source receives a higher dose than the surface<br />
2 The total dose to be measured consisted of two principal components:<br />
unknown lifetime dose of the tooth donor and the known dose added at the<br />
laboratory.<br />
15
Retrospective Reconstruction of Radiation Doses by EPR<br />
opposite the source, which is typical for dental x-ray examinations) and<br />
non-zero readouts in nonirradiated parts of the teeth. The findings<br />
(table 3) are not statistically significant and are given only as guidance<br />
to demonstrate general tendencies. Unfortunately, x-ray doses for all<br />
the teeth were below the threshold of reliable dose reconstruction with<br />
the EPR technique, and therefore these figures could not be used to<br />
correct the results of the intercalibration.<br />
Table 3.<br />
(in mGy)<br />
Dose assessments for different parts of teeth (SCRM)<br />
Tooth Inner part External part Inner part External part Mean dose of<br />
number (unexposed) (unexposed) (exposed) (exposed) exposed half<br />
T4 30 50 230 240 230±50<br />
T5 30 50 230 270 250±50<br />
T6 20 50 290 300 300±70<br />
UU measurements were taken at a microwave power of 2 mW (a<br />
method chosen after a series of tests as the most promising one to<br />
minimize noise and, therefore, uncertainty of the dose determination).<br />
An analysis of expected uncertainties was performed before analysis,<br />
and the number of spectra to be collected at each power was set at 42<br />
for the irradiated samples, 12 at the additive dose level of 1 Gy, and 6<br />
at the 10-Gy additive dose level. Samples were stored for a minimum<br />
of 12 hours at room temperature following each irradiation in a 6 0 Co<br />
irradiator at a dose rate of 10 Gy/h (5 mm of aluminum was used for<br />
electron buildup). Dose increments were 210, 435, 435, 435, 870 mGy.<br />
Both sets of dose determinations are presented in table 4. As expected,<br />
the results of the second stage intercalibration were not as clear-cut as<br />
those of stage 1, and they could not be interpreted definitely. Both<br />
laboratories demonstrated good agreement (within 17%) with nominal<br />
Table 4.<br />
Intercomparison of whole teeth exposed in vitro<br />
Laboratory-<br />
Group Sample added dose Laboratory Measured dose<br />
mGy<br />
mGy<br />
1 T1 171 UU 190±50<br />
T4 SCRM 230±50<br />
2 T2 256 UU 180±50<br />
T5 SCRM 250±50<br />
3 T3 200 UU 190±100<br />
T6 SCRM 300±70<br />
16
Quality Assurance Program<br />
dose values, although the results produced by the SCRM technique<br />
tended to overestimate the doses. The latter may be explained by the<br />
contribution of the lifetime dose (particularly medical x-ray exposure)<br />
to the total dose to the tooth, which was determined by the dose<br />
reconstruction. Neither data on x-ray examination nor age of patients<br />
was available, making assessment of this component of the total dose<br />
impossible. Since the detected doses corresponding to the preintercalibration<br />
history of the teeth were found to be below the threshold<br />
of reliable dose reconstruction, the correction of the results was,<br />
unfortunately, impossible.<br />
Stage 3. Intercomparison of Teeth From Liquidators<br />
The third stage of cross-calibration was designed to test the capability<br />
of the two techniques to perform dose reconstruction using teeth<br />
exposed in vivo. Since the actual doses were unknown and, therefore,<br />
absolute validation of the results is impossible, the expected yield of<br />
this effort was a consistency check.<br />
Intercomparison Design<br />
The initial intention was to provide both laboratories with identical<br />
samples exposed in vivo. Three groups of samples from liquidators<br />
totaling 34 specimens were shipped to the United States:<br />
* 13 halves of large teeth (molars)-the remainders were retained at<br />
SCRM<br />
* 15 teeth from pairs extracted simultaneously from the same individual<br />
and therefore presumably having the same doses<br />
* 6 samples in the form of pieces of mechanically separated tooth<br />
enamel<br />
Due to the limited time that could be allocated by UU for examination<br />
of the samples, the number of dose reconstructions was reduced to<br />
five. The specimens selected (numbers X23, X24, X25, X26, and X28)<br />
were represented by granular samples of tooth enamel. For each<br />
sample, the initial separation of tooth enamel was performed at SCRM<br />
using a steel dental drill. After the removal of dentine, the pieces of<br />
enamel were collected and the whole sample was divided into parts of<br />
about 100 mg each for independent determination of dose by the<br />
participants. The characteristic size of enamel grains was about 500<br />
micrometers, although the dimensions of individual particles varied<br />
from hundreds of microns to several millimeters.<br />
17
Retrospective Reconstruction of Radiation Doses by EPR<br />
Methods and Results<br />
SCRM used basically the same standard technique described above.<br />
The technique used at UU was somewhat modified. The parameters<br />
used for the second EPR intercomparison of the liquidators' teeth were<br />
as follows: 8-mW microwave power, 5-Gauss modulation amplitude,<br />
20-second conversion time, 35-Gauss sweep width, 168-ms time constant,<br />
and 105 gain.<br />
The dose reconstruction was done using the spectra from the samples,<br />
a baseline enamel sample with negligible dose, and an empty EPR<br />
tube. The spectrum of the tube was subtracted from each spectrum<br />
taken of all the samples and the baseline samples. This was done in<br />
proportion to the number of scans taken in each enamel spectrum,<br />
that is, an enamel spectrum composed of 12 sweeps was corrected by<br />
subtracting the spectrum of the empty tube, which itself was normalized<br />
to 12 sweeps. The spectra of the samples were then precisely<br />
normalized to the standard of 10 sweeps and 100 mg per spectrum by<br />
the normalization factor:<br />
(10 sweeps/# of sweeps taken) * (100 mg sample weight)<br />
This adjustment was not necessary for the baseline sample as it was<br />
precisely weighed before measurement. The spectrum of the baseline<br />
sample was then subtracted from all the spectra of the irradiated and<br />
unirradiated enamel samples. From these resulting background and<br />
baseline-free spectra, we did the dose reconstruction on the g-perpendicular<br />
signal extremes using standard least squares fitting and error<br />
propagation for the dose estimates and errors, respectively. The additive<br />
dose technique was employed using only one applied dose of 5 or<br />
10 Gy (10 Gy if the sample mass was less than 35 mg). The number of<br />
spectra taken at each dose was 20 to 25 for the zero dose and 12 for the<br />
one applied dose (25 if the sample mass was less than 35 mg).<br />
18<br />
In this intercomparison, an additional check of the purity of the<br />
samples was performed using the EPR spectra recorded before additional<br />
irradiation and the spectrum of a milk tooth as the standard of<br />
the background signal. Two (X24 and X25) of five samples had satisfactory<br />
purity. Three others were subjected to treatment with an NaOH<br />
solution. After the chemical treatment, the shape of the signal had<br />
improved. Two of the samples (X26 and X28) were considered to be<br />
purified completely. Although the purity of the third one (X23) had<br />
improved, it still had some distortions in the spectrum. Because of<br />
significant loss of mass (it had decreased from 70 to 33 mg), we decided<br />
to refrain from further purification and proceed with EPR<br />
measurements.
Quality Assurance Program<br />
This third stage of cross-calibration had relatively poor results (table<br />
5). The doses, determined in different laboratories, coincided within<br />
declared uncertainty ranges for only two individuals out of five. The<br />
results from the two laboratories differed significantly (up to 60%) for<br />
some of the samples. The results of this intercomparison are discussed<br />
elsewhere [6].<br />
At the present stage of the intercomparison, it is impossible to determine<br />
the major reasons for these discrepancies. Adequate interpretation<br />
of the results requires additional investigation and, possibly,<br />
conducting the intercomparison with a partially modified design. One<br />
possibility could be to use teeth from individuals whose doses have<br />
been assessed by independent methods of retrospective dosimetry<br />
(such as a FISH test or analytical dose reconstruction).<br />
Table 5. Results of intercomparison with teeth of liquidators<br />
exposed in vivo<br />
Sample SCRM, Gy UU, Gy<br />
X23 0.36±0.05 0.67±0.10<br />
X24 1.42±0.14 1.60±0.24<br />
X25 1.08±0.11 1.56±0.23<br />
X26 1.50±0.15 1.56±0.23<br />
X28 0.48±0.05 1.18±0.18<br />
Discussion<br />
The cross-calibration performed within Task 2 of the project was the<br />
first international, full-scale effort to harmonize EPR-dosimetric techniques<br />
developed in Ukrainian and US laboratories as well as to<br />
perform quality assurance of this method. The three stages of cross-calibration,<br />
for the most part, covered all degrees of complexity and<br />
adequacy of approaches to dose reconstruction. Generally positive, the<br />
results of the cross-calibration have proven the applicability of EPR<br />
dosimetry to practical reconstruction of individual doses.<br />
The reproducibility of the results of the different versions of EPR<br />
technique that were designed and used on different continents is, at<br />
worst, within the 60% standard deviation interval. Clearly, even with<br />
this conservative and potentially improvable uncertainty interval, EPR<br />
dosimetry could produce more accurate dose assessments than any<br />
other method of retrospective dose reconstruction to be used in a<br />
post-<strong>Chernobyl</strong> epidemiological followup. Moreover, stage 2 of the<br />
cross-calibration experimentally demonstrated that lifetime diagnostic<br />
x-ray examinations may lead to overestimation of doses within only<br />
19
Retrospective Reconstruction of Radiation Doses by EPR<br />
30% limits. This important point, which needs additional investigation,<br />
could resolve positively the greatest concern presently associated with<br />
the use of EPR dosimetry for reconstruction of individual doses among<br />
the liquidators.<br />
20
Task 3<br />
Test of Practical Dose Determination<br />
A system for retrospective reconstruction of doses received by the<br />
<strong>Chernobyl</strong> liquidators was tested in Task 3.<br />
Dose Reconstruction<br />
The teeth from liquidators were collected in the course of dental<br />
surgical practice in the Kiev central liquidators' polyclinics. Extracted<br />
teeth were accompanied by special ID forms (see appendix) reflecting<br />
the personal data necessary for tracing the individual, information<br />
about occupational contacts with ionizing radiation, lifetime medical<br />
x-ray examination of the head, data on location of the extracted tooth,<br />
and the diagnosis leading to extraction. After extraction, the teeth were<br />
preserved in formalin in small bottles. Periodically (approximately<br />
once a month), the teeth were transported to the Laboratory of External<br />
Exposure Dosimetry for storage, processing, and determination of<br />
radiation doses.<br />
Upon arrival, all teeth were subjected to preprocessing, including<br />
rinsing in distilled water and drying at 80 'C. The tooth root was<br />
separated and, if necessary, residuals of soft tissues and damaged areas<br />
of teeth were removed. Then, teeth were placed in intermediate storage<br />
under room conditions.<br />
The dose determination cycle began with chemical treatment as discussed<br />
in Task 1 and illustrated in figure 2. The samples of pure tooth<br />
enamel were subjected to measurements, including recording of EPRspectra<br />
(with parameters as indicated in table 1) and laboratory irradiation<br />
with preset doses. Individual calibration curves were plotted for<br />
all measured samples, and doses were determined accounting for<br />
individual radiosensitivity of enamel. It was found that calibration<br />
21
Retrospective Reconstruction of Radiation Doses by EPR<br />
- A curves may not always<br />
200 -linear regression<br />
2polynome of the<br />
be fitted by linear regression.<br />
In some cases,<br />
100 f dose-response curves<br />
were superlinear (fig-<br />
0<br />
050 0<br />
50<br />
50 100 150 200 250<br />
ure5b)orsublinear(figure<br />
5c). Although the<br />
S200 B<br />
Dose, cGy<br />
total fraction of teeth<br />
with nonlinear dose-<br />
150 response curves was<br />
._ rather small (about 5%),<br />
S100<br />
this phenomenon needs<br />
to be accounted for in<br />
"order to avoid signifi-<br />
0 * cant under- or overesti-<br />
-40 0 40 80 120 160 mation of dose values.<br />
Dose, cGy<br />
The sources of this ef-<br />
300 - C fect need to be studied<br />
200r<br />
100<br />
and localized.<br />
The results of dose reconstruction<br />
of 146<br />
0 0 50 * ' 200 teeth from 135 liquida-<br />
-25 0 50 100 150 2'00 tors are shown in table<br />
Dose, cGy<br />
6. The cumulative doses<br />
and dose values are not<br />
Figure 5. Different types of dose-response corrected for lifetime<br />
(calibration) curves for tooth enamel sam- exposure. Age of the paples.<br />
(A) linear dose response (95% of sam- tient and type of tooth<br />
ples), (B) superlinear dose response (4% of give a clue to the<br />
samples), (C) sublinear dose response (1% amount of the dose due<br />
of samples).<br />
to natural background.<br />
The type of tooth is also<br />
important for the possible<br />
contribution of ultraviolet<br />
(UV) light to<br />
the generation of paramagnetic centers, which may be more pronounced<br />
for the front teeth (classes A and B).<br />
22<br />
The frequency distribution of individual doses measured with EPR<br />
dosimetry is presented in figure 6. It may be seen that the shape of<br />
the distribution is close to lognormal-mean dose is 287 mGy,<br />
geometric mean is 205 mGy. Median dose value is 200 mGy. The<br />
reconstructed dose values range from 30 to 2220 mGy. The individual<br />
with the highest dose is a policeman who performed his guard<br />
mission outdoors during the first days after the accident. On some<br />
occasions, several teeth used in the investigation came from the<br />
same individual.
Test of Practical Dose Determination<br />
Table 6. Individual dose values reconstructed in course of EPR<br />
dosimetric exercise<br />
Sample Age at Beginning of<br />
number/ time of tooth X-ray cleanup work Type of Total dose<br />
ID code extraction examinations (year) tooth* (cGy)<br />
1 2 3 4 5 6<br />
177/13862 60 - 86 A 39<br />
178/16042 53 - 86 C 16<br />
180/17568 25 - 86 C 5.5<br />
182/2069 56 + 86 C 13<br />
183/59 45 - 86 C 24<br />
185/10709 43 + 86 C 12<br />
187/15961 61 - 86 C 19<br />
190/17871 64 - 86 C 14<br />
C 29<br />
192/17099 60 - 86 C 25<br />
194/20457 45 - 86 C 10<br />
195/13337 58 - 86 A 34<br />
198/12006 55 + 86 B 15<br />
199/20491 44 + 86 C 11<br />
200/7107 46 - 86 C 10<br />
201/4385 43 - 86 C 17<br />
202/1304 56 + 86 C 28<br />
204/1801 47 + 86 C 8<br />
205/ 40 + 86 C 13<br />
208/9731 56 - 86 C 12<br />
301/15518 60 + 86 C 19<br />
302/16012 58 - 86 C 28<br />
303/13827 39 86 U 6<br />
304/13473 55 + 86 C 15<br />
305/7826 39 86 A 31<br />
307/4513 48 - 86 C 8<br />
278/14877 40 - 86 C 4<br />
28/10058 55 + 86 C 12<br />
281/14571 57 - 86 C 11<br />
284/17631 57 - 86 A 31<br />
A 16<br />
287/19245 57 - 86 C 18<br />
298/8822 66 - 86 B 34<br />
33/483 57 - 86 A 46<br />
8/4416 44 + 86 U 18<br />
A 21<br />
4/13930 53 + 86 C 45<br />
C 67<br />
39/15579 62 + 86 B 78<br />
16/4689 52 + 86 C 38<br />
9/7821 55 - 86 C 53<br />
*A - incisors and canines, B - premolars, C - molars, U - unknown<br />
23
Retrospective Reconstruction of Radiation Doses by EPR<br />
24<br />
Table 6. Continued<br />
1 2 3 4 5 6<br />
10/11338 53 + 86 A 40<br />
15/13919 59 - 86 C 22<br />
29/16047 53 + 86 C 19<br />
24/6902 55 - 86 A 64<br />
53/1926 54 + 86 C 18<br />
46/7125 46 - 87 C 13<br />
97/10075 56 - 86 A 55<br />
98/199 52 - 86 A 59<br />
43/17164 55 + 87 C 120<br />
67/16344 38 + 86 A 66<br />
138/17320 41 + 87 C 9<br />
151/3598 64 + 86 C 12<br />
132/3915 66 - 86 A 23<br />
64/3978 40 + 86 C 27<br />
84/2743 54 - 87 C 20<br />
181/9287 63 + 86 B 14<br />
6C 53 - 86 C 142<br />
19C 62 - 86 C 25<br />
105/11735 44 - 86 C 16<br />
106/9138 44 - 86 C 20<br />
108/573 62 + 86 C 30<br />
C 23<br />
7/1670 62 + 86 A 30<br />
71/763 64 - 86 C 18<br />
72/18249 42 - 86 C 3.5<br />
78/18414 30 + 86 C 5<br />
C 8<br />
79/15694 79 - 86 C 13<br />
82/7452 49 - 86 C 5<br />
83/18187 49 + 86 C 13<br />
86/17174 50 - 87 C 12<br />
87/15171 55 - 86 C 6<br />
35/3602 28 + 86 C 20<br />
41/4068 37 + 86 C 18<br />
42/3304 55 - 86 C 7<br />
44/8329 47 - 86 B 18<br />
61/8329 B 28<br />
45/8203 60 + 86 B 20<br />
49/9037 65 + 86 C 19<br />
56/9737 56 - 86 A 29<br />
59/7727 45 - 86 C 8<br />
50/7727 C 6<br />
65/4299 50 + 86 C 9<br />
68/10373 50 + 87 C 8<br />
"A- incisors and canines, B - premolars, C - molars
Test of Practical Dose Determination<br />
Table 6. Continued<br />
1 2 3 4 5 6<br />
69/17092 60 - 86 C 8<br />
109/17545 45 - 87 C 21<br />
113/8025 29 - 87 C 6<br />
123/13820 52 + 86 A 50<br />
125/10396 55 + 86 C 14<br />
129/13494 57 + 86 A 24<br />
130/15240 64 - 86 A 60<br />
142/5009 68 - 86 A 71<br />
145/4768 56 - 86 A 35<br />
148/16267 49 + 87 C 67<br />
153/13349 60 + 86 C 96<br />
209/1408 40 - 86 U 47<br />
212/8456 46 - 86 A 39<br />
216/8544 45 - 86 C 25<br />
217/13506 43 - C 13<br />
218/13953 59 86 U 20<br />
219/15067 64 + 86 U 13<br />
223/3467 57 - 86 C 23<br />
224/18355 71 - 87 C 65<br />
225/17113 56 - 87 C 14<br />
227/18318 55 - 86 C 10<br />
228/8369 53 + 86 C 14<br />
249/8369 54 C 39<br />
230/2721 57 - 86 C 12<br />
231/13381 63 - 86 C 195<br />
234/16478 63 - 86 C 16<br />
235/8152 56 + 86 C 23<br />
236/4325 47 - 86 A 27<br />
238/9123 62 - 86 A 70<br />
247/14939 43 - 86 C 18<br />
250/6863 55 + 86 C 21<br />
C 15<br />
251/7583 45 + 86 A 23<br />
263/8939 64 - 86 A 35<br />
300/17933 66 + 86 C 40<br />
81/10298 33 - 89 C 8<br />
308/19933 51 - 87 C 7<br />
309/8120 56 - 88 C 13<br />
312/2043 61 + 86 C 23<br />
313/16137 68 - 87 C 26<br />
314/4571 58 + 86 U 64<br />
317/21873 55 - 86 C 50<br />
320/7004 48 86 C 6<br />
321/20577 44 - 87 C 7<br />
*A- incisors and canines, B - premolars, C - molars, U - unknown<br />
25
Retrospective Reconstruction of Radiation Doses by EPR<br />
Table 6. Continued<br />
1 2 3 4 5 6<br />
322/15777 38 - 86 A 36<br />
B 31<br />
325/23887 42 + 86 C 8<br />
326/15707 64 - 86 A 42<br />
330/17012 61 + 86 C 12<br />
331/8719 33 + 86 C 10<br />
332/14533 59 - 86 C 35<br />
334/3713 57 + 86 C 12<br />
335/15336 51 - 86 B 3<br />
337/4722 68 - 86 A 21<br />
340/13695 52 - 86 C 32<br />
C 20<br />
2C 86 U 57<br />
17C 40 86 U 30<br />
18C 35 86 U 222<br />
20C 41 86 U 40<br />
21C 86 U 30<br />
22C 37 86 U 15<br />
*A - incisors and canines, B - premolars, C - molars, U - unknown<br />
The doses generally depend on the amount of time spent working in<br />
the 30-km zone. As may be seen from table 6, most of the liquidators<br />
involved in the current dose reconstruction effort began their work in<br />
1986. Median dose of this group is 211 mGy, while doses to liquidators<br />
of 1987 and later years are lower-164 mGy. Maximum doses to liquidators<br />
from 1986 and 1987 were 2220 mGy and 1200 mGy, respectively.<br />
This observation is in good agreement with the fact that the most<br />
dose-intensive activities were performed during the first months after<br />
50<br />
_T 40<br />
0-<br />
0<br />
a)<br />
0<br />
S20<br />
E<br />
z 10<br />
0'= 6= 1 W I I<br />
10 30 50 70 90 110 130 150 170 190 210 230<br />
Dose, cGy<br />
Figure 6. Frequency distribution of individual doses to liquidators<br />
determined by EPR dosimetry of teeth.<br />
26
Test of Practical Dose Determination<br />
the accident, when dose rate levels were much higher and most of the<br />
cleanup work took place.<br />
System Development<br />
Practical demand dictates a need for reconstruction of radiation doses<br />
to the large groups of liquidators included in cohorts studied for<br />
epidemiological followup. The possibility of long-term storage of tooth<br />
samples together with increasing performance of the EPR-dosimetric<br />
technique make this task feasible.<br />
However, since the teeth used for dose reconstruction are extracted for<br />
medical reasons only, sampling is a random and relatively infrequent<br />
event. Besides, the process of natural tooth loss is an important factor,<br />
reducing the available sampling population over time.<br />
Therefore, a systematic approach to dose reconstruction from teeth,<br />
including sample acquisition, is required. For longitudinal epidemiological<br />
followup of an exposed population, the problem of availability<br />
of samples may be solved by organizing a widespread network for<br />
acquisition of teeth extracted from the members of the studied cohort.<br />
This network should be based on centers with a high density of<br />
liquidators and other heavily exposed populations. Since the productivity<br />
of EPR dosimetry is limited and not yet sufficient to process all<br />
the potential influx of samples in real time, a central bank of bioprobes<br />
should be established for acquisition, storage, processing, and retrieval<br />
of tooth material. Potentially, every participant of this studied cohort<br />
sooner or later would be covered by this effort, yielding tooth samples<br />
to the bioprobe bank. Teeth from those individuals who were included<br />
in the study cohort and have died could be received in the course of<br />
autopsy. Dose values, reconstructed by means of EPR, could be entered<br />
in the personal dosimetric file of the individual for access by<br />
radioepidemiologists.<br />
Development of such an infrastructure for dose reconstruction is<br />
underway now in Ukraine. The acquisition network (figure 7) would<br />
be based on special liquidators' hospitals in seven regional centers,<br />
covering about 45% of the heavily exposed cleanup workers. The samples,<br />
along with ID forms, would be transferred to the central bioprobe<br />
bank for long-term storage and subsequent processing.<br />
The role of the central bioprobe bank is to coordinate activities in<br />
acquisition of teeth, EPR dosimetry, and data management on a national<br />
scale. Results of the ongoing EPR dose reconstruction will be<br />
forwarded to the National Registry and sent in parallel to the local<br />
health care bodies in order to provide feedback to individuals whose<br />
teeth had been submitted for examination. Access to the individual<br />
27
Retrospective Reconstruction of Radiation Doses by EPR<br />
dose records will also be provided to the researchers involved in<br />
post-<strong>Chernobyl</strong> followup studies.<br />
<strong>Chernobyl</strong> Ministry<br />
of Ukraine<br />
lDnipropetrovsk i Donetsk Kharkiv I Kir°vgr<br />
P •, va Health, •Ministryl F Zo rtzhiya<br />
, IoUraino of Uov L• z 1]<br />
Kiev<br />
Central bioprobe bank<br />
Consumers of<br />
Bioprobe section Database section dosimetric information<br />
"* conservation * registration of incoming probes * epidemiology<br />
"• storage * tracing of the samples 1..- • medicine<br />
"* sample preparation * storage of reconstructed doses * social service<br />
"* access for analysis e retrieval and output of doses o exposed individuals<br />
t<br />
t<br />
methodological EPR-dosimetry lab 1 1EPR-dosimetry lab 2 EPR-dosimetry lab3<br />
center H<br />
Figure 7. Infrastructure of the system for retrospective reconstruction<br />
of doses received by the <strong>Chernobyl</strong> liquidators<br />
28
Discussion<br />
Although performance and capability were tested in a series of crosscalibrations<br />
and routine dose reconstructions, several issues of key<br />
importance need to be resolved prior to extensive use of EPR dosimetry<br />
with teeth for followup studies. These yet unresolved problems may<br />
threaten the utility of EPR dosimetry. Some of these effects have been<br />
known for a long time; others were discovered recently. Among these<br />
are the well-known effect of enhanced sensitivity to low-energy photons<br />
and the recently reported generation of paramagnetic centers by<br />
UV light [7]. Nonlinearity of dose-response curves in the dose range<br />
below 1 Gy was observed by the authors of this report only during the<br />
dose reconstruction exercise and is yet unpublished.<br />
Effects of Medical X Rays<br />
Irradiation of tooth enamel with low-energy photons may lead to<br />
substantial (up to seven times) overestimation of the tissue-absorbed<br />
dose. This effect has pronounced energy dependence, with the highest<br />
oversensitivity at 60 keV. The signals from paramagnetic centers produced<br />
by high-energy (accidental) and low-energy (medical x-ray) photons<br />
are identical, making discrimination of these signals by means of<br />
EPR impossible. As a result, a dose measured by EPR is the sum of an<br />
accidental component (dose of interest) and a component due to<br />
medical exposure. The degree of significance of the latter depends on<br />
the relative value, which is a function of incidence energy, dose per<br />
examination, and number of examinations. This means that the type<br />
of x-ray apparatus used in the dental practice is very important, determining,<br />
after all, the degree of significance of the x-ray component.<br />
In order to clarify this issue, it is necessary to conduct a systematic<br />
investigation of the effects connected with x-ray exposure. This inves-<br />
29
Retrospective Reconstruction of Radiation Doses by EPR<br />
tigation should include both experimental and theoretical evaluation<br />
of dose responses prompted by different types of x-ray examination,<br />
including different geometry, x-ray apparatus, and dose per examination.<br />
The issue of doses deposited on neighboring and opposite teeth<br />
should be studied also. The contribution to the total tooth dose from<br />
different types of x-ray examination (e.g., gamma-tomography, panoramic<br />
diagnostics of the mandible, skull and sinus films) and radiotherapy<br />
is still unknown and requires special investigation. This work<br />
will demand the use of both mathematical and physical phantoms for<br />
the simulation of realistic situations.<br />
Since x-ray practices are different in Ukraine and the United States,<br />
special attention in this research should be paid to comparison of these<br />
two cases and the development of approaches to the solution of this<br />
problem. Intercomparisons using teeth exposed to x rays or exposed<br />
to mixed fields would be useful as well. Such studies should either<br />
conclude that the x-ray contribution to the tooth dose is insignificant<br />
(and establish the limits of the application of this assumption) or else<br />
recommend how to mitigate or account for the effect of x-ray<br />
examination.<br />
Analysis of different teeth from the same person may be useful for<br />
understanding of the effects of x-ray examination on different types of<br />
teeth and other factors. Among the 135 individuals studied in the<br />
present research, multiple teeth were available from 11 persons. In<br />
nine cases, teeth were extracted simultaneously, with presumably<br />
equal accidental doses. In two cases (individuals 190/17871 and<br />
284/17631), the doses determined using different teeth showed a<br />
discrepancy above the 40% standard deviation accuracy granted by the<br />
technique. Since the teeth in both cases were of the same type, and<br />
x-ray examination was not reported on the ID form, this phenomenon<br />
could not be explained by the contribution of x-ray exposure or variations<br />
in the type of tooth. Some as yet unknown effects may be<br />
responsible for such deviation. Unfortunately, the limited scope of<br />
research and similarities in lifetime exposure and type of teeth give<br />
little material for analysis. However, so far, the largest deviation of<br />
doses determined for similar teeth was 52%, which is not a particularly<br />
large error, considering errors typical for other methods of retrospective<br />
dosimetry.<br />
Effects of UV Light<br />
30<br />
Another phenomenon which may affect the reliability of dose reconstruction<br />
with tooth enamel is the generation of paramagnetic centers<br />
by UV light. Information about the role and the qualitative and quantitative<br />
characteristics of this effect is quite contradictory. This effect<br />
was first discovered and reported by Ivannikov et al. [7] in 1995. The
Discussion<br />
series of experiments conducted worldwide to study this effect brought<br />
no clarification.<br />
According to existing information and our own data, the centers<br />
generated in tooth enamel have position and shape very similar to<br />
those of radiation-induced centers. That means that discrimination of<br />
UV- and radiation-induced signals by spectrometric means may be<br />
difficult. It is expected that UV irradiation effects are most pronounced<br />
for front teeth; such factors as time spent outdoors and elevation of the<br />
living area above sea level may also influence the degree of this effect.<br />
At present, the problem of UV irradiation needs to be approached in a<br />
systematic way; this phenomenon must be studied from the point of<br />
view of its physical, spectrometric, and kinetic (half-life) properties.<br />
Processes of generation of paramagnetic centers as a function of wavelength<br />
and intensity of UV light and decay of these centers should be<br />
investigated in order to obtain a clear view of this effect.<br />
Study of spectrometric properties (e.g., saturation of the signals) may<br />
yield an approach to discrimination of UV- and radiation-induced<br />
signals by means of EPR technique. Investigation of depth profiles of<br />
UV-generated signals in teeth for different energies of UV photons and<br />
the UV component of daylight should help explain attenuation of UV<br />
light in enamel and could be used for target etching of exposed<br />
fractions of tooth enamel. Recommendations concerning accounting<br />
for and mitigating this effect should be issued as a final point of this<br />
research.<br />
Nonlinearity of Dose Response Curves<br />
Nonlinearity of dose response curves in the dose range below 1 Gy was<br />
observed in some teeth in the course of the dose reconstruction exercise<br />
in this project. Before, saturation of the dose-response curve was<br />
observed only at doses above 10 Gy; below this range, the curve was<br />
considered to be linear, and this property is widely used for extrapolation<br />
of calibration curves in the low-dose regions. Moreover, the techniques<br />
based on the utility of a single calibration factor (without<br />
additive dose) critically depend on linearity of the dose-response<br />
function.<br />
Nonlinearity of dose response curves may have a significant influence<br />
on the results of dose reconstruction. Not accounting for nonlinearity<br />
of calibration curves leads to substantial under- or overestimation of<br />
individual doses (as illustrated in figure 5). Advanced study of this<br />
effect, investigation of factors having impact on the dose-response<br />
curve, and the development of methods for extrapolating additive-dose<br />
curves are necessary for accurate and reliable retrospective dosimetry<br />
31
Retrospective Reconstruction of Radiation Doses by EPR<br />
using teeth as a natural dosimeter. Since this effect takes place in only<br />
about 5% of cases, the scope of dose reconstruction should be large<br />
enough to provide consistent and statistically significant conclusions.<br />
32
Summary<br />
During the 14-month period covered by this contract, extensive research<br />
and technological developments were performed at SCRM AMS<br />
Ukraine in close collaboration with the University of Utah, USA. As a<br />
result of this effort, EPR dosimetry with teeth was brought to the level<br />
of a semiroutine technique for evaluation of doses received by individuals<br />
heavily exposed after the <strong>Chernobyl</strong> accident.<br />
Special attention was paid to quality assurance for this high-technology<br />
method in order to provide accurate and reliable individual dose<br />
assessments. The quality assurance program included several international<br />
cross-calibrations using a variety of specimens, from pulverized<br />
tooth enamel in the beginning to whole teeth from liquidators exposed<br />
in vivo during the final phase of intercomparison.<br />
Since the limited availability of samples from the individuals of interest<br />
is one of the important bottlenecks of EPR dosimetry now, a<br />
complete system for the reconstruction of doses to liquidators must<br />
include a means for acquiring the samples. Therefore, an organization<br />
pattern for acquisition of teeth extracted by medical prescription from<br />
the <strong>Chernobyl</strong> liquidators was presented. This infrastructure is being<br />
implemented in Ukraine now.<br />
The semiroutine technique developed and adopted over the period of<br />
consideration was used for retrospective dosimetry of a considerable<br />
group of liquidators. In total, doses were reconstructed for 135 individuals<br />
who took part in the <strong>Chernobyl</strong> cleanup in 1986-87. The cohort of<br />
liquidators studied was assembled randomly in the course of dental<br />
surgery in the Kiev central liquidators' polyclinic.<br />
Analysis of the data obtained revealed that the mean dose of this group<br />
is 287 mGy, ranging to the highest value of 2220 mGy. This is signifi-<br />
33
Retrospective Reconstruction of Radiation Doses by EPR<br />
cantly higher than the officially reported mean dose of 110 mGy.<br />
Therefore, the widely accepted opinion that the official records are of<br />
low quality and underestimate the actual doses was supported by this<br />
first retrospective dosimetry effort involving an appreciable number<br />
of subjects. The fact that doses reconstructed instrumentally are much<br />
higher than those officially recorded gives additional justification for<br />
the investment in development and performance of retrospective<br />
dosimetry, particularly EPR.<br />
However, recent research and developments in the field of EPR dosimetry<br />
make obvious a need for further investigations. From the pragmatic<br />
point of view, these investigations should be conducted along<br />
the following lines:<br />
"* Investigation and development of approaches to account for EPR<br />
signals induced by lifetime medical x-ray exposure<br />
"* Comprehensive study of the effects in tooth enamel caused by UV<br />
light<br />
"* Investigation of the factors causing nonlinearity of the dose-response<br />
function in the dose range below 1 Gy and development of<br />
approaches to account for this effect in dose determination<br />
" Cross-validation of EPR dosimetry with independent methods of<br />
retrospective dosimetry; this may be achieved by parallel application<br />
of different methods (e.g., EPR, FISH, and analytical) to the<br />
same objects<br />
"* Methodological research aimed at improving the technological<br />
capabilities of EPR dosimetry and enhancing the productivity of<br />
the technique.<br />
Completion and success of the outlined efforts will bring EPR dosimetry<br />
from a quite exotic methodology to an ordinary dosimetric routine<br />
like gamma-spectroscopy and alpha counting.<br />
34
References<br />
1. Romanyukha AA, Ignatiev EA, Degteva MO, Kozheurov VP, Wieser<br />
A, Jacob P (1996) Radiation doses from Ural Region. Scientific<br />
Correspondence. Nature 381:199-200<br />
2. Chumak V, Sholom S, Likhtarev 1 (1995) Semi-routine ESR-dosimetry<br />
technique currently used in Ukraine. Presented at the 4th<br />
International Symposium on ESR Dosimetry and Application, Munich,<br />
Germany, May 15-19, 1995<br />
3. Chumak V, Sholom S, Pasalskaya L, Pavlenko Yu (1995) Ukrainian<br />
version of the EPR-dosimetric technique: An approach to the routine<br />
dose reconstruction. Second Workshop on Dose Reconstruction,<br />
Bad-Honnef, Germany, November 20-22, 1995<br />
4. Sholom S, ChumakV, Pavlenko Yu (1995) An account of diagnostic<br />
x-ray exposure in the problem of retrospective ESR dosimetry.<br />
Presented at the 4th International Symposium on ESR Dosimetry<br />
and Application, Munich, Germany, May 15-19, 1995.<br />
5. Chumak V, Baran N, Bugai A, Dubovsky S, Fedosov I, Finin V,<br />
Haskell E, Hayes R, Ivannikov A, Kenner G, Kirilov V, Khamidova<br />
L, Kolesnik S, Liidja G, Lippmmaa E, Maksimenko V, Meijer E,<br />
Pasalskaya L, Past J, Puskar J, Sholom S, Skvortzov V, Vaher U,<br />
Wieser A (1995) The first international intercomparison of EPR-dosimetry<br />
with teeth: First results. Presented at the 4th International<br />
Symposium on ESR Dosimetry and Application, Munich, Germany,<br />
May 15-19, 1995<br />
6. Haskell EH, Kenner GH, Hayes RB, Sholom S, ChumakV (1995) An<br />
EPR intercomparison using teeth irradiated prior to crushing. Sec-<br />
35
Retrospective Reconstruction of Radiation Doses by EPR<br />
ond Workshop on Dose Reconstruction, Bad-Honnef, Germany,<br />
November 20-22, 1995<br />
7. Ivannikov A, Skvortzov V, Khamidova L, Eichhoff U (1995) Development<br />
of tooth enamel EPR spectroscopy method for individual<br />
dosimetry. Presented at the 4th International Symposium on ESR<br />
Dosimetry and Application, Munich, Germany, May 15-19, 1995<br />
36
Appendix<br />
Identification Form for Tooth Sampling<br />
1. Complete affiliation of the hospital which<br />
performed extraction<br />
2. ID number 3. Date of extraction ___<br />
N General information Fragment 1<br />
1 Family name<br />
2 First name<br />
3 Second name<br />
4 Sex( male - 1, female - 2)<br />
5 Date of birth<br />
6 Liquidators pass (series and number)<br />
7 Year of work in <strong>Chernobyl</strong><br />
8 Dose value, officially recorded (if available)<br />
9 Date of evacuation from the 30-km zone<br />
10 From what settlement<br />
N Postal address at present time Fragment 2<br />
1 ZIP code<br />
2 Region<br />
3 District<br />
4 Town<br />
5 Street<br />
6 House<br />
7 Building<br />
8 Appartment<br />
37
Retrospective Reconstruction of Radiation Doses by EPR<br />
4. Places of stay since the accident (region, district, settlement)<br />
(1986 in all details, afterwards - reflect locations with period of stay more than 3 months).<br />
Year Settlement Period of stay<br />
Arrival Departure<br />
5. Professional contact with radiation (including military service)<br />
6. Information about x-ray examinations of skull, jaws, teeth (dates, type, approximate number during<br />
life span):<br />
7. General deseases affecting solid tissues of tooth<br />
8. Location of the tooth and reason of extraction:<br />
8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8<br />
38
Appendix<br />
9. Affiliation during the <strong>Chernobyl</strong> recovery activities<br />
10. Notes<br />
11. Name of physician who extracted the tooth<br />
39
DISTRIBUTION LIST<br />
DEPARTMENT OF DEFENSE<br />
U.S. ARMY RESEARCH INSTITUTE OF ENVIRONMENTAL<br />
MEDICINE<br />
ARMED FORCES RADIOBIOLOGY RESEARCH INSTITUTE ATTN: DIRECTOR OF RESEARCH<br />
ATTN: PUBLICATIONS BRANCH<br />
ATTN: LIBRARY U.S. ARMY RESEARCH LABORATORY<br />
ATTN: DIRECTOR<br />
ARMY/AIR FORCE JOINT MEDICAL LIBRARY<br />
ATTN: DASG-AAFJML WALTER REED ARMY INSTITUTE OF RESEARCH<br />
ATTN: DIVISION OF EXPERIMENTAL THERAPEUTICS<br />
ASSISTANT TO THE SECRETARY OF DEFENSE<br />
ATTN: AE<br />
ATTN: HA(IA) DEPARTMENT OF THE NAVY<br />
DEFENSE SPECIAL WEAPONS AGENCY<br />
BUREAU OF MEDICINE & SURGERY<br />
ATTN: TITL ATTN: CHIEF<br />
ATTN: DDIR<br />
ATTN: RAEM NAVAL AEROSPACE MEDICAL RESEARCH LABORATORY<br />
ATTN: MID ATTN: COMMANDING OFFICER<br />
DEFENSE TECHNICAL INFORMATION CENTER<br />
NAVAL MEDICAL RESEARCH AND DEVELOPMENT COMMAND<br />
ATTN: ACQUISITION ATTN: CODE 42<br />
ATTN: ADMINISTRATOR<br />
NAVAL MEDICAL RESEARCH INSTITUTE<br />
FIELD COMMAND DEFENSE SPECIAL WEAPONS AGENCY ATTN: LIBRARY<br />
ATTN: DASIAC<br />
ATTN: FCIEO NAVAL RESEARCH LABORATORY<br />
ATTN: LIBRARY<br />
INTERSERVICE NUCLEAR WEAPONS SCHOOL<br />
ATTN: DIRECTOR OFFICE OF NAVAL RESEARCH<br />
ATTN: BIOLOGICAL & BIOMEDICAL S&T<br />
LAWRENCE LIVERMORE NATIONAL LABORATORY<br />
ATTN: LIBRARY<br />
DEPARTMENT OF THE AIR FORCE<br />
UNDER SECRETARY OF DEFENSE (ACQUISITION)<br />
ATTN: OUSD(A)/R&E BROOKS AIR FORCE BASE<br />
ATTN: AL/OEBZ<br />
UNIFORMED SERVICES UNIVERSITY OF THE HEALTH SCIENCES ATTN: OEHL/RZ<br />
ATTN: LIBRARY ATTN: USAFSAM/RZB<br />
OFFICE OF AEROSPACE STUDIES<br />
DEPARTMENT OF THE ARMY ATTN: OAS/XRS<br />
HARRY DIAMOND LABORATORIES<br />
OFFICE OF THE SURGEON GENERAL<br />
ATTN: SLCSM-SE ATTN: HQ AFMOA/SGPT<br />
A-TN: HQ USAF/SGES<br />
OFFICE OF THE SURGEON GENERAL<br />
ATTN: MEDDH-N U.S. AIR FORCE ACADEMY<br />
ATTN: HQ USAFA/DFBL<br />
U.S. ARMY AEROMEDICAL RESEARCH LABORATORY<br />
ATTN: SCIENCE SUPPORT CENTER U.S. AIR FORCE OFFICE OF SCIENTIFIC RESEARCH<br />
ATTN: DIRECTOR OF CHEMISTRY & LIFE SCIENCES<br />
U.S. ARMY CHEMICAL RESEARCH, DEVELOPMENT, &<br />
ENGINEERING CENTER<br />
ATTN: SMCCR-RST OTHER FEDERAL GOVERNMENT<br />
U.S. ARMY INSTITUTE OF SURGICAL RESEARCH<br />
ARGONNE NATIONAL LABORATORY<br />
ATTN: COMMANDER ATTN: ACQUISITIONS<br />
U.S. ARMY MEDICAL DEPARTMENT CENTER AND SCHOOL<br />
BROOKHAVEN NATIONAL LABORATORY<br />
ATTN: MCCS-FCM ATTN: RESEARCH LIBRARY, REPORTS SECTION<br />
U.S. ARMY MEDICAL RESEARCH AND MATERIEL COMMAND<br />
CENTER FOR DEVICES AND RADIOLOGICAL HEALTH<br />
ATTN: COMMANDER ATTN: DIRECTOR<br />
U.S. ARMY MEDICAL RESEARCH INSTITUTE OF CHEMICAL<br />
GOVERNMENT PRINTING OFFICE<br />
DEFENSE ATTN: DEPOSITORY ADMINISTRATION BRANCH<br />
ATTN: MCMR-UV-R ATTN: CONSIGNED BRANCH<br />
U.S. ARMY NUCLEAR AND CHEMICAL AGENCY<br />
LIBRARY OF CONGRESS<br />
ATTN: MONA-NU ATTN: UNIT X
LOS ALAMOS NATIONAL LABORATORY<br />
BRITISH LIBRARY<br />
ATTN: REPORT LIBRARY ATTN: ACQUISITIONS UNIT<br />
NATIONAL AERONAUTICS AND SPACE ADMINISTRATION<br />
CENTRE DE RECHERCHES DU SERVICE DE SANTE DES ARMEES<br />
ATTN: RADLAB ATTN: DIRECTOR<br />
NATIONAL AERONAUTICS AND SPACE ADMINISTRATION<br />
FEDERAL ARMED FORCES DEFENSE SCIENCE AGENCY FOR NBC<br />
GODDARD SPACE FLIGHT CENTER<br />
PROTECTION<br />
ATTN: LIBRARY ATTN: LIBRARY<br />
NATIONAL CANCER INSTITUTE<br />
FOA NBC DEFENCE<br />
ATTN: RADIATION RESEARCH PROGRAM ATTN: LIBRARY<br />
NATIONAL DEFENSE UNIVERSITY<br />
INHALATION TOXICOLOGY RESEARCH INSTITUTE<br />
ATTN: LIBRARY TECHNICAL SERVICES ATTN: LIBRARY<br />
U.S. DEPARTMENT OF ENERGY<br />
INSTITUTE OF NUCLEAR MEDICINE AND ALLIED SCIENCES<br />
ATTN: LIBRARY ATTN: DIRECTOR<br />
U.S. EMBASSY, MOSCOW<br />
INSTITUTE OF RADIOBIOLOGY, ARMED FORCES<br />
ATTN: DEPARTMENT OF ENVIRONMENT, HEALTH, MEDICAL ACADEMY<br />
OCEANS, AND FISHERIES ATTN: DIRECTOR<br />
U.S. FOOD AND DRUG ADMINISTRATION<br />
OAK RIDGE ASSOCIATED UNIVERSITIES<br />
ATTN: WINCHESTER ENGINEERING AND ATTN: MEDICAL LIBRARY<br />
ANALYTICAL CENTER<br />
RESEARCH CENTER OF SPACECRAFT RADIATION SAFETY<br />
U.S. NUCLEAR REGULATORY COMMISSION ATTN: DIRECTOR<br />
ATTN: LIBRARY<br />
RUTGERS UNIVERSITY<br />
ATTN: LIBRARY OF SCIENCE AND MEDICINE<br />
RESEARCH AND OTHER ORGANIZATIONS<br />
UNIVERSITY OF CALIFORNIA<br />
ACADEMY OF MEDICAL SCIENCES, SCIENTIFIC CENTER OF ATTN: DIRECTOR, INSTITUTE OF TOXICOLOGY &<br />
RADIATION MEDICINE<br />
ENVIRONMENTAL HEALTH<br />
ATTN: V. CHUMAK ATTN: LIBRARY, LAWRENCE BERKELEY LABORATORY<br />
AUSTRALIAN DEFENCE FORCE<br />
UNIVERSITY OF CINCINNATI<br />
ATTN: SURGEON GENERAL ATTN: UNIVERSITY HOSPITAL, RADIOISOTOPE<br />
LABORATORY<br />
AUTRE, INC.<br />
ATTN: PRESIDENT XAVIER UNIVERSITY OF LOUISIANA<br />
ATTN: COLLEGE OF PHARMACY
REPORT DOCUMENTATION PAGE<br />
I<br />
FormApproved<br />
OMB No. 0704-OIBB<br />
Public reworting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructionh. searching existing data sources.<br />
gathering and maintaining the data needed, and completing and reviewing th• collection of information Send comnments regarding this burden estimate or any other aspect of this<br />
collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for information Operations and <strong>Reports</strong>. 1215 Jefferson<br />
Davis Highway, Suite 1204, Arlington, VA 22202-4302. and to the Office of Management and Budget, Paperwork Reduction Project 10704-0188). Washington. DC 20503<br />
1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED<br />
I December 1997 Contract Report<br />
4. TITLE AND SUBTITLE 5. FUNDING NUMBERS<br />
Retrospective Reconstruction of Radiation Doses of<br />
<strong>Chernobyl</strong> Liquidators by Electron Paramagnetic Resonance NWED QAXM<br />
6. AUTHOR(S)<br />
Chumak, V.V., Likhtarev, I.A., Sholom, S.S.,<br />
Pasalskaya, L.F., Pavlenko, Y.V.<br />
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION<br />
Armed Forces Radiobiology Research Institute<br />
REPORT NUMBER<br />
8901 Wisconsin Avenue CR97-2<br />
Bethesda, MD 20889-5603<br />
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING<br />
AGENCY REPORT NUMBER<br />
11. SUPPLEMENTARY NOTES<br />
12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE<br />
Approved for public release; distribution unlimited.<br />
13. ABSTRACT (Maximum 200 words)<br />
Accurate, reliable dose reconstruction is a critical component in the epidemiological followup of<br />
liquidators. Dosimetry of teeth by electron paramagnetic resonance (EPR) is a state-of-the-art laboratory<br />
technique that is key to this effort. The Scientific Center of Radiation Medicine (SCRM) has developed<br />
and refined this technique in order to meet the practical demands of large-scale epidemiologic followup of<br />
the <strong>Chernobyl</strong> liquidators. Independent analysis using similar technology was performed by investigators<br />
at the University of Utah and showed good correlation with the SCRM results. The lower limit of detection<br />
for reliable dose reconstruction was 100 mGy. Techniques were applied to samples from approximately<br />
135 liquidators involved in cleanup activities within the first 2 years after the <strong>Chernobyl</strong> accident in 1986.<br />
Mean dose was 287 mGy, geometric mean was 205 mGy, and median dose value was 200 mGy. The<br />
reconstructed dose values range from 30 to 2220 mGy. Correlation of results between the two institutions<br />
was generally within 17%. This report also addresses some confounding factors (previous medical x-ray<br />
exposures, ultraviolet light effects on anterior teeth, nonlinearity of dose response curves below 100 mGy)<br />
and how to deal with them.<br />
14. SUBJECT TERMS 15. NUMBER OF PAGES<br />
56<br />
16. PRICE CODE<br />
17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF<br />
OF REPORT OF THIS PAGE OF ABSTRACT ABSTRACT<br />
UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED UL<br />
NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89)<br />
Prescribed by ANSI Sts. 239-18<br />
298-102
DTRA-TR -00-14<br />
,,:<br />
t I .<br />
~I<br />
I Neurocognitive & Physical Abilities Assessments<br />
Twelve Years After the Chemobyl <strong>Nuclear</strong> <strong>Accident</strong><br />
. Approved for public release; distribution is unlimited.<br />
March 2001<br />
Prepared for:<br />
<strong>Defense</strong> Threat Reduction Agency<br />
8725 JohnJ. Kingman Road, Mail Stop 6201<br />
Fort Belvoir, VA 22060-6201<br />
20010504 158<br />
DSWA 01-97-C-0047<br />
Gerald L. Gamache, et. aI.<br />
Prepared by: KGA International<br />
8 Althea Street<br />
St. Augustine, FL 32084-2181
DESTRUCTION NOTICE:<br />
Destroy this report when it is no longer needed.<br />
return to sender.<br />
Do not<br />
PLEASE NOTIFY THE DEFENSE THREAT REDUC<br />
TION AGENCY, ATTN: ADM, 8725 JOHN 1. KING<br />
MAN ROAD, MS6201, FORT BELVOIR, VA 22060-<br />
6201 ,IF YOUR ADDRESS IS INCORRECT, IF YOU<br />
WISH IT DELETED FROM THE DISTRlBUTION<br />
LIST, OR IF THE ADDRESSEE IS NO LONGER<br />
EMPLOYED BY YOUR ORGANIZATION.
I:<br />
,:<br />
I<br />
I:<br />
I:<br />
DISTRIBUTION LIST UPDATE<br />
I:<br />
I; Ibis mailer is provided to enable DTBA to maintain current distribution lists for reports, (We would<br />
I; appreciate you providing the reguested information,)<br />
I:<br />
I:<br />
Ct Add the individual listed to your cflStnbution list<br />
Cl<br />
Cl<br />
Delete the cited organizationlindividual.<br />
Change of address,<br />
Nota:<br />
Please retum the maJDng label from the<br />
document so that any additions, changes,<br />
corrections or deletions can be made easily,<br />
For distribution cancellation or mora<br />
Information can DTRAlADM (703) 325-1036.<br />
NAME: __________________________________________________ __<br />
·ORGANIZATION: ______________________________________________ __<br />
z<br />
a:<br />
~<br />
a:<br />
o<br />
z<br />
<<br />
w<br />
a:<br />
w<br />
::I:<br />
l<br />
=><br />
U<br />
OLD ADDRESS<br />
NEW ADDRESS<br />
TELEPHONE NUM8ER: ~(~~) ____________________________________ __<br />
DTRA PUBLICATION NUMBERITITLE<br />
CHANGESIDELETlONSlADOmONS, etc.)<br />
(Allach ShHt If mot8 ~,. ~<br />
DTRAoro~erGOVERNMENTCONTRACTNUM8ER: ________________________ __<br />
CERTIFICATION of NEED-TO-KNOW BY GOVERNMENT SPONSOR [If other than DTRA):<br />
SPONSORING ORGANIZATION:<br />
------------------------------------------<br />
CONTRACTING OFFICER or REPRESENTATIVE:<br />
SIGNATURE:
DEFENSE THREAT REDUCTION AGENCY<br />
ATTN: ADM<br />
45045 AVIATION DRIVE<br />
DULLES, VA 20156-7517<br />
DEFENSE THREAT REDUCTION AGENCY<br />
ATTN: ADM<br />
6801 TELEGRAPH ROAD<br />
ALEXANDRIA, VA 22310-3398
REPORT DOCUMENTATION PAGE<br />
Form Approved<br />
OMB No. 0704-0811<br />
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing dala sources, gathering and<br />
mainlaining the date needed, and completing and reviewing the collection of information. Send commenls regarding this burden estimate or any other aspect of this collection of information,<br />
including suggestions for reducing this burden, to Washington Headquarters Se!vices, Directorate for Information Operations and <strong>Reports</strong>. 1215 Jefferson Davis Highway, Suite 1204, Arlington,<br />
VA 22202-4302, and to the OffICe of Manaaement and Budaet. Paperwork Reduction Pro· ect (0704-0188), Washinirton DC 20503.<br />
1. AGENCY USE ONLY (Leave Blank) 2. REPORT DATE 13. REPORT TYPE AND DATES COVERED<br />
990930 Technical 970519 - 990930<br />
4. TITLE AND SUBTITLE 4. FUNDING NUMBERS<br />
Neurocognitive & Physical Abilities Assessments Twelve Years After C -DSWA01 - 97 - C - 0047<br />
the Chemobyl <strong>Nuclear</strong> <strong>Accident</strong>: Final Report<br />
PE -RTD&E<br />
PR -AF<br />
5. AUTHOR(S)<br />
Gerald L. Gamache, Dennis L. Reeves, Daniel M. Levinson, and<br />
TA-AD<br />
Peter I. Bidiouk<br />
WU-DH54909<br />
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION<br />
REPORT NUMBER<br />
KGA International<br />
8 Althea Street KGAI Report 97-1<br />
St. Augustine, FL 32095-2181<br />
9. SPONSORINGIMONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORINGIMONITORING<br />
AGENCY REPORT NUMBER<br />
<strong>Defense</strong> Threat Reduction Agency<br />
8725 John J. Kingman Road, MS-6201 DTRA-TR-00-14<br />
Fort Belvoir, VA 22060-6201<br />
TDACClKehlet<br />
11. SUPPLEMENTARY NOTES<br />
This work was sponsored by the <strong>Defense</strong> Threat Reduction Agency under RDT &E RMC code B 4662 D<br />
AF AD 549094WlO A AF 25904D.<br />
12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE<br />
Approved for public release; distribution is unlimited<br />
13. ABSTRACT(Maximum 200 words)<br />
In an effort to assess the effects of exposure to ionizing radiation on neuropsychological and physical<br />
abilities, a longitudinal study in and near Chemobyl, Ukraine was conducted. In this report are findings<br />
from 1995 to 1998. Participants were volunteers who resided in Ukraine during and since the Chemobyl<br />
<strong>Nuclear</strong> Power Plant accident. A translated subset of the Automated Neuropsychological Assessment<br />
Metrics battery and the Gamache Physical Abilities Battery were administered to a control and three<br />
experimental groups. Controls were healthy volunteers who resided well outside of the exposed area.<br />
Eliminators were decontamination and reconstruction workers with known levels of exposure. Forestry<br />
and Agricultural workers resided and worked in contaminated areas. Analyses of 1995 - 1998 4-year<br />
averaged results indicated the Eliminators were significantly impaired on all measures of neurocognitive<br />
14. SUBJECT TERMS<br />
Chemobyl GPAB radiation effects<br />
ANAM<br />
radionuclides<br />
17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION<br />
OF REPORT OF THIS PAGE OF ABSTRACT<br />
UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED<br />
15. NUMBER OF PAGES<br />
117<br />
16. PRICE CODE<br />
20. LIMITATION OF<br />
ABSTRACT<br />
SAR<br />
NSN 754M1-280-5500<br />
Standard Form 298 (Rev.2-89)<br />
Prescribed by ANSI Sta. 239-18<br />
298-102
~ ___ UNCLASS~D ____ _<br />
SECURITY CLASSIFICATION OF TInS PAGE<br />
CLASSIFIED BY:<br />
NI A since Unclassified<br />
DECLASSIFY ON:<br />
NI A since Unclassified<br />
13. ABSTRACT (Continued)<br />
and physical performance as compared to Controls. Forestry and Agricultural workers were impaired<br />
on subsets of the neurocognitive and physical batteries. Significant correlations between levels of<br />
radiation dosage and 4-year averaged physical and cognitive performance were obseIVed on 21 of24<br />
tasks for the combined exposure groups. The results appear to reflect the existence of clinically<br />
meaningful neurotoxic effects of both acute and chronic exposure to radionuclides.<br />
ii<br />
SECURITY CLASSIFICATION OF THIS PAGE<br />
UNCLASSIFIED
PREFACE<br />
This document represents the Final Report of a four year effort sponsored by the <strong>Defense</strong> Threat<br />
Reduction Agency (DTRA). This report provides the Tables and Figures to substantiate the effects of<br />
low-dosage radiation (less than 70 rads) on a Ukrainian population exposed as a result of the <strong>Chernobyl</strong><br />
disaster.<br />
The success of this study was dependent on the active participation of a group of subject matter experts<br />
from military and civilian services on an international basis. Our first line acknowledgment and sincere<br />
appreciation is extended to our sponsor, the United States <strong>Defense</strong> Threat Reduction Agency (DTRA).<br />
Further, special thanks goes to Robert A. Kehlet, our Program Manager at DTRA. He has been our<br />
champion in this project, and successes to date are due to his dedicated scientific and program guidance.<br />
Development of the TWB/ANAM, which has served as the neuropsychological test instrument in this<br />
project, was sponsored by the U.S. Army Office of Military Performance Assessment Technology,<br />
Walter Reed Army Institute of Research, Washington, DC, Dr. F.W. Hegge, Program Director. The<br />
TWB/ANAM system was constructed at the Naval Computer & Telecommunications Station, NAS,<br />
Pensacola, FL, K. P. Winter, Principal Investigator. Authors of the ANAMUKR Battery were D. Reeves,<br />
G. Gamache, A. Chervinsky, & P. Bidiouk. Finally, the authors would like to express a special note of<br />
appreciation for the exceptional volunteer technical assistance provided by Dr. J. Wood, Krug Life<br />
Sciences during the English to RussianlUkrainian portion of this project. Her knowledge and experience<br />
with the NASA-MIR projects proved invaluable in helping us to not "re-invent" the wheel and launch<br />
our project "on-schedule."<br />
On-site Supervisors in the Ukraine were: Damian V. Kolisnyk, who supervised all aspects of testing the<br />
control group, Nikolay N. Kaletnik, who supervised all aspects of testing the forestry workers, and<br />
Victor G. Bondarenko, who supervised all aspects of testing the eliminators. Dr. Peter I. Bidiouk was<br />
the on-site project scientist and over-site supervisor for all aspects of the selection of test sites and the<br />
selection of participants, as well as being the on-site project administrator in Ukraine (see Appendix A).<br />
Finally, the authors would like to express sincere appreciation to Dr. A. J. Glasner, National Cognitive<br />
Recovery Foundation Editorial Board Member, who has been instrumental in production of this report<br />
and helping us remain "on-track" and in adherence with NCRF lAP A guidelines.<br />
111
CONVERSION TABLE<br />
Conversion factors for U.S. Customary to metric (SI) units of measurement.<br />
MULTIPLY • BY ~ TOGET<br />
TO GET ~ BY ~ DIVIDE<br />
angstrom 1.000 000 X E -10 meters<br />
atmosphere (normal) 1.013 25 X E +2 kilo pascal (kPa)<br />
bar 1.000 000 X E +2 kilo pascal (kPa)<br />
bam 1.000 000 X E -28 meter (mz)<br />
British thermal unit (thermochemical) 1.054 350 X E +3 joule (1)<br />
calorie (thermochemical) 4.184 000 joule (1)<br />
cal (thermochemical/em?) 4.184 000 X E -2 mega joule/mz (MJ/m 2 )<br />
curie 3.700 000 X E +l *giga becquerel (GBq)<br />
degree (angle) 1.745 329 X E -2 radian (rad)<br />
degree Fahrenheit !I< = (tOf + 459.67)/1.8 degree kelvin (K)<br />
electron volt 1.602 19 X E -19 joule (1)<br />
erg 1.000 000 X E -7 joule (1)<br />
erg/second 1.000 000 X E -7 watt{W)<br />
foot 3.048 000 X E -1 meter(m)<br />
foot-pound-force 1.355 818 joule (1)<br />
gallon (u.s. liquid) 3.785 412 X E -3 mete~ (m 3 )<br />
inch 2.540 000 X E -2 meter (m)'<br />
jerk 1.000 000 X E +9 joule (1)<br />
joulelkilogram (j/kg) radiation dose<br />
absorbed 1.000 000 Gray(Gy)<br />
kilotons 4.183 terajoules<br />
kip (1000 lbf) 4.448 222 X E +3 newton(N)<br />
kiplinch 2 (ksi) 6.894 757 X E +3 kilo pascal (kPa)<br />
ktap 1.000 000 X E +2 newton-second! m Z (N-slm 2 )<br />
micron 1.000 000 X E -6 meter(m)<br />
mil 2.540 000 X E -5 meter(m)<br />
mile (international) 1.609 344 X E +3 meter(m)<br />
ounce 2.834 952 X E -2 kilogram (kg)<br />
pound-force (lbs avoirdupois) 4.448 222 newton(N)<br />
pound-force inch 1.129 848 X E -1 newton-meter (N"m)<br />
pound-force/inch 1.751 268 X E +2 newton-meter (N/m)<br />
pound-force/foof 4.788 026 X E -2 kilo pascal (kPa)<br />
pound-force/inch 2 (PSi) 6.894 757 kilo pascal (kPa)<br />
pound-mass (Ibm avoirdupois) 4.535 924 X E -1 kilogram (kg)<br />
pound-mass-foof (moment of inertia) 4.214 011 X E -2 kilogram-meter (kg. m 2 )<br />
pound-massIfoof 1.601 846 X E +l kilogram-mete~ (kg/m 3 )<br />
rad (radiation dose absorbed) 1.000 000 X E -2 **Gray(Gy)<br />
roentgen 2.579 760 X E -4 ooulomblkilogram (CIkg)<br />
shake 1.000 000 X E -8 seconds (s)<br />
slug 1.459 390 X E +1 kilogram (kg)<br />
torr ~mm Hg, (Y C) 1.333 22 X E -1 kilo pascal (kPa)<br />
• The becquerel (Bq) is the SI unit of radioactivity; 1 Bq = 1 event/so<br />
**The Gray (Gy) is the SI unit of absorbed radiation.<br />
IV
TABLE OF CONTENTS<br />
Section<br />
Page<br />
PREFACE ................................................................................................................................. iii<br />
CONVERSION TABLE ........................................................................................................... iv<br />
FIGURES ................................................................................................................................. vii<br />
TABLES ..................................................................................................................................... x<br />
1 INTRODUCTION ..................................................................................................................... 1<br />
2 METHOD ................................................................................................................................ 20<br />
2.1 PARTICIPANTS ......................................................................................................... 20<br />
2.2 INSTRUMENTS .......................................................................................................... 21<br />
2.2.1 Gamache Physical Abilities Battery (GP AB) ............................................... 21<br />
2.2.2 Automated Neuropsychological Assessment Battery-Ukraine .................... 21<br />
2.3 ASSESSMENT SITES AND ENVIRONMENT ........................................................ 22<br />
2.4 PROCEDURE .............................................................................................................. 22<br />
3 RESULTS ................................................................................................................................. 24<br />
3.1 OVERVIEW OF 4-YEAR RESULTS: 1995-1998 ..................................................... 24<br />
3.1.1 GPAB ............................................................................................................ 22<br />
3.1.2 ANAMUKR: Accuracy ................................................................................. 27<br />
3.1.3 ANAMUKR: Efficiency ............................................................................... 27<br />
3.1.4 ANAMUKR: Additional Measures .............................................................. 28<br />
3.1.5 A Clinical Neuropsychological Interpretation ofChemobyl-ANAM<br />
Data ............................................................................................................... 30<br />
3.2 CORRELATIONS BETWEEN DOSAGE OF RADIATION AND 4-YEAR<br />
AVERAGED PERFORMANCE LEVELS ................................................................ .31<br />
3.3 RESULTS OF 1995 INITIAL TEST SESSION ......................................................... .43<br />
3.3.1 GPAB ............................................................................................................ 44<br />
3.3.2 ANAMUKR: Accuracy of Performance .......................................................47<br />
3.3.3 ANAMUKR: Efficiency of Performance (Throughput) ............................... 50<br />
3.3.4 ANAMUKR: Additional Measures .............................................................. 53<br />
3.3.5 GPAB-ANAMUKR: Correlational Analyses ...............................................53<br />
3.4 RESULTS OF 1996 RETEST SESSION .................................................................... 56<br />
v
TABLE OF CONTENTS (Continued)<br />
Section<br />
Page<br />
3.5 GLOBAL ASSESSMENTS OF DECLINES BY EXPOSURE GROUPS ................. 56<br />
3.6 SPECIFIC ASSESSMENTS OF DECLINES IN EXPOSURE GROUPS ................. 60<br />
3.6.1 GPAB ............................................................................................................ 60<br />
3.6.2 ANAMUKR: Accuracy ................................................................................. 60<br />
3.6.3 ANAMUKR: Efficiency .............................................................................. ~61<br />
3.6.4 ANAMUKR: Additional Measures .............................................................. 62<br />
3.7 RESULTS OF 1997 RETEST SESSION .................................................................... 64<br />
3.7.1 Global Assessments of Declines in Exposure Groups .................................. 64<br />
3.7.2 Specific Assessments of Declines in Exposure Groups ................................ 67<br />
3.7.3 GPAB ............................................................................................................ 69<br />
3.7.4 ANAMUKR: Accuracy ................................................................................. 69<br />
3.7.5 ANAMUKR: Efficiency ............................................................................... 69<br />
3.7.6 ANAMUKR: Additional Measures ............................................................. 70<br />
3.8 RESULTS OF 1998 RETEST SESSION .................................................................... 70<br />
3.8.1 Assessments of Declines by Exposure Groups ............................................. 71<br />
3.8.2 GPAB ............................................................................................................ 71<br />
3.8.3 ANAMUKR: Accuracy ................................................................................. 71<br />
3.8.4 ANAMUKR: Efficiency ............................................................................... 72<br />
3.8.5 ANAMUKR: Additional Measures .............................................................. 73<br />
4 CONCLUSIONS ...................................................................................................................... 86<br />
5 REFERENCES ........................................................................................................................ 88<br />
Appendix<br />
A UKRANIAN PROJECT DEPARTMENTS AND PERSONNEL CONTACTED .............. A-I<br />
B PHOTOGRAPHS OF CHERNOBYL AND TESTING SITES ............................................ B-1<br />
C CONSENT FORM ................................................................................................................. C-1<br />
D GLOSSARY ......................................................................................................................... D-1<br />
DISTRIBUTION LIST ....................................................................................................... DL-l<br />
VI
FIGURES<br />
Figure<br />
Page<br />
3-1 4-year averaged performances on GPAB: BROADJUMP ........................................................ 25<br />
3-2 4-year averaged performances on GPAB: CARRYING WEIGHT ...........................................26<br />
3-3 4-year averaged performances on GPAB: SQUAT THRUSTS ...............................................26<br />
3-4 4-year averaged performances on GPAB: BALANCE BEAM ................................................27<br />
3-5 4-year averaged performances on ANAMUKR: 2CH-ACC ....................................................33<br />
3-6 4-year averaged performances on ANAMUKR: CDS-ACC .....................................................33<br />
3-7 4-year averaged performances on ANAMUKR: CDI-ACC ..................................................... 34<br />
3-8 4-year averaged performances on ANAMUKR: CDD-ACC ...................................................34<br />
3-9 4-year averaged performances on ANAMUKR: CPT -ACC ....................................................35<br />
3-10 4-year averaged performances on ANAMUKR: DGS-ACC ....................................................35<br />
3-11 4-year averaged performances on ANAMUKR: MSP-ACC ....................................................36<br />
3-12 4-year averaged performances on ANAMUKR: SPD-ACC ....................................................36<br />
3-13 4-year averaged performances on ANAMUKR: SRT-EFF .......................................................37<br />
3-14 4-year averaged performances on ANAMUKR: 2CH-EFF ......................................................37<br />
3-15 4-year averaged performances on ANAMUKR: CDS-EFF ......................................................38<br />
3-16 4-year averaged performances on ANAMUKR: CDI-EFF .......................................................38<br />
3-17 4-year averaged performances on ANAMUKR: CDD-EFF ...................................................... 39<br />
3-18 4-year averaged performances on ANAMUKR: CPT-EFF .......................................................39<br />
3-19 4-year averaged performances on ANAMUKR: DGS-EFF ......................................................40<br />
3-20 4-year averaged performances on ANAMUKR: MSP-EFF ......................................................40<br />
3-21 4-year averaged performances on ANAMUKR: SPD-EFF .......................................................41<br />
VB
FIGURES (Continued)<br />
Figure<br />
Page<br />
3-22 4-year averaged performances on ANAMUKR: T APPING-RlGHT .......................................41<br />
3-23 4-year averaged performance on ANAMUKR: TAPPING LEFT ...........................................42<br />
3-24 4-year averaged scores on ANAMUKR: SLEEP SCALE .......................................................42<br />
3-25 Mean % Decrement for Exposure Groups Relative to Controls-1995 .....................................43<br />
3-26 Mean % Decrement for Exposure Groups-1996 Relative to Controls-l 995 ............................ 57<br />
3-27 Mean % Decline for Exposure Groups: 1995-1996 ................................................................. 59<br />
3-28 Mean % Decrement for Exposure Groups-1997 Relative to Controls-1995 ............................ 64<br />
3-29 Mean % Decline for Exposure Groups: 1996-1997 ................................................................. 66<br />
3-30 Mean Performance on GP AB: BROAD JUMP ......................................................................... 74<br />
3-31 Mean Performance on GPAB: CARRYING WEIGHT ........................................................... 74<br />
3-32 Mean Performance on GPAB: SQUAT THRUSTS ................................................................. 75<br />
3-33 Mean Performance on GPAB: BALANCE BEAM .................................................................. 75<br />
3-34 Mean Performance on ANAMUKR: 2CH-ACC ...................................................................... 76<br />
3-35 Mean Performance on ANAMUKR.: CDS-ACC ...................................................................... 76<br />
3-36 Mean Performance on ANAMUKR: CDI-ACC ....................................................................... 77<br />
3-37 Mean Performance on ANAMUKR: CDD-ACC ..................................................................... 77<br />
3-38 Mean Performance on ANAMUKR: CPT-ACC ...................................................................... 78<br />
3-39 Mean Performance on ANAMUKR: DGS-ACC ..................................................................... 78<br />
3-40 Mean Performance on ANAMUKR: MSP-ACC ..................................................................... 79<br />
3-41 Mean Performance on ANAMUKR: SPD-ACC ...................................................................... 79<br />
Vlll
FIGURES (Continued)<br />
Figure<br />
Page<br />
3-42 Mean Performance onANAMUKR: SRT-EFF ....................................................................... 80<br />
3-43 Mean Performance on ANAMUKR: 2CH-EFF ....................................................................... 80<br />
3-44 Mean Performance on ANAMUKR: CDS-EFF ....................................................................... 81<br />
3-45 Mean Performance on ANAMUKR: CDI-EFF ........................................................................ 81<br />
3-46 Mean Performance on ANAMUKR: CDD-EFF ...................................................................... 82<br />
3-47 Mean Performance onANAMUKR: CPT-EFF ....................................................................... 82<br />
3-48 Mean Performance on ANAMUKR: DGS-EFF ....................................................................... 83<br />
3-49 Mean Performance on ANAMUKR: MSP-EFF ....................................................................... 83<br />
3-50 Mean Performance on ANAMUKR: SPD-EFF ....................................................................... 84<br />
3-51 Mean Performance on ANAMUKR: TAPPING-RIGHT ......................................................... 84<br />
3-52 Mean performance on ANAMUKR: TAPPING-LEFT ........................................................... 85<br />
3-53 Mean Ratings on ANAMUKR: SLEEP SCALE ...................................................................... 85<br />
lX
TABLES<br />
Table<br />
Page<br />
1-1 Comparison of radionuclide content released into the environment as a nuclear<br />
weapons tests with levels resulting from the Chemobyl accident ................................................ 3<br />
1-2 Volume ofCs J37 in the result of Dnieper reservoirs, measured in Curies .................................... .4<br />
1-3 Distribution of radio nuclides in Kiev and suburbs ....................................................................... 5<br />
1-4 Contamination by Cs 137 ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••• 6<br />
1-5 Contamination by Sr 90 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••• 7<br />
1-6 Radionuclide activity in suburbs of Kiev: May, 1986 .................................................................. 7<br />
1-7 Contamination of water by tritium ............................................................................................... 8<br />
1-8 Measurement of gamma radiation for 30-km zone and adjacent areas, May 16, 1986 ................ 9<br />
1-9 Gamma radiation in forest south of Chemobyl at various distances: May, 1 Device:<br />
DP-58 (milliradslhr) .................................................................................................................... 1 0<br />
1-10 Result of forest damage by radiation within the 30-kilometer zone ........................................... 10<br />
1-11 Radioactive particles from the Brown F orest ............................................................................. 11<br />
1-12 Contamination of forests by CS 137 ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••• 12<br />
1-13 Concentration ofSro and Pu in 1991 Kiev foliage ..................................................................... 13<br />
1-14 Contamination ofCs!37 in berries, mushrooms, and medical herbs ............................................ 13<br />
1-15 Contamination of Cs 137 in wood .................................................................................................. 14<br />
1-16 Disease rates for relocated individuals ........................................................................................ 16<br />
1-17 Disease rates among individuals relocated to Kiev .................................................................... 16<br />
1-18 Percentages of individuals who are considered healthy .............................................................. 17<br />
2-1 Demographic information and mean dose of radiation for the 4 groups .................................... 20<br />
2-2 Description of Gamache Physical Abilities Battery (GPAB) ..................................................... 21<br />
x
TABLES (Continued)<br />
Table<br />
Page<br />
3-1 4-Year averaged performance on GP AB: physical tasks ............................................................ 25<br />
3-2 4-Year averaged performances on ANAMUKR: Accuracy ....................................................... 28<br />
3-3 4-Year averaged performances on ANAMUKR: Efficiency ...................................................... 29<br />
3-4 4-Year averaged performances on ANAMUKR: Additional Measures ..................................... 29<br />
3-5 Correlations between dosage of radiation and performance levels ............................................ .32<br />
3-6 Mean % performance decrement for exposure groups relative to controls-1995 ...................... .43<br />
3-7 GP AB: 1995 Means (and Standard Deviations) for the four groups ......................................... .44<br />
3-8 GP AB: Results ofMANOV A and Univariate ANOV AS ......................................................... .45<br />
3-9 GPAB: Groups significantly lower on physical abilities tasks thanAC .................................... .45<br />
3-10 GPAB: Results of discriminant function analysis for groupsAC andAE ................................ .45<br />
3-11 GPAB: Means (and Standard Deviations) for females and males, either not exposed<br />
(Controls) or exposed to radiation ............................................................................................. .46<br />
3-12 GPAB: MANOVA and univariate tests for females and males, either exposed or not<br />
exposed (Controls) to radiation .................................................................................................. .47<br />
3-13 Accuracy (% Correct): 1995 Means (and Standard Deviations) ................................................ .48<br />
3-14 Accuracy: Results ofMANOVA and univariate ANOVAs ...................................................... .48<br />
3-15 Accuracy: Groups significantly less accurate thanAC .............................................................. .49<br />
3-16 Accuracy: Results of discriminant function analysis for groupsAC andAE ............................ .49<br />
3-17 Efficiency (Correct responses/min): 1995 Means (and Standard Deviations) ............................ 51<br />
3-18 Efficiency: Results ofMANOVA and univariate ANOVAs (and MANCOVA and<br />
univariate ANCOVAs with SRI as a covariate) ........................................................................ 52<br />
3-19 Efficiency: Groups significantly less efficient thanAC. ............................................................. 52<br />
xi
TABLES (Continued)<br />
Table<br />
Page<br />
3-20 Efficiency: Results of discriminant function analysis for groupsAC andAE ........................... 53<br />
3-21 ANAMUKR Additional Measures: Means (and Standard Deviations) ..................................... .53<br />
3-22 Pearson correlations between GPAB and ANAMUKR: Accuracy (N=127) ............................. .54<br />
3-23 Pearson correlations between GPAB and ANAMUKR: Efficiency (N=127) ............................ 55<br />
3-24 Pearson correlations between GP AB and ANAMUKR: Additional Measures<br />
(N=127) ....................................................................................................................................... 55<br />
3-25 Mean % Performance decrement for Exp. Groups-1996 relative to Controls-1995 ................... 56<br />
3-26 Groups not significantly lower thanAC in 1995, but significantly lower in 1996 ..................... 58<br />
3-27 Mean % performance decline for exposure groups: 1995-1996 ................................................. 58<br />
3-28 Significant multivariate declines by exposure groups ................................................................ 59<br />
3-29 GPAB: 1996 Means (and Standard Deviations) for the exposure groups ................................. 60<br />
3-30 ANAMUKR Accuracy: 1996 Means (and Standard Deviations) for the<br />
exposure groups ........................................................................................................................... 61<br />
3-31 ANAMUKR Efficiency: 1996 Means (and Standard Deviations) for the<br />
exposure groups ........................................................................................................................... 62<br />
3-32 ANAMUKR Additional Measures: 1996 Means (and Standard Deviations) ............................. 62<br />
3-33 Significant declines in performance by the exposure groups: 1995 to 1996 .............................. 63<br />
3-34 Mean % Performance decrement for expo groups-1997 relative to Controls-1995 .................... 64<br />
3-35 Groups not significantly lower thanAC in 1995 or 1996, but significantly lower<br />
in 1997 ......................................................................................................................................... 65<br />
3-36 Mean % performance decline for exposure groups: 1996-1997 ................................................. 66<br />
3-37 Significant multivariate declines by exposure groups from 1996 to 1997 ................................. 67<br />
3-38 GPAB: 1997 Means (and Standard Deviations) for the exposure groups .................................. 67<br />
xii
TABLES (Continued)<br />
Table<br />
Page<br />
3-39 ANAMUKR Accuracy: 1997 Means (and Standard Deviations) for the<br />
exposure groups .......................................................................................................................... 68<br />
3-40 ANAMUKR Efficiency: 1997 Means (and Standard Deviations) for the<br />
exposure groups ........................................................................................................................... 68<br />
3-41 ANAMUKRAdditional Measures: 1997 Means (and Standard Deviations) ............................. 69<br />
3-42 Significant declines in performance by the exposure groups: 1996 to 1997 .............................. 70<br />
3-43 GPAB: 1998 Means (and Standard Deviations) for the exposure groups .................................. 71<br />
3-44 ANAMUKR Accuracy: 1998 Means (and Standard Deviations) for the<br />
exposure groups .......................................................................................................... , ................ 72<br />
3-45 ANAMUKR Efficiency: 1998 Means (and Standard Deviations) for the<br />
exposure groups ........................................................................................................................... 73<br />
3-46 ANAMUKR Additional Measures: 1998 Means (and Standard Deviations) ............................. 73<br />
xiii/xiv
SECTION 1<br />
INTRODUCTION<br />
The nuclear reactor accident at Chemobyl in 1986 resulted in geoecophysical damage and polluted<br />
farmlands and forests with radioactive contamination in im<br />
mediate and outlying areas. It has been estimated that 72% of the land mass of Ukraine is contaminated,<br />
and will be so for thousands of years (Yakovlev, 1992). By 1988, the Ukrainian registry contained<br />
names of347,619 civilians who had experienced medical symptoms that are frequently associated with<br />
exposure to the ionizing radiation fallout. In addition, 36,000 military personnel (Yakovlev, 1992) were<br />
listed as adversely affected. By 1992, over 1.5 million individuals were on government registries as<br />
having suffered medical problems associated with radiation exposures. This group included 350,225<br />
children and 180,144 persons assigned to "clean-up" duties at or very near the accident site<br />
(Awramenko, 1992). By 1993, about 7,000 people had died from apparent radiation-related illnesses,<br />
including heart, vascular, respiratory, and digestive diseases. One area of investigation that has been<br />
neglected during the aftermath of the Chemobyl accident has been the possible long term and subtle<br />
effects on neuropsychological functions.<br />
According to Chemousenko (1991) and Gittus et al. (1988), construction started on the Chemobyl<br />
<strong>Nuclear</strong> Power Plant (CNPP) in March, 1970. Plans called for construction of six High Power Channel<br />
Reactors, each having the capability to generate 1,000 megawatts of electricity. These are single-loop<br />
reactors, which means that steam travels directly to the turbine for electric generation. On September<br />
28, 1977, the fIrst of four reactors for CNPP became operational. Engineers working at the site warned<br />
of substantial problems because the reactor was unstable and emitting small doses of radioactivity.<br />
Despite these problems, more reactors were built, all operational until Saturday, April 26, 1986, when at<br />
1 :23 AM the fourth reactor exploded. The blast ripped off the roof and radioactive waste in the form of<br />
plutonium, cesium, and uranium dioxide were released (see Figure B-1 in Appendix B).<br />
At the reactor site, a 30-kilometer exclusion zone was established by the government to keep out nonscientifIc<br />
personnel. Cleanup and containment crews were dispatched to the scene where approximately<br />
660,000 volunteers and soldiers were employed. Only the Russian radiation monitors were outfItted for<br />
the task at hand (Figure B-2). The rest were "eliminators" composed of military units (infantry,<br />
helicopter crews, and engineer units), Ukrainian police, tractor and truck drivers from all Ukrainian<br />
provinces, medical doctors and nurses, scientists and engineers from Ukraine, Belarus, and Russia, and<br />
farm laborers (men and women) from Ukraine and Russia. The Ukrainian volunteers were<br />
approximately 396,515, while those living in Belarus and Russia comprised approximately 264,343.<br />
These cleanup crews were dressed only in surgical masks and lead aprons (Chemousenko, 1991). In<br />
neighboring Pripyat the population of 55,000 was not evacuated until 36 hours post-accident.<br />
Meanwhile in Kiev, government officials detected large doses of radiation; however, they chose not to<br />
cancel a planned May Day parade.<br />
Several years later, the accident continued to bring suffering to millions of people in the Ukraine,<br />
Belarus, and Russia. According to Baranov and Guskova (1988), and Laupa and Anno (1989), thirtyone<br />
workers at Chemobyl <strong>Nuclear</strong> Power Plant died a little over three months post-accident as a result of<br />
acute radiation sickness.<br />
1
Baryahtar (1991) reported that as many as 7,000 have died since then of radiation-related illnesses, and<br />
The International <strong>Chernobyl</strong> Project claimed that the scientific community lacked research and<br />
information on the medical consequences of radioactive pollution. Read (1993) stated that the number<br />
of fatalities ranged from thirty-one to "a projection that <strong>Chernobyl</strong> will ultimately claim more victims<br />
than did World War II."<br />
According to the newspaper Vestnik <strong>Chernobyl</strong>ia, Ukraine produces 54.4 million kilowatts-year of<br />
energy annually. Over 62% is generated from coal or oil burning plants, 8.3% from hydroelectric plants,<br />
23.5% from other nuclear power stations, and 5.5% from the two operating reactors at the <strong>Chernobyl</strong><br />
<strong>Nuclear</strong> Power Plant. In the winter of 1992-1993, nuclear power plants accounted for 40% of all electric<br />
power generated. In 1990,5.8% of this power was generated by <strong>Chernobyl</strong>; in 1991 the rate droppecl to<br />
5.2% while in 1992 the rate was only 2.2%.<br />
Two reactors are still operational at <strong>Chernobyl</strong> (Figure B-8). The third reactor was destroyed in a fire in<br />
October, 1991. Each reactor is programmed for 30 years of service; however, the two remaining<br />
reactors were scheduled to be deactivated in 1993. Decommissioning will cause substantial problems<br />
because the four reactors have generated over 244 million kilowatts-year of energy, including 94 million<br />
kilowatts-year since the accident. The two remaining reactors have the capability of generating 10.3<br />
million kilowatts-year of electrical energy at a cost of $515 million dollars. Utilizing them would save<br />
2.6 million tons of oil or 6.1 million tons of coal, purchases the Ukrainians cannot afford. In October,<br />
1993, the law which would have closed <strong>Chernobyl</strong> by the end of 1993 was repealed.<br />
For the year 1989, the absorbed dose for people at <strong>Chernobyl</strong> was 1.3 rads compared with a lifetime<br />
dose of35 rads. In 1992, the absorbed dose was only .97 rads. The amounts released into the<br />
atmosphere by the cracking sarcophagus comprise only 20% of the rate of the fully functional reactor.<br />
Decommissioning will involve four steps. First, the reactors are shutdown with a one-year cooling<br />
period. In step two, the nuclear fuel is removed and kept in water for one to two years before placing it<br />
in special storage. In stage three, <strong>Chernobyl</strong> will be temporary closed for a period of 20 to 30 years to<br />
allow any radionuclides present in the facility to decay. Finally, in stage four, <strong>Chernobyl</strong> will be<br />
dismantled.<br />
When <strong>Chernobyl</strong> was fully operational, a total of 50,000 persons worked at four reactors. By 1987-<br />
1988, 90% of these workers had been replaced. Of the replaced workers, more than 90% were under the<br />
age of 45 and 60% graduated from a university or technical colleges. Today, the 25,000 present staff<br />
lived about 50 kilometers from <strong>Chernobyl</strong> in the town of Slavutych, which was constructed in 1986 after<br />
the town of Prypiat was evacuated.<br />
Debate over the future of <strong>Chernobyl</strong> is still going on in the Ukrainian Supreme Soviet. On one side of<br />
the debate are those who believe the economy of Ukraine will not improve by decommissioning. They<br />
point out the replacement cost of alternative methods of generating<br />
electricity. They also point out the Dnieper basin has seven other reactors with the same design as<br />
<strong>Chernobyl</strong>, and another similar reactor will soon be operational near Kursk. They argue that <strong>Chernobyl</strong><br />
has a highly trained staff which will go elsewhere if the reactors are decommissioned, and each year<br />
savings from nuclear power can be turned to clean-up efforts and treating those still suffering from the<br />
accident.<br />
2
Finally, they argue that other third-world governments are not decommissioning Russian-made RBMK<br />
reactors; in fact, Russia is planning to double energy production by nuclear power by the year 2010,<br />
although they will not rely on the RBMK reactor.<br />
According to Yakovlev (1991), the Chemobyl incident resulted in waste emission of approximately 50-<br />
80 million Curies (Ci), including over a million Ci of Cs137, 200,000 Ci of Sr 90 and between 3,500 and<br />
5,500 Ci ofPu239124o into the environment.<br />
His research states that as of January 1, 1992, forty three thousand square kilometers, or 15% of the<br />
Ukraine, was contaminated by Cs137, with doses greater than one curie per square kilometer. Over 72%<br />
of Ukraine is contaminated above background radiation. This area has 3,200 towns and villages and a<br />
population, excluding Kiev, of over 4,000,000 people. Migratory birds and animals carry radiationrelated<br />
diseases along their routes, increasing the contaminated area. Table 1-1 (Sobodovych et al,<br />
1992) shows contamination from Chemobyl as compared with that resulting from nuclear weapons tests.<br />
While Chemobyl nuclear contamination is low compared with Russian nuclear weapons tests, the<br />
scientific community cannot afford to forget the fact that Chemobyl is adjacent to a heavily populated<br />
area.<br />
Table 1-1. Comparison of radionuclide content released into the environment as a<br />
result of nuclear weapons tests with levels resulting from the <strong>Chernobyl</strong><br />
accident.<br />
Type of Half-life in <strong>Nuclear</strong> Total in <strong>Chernobyl</strong> <strong>Chernobyl</strong> %of<br />
radio- years weapons test reactor outburst % outburst NWT<br />
nuclide (million Ci) (million Ci) (million Ci)<br />
Sr 90 28.60 57.5 6.00 6.0 0.30 0.50<br />
Cs 137 30.17 87.0 7.02 30.0 2.10 2.40<br />
PU 238 87.74 0.0055 0.0254 9.0 0.00076 13.80<br />
PU 239 24118.00 0.96 0.0256 9.0 0.00077 0.20<br />
PU 240 6570.00 0.50 0.040 9.0 0.0012 0.20<br />
PU 24 ! 1435.00 23.00 4.97 9.0 0.15 0.70<br />
PU 242 .763x10 5 0.00045 0.000056 9.0 0.000002 0.40<br />
It should be noted that 50-80 million curies of radiation were released into the environment from the<br />
Chemobyl accident, as compared to only 14-20 from the Three Mile Island incident.<br />
During the accident, reactor cooling water was flushed into the Prypiat River, a tributary of the Dnieper,<br />
rather than allowing more radioactive steam to escape into the atmosphere (Chemousenko, 1991). The<br />
result of this action, and the break in the sewage system that serviced the Chemobyl cleanup effort, are<br />
the principle causes of radionuclide pollution in the Dnieper and Prypiat rivers, as well as their water<br />
reservoirs which serve to irrigate the region. Thus the contamination is spread. At present, the vertical<br />
migration of radionuclides approaches a depth of one meter from the surface.<br />
3
According to Professor V. Kopeikin (1993), water near the burial place of<strong>Chernobyl</strong> debris is about 4<br />
meters underground. Ten wells with filters installed were constructed on-site to monitor ground water<br />
contamination. These wells are monitored daily and were placed every four meters, with depths ranging<br />
from 8-9 meters.<br />
The damaged reactor is covered by a sarcophagus composed of220,000 m 3 of concrete and 15,000 m 3 of<br />
steel, around which there are 5 to 6 meters of "clean topsoil" (Yakovlev, 1991).<br />
However, the sarcophagus has approximately 700 square meters of heat cracks caused by the damaged<br />
reactor core underground. The cracks cause venting of radioactive materials from fragments of the<br />
reactor core and the graphite bed underneath it. About 75% of the residual nuclear fuel is composed of<br />
clinker (135 tons) and nuclear dust (10 tons). The maximum temperature at the surface is 60 degrees<br />
centigrade, while the temperature underground approaches 200 degrees centigrade.<br />
On the surface radioactive contamination is about 3,000 roentgen per hour. In 1990-1991 the outburst<br />
from radioactive dust was 1,000 times less than that of a working reactor. However, with time and<br />
clinker decomposition, the amount of nuclear dust will increase. On May 30-31, 1990, <strong>Chernobyl</strong> was<br />
struck by an earthquake measuring 4.0 on the Richter scale. No damage to the sarcophagus was<br />
reported. In 1992 the Ukrainian government opened bidding on a replacement sarcophagus. In June,<br />
1993, the lowest bidder, a French company, began studying the problem. During 1987-1990,270,000<br />
rubles were spent to increase safety of all reactors in the Ukraine.<br />
According to Woytsehowich (1991), as shown in Table 1-2, between 1986-1990 the amount ofCs l37 in<br />
the Kiev water reservoir increased more than 17 times. In the Kremenchug reservoir, approximately 170<br />
miles south of Kiev, the pollution has doubled; this signifies a spreading of radio nuclides along the<br />
Dnieper basin.<br />
Table 1-2. Volume orcs!37 in the Dnieper reservoirs, measured in curies.<br />
Reservoir 1986 1987 1988 1989 1990<br />
Kiev 413 850 -- 1000 7200<br />
Kanev 60 -- -- 570 2200<br />
Kremenchug 150-200 -- 218 294 294<br />
The maximum values of specific activity of radionuclides as well as heavy metals are in riverbeds and<br />
shallow bays, while the minimum values are in the floodlands and irrigated fields. From CNPP, along<br />
the Prypiat and Dnieper rivers, down to Kiev Reservoir, three contamination zones have been<br />
established (Woytsehowich, 1991) for those areas affected by Cs 137 • The Migration Zone lies along the<br />
Prypiat river between CNPP and its entrance into the Dnieper. The Accumulation Zone is from the<br />
mouth to 15 kilometers upstream from the reservoir dam. The Wash Away Zone is from the point 15<br />
kilometers from the dam to the dam itself.<br />
4
Table 1-3 (Sobodovych et al., 1992) shows soil contamination in Kiev and its suburbs to be .015 to 5.31<br />
Ci/Km 2 for Csl37, .02 to .80 Cilkm 2 for Ce l44 , and .03 to 1.57 Ci/km 2 for RU 106 • Not shown in Table 1-3<br />
is the contamination by Sro, which is 50-750 Ci/Km 2 . The concentration of plutonium was reported by<br />
two sources. Source 1 (Ukrainian) shows the level of contamination to be 5 to 10 Ci/km 2 , while Source<br />
2 (Russian) shows the level of contamination to be much lower--0.1 to 2.7 Ci/km 2 . The authors were<br />
given no explanation for these disparities.<br />
Table 1 -3. Distribution of radionuclides in Kiev and suburbs.<br />
Note:<br />
Radionuclide Activity (CiIkg) Density of contamination<br />
Cs 137 2.6 x 10- 10 17.94 x 10- 8<br />
Ce l44 3.06 x 10- 10 11.12 x 10- 8<br />
RU lO6 4.31 x 10- 1 °/2.19 x 10- 8<br />
(1) Maximum Contamination Levels in Kiev<br />
1880 mCilkm 2 (Kudry, Pechersk)<br />
4524 mCilkm 2 (Montazbnikov, Sovley)<br />
1182 mCi/km2 (Davydova, Rusanovka)<br />
1367 mCi/km2 (Kybalchiucha, Voskresenka)<br />
1105 mCi/km2 (Kosmomantov, Otradny)<br />
(2) Pu Contamination:<br />
Source 1: 5 - 10 Ci/km 2<br />
Source 2: 0_1 - 2.7 Ci/km 2 (Ci/km 2 )<br />
0.015-5.31<br />
0.02-0.80<br />
0_03-1.57<br />
The city of Kiev comprises 340 km 2 ; however, only 7% is contaminated, mostly with Cs 137 • As shown in<br />
Tables 1-4 and 1-5 (Sobodovych et aI., 1992), in the largest area in the Kiev region which includes the<br />
city plus surrounding countryside, fifty percent of the total landmass has received greater than 40 Ci/km 2<br />
of contaminated radiation. This exceeds the recognized standard for lifetime exposure rate of 40 Ci/km 2 ;<br />
however, the city itself is quite habitable.<br />
Dr. Kaletnik, Head of the Scientific and Technical Office for the Ministry of Forest, was questioned why<br />
the Kiev region, rather than the city itself, was so highly contaminated. He explained that this<br />
phenomenon was due to excessive forest fires in the area, and that rising smoke and soot which blankets<br />
an area after rainfall would increase already high levels of contamination.<br />
5
Table 1-4. Contamination by CS137.<br />
Concentration<br />
Intervals<br />
(CiJKm 2 )<br />
0.5-1.0 1.0-5.0 5.0-15.0 15.0-40.0
Table 1-5. Contamination by Sr>°.<br />
Concentration<br />
Intervals<br />
CilKm2)<br />
0.05-.50 0.50-1.0 1.0-3.0
Finally, tritium, a radioactive isotope of hydrogen, was present at Chemobyl (Woytsenhowich, 1991).<br />
Tritium emits negative beta particles of 19,000 electron volts of energy and has a half-life of 12.5 years.<br />
Table 1-7 (Sobodovych et aI., 1992) shows the testing of water contaminated by tritium. Since tritium<br />
naturally occurs in water, probably the action of cosmic rays on atmospheric hydrogen, one would need<br />
prior testing to determine if these measurements were significant. If they were significant, the dates of<br />
these measurements might indicate serious problems.<br />
Table 1-7. Contamination of water by tritium.<br />
Type of Date of Number of<br />
Concentration (BklLiter)<br />
Water test test Range Median<br />
1. Atmospheric Jan 92-Feb 92 51 2.7 to 6.4 3.5<br />
(Snow, Water)<br />
2. Surface Water Nov 91-Jan 92 124 2.7 to 11.8 4.8<br />
3. Ground Water Nov 91-Jan 92 101 2.7 to 12.7 5.1<br />
4. Underground Sep 91-Jan 92 68 2.2 to 5.2 2.2<br />
Water<br />
Within weeks following the Chemobyl disaster, measurements to assess damage to forests surrounding<br />
the area were taken. Ultimately, about 500 hectars of forests were destroyed as a result of the accident.<br />
The government established a 30-kilometer fenced zone around the reactor with military patrols, whose<br />
purpose was to prevent unauthorized access. Even outside this zone, the density of contamination was<br />
10 to 80 Ci/km 2 • At least 28,000 km 2 of forest were contaminated.<br />
The first reported measurement of gamma radiation occurred on May 16, 1986,20 days post-accident<br />
(Sobodovych, 1992). Two devices were used for measurements. The first device, the Russian Army<br />
DP-5B, was available to units working in nuclear contaminated battlefields where it was used for rough<br />
estimates of nuclear contamination. The second device, SRP-6801, yielded more precise measurements<br />
and was carried by engineers who surveyed the site. Table 1-8 shows the results for measurements inside<br />
the 30-kilometer zone and adjacent forest areas. Approximately 100 hectare grids were laid out on maps<br />
by the Ministry of Forest. A sampling technique was then utilized. The number in parentheses<br />
following the forest name in the first column denotes the grid area.<br />
8
Table 1-8.<br />
Measurement of gamma radiation for 30-km zone and adjacent areas,<br />
May 16, 1986.<br />
DP-58<br />
SRP-6801<br />
Location of the measurement (milliradslhr) ( milliradslhr)<br />
Army device (more precise)<br />
1. Polissia Forest (18)<br />
Meadows 0.14 0.20<br />
Free in air 0.10 0.25<br />
Grass 0.12 0.25<br />
Carpet 0.20 Not Measured<br />
2. Radynskoye forest (12)<br />
Forest edge 0.13 0.25<br />
Meadows 0.20 0.41<br />
Trees 0.26 0.44<br />
Carpet 0.22 0.42<br />
Village of Cheremoshe 0.20 0.38<br />
3. Radynskoye forest<br />
Edge of pine trees 0.75 1.25<br />
Young mixed trees 0.75 1.50<br />
Free in air 0.75 1.40<br />
Crown of trees 0.75 1.45<br />
Grass 0.80 1.50<br />
Moss 1.50 1.50<br />
Oat field 0.45 1.25<br />
Ground 0.70 0.70<br />
Note: numbers in parentheses show grid number denoting location of sampling squares.<br />
Table 1-9 (Sobodovych, 1992) shows gamma radiation in forests south of the reactor. Measurements<br />
were also made in May, 1986. These forests contain coniferous trees, which are especially vulnerable to<br />
radiation. This table is useful because the distance from the reactor is measured, as well as the growth of<br />
pine trees in a particular forest.<br />
9
Table 1-9. Gamma radiation in forest south of <strong>Chernobyl</strong> at various distances:<br />
May,1986; Device: DP-58 (milliradslhr).<br />
Location Free in air Carpet<br />
1. Dymer forest (100 km*)<br />
Pine trees-50 years old 0.6 1.0<br />
2. Ivankov forest (80 km*)<br />
Pine trees-80 years old 0.7 0.7<br />
Pine trees-50 years old 1.5 2.8<br />
Pine trees-l 8 years old 1.3 3.2<br />
3. Chemobyl forest (20 km*)<br />
Pine trees-30 years old 3.5 10.0<br />
4. Novoshepelychi forest (l0 km*)<br />
Pine trees-30 years old 10.0 30.0<br />
Note: * denotes kilometers from reactor to center of forest.<br />
Table 1-10 is extremely interesting, since it shows the result of forest damage within the 30- kilometer<br />
zone. It was constructed based on conversations with Mr. Kaletnik. This table was cited in the Pacific<br />
Sierra Research Corporation's analysis of Landsat imagery (McClellan, G.E. et al., 1994). The distance<br />
from the reactor site is approximately one kilometer, and as Table 10 reflects, 100% of the trees located<br />
350 meters from the edge of the forest were recovered and sent to mills to be used as lumber. The rest of<br />
the trees were bulldozed into large pits and covered with topsoil. This action has increased ground<br />
contamination, and the Ukrainians are presently reviewing options to deal with this situation.<br />
Table 1-10. Result of forest damage by radiation within the 30-kilometer zone.<br />
Distance from Calculated absorbed % of tree Degree of 0/0 of<br />
the edge of the Dose-Rad x 10 3 crown harm recovered<br />
forest damage trees<br />
Edge of forest 10 100 Completely dry 0<br />
Wood<br />
35 meters 6.5 50 Very strong 25<br />
90 meters 4.9 20-30 Medium 50<br />
350 meters 0.5 Up to 10 Small 100<br />
10
In 1986, the crowns of trees contained about 50% of the radionuclides (Yakovlev, 1992). By 1988,95%<br />
of the radionuclides were in humus or carpet. Today, most radionuclide content of the foliage has<br />
migrated through the root system.<br />
Beginning in the summer of 1988 measurements were made of radioactive particles within the 30-<br />
kilometer zone in an area called the "Brown Forest", so named because the leaves and vegetation were<br />
discolored by radiation. Table 1-11 (Sobodovych, 1992) shows the results of these measurements, with<br />
the percentages of each isotope listed. The table also displays the particle properties as either irregular<br />
shaped flakes or round balls with varying composition of beryllium, cuprum, lead, silicon, tantalum, and<br />
iron. High concentrations of cerium, cesium, and ruthenium are shown.<br />
Table 1-11. Radioactive particles from the brown forest.<br />
Date of Form, size (mcm), and Basis of Element (percentages)<br />
Test properties of particles particle Ce l44 Cs 134 CS l37 Ru lo6 Co 6O<br />
Jun 1988 Black, hard, magnetic Oxides of 2.0 - 2.0 94.0 2.0<br />
balls, 46.7 mcm<br />
Fe<br />
Aug 1988 Irregular, 1.2-6.5 mcm and Be, Pb, Cu 54.2 5.4 15.5 24.9 -<br />
Balls, 4.1-5 mcm, dark br<br />
Jan 1989 Irregular, 1.6-88.0 mcm, Fe, Si, Pb 50.0 4.5 22.3 29.0 -<br />
dark br, nonmagnetic<br />
Jan 1989 Irregular, fragile, black, Fe, Si 4.0 8.9 92.6 24.5 -<br />
Nnnmagnetic, 2.0-20 mcm,<br />
Balls, 1.0-4.0 mcm<br />
Jan 1989 Irregular, 1.2-6.4 mcm, Si, Ta 88.2 0.8 3.8 7.2 -<br />
dense, black, with balls,<br />
0.6-2.4mcm<br />
Table 1-12 (Sobodovych, 1992) displays a list of forests contaminated by Cs 137 and the levels of<br />
contamination. These measurements were made in 1990 and 1991. Figures for 1992 were not available.<br />
Table 1-13 (Woytsehowich, 1991) shows measurements taken from Kiev in 1991 in which foliage<br />
samples were collected, burned, and analyzed for Sr 90 and Pu. These measurements were taken in three<br />
parks in Kiev proper. Leningradskaya Square is in the center of downtown Kiev, and the hydropark lies<br />
along the Dneiper River to the east of downtown.<br />
11
Finally, Tables 1-14 and 1-15 (Baryahtar & Bobyleva, 1991) reflect the Ukrainians' concern with<br />
individuals growing, harvesting, and consuming food from contaminated areas.<br />
Blackberries, mushrooms, and medical herbs were chosen for analysis because they are both plentiful<br />
and susceptible to the effects of radioactive contamination. The transition coefficient presented in these<br />
tables was designed to give a "density rating" by converting measurements taken in square meters to<br />
more useful kilograms. The specific activity for CS 137 is listed in becquerels per kilogram, with a level of<br />
confidence as shown.<br />
Table 1-12. Contamination of forests by CS137.<br />
Regio Year Total Levels of Contamination CiIkm~)<br />
n area Studied
Table 1-13. Concentration ofSro and Pu in 1991 Kiev foliage.<br />
Location Weight of Weight of Sro Pu<br />
dry sample ash (Bk/kg) (BkIkg)<br />
Leningradskaya square 987 gr 120 gr 2.5-22.9 0.03-0.10<br />
Lesnoy district 638 gr 67 gr 4.6-49.4 0.05-0.22<br />
Hydropark 836 gr 90 gr - 0.05-0.39<br />
Note: Free-in-Air: Pu = 10-70 Bklm 3 or 10- 13 to 10- 12 Ci/liter<br />
Table 1-14. Contamination OfCS 137 in berries, mushrooms, and medical herbs.<br />
Contamination in Ci/km 2<br />
Table 1-15. Contamination of CS137 in wood.<br />
Contamination in Ci/km l<br />
Prior to 1991 there were no cases of internal organ cancers among children, yet today it ranks as the<br />
second largest cause of infant mortality. In 1991, 37.1 % of newborns evidenced some kind of pathology<br />
and in Kiev there were 2,500 premature deliveries. Anemia increased four times among pregnant<br />
women post-<strong>Chernobyl</strong>.<br />
According to Ukrainian law, individuals who were employed at <strong>Chernobyl</strong> and those involved in the<br />
clean-up efforts were divided into five categories depending on the amount of radiation absorbed.<br />
Category I individuals were assigned to <strong>Chernobyl</strong> when the accident occurred and received at least 25<br />
Gy of radiation. Category II were men who were assigned clean-up duties who also received 25 Gy or<br />
more of radiation. Category III were individuals who received between 10 and 24.9 Gy of radiation.<br />
Category IV were individuals who received between 5.0 and 9.9 Gy, and Category V were those<br />
individuals who received between 0.1 and 4.9 Gy. One Gy is equivalent to 100 rads.<br />
According to the Office of <strong>Chernobyl</strong> Affairs in Kiev, the new disease rates for individuals who received<br />
at least 25 Gy has doubled over those receiving less dosages. The relationship between absorbed dose<br />
and effect has only been investigated since 1990. In 1990, it was noted that increased oncological rates<br />
were attributed to men who received 25 Gy or more and to women who received lOGy or more.<br />
Generally, endocrine disease rates among male clean-up crew workers increased almost 4 times the 1988<br />
rate. The endocrine disease rate among women who received lOGy or more has doubled each year<br />
between 1989 to 1990. Category II personnel were 80 - 100% more likely to suffer from digestive and<br />
nervous system diseases than individuals in other categories. In the Kiev, Zhytomir, and Chernigov<br />
regions the rates for newly acquired diseases for Category V are 26.2% hematic system, 18.2%<br />
respiratory system, and 12.6% nervous and digestive system (Baryahtar & Bobyleva, 1991). The<br />
maximum disease rates are in Ivankov, Polissia, Narodichi, and Ovruch. In the Rovno region, the lowest<br />
newly acquired disease rates were reported among children born in 1984 and 1985. Disease rates for<br />
children born after the <strong>Chernobyl</strong> accident were 1.5 to 3.0 times higher than those born prior. In the<br />
same region, respiratory illness among children accounts for 25 - 40% of all new illnesses, especially<br />
among children 1 - 3 years of age where they present with respiratory problems 8 - 10 times per year.<br />
The other sixty percent of children present with thyroid gland problems. Anemia has increased among<br />
children 2.5 to 3.2 times between 1985 and 1988. .<br />
Ukrainian law also divides people who suffered from <strong>Chernobyl</strong> into another five-group categorization.<br />
Group A includes 5,237 disabled individuals, 187 people diagnosed with acute radiation symptoms, and<br />
15,000 people who suffered diseases directly attributed to the <strong>Chernobyl</strong> accident. Group B includes<br />
180,000 personnel who took part in the clean-up efforts, 130,000 people who received doses in excess of<br />
250 mSv and were relocated from areas inside the 30 kilometer zone, and 12,000 children born to<br />
parents involved in the clean-up effort. Group C includes children who have thyroid gland radiation in<br />
excess of allowable standards, 60,000 people who took part in the clean-up effort from 1988 until 1990<br />
who received less dosages than those who were first on the scene, and approximately one million people<br />
who currently live in contaminated areas but who await relocation. Among them are 350,000 children,<br />
65,000 of whom were born after the accident. Group D includes 1.5 million persons who work or live<br />
permanently in areas still receiving radio-ecological monitoring, in addition to their 400,000 children.<br />
Some government figures include in Group D all the inhabitants of the Kiev, Chernigov, and Zhytomir<br />
regions, or approximately another 4.5 million people.<br />
15
In 1990, the fifth group was added, comprised of women, when it was realized they showed the highest<br />
rates for new respiratory and digestive systems diseases for ages of30 - 39. Today, the Ukrainian<br />
registry contains the names of347,619 civilians who suffered direct medical problems as a result of<br />
Chemobyl, plus 36,000 military personnel who were also affected.<br />
Table 1-16 (Awramenko, 1992) shows the newly acquired disease rates for relocated persons. These<br />
individuals lived in contaminated areas but were forced to relocate because of excessive levels of<br />
radiation. Most of these individuals belong to Group D and the majority of them have been relocated.<br />
Table 1-16. Disease rates for relocated individuals.<br />
Disease 1986 1990 Percent increase<br />
Heart and blood 0.74/1000 6.9411000 938%<br />
Endocrine, digestion, 12.67/1000 171.11/1000 1350%<br />
Immune<br />
Respiratory 23.6711000 136.6811000 577%<br />
Nervous 21.25/1000 106.28/1000 500%<br />
Table 1-17 (A wramenko, 1992) shows the disease rates for those people relocated to Kiev.<br />
Table 1-17. Disease rates among individuals relocated to Kiev.<br />
Disease 1986 1990 Percentage<br />
Increase<br />
Endocrine system 11.7911000 119.7311000 1016 %<br />
Respiratory system 26,8911000 163.1611000 607%<br />
16
Finally, Table 1-18 (Awramenko, 1992) shows the relative health of these four groups from 1988 to<br />
1991. Year-to-year percentages are decreasing due to survival rate.<br />
Table 1-18. Percentages of individuals who are considered healthy.<br />
Groups 1988 1989 1990 1991<br />
Group A 74.0 66.4 52.8 33.8<br />
Group B<br />
Adults 61.5 44.1 35.3 28.8<br />
Children - 43.9 35.2 29.1<br />
Group C<br />
Adults 35.4 35.4 26.0 31.7<br />
Children - 52.9 40.7 39.8<br />
GroupD<br />
(Only 400,000 children - 77.7 62.9 48.5<br />
were tested)<br />
As of January 1, 1992, 1,536,270 persons were registered by the Ukrainian government as having<br />
suffered medical problems as a result of<strong>Chernobyl</strong>. Among those were 350,225 children and 180,144<br />
personnel assigned clean-up duties after the <strong>Chernobyl</strong> accident (A wramenko, 1992).<br />
Seventy percent of workers in the Narodichi forest in the Zhytomir region received 0.44 rads per year of<br />
radiation, and 10% received more than 2.3 rads per year (Yakovlev, 1992). The highest content of CS137<br />
was absorbed by woodcutters and forestry workers in the Rovno region which contains two forests, the<br />
Vladimiretsky and Dubrovitsky. These workers received between 31.8 and 74.2 Bk/kg.<br />
Ukrainian law divides all forests into four categories, dependent on CS137 dose. Category I includes the<br />
Pollessky and Narodichi forests, where contamination exceeds 40 Cilkm? Category II contains the<br />
Dymer, Ivankov, Ovruch, Luginsk, and Slovechansky forests, where active monitoring shows<br />
contamination between 15 and 40 Ci/km 2 • Category III includes forests with contamination between 5<br />
and 15 Ci/km 2 • These forests require continuous monitoring. Category IV consists of forests containing<br />
no more than 5 Ci/km 2 • The Ministry of Forests in Kiev is concerned about forest workers who have<br />
monitoring and woodcutting duties in contaminated forests.<br />
In the 10 years following the <strong>Chernobyl</strong> nuclear accident (as of 1996), the number of healthy individuals<br />
living in contaminated areas decreased from 67.1 % to 33.1 %. Chronic pathologies increased from<br />
31.5% to 66.0%.<br />
17
Most cases reflect pathologies of the endocrine system, blood and blood-generating systems, nervous<br />
system, and gastroenterological system. The most substantial growth of this dangerous statistics is related<br />
to young people aged 15-17 years--6.6 times normal (8.1 times for boys and 5.6 times among girls).<br />
Death rates and disability rates have increased substantially (about 2.5-3.5 times), as compared to those<br />
of people who lived in normal conditions. Pathology of the thyroid gland constitutes 72.7% of all<br />
endocrine cases for women, and 62.5% for men.<br />
As of 1998, the official death rate among the "Eliminators" (who are still alive) is 1.8 (80% higher than<br />
normal). This is primarily due to higher incidences of cancer, diseases of blood and blood-generating<br />
organs, and pneumonia.<br />
The present study, an ongoing longitudinal project which commenced in 1995, entails assessments of<br />
neuropsychological and physical capabilities of four independent volunteer participant groups. The<br />
control group (Controls) consists of healthy volunteers that reside outside the immediate radiation<br />
exposure area. The second group consists of "Eliminators," who are individuals who were involved in<br />
the tasks of removing nuclear debris and assisting in construction of the containment chamber for the<br />
defective reactor facility. The third group of volunteers consists of Forestry workers who perform<br />
monitoring, woodcutting, and other related activities in the Narodichi forest, which is in close proximity<br />
of <strong>Chernobyl</strong>. It is known that 70% of workers in the Narodichi forest (in the Zhytomir region) received<br />
approximately 0.44 rads per year of radiation, and 10% received more than 2.3 rads per year (Yakovlev,<br />
1992). Finally, the fourth group is comprised of Agricultural workers from Rozumnytsia, which is<br />
approximately 150 km south of Kiev, and for whom knowledge of the level of radio nuclide is known.<br />
The instrument that was chosen to be the primary measure of cognitive performance is the Automated<br />
Neuropsychological Assessment Metrics (ANAM) Battery, which is a subset of the Office of Military<br />
Performance Assessment Technology (OMPAT) Tester's Workbench (TWB). The TWB is a library of<br />
automated tests that has been constructed to meet the need for precise measurements of cognitive<br />
processing efficiency. The ANAM batteries are unique combinations ofTWB tests that have been<br />
configured for neurocognitive assessment and evaluation of functioning in a variety of<br />
neuropsychological domains. Many of the component tests in ANAM were derived from the Unified<br />
Tri-Service Cognitive Performance Assessment Battery (UTCPAB; Reeves et al. 1991) and the Walter<br />
Reed Performance Assessment Battery (Thorne et al. 1985). The Ukrainian subset of ANAM<br />
(ANAMUKR) was designed by Reeves and Gamache (1994), and constitutes a specialized subset of the<br />
TWB-ANAM batteries. It consists of tests that have been configured for repeated measures testing for<br />
neurocognitive impairment due to exposure to radionuclides. It has been designed to assess levels of<br />
neurocognitive function ranging from superior to moderately impaired. ANAMUKR subtests also<br />
include a stand-alone module for assessing sustained attention (Running Memory Continuous<br />
Performance Test).<br />
In the present study, the ANAMUKR battery was combined with the Gamache Physical Abilities<br />
Battery (GP AB; Gamache, 1993), for testing the physical capabilities of individuals exposed to<br />
radionuclides. This battery, composed of tests derived from Fleishman and Quaintance (1984), is<br />
especially sensitive to the physical decrements in performance resulting from exposure to radionuclide<br />
contamination. The GP AB consists of tests designed to measure explosive strength (broadjump), static<br />
strength (carrying weight), dynamic strength (squat thrusts), and gross body equilibrium (balance beam).<br />
18
In this report, we describe data on the GP AB and ANAMUKR obtained from four independent groups<br />
from Ukraine in 1995 (initial test), and in 1996,1997, and 1998 (repeated tests). These data provide a<br />
reference point from which to gauge physical and cognitive performance of individuals who mayor may<br />
not have been exposed to varying levels of ionizing radiation resulting from the nuclear accident at<br />
Chemobyl in 1986.<br />
19
SECTION 2<br />
METHOD<br />
2.1 PARTICIPANTS.<br />
The participants in the initial phase of the study consisted of 127 volunteers (24 females, 103 males)<br />
who lived in Ukraine prior to 1986. Ages ranged from 11-61 years, averaging 40.21 years. The four<br />
groups into which they were divided included a non-exposed Control group (AC); and three exposed<br />
(exposure) groups: Eliminators (AE), Forestry Workers (AF), and Agricultural Workers (AG).<br />
Demographic information is presented in Table 2-1. Mean dose levels of exposure to radiation (in rads)<br />
for each group are included; these are based on the medical records of the individuals.<br />
Table 2-1. Demographic information and mean dose of radiation for the 4 groups -<br />
above background radiation.<br />
GROUP~ AC(n=31) AE(n=36) AF(n=29) AG (n=31)<br />
Age<br />
Mean (S.D.) 33.23 ( 7.86) 40.47 ( 6.81) 50.83 ( 7.83) 36.32 (14.26)<br />
Gender<br />
Male 24 33 29 17<br />
Female 7 3 0 14<br />
Mean dose<br />
In rads 0 62.95 12.61 8.81<br />
(FIA)<br />
These individuals were randomly assigned to their groups, which were counterbalanced by occupation.<br />
Further, participants in the control group were assigned to match by occupation, as closely as possible,<br />
the exposure participants. For example, if there was a truck driver in any of the three exposure groups, a<br />
truck driver was sought as a control. In addition, participants in the control group were matched for age<br />
and gender to those in the exposure groups.<br />
20
2.2. INSTRUMENTS.<br />
2.2.1. Gamache Physical Abilities Battery (GP AB).<br />
Physical testing involved two stations, each equipped with a stop watch and tape measure. Three tests<br />
were timed with a stopwatch: balance beam, squat thrusts, and carrying weights. A description of each<br />
test in the GPAB is presented in Table 2-2.<br />
Table 2-2. Description of Gamache Physical Abilities Battery (GP AB).<br />
Test<br />
Broad jump<br />
(BROADJMP)<br />
Carrying weight<br />
(CARRYWGT)<br />
Squat thrusts<br />
(SQUATTHR)<br />
Balance beam<br />
(BALBEAM)<br />
Measurement<br />
Distance covered in one broad<br />
jump (meter)<br />
Distance covered in 30 sec.<br />
(meter)<br />
Number of squat thrusts in 2<br />
mm.<br />
Distance covered in 20 sec.<br />
(meter)<br />
Apparatus<br />
Designated starting point.<br />
Tape measure.<br />
10-meter course. participants run<br />
back and forth in straight lines<br />
carrying 15 kilograms of sand<br />
(men) or 10 kilograms (women,<br />
children). Tape measure, stop<br />
watch, buckets of sand<br />
Any area where squat thrusts can<br />
be don. stop watch<br />
4-meter board, 12 centimeters<br />
wide, 15 centimeters from<br />
ground. Tape measure, stop<br />
watch.<br />
All physical test scores were recorded in laboratory notebooks. Two observers participated as test<br />
administrators for the physical abilities battery. One administrator read instructions to participants and<br />
ensured their understanding. The other timed and/or took measurements as appropriate for each test.<br />
Independent observer confirmation was required prior to recording scores. All subjects were tested<br />
according to standard procedures, with one exception. The hospital where the eliminators resided did<br />
not allow buckets of sand on the premises. Therefore, hand-held weights equivalent to the weight<br />
carried by other groups were substituted.<br />
2.2.2 Automated Neuropsychological Assessment Battery-Ukraine (ANAMUKR).<br />
The tests in the ANAMUKR battery includes the Stanford Sleepiness Scale (SLP), Code Substitution<br />
(visual search, immediate recall, and delayed recall: CDS, CDI, CDD), Running Memory Continuous<br />
Performance Task (CPT), Digit Symbol (DGS), Matching to Sample (MSP), Spatial Processing (SPD),<br />
Simple Reaction Time (SRT), Tapping-Right and Left Index Fingers (TAPR & TAPL), and Twochoice<br />
Reaction Time (2CH). These subtests of ANAM have been described previously (Reeves &<br />
Winter,1992; Levinson & Reeves, 1994). Each session required approximately 60 minutes.<br />
21
2.3 ASSESSMENT SITES AND ENVIRONMENTS.<br />
Participants in the Control group resided in Temopil (pop. 250,000). This city is located approximately<br />
450 km west of Kiev, the capital of Ukraine. All testing was conducted in High School Number 22 (Fig.<br />
B-7). The Eliminators were tested in the Ukrainian Center for the Radiation Protection of the Population,<br />
which is essentially a hospital environment (Fig. B-3). This special hospital is in the suburbs of Kiev,<br />
and was established to attend to the medical needs of these individuals. The Forestry Workers were<br />
tested in the Ovruch forest, approximately 250km northwest of Kiev. All testing was conducted in their<br />
barracks (see Fig. B-8). The Agricultural workers were tested in the village ofRozumnytsia, in the Kiev<br />
region, approximately 150 km south of Kiev. The testing site was a farmhouse (Figs B-4, B-5, B-6).<br />
2.4 PROCEDURE.<br />
On days assigned to specific groups, researchers prepared the testing site by installing laptop computers<br />
for administration of the ANAMUKR battery (Fig. B-4), and by setting up the apparatus for<br />
administration of the physical abilities test battery. A table and two chairs were situated at each of three<br />
ANAMUKR testing stations. Each participant sat in one chair and the test administrator sat in the other.<br />
This enabled the administrator to ensure that the participant understood instructions and was prepared<br />
for testing. Participants were instructed to ask as many questions as necessary to ensure full<br />
understanding prior to testing. One table was placed at the entrance for participant registration and<br />
orientation, which included having each participant read and sign an informed Consent Form in Russian<br />
and English (see Appendix C -only the English version is shown).<br />
During periods when all computer test stations were occupied, participants first completed the GP AB.<br />
The rotation of physical testing (i.e., order of balance beam, squat thrusts, etc.) was randomized to the<br />
extent that space was available. A rest area was established for participants while awaiting further<br />
testing. During test sessions the administrator ensured that there was no discussion among participants<br />
about up-coming tests. At the conclusion of the testing sessions, all participants were thanked and<br />
given the equivalent of two USA dollars.<br />
The ANAM battery test scores were stored on the hard disk drive of each computer. At the end of the<br />
day, all scores were copied to backup 3 Yz" floppy disks and marked for that group and year of testing.<br />
The backup disks were then re-copied to a second diskette.<br />
Participants in the four groups were tested on the GPAB and ANAMUKR in 1995, 1996, 1997, and<br />
again in 1998. All 1995 measures were deemed valid, as were the 1996 GPAB data obtained on the<br />
Controls.<br />
Due to extraneous factors, a major portion of the 1996 ANAMUKR data obtained from the Controls was<br />
not valid. As a result, these data were not included in analyses of the 1996 ANAMUKR results.<br />
Instead, all analyses of the 1996 ANAMUKR performances of the exposure groups relative to the<br />
Controls were based on the 1995 Control data. Further, the 1996 GPAB data from the Controls were<br />
virtually identical to those obtained in 1995.<br />
22
Therefore, the same procedure was used in analyses of the 1996 OP AB performances of the exposure<br />
groups relative to those of the Controls; i.e., 1995 Control data were also used for these comparisons.<br />
For the same reasons, similar analyses were performed on the 1997 and 1998 data. However, for<br />
purposes of data analyses based on 4-year averages, all data from all groups were used.<br />
23
SECTION 3<br />
RESULTS<br />
3.1 OVERVIEW OF 4-YEAR RESULTS: 1995-1998.<br />
The ANAM data files were first consolidated in a computerized spreadsheet and then inspected for<br />
completeness and invalid data. Invalid data were defined as "premature responses" which occurred in<br />
less than 100 ms, and/or an inordinate number of lapses which indicated that the participant did not<br />
understand the instructions for a test prior to administration. This initial screening resulted in unequal<br />
numbers of participants associated with each test, but it ensured that the preliminary data presented<br />
herein were derived from complete and valid test administrations.<br />
An ANOV A revealed a significant age difference among the groups. Subsequent Scheffe tests<br />
indicated that this was due to the higher ages of Group AF, as this group differed significantly from each<br />
of the others. No other significant differences in age were observed. Gender composition of the four<br />
groups also significantly differed, as revealed by results from a Chi-square test. This is most likely a<br />
result of the higher number of females in GroupAG, as none of the other groups differed from each<br />
other: all were predominantly male. Possible differences between the ACs and AEs on the demographic<br />
variables were of particular concern, but none were revealed.<br />
Analyses of 4-year averages for all groups on all measures were performed so as to obtain an overview<br />
of how the exposure groups performed relative to the controls on the physical and cognitive measures.<br />
Multivariate analyses of variance (MANOVAs) revealed significant differences among the 4 groups on<br />
all measures; further, pairwise comparisons indicated that with only a few exceptions, the average levels<br />
of performance of the exposure groups were significantly lower than those of the controls.<br />
Graphic illustrations of actual performance levels of the exposure groups on each task across the 4 years<br />
are presented in Figures 3-30 through 3-53, in the section describing the 1998 retest. In each figure, the<br />
4-year averaged performance level of the controls is used as a referent; it is denoted by a dotted line<br />
(typically across the top) on each figure.<br />
3.1.1 GPAB.<br />
The 4-year averages for the 4 groups on the physical tasks are presented in Table 3-1. The difference<br />
among the groups on BRODJMP was significant at the .01 level, while the other differences were<br />
significant at the .001 level. Pairwise comparisons of the 4-year averages revealed that with only one<br />
exception, all exposure groups performed at significantly lower levels than the controls. The 4-year<br />
averages for all groups on all GP AB tasks are graphically illustrated in Figures 3-1 through 3-4.<br />
24
Table 3-1. Four-year averaged performance on GPAB: physical tasks.<br />
GROUP~ AC AE AF AG<br />
TASK<br />
BRODJMP (meters) 1.59 1.33 1.43 1.50*<br />
CARRYWGT (meters) 51.61 30.96 40.45 42.51<br />
SQUATTHR (number) 66.66 21.24 40.35 48.41<br />
BALBEAM (meters) 22.19 15.96 20.26 19.13<br />
Note: *not significantly lower than controls.<br />
1.6<br />
en 1.5<br />
a::<br />
w<br />
I- w .AC<br />
:E 1.4<br />
z<br />
.AE<br />
~<br />
w<br />
:E 1.3<br />
.AF<br />
DAG<br />
1.2<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-1. 4-year averaged performances on GPAB: BROADJUMP.<br />
25
en<br />
a::<br />
w<br />
I W<br />
:E<br />
z<br />
«<br />
w<br />
:E<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-2. 4-year averaged performances on GPAB: CARRYING WEIGHT.<br />
a:<br />
w<br />
m<br />
70<br />
60<br />
== ::;) 40<br />
Z<br />
Z 30<br />
en<br />
a:<br />
60<br />
50<br />
w 40<br />
I-<br />
w<br />
.AC<br />
:aE 30<br />
z<br />
.AE<br />
«<br />
w 20 .AF<br />
:aE<br />
DAG<br />
10<br />
0<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-4. 4-year averaged performances on GPAB: BALANCE BEAM.<br />
3.1.2 ANAMUKR: Accuracy.<br />
The 4-year averages for the 4 groups on ANAMUKR-accuracy are presented in Table 3-2. Table 3-2<br />
represents the percentage of correct responses. All differences among groups were significant at the .001<br />
level. Pairwise comparisons of the 4-year averages revealed that, with only 3 exceptions, the exposure<br />
groups performed at significantly (most at .001) lower levels than the controls. Figures 3-5 through 3-<br />
12 graphically illustrate these findings.<br />
~.1.3 ANAMUKR: Efficiency.<br />
The 4-year averages for the 4 groups on ANAMUKR-efficiency are presented in Table 3-3. Table 3-3<br />
represents correct responses per minute. All differences among groups were significant at the .001 level.<br />
Pairwise comparisons of the 4-year averages revealed that, with only 1 exception, the exposure groups<br />
performed at significantly (most at .001) lower levels than the controls. Figures 3-13 through 3-21<br />
graphically illustrate these findings.<br />
27
3.1.4 ANAMUKR: Additional Measures.<br />
Performances averaged over 4 years on the tapping task (right, left), and on the sleepiness scale are<br />
presented in Table 3-4. Although all group differences were significant at the .001 level, the levels of<br />
the agricultural workers did not differ from those of the controls on any of these measures. The tapping<br />
rates for the other exposure groups were significantly (.001) lower that those of the controls, and levels<br />
of sleepiness were significantly higher for the Eliminators. These findings are graphically illustrated in<br />
Figures 3-22 through 3-24.<br />
Table 3-2. 4-Year averaged performances on ANAMUKR: Accuracy (Percentage of correct<br />
responses).<br />
GROUP -7 AC AE AF AG<br />
TASK<br />
2CH 97.36 92.76 92.91 93.98<br />
CDS 96.17 91.14 90.66 95.18*<br />
eDI 91.42 73.50 80.49 89.01 *<br />
-- --<br />
eDD 89.24 72.35 78.99 85.32<br />
-- --<br />
CPT 93.66 77.06 86.08 89.72<br />
-- --<br />
DGS 87.96 73.00 81.32 80.40<br />
--<br />
MSP 92.88 75.33 84.22 87.68<br />
--<br />
SPD 88.21 81.18 83.96 91.94*<br />
Note: *not significantly lower than controls. Italicized, bolded, and underlined numbers are<br />
considered to suggest clinically meaningful impairment.<br />
28
Table 3-3. 4-Year averaged performances on ANAMUKR: Efficiency (Correct responses<br />
per minute).<br />
GROUP~ AC AE AF AG<br />
TASK<br />
SRT 163.51 111.58 111.67 147.24<br />
2CH 110.62 76.42 80.95 101.27<br />
-- --<br />
CDS 39.58 24.22 26.12 34.33<br />
-- --<br />
CDI 36.42 18.76 24.07 29.46<br />
-- --<br />
CDD 38.19 20.77 26.46 31.57<br />
-- --<br />
CPT 85.22 58.22 71.69 79.00<br />
--<br />
DGS 36.19 24.28 30.54 29.30<br />
--<br />
MSP 38.63 20.23 24.56 30.31<br />
-- --<br />
SPD 27.67 19.11 22.81 26.38*<br />
--<br />
Note: *not significantly lower than controls. Italicized, bolded, and underlined numbers are<br />
considered to suggest clinically meaningful impairment.<br />
Table 3-4. 4-Year averaged performances on ANAMUKR: Additional measures.<br />
GROUP~ AC AE AF AG<br />
TASK<br />
TAP-R (Mean number of 57.57<br />
--<br />
47.00 49.20 55.23*<br />
--<br />
responses in 10 seconds)<br />
TAP-L (Mean number of 51.50 41.23 42.63 48.70*<br />
-- --<br />
responses in 10 seconds)<br />
SLP** (Scores from 1-7) 1.63 2.46 1.77* 1.69*<br />
--<br />
Note: *not significantly different than controls.<br />
**higher score = more sleepy. Italicized, bolded, and underlined numbers are<br />
considered to suggest clinically meaningful impairment.<br />
29
3.1.5 A Clinical Neuropsychological Interpretation of <strong>Chernobyl</strong>-ANAM data.<br />
Overall results from the 4-year averages of ANAM test results, presented in Tables 3-2 through 3-4,<br />
provide evidence of clinically meaningful neurocognitive impairment associated with the Eliminator and<br />
Forestry groups. The Agricultural group was generally comparable to the Control group. Their test<br />
performance did not reveal any meaningful evidence of neuropsychological impairment, and scores were<br />
generally within normal limits.<br />
With respect to the Eliminator and Forestry groups, it appears that they have clinically significant<br />
neuropsychological deficits in several areas. These include deficits in the ability to sustain high levels<br />
of attention/concentration, to encode new information (i.e., learning ability), working memory (i.e., the<br />
ability to hold new information in memory for short periods of time), mental flexibility (i.e., the ability<br />
to shift mental sets rapidly), and psychomotor speed. The Eliminators are suffering the most severe<br />
impairment of neurocognitive and psychomotor abilities. Their test scores revealed impairment of<br />
mental power (the ability to produce correct responses to test items) and neurocognitive efficiency (the<br />
ability to do both quickly and accurately). In this study mental power was indexed by using percent<br />
accuracy scores, which are presented in Table 3-2. Neurocognitive efficiency was indexed by the<br />
ANAM "thruput" score, which literally translates to "number of correct responses per minute." These<br />
data are presented in Table 3-3. Psychomotor speed data are presented in Table 3-4. These data<br />
represent the average number of "Finger Taps" a subject is able to make on a mouse key during 5<br />
consecutive 10 second response trials. Finally, each subject's fatigue level was measured by a<br />
Sleepscale score that ranges from 1 =very alert and energetic to 7=sleep onset soon. A summary of<br />
clinically meaningful deficits for the eliminators is presented below.<br />
3.1.5.1 Learning and Memory. The Code Substitution, Learning, Immediate, and Delayed recall<br />
subtests were the primary instruments for assessing a subject's ability to learn and retain new<br />
information. This test entails having the subjects learn 8 pairs of associated digits and symbols.<br />
Following the learning trial, subjects are required to demonstrate the ability to remember the correct<br />
pairings immediately and then after a delayed time interval of approximately 20 minutes. Memory<br />
scores for immediate and delayed recall trials for Eliminators were 73 & 72% respectively as compared<br />
to 91 & 89% for the controls. These results suggest that the Eliminators have an impaired ability to<br />
encode, i.e., learn and store new information in short-term memory (e.g., CDI=73%). Although their<br />
learning ability is impaired, they are still able to learn and store new information to a certain degree.<br />
Further, they are able to retain and retrieve (i.e., remember) the newly learned information over<br />
meaningful time intervals. This is indicated by a delayed recall score (i.e., CDD=72%) that nearly<br />
matches the immediate recall score (i.e., CDI=73%). These results indicate that they do NOT have a<br />
rapid rate of forgetting over retention intervals, as would be the case in Alzheimer's and Alcohol<br />
Korsakoff s Dementias. The implication is that observed impairments in neurocognitive and memory<br />
processes in this sample are NOT a result of chronic alcohol abuse or an Alzheimer's-like eNS<br />
disorders.<br />
3.1.5.2 Attention/Working Memory. The Digit Set Comparison Test (DGS) was the primary<br />
ANAM subtest used to assess attention and working memory. It is comparable to the traditional<br />
Wechsler Adult Intelligence Scale Digit Span subtest. The DGS requires the subject to remember a<br />
series of numbers for a few seconds and then determine if a comparison sample is the same or different.<br />
30
Results from this test indicated that the Eliminators were meaningfully impaired with respect to both<br />
percent accuracy (power) and efficiency measures. For example, the Eliminators had an averaged<br />
accuracy score of73% as compared to a Control group score of 87%. Further, the Eliminators had an<br />
averaged efficiency score of24% vs. a score of36% for the Controls. These results suggest that the<br />
Eliminators are experiencing significant difficulties with the ability to attend to and retain information<br />
that is not personally meaningful; even for brief periods of time.<br />
3.1.5.3 Mental FlexibilitylExecutive Function. The Continuous Performance Task was the primary<br />
ANAM subtest used for assessing the ability to sustain high levels of concentration and rapidly shift<br />
mental sets. These are important "executive" functions that relate to frontal lobe functions. The test<br />
requires the subject to rapidly determine if a letter that is displayed on the computer screen is the "same<br />
or different" from the letter presented immediately before. Results from this test indicate that the<br />
Eliminators have serious deficits regarding the ability to sustain concentration and exercise mental<br />
flexibility. For example, the Eliminator's accuracy scores were 77% vs. Control's score of93%.<br />
Further, the Eliminators' efficiency scores were 52 vs. the Control's scores of85.<br />
3.1.5.4 Level of Subjective Energy. The ANAM Sleep Scale was used to assess the subjects' level of<br />
fatigue ... i.e., how energized or tired did they feel on a 1-7 scale. The results indicate that the<br />
Eliminators felt slightly more tired than the other groups ... however, the difference was barely 1 point<br />
higher (i.e., the Eliminator average was 2.46). This means that they did not really feel tired. This<br />
suggests that observed attention and memory deficits were not due to fatigue.<br />
3.1.5.5 Psychomotor Ability. The ANAM Finger Tapping Test was implemented as the primary test of<br />
psychomotor speed. The test requires the subject to "tap" on a mouse key as fast as possible during ten<br />
second test trials. The outcome measure is the average number of taps for a ten-second interval. The<br />
ANAM test results on this subtest revealed that the Eliminators were much slower that the other groups.<br />
For example, their averaged tapping scores for RT and LT index finger tapping were 47 & 41 vs. the<br />
Control group scores of 57 & 51. Since their subjective level of fatigue was minimal, results suggest<br />
that this psychomotor slowing was not due to being tired.<br />
3.1.5.6 Final Conclusion. The overall results that include loss of mental power and cognitive efficiency<br />
and psychomotor slowing strongly suggest impaired brain function. The pattern of impairment is similar<br />
to that commonly associated with white matter disease (white matter disease effects the myelin sheaths<br />
as a part of neurological functioning).<br />
3.2 CORRELATIONS BETWEEN DOSAGE OF RADIATION AND 4-YEAR AVERAGED<br />
PERFORMANCE LEVELS.<br />
For each participant in the study, the level of radiation exposure was obtained from medical records. The<br />
original dosage presented on Table 2-1, on page 21, included all subjects. However, our statement of<br />
work specifies that the government is interested in low dosage defined as below 70 rads. Therefore we<br />
eliminated from Table 2-1 all dosages higher than 70 rads. The following reflects only those dosages<br />
less than 70 rads. Mean dosage (and standard deviation) in rads for the four groups were as follows: AC:<br />
0.00 (.OO);AE: 43.41 (19.82);AF: 12.61 (2.10); andAG: 8.81 (5.63).<br />
31
Since the Controls received virtually no radiation, correlations were based only on participants in the 3<br />
combined exposed groups for whom measures on all tasks were obtained for all 4 years of testing,<br />
excluding Eliminators receiving over 70 rads. The 4-year average for each exposed individual on each<br />
measure was calculated, and these were used to obtain Pearson correlations between each measure and<br />
dosage of radiation. In addition, standard multiple regressions were used to calculate the correlations<br />
between combined groups of tasks and dosage. The results of these analyses are presented in Table 3-5.<br />
Table 3-5. Correlations between dosage of radiation and performance levels.<br />
TASK CORRELATION SIGNIFICANCE<br />
GPAB .70 .001<br />
BRODJMP -.18 --<br />
CARRYWGT -.34 .01<br />
SQUATTHR -.55 .001<br />
BALBEAM -.56 .001<br />
ANAMUKR:ACCURACY .71 .001<br />
2CH -.07 --<br />
CDS -.38 .01<br />
CDI -.62 .001<br />
CDD -.56 .001<br />
CPT -.47 .001<br />
DGS -.35 .01<br />
MSP -.54 .001<br />
SPD -.52 .001<br />
ANAMUKR: EFFICIENCY .68 .001<br />
SRT -.25 --<br />
2CH -.44 .001<br />
CDS -.37 .01<br />
CDI -.50 .001<br />
CDD -.50 .001<br />
CPT -.48 .001<br />
DGS -.49 .001<br />
MSP -.45 .001<br />
SPD -.51 .001<br />
ANAMUKR:OTHERTASKS<br />
TPR -.34 .01<br />
TPL -.37 .01<br />
SLP .60 .001<br />
Significant correlations were revealed for 21 of the 24 measures. The only tasks not significantly<br />
correlated with dosage were broad jump, simple reaction time, and 2-choice accuracy.<br />
32
Unlike the other GP AB tasks, broadjump does not require sustained energy; therefore, it is not surprising<br />
that it is not related to dose level. The two ANAMUKR measures not related to dose are the simplest of<br />
all the tasks, requiring little effort to complete. Thus one would not expect them to be related to levels<br />
of radiation. Although tapping is also relatively simplistic, it requires fine motor coordination. Such<br />
coordination is reflective not only of integrity of the cerebral precentral gyri, but also of the cerebellum.<br />
Involvement of either of these areas would compromise the ability to perform this task.<br />
33
I<br />
U<br />
w<br />
a:<br />
a:<br />
o<br />
u<br />
eft.<br />
z<br />
«<br />
w<br />
~<br />
lilAC<br />
.AE<br />
EAF<br />
DAG<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-5. 4-year averaged performances on ANAMUKR: 2CH-ACC.<br />
I<br />
U<br />
w<br />
a:<br />
a:<br />
o<br />
u<br />
eft.<br />
z<br />
«<br />
w<br />
~<br />
iliAC<br />
.AE<br />
rmAF<br />
DAG<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-6. 4-year averaged performances on ANAMUKR: CDS-ACC.<br />
34
95<br />
I- 0 90<br />
w<br />
a:<br />
a: 85<br />
0<br />
.AC<br />
0 ~<br />
80 .AE<br />
Z<br />
.AF<br />
«<br />
w 75<br />
DAG<br />
:l5<br />
70<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-7. 4-year averaged performances on ANAMUKR: CDI-ACC.<br />
90<br />
I- 0<br />
W 85<br />
a:<br />
a:<br />
0<br />
iliAC<br />
80<br />
0 ~<br />
.AE<br />
Z<br />
.AF<br />
«<br />
w 75<br />
DAG<br />
:l5<br />
70<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-8. 4-year averaged performances on ANAMUKR: CDD-ACe.<br />
35
....<br />
o<br />
w<br />
c::<br />
c::<br />
o<br />
o<br />
'ifz<br />
«<br />
w<br />
::2E<br />
I1AC<br />
IIAE<br />
I!!I AF<br />
DAG<br />
AC AE AF<br />
GROUP<br />
AG<br />
Figure 3-11. 4-year averaged performances on ANAMUKR: MSP-ACC.<br />
95<br />
90<br />
85<br />
80<br />
75<br />
&lAC<br />
I!IAE<br />
IIAF<br />
DAG<br />
70<br />
AC AE AF<br />
GROUP<br />
AG<br />
Figure 3-12. 4-year averaged performances on ANAMUKR: SPD-ACC.<br />
36
170<br />
c 160 . Ē<br />
....... 150 a:<br />
a:<br />
0 140 .AC<br />
()<br />
=1:1: 130 .AE<br />
Z<br />
« 120<br />
IIAF<br />
w<br />
DAG<br />
~ 110<br />
100<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-13. 4-year averaged performances on ANAMUKR: SRT -EFF.<br />
c<br />
E<br />
a:<br />
a:<br />
o<br />
()<br />
=1:1:<br />
Z<br />
«<br />
W<br />
~<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-14. 4-year averaged performances on ANAMUKR: 2CH-EFF.<br />
37
40<br />
t: 35<br />
. Ē 30<br />
......<br />
a::<br />
a:: 25<br />
0<br />
20<br />
iliAC<br />
:#: .AE<br />
Z<br />
15<br />
Pfl AF<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-17. 4-year averaged performances on ANAMUKR: CDD-EFF.<br />
c .-<br />
.E<br />
CC<br />
CC<br />
o<br />
:t:I::<br />
Z<br />
40<br />
c 35<br />
. Ē '- 30<br />
a:<br />
a: 25<br />
0<br />
(.) 20<br />
:t:I:<br />
Z<br />
15<br />
« 10<br />
w<br />
~<br />
5<br />
0<br />
AC AE AF AG<br />
GROUP<br />
lilAC<br />
I!!!IAE<br />
fmAF<br />
DAG<br />
Figure 3-19. 4-year averaged performances on ANAMUKR: DGS-EFF.<br />
c<br />
E<br />
a:<br />
a:<br />
o<br />
(.)<br />
:t:I:<br />
Z<br />
«<br />
W<br />
~<br />
iliAC<br />
.AE<br />
I1AF<br />
DAG<br />
AC AE AF<br />
GROUP<br />
AG<br />
Figure 3-20. 4-year averaged performances on ANAMUKR: MSP-EFF.<br />
40
30<br />
c 25<br />
E<br />
-a:: 20<br />
a:<br />
0 .AC<br />
0<br />
=I*:<br />
.AE<br />
Z 10<br />
«<br />
.AF<br />
w<br />
5<br />
DAG<br />
==<br />
0<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-21. 4-year averaged performances on ANAMUKR: SPD-EFF.<br />
iliAC<br />
II AE<br />
.AF<br />
DAG<br />
AC AE AF<br />
GROUP<br />
AG<br />
Figure 3-22. 4-year averaged performances on ANAMUKR: TAPPING-RIGHT.<br />
41
en<br />
Q.<br />
«<br />
l-<br />
#:<br />
Z<br />
«<br />
w<br />
:E<br />
liAC<br />
.AE<br />
IIAF<br />
DAG<br />
AC AE AF AG<br />
GROUP<br />
Figure 3-23. 4-year averaged performance on ANAMUKR: TAPPING LEFT.<br />
en<br />
w<br />
a:<br />
o<br />
en<br />
z<br />
«<br />
w<br />
:E<br />
iliAC<br />
.AE<br />
IlAF<br />
DAG<br />
AC AE AF<br />
GROUP<br />
AG<br />
Figure 3-24. 4-year averaged scores on ANAMUKR: SLEEP SCALE.<br />
42
3.3 RESULTS OF 1995 INITIAL TEST SESSION.<br />
Because perfonnance on the tasks within and among the test batteries is indexed in a variety of different<br />
ways, a measure was required which would describe comparisons between the exposure groups and the<br />
control group in similar tenns across the tasks and test batteries. This measure consisted of determining<br />
the proportion of the control group's mean on a given task equivalent to an exposure group's mean on<br />
that task (i.e. mean-E / mean-C). The complement of this proportion (I-prop.) reflects the proportional<br />
difference between the means of the two groups on the task, and when multiplied by 100 it reflects the<br />
percent difference between the means of the two groups. The mean % difference for a given test battery<br />
was then calculated by obtaining the mean of the % differences for all the tasks in that battery. In all four<br />
cases the mean of a given exposure group on any given task was either equal to, or (most typically) less,<br />
than that of the control group. Therefore, the mean differences of all exposure groups on all test<br />
batteries reflected decrements in perfonnance, as compared to the control group. Thus it seemed<br />
appropriate to describe the composite results of the 1995 test session in tenns of mean % decrements as<br />
shown by the exposure groups relative to the control group, on all test batteries. These are presented in<br />
Table 3-6 and graphically in Figure 3-25.<br />
Table 3-6. Mean % performance decrement for exposure groups relative to<br />
controls-1995.<br />
GROUP<br />
GPAB<br />
ANAMUKR-ACC<br />
ANAMUKR-EFF<br />
AE<br />
AF<br />
AG<br />
27.00<br />
16.88<br />
12.49<br />
11.00<br />
7.35<br />
3.09<br />
40.52<br />
23.91<br />
22.51<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
o +==~_--lio<br />
GPAB<br />
ANAM-ACC<br />
ANAM-EFF<br />
DAG<br />
IIAF<br />
.AE<br />
GROUP<br />
Figure 3-25. Mean % Decrement for Exposure Groups Relative to Controls-1995.<br />
43
All of these decrements were significant, as evidenced by the results of multivariate analyses of variance<br />
(MANOVAs) described later. As dramatically illustrated in Figure 3- 25, the Eliminators were most<br />
drastically affected by their exposure to the radiation in the power station, on all test batteries. The other<br />
groups were also affected (although to a lesser degree) by the radiation in the Chemobyl region and<br />
showed sizeable decrements in performance on the test batteries as well.<br />
3.3.1 GPAB.<br />
Means and standard deviations for the groups on the GPAB are presented in Table 3-7.<br />
Table 3-7. GPAB: 1995 Means (and standard deviations) for the four groups.<br />
GROUP~ AC AE AF AG<br />
TASK<br />
BROADJMP 1.57 (.18) 1.40 (.17) 1.39 (.16) 1.43 (.45)<br />
(meter)<br />
CARRYWGT 43.42 (3.85) 36.61 (7.24) 37.17 (6.09) 40.58 (5.57)<br />
(meter)<br />
SQUATTHR 62.65 (10.01) 22.42 (8.78) 35.52 (12.72) 44.10 (17.39)<br />
(number)<br />
BALBEAM 19.61 (1.19) 16.22 (2.30) 19.93 (1.51) 18.65 (2.22)<br />
(meter)<br />
The results of a MANOV A (Table 3-28) revealed a significant difference on the composite<br />
MEASURES. Univariate ANOVAs and post-hoc Dunnett tests, the results of which are presented in<br />
Tables 3-8 and 3-9, indicated that theAEs were significantly impaired on all 4 tasks as compared to the<br />
ACs. TheAFs were significantly lower than theACs on BROADJMP, CARRYWGT and<br />
SQUATTHR, while theAGs were significantly lower on SQUATTHR only. Interestingly, theAFs<br />
performed somewhat better than the A Cs on BALBEAM; this may well have resulted from their<br />
training in forestry work, including balancing on logs.<br />
44
Table 3-8. GP AB: Results of MANOV A and univariate ANOV AS.<br />
TASK F p<<br />
COMP* 21.10 .001<br />
BROADJMP 2.92 .05<br />
CARRYWGT 9.33 .001<br />
SQUATTHR 59.65 .001<br />
BALBEAM 26.53 .001<br />
Note: *Wilks' Lambda = .21<br />
Table 3-9. GP AB: Groups significantly lower on physical abilities tasks than AC.<br />
TASK GROUP p<<br />
BRODJMP AE,AF .05, .05<br />
CARRYWGT AE,AF .001, .001<br />
SQUATTHR AE,AF,AG .001, .001, .001<br />
BALBEAM AE .001<br />
A discriminant function analysis, the results of which are presented in Table 3-10, indicated that the<br />
GP AB is extremely sensitive to the effects of exposure to ionizing radiation: 98.51 % of ACs and AEs<br />
were correctly classified. Table 3-10 represents numbers of subjects.<br />
Correlational analyses were performed for dosage of radiation and measures of physical performance for<br />
theAEs. Significant negative correlations ofrads with performance on BRODJMP and SQUATTHR<br />
(both rs>-.35, ps
Subsequent analyses included comparisons of performance scores of females and males as a function of<br />
non-exposure or exposure to radiation. The control group included only 7 females; thus 7 males from<br />
this group were selected for comparison purposes. The 3 exposed groups included a total of 17 females,<br />
so 17 males (matching the numbers offemales from each of these groups) were selected. Selection of<br />
males was not altogether random, since an attempt was made to ensure equivalence of ages between<br />
females and males. Means and standard deviations for the groups resulting from the combination of<br />
these variables are presented in Table 3-11, for each of the 4 physical tasks. The results of a MANOVA<br />
and univariate tests are presented in Table 3-12. MANOVA creates a composite measures based upon<br />
the corrolation relationship between individual measures.<br />
Significant differences on the composite measures were revealed for the main effects of gender and<br />
exposure; however, the multivariate interaction was not significant. The univariate tests indicated that<br />
overall, males performed significantly better than females on BRODJMP, CARRYWGT and<br />
SQUATTHR, but not on BALBEAM. Further, the controls were significantly better than the exposed<br />
people on SQUATTHR and BALBEAM. None of the univariate interactions were significant,<br />
however, indicating that the magnitudes of the female-male differences did not differ as a function of<br />
exposure to radiation.<br />
Table 3-11.<br />
GPAB: Means (and standard deviations) for females and males,<br />
either not exposed (controls) or exposed to radiation.<br />
TASK FEMALES- FEMALES- MALES- MALES-<br />
CONTROL EXPOSED CONTROL EXPOSED<br />
BROADJMP 1.36 (.11) 1.24 (.31) 1.72 (.13) 1.58 .37)<br />
(meter)<br />
CARRYWGT 38.14 (2.04) 33.88 (6.38) 44.71 (3.25) 39.88 (7.27)<br />
(meter)<br />
SQUATTHR 46.86 (5.49) 36.41 (15.12) 66.29 (4.19) 44.82 (22.55)<br />
(number)<br />
BALBEAM 19.21 (.91) 17.19 (1.91) 20.21 (1.43) 18.44 (2.79)<br />
(meter)<br />
46
Table 3-12.<br />
GP AB: MANOVA and univariate tests for females and males,<br />
either exposed or not exposed (controls) to radiation.<br />
SOURCE<br />
TASK<br />
F p<<br />
FEMALE-<br />
MALE<br />
CONTROL-<br />
EXPOSED<br />
INTERACTION<br />
COMP*<br />
BRODJMP<br />
CARRYWGT<br />
SQUATTHR<br />
BALBEAM<br />
COMP**<br />
BRODJMP<br />
CARRYWGT<br />
SQUATTHR<br />
BALBEAM<br />
COMP***<br />
BRODJMP<br />
CARRYWGT<br />
SQUATTHR<br />
BALBEAM<br />
4.28 .01<br />
13.42 .001<br />
5.05 .05<br />
7.00 .01<br />
2.76<br />
3.34 .05<br />
1.59<br />
1.78<br />
9.19 .01<br />
7.85 .01<br />
1.25<br />
.002<br />
1.44<br />
1.10<br />
.03<br />
Notes: * Wilks' Lambda = .71<br />
** Wilks' Lambda = .75<br />
*** Wilks' Lambda = .89<br />
3.3.2 ANAMUKR: Accuracy.<br />
Means and standard deviations of accuracy scores for the 9 ANAM tasks are presented in Table 3-13.<br />
Since all scores on SRT were 100%, this task was not included in data analyses. The results of a<br />
MANDV A revealed a significant difference on the composite measures. Univariate ANOV As and posthoc<br />
Dunnett tests indicated that these results were primarily due to the AEs being significantly less<br />
accurate than the ACs on 6 of the tasks: CDS, CDI, CDD, CPT, DGS, and MSP. These data indicate<br />
that the AEs are experiencing remarkable decrements in short-term (working) memory, as 5 of these<br />
tasks assess integrity of this cognitive skill. TheAFs performed less accurately than theACs on 5 tasks:<br />
2CH, CDI, CDD, DGS, and MSP. Apparently, they are also experiencing some difficulties in shortterm<br />
memory.<br />
The A Gs were significantly less accurate on 2 tasks: DGS and MSP. Thus all the exposed groups are<br />
being affected by the ionizing radiation in varying degrees, as compared to the control group. The<br />
results of these analyses are presented in Tables 3-14 and 3-15.<br />
47
Table 3-13. Accuracy (% Correct): 1995 Means (and Standard Deviations).<br />
GROUP~ AC AE AF AG<br />
TASK<br />
SRT 100.00 (0.00) 100.00 (0.00) 100.00 (0.00) 100.00 (0.00)<br />
2CH 97.52 (3.51) 96.94 (3.19) 94.10 (7.05) 96.97 (2.83)<br />
CDS 97.74 (1.84) 95.39 (3.77) 95.97 (3.55) 96.35 (3.05)<br />
CDI 97.84 (4.10) 83.69 (15.48) 87.90 (8.33) 93.32 (6.67)<br />
CDD 97.10 (3.42) 84.00 (16.84) 88.14 (9.23) 93.00 (7.75)<br />
CPT 96.19 (4.29) 76.28 (15.72) 90.55 (8.28) 91.74 (5.56)<br />
DGS 89.84 (8.52) 73.69 (13.46) 82.03 (11.28) 83.13 (7.96)<br />
MSP 98.42 (2.98) 84.83 (10.64) 82.03 (11.28) 89.71 (15.18)<br />
SPD 90.16 (5.24) 87.08 (8.97) 87.76 (8.41) 96.97 (2.83)<br />
Table 3-14. Accuracy: Results of MANOV A and Univariate ANOVAs.<br />
TASK F p<<br />
COMP* 7.81 .001<br />
2CH 3.67 .01<br />
CDS 3.25 .05<br />
CDI 12.74 .001<br />
CDD 9.14 .001<br />
CPT 26.03 .001<br />
DGS 12.98 .001<br />
MSP 13.35 .001<br />
SPD 13.55 .001<br />
Note: *Wilks' Lambda = .28<br />
48
Table 3-15. Accuracy: Groups significantly less accurate thanAC.<br />
TASK GROUP p<<br />
2CH AF .01<br />
CDS AE .01<br />
CDI AE,AF .001, .001<br />
CDD AE,AF .001, .01<br />
CPT AE .001<br />
DGS AE,AF,AG .001, .05, .05<br />
MSP AE,AF,AG .001, .001, .01<br />
SPD<br />
NONE<br />
The results of a discriminant function analysis for theACs andAEs are presented in Table 3-16. The<br />
high percentage of correct classification (93%) is mainly a result of all the ACs being correctly<br />
classified, as only 14% oftheAEs were misclassified. Nonetheless, accuracy of performance would<br />
appear to be a good indicator of the effects of the hazardous environment surrounding Chemobyl, and it<br />
is as sensitive to these effects as to those resulting from traumatic brain injury (Levinson & Reeves,<br />
1997) and stroke (Goldstone et al. 1995).<br />
Levels of exposure to varying dosages of radiation were obtained for each of the AEs. Mean dose was<br />
62.95 (SD=32.64); these ranged from 18 rads to 144 rads. No significant correlations between rads dose<br />
and accuracy on any of the tasks were found.<br />
Table 3-16. Accuracy: Results of discriminant function analysis for groups AC and AE.<br />
ACTUAL PREDICTED PREDICTED<br />
GROUP GROUP--AC GROUP--AE<br />
AC (numbers) 31 0<br />
AE (numbers) 5 31<br />
49
3.3.3 ANAMUKR: Efficiency.<br />
Means and standard deviations of efficiency scores for the 9 ANAM tasks are presented in Table 3-17.<br />
Scores on SRT were included in these analyses. Once again, the results of a MANOV A (Table 38)<br />
revealed a significant difference on the composite measures. Univariate ANOV As and post -hoc<br />
Dunnett tests, the results of which are presented in Tables 3-18 and 3-19, indicated that theAEs were<br />
significantly impaired on all 9 tasks as compared to the ACs. These effects appear to be extremely<br />
profound: the mean response time on SRT was greater than that of a group of 70-year olds (Goldstone,<br />
et al. 1995); in fact, a discriminant function analysis, the results of which are presented in Table 3-20,<br />
resulted in over 91 % correct classification oftheACs andAEs, with 34 of the 36AEs being correctly<br />
classified. The global impairment seen here is reminiscent of that observed in individuals who have<br />
suffered moderate traumatic brain injury (Levinson & Reeves, 1997). As with accuracy, the effects were<br />
not limited to theAEs. TheAFs were significantly less efficient than theACs on 7 tasks: SRT, CDS,<br />
CDI, CDD, CPT, DGS, and MSP. The efficiency oftheAGs was similarly affected; they were<br />
significantly less efficient than theACs on 6 tasks: CDS, CDI, CDD, CPT, DGS, and MSP.<br />
Since efficiency is partially based upon response speed, and since it was significantly lower for the AEs<br />
on all tasks (and for theAFs andAGs on most of them), a MANCOVA was performed in which SRT<br />
was entered as a covariate. It was believed that adjusting for differences on this primarily motor<br />
measure would yield a clearer assessment of CNS-processing of information. Although a multiple<br />
regression analysis indicated that SRT was significantly correlated with the other measures combined<br />
(Multiple R=.72,p
Table 3-17. Efficiency (Correct responses/min): 1995 means (and standard deviations).<br />
GROUP~ AC AE AF AG<br />
TASK<br />
SRT 167.26 (46.34) 117.78 (43.14) 131.76 (47.05) 147.71 (48.24)<br />
2CH 110.87 (25.49) 86.94 (27.06) 98.62 (32.97) 116.42 (19.63)<br />
CDS 49.42 (14.44) 28.64 (8.74) 31.52 (18.69) 38.58 (13.95)<br />
CDI 49.77 (15.39) 21.72 (8.44) 30.55 (17.67) 32.64 (12.13)<br />
CDD 52.61 (13.04) 27.89 (10.94) 35.69 (19.85) 38.29 (10.81)<br />
CPT 96.74 (19.68) 58.86 (16.88) 82.90 (22.63) 81.35 (18.53)<br />
DGS 44.65 (10.45) 26.17 (8.07) 35.14 (10.11) 29.94 (8.06)<br />
MSP 51.55 (16.77) 26.39 (10.49) 32.59 (19.14) 29.29 (11.20)<br />
SPD 32.61 (9.85) 20.33 (6.05) 32.14 (11.88) 26.03 (7.27)<br />
51
Table 3-18.<br />
Efficiency: Results ofMANOVA and univariate ANOVAs (and<br />
MANCOV A and univariate ANCOVAs with SRT as a covariate).<br />
TASK F p<<br />
COMP* 6.79 (6.51) .001 (.001)<br />
SRT 7.00 .001<br />
2CH 8.15 (4.58) .001 (.01)<br />
CDS 13.77 (6.55) .001 (.001)<br />
CDI 24.27 (16.14) .001 (.001)<br />
CDD 17.98 (11.29) .001 (.001)<br />
CPT 22.14 (14.77) .001 (.001)<br />
DGS 24.75 (18.46) .001 (.001)<br />
MSP 19.20 (13.55) .001 (.001)<br />
SPD 14.11 (11.42) .001 (.001)<br />
Notes: *Wilks' Lambda = .28 (.33)<br />
Table 3-19. Efficiency: Groups significantly less efficient thanAC.<br />
TASK GROUP p<<br />
SRT AE,AF .001, .01<br />
2CH AE .001<br />
CDS AE,AF,AG .001, .001, .01<br />
CDI AE,AF,AG .001, .001, .001<br />
CnD AE,AF,AG .001, .001, .001<br />
CPT AE,AF,AG .001, .01, .01<br />
DGS AE,AF,AG .001, .001, .001<br />
MSP AE,AF,AG .001, .001, .001<br />
SPD AE,AG .001, .01<br />
52
Table 3-20. Efficiency: Results of discriminate function analysis for groups AC and AE.<br />
ACTUAL PREDICTED PREDICTED<br />
GROUP GROUP--AC GROUP--AE<br />
AC (numbers) 27 4<br />
AE (numbers) 2 34<br />
Correlational analyses were also perfonned for dose of radiation and efficiency of perfonnance for the<br />
AEs. Unlike accuracy, significant negative correlations ofrads dose with efficiency on CDI, CDS,<br />
CPT, DGS, and MSP (all of which assess short-tenn memory), as well as with SRT (all 7s>-.36, all<br />
ps 4.36,ps < .01). Compared<br />
to the ACs, tapping rates for both hands were significantly lower for theAEs (.001) and for theAFs<br />
(.05); while levels of sleepiness were higher for theAEs (.001).<br />
3.3.5 GP AB-ANAMUKR: Correlational Analyses.<br />
Significant correlations were found between many of the measures on the GPAB; these are presented in<br />
Table 3-22, along with intercorrelations between perfonnance on the tasks comprising the GPAB and<br />
accuracy ofperfonnance on the tasks in ANAMUKR for the combined groups (N=127).<br />
53
Table 3-22.<br />
Pearson correlations between GP AB and ANAMUKR: Accuracy<br />
(N=127).<br />
TASK<br />
BRODJMP<br />
CARRYWGT<br />
SQUATTHR<br />
BALBEAM<br />
CDD<br />
CDI<br />
CDS<br />
CPT<br />
DGS<br />
MSP<br />
SPD<br />
SRT<br />
2CH<br />
BRODJMP<br />
.59**<br />
.62**<br />
-.04<br />
.01<br />
.14<br />
.02<br />
.18<br />
.15<br />
.29**<br />
-.06<br />
.03<br />
-.07<br />
CARRY SQUATTHR BALBEAM<br />
WGT<br />
.65**<br />
.14 .30**<br />
.11 .27** .19*<br />
.32** .37** .17<br />
.06 .11 .21*<br />
.21 * .48** .45**<br />
.14 .37** .26**<br />
.31 ** .41 ** .05<br />
.11 .08 .14<br />
-.08 .06 .09<br />
-.02 -.05 -.01<br />
Notes:<br />
* p
Table 3-23. Pearson correlations between GP AB and ANAMUKR: Efficiency<br />
(N=127).<br />
TASK<br />
BRODJMP<br />
CARRYWGT SQUATTHR BALBEAM<br />
CDD<br />
CDI<br />
CDS<br />
CPT<br />
DGS<br />
MSP<br />
SPD<br />
SRT<br />
2CH<br />
.04<br />
.21*<br />
.23*<br />
.28**<br />
.24**<br />
.27**<br />
.16<br />
.35**<br />
.10<br />
.07 .42** .26**<br />
.20* .54** .24**<br />
.19* .45** .16<br />
.19* .49** .40**<br />
.23** .48** .27**<br />
.29** .49** .22*<br />
.15 .31 ** .36**<br />
.21* .48** .13<br />
.11 .24** .23*<br />
Notes:<br />
* p
3.3.5.1 Age. Since ages of all 127 participants were obtained, correlations between it and all other<br />
variables were calculated. Significant negative correlations were observed between age and three of the<br />
GPAB tasks (BRODJMP, CARRYWGT, and SQUATTHR: all rs>.36, allps
GPAB ANAM-ACC ANAM-EFF<br />
GROUP<br />
Figure 3-26. Mean % Decrement for Exposure Groups-1996 Relative to Controls-1995.<br />
The data of Figure 3-26 are similar to those illustrated in Figure 3-25; however,except for Groups AF<br />
andAG on the GPAB, the magnitude of the 1996 differences between the exposure groups and the<br />
Controls is noticeably greater than those of 1995. In addition to declines in performance in groups<br />
significantly lower than the Controls in 1995, the performance of several groups not significantly lower<br />
than the Controls on certain tasks in 1995 declined to significant levels in 1996; these are listed in<br />
Table 3-46.<br />
Using a procedure similar to that for calculating mean percent decrement of the exposure groups relative<br />
to the Controls, the mean percent declines for each of these groups on each battery of tests were<br />
calculated by using 1995 levels of performance as a baseline by which to gauge 1996 levels (i.e., mean -<br />
1996/ mean-1995). These declines are presented in Table 3-27 and illustrated graphically in Figure 3-<br />
27. With the exception of Groups AF and AG on the GP AB, all groups showed significant declines as<br />
revealed by MANOV As, the results of which are presented in Table 3-28, on all test batteries from 1995<br />
to 1996. These findings indicate that both the physical (in the case of Group AE) and cognitive<br />
performance levels in these groups of individuals are worsening over time.<br />
57
Table 3-26. Groups not Significantly Lower than AC in 1995, but significantly lower in 1996.<br />
BATTERY: GROUP p<<br />
TASK<br />
GPAB<br />
BRODJUMP AE .001<br />
ANAM-ACCURACY<br />
2CH AE,AG .01, .01<br />
CDS AF .01<br />
CDI AG .001<br />
CDD AG .001<br />
SPD AE,AF .001, .01<br />
ANAM-EFFICIENCY<br />
SPD AF .001<br />
Table 3-27. Mean % Performance Decline for Exposure Groups: 1995-1996.<br />
GROUP GPAB ANAMUKR-ACC ANAMUKR-EFF<br />
AE 9.78 1.95 11.89<br />
AF 0.00 2.84 15.22<br />
AG 0.00 3.18 9.03<br />
58
GPAB ANAM-ACC ANAM-EFF<br />
GROUP<br />
Figure 3-27. Mean % decline for exposure groups: 1995-1996.<br />
Table 3-28. Significant multivariate declines by exposure groups.<br />
TEST GROUP WILKS' F p<<br />
BATTERY<br />
LAMBDA<br />
GPAB AE .39 11.95 .001<br />
ANAM-ACC AE .51 3.07 .01<br />
AF .42 3.49 .01<br />
AG .38 4.60 .01<br />
ANAM-EFF AE .39 4.39 .001<br />
AF .28 5.01 .001<br />
AG .32 5.32 .001<br />
59
3.6 SPECIFIC ASSESSMENTS OF DECLINES IN EXPOSURE GROUPS.<br />
Means and standard deviations for the exposure groups on the 1996 GP AB retest are presented in Table<br />
3-29. Similar data for accuracy, efficiency, and the additional measures performance on the 1996<br />
ANAMUKR retest are presented in Tables 3-30, through 3-32, respectively. Significant declines, as<br />
revealed by the results of univariate analyses comparing the 1996 levels of performance of the exposure<br />
groups on the various tasks in the test batteries to their 1995 levels, are presented in Table 3-33, along<br />
with the figure number from Figures 3-30 through 3-53 illustrating these declines.<br />
3.6.1 GPAB.<br />
Already significantly weaker than the ACs (1995), the AEs continued to decline on all measures of<br />
strength, including explosive, static and dynamic. These findings indicate that they are physically<br />
deteriorating, even 10 years following their exposure to the radiation in the power station. In contrast,<br />
the AFs and A Gs maintained their 1995 levels--perhaps because they are working in occupations<br />
requiring physical exertion and were therefore able to "stay in shape".<br />
TABLE 3-29. GPAB: 1996 means (and standard deviations) for the exposure groups.<br />
TASK AE AF AG<br />
·BRODJMP 1.30 (.28) 1.37 (.14) 1.41 (.42)<br />
CARRYWGT 30.56 (11.68) 38.89 (5.93) 41.48 (6.03)<br />
SQUATTHR 19.47 (7.44) 35.39 (11.97) 43.74 (17.07)<br />
BALBEAM 15.85 (3.16) 20.02 (1.76) 19.11 (1.20)<br />
3.6.2 ANAMUKR: Accuracy.<br />
All three exposure groups showed significant declines in accuracy of performance on a variety of the<br />
tasks. Most of these declines occurred in Groups AF and A G, on tasks assessing both attention and<br />
memory skills. Nonetheless, theAEs declined as well, on tasks requiring sustained attention.<br />
Surprisingly, all three groups showed significant declines on SPD, indicating a developing difficulty in<br />
skills related to assessment of spatial relations.<br />
60
Table 3-30. ANAMUKR Accuracy (% correct): 1996 means (and standard deviations)<br />
for the exposure groups.<br />
TASK AE AF AG<br />
SRT 100.00 (.00) 100.00 (.00) 100.00 (.00)<br />
2CH 94.26 (5.30) 94.79 (4.53) 93.90 (7.19)<br />
CDS 93.59 ( 3.66) 94.07 (3.21) 95.39 (3.32)<br />
CDI 79.85 (8.07) 82.96 (8.54) 88.71 (10.55)<br />
CDD 79.53 (9.96) 81.11 (6.79) 84.00 (10.75)<br />
CPT 83.24 (16.15) 89.11 (6.96) 91.61 (6.80)<br />
DGS 74.82 (10.26) 83.43 (7.03) 85.16 (8.56)<br />
MSP 80.94 (13.30) 81.18 (16.88) 89.58 (9.14)<br />
SPD 80.88 (7.83) 81.79 (6.56) 88.55 (6.73)<br />
3.6.3 ANAMUKR: Efficiency.<br />
As with accuracy, all exposure groups exhibited significant declines in efficiency of performance on the<br />
majority of the tasks. Most of these were the same tasks upon which their accuracy declined. Thus the<br />
declines in efficiency are not surprising, since efficiency is based in part on accuracy of performance.<br />
Nonetheless, it should be noted that none of the groups showed significant declines in SRT, and<br />
therefore the declines in efficiency cannot be explained by slowing of response speed (the other factor<br />
upon which efficiency of performance is based). The fmding that declines in efficiency for the most part<br />
accompanied corresponding declines in accuracy would indicate that these individuals are experiencing<br />
difficulty in processing cognitive information on tasks entailing attention, memory and spatial abilities.<br />
61
Table 3-31.<br />
ANAMUKR Efficiency (correct responses/min): 1996 means (and standard<br />
deviations) for the exposure groups.<br />
TASK AE AF AG<br />
SRT 117.74 (34.16) 136.93 (52.23) 139.00 (52.65)<br />
2ca 76.47 (29.60) 90.39 (31.89) 98.13 (33.56)<br />
CDS 23.35 (8.03) 31.79 (21.41) 30.45 (11.55)<br />
CDI 17.35 (5.71) 27.57 (16.19) 26.29 (11.48)<br />
CDD 20.74 (8.47) 25.50 (13.26) 29.13 (12.77)<br />
CPT 61.26 (1S.0S) 74.14 (24.49) 79.74 (27.67)<br />
DGS 25.26 (7.57) 29.39 (7.4S) 30.S7 (10.S7)<br />
MSP 19.52 (S.52) 21.82 (11.30) 31.81 (15.30)<br />
SPD 18.41 (7.22) 21.18 (9.29) 24.97 (9.53)<br />
3.6.4 ANAMUKR: Additional Measures.<br />
Means and standard deviations for the exposure groups are shown in Table 3-32.<br />
Table 3-32. ANAMUKR Additional measures: 1996 means (and standard deviations).<br />
TASK AE AF AG<br />
TAP-R (mean n 42.82 (9.74) 48.351 (11.38) 55.58 (13.37)<br />
of responses in 10<br />
sec)<br />
TAP-L (mean n 37.97 (8.11) 41.00 (10.04) 49.30 (12.07)<br />
of responses in 10<br />
sec)<br />
SLP (scores from 2.59 (.71) 1.27 (.46) 1.46 (.59)<br />
1-7)<br />
No significant changes were observed from 1995 to 1996 in rates of tapping for either hand by any<br />
group; however, all 3 groups showed significant decreases in levels of sleepiness (all Fs > 4.29, allps<br />
< .05).<br />
62
Table 3-33 shows the significant declines in performance from 1995 to 1996 for the GPAB tasks, and for<br />
accuracy and efficiency on ANAMUKR; the figure number refers to the figure which specifically<br />
illustrates the decline.<br />
Table 3-33. Significant declines in performance by the exposure groups: 1995 to 1996.<br />
BATTERY: GROUP F p< FIG.<br />
TASK #<br />
GPAB<br />
BRODJUMP AE 8.29 .01 30<br />
CARRYING WGT AE 31.58 .001 31<br />
SQUATTHRUSTS AE 16.32 .001 32<br />
ANAM-ACCURACY<br />
2CH AE,AG 7.90, 4.37 .01, .05 34<br />
CDS AE,AF 5.71, 7.90 .05, .01 35<br />
CDI AF,AG 5.52, 5.52 .05, .05 36<br />
CDD AF,AG 10.30, 12.53 .01, .001 37<br />
SPD AE,AF, 10.05, 12.04, 35.40 .01, .01,001 41<br />
AG<br />
ANAM-EFFICIENCY<br />
2CH AG 7.45 .01 43<br />
CDS AE,AG 8.01, 10.96 .01, .01 44<br />
CDI AE,AG 6.15, 7.90 .01, .01 45<br />
CDD AE,AF, 11.02, 4.62, 10.89 .01, .05, .01 46<br />
AG<br />
DGS AF 7.51 .01 48<br />
MSP AE,AF 10.30, 8.01 .01, .01 49<br />
SPD AF 14.67 .001 50<br />
63
3.7 RESULTS OF 1997 RETEST SESSION.<br />
Because of unavailability for testing resulting from relocation, illness, or death, 1997 data were obtained<br />
on 34 Eliminators, 20 Forestry workers, and 28 Agricultural workers. However, some of these had not<br />
been available for testing on the GPAB and/or ANAMUKR in 1996, so the 1996-97 comparisons were<br />
based on Ns as follows: GPAB-Eliminators-32, Forestry workers-l 8, Agricultural workers-28;<br />
ANAMUKR-Eliminators-30, Forestry workers-15, Agricultural workers-28. Because many of the<br />
original forestry workers were unavailable for testing in 1997, a new group of 13 foresters was tested.<br />
However, since this was their first year of testing, their data could not be included in the analyses.<br />
3.7.1 Global Assessments of Declines By Exposure Groups.<br />
Mean percent decrements in performance for the exposure groups on the 1997 retest relative to the 1995<br />
control data are presented in Table 3-34 and graphically illustrated in Figure 3-28. The data of Figure 3-<br />
28 are similar to those illustrated in Figures 3-25 and 3-26, except that decrements in ANAMUKR<br />
accuracy for Groups AE and AF are more pronounced.<br />
Table 3-34. Mean % performance decrement for expo groups-1997 relative to<br />
controls-1995.<br />
GROUP<br />
GPAB<br />
ANAMUKR-ACC<br />
ANAMUKR-EFF<br />
AE 35.97<br />
AF 9.93<br />
AG 11.60<br />
23.24<br />
14.96<br />
8.46<br />
50.54<br />
43.68<br />
34.57<br />
DAG<br />
ImAF<br />
.AE<br />
GPAB<br />
ANAM-ACC<br />
GROUP<br />
ANAM-EFF<br />
Figure 3-28. Mean % decrement for exposure groups-1997 relative to controls-199S.<br />
64
In addition to continuing declines exhibited by all the groups on at least some of the measures, the<br />
performance of Groups AF and A G not significantly lower than that of the Controls in 1995 or 1996<br />
declined to significant levels on several tasks in 1997. These are listed in Table 3-35.<br />
Table 3-35. Groups not significantly lower thanAC in 1995 or 1996, but significantly<br />
lower in 1997.<br />
BATTERY: GROUP p<<br />
TASK<br />
GPAB<br />
BALANCE BEAM AF .05<br />
ANAM-ACCURACY<br />
CDS AG .05<br />
CPT AF .001<br />
ANAM-EFFICIENCY<br />
SRT<br />
AG<br />
.01<br />
2CH AF,AG .001, .05<br />
U sing a procedure similar to that for calculating mean percent decrement of the exposure groups relative<br />
to the Controls, the mean percent declines for each of these groups on each battery of tests were<br />
calculated by using 1995 levels of performance as a baseline by which to gauge 1997 levels (i.e., mean -<br />
1997/ mean-1995). These declines are presented in Table 3-36 and illustrated graphically in Figure 3-<br />
29. With the exception of Groups AF and A G on the GP AB, all groups showed significant declines as<br />
revealed by MANOVAs, the results of which are presented in Table 3-37, on all test batteries from 1995<br />
to 1997. These findings indicate that both the physical (in the case of Group AE) and cognitive<br />
performance levels in these groups of individuals are worsening over time.<br />
Mean percent declines in performance by the 3 exposure groups from 1996 to 1997 are presented in<br />
Table 3-36 and graphically illustrated in Figure 3- 29. As revealed by MANOV As, the decline of Group<br />
AE on the GP AB was significant, as were those of all 3 groups on ANAM accuracy. The declines in<br />
ANAM efficiency, although noteworthy, were not significant.<br />
65
Table 3-36. Mean % performance decline for exposure groups: 1996-1997.<br />
GROUP<br />
GPAB<br />
ANAMUKR-ACC<br />
ANAMUKR-EFF<br />
AE 3.00<br />
AF 0.00<br />
AG 0.00<br />
12.43<br />
8.44<br />
2.43<br />
5.20<br />
14.52<br />
6.41<br />
DAG<br />
IIAF<br />
.AE<br />
GPAB<br />
ANAM-ACC<br />
GROUP<br />
ANAM-EFF<br />
Figure 3-29. Mean % Decline for Exposure Groups: 1996-1997.<br />
These findings indicate that the physical performance of Group AE and the accuracy of performance of<br />
all 3 exposed groups are continuing to significantly decline. Although a MANOV A indicated that the<br />
decline of Group AE on efficiency was also significant, a MANCOV A using SRT as a covariate<br />
revealed that the significance was a result of slowing of reaction time only (see Table 3-41). The results<br />
of the MANOV As assessing the significance of the changes in performance for all 3 groups from 1996<br />
to 1997 are presented in Table 3-37.<br />
66
Table 3-37. Significant multivariate declines by exposure groups from 1996 to 1997.<br />
TEST BATTERY GROUP WILK'S F p<<br />
LAMBDA<br />
GPAB AE .52 6.48 .001<br />
ANAM-ACC AE .14 17.31 .001<br />
AF .09 8.62 .01<br />
AG .47 2.84 .05<br />
ANAM-EFF AE .39 3.63 .01*<br />
Note: *n.s. when analyzed via MANCOVA using SRT as a covariate.<br />
3.7.2 Specific Assessments of Declines in Exposure Groups.<br />
Means and standard deviations for the exposure groups on the 1997 GP AB retest are presented in Table<br />
3-38. Similar data for accuracy and efficiency of performance on the 1997 ANAMUKR retest, and for<br />
the additional ANAMUKR measures, are presented in Tables 3-39, through 3-41 respectively. Figures<br />
3-30 through 3-53 illustrate changes in performance for the three exposure groups from 1996 to 1997 (as<br />
well as from 1995-1996). On each figure, the 1995 control data are again included as a point of<br />
reference. Significant declines, as revealed by the results of univariate analyses comparing the 1997<br />
levels of the exposure groups on the various tasks in the test batteries to their 1996 levels, are presented<br />
in Table 3-42 along with the figure number from Figures 3-30 through 3-53 illustrating these declines.<br />
Table 3-38. GPAB: 1997 means (and standard deviations) for the exposure groups.<br />
TASK AE AF AG<br />
BRODJMP 1.24 (.28) 1.42 (.20) 1.40 (.44)<br />
(meter)<br />
CARRYWGT 28.32 (12.13) 42.05 (7.99) 41.00 (6.77)<br />
(meter)<br />
SQUATTHR 19.71 (8.39) 41.65 (13.81) 45.04 (19.30)<br />
(number)<br />
BALBEAM 15.76 (2.55) 20.95 (1.94) 19.25 (1.85)<br />
(meter)<br />
67
Table 3-39.<br />
ANAMUKR: Accuracy (% correct): 1997 means (and standard deviations)<br />
for the exposure groups.<br />
TASK AE AF AG<br />
SRT 100.00 (.00) 100.00 (.00) 100.00 (.00)<br />
2CH 89.86 (6.91) 91.45 (12.00) 94.07 (16.82)<br />
CDS 87.06 (8.50) 86.92 (9.18) 93.80 (3.91)<br />
CDI 65.63 (14.29) 73.36 (11.19) 88.13 (8.74)<br />
CDD 61.33 (10.82) 68.44 (11.91) 84.17 (9.39)<br />
CPT 73.03 (14.13) 77.06 (10.84) 86.88 (12.91)<br />
DGS 68.70 (11.67) 72.02 (10.66) 76.39 (12.35)<br />
MSP 64.79 (17.04) 80.33 (11.34) 85.78 (13.75)<br />
SPD 75.63 (10.22) 81.00 (9.12) 90.67 (8.48)<br />
Table 3-40. ANAMUKR: Efficiency (correct responses/min): 1997 means (and<br />
standard deviations) for the exposure groups.<br />
TASK AE AF AG<br />
SRT 94.69 (30.61) 103.23 (42.00) 133.87 (37.75)<br />
2CH 67.28 (32.01) 75.70 (30.03) 91.45 (31.32)<br />
CDS 23.89 (11.47) 20.03 (9.29) 30.44 (15.20)<br />
CDI 19.28 (12.97) 21.26 (10.27) 27.19 (12.11)<br />
CDD 18.55 (10.87) 25.00 (9.79) 27.13 (13.06)<br />
CPT 54.01 (20.43) 58.27 (13.87) 71.16 (19.45)<br />
DGS 23.00 (9.13) 26.12 (8.81) 25.80 (6.72)<br />
MSP 18.43 (12.29) 24.09 (11.13) 27.04 (12.65)<br />
SPD 20.66 (12.68) 20.16 (6.67) 24.63 (8.42)<br />
68
Table 3-41. ANAMUKR: Additional measures: 1997 means (and standard deviations).<br />
TASK AE AF AG<br />
TAP-R (mean n 42.97 (11.99) 51.20 (11.89) 57.47 (11.01)<br />
of responses in<br />
10 sec)<br />
TAP-L (mean n 37.81 (9.30) 46.16 (12.40) 51.43 (10.79)<br />
of responses in<br />
10 sec)<br />
SLP (scores from 2.12 (.60) 1.87 (.64) 1.46 (.59)<br />
1-7)<br />
3.7.3 GPAB.<br />
From 1996 to 1997, Group AE declined on measures of explosive and dynamic strength, reflecting their<br />
continuing physical deterioration (see Table 3-42). Conversely, GroupsAF andAG showed some<br />
improvement in physical abilities.<br />
3.7.4 ANAMUKR: Accuracy.<br />
As seen in Table 3-42, GroupAE showed significant declines in accuracy from 1996 to 1997 on all<br />
tasks. This global decline in accuracy is most likely reflecting a general deterioration of their<br />
neurocognitive abilities; in fact, their levels of performance are similar to those observed in individuals<br />
with moderate-severe traumatic brain injuries (TBIs) (Levinson & Reeves, 1997; Levinson, et al, 1998).<br />
Similar declines were observed in GroupAF, on 6 of8 tasks. Accuracy of performance of Group AG<br />
also declined on 3 tasks, although not as sharply as that of the other exposed groups.<br />
3.7.5 ANAMUKR: Efficiency.<br />
GroupAE showed a significant decline in efficiency on SRT only, and when this was entered into a<br />
MANCOV A as a covariate, the multivariate difference in efficiency revealed by the MANOV A (see<br />
Table 3-37) was no longer significant. Group AF also showed a significant slowing of reaction time,<br />
and this most likely explains their corresponding decline in efficiency on CPT. Group A G did not<br />
exhibit any apparent slowing of reaction time, however, and therefore the significant declines on CPT<br />
and DGS are more likely reflecting difficulty in sustaining attention. At this point in time, the<br />
performance of all exposed groups on all tasks is significantly lower than that of the Controls.<br />
69
Table 3-42. Significant declines in performance by the exposure groups: 1996 to 1997.<br />
BATTERY: GROUP F p< FIG. #<br />
TASK<br />
GPAB<br />
BROADJUMP AE 7.87 .01 30<br />
SQUATTHRUSTS AE 3.49 .05* 32<br />
ANAM-ACCURACY<br />
2CH AE 14.59 .001 34<br />
CDS AE,AF 18.77, 11.88 .001, .001 35<br />
CDI AE,AF 23.02, 14.43 .001, .001 36<br />
CDD AE,AF 69.38,14.21 .001, .001 37<br />
CPT AE,AF, 11.76, 18.07, .001, .001, .05 38<br />
AG 4.07<br />
DGS AE,AF, 6.32, 16.09, .01, .001, .001 39<br />
AG 15.49<br />
MSP AE,AF, 12.26, 9.29, .001, .01, .05 40<br />
AG 4.04<br />
SPD AE 5.98 .05 41<br />
ANAM-EFFICIENCY<br />
SRT AE,AF 11.93, 7.24 .001, .01 42<br />
CPT AF,AG 5.56, 8.79 .05, .01 47<br />
DGS AG 13.89 .001 48<br />
Note: * I-tailed<br />
3.7.6 ANAMUKR: Additional Measures.<br />
No significant declines in tapping rates for either hand were observed. GroupAE showed a significant<br />
(.001) decrease in levels of sleepiness, while Group AF showed a significant increase.<br />
3.8 RESULTS OF 1998 RETEST SESSION.<br />
For similar reasons described for the 1997 retest, 1998 data were obtained on 22 Eliminators, 21<br />
Forestry workers, and 29 Agricultural workers. Since some of these had not been available for testing<br />
on the GPAB and/or ANAMUKR in 1996, but were in 1997; therefore, the 1997-98 comparisons were<br />
based on Ns as follows: GPAB-Eliminators-22, Forestry workers-19, Agricultural workers-29;<br />
ANAMUKR-Eliminators-22, Forestry workers-20, Agricultural workers-29. Of the new group of 13<br />
foresters from 1997,8 were retested. Since this was only their second year of testing, their data were not<br />
included in the analyses.<br />
70
3.8.1 Assessments of Declines by Exposure Groups.<br />
Multivariate and univariate tests revealed no significant declines from 1997 levels on any measure in<br />
any of the test batteries, for any group. In fact, some of the groups showed significant increases in levels<br />
of performance compared to the 1995 control groups; these will be described in the relevant sections of<br />
this report.<br />
3.8.2 GPAB.<br />
Means and standard deviations on the 1998 retest are presented in Table 3-43. TheAEs showed<br />
significant increases in performance on CARRYWGT, SQUATTHR and BALBEAM (ps < .01), while<br />
theAGs improved on CARRYWGT (.05).<br />
Table 3-43. GP AB: 1998 means (and standard deviations) for the exposure groups.<br />
TASK AE AF AG<br />
BRODJMP 1.29 (.16) 1.42 (.24) 1.51 (.40)<br />
(meter)<br />
CARRYWGT 31.82 (12.40) 41.62 (8.56) 44.58 (6.98)<br />
(meter)<br />
SQUATTHR 23.91 (7.51) 40.52 (12.02) 51.35 (19.01)<br />
(number)<br />
BALBEAM 17.36 (1.59) 19.93 (1.88) 19.96 (2.81)<br />
(meter)<br />
3.8.3 ANAMUKR: Accuracy.<br />
Means and standard deviations for the 1998 accuracy retest are presented in Table 3-44. Although some<br />
improvements over the 1997 levels were observed in some groups, none of these was significant.<br />
71
Table 3-44.<br />
ANAMUKR: Accuracy (% correct): 1998 means (and standard<br />
deviations for the Exposure Groups.<br />
TASK AE AF AG<br />
SRT 100.00 (.00) 100.00 (.00) 100.00<br />
2CH 89.90 (5.56) 88.89 (10.25) 92.86<br />
CDS 89.70 (6.62) 88.20 (2.75) 94.99<br />
CDI 63.83 (8.99) 74.06 (11.52) 85.29<br />
CDD 64.20 (11.92) 72.19 (10.23) 79.53<br />
CPT 73.07 (16.74) 82.29 (9.35) 87.96<br />
DGS 68.75 (10.11) 72.29 (9.59) 76.01<br />
MSP 67.57 (16.75) 79.63 (8.39) 85.63<br />
SPD 82.05 (10.98) 83.25 (5.45) 91.72<br />
(.00)<br />
(14.32)<br />
(3.72)<br />
(10.92)<br />
(13.56)<br />
(11.58)<br />
(13.94)<br />
(17.57)<br />
(5.39)<br />
3.8.4 ANAMUKR: Efficiency.<br />
Means and standard deviations in efficiency are presented in Table 3-45 compared to the 1995 control<br />
group. Significant (.05) increases in efficiency of performance were made by the A Gs on SRT and<br />
SPD. Although other increases in levels of performance were observed (see, e.g., Figures 3-43,3-47,<br />
and 3-49), none were significant.<br />
72
Table 3-45. ANAMUKR: Efficiency (correct responses/min) : 1998 means<br />
(and standard deviations) for the exposure groups.<br />
TASK AE AF AG<br />
SRT 96.72 (38.46) 86.39 (20.87) 150.20<br />
2CH 60.31 (30.64) 63.69 (22.84) 100.70<br />
CDS 18.62 (7.61) 21.56 (13.28) 33.82<br />
CDI 15.97 (9.63) 18.32 (13.11) 28.06<br />
CDD 14.25 (6.10) 16.58 (8.19) 29.02<br />
CPT 53.60 (16.77) 63.26 (13.80) 76.97<br />
DGS 19.64 (5.66) 22.73 (8.38) 27.36<br />
MSP 15.38 (9.89) 19.32 (13.10) 30.10<br />
SPD 16.32 (9.07) 19.45 (9.27) 28.15<br />
(31.82)<br />
(26.94)<br />
(11.59)<br />
(12.38)<br />
(10.22)<br />
(15.54)<br />
(7.04)<br />
(13.31)<br />
(6.85)<br />
3.8.5 ANAMUKR: Additional measures.<br />
Means and standard deviations for these are presented in Table 3-46. GroupAG showed a significant<br />
increase (.05) in TAP-L.<br />
Table 3-46. ANAMUKR: Additional Measures: 1998 Means (and Standard Deviations).<br />
TASK AE AF AG<br />
TAP-R (mean n 47.64 (6.48) 49.65 (6.09) 59.48 (10.29)<br />
of responses in<br />
10 sec)<br />
TAP-L (mean n 40.80 (5.36) 43.53 (6.62) 53.84 (10.14)<br />
of responses in<br />
10 sec)<br />
SLP (scores from 2.32 (.72) 1.70 (.66) 1.21 (.49)<br />
1-7)<br />
Figures 3-30 through 3-53 graphically illustrate yearly changes in perfonnance from 1995-1998 on all<br />
measures. The 4-year averaged levels of the Controls are represented as a straight dotted line (typically<br />
across the top) on each figure. Each graph represents the mean and the standard deviation are identified<br />
in their appropriate tables.<br />
73
BROADJUMP<br />
..........................................<br />
1.6.,...---------------------.<br />
C/)<br />
a:::<br />
w<br />
I- W<br />
~<br />
c:<br />
CO<br />
Q)<br />
~<br />
1.5<br />
1.4<br />
1.3<br />
..--.-....... ......-.. --.....-<br />
.<br />
~-~.------.------<br />
GROUP<br />
• AE<br />
..........------r--<br />
• AF<br />
1.2<br />
... AG<br />
~-----_r------~-------~<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-30. Mean performance on GPAB: BROADJUMP.<br />
fQ<br />
W<br />
tu<br />
~<br />
c:<br />
CO<br />
Q)<br />
~<br />
CARRYING WEIGHT<br />
55~-------------------------~<br />
50<br />
45<br />
40<br />
35<br />
30<br />
..........................................<br />
--<br />
..----.......--.. ........ -.-......~..,...-.---<br />
.... ---.---<br />
------.. --<br />
• AE<br />
25<br />
• AF<br />
20<br />
... AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-31. Mean performance on GP AB: CARRYING WEIGHT.<br />
74
SQUATTHRUSTS<br />
70~------------------------------------.<br />
60<br />
.. - -- .. -.--<br />
~ 50 __ -<br />
W IP------..------- ---<br />
III<br />
2 40 ..------..--- -- ... ~<br />
c:<br />
m<br />
~ 30<br />
GROUP<br />
~ -_ .... ---- ..<br />
10~ __________ ~ __________ ~~ ________ ~ .... AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-32. Mean performance on GPAB: SQUAT THRUSTS.<br />
BALANCE BEAM<br />
23~------------------------------------,<br />
~<br />
UJ<br />
tu<br />
:2<br />
c:<br />
m<br />
Q)<br />
:2<br />
22 ••••••••••••••••••••••••••••••••••••••••••<br />
21<br />
20<br />
19<br />
18<br />
17<br />
-_ .......... -<br />
_..._-<br />
............. ....- ....<br />
--<br />
....... _._._.JJr--_.--<br />
.----<br />
• AE<br />
16<br />
• AF<br />
15<br />
.... AG<br />
~----------~-----------,------------4<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-33. Mean performance on GP AB: BALANCE BEAM.<br />
75
2CH-ACC<br />
..........................................<br />
98~------------------------------------~<br />
96 ~<br />
•<br />
0::: 94 .<br />
0:::<br />
""<br />
t;<br />
" •<br />
W<br />
---~-<br />
0<br />
()<br />
COI-ACC<br />
100~------------------------------------------------------~<br />
.. -: :-:-. .. ....,.............................. --- -<br />
.<br />
90 .... -~-<br />
....<br />
..<br />
........ .......... 80<br />
" .-.-<br />
~<br />
"<br />
" " '-... GROUP<br />
-----<br />
70 e AE<br />
--........................................ -l. AF<br />
60~ _______________ ~ __________________ ~ _______________ ~ ... AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-36. Mean performance on ANAMUKR: CDI-ACC.<br />
COO-ACC<br />
100~------------------------------------------------------~<br />
b<br />
w<br />
c::<br />
c::<br />
8<br />
::R 0<br />
~<br />
90<br />
80<br />
70<br />
60<br />
..... ...,"="-,.,-.......<br />
•••••••••...••••.•••••.••.••••••••<br />
, . .........." ------ .......<br />
. .......<br />
1~--------~1L .-. .........<br />
' .... , ....<br />
--<br />
' ... ---<br />
50<br />
1995 1996 1997<br />
-.._---,eAE<br />
1998<br />
• AF<br />
... AG<br />
YEARS<br />
Figure 3-37. Mean performance on ANAMUKR: CDD-ACC.<br />
77
CPT-ACC<br />
100~----------------------------------~<br />
90<br />
80<br />
.........................................<br />
......<br />
~.-<br />
.... - .....,' ..<br />
"<br />
' ..,<br />
, ,<br />
.......-._.-<br />
" , ,<br />
, --' .... -<br />
lI'--<br />
• AE<br />
• AF<br />
70~ __________ ~ __________-,__________--4<br />
.... AG<br />
1995 1996 1997<br />
1998<br />
YEARS<br />
Figure 3-38. Mean performance on ANAMUKR: CPT -ACC.<br />
OGS-ACC<br />
t;<br />
W 80<br />
0:::<br />
0:::<br />
8<br />
'$.<br />
~ 70<br />
.............................................<br />
---- ._111 -.~'" ~<br />
90~----------------------------------~<br />
~~<br />
" ".~<br />
, .<br />
..<br />
,~.-.-.<br />
-----<br />
• AE<br />
• AF<br />
60~ __________ ~ __________ ~~ __________ ~<br />
.... AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-39. Mean performance on ANAMUKR: DGS-ACC.<br />
78
MSP-ACC<br />
t;<br />
w<br />
0:::<br />
0:::<br />
o<br />
()<br />
"#.<br />
z<br />
U)<br />
~<br />
100~---------------------------------------------------~<br />
90<br />
80<br />
70<br />
.........................................<br />
-"'~~<br />
~ .........<br />
" ..............-.-.-<br />
" " , " ...... -<br />
1I ...... -<br />
GROUP<br />
• AE<br />
'-___---------1· AF<br />
60~---------------~---------------~---------------~<br />
1995 1996 1997 1998<br />
... AG<br />
YEARS<br />
Figure 3-40. Mean performance on ANAMUKR: MSP-ACC.<br />
SPD-ACC<br />
100~---------------------------------------------------~<br />
90<br />
.'-, .<br />
., . - .--_<br />
.A.-----<br />
............ ~ ..........................<br />
80<br />
ROUP<br />
• AE<br />
• AF<br />
70~ _______________ ~ _______________ ~ _______________ --4 ... AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3- 41. Mean performance on ANAMUKR: SPD-ACC.<br />
79
Z<br />
~ -..<br />
t;<br />
LU<br />
c::<br />
c::<br />
0<br />
(,)<br />
=*I:<br />
~<br />
~<br />
SRT-EFF<br />
.........................................<br />
---.--. ,-<br />
~ ,,'"<br />
170~--------------------------------------------------------~<br />
160<br />
150<br />
140<br />
130<br />
120<br />
110<br />
100<br />
90<br />
80<br />
70<br />
1995<br />
1996<br />
--.. -<br />
~.~.<br />
--"- "<br />
GROUP<br />
...----...·AE<br />
11-__ •<br />
--- AF<br />
... AG<br />
1997 1998<br />
YEARS<br />
Figure 3-42. Mean performance on ANAMUKR: SRT -EFF.<br />
110<br />
Z<br />
~ 100<br />
~<br />
90<br />
~<br />
0<br />
(,)<br />
80<br />
'*I:<br />
~ 70<br />
~<br />
2CH-EFF<br />
120~---------------------------------------------------~<br />
.. ~ ............................................ .<br />
.. """ ........<br />
"-..<br />
'. ,.<br />
, ."",--<br />
.. , .... ,.<br />
~ - .... ......... GROUP<br />
....... -<br />
----- • AE<br />
~ • AF<br />
50 .... AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-43. Mean performance on ANAMUKR: 2CH-EFF.<br />
80
CDS-EFF<br />
50~------------------------------------~<br />
z<br />
~ 40 .., §<br />
........<br />
O~<br />
30<br />
()<br />
~<br />
~<br />
-- ....... '..,e--._._....-<br />
.. .............<br />
• •<br />
"<br />
'....<br />
GROUP<br />
::2: 20 • AE<br />
• AF<br />
10~ __________ ~ ____________ ~ __________ -4 A AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-44. Mean performance on ANAMUKR: CDS-EFF.<br />
CDI-EFF<br />
40~------------------------------------~<br />
"' ..<br />
.... .,<br />
....... ,<br />
'....... . ..--<br />
-... ~.-.-.~-.--<br />
'-.... ........ ........<br />
.....<br />
........<br />
..........<br />
............. • AE<br />
GROUP<br />
• AF<br />
10~ __________ ~ __________ ~~ __________ 4 A AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-45. Mean performance on ANAMUKR: CDI-EFF.<br />
81
CDD-EFF<br />
50~--------------------------------------------------------,<br />
Z 40<br />
~<br />
W<br />
0:::<br />
0::: 30<br />
o<br />
=*I:<br />
~<br />
~ 20<br />
.........................................<br />
"<br />
, ..<br />
".<br />
,, '.<br />
,<br />
~--.<br />
---...------<br />
, , --_ ......<br />
.--- --<br />
• AE<br />
~-.. .... AF<br />
10~ _______________-,__________________ ~ _______________ ~ ... AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-46. Mean performance on ANAMUKR: CDD-EFF.<br />
CPT-EFF<br />
100~------------------------------------------------------~<br />
90<br />
80<br />
70<br />
. --..,._ .......<br />
• ~.
DGS-EFF<br />
50~------------------------------------------------------~<br />
Z 40<br />
~ . ~."': ~ ......................................<br />
-----... ~ "'-'~- _<br />
0::: .... 'It----L.<br />
0::: ..-.- ~ ........<br />
o 30 - - ::::-.....<br />
_ .. ..<br />
.. ()<br />
'*I: ---... -.:. _ GROUP<br />
~ ---<br />
~ 20 ~----_.J.. •<br />
AE<br />
• AF<br />
10~ _______________ ~ __________________ ~ _______________ -4 • AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-48. Mean performance on ANAMUKR: DGS-EFF.<br />
z<br />
~<br />
hl 30<br />
~<br />
()<br />
'*I:<br />
~ 20<br />
:2<br />
MSP-EFF<br />
..........................................<br />
40~-------------------------------------------------------,<br />
, -_.-4 ......<br />
, -.. .~<br />
, -- ,~<br />
....-....-- .<br />
, ...... ......... ......<br />
~<br />
, , , -, .. ,<br />
'---- ~<br />
,.- ' .... GROUP<br />
' ....<br />
~-..-..-..-..-..-e~ __________ I·AE<br />
• AF<br />
10~ _______________ ~ __________________ ~ _______________-4<br />
& AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-49. Mean performance on ANAMUKR: MSP-EFF.<br />
83
SPD-EFF<br />
40~------------------------------------,<br />
z<br />
~ ,<br />
j:: 30 ,<br />
() ,<br />
w •••• ~ ••••••••••••••••••••••••••••••• ..........<br />
~ ~-<br />
,<br />
~ .-~-. .-...-. ",-<br />
O<br />
()<br />
'<br />
,<br />
.......-.,.-.<br />
'*I: ,<br />
~ 20"' ___ ~",,-=-:-:::::-:~<br />
:2<br />
• AF<br />
10 ... AG<br />
~----------~----------~------------4<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-50. Mean performance on ANAMUKR: SPD-EFF.<br />
TAPPING-RIG HT<br />
S2~------------------------------------~<br />
so<br />
,.,. ....<br />
58 ~ ..<br />
••• • • •• •• • ••• ••• • ••• •• •• ••• • •• -:..;..r-~•••••••<br />
~ 56<br />
-M-"<br />
/_.-.--<br />
~ 54<br />
_ -<br />
#<br />
'*I: #/<br />
~ 52 ~#/<br />
itJ, _ .... ---__ GROUP<br />
:2 50)./. .._ .... - _<br />
~ --1~.<br />
48 .". .... - ~ ..__<br />
AE<br />
....-..........-----·-------T..<br />
~~--<br />
---<br />
4S I '" • AF<br />
44~----------__ ------------__ ----------_4 A AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-51. Mean performance on ANAMUKR: TAPPING-RIGHT.<br />
84
TAPPING-LEFT<br />
54~-----------------------------------.<br />
52<br />
................................... ..•. ,;,;. ..<br />
50<br />
------<br />
~ 48 .-.Jit---<br />
I-<br />
=II: 46<br />
.<br />
_ .. ..ttl!"./ '"<br />
11--<br />
~<br />
~ / ---...<br />
:2: 44 ~<br />
GROUP<br />
,<br />
42<br />
40<br />
~<br />
~<br />
, • AE<br />
~~,:-................ """" ........ ~----------~. AF<br />
38~ __________ ~ __________ -r __________ -4 .6. AG<br />
1995 1996 1997 1998<br />
YEARS<br />
Figure 3-52. Mean performance on ANAMUKR: TAPPING-LEFT.<br />
SLEEPINESS<br />
UJ<br />
~<br />
OC/)°<br />
~<br />
:2:<br />
2.0<br />
1.5<br />
, •<br />
, , •<br />
,, .,<br />
,, ., ------<br />
~--<br />
, . ~<br />
, . --<br />
~<br />
, .... ~.-.~--.<br />
........, , .. ......... ".. ..................<br />
... ~ ........<br />
~~ ..........<br />
~ --<br />
1.0 ....------------.------------r-------------j,<br />
1995 1996 1997 1998<br />
• AE<br />
• AF<br />
.6. AG<br />
YEARS<br />
Figure 3-53. Mean ratings on ANAMUKR: SLEEP SCALE.<br />
85
SECTION 4<br />
CONCLUSIONS<br />
Taken collectively, the results of the data analyses are rather frightening. Initial dosages were from 1 rad<br />
to 183 rads. Our research suggests neurocognitive and physical decrements in performance 12 years<br />
AFTER a nuclear accident. They indicate that not only have cognitive and physical functioning of the<br />
AEs been severely compromised by exposure to the environmental effects of the ionizing radiation being<br />
emitted by the Chemobyl power station, but also that cognitive and physical performance of the other<br />
groups (AFs, AGs) in the vicinity of Chemobyl have been affected as well, although to a lesser degree.<br />
At this point it is hard to specify whether the observed impairments are permanent or temporary<br />
(although they would appear to be permanent in the AEs). It is also difficult to determine whether they<br />
are a result of direct effects of radiation on the CNS itself, as opposed to being an indirect result of<br />
bodily illness resulting from chronic radiation perhaps paralleling that experienced by people diagnosed<br />
as suffering from reporting heat exhaustion (Gestaldo, et. al., 1997). It is also possible that the<br />
individuals in and around the Chemobyl area are experiencing symptoms somehow similar to those of<br />
the "neurasthenic syndrome," which was reported by Soviet workers exposed to non-ionizing microwave<br />
radiation in the 1980's. As described by Akoyev and Justesen (personal communication), this syndrome<br />
included fatigue, malaise, and achiness. Such factors would most certainly result in compromised<br />
performance on neurocognitive tasks requiring attention and working memory, and on physical tasks<br />
requiring explosive and sustained energy. Since this study was not a "medical" study, we were not<br />
equipped to identify aplastic anemia (pancytopenia).<br />
The results of the 1996 retest indicated that both the physical and cognitive abilities of the individuals<br />
initially exposed (Eliminators) to the ionizing radiation resulting from the Chemobyl nuclear accident<br />
were seriously declining. Although the 1997 retest indicated that the forestry and agricultural workers<br />
were actually improving on the physical tasks, the cognitive performance of these groups was becoming<br />
globally impaired; i.e., they showed significant impairments on the majority of the ANAMUKR tasks.<br />
This global impairment is reminiscent of that observed in survivors of moderate-to-severe traumatic<br />
brain injuries (Levinson & Reeves, 1997; Levinson, et al, 1998). Unlike those people, however, whose<br />
cognitive performance improved over time, these individuals continued to experience increasing<br />
difficulty in neurocognitive function. As of eleven years after the accident, they continued to decline<br />
and had not yet plateaued. The results of the 1998 retest indicate that the declines of the exposure<br />
groups appeared to be leveling off, and improvements in performance were observed in some cases.<br />
Nevertheless, the results of analyses of the 4-year averaged scores indicate that the effects of exposure to<br />
radionuclides in and around the area of Chemobyl have resulted in clinically meaningful impairments in<br />
both physical and cognitive performance. Further, the fmding that significant correlations between<br />
dosage and 4-year averaged performance occurred on 21 of24 tasks for the combined exposure groups is<br />
extremely disconcerting. Retests performed during the next several years will be extremely valuable in<br />
determining whether the physical and neurocognitive performance of these individuals resumes to<br />
decline, continues to plateau, or begins to improve.<br />
There are in the literature, several views on low-dosage radiation. One view holds that low-dosage does<br />
not cause an increase in cancer. However, our end was "performance" and physical and<br />
neuropsychological performance, according to our study, has been severely compromised. We believe it<br />
is time to take into account performance as well as health consequences.<br />
86
Ukrainians have suffered serious medical problems as a result of the <strong>Chernobyl</strong> disaster, and their<br />
country has suffered economically since the accident. The burden of medical treatment is enormous and<br />
containment of nuclear pollution is almost impossible. Ukrainian scholars and scientists are generally in<br />
agreement that <strong>Chernobyl</strong> was one of the causes for Ukrainian independence from the former Soviet<br />
Union. For Russia the economic expenditures for environmental clean-up and treatment of the<br />
population were prohibitive.<br />
In 1991, the Ukrainian Supreme Soviet enacted a law, which would decommission <strong>Chernobyl</strong> by the end<br />
of 1993; however, in October 1993, this law was repealed. The cost of decommissioning <strong>Chernobyl</strong>,<br />
especially in light of obtaining alternate energy sources, remains too great. The Chairman of the<br />
Ukrainian Supreme Soviet Committee on <strong>Chernobyl</strong>, Volodymyr Javorivsky said, "We will be<br />
extracting problems from the well of<strong>Chernobyl</strong> for a very long time." In light of this report, that is an<br />
understatement.<br />
To put this report into some perspective it needs to be stated that while contamination of the ecostructure<br />
is considerable, much of the contamination affecting the Ukraine is below international<br />
standards for life-time dosages. Poor diet, the stress of living in a country where the minimum wage is<br />
$1.50 per month, and where food is a daily preoccupation, are not adequately addressed in medical<br />
reports which blame <strong>Chernobyl</strong> for all increases in disease. Yet research has shown that exposure to<br />
ionizing radiation is harmful to humans and the environment. The effects of living in contaminated<br />
areas, as well as growing and consuming contaminated food, have received little research support. The<br />
psychological concomitants of Post-Traumatic Stress Disorder are being considered in terms of recent<br />
child development studies in the Ukraine.<br />
The Ukrainians are desirous for research support. The Ministry of Forests is particularly interested in<br />
research support to develop models for cognitive and physical decrements in performance associated<br />
with forestry workers exposed to contaminated forests. The Director of the Ukrainian Psychological<br />
Research Institute is also desirous of research support dealing with psycho-sociological problems<br />
associated with adults and children living in contaminated areas, or who have relocated from these areas.<br />
The need for assistance to help plan and guide longitudinal research has been expressed by those who<br />
have been working with the children affected by the <strong>Chernobyl</strong> disaster.<br />
The benefits of these research proposals would be enormous for the United States. <strong>Nuclear</strong> energy is a<br />
fact worldwide. Data gathered by research efforts involving <strong>Chernobyl</strong> could impact Federal<br />
Emergency Management Administration procedures for relocating victims of similar accidents, and their<br />
medical and psychological treatment. Unfortunately, in the conduct of this research, a wealth of data<br />
was obtained that is available, however, funds were not allocated for analyses. In addition, little is<br />
known about cognitive and physical decrements in performance or stress associated with living and<br />
working in contaminated areas. Preparedness should be the bottom line in research.<br />
Read (1993) mentions that because radiation technology was so new, there was no way to be prepared<br />
for all possible problems resulting from its use. Nonetheless, it is important that people learn from such<br />
accidents about the effects of radiation, especially considering the aftermath of nuclear war or terrorist<br />
attack. This remains one of the objectives of the present, ongoing study of the physical and<br />
neurocognitive effects of the <strong>Chernobyl</strong> accident.<br />
87
SECTION 5<br />
REFERENCES<br />
Awramenko, O. (1992). Health is deteriorating: Problems remain. Your Health, 9, 23.<br />
(UNCLASSIFIED)<br />
Baranov, A.E. & Guskova, A. K. (1988). Acute Radiation Disease in <strong>Chernobyl</strong> <strong>Accident</strong>s Victims.<br />
Paper presented at the Institute of Biophysics, Ministry of Health of the USSR, Moscow, USSR.<br />
(UNCLASSIFIED)<br />
Baryahtar, V. & Bobyleva, O. (1991). Generalized Data of the Academy of Sciences of the<br />
Ukraine and the Ministry of Health Characterizing the Consequences of the <strong>Chernobyl</strong> Disaster with<br />
Respect to the Health ofthe Population. Ukrainian Ecological Bulletin, 4. (UNCLASSIFIED)<br />
Chernousenko, V. M. (1991). <strong>Chernobyl</strong>: Insightfrom the inside. Berlin: Springer-Verlag.<br />
(UNCLASSIFIED)<br />
Fleishman, E.A. & Quaintance, M.K. (1984). Taxonomies of human performance: The<br />
description of human tasks. Orlando, FL: Academic Press. (UNCLASSIFIED)<br />
Gamache, G. (1993). Gamache physical abilities battery. (Report No. KGAI-93-97-TR).<br />
St. Augustine, FL: KGA International. (UNCLASSIFIED)<br />
Gastaldo, E. Reeves, D. Levinson, D., & Winger, B. (1996). ANAM USMC normative data, Series I: The<br />
effects of hyponatremia on neurocognitivefunction. (Report No. NCRF-TR-96- 01). San, Diego, CA:<br />
National Cognitive Recovery Foundation. (UNCLASSIFIED)<br />
Gittus, J., Hicks, D., Bonnell, P., Clough, P., Dunbar, I., Egan, M., Hall, A., Hayns, M., Nixon, W.,<br />
Bullock, R., Luckhurst, D., Maccabee, A., and Edens, D. (1988). The <strong>Chernobyl</strong> <strong>Accident</strong> and its<br />
Consequences. London: Atomic Energy Authority. (UNCLASSIFIED)<br />
Goldstone, A., Reeves, D., Levinson, D.M., & Pelham, M. (1995). Effects of stroke on cognitive<br />
performance as assessed by Automated Neuropsychological Assessment Metrics (ANAM V3.11 a).<br />
Report No. NCRF-95-01). San Diego, CA: National Cognitive Recovery Foundation.<br />
(UNCLASSIFIED)<br />
Guskova, A. et al. (1988). Medical Aspects of the <strong>Chernobyl</strong> <strong>Accident</strong>. Proceedings of All-Union<br />
Conference organized by the USSR Ministry of Health and the All-Union Scientific Centre of Radiation<br />
Medicine, USSR Academy of Medical Sciences, held in Kiev, Ukraine, 11-13 May, 1988.<br />
(UNCLASSIFIED)<br />
Kay, G. (1995). Cogscreen: Aeromedical edition; professional manual. Odessa, FL: Psychological<br />
Assessment Resources, Inc. (UNCLASSIFIED)<br />
88
Kopeikin, V. (January, 1993). Answersfrom Professor Kopeikinfrom Kazan. Vestnik <strong>Chernobyl</strong>ia<br />
9 (442). (UNCLASSIFIED)<br />
Laupa, A. & Anno, G. (1989). <strong>Chernobyl</strong> <strong>Accident</strong> Fatalities and Causes. Pacific-Sierra Research<br />
Corporation Technical Report 1865 prepared for the <strong>Defense</strong> <strong>Nuclear</strong> Agency, Washington, D.C.<br />
(UNCLASSIFIED)<br />
Levinson, D.M. (1990, August). Normative data on a subset of the Unified Tri-Service Cognitive<br />
Performance Assessment Rattery (UTC-PAR). Presented at the 98th American Psychological<br />
Association Annual Convention, Boston, MA. (UNCLASSIFIED)<br />
Levinson, D. M., & Reeves, D. L.(1994). Automated Neuropsychological Assessment Metrics<br />
(ANAM): ANAMVi.ONormative Data. (Report No. NCRF-TR-94-01). San, Diego, CA: National<br />
Cognitive Recovery Foundation. (UNCLASSIFIED)<br />
Levinson, D.M., & Reeves, D.(1997). Monitoring recovery from traumatic brain injury using<br />
Automated Neuropsychological Assessment Metrics (ANAM Vl.O). Archives of Clinical<br />
Neuropsychology, 12, 155-r66. (UNCLASSIFIED)<br />
Levinson, D.M, Reeves, D.L., Wild, M.R, & Lewandowski, A.G. (1998). Classifying level of<br />
neurocognitive impairment in individuals with acquired brain injury. Archives of Clinical<br />
Neuropsychology, 13, 73. (UNCLASSIFIED)<br />
Lewandowski, A.G., Reeves, D.L., & Dietz, A.J. (1994, April). Assessing the pharmacodynamics of<br />
CNS compounds through repeated neuropsychological measures under clinical trial conditions.<br />
Presented at the 14th Frontiers Symposium of the American College of Clinical Pharmacology,<br />
Rockville, MD. (UNCLASSIFIED)<br />
McClellin, G.E., Anno, G.H., Whicker, F.W. (1994). <strong>Chernobyl</strong> Doses Volume i-Analysis of Forest<br />
Canopy Radiation Response from Multispectral Imagery and the Relationship to Doses. (Technical<br />
Report DNA 001-87-C-0104). Washington DC <strong>Defense</strong> <strong>Nuclear</strong> Agency. (UNCLASSIFIED)<br />
Read, P.P. (1993). Ablaze: The story of the heroes and victims of<strong>Chernobyl</strong>. New York: Random<br />
House. (UNCLASSIFIED)<br />
Reeves, D. & Gamache, G. (1994). Ukrainian subset of ANAM battery: ANAMUKR. (Report No.<br />
KGAI-94-106-TR). St. Augustine, FL: KGA International. (UNCLASSIFIED)<br />
Reeves, D., Kane, R., Winter, K., Raynsford, K., & Pancella, T. (1993). Automated<br />
NeuropsychologicalAssessment Metrics (ANAM): Test administrator's guide Version i.O. St. Louis:<br />
Missouri Institute of Mental Health. (UNCLASSIFIED)<br />
Reeves, D., Kane, R., & Winter, K. (1995). Automated Neuropsychological Assessment Metrics<br />
(ANAM): Test administrator's guide Version 3.11. (Report No. NCRF-SR-95-01). San Diego, CA:<br />
National Cognitive Recovery Foundation. (UNCLASSIFIED)<br />
89
Reeves, D., Schlegel, R, Gilliland, K., & Crabtree, M. (1991). UTC-PAB and the NATO/AGARD<br />
STRES Battery: Results from standardization studies. Proceedings of the 1991 Medical <strong>Defense</strong><br />
Bioscience Review. Aberdeen Proving Grounds, MD. (UNCLASSIFIED)<br />
Reeves, D.L., Thome, D.R, Winter, S.L., & Hegge, F.W. (1989). The Unified Tri-Services<br />
CognitivePerformance Battery (UTCPAB): II Hardware and software design and specifications. (Report<br />
No. SR 89-1). Pensacola, FL: Naval Aerospace Medical Research Laboratory. (UNCLASSIFIED)<br />
Reeves, D.L. & Winter, K.P. (1992). ANAM documentation volume I: Test administrator's<br />
guide. Washington, DC: Office of Military Performance Technology, Scienceboard. (UNCLASSIFIED)<br />
Sorodovych, E., Dudkin, V., Kaletnik, N., Los, I., Guclzenko, V., Bobyleva, D., & Tabachny, L.<br />
(1992). <strong>Chernobyl</strong> Catastrophe: Reasons and Consequence. Kiev, Ukraine: Academy of Science.<br />
(UNCLASSIFIED)<br />
Thome, D.R (1990). Throughput: a simple performance index with desirable characteristics.<br />
Washington, DC: Office of Military Performance Technology, Scienceboard. (UNCLASSIFIED)<br />
Thome, D., Genser, S., Sing, H., & Hegge, F. (1985). The Walter Reed Performance Assessment<br />
Battery. Neurobehavioral Toxicology and Teratology, 7,415-418. (UNCLASSIFIED)<br />
Vestnik Chemobylia. (May, 1993). Energy Generated by the Ukraine. 37 (472). (UNCLASSIFIED)<br />
Woytsehowich, O. (1991). The Problems of Radioactive Pollution of the Dnipro River System After<br />
Five Years of the Chemobyl Catastrophe. Ukrainian Ecological Bulletin, 5. (UNCLASSIFIED)<br />
Yakovlev, E. A. (1991). Geological and Economical Aspects of Decommissioning the Chemobyl<br />
<strong>Nuclear</strong> Power Plant. Ukrainian Ecological Bulletin, 2. (UNCLASSIFIED)<br />
Yakovlev, E.A. (1992). We are hewing the bough. Rabochaya Gaseta (August 12). (UNCLASSIFIED)<br />
90
APPENDIX A<br />
UKRAINIAN PROJECf DEPARTMENTS AND PERSONNEL<br />
CONTACTED<br />
Ukrainian Psychological Institute<br />
2 Pankovskaya Street<br />
Kiev, Ukraine 252033<br />
Director, Professor A. K. Wasilewich<br />
Director of the Children's Center for <strong>Chernobyl</strong>, Dr. S. Yakovenko<br />
Project Manager, Adults of <strong>Chernobyl</strong>, Dr. Y. Sewalb<br />
Child Psychologist, Natalia A Bastun<br />
Director, Ukrainian Guidance Center, Dr. V. Panok<br />
Institute for Psychiatry<br />
103 Frunze Street<br />
Kiev, Ukraine 252053<br />
Director, A. P. Chuprinov, M.D.<br />
Ministry of Forests<br />
5 Kreshechatik Street<br />
Kiev, Ukraine 252601<br />
Head of the Scientific and Technical Office, N. Kaletnik<br />
Minister of <strong>Chernobyl</strong> Affairs<br />
1 Lesi Ukrainki Square<br />
Kiev, Ukraine 252196<br />
Chiet: Protection of the Population <strong>Department</strong>, L. Tabachny<br />
First, Vice-Minister, Boris S. Prister<br />
Director, International Cooperation, Y. Pavlov<br />
<strong>Department</strong> Head, Statistics, N. Repina<br />
Shevchenko University<br />
64 VJadimirskaya Street<br />
Kiev, Ukraine 252601<br />
Chairman, Psychodiagnostics and Medical Psychology, Prof L. Burlatchuk<br />
First, Vice-Rector, Prof O. Tretyak<br />
Vice-Rector of Education, Prof L. Gubersky<br />
A-I
Ukrainian Academy of Sciences<br />
Dr. V. V. Gudzenko<br />
Dr. E. V. Sobodovych<br />
A-2
APPENDIXB<br />
PHOTOGRAPHSOFCHERNOBYL<br />
AND<br />
TESTING SITES<br />
Figure B-1. ChernobyI site after explosion, 26 April 1986.<br />
B-1
Figure B-2. Russian monitors, 26 April 1986.<br />
B-2
Figure B-3. Setting up to test eliminators.<br />
B-3
Figure B-4. Agricultural workers ready for testing on At~AMUKR<br />
B-4
Figure B-S. Agricultural workers walking on balance beam.<br />
B-5
Figure B-6. Agricultural workers performing broad jump.<br />
B-6
Figure B-7. Control person carrying weights.<br />
B-7
Figure B-8. Forestry workers performing ANAM.<br />
B-8
Figure B-9. Dr. Gamache at <strong>Chernobyl</strong> <strong>Nuclear</strong> Power Plant, 1997.<br />
B-9/B-I0
APPENDIXC<br />
CONSENT FORM<br />
Dear<br />
----------------------------------------------------------------<br />
You are invited to participate as a VOLUNTEER in a 4-year MINIMAL RISK research project<br />
designed to test your physical and cognitive perfonnance. This research is conducted by KGA<br />
International, Kiev Polytechnic Institute, and the Ukraine Center for Radiation Medicine. The research<br />
is planned for 1995-1998, and the testing will be carried out in the summer months and will require you<br />
to be tested, as scheduled. Complete testing will take one-half day.<br />
Prior to testing you will be provided with instructions how to perfonn the tests. The physical<br />
part of testing is based on simple exercises that are easy to perfonn for any individual. The cognitive<br />
test will be perfonned on a computer with appropriate instructions in Russian.<br />
If, during the testing, you feel ill or want to ask a question, notify your instructor immediately.<br />
If you are ill, the instructor will refer you to the medical staff. Inquries regarding instructions given<br />
MAY necessitate starting testing over again. Make sure to ask ALL questions prior to commencing the<br />
test procedure.<br />
I, certify that I am a volunteer and all procedures and risks have been thoroughly explained to<br />
me. I also have the choice NOT to participate at any time.<br />
Signature __________________________________ _<br />
Dare _________________ _<br />
C-lIC-2
APPENDIXD<br />
GLOSSARY<br />
2CH<br />
AC<br />
ACC<br />
AE<br />
AF<br />
AG<br />
ANAM<br />
ANAM-ACC<br />
ANAM-EFF<br />
ANAMUKR<br />
BALBEAM<br />
BROADJMP<br />
CARRYWGT<br />
CDD<br />
CDI<br />
CDS<br />
COMP<br />
CPT<br />
DECL<br />
DECR<br />
DGS<br />
EFF<br />
GPAB<br />
MSP<br />
SLP<br />
SPD<br />
SQUAI"I'HR<br />
SRT<br />
TAP-L<br />
TAP-R<br />
Twq-choice Reaction Time<br />
Control group<br />
Accuracy<br />
Eliminator group<br />
Forester group<br />
Agricultural group<br />
Automated Neuropsychological Assessment Matrices<br />
ANAMUKR - accuracy scores<br />
ANAMUKR - efficiency sores<br />
Special subset of ANAM created for this study<br />
Balance Beam<br />
Broad jump<br />
Carrying weights<br />
Code Substitution - delayed recall<br />
Code Substitution - immediate recall<br />
Code Substitution - visual search<br />
Composite measure<br />
Running Memory Continuous Perfonnance Task<br />
Decline<br />
Decrement<br />
Digit Symbol<br />
Efficiency<br />
Gamache Physical Abilities Battery<br />
Matching to Sample<br />
Stanford Sleepiness Scale<br />
Spatial Processing<br />
Squat thrusts<br />
Simple Reaction Time<br />
Tapping -left index finger<br />
Tapping - right index finger<br />
D-lill-2
DISTRIBUTION LIST<br />
DEPARTMENT OF DEFENSE<br />
DIRECTOR<br />
ARMED FORCES RADIOBIOLOGY RSCH INST<br />
INFORMATION SERVICES DEPARTMENT<br />
8901 WISCONSIN AVENUE<br />
BETHESDA, MD 20889-5603<br />
ATTN: BHS<br />
ATTN: EXH<br />
ATTN: MRA<br />
ATTN: RSD<br />
ATTN: TECHNICAL LIBRARY<br />
DEFENSE TECHNICAL INFORMATION CENTER<br />
8725 JOHN J. KINGMAN ROAD, SUITE 0944<br />
FORT BELVOIR, VA 22060-6218<br />
ATTN: DTIC/OCP<br />
DEFENSE THREAT REDUCTION AGENCY<br />
6801 TELEGRAPH ROAD<br />
ALEXANDRIA, VA 22310-3398<br />
ATTN: TDA,C.MCFARLAND<br />
ATTN: TDAC<br />
ATTN: TDACC<br />
ATTN: TDANP, TRC<br />
DEFENSE THREAT REDUCTION AGENCY<br />
ALBUQUERQUE OPERATIONS<br />
1680 TEXAS STREET, SE<br />
KIRTLAND AFB, NM 87117-5669<br />
ATTN: TDTO<br />
DIRECTOR<br />
DISA<br />
TECHNICAL RESOURCES CENTER<br />
7010 DEFENSE PENTAGON<br />
WASHINGTON, DC 20301-7010<br />
ATTN: JNGO<br />
PRES~DENT<br />
NATIONAL DEFENSE UNIVERSITY<br />
FORT LESLEY J. MCNAIR<br />
WASHINGTON, DC 20319-6000<br />
ATTN: ICAF, TECH LIB<br />
ATTN: LIBRARY, STOP 315<br />
DIRECTOR<br />
NET ASSESSMENT<br />
OFFICE OF THE SEC OF DEFENSE<br />
ROOM 3A930, THE PENTAGON<br />
WASHINGTON, DC 20301<br />
ATTN: DOCUMENT CONTROL<br />
HEADQUARTERS<br />
US EUROPEAN COMMAND<br />
UNIT 30400 BOX 1000<br />
APOAE 09128<br />
ATTN: ECJ5-N<br />
USSTRATCOMlJ531T<br />
901 SAC BOULEVARD, SUITE-BB11<br />
OFFUTT AFB, NE 68113-5160<br />
ATTN: J-521<br />
DEPARTMENT OF DEFENSE CONTRACTORS<br />
INSTITUTE FOR DEFENSE ANALYSES<br />
CONTROL & DISTRIBUTION<br />
1801 N. BEAUREGARD STREET<br />
ALEXANDRIA, VA 22311<br />
ATTN: D. SCHULTZ<br />
ITT INDUSTRIES<br />
ITT SYSTEMS CORPORATION<br />
ATTN: AODTRAlDTRIAC<br />
1680 TEXAS STREET, SE<br />
KIRTLAND AFB, NM 87117-5669<br />
ATTN: DTRIAC<br />
ATTN: DTRIAC/DARE<br />
JAYCOR<br />
1410 SPRING HILL ROAD, SUITE 300<br />
MCLEAN, VA 22102<br />
ATTN: DR C. P. KNOWLES<br />
PACIFIC-SIERRA RESEARCH<br />
OPERATING COMPANY OF VERID<br />
29801 28TH STREET, 2ND FLOOR<br />
SANTA MONICA, CA 90405<br />
ATTN: G. ANNO<br />
ATTN: H. BRODE<br />
ATTN: L. MATHESON<br />
PACIFIC-SIERRA RESEARCH CORPORATION<br />
WASHINGTON OPERATIONS<br />
1400 KEY BOULEVARD, SUITE 700<br />
ARLINGTON, VA 22209<br />
ATTN: G. MCCLELLAN<br />
PSYCHOLOGICAL RESEARCH CENTER<br />
8 ALTHEA STREET<br />
ST AUGUSTINE, FL 32095<br />
ATTN: D. L. REEVES<br />
ATTN: D. M. LEVINSON<br />
ATTN: L. GAMACHE<br />
ATTN: P. I. BIDIOUK<br />
DL-1
SCIENCE APPLICATIONS INT'L CORPORATION<br />
10260 CAMPUS POINT DRIVE<br />
SAN DIEGO, CA 92121-1578<br />
ATTN: D. KAUL, MS-33<br />
ATTN: E. SWICK, MS-33<br />
UNIVERSITY OF CINCINNATI MEDICAL CENTER<br />
MAIL LOCATION 577<br />
CINCINNATI, OH 45267<br />
ATTN: E. SILBERSTEIN<br />
DEPARTMENT OF ENERGY<br />
UNIVERSITY OF CALIFORNIA<br />
LAWRENCE LIVERMORE NATIONAL LAB<br />
P. O. BOX 808<br />
LIVERMORE, CA 94551-9900<br />
ATTN: Z. DIVISION LIBRARY<br />
SANDIA NATIONAL LABORATORIES<br />
ATTN: MAIL SERVICES<br />
P. o. BOX 5800<br />
ALBUQUERQUE, NM 87185-0459<br />
ATTN: TECH LIB, MS-0899<br />
DEPARTMENT OF THE AIR FORCE<br />
AIR UNIVERSITY LIBRARY<br />
600 CHENNAULT CIRCLE<br />
BUILDING 1405 - ROOM 160<br />
MAXWELL AFB, AL 36112-6424<br />
ATTN: AUL-LSE<br />
ATTN: LIBRARY<br />
ATTN: STRATEGIC STUDIES<br />
ASSISTANT CHIEF OF STAFF<br />
STUDIES AND ANALYSES AGENCY<br />
1570 AIR FORCE PENTAGON<br />
WASHINGTON, DC 20330<br />
ATTN: AFSAAISAMI<br />
DIRECTORATE OF NUCLEAR<br />
AND COUNTERPROLIFERA TION<br />
DEPARTMENT OF THE AIR FORCE<br />
WASHINGTON, DC 20330<br />
ATTN: XOOSS<br />
US AIR FORCE ACADEMY<br />
2354 FAIRCHILD DRIVE, SUITE 3A22<br />
USAF ACADEMY, CO 80840-6214<br />
ATTN: LIBRARY<br />
DEPARTMENT OF THE ARMY<br />
DEPARTMENT OF THE ARMY<br />
DEPUTY CHIEF OF STAFF FOR<br />
OPERATIONS AND PLANS<br />
THE PENTAGON<br />
WASHINGTON, DC 20310-0460<br />
ATTN: DAMO-SWN<br />
ATTN: DAMO-ZXA<br />
US ARMY MEDICAL RESEARCH &<br />
DEVELOPMENT COMMAND<br />
FORT DETRICH, BUILDING 722<br />
FREDRICK, MD 21702-5012<br />
ATTN: SGRD-PLE<br />
US ARMY MODEL IMPROVEMENT STUDY<br />
MANAGEMENT AGENCY<br />
CRYSTAL SQUARE 2<br />
1725 JEFFERSON DAVIS HIGHWAY SUITE 808,<br />
ARLINGTON, VA 22202<br />
ATTN: SFUS-MIS<br />
COMMANDER<br />
US ARMY NUCLEAR & CHEMICAL AGENCY<br />
7150 HELLER LOOP, SUITE 101<br />
SPRINGFIELD, VA 22150-3198<br />
ATTN: MONA-NU<br />
US ARMY RESEARCH LAB<br />
5232 FLEMING ROAD<br />
ABERDEEN PROVING GROUND, MD 21001-5067<br />
ATTN: DIRECTOR<br />
ATTN: SLCBR-D<br />
ATTN: SLCBR-DD-T, TECH LIB<br />
ATTN: SLCBR-TB<br />
US ARMY RESEARCH LABORATORIES<br />
2800 POWDER MILL ROAD<br />
ADELPHI, MD 20783-1197<br />
ATTN: TECH LIB<br />
COMMANDER<br />
US ARMY TEST & EVALUATION COMMAND<br />
BUILDING 860<br />
ABERDEEN PROVING GROUND, MD 21005<br />
ATTN: STECS-NE,70<br />
DIRECTOR<br />
US ARMY VULNERABILITYILETHALITY<br />
ASSESSMENT MANAGEMENT OFFICE<br />
ABERDEEN PROVING GROUND, MD 21005-5001<br />
ATTN: AMSLC-VL-NE, DR J. FEENEY<br />
DL-2
SUPERINTENDENT<br />
US MILITARY ACADEMY<br />
WEST POINT, NY 10996<br />
ATTN: DEPT OF BEHAVORIAL<br />
SCI & LEADERSHIP<br />
DIRECTOR<br />
WALTER REED ARMY INSTITUTE OF RESEARCH<br />
6825 16TH STREET, NW<br />
WASHINGTON, DC 20307-5100<br />
ATTN: SLRD-UWI<br />
DEPARTMENT OF THE NAVY<br />
PRESIDENT<br />
NAVAL WAR COLLEGE<br />
686 CUSHING ROAD<br />
NEWPORT, RI 02841-1207<br />
ATTN: LIBRARY, CODE 1E22<br />
OTHER GOVERNMENT<br />
NATIONAL ARCHIVES & RECORDS<br />
ADM I NISTRA TION<br />
8601 ADELPHI ROAD, ROOM 3360<br />
COLLEGE PARK, MD 20740-6001<br />
ATTN: USER SERVICE BRANCH<br />
DL-3
SIMULATING WET DEPOSITION<br />
OF RADIOCESIUM<br />
FROM THE CHERNOBYL ACCIDENT<br />
THESIS<br />
Aaron M. Kinser, Captain, USAF<br />
AFIT/GM/ENP/OIM-05<br />
DEPARTMENT OF THE AIR FORCE<br />
AIR UNIVERSITY<br />
AIR FORCE INSTITUTE OF TECHNOLOGY<br />
Wright-Patterson Air Force Base, Ohio<br />
APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED.<br />
I'J<br />
C=»<br />
C=»<br />
~<br />
C=»<br />
-........<br />
«....PrI<br />
C=»<br />
C=»<br />
«....PrI<br />
'-C)<br />
/
The views expressed in this thesis are those of the author and do not reflect the official<br />
policy or position of the <strong>Department</strong> of <strong>Defense</strong> or the United States Government.
AFIT /GM/ENP /01M-5<br />
SIMULATING WET DEPOSITION OF RADIOCESIUM<br />
FROM THE CHERNOBYL ACCIDENT<br />
THESIS<br />
Presented to the Faculty of the <strong>Department</strong> of Engineering Physics<br />
Graduate School of Engineering and Management<br />
of the Air Force Institute of Technology<br />
Air University<br />
In Partial Fulfillment of the<br />
Requirements for the Degree of<br />
Master of Science<br />
Aaron M. Kinser, B.S.<br />
Captain, USAF<br />
March 2001<br />
Approved for public release; distribution unlimited
AFIT /GM/ENP /OlM-5<br />
SIMULATING WET DEPOSITION OF RADIOCESIUM<br />
FROM THE CHERNOBYL ACCIDENT<br />
Aaron M. Kinser, B.S.<br />
Captain, USAF<br />
Approved:<br />
~1f{~<br />
Lt. Col. Michael K. Walters (Chairman)<br />
~<br />
i 2fo..tR 0 I<br />
Date<br />
/2.. JI1~ {J/<br />
Date<br />
Dr. Dennis W. Quinn (Member)<br />
Date
· ----------------------------------------------------------------------------------------~<br />
Acknowledgements<br />
Craig Sloan from AFTAC offered the initial inspiration for this research,<br />
and his continued coordination has been crucial to the work. He also personally<br />
introduced me to Roland Draxler, paving the way for our productive<br />
relationship. The greatest and most generous contributions to this<br />
work came from the lead author of HySPLIT software, Roland Draxler of<br />
NOAA's Air Resources Laboratory. He has convincingly earned the title<br />
of Chief Scientific and Technical Advisor for this thesis. The amount<br />
and variety of work, knowledge, and collaboration he contributed (almost<br />
totally via e-mail) merits serious consideration as co-author. Thank you,<br />
Roland.<br />
Thesis Committee Chairman and inspiring mentor, Lt Col Michael Walters,<br />
laid critical groundwork of scientific principles and theory during my<br />
graduate level studies at AFIT. Major Vince Jodoin helped me catch<br />
the excitement of modeling nuclear events, and quickly got me back on<br />
the track whenever I derailed my "nuclear" train of thought. Dr. Dennis<br />
Quinn introduced sound, fundamental approaches to the data at hand,<br />
and I have fed on his contagious optimism and enthusiasm.<br />
My appreciation goes also to Peter J. Rahe, AFIT Meteorology Laboratory<br />
Technician, for equipment set-up and software installation. I<br />
am grateful to Capt Lisa C. Shoemaker, Air Force weather officer and<br />
AFTAC Meteorologist, for technical advice, data templates, and "been<br />
there" encouragement. Ginger Caldwell from UCAR granted and coordinated<br />
access to ECMWF data at NCAR. Joey Comeaux from NCAR<br />
supplied download and conversion instructions for their mass storage system.<br />
He also assembled a program for reading and processing GRIB format<br />
data. Thanks to the European Center for Medium-range Weather<br />
Forecasting (ECMWF) for reanalyzing and sharing ECMWF model data<br />
with the global weather research community.<br />
Noted editor and scientist, Giovanni Graziani, from the European Commission's<br />
Joint Research Centre (JRC) in Ispra, Italy contributed deposition<br />
data files, key information on handling their contents, the full<br />
Radioactivity Environmental Monitoring (REM) <strong>Chernobyl</strong> database report,<br />
and a handsome atlas of European caesium deposition.<br />
I dedicate this thesis to a hard-working heroine, my beautiful wife,<br />
Aaron M. Kinser<br />
iii
Table of Contents<br />
Page<br />
Acknowledgements<br />
III<br />
List of Figures<br />
vii<br />
List of Tables<br />
xii<br />
Abstract ....... .<br />
Xlll<br />
I.<br />
Introduction .......... .<br />
1.1 Problem and Objective.<br />
1.2 Thesis Organization<br />
1<br />
2<br />
2<br />
II.<br />
Background . . . . . . . . . .<br />
2.1 Background Overview<br />
2.2 The <strong>Chernobyl</strong> <strong>Accident</strong> as a Wet Deposition Case Study<br />
3<br />
3<br />
3<br />
2.3 Weather Patterns During the <strong>Chernobyl</strong> <strong>Accident</strong> . . 5<br />
2.4 Cesium-137 Transport from the <strong>Chernobyl</strong> <strong>Accident</strong> . 5<br />
2.5 Other Long-Range Transport Modeling Exercises 8<br />
2.6 HySPLIT Model Description. . . . . . . . . . . . 11<br />
III.<br />
Methodology . . . . . . . . . . . . . .<br />
3.1 Methodology Chapter Overview .<br />
3.2 Incorporation of Meteorological Input Fields<br />
3.3 Comparison to <strong>Chernobyl</strong> Simulation by ARL<br />
3.4 In-Cloud Wet Scavenging Rate Sensitivity Runs<br />
3.5 Modeled Cloud Base Modification. . . . . . .<br />
3.5.1 Daily Phases of <strong>Chernobyl</strong> Emissions<br />
14<br />
14<br />
14<br />
16<br />
16<br />
19<br />
19<br />
IV
Page<br />
3.5.2 Modified-Cloud-Base Motivation and Procedure 22<br />
3.5.3 Modified-Cloud-Base Procedures, Simulation of<br />
Daily Deposition . . . . . . . . . . . . . . . . 25<br />
3.5.4 Modified-Cloud-Base Procedures, Simulation of<br />
April Deposition in Germany and Austria .. 28<br />
IV.<br />
Results<br />
37<br />
4.1 In-Cloud Scavenging Sensitivity Test Results.<br />
37<br />
4.1.1 ICS Sensitivity Over Germany, 86.04.26.06Z 37<br />
4.1.2 ICS Sensitivity Over Germany, 86.04.26.12Z 44<br />
4.2 Modified-Cloud-Base Height Simulation Results . . . 44<br />
4.2.1 Modified-Cloud-Base Performance Over Time 44<br />
4.2.2 Modified-Cloud-Base Performance in April Over<br />
Germany / Austria . . . . . . . . . . . . . . . . 62<br />
V.<br />
Conclusions . . . . . . . .<br />
5.1 Sensitivity Runs<br />
5.2 Cloud Base Modification Runs<br />
5.3 Future Research Opportunities<br />
69<br />
69<br />
69<br />
73<br />
Appendix A. Glossary of Acronyms 75<br />
Appendix B.<br />
Radioactivity Primer.<br />
B.1 Ionizing Radiation<br />
B.2 Cesium-137 ....<br />
77<br />
77<br />
78<br />
Appendix C.<br />
Reanalyzed Precipitation Fields from the ECMWF Model<br />
80<br />
Appendix D. HySPLIT Settings . . . . . . . .<br />
D.1 Release Height Sensitivity Runs.<br />
D.2 Comparison to ARL <strong>Chernobyl</strong> Simulation.<br />
87<br />
87<br />
91<br />
v
-----------------------------------------------------------------------------------<br />
Page<br />
D.3 In-Cloud Scavenging Sensitivity Control Run 91<br />
D.4 Daily Deposition Control Run. . . . . . . . . 91<br />
D.5 <strong>Chernobyl</strong> Control Run - Cumulative Deposition on Germany<br />
and Austria to OOZ, 1986MayOl ......... 92<br />
D.6 Greece Diagnostic Run HySPLIT Settings - Emission<br />
10m to 1750m ...................... 92<br />
Appendix E. Political Map of Europe 93<br />
Appendix F. HySPLIT Source Code Modification 94<br />
Appendix G.<br />
Investigation of Greece Exclusion from Modeled April<br />
Cs-137 Deposition ........... .<br />
96<br />
G.1 Evaluation of Source Term Height - Wet<br />
G.2 Evaluation of Source Term Height - Dry<br />
97<br />
97<br />
Bibliography<br />
105<br />
Vita .....<br />
108<br />
vi
Figure<br />
1.<br />
List of Figures<br />
Simplified 12Z Surface Weather Maps 25 - 28 Apr, 1986<br />
Page<br />
6<br />
2. Cumulative Cs-137 Deposition Measurements 27Apr1986 to lOMay1986,<br />
from ATMES Report .................. 10<br />
3.<br />
4.<br />
5.<br />
6.<br />
7.<br />
8.<br />
9.<br />
10.<br />
11.<br />
12.<br />
13.<br />
14.<br />
15.<br />
16.<br />
17.<br />
18.<br />
19.<br />
20.<br />
Illustration of HySPLIT Cloud Base Parameterization<br />
Domain of ECMWF Input Meteorological Data . . . .<br />
ARL's <strong>Chernobyl</strong> Cs-137 Deposition Using HySPLIT .<br />
Attempted Duplicate of ARL's <strong>Chernobyl</strong> Cs-137 Deposition<br />
Using HySPLIT ........................ .<br />
<strong>Chernobyl</strong> Cs-137 Emissions in Twelve Phases, Summary Bargraph<br />
.......................... .<br />
Emission Height Sensitivity Run, 1500-meter Release.<br />
Emission Height Sensitivity Run, 3000-meter Release.<br />
Emission Height Sensitivity Run, 4000-meter Release.<br />
Emission Height Sensitivity Run, 5000-meter Release.<br />
Daily Deposition Cities with Measurement Start Dates .<br />
Surface-Based Cesium-137 Measurement Sites in Europe.<br />
Modeled April <strong>Chernobyl</strong> Trajectories from 2100Z, 25Apr1986<br />
Modeled April <strong>Chernobyl</strong> Trajectories from 0400Z, 26Apr1986<br />
Modeled April <strong>Chernobyl</strong> Trajectories from OOOOZ, 27 Apr1986<br />
Modeled April <strong>Chernobyl</strong> Trajectories from OOOOZ, 28Apr1986<br />
Modeled April <strong>Chernobyl</strong> Trajectories from OOOOZ, 29Apr1986<br />
Modeled April <strong>Chernobyl</strong> Trajectories from OOOOZ, 30Apr1986<br />
Deposition of Phase I Emissions Over 04.25.21Z - 05.01.00Z .<br />
12<br />
15<br />
17<br />
18<br />
20<br />
23<br />
23<br />
24<br />
24<br />
27<br />
29<br />
31<br />
32<br />
33<br />
34<br />
35<br />
36<br />
36<br />
21. ECMWF 6-hr Model Precipitation During 86.04.26.06Z Sensitivity<br />
Run ............................ 38<br />
Vll
Figure<br />
Page<br />
22. In-Cloud Scavenging Sensitivity Test, Control Run Deposition,<br />
86.04.26.06Z . . . . . . . . . . . . . . . . . . . . . . . . . . . 39<br />
23. In-Cloud Scavenging Sensitivity, 86.04.24.06Z, Deposition Difference<br />
with 1% ICS Efficiency Boost ............. 40<br />
24. In-Cloud Scavenging Sensitivity, 86.04.24.06Z, Deposition Difference<br />
with 1% ICS Efficiency Reduction . . . . . . . . . . . 41<br />
25. In-Cloud Scavenging Sensitivity, 86.04.24.06Z, Deposition Difference<br />
with 5% ICS Efficiency Boost ............. 42<br />
26. In-Cloud Scavenging Sensitivity, 86.04.24.06Z, Deposition Difference<br />
with 5% ICS Efficiency Reduction . . . . . . . . . . . 42<br />
27. In-Cloud Scavenging Sensitivity, 86.04.24.06Z, Deposition Difference<br />
with 10% ICS Efficiency Boost 43<br />
28. In-Cloud Scavenging Sensitivity, 86.04.24.06Z, Deposition Difference<br />
with 10% ICS Efficiency Reduction .......... 43<br />
29. ECMWF 6-hr Model Precipitation During 86.04.26.12Z Sensitivity<br />
Run ............................ 45<br />
30. In-Cloud Scavenging Sensitivity Test, Control Run Deposition,<br />
86.04.26.12Z .......................... . 46<br />
31. In-Cloud Scavenging Sensitivity, 86.04.24.12Z, Deposition Difference<br />
with 1 % ICS Efficiency Boost ............. 46<br />
32. In-Cloud Scavenging Sensitivity, 86.04.24.12Z, Deposition Difference<br />
with 1% ICS Efficiency Reduction . . . . . . . . . . . 47<br />
33. In-Cloud Scavenging Sensitivity, 86.04.24.12Z, Deposition Difference<br />
with 5% ICS Efficiency Boost ............. 47<br />
34. In-Cloud Scavenging Sensitivity, 86.04.24.12Z, Deposition Difference<br />
with 5% ICS Efficiency Reduction . . . . . . . . . . . 48<br />
35. In-Cloud Scavenging Sensitivity, 86.04.24.12Z, Deposition Difference<br />
with 10% ICS Efficiency Boost 48<br />
36. In-Cloud Scavenging Sensitivity, 86.04.24.12Z, Deposition Difference<br />
with 10% ICS Efficiency Reduction .......... 49<br />
viii
Figure<br />
37.<br />
38.<br />
Page<br />
Summary of Helsinki Daily Cs-137 Deposition 1986Apr27 - 1986May14,<br />
Measured, Control Run, and Modified Run . . . . . . . . . . 51<br />
Summary of Bratislava Daily Cs-137 Deposition 1986Apr27 -<br />
1986May14, Measured, Control Run, and Modified Run . . . 51<br />
39. Summary of Moravsky Krumlov Daily Cs-137 Deposition 1986Apr27<br />
- 1986May14, Measured, Control Run, and Modified Run .. 52<br />
40. Summary of Hof Daily Cs-137 Deposition 1986Apr27 -1986May14,<br />
Measured, Control Run, and Modified Run . . . . . . . . . . 52<br />
41. Summary of Pass au Krumlov Daily Cs-137 Deposition 1986Apr27<br />
- 1986May14, Measured, Control Run, and Modified Run .. 53<br />
42. Summary of Schwandorf Krumlov Daily Cs-137 Deposition 1986Apr27<br />
- 1986May14, Measured, Control Run, and Modified Run .. 53<br />
43. Summary of Hradec Kralov Daily Cs-137 Deposition 1986Apr27<br />
- 1986May14, Measured, Control Run, and Modified Run .. 54<br />
44. Summary of Kosice Daily Cs-137 Deposition 1986Apr27 - 1986May14,<br />
Measured, Control Run, and Modified Run . . . . . . . . . . 54<br />
45. Summary of Budapest Daily Cs-137 Deposition 1986Apr27 -<br />
1986May14, Measured, Control Run, and Modified Run . . . 55<br />
46. Summary of Mol Daily Cs-137 Deposition 1986Apr27 -1986May14,<br />
Measured, Control Run, and Modified Run . . . . . . . . . . 55<br />
47. Summary of Harwell Daily Cs-137 Deposition 1986Apr27 - 1986May14,<br />
Measured, Control Run, and Modified Run . . . . . . . . . . 56<br />
48. Summary of Aachen Daily Cs-137 Deposition 1986Apr27 -1986May14,<br />
Measured, Control Run, and Modified Run . . . . . . . . . . 56<br />
49. Summary of Emden Daily Cs-137 Deposition 1986Apr27 - 1986May14,<br />
Measured, Control Run, and Modified Run . . . . . . . . . . 57<br />
50. Summary of Koblenz Daily Cs-137 Deposition 1986Apr27 -1986May14,<br />
Measured, Control Run, and Modified Run . . . . . . . . . . 57<br />
51. Summary of Schleswig Daily Cs-137 Deposition 1986Apr27 -<br />
1986May14, Measured, Control Run, and Modified Run . . . 58<br />
ix
Figure<br />
Page<br />
52. Summary of Bilthoven Daily Cs-137 Deposition 1986Apr27 -<br />
1986May14, Measured, Control Run, and Modified Run . . . 58<br />
53. Summary of Offenbach Daily Cs-137 Deposition 1986Apr27 -<br />
1986May14, Measured, Control Run, and Modified Run . . . 59<br />
54. Summary of Glasow Daily Cs-137 Deposition 1986Apr27 - 1986May14,<br />
Measured, Control Run, and Modified Run . . . . . . . . . . 59<br />
55. Summary of Berlin Daily Cs-137 Deposition 1986Apr27 - 1986May14,<br />
Measured, Control Run, and Modified Run . . . . . . . . . . 60<br />
56. Summary of Giessen Daily Cs-137 Deposition 1986Apr27 - 1986May14,<br />
Measured, Control Run, and Modified Run . . . . . . . . . . 60<br />
57. Summary of Muenchen Daily Cs-137 Deposition 1986Apr27 -<br />
1986May14, Measured, Control Run, and Modified Run . . . 61<br />
58. Summary of Berkeley Daily Cs-137 Deposition 1986Apr27 -<br />
1986May14, Measured, Control Run, and Modified Run . . . 61<br />
59. Summary of Risoe Daily Cs-137 Deposition 1986Apr27 -1986May14,<br />
Measured, Control Run, and Modified Run<br />
62<br />
60.<br />
Cloud Base Modification Test, Control Run<br />
63<br />
61. Cloud Base Modification Test, 75%-RH Continental Cloud Base<br />
Run. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64<br />
62. April <strong>Chernobyl</strong> Deposition Difference: (April Deposition Modified-<br />
Cloud-Base Run - April Deposition Control Run) . . . . . . . 65<br />
63. Normal Probability Plot of April Cumulative Cs-137 Deposition<br />
Measurements in Germany and Austria . . . . . . . . . . . . 67<br />
64.<br />
65.<br />
66.<br />
67.<br />
Set 1. Pearson Correlation Coefficients of Daily Deposition by<br />
City ............................... .<br />
Set 2. Pearson Correlation Coefficients of Daily Deposition by<br />
City ............................... .<br />
Typical Shielding Requirements for Different Ionizing Radiation<br />
Types from VIC, 00 . . . . . . .<br />
Current Political Map of Europe<br />
71<br />
71<br />
78<br />
93<br />
x
Figure<br />
68.<br />
69.<br />
70.<br />
Greece Diagnostic Run, Cumulative April Deposition [Bq/m2],<br />
Release Height Profile from 10m to 1750m . ......... .<br />
Greece Diagnostic Run, Cumulative April Deposition [Bq/m2] ,<br />
Release Height Profile from 10m to 2990m. . . . . . . . . . .<br />
Greece Diagnostic Run, Cumulative April Deposition [Bq/m2],<br />
Release Height Profile from 10m to 9000m . ......... .<br />
Page<br />
98<br />
99<br />
100<br />
71. Five-Day Deposition from Exaggerated Phase I Emissions, Precipitation<br />
Turned Off . . . . . . . . . . . . . . . . . . . . . . 102<br />
72. Five-Day 1000-m Average from Exaggerated Phase I Emissions,<br />
Precipitation Turned Off . . . . . . . . . . . . . . . . . . . . 102<br />
73. Five-Day 2000-m Average from Exaggerated Phase I Emissions,<br />
Precipitation 'furned Off .. . . . . . . . . . . . . . . . . . . 103<br />
74. Five-Day 4000-m Average from Exaggerated Phase I Emissions,<br />
Precipitation 'furned Off . . . . . . . . . . . . . . . . . . . . 103<br />
75. Five-Day 7000-m Average from Exaggerated Phase I Emissions,<br />
Precipitation 'furned Off .... . . . . . . . . . . . . . . . . 104<br />
76. Five-Day 10000-m Average from Exaggerated Phase I Emissions,<br />
Precipitation 'furned Off . . . . . . . . . . . . . . . . . 104<br />
xi
List of Tables<br />
Table<br />
Page<br />
l. Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr25. 81<br />
2. Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr26. 82<br />
3. Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr27. 83<br />
4. Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr28. 84<br />
5. Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr29. 85<br />
6. Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr30. 86<br />
Xll
AFIT /GM/ENP /OlM-5<br />
Abstract<br />
In response to the <strong>Chernobyl</strong> nuclear power plant accident of 1986, cesium-137<br />
deposition was measured in Europe at sites equipped to do so. The resulting deposition<br />
dataset is uniquely applicable to atmospheric transport model validation.<br />
Most of the airborne <strong>Chernobyl</strong> cesium was wet deposited, i.e., either via interception<br />
by falling raindrops (below-cloud scavenging) or via absorption into cloud<br />
droplets destined to become raindrops (in-cloud scavenging). The model used in<br />
this work is the Hybrid Single-Particle Lagrangian Integrated Transport (HySPLIT)<br />
model developed at Air Resources Laboratory. A cloud base modification is tested<br />
and appears to slightly improve the accuracy of one HySPLIT simulation of daily<br />
<strong>Chernobyl</strong> cesium-137 deposition over the course of the accident at isolated European<br />
sites, and degrades the accuracy of another HySPLIT simulation of deposition in Germany<br />
and Austria accumulated in the month of April, 1986. Large uncertainties in<br />
the emission specifications, model precipitation fields, and deposition measurements<br />
prevent designating the results as conclusive, but most evidence points to improved<br />
performance within 500km of the emission source. Trial and error lessons learned<br />
from hundreds of preliminary model runs are documented, and the exact HySPLIT<br />
settings of successful and meaningful simulations are appended.<br />
xiii
SIMULATING WET DEPOSITION OF RADIOCESIUM<br />
FROM THE CHERNOBYL ACCIDENT<br />
1. Introduction<br />
The United States Air Force Technical Applications Center (AFTAC) is charged<br />
with observing global environmental conditions to detect and identify activities peculiar<br />
to nuclear weapons testing. AFTAC's global array of seismic, atmospheric,<br />
and other environmental sensors makes up the U.S. Atomic Energy Detection System<br />
(USAEDS). By means of USAEDS and a full complement of world-class analytical<br />
laboratories, AFTAC monitors signatory nations' compliance with international<br />
nuclear test ban treaties (Hagans 00). In a role supporting this mission, AFTAC<br />
meteorologists generate routine and special atmospheric pollutant transport simulations.<br />
The simulations can, from a given source, gauge how much pollutant will<br />
arrive where and when. Long-range meteorological simulations require accounting<br />
for precipitation scavenging, both in-cloud and below-cloud. In-cloud scavenging<br />
(or rain-out), hereafter referred to as ICS, is the process of cloud droplets or ice<br />
crystals assimilating pollutant within clouds, aggregating, and falling to the ground.<br />
Below-cloud scavenging (or wash-out), hereafter referred to as BCS, is the process<br />
of pre-formed precipitation cleansing pollutant from the air below clouds on its way<br />
to the surface. The combined processes of ICS and BCS result in wet deposition<br />
at the earth's surface, and playa significant role in removing long-term pollutants<br />
from the atmosphere. So, improvements to wet deposition modeling are important<br />
to improving the accuracy of long-range transport simulations. Because wet deposition<br />
dominated the other long-range <strong>Chernobyl</strong> fallout deposition mechanisms,<br />
the <strong>Chernobyl</strong> case, though severely limited by uncertainty in initial conditions and<br />
deficiencies in measurement data, is uniquely applicable to wet deposition scheme<br />
1
validation. AFTAC meteorologists have proposed tests of wet deposition schemes<br />
in <strong>Chernobyl</strong> deposition simulations using the Hybrid Single-Particle Lagrangian Integrated<br />
Trajectory (HySPLIT) model (Draxler 98b). The rest of this document<br />
expands on the motivations, procedures, and results of requested HySPLIT simulations<br />
of the <strong>Chernobyl</strong> accident fallout deposition.<br />
1.1 Problem and Objective<br />
Reasonable cloud base parameterization is crucial to realistic model assignments<br />
of ICS and BCS. To that end, the sensitivity of wet deposition modeling to<br />
ICS is tested, then used to interpret results of a test of a HySPLIT modified-cloudbase<br />
scheme, namely, reducing cloud bases to 75% relative humidity over land masses<br />
from 80%, in pursuit of improved HySPLIT wet deposition parameterization.<br />
1.2 Thesis Organization<br />
The chapters of this thesis are structured so as to clarify specific challenges<br />
to modeling wet deposition of <strong>Chernobyl</strong> Cs-137. Chapter II provides background<br />
on aspects of the <strong>Chernobyl</strong> accident relevant to wet deposition modeling, including<br />
emission characteristics and prevalent weather patterns. Chapter II also puts<br />
major atmospheric transport studies in perspective with respect to wet deposition<br />
modeling. Chapter III describes the methods used to introduce weather variables,<br />
validates the basic simulation parameters by comparison to previous work, and details<br />
the methods used for further simulation runs of ICS sensitivity and cloud base<br />
modification. Chapter IV presents separately the results of said sensitivity runs<br />
and cloud base modification runs. Chapter V ties the other chapters together and<br />
gives the reader direction for further wet deposition investigation. The appendix is<br />
designed to aide the reader in reconstructing and customizing the simulations herein.<br />
The reader may find the Glossary of Acronyms in Appendix A frequently useful.<br />
2
---------------,<br />
II. Background<br />
2.1 Background Overview<br />
This chapter includes a review of the <strong>Chernobyl</strong> nuclear power plant accident of<br />
April 1986 with emphasis on the role of Cs-137 deposition, a brief description of prevailing<br />
weather conditions, a short discussion of major long-range transport studies,<br />
and an overview of HySPLIT, the transport model used for all <strong>Chernobyl</strong> simulations<br />
in this thesis. These topics provide a foundation both for an ICS sensitivity<br />
study described in Section 3.4, and for a cloud base modification study described in<br />
Section 3.5.<br />
2.2 The Chemobyl <strong>Accident</strong> as a Wet Deposition Case Study<br />
At 2123 UTC (0123L) on 25 April 1986, during a sequence of tests, reactor<br />
unit number four at the <strong>Chernobyl</strong> nuclear power plant experienced an unmanageable<br />
increase in power due to a number of "design deficiencies and operator errors"<br />
(DeCort 98: 11). Emergency actions by the attendant technicians were fruitless as<br />
increasing temperatures in the cooling system brought on two violent steam explosions,<br />
ejecting reactor components into the reactor room, blowing apart portions of<br />
the building including the roof, and setting dozens of fires at the site. Heat from<br />
the initial steam explosion and subsequent graphite fire lifted a cloud of radioactive<br />
particulates at least a kilometer up into the atmosphere. The emission rate gradually<br />
tapered off until 2 May when heroic efforts to contain the fire caused, instead,<br />
increasing emissions over the next four days. The ruptured unit was finally sealed<br />
in a concrete sarcophagus on 6 May (Klug 92:2). In total, the event released 6000-<br />
8000kg of radioactive material and probably much more inactive material into the<br />
atmosphere. About one-third of the particles were transported more than 20km<br />
from the power plant (Pollanen 97). Over the course of the 10-day emission an<br />
estimated 85 ± 26P Bq of Cs-137 was released (Metivier 95). Later, in Section 3.5.1,<br />
3
complete <strong>Chernobyl</strong> emission specifications for simulations used in this thesis are<br />
described.<br />
Wet deposition played a dominant role in long-range cesium deposition during<br />
the <strong>Chernobyl</strong> accident. Based on estimates of time-integrated concentrations<br />
and measured dry deposition velocities of cesium (or "caesium") after the accident,<br />
Lauritzen and Mikkelsen suggest "that only approx. 10% of the total deposition of<br />
caesium is due to dry deposition, while the remaining 90% stems from wet deposition"<br />
(Lauritzen 99). Wet deposition modeling, then, must be included in any<br />
realistic simulation of <strong>Chernobyl</strong> cesium deposition. Long-range (dry) transport is<br />
itself an inexact science. Superimposing another process as complicated as rain or<br />
snow modeling onto the dry transport process takes transport modeling to a new<br />
level of uncertainty. Lauritzen and Mikkelsen call the distribution of atmospheric<br />
transport deposition "multi-fractal" and believe that this randomness on all scales<br />
"implies that standard atmospheric dispersion models (i.e., deterministic models)<br />
cannot explain details of the deposition pattern, but only its gross, average structure"<br />
(Lauritzen 99:3271). Case studies of wet deposition are difficult and rare because of<br />
this compounded uncertainty. Severe patchiness in deposition measurements from<br />
the localizing effects of precipitation scavenging oblige transport experiment designers<br />
to carefully schedule experiments so as to avoid the complications of precipitation<br />
effects. Likewise, nuclear weapons tests are performed on clear days, avoiding dangerous<br />
radioactive hot spots associated with precipitation (Glasstone 77:418). The<br />
<strong>Chernobyl</strong> accident is a unique case study in that it involves a massive quantity of<br />
radioactive tracer wet-deposited and measured hundreds and even thousands of miles<br />
away. Because the <strong>Chernobyl</strong> case is a unique case of measured wet deposition, it<br />
presents a unique opportunity to validate wet deposition in transport and dispersion<br />
models.<br />
4
2.3 Weather Patterns During the <strong>Chernobyl</strong> <strong>Accident</strong><br />
Although <strong>Chernobyl</strong> pollutants were eventually detected throughout the northern<br />
hemisphere, much of <strong>Chernobyl</strong>'s emissions were deposited in Europe because<br />
of low level circulations and widespread precipitation typical for the season. The<br />
weather patterns during the accident provided a range of changing conditions throughout<br />
the continent. Figure 1 presents simplified surface weather charts for the first<br />
four days of the accident from Knap, 1988 (Knap 88:151). Synoptic weather analysis<br />
reveals a prevailing cold continental high pressure system to the northeast of <strong>Chernobyl</strong>.<br />
Meanwhile, a North Atlantic semi-permanent low pressure system off the<br />
west coast of Great Britain spawned a series of precipitating troughs across western<br />
and central Europe during all phases of the <strong>Chernobyl</strong> accident emissions. While<br />
the effects of large-scale features west and east of Central Europe on the <strong>Chernobyl</strong><br />
plume were apparent during the time of the accident, the plume was often directly<br />
steered by smaller, weaker weather features such as the shallow fronts associated<br />
with these short-wave troughs.<br />
2.4 Cesium-137 Transport from the <strong>Chernobyl</strong> <strong>Accident</strong><br />
For particles to travel far enough (100's to 1000's of km) to be considered<br />
long-range emissions, they must be aerodynamically small enough for turbulence to<br />
keep them suspended and carried on the wind for days. The <strong>Chernobyl</strong> fire generated<br />
massive quantities of sub-micron particles carrying Cs-137 (cesium-137). Small<br />
particles, 111m in aerodynamic diameter and smaller, have a fall speed of less than<br />
about l.Omml s. Accordingly, the dry processes that have the greatest influence on<br />
their transport and deposition are turbulent eddies and Brownian diffusion (small<br />
particles spread out by random collisions with air molecules) (P611anen 97). Particles<br />
that tiny remain suspended long enough for precipitation, when present, to<br />
play a major role in their deposition. Large particles, 20l1m in aerodynamic diameter<br />
and larger, have a fall speed of greater than about 1O.0mml s in the lower<br />
5
1030_-.::!..,-__<br />
~<br />
Figure 1 Simplified 12Z surface weather maps for 25 - 28 Apr, 1986 (Knap 88)<br />
6
atmosphere. As a result, the processes that have the greatest influence on their<br />
deposition are gravitational settling (large particles pulled earthward) and turbulent<br />
dispersion (particle-bearing air mass grows by entrainment). The relatively short<br />
transport life of large particles prevents rain from playing a significant role in their<br />
deposition. Particles with a size between small and large present a special challenge<br />
to the transport and deposition modeler. A single fall speed parameterization for<br />
medium-sized particles is elusive. Which dry transport mechanisms determine the<br />
effective fall speed of medium-size particles depends on whether the particles are<br />
large-medium or small-medium and on several specific weather conditions at each<br />
location (Seinfeld 86). Fortunately for <strong>Chernobyl</strong> plume modelers, particles from<br />
the <strong>Chernobyl</strong> accident containing Cs-137 are confined mainly to the small category<br />
because of how they were formed in the fire and smoke at <strong>Chernobyl</strong>.<br />
Cs-137 and Sr-90 (strontium-90) are signature long-range fallout isotopes both<br />
of nuclear power production and of nuclear weapons testing. The human body is<br />
less susceptible to harm from ingesting Cs-137 than from ingesting the same amount<br />
of Sr-90. The biological half-life of Cs-137 is 50 to 200days. Sr-90, on the other<br />
hand, is chemically similar to calcium, so the body tends to concentrate the isotope<br />
in bone tissue where it remains in the body much longer. So, even though Cs-137 is<br />
just as easy to measure, has a slightly longer radioactive half-life, and is slightly more<br />
abundant than Sr-90 in nuclear weapons fallout, nuclear scientists normally characterize<br />
long-range weapons fallout by patterns of Sr-90 deposition (Glasstone 77:604).<br />
The accident at <strong>Chernobyl</strong>, due to the nature of the explosion and fire, produced<br />
relatively little Sr-90 outside the 30-km evacuation zone (DeCort 98:13). Therefore,<br />
Cs-137 is the best species for characterizing the long-range radioactive deposition<br />
pattern from the <strong>Chernobyl</strong> accident as a whole (Klug 92).<br />
Because Cs-137 is radioactive, it is detectable in very small concentrations,<br />
an ideal property for a long-range plume tracer. The radionuclide itself has a'<br />
radioactive decay half-life of more than thirty years (Serway 92). Such a long half-<br />
7
life affords meaningful cumulative measurements over periods of months and even<br />
years. More details are available in Appendix B, a primer on radioactivity and<br />
Cs-137. An experiment releasing sub-micron particles bearing Cs-137 would, in<br />
theory, be ideal for validating and improving operational wet deposition modeling.<br />
Atmospheric nuclear weapon tests in the 1950's and 1960's injected radiocesium<br />
into the stratosphere which, to this day, continues to trickle radioactive particles<br />
back into the troposphere, especially at mid-latitudes near the jetstream, although<br />
one could argue that the amount is negligible (Glasstone 77:448). At any rate, a<br />
large and hazardous emission would be required to discern long-range experimental<br />
concentrations above measurement background noise, and the political repercussions<br />
of such an experiment would be prohibitive. An experimental case study using Cs-<br />
137 as a tracer is not feasible. So, the <strong>Chernobyl</strong> case is likely to remain a unique<br />
opportunity to model and compare Cs-137 deposition on a large scale.<br />
2.5 Other Long-Range Transport Modeling Exercises<br />
In November 1986, an international effort emerged to coordinate a transport<br />
modeling study within the context of the <strong>Chernobyl</strong> accident. In response to the accident<br />
and its environmental repercussions, the IAEA (International Atomic Energy<br />
Agency) collaborated with the WMO (World Meteorological Organization) to develop<br />
the Atmospheric Transport Model Evaluation Study (ATMES). The study was<br />
designed both to test the emergency response capability of current transport modeling<br />
agencies, and to "intercalibrate" their various models (Klug 92:v). Twenty-two<br />
agencies from fourteen countries volunteered their transport models for the study.<br />
Each model was developed independently, some for purposes other than nuclear accident<br />
response, so results varied widely. Some models were designed for meso-scale<br />
application and were specially adapted for the ATMES exercise. Some simulated<br />
puff emissions, others tracked individual particles. Some were based on integrations<br />
in an Eulerian frame, others were based on Lagrangian integrations. Each<br />
8
model's configuration and results were compiled in the ATMES Report along with<br />
comparisons to each other and to available measured deposition data. The writers<br />
of the ATMES Report constructed a contour plot of accumulated Cs-137 deposition<br />
(Figure 2) from available surface-based measurement data. Though based on all<br />
available deposition measurements at the time, the figure is not representative of<br />
the whole pattern of <strong>Chernobyl</strong> Cs-137 deposition. For instance, even though the<br />
highest concentrations of Cs-137 were found near <strong>Chernobyl</strong> at 51.38° latitude, 30.1°<br />
longitude, Figure 2 suggests a minimum there. So, due to large data sparse regions,<br />
it appears Figure 2's content may be less representative of area-averaged Cs-137 deposition<br />
values than of the geographical density of observation sites in the ATMES<br />
Cs-137 deposition dataset (map in Section 3.5.4).<br />
A key result of the ATMES project was the realization of the strong need for<br />
an experimental case, i.e. a transport experiment with known emission specifications<br />
and synchronized, homogeneous measurements, to confidently evaluate even<br />
the relative performance of long-range transport models (Klug 92). Controlled experiments<br />
have several advantages over accidental cases. To date, major long-range<br />
atmospheric transport modeling experiments have all released non-depositing tracers<br />
to maintain detectable pollutant concentrations over distance and remove uncertainties<br />
involved in deposition. Controlled experiment observations are planned<br />
at regular time and space intervals to generate output grids that are homogeneous<br />
(Rodriguez 95:800). Perhaps most importantly, the source rate and height are<br />
known precisely in a controlled experiment, in stark contrast to the typically vague<br />
specifications of accidental emissions.<br />
Following the guidance from ATMES conclusions, and the lessons learned from<br />
the Across North America Tracer EXperiment (ANATEX), the same agencies that<br />
organized the ATMES Report designed and executed the European Tracer Experiment,<br />
or ETEX (Rodriguez 95). This time, NOAA's Air Resources Laboratory<br />
(ARL) was a participant in the study, contributing deposition simulations created<br />
9
70.oo,..------------::4'i~-~"""'--_,<br />
61.25<br />
(J)<br />
~ 52.50<br />
~<br />
43.75<br />
I<br />
............-<br />
(,/<br />
I<br />
I<br />
I<br />
I<br />
I<br />
I<br />
//)<br />
, ,<br />
I<br />
I<br />
I<br />
I<br />
I<br />
I<br />
I<br />
I<br />
I<br />
I<br />
I<br />
I<br />
.... ~~ ...... '<br />
~.oo~~~~~~~~.-~~~~~~~~<br />
-10.0 2.5 15.0 27.5 40.0<br />
Longltlx:le<br />
MEAS ------. 2 ----. 5 --- 10 -- 50<br />
Figure 2 Geographic plot of contours of accumulated Cs-137 deposition [kBqjm 2 ]<br />
from 27 Apr1986 to 10May1986. ATMES Report Figure 10 (Klug 92).<br />
with HySPLIT Version 4.0 (ARL OOb).<br />
The organizers of ETEX, like the ANA<br />
TEX designers, chose a non-soluble perfiuorcarbon chemical as a tracer species.<br />
The chemical resists deposition by both wet and dry mechanisms, optimizing the<br />
homogeneity of the tracer's transport pattern, and so making for better simulation<br />
comparisons.<br />
Both ANATEX and ETEX were initiated in part to provide a<br />
dataset for future model evaluations. The careful completeness of each experiment's<br />
design makes them ideal for long-range (dry) transport model evaluations and improvements.<br />
However, since the tracer could not be rained out, their datasets are<br />
not suited for a wet deposition study (Graziani 97). Again, the <strong>Chernobyl</strong> accident<br />
10
stands alone as a case study for validating long-range wet deposition schemes against<br />
in situ measurements.<br />
2.6 HySPLIT Model Description<br />
HySPLIT, the Hybrid Single-Particle Lagrangian Integrated Transport model<br />
calculates either the trajectories of air parcels, or the transport, dispersion and deposition<br />
of pollutant particles or puffs. User-supplied inputs for HySPLIT calculations<br />
are pollutant species characteristics, emission parameters, gridded meteorological<br />
fields, and output deposition grid specifications. Input meteorological fields can be<br />
on Polar Stereographic, Lambert Conformal, or Mercator map projections. The<br />
horizontal deformation of the wind field, the wind shear, and the vertical diffusivity<br />
profile are used to compute dispersion rates. The model can be configured to treat<br />
the pollutant as particles, or as Gaussian puffs, or as top-hat puffs. The term hybrid<br />
refers to the additional capability of HySPLIT to treat the pollutant as a Gaussian<br />
or top-hat puff in the horizontal, while treating the pollutant as a particle for the<br />
purposes of calculating vertical dispersion. An advantage to the hybrid approach<br />
is that the higher dispersion accuracy of the vertical particle treatment is combined<br />
with the spatial resolution benefits of horizontal puff-splitting. All model runs for<br />
this work were made in the default hybrid particle/top-hat mode.<br />
HySPLIT calculates wet deposition by scavenging pollutant from portions of<br />
the plume in (ICS) and below (BCS) precipitating model clouds. All of the scavenged<br />
pollutant is assumed to deposit on the ground directly below the clouds. To identify<br />
precipitating model clouds, HySPLIT's wet deposition algorithm checks the input<br />
meteorological data at each surface gridpoint for precipitation. Where precipitation<br />
is non-zero, it searches upward from the top of the surface layer (i.e., no fog modeling)<br />
for the lowest model level with an RH (relative humidity) greater than or equal to<br />
80%. This 80% threshold establishes the modeled cloud base. A 75% threshold is<br />
tested later in Chapter III. The cloud top is determined by the lowest level above<br />
11
Figure 3<br />
* *<br />
Illustration of cloud representation in HYSPLIT. Stars represent pollutant<br />
plume. 80% RH is the default cloud base, 75% RH is the tested<br />
modification.<br />
the base where the RH is below 60%. Above the first (lowest) cloud layer, HySPLIT<br />
diagnoses no more clouds. Figure 3 illustrates a precipitating cloud, both modeled<br />
cloud bases, and a pollutant plume.<br />
Below the bases of precipitating clouds HySPLIT scavenges the pollutant<br />
plume, reducing its concentration by an amount equal to the product of the pollutant<br />
concentration, the user-specified BGS rate [8- 1 ] (below-cloud scavenging rate),<br />
and the time increment [8]. The amount of below-cloud pollutant reduction (concentration<br />
reduction times plume volume) is then added to the surface deposition<br />
output grid. Deposition from IGS (in-cloud scavenging) is calculated (and in-cloud<br />
pollutant concentration is reduced accordingly) using a user-specified IGS efficiency<br />
[L/ L] defined as the ratio of pollutant concentration in air (grams of plume pollutant<br />
per liter of air) to pollutant concentration in rain (grams of deposited pollutant per<br />
liter of precipitated water). The amount of deposition from IGS is found by multiplying<br />
the in-plume pollutant concentration [L- 1 ] by the rain accumulation [mm]<br />
and dividing by the IGS efficiency as derived in the equations below (unit conversion:<br />
1m 2 x 1mm = 1Liter = 0.001m 3).<br />
This approach to IGS requires rain rate [mm]<br />
12
from the input meteorological grid. HySPLIT is not able to calculate wet deposition<br />
without input precipitation fields (Draxler 98a:16).<br />
In-Cloud Plume Concentration ICS Effi .<br />
. . = C18ncy<br />
Ram ConcentratIOn<br />
R · P 11 [ IL] _ Air Pollutant[gl L]<br />
am 0 utant 9 - ICS Effi<br />
C18ncy<br />
.<br />
R · P 11 [I 2] I R . [ ] _ 0.001 x Air Pollutant[glm3 ]<br />
am 0 utant 9 m am mm - ICS Effi .<br />
clency<br />
[I 2] _ O.OOl(Air Pollutant[glm3 ])(Rain[mm])<br />
R am · P 11<br />
0 utant 9 m - ICS Effi<br />
clency<br />
.<br />
13
III. Methodology<br />
3.1 Methodology Chapter Overview<br />
HySPLIT has seen frequent algorithm updates to assimilate current findings in<br />
the field of atmospheric transport modeling, and HySPLIT's user interface has been<br />
continually enhanced to improve its usability as an operational tool. This chapter<br />
supplies the details on exactly how to use HySPLIT to perform selected atmospheric<br />
transport simulations. Section 3.2 describes how meteorological fields including<br />
wind, temperature, pressure, humidity, and precipitation data were incorporated in<br />
simulations for this thesis. To ensure that correct meteorological fields and other<br />
inputs are being used, a duplicate Cs-137 deposition simulation is attempted and<br />
compared to results produced by ARL. Section 3.3 explains how the attempted<br />
duplicate simulation is performed. Section 3.4 describes the method used to evaluate<br />
the sensitivity of a <strong>Chernobyl</strong> simulation to various scavenging rates. Finally, Section<br />
3.5 describes the procedures used to validate a proposed cloud base modification in<br />
the model against <strong>Chernobyl</strong> deposition measurements. Results of sensitivity runs<br />
and of cloud base modification runs are presented later, in Chapter IV.<br />
3.2 Incorporation of Meteorological Input Fields<br />
The meteorological input fields for all simulations are reanalyzed ECMWF<br />
data from NCAR. Using HySPLIT for <strong>Chernobyl</strong> plume transport and deposition<br />
calculations requires conversion of ECMWF GRIB format meteorological fields to<br />
ARL packed format (Draxler 99). The conversion utility program provided with<br />
HySPLIT requires platform-dependent GRIB decoder libraries typically available<br />
from the source of raw GRIB data, in this instance NCAR. As it converts a file to<br />
ARL-packed format, the utility interpolates the data linearly to a polar stereographic<br />
lat/lon grid in the horizontal, and to internal terrain-following sigma levels in the<br />
vertical. HySPLIT uses the smallest domain of input meteorological fields as the<br />
14
Figure 4<br />
Model domain and resolution of ECMWF meteorological input data grid,<br />
also used for concentration calculations in HySPLIT. Crosses at every<br />
fourth gridpoint for clarity. .<br />
total domain for a given simulation. The domain and resolution of the ECMWF<br />
input files for all simulations in this work is depicted in Figure 4. Upper air data<br />
fields for all simulations include temperature in [0C], u and v wind components in<br />
[m/s], w wind component in [hPa/hr], and specific humidity in [g/kg]. The surface<br />
data fields provided are 2-m temperature in [0C], 10-m u and v wind components in<br />
[m/ s], surface pressure in [hPa], and 6-hr prior accumulated precipitation in [mm].<br />
For comparison to simulations and for informal diagnosis of wet deposition effects, a<br />
full set of six-hourly ECMWF re-analyzed precipitation fields over the model domain<br />
for the first five days of the accident are provided as shaded plots in Appendix C.<br />
15
3.3 Comparison to <strong>Chernobyl</strong> Simulation by ARL<br />
Before addressing the methods for sensitivity runs and cloud base modification<br />
runs (Sections 3.4 and 3.5), evidence is presented here to show that the<br />
working installation of HySPLIT is performing as designed, and that appropriate<br />
meteorological fields and user-specified parameterization settings are being utilized<br />
properly. The evidence takes the form of results from an attempted duplicate<br />
of web-published <strong>Chernobyl</strong> deposition contours generated at ARL (Air Resources<br />
Laboratory) (ARL OOa). The attempted duplicate simulation uses an abbreviated<br />
<strong>Chernobyl</strong> source term, releasing pollutant at a constant rate for 24hrs only. Gross<br />
features of the 84-hr deposition patterns from the ARL simulation (Figure 5) and<br />
the attempted duplicate (Figure 6) are in agreement, suggesting that HySPLIT is<br />
functioning properly, the proper time period of meteorological data has been applied,<br />
wet scavenging is actually being modeled, etc. Differences (e.g., deposition southeast<br />
of <strong>Chernobyl</strong>) between the ARL simulation and the attempted duplicate are<br />
attributable to ARL's undocumented inclusion of some emissions beyond the first 24<br />
hours (Draxler OOa). A copy of the control file settings used to create the HySPLIT<br />
duplicate simulation is furnished in Section D.2 of Appendix D. The setup for this<br />
simulation serves as a baseline for simulation setups for the remainder of this work.<br />
3.4 In-Cloud Wet Scavenging Rate Sensitivity Runs<br />
To gauge the relative importance of ICS in wet deposition modeling, a simplified<br />
scenario is required mainly because the actual emissions from <strong>Chernobyl</strong> were<br />
continuous for days, making it difficult to attribute given deposition to a particular<br />
release time. So, an abbreviated emission is used in the sensitivity run, and the<br />
country of Germany is chosen as the deposition domain because the weather conditions<br />
modeled there also apply to cloud base modification studies in Section 3.5.<br />
The sensitivity run emission rate, 6.65 x 10 14 Bq / hr, and the emission's uniform vertical<br />
profile from 1250m to 1750m, mirror the first phase of the <strong>Chernobyl</strong> emission<br />
16
DEPOSITION FROM 002 27 APR TO 12Z 30 APR (UTe)<br />
12:2 215 APR CHNB FOREeAS!' IN1TIALIZATION<br />
GROUND-LEVEL DEPOSlTlON t ;1112)<br />
·1.0E+04 • LOE+02 1.0E+OO >0.0<br />
7.0E+D4 llAXIMUM A.T SQUARE<br />
Figure 5<br />
<strong>Chernobyl</strong> Cs-137 deposition as modeled by ARL. Deposition velocity<br />
set at 0.lcm/8. In-cloud scavenging ratio set at 3.2 x 10 5 1/1. Belowcloud<br />
scavenging rate set at 5.0 x 10- 5 8- 1 . Deposition contoured on a<br />
logarithmic scale<br />
simply because it is useful to verify the model's ability to accommodate values with<br />
these magnitudes. Apr 26 is the chosen time period because precipitation is present<br />
that day. Six-hourly accumulations of pollutant deposition are recorded to coincide<br />
with the time resolution of the precipitation fields. The coordinate 48.0 0 latitude,<br />
11.0 0 longitude is the chosen release location because the spot is immediately upstream<br />
from Germany during the chosen period. If the modeled release were chosen<br />
at <strong>Chernobyl</strong>, it would not be possible to observe the immediate influence of ICS.<br />
ARL suggests a value of 3.2 x 10 5 for the user-specified ICS efficiency after<br />
Hicks (Hicks 86) and an empirical mean value of 5.0 x 10- 5 8- 1 for the BCS rate.<br />
To evaluate the sensitivity of HySPLIT's wet deposition scheme to ICS parameters,<br />
17
Deposition from OOz 27 Apr to 12z 30 Apr (UTe)<br />
12Z 25 Apr 86 ECMF FORECAST INITIALIZATION<br />
w<br />
o<br />
.....<br />
a<br />
CO?<br />
Z<br />
.. I".~j""." __ '/_" __<br />
o<br />
......<br />
1;;<br />
* c<br />
o<br />
.~<br />
u<br />
.3<br />
Q)<br />
~<br />
::J<br />
o<br />
if)<br />
.1.0E+04<br />
3.2E+Q4 MAXIMUM AT SQUARE<br />
1.0E-02<br />
Figure 6 Attempted duplicate of ARL's <strong>Chernobyl</strong> Cs-137 deposition. Source<br />
modeled as uniform vertical line source from 750m to 1500m at a rate of<br />
10 15 Bq/hr for 24 hours. Deposition velocity set at 0.1cm/8. In-cloud<br />
scavenging ratio set at 3.2 x 10 5 [/1. Below-cloud scavenging rate set at<br />
5.0 x 10- 5 8- 1 . Deposition contoured on a logarithmic scale.<br />
these ARL-recommended values are employed as the baseline values for the sensitivity<br />
control run. In addition to the sensitivity control run, a simulation is performed<br />
for each of these ICS efficiencies: 3.232 x 10 5 , 3.36 X 10 5 , 3.52 X 10 5 , 3.168 X 10 5 ,<br />
3.04 X 10 5 , 2.88 X 10 5 . These values reflect boosts and reductions of the ICS efficiency<br />
by 1%, 5%, and 10%. HySPLIT control file settings for the sensitivity control run<br />
are recorded in Section D.3 of Appendix D. HySPLIT output concentration grid<br />
files are converted using HySPLIT utility program, 'con2bin.exe' to GRADS format<br />
for field differencing. Results of the sensitivity simulations accompanied by plots of<br />
simultaneous accumulated model precipitation are presented in Section 4.1.<br />
18
3.5 Modeled Cloud Base Modification<br />
In this Section, a method is presented for assessing performance of a modifiedcloud-base<br />
height parameterization in simulations of <strong>Chernobyl</strong> Cs-137 deposition.<br />
Subsection 3.5.1 develops a reasonable source term (emission specification). Subsection<br />
3.5.2 gives the motivation and procedure for the specific cloud base modification<br />
tested later. The available dataset lends itself best to two main model comparisons:<br />
a comparison to simulations of daily deposition 1986Apr28 - 1986May15 at 21 measurement<br />
sites spread across Europe described in Subsection 3.5.3, and a separate<br />
comparison to simulations of April-cumulative deposition from the onset of <strong>Chernobyl</strong><br />
emissions at 2123Z, 1986Apr25, up to OOOOZ, 1986May01 at a cluster of 395<br />
measurement sites in Germany and Austria, described in Subsection 3.5.4.<br />
3.5.1 Daily Phases of <strong>Chernobyl</strong> Emissions. The best-guess <strong>Chernobyl</strong><br />
source term (i.e., the emission specifications) is segmented into daily phases, except<br />
that the plume is treated separately during the first seven hours because it is believed<br />
to have risen significantly higher than subsequent emissions. The twelve phases, I<br />
through XII, used in this research are adapted from Table 1 of the ATMES Report<br />
(Klug 92:358). The ATMES Report provided a daily Cs-137 emission rate, specified<br />
a center of mass height for each phase of emission, and required that a step-function<br />
be used for the project's simulations. However, the ATMES organizers "still gave<br />
a certain degree of freedom to the participants, e.g. on the mass distribution with<br />
height" (Klug 92:2). According to the ATMES report, revised Russian release height<br />
estimates presented to the ATMES Steering Committee in January, 1989 are the only<br />
authoritative estimates (Klug 92:1,2). Today, though some documented evidence<br />
supports a change to the official release height, the source term estimate has not yet<br />
been updated by consensus (Graziani 00). Generally, it is accepted that the initial<br />
plume escaped the boundary layer, and that, after the first two days, the initial<br />
plume did not exceed an altitude of 400m (Persson 87). For this research, Phases I<br />
and II were recalculated (based on equal total emission amounts to OOOOZ, 27 Apr)<br />
19
<strong>Chernobyl</strong> Emission Rate in Twelve Phases<br />
+/- 50% Error Bars<br />
IOCS-137 [TBq/hr] I<br />
1400<br />
1200 +-----~~~------------------------------------------<br />
1000 ~~--------------------------------------------<br />
800<br />
600<br />
400<br />
200<br />
o ~~,-LJ-,L-~~~~~,L-L~~~LJ"L-~~L,~~,-~<br />
Figure 7<br />
Complete <strong>Chernobyl</strong> Cs-137 modeled source term in twelve phases.<br />
Plume modeled as uniform vertical line source at an hourly rate in becquerels.<br />
Phase I initial plume 1250-1750m, Phases II and III initial<br />
plumes 350-850m, Phases IV through XII initial plumes 200-400m.<br />
as a compromise between ATMES' 26.0000Z 6-hr initial plume and the known time<br />
of <strong>Chernobyl</strong>'s initial explosion, 25.2123Z. Each phase was modeled in HySPLIT<br />
as a uniform vertical line source. Figure 7 displays the resulting twelve <strong>Chernobyl</strong><br />
emission phases.<br />
The first few hours of <strong>Chernobyl</strong> emissions are the release time period with the<br />
greatest vertical location uncertainty. It is agreed that the initial steam explosion<br />
and ensuing fire at <strong>Chernobyl</strong>launched radioactive particles well above the accidentaveraged<br />
boundary layer top at roughly 500m. Pollanen et al. present evidence for<br />
a higher release height based on large particle trajectory calculations.<br />
"In northeastern Poland, 500-700 km from <strong>Chernobyl</strong>, particles up to<br />
rv 60 microns in aerodynamic diameter were found. Their sedimentation<br />
velocity is so large (up to rv O.lms- I ) that turbulent dispersion, rapid<br />
transport in a prefrontal low-level jet, warm frontal conveyor belt ... or<br />
20
even effective release height of 3000 m cannot explain these findings"<br />
(Pollanen 97:3581).<br />
A series of four preliminary 84-hr <strong>Chernobyl</strong> simulations is accomplished here<br />
to demonstrate how much the pattern of modeled pollutant deposition in Europe<br />
changes for a range of distinct, reasonable point source heights, namely, 1500m,<br />
3000m, 4000m, and 5000m. These preliminary simulations are not referenced outside<br />
of this section of the thesis.<br />
Each simulation's emission mirrors <strong>Chernobyl</strong>'s<br />
initial emission rate of 6.65 x 10 14 Bq/hr (see Subsection 3.5.1 for a complete time<br />
profile of best-guess <strong>Chernobyl</strong> emissions).<br />
Other particle parameterizations are<br />
empirical estimates of those properties typical of Cs-137-bearing particles from a<br />
nuclear reaction. The four preliminary simulations are identical except for release<br />
height. Exact HySPLIT settings used for the 1500-m run are presented in Section<br />
D.1 of Appendix D along with detailed descriptions of each setting. These descriptions<br />
serve to familiarize the reader with HySPLIT concentration model setup.<br />
Should the reader consider downloading and using HySPLIT, further instructions<br />
are available on the internet from ARL (and more details are in Appendix D).<br />
surface deposition concentration grid is computed in each simulation over the lat/lon<br />
grid centered at 48°/13° and spanning 26° of latitude, 36° of longitude. The main<br />
difference between a 1500-m release (Figure 8) and a 3000-m release (Figure 9) is<br />
an overall decrease in deposition, presumably because the particles take longer to<br />
settle from a higher release point, and because greater wind speeds at 3000m carry<br />
more particles beyond deposition grid boundaries.<br />
A<br />
Changing the release height to<br />
4000m (Figure 10) produces a distinct southward shift in the 84-hr deposition pattern.<br />
Less pollutant deposits in the Nordic countries (e.g., none in Finland), while<br />
higher and more widespread pollutant concentrations are modeled from Ukraine and<br />
Romania to Italy, France and even Algeria. A release height of 5000m (Figure 11)<br />
produces a further southward shift in the deposition pattern: less deposition from<br />
Lithuania and Belarus to Sweden, more deposition from Ukraine and Romania to<br />
21
Italy and Algeria. For the reader's reference, Appendix E offers a map of Europe<br />
with current political boundaries.<br />
The pollutant is transported completely across the Mediterranean Sea into<br />
Algeria when the initial plume exceeds 3000m. The concentration of modeled<br />
deposition is on the order of only IBq/m2, well below background levels of 2000<br />
- 3000Bq/m2. However, modeled deposition underestimates hot spots due to precipitation<br />
field smoothing. So, if the model is broadly accurate in the region, detection<br />
would not be impossible, especially if daily measurements are available within<br />
deposition hot spots. Though beyond the scope of this work, this clue would be<br />
of special interest to those interested in refining the <strong>Chernobyl</strong> source term. There<br />
is a distinct shift in the deposition pattern to include deposition south and east of<br />
<strong>Chernobyl</strong> for release heights above 3500m. This change in general plume direction<br />
supports the view that release heights above 3000m would seriously alter the<br />
deposition pattern of <strong>Chernobyl</strong>'s day one emissions.<br />
3.5.2 Modified-Cloud-Base Motivation and Procedure. Since empirical values<br />
for Cs-137 ICS efficiency and BCS rate have been documented, a logical place<br />
to look for wet deposition improvement is the cloud base parameterization since it<br />
directly determines vertically where modeled BCS stops and modeled ICS begins.<br />
The complexities of cloud formation are immense and many. Sophisticated prognostic<br />
cloud models are available, but are computationally expensive and gain little<br />
accuracy over diagnostic parameterizations since large uncertainties remain in the<br />
accounting of "advective transports of cloud variables, sub-grid scale processes, cloud<br />
microphysics, and cloud optical properties" (Tiedke 93:3040). Most current transport<br />
models and even some global meteorological models still use simple diagnostic<br />
schemes to model clouds. Future generations of transport models may just accommodate<br />
liquid and ice cloud fields from the input meteorological model rather than<br />
calculating their own cloud limits. For now, a parameterization is still needed and<br />
HySPLIT's simple scheme of cloud diagnosis from relative humidity is considered<br />
22
• 1.0E+04 • 1.0E+02 " 1.0E+OO 1.0E-02<br />
9,4E104 MAXIMUM AT SQUARE<br />
Figure 8<br />
<strong>Chernobyl</strong> deposition [Bq/m2] from OOZ 1986Apr26 to OOZ 1986MayOl.<br />
Emission from 2123Z Apr25 for 24hrs at <strong>Chernobyl</strong> (star in the graphic)<br />
at 6.65 x 10 14 Bq / hr from 1500-m point source height.<br />
w<br />
~ ',,1,;<br />
c:i<br />
C'l<br />
1ll<br />
* c:<br />
o<br />
]<br />
(I)<br />
[J<br />
~<br />
"40<br />
/----<br />
C"><br />
III<br />
:0<br />
(tI<br />
ifi<br />
III<br />
III<br />
(tI<br />
(J)<br />
w<br />
~<br />
(tI<br />
0..<br />
~<br />
~<br />
N<br />
~<br />
»<br />
"Q<br />
~<br />
g<br />
·1.0E+04<br />
4,8E; 04 MAXIMUM AT SQUARE<br />
1.0E-02<br />
Figure 9<br />
<strong>Chernobyl</strong> deposition [Bq/m2] from OOZ 1986Apr26 to OOZ 1986MayOl.<br />
Emission from 2123Z Apr25 for 24hrs at <strong>Chernobyl</strong> (star in the graphic)<br />
at 6.65 x 10 14 Bq/hr from 3000-m point source height.<br />
23
LlJ<br />
0<br />
r-<br />
c:i<br />
C'l<br />
Z<br />
co<br />
iD<br />
·f5<br />
III<br />
to'><br />
~<br />
~<br />
Iii ;:+<br />
iii<br />
*<br />
e: ~<br />
~<br />
] }:-<br />
Q) "Q<br />
u<br />
0<br />
'.p<br />
N<br />
:s<br />
~<br />
(")<br />
{IJ<br />
:0<br />
(tI<br />
(ll<br />
ID<br />
0-<br />
W<br />
M-<br />
C<br />
~<br />
9<br />
1 ,2E t 04 MAXIMUM AT SQUARE<br />
1.0E-02<br />
Figure 10<br />
<strong>Chernobyl</strong> deposition [Bq/m2] from OOZ 1986Apr26 to OOZ 1986MayOl.<br />
Emission from 2123Z Apr25 for 24hrs at <strong>Chernobyl</strong> (star in the graphic)<br />
at 6.65 x 10 14 Bq/hr from 4000-m point source height,<br />
UJ<br />
0<br />
..-<br />
c:i<br />
:0<br />
(tI<br />
C'l "<br />
iD<br />
4:5 III<br />
Z<br />
{IJ<br />
41<br />
co<br />
rJ)<br />
~<br />
Iii<br />
iii<br />
*<br />
e: ~<br />
0<br />
'.p<br />
N<br />
~<br />
...J }:-<br />
al "Q<br />
~<br />
:J<br />
~ it<br />
(")<br />
{/)<br />
iif<br />
g<br />
0-<br />
~<br />
~<br />
~<br />
;;,OEto.1 MAX.lMUM AT SQUARE<br />
1.0E-02<br />
Figure 11<br />
<strong>Chernobyl</strong> deposition [Bq/m2] from OOZ 1986Apr26 to OOZ 1986MayOl.<br />
Emission from 2123Z Apr25 for 24hrs at <strong>Chernobyl</strong> (star in the graphic)<br />
at 6.65 x 10 14 Bq/hr from 5000-m point source height.<br />
24
an over-generalization (Draxler OOb). To attempt an improvement to HySPLIT's<br />
cloud base parameterization simple enough to test in a short time, and as a starting<br />
point for exploring cloud base parameterization, a simple HySPLIT modification<br />
is proposed following work at NCEP (National Center for Environmental Prediction).<br />
As part of a comprehensive cloud model algorithm in NOAA's Meso ETA<br />
model, meteorologists at NCEP split the cloud base scheme into a marine part and<br />
a terrestrial part. It is believed that an 80% RH cloud base over continents represents<br />
too little cloud condensation. Cloud bases over land are modeled at 75%<br />
RH while cloud bases over water are modeled at 80% RH (Staudenmaier 96). HyS<br />
PLIT's hard-wired 80% RH cloud base is not unreasonable. If limited to a single<br />
value, long-range transport models should weight a unified scheme in favor of the<br />
marine environment since the earth's surface is mostly water. So, if NCEP's scheme<br />
approximates reality, 80% RH-modeled global cloud bases should outperform 75%<br />
RH-modeled cloud bases on a global scale. However, since HySPLIT has the ability<br />
to distinguish land use types, there is no need to compromise. The requisite cloud<br />
base modification in HySPLIT requires a change to the model source code as given in<br />
Appendix F. HySPLIT source code was provided for this thesis courtesy of Roland<br />
Draxler at ARL. Once the source code is edited and recompiled, the comparison<br />
runs can be accomplished identically to the April deposition control runs. Test results<br />
of April deposition control runs and April deposition modification runs against<br />
the April deposition data are presented in Section 4.2.<br />
3.5.3 Modified-Cloud-Base Procedures, Simulation of Daily Deposition.<br />
The available dataset used for comparing surface-based Cs-137 measurements to simulation<br />
data is from the REM (Radioactivity Environmental Monitoring) <strong>Chernobyl</strong><br />
archives at the JRC /Ispra (Joint Research Centre - Ispra, Italy) of the EC (European<br />
Commission) (DeCort 90). Deposition from above-ground nuclear weapon tests in<br />
the mid 1950's and early 1960's has blanketed the surface of the entire globe with a<br />
thin layer of Cs-137. Just before the <strong>Chernobyl</strong> accident, typical Cs-137 concentra-<br />
25
tions on the ground in Europe were between 2000 and 3000Bq/m 2 (DeCort 98:15).<br />
Post-accident deposition measurements on the ground near that range of values cannot<br />
be attributed to <strong>Chernobyl</strong> with confidence unless measurements were recorded<br />
at the same site before the accident. Since measurement records were mainly near<br />
large cities and near other nuclear power plants, pre-accident measurement data is<br />
sparse and geographically irregular. These limitations prevent construction of a<br />
more complete surface-based measurement dataset. Aerial gamma spectrometry<br />
measurements were taken over Eastern Europe and Western Russia weeks after the<br />
accident and deposition maps of these measurements exist (DeCort 98). However,<br />
these data were not available for this research, so this work is based entirely on<br />
the available surface-based Cs-137 measurements (DeCort 90). The surface-based<br />
deposition data for the <strong>Chernobyl</strong> case, available from the JRC, are segregated into<br />
two distinct datasets: a daily deposition dataset described in this section, and a<br />
cumulative deposition dataset described in Section 3.5.4. Figure 12 identifies, for<br />
the period during and just after <strong>Chernobyl</strong>'s emission, the 23 sites where daily deposition<br />
readings are available and the first date at each site that daily measurements<br />
were recorded. Section 4.2.1 presents a bar graph for each city with daily deposition<br />
measurements.<br />
Daily deposition data in the REM dataset is recorded at most sites in [Bq/m2].<br />
Passau, Koblenz, Glasgow, and Berkeley reported deposition in [Bq/ LJ, i.e., becquerels<br />
per liter of rain. For calculation of daily deposition totals at these sites, the<br />
REM dataset supplies daily precipitation amounts in [mm] for some sites. For those<br />
sites with measurements in [Bq/ L] and no precipitation, precipitation at the nearest<br />
weather station is used to convert to [Bq/m2]. No weather station precipitation<br />
was recorded for Koblenz, so model precipitation amounts are used to convert measurements<br />
from [Bq/ L] to [Bq/m2] for the Koblenz data only. European weather<br />
station precipitation records are courtesy of the Air Force Combat Climatology Center<br />
(AFCCC). Precipitation is missing from the Koblenz record in the REM dataset<br />
26
;<br />
'"<br />
><br />
,',-'<br />
..... '-~- _." -: ....... -"<br />
Figure 12<br />
European cities where daily Cs-137 deposition measurements were taken<br />
with measurement start dates. Dates 29-30 are April 1986, 01-08 are<br />
May 1986<br />
and from precipitation records supplied by AFCCC. Absent human-recorded precipitation,<br />
ECMWF model precipitation is substituted for the conversion of Koblenz<br />
deposition from [Bqj L] to [Bqjm 2 ]. HySPLIT limits pollutant release specifications<br />
to a constant emission rate, so twelve separate runs are required to model emissions<br />
from the entire accident, one for each phase of emission. Time series of surface<br />
deposition at measurement sites are extracted from the daily deposition control run<br />
output grids of each run and summed. To produce corresponding deposition time<br />
series modeled with the proposed cloud base modification, the process is repeated<br />
exactly as the daily deposition control run, but executed with a recompiled HyS<br />
PLIT model. The resulting daily deposition control and modification time series<br />
are presented and examined in Section 4.2.1. To allow the reader to reproduce the<br />
27
simulations, HySPLIT settings used to produce Phase I model output from the daily<br />
deposition control run are given in Section D.4 of Appendix D.<br />
3.5.4 Modified-Cloud-Base Procedures, Simulation of April Deposition in Germany<br />
and Austria. Measurement sites from the REM cumulative deposition<br />
dataset are depicted geographically in Figure 13, from ATMES Report Figure 8. It<br />
is evident from Figure 13 that the dataset's data points in Germany, Austria, and<br />
Greece are uniquely dense and homogeneous. Further examination reveals that the<br />
German, Austrian, and Grecian data in the cumulative dataset is also simultaneous.<br />
Because isolated measurement data points are not generally representative of a region<br />
due to unpredictable hot spots and holes in the long-range deposition pattern, the<br />
portion of the REM cumulative Cs-137 deposition dataset in these three countries is<br />
extracted for cloud modification run comparisons in this thesis. Unfortunately, the<br />
HySPLIT April deposition control run with best-guess <strong>Chernobyl</strong> source term and<br />
reanalyzed meteorological input data (Section 4.2.2) does not yield any deposition in<br />
Greece up to 86.05.01.00Z, while several separate measurements taken on that day<br />
in Greece indicate cumulative Cs-137 concentrations abqve 10 5 Bq/m 2 • Appendix G<br />
addresses possible reasons for the exclusion of Greece from modeled April deposition<br />
patterns in this thesis. May 1 German and Austrian deposition data remains the<br />
most homogeneous cumulative <strong>Chernobyl</strong> deposition data and is used exclusively for<br />
the April-cumulative model comparison (results in Section 4.2.2).<br />
Because HySPLIT only accommodates a constant emission rate, a separate<br />
model run must be accomplished for each phase of <strong>Chernobyl</strong> emissions to account<br />
for the total Cs-137 emission, then the deposition from each phase can be added<br />
together. To minimize the number of deposition simulations required to model Aprilcumulative<br />
<strong>Chernobyl</strong> deposition on Germany and Austria, air parcel trajectories<br />
are calculated for each of the first six phases of <strong>Chernobyl</strong> emissions. Figures 14<br />
through 19 are the modeled atmospheric trajectories of air parcels from <strong>Chernobyl</strong><br />
during April (86.04.25.21Z - 86.04.30.24Z). The six figures correspond to the first six<br />
28
·
15 trajectories from 350m and 500m and Figure 16 trajectory from 850m imply that<br />
the only modeled <strong>Chernobyl</strong> accident air parcel trajectories that cross Germany or<br />
Austria in April originate from Phase II emissions (1986.04.26.04Z - 27.00Z) and<br />
Phase III emissions (27.00Z - 28.00Z). These three trajectories trace the plume's<br />
path northwestward toward the Baltic Sea under the influence of northeast high pressure<br />
before getting wrapped southwestward through Germany and Austria around<br />
the back side of a shallow, transient trough. The remainder of modeled April <strong>Chernobyl</strong><br />
trajectories once again exhibit anticyclonic curvature, characteristic of a high<br />
pressure system, and curve away from Germany and Austria (Figures 17, 18, and<br />
19). April 28 and 29, the northeast high retreats eastward leaving a loose arrangement<br />
of very weak frontal boundaries. Without a strong pressure gradient to boost<br />
winds, the plume trajectories slow and meander more. By April 30, high pressure<br />
to the west begins to build and move eastward.<br />
Since only Phases II and III produced plumes over the area of interest (Germany<br />
and Austria), only two runs are required for each April cumulative deposition<br />
simulation. April deposition control run cumulative output concentration grids from<br />
HySPLIT are converted to GRADS format and summed in the GRADS program,<br />
available from the Institute of Global Environment and Society (IGES 01). The<br />
process is repeated using the recompiled HySPLIT model, and results for April deposition<br />
control run and April deposition modification runs are presented in Section<br />
4.2.2. Exact HySPLIT settings for April cumulative Phase II model output from<br />
the April deposition control run over Germany and Austria are given in Section D.5<br />
of Appendix D. To confirm that Phase I emissions did not diffuse from their mean<br />
path (i.e., the trajectories in Figure 14) all the way to Germany, the results of a<br />
full deposition simulation for Phase I appears in Figure 20. The results show that<br />
modeled Phase I emissions do not contribute in April to the initial 5-day deposition<br />
on Germany and Austria, but deposit instead largely in Belarus and Lithuania.<br />
30
w<br />
0<br />
ci<br />
C')<br />
z<br />
~<br />
10<br />
-ru<br />
• t:<br />
0<br />
~<br />
0<br />
0<br />
-I<br />
Ql<br />
!:?<br />
::I<br />
0<br />
C/)<br />
.. , .:<br />
.0<br />
,<br />
o<br />
s:<br />
m<br />
G)<br />
»<br />
! L ii . d • ...--;: ..<br />
Figure 14<br />
Modeled trajectories of <strong>Chernobyl</strong> air parcels in 6-hr increments. Originating<br />
at 2100Z on 25Apr19S6 from 500m (triangles to Sweden), 1000m<br />
(circles), 1500m (squares), and 2000m(triangles to Russia)<br />
31
--------------------------------,<br />
w<br />
0<br />
{)<br />
.... <<br />
ci<br />
l'TI<br />
t")<br />
(:> i:!;<br />
Z ~<br />
co s:<br />
~ 0<br />
,... c!'.<br />
Lf) 0<br />
:l<br />
iii s:<br />
it ~<br />
0<br />
r:: a.<br />
0 I<br />
~<br />
0<br />
...J<br />
m<br />
G)<br />
Q)<br />
~ :t><br />
:::I<br />
0<br />
en<br />
-....<br />
& 1 100<br />
.r:: 900<br />
a<br />
?:<br />
Figure 15<br />
Modeled trajectories of <strong>Chernobyl</strong> air parcels in 6-hr increments. Originating<br />
at 0400Z on 26Apr1986 from 350m (triangles), 500m (squares),<br />
and 850m (circles)<br />
32
LU<br />
o<br />
ci<br />
C')<br />
c<br />
o<br />
.,0:;<br />
ttl<br />
o<br />
....J<br />
QJ<br />
1\<br />
,<br />
o<br />
s::<br />
m<br />
G)<br />
»<br />
0.. ttl ~700 800<br />
..c: 900<br />
000 * ,. .Ic"<br />
Figure 16<br />
Modeled trajectories of <strong>Chernobyl</strong> air parcels in 6-hr increments. Originating<br />
at OOOOZ on 27 Apr 1986 from 350m (triangles), 500m (squares),<br />
and 850m (circles)<br />
33
w<br />
- -56<br />
0<br />
<<br />
ci (1)<br />
en ?a<br />
40 42 ::l-<br />
34 36 ;38 n'<br />
z !!.<br />
C(I<br />
s;:<br />
~<br />
!l<br />
..- O·<br />
It)<br />
:::I<br />
1U<br />
ir :T<br />
0<br />
c<br />
c..<br />
0<br />
".j:;<br />
t'II<br />
0<br />
0 s::<br />
0<br />
.J<br />
m<br />
(;)<br />
GI<br />
~<br />
»<br />
:::I<br />
0<br />
(/J<br />
.. ~'.<br />
s;:<br />
(1)<br />
. ~ ~900 #I a I • • • • • I '<br />
L: 000<br />
L a •<br />
•<br />
Figure 17<br />
Modeled trajectories of <strong>Chernobyl</strong> air parcels in 6-hr increments. Originating<br />
at OOOOZ on 28Apr1986 from 200m (triangles), 300m (squares),<br />
and 400m (circles).<br />
34
w<br />
0<br />
ci<br />
(')<br />
Z<br />
~<br />
10<br />
26<br />
'--<br />
36 40<br />
"Iii<br />
ir<br />
t:<br />
0<br />
~<br />
0<br />
0<br />
--l<br />
CII<br />
~<br />
::J<br />
0<br />
f/)<br />
o<br />
So:<br />
m<br />
G)<br />
»<br />
..c ~~~ 000<br />
.. J • £! 3<br />
Figure 18<br />
Modeled trajectories of <strong>Chernobyl</strong> air parcels in 6-hr increments. Originating<br />
at OOOOZ on 29Apr1986 from 200m (triangles), 300m (squares),<br />
and 400m (circles)<br />
35
ill<br />
o<br />
d<br />
M<br />
31 32<br />
33<br />
-+c<br />
c<br />
o<br />
'';:::;<br />
t1I<br />
o<br />
...J<br />
~<br />
:;:)<br />
o<br />
C/J<br />
.................... ··········50··<br />
o<br />
s::<br />
m<br />
G)<br />
»<br />
c I.<br />
•<br />
IS<br />
••<br />
Figure 19<br />
Modeled trajectories of <strong>Chernobyl</strong> air parcels in 6-hr increments. Originating<br />
at OOOOZ on 30Apr1986 from 200m (triangles), 300m (squares),<br />
and 400m (circles).<br />
UJ 0<br />
w<br />
0<br />
0<br />
;u<br />
(')<br />
!l!.<br />
if)<br />
::z 1}:<br />
co<br />
III<br />
I"l [g<br />
lti<br />
~<br />
'iii a.<br />
* <br />
~<br />
~<br />
13 -f<br />
(f)<br />
.D<br />
~<br />
N<br />
C<br />
1.SE-(I5 \1A)(n~tJ\1 ATf.QI)ARF<br />
Figure 20<br />
Modeled Deposition of Phase I <strong>Chernobyl</strong> Emissions Accumulated Over<br />
04.25.21Z - 05.01.00Z<br />
36
IV. Results<br />
4.1 In-Cloud Scavenging Sensitivity Test Results<br />
The results of seven diagnostic HySPLIT sensitivity test runs are explored in<br />
this section. Design of the runs follows the method for diagnosing ICS sensitivity<br />
presented in Section 3.4. The seven runs include the sensitivity control run,<br />
three runs with ICS efficiencies boosted by different amounts, and three runs with<br />
ICS efficiencies reduced by different amounts. The first time period, 86.04.26.00Z -<br />
86.04.26.06Z, is examined in Subsection 4.1.1, then the second time period, 26.06Z-<br />
26.12Z is examined in Subsection 4.1.2. In the sensitivity test scenario, a plume of<br />
particles like those carrying Cs-137 from a nuclear event is initiated in southernmost<br />
Germany at 48.0 o N, 1l.ooE and travels north-northeast as evidenced in deposition<br />
plots to follow. In HySPLIT, the units of emission per hour translate to the units of<br />
surface deposition per square meter. So, emissions in [Bq/hr] translate to deposition<br />
in [Bq/m2].<br />
4.1.1 ICS Sensitivity Over Germany, 86. 04.26. 06Z. Figure 21 provides<br />
6-hr-accumulated precipitation from the model to aide interpretation of deposition<br />
plots in Figures 22, 23, 24, 25, 26, 27, and 28. For example, Figure 21 identifies a<br />
dry area (no rain) in southeast Germany that corresponds to a deposition minimum<br />
between two maximums in the sensitivity control run deposition plot in Figure 22.<br />
The effects of boosting and reducing default ICS efficiency by various percentages<br />
in deposition model runs can be seen clearly by subtracting sensitivity control run<br />
deposition from each test run's results. The amount of deposition difference from<br />
the sensitivity control run for each boosted-ICS or reduced-ICS test run appears in<br />
Figures 23 through 28.<br />
In Figure 23 subtracting the sensitivity control run deposition from the (1%<br />
ICS efficiency boost) test run deposition yields a change in deposition on the order<br />
37
TPP6 ( mm) AT HEIGHT: 1.000<br />
• 1.00E+OO • 2.00E+OO • 3.00E+OO<br />
II 4.00E+OO 5.00E+OO r 6.00E+OO<br />
Figure 21<br />
Reanalyzed ECMWF model precipitation in [mm] accumulated from<br />
86.04.26.00Z to 86.04.26.06Z for comparison to in-cloud scavenging sensitivity<br />
run. Contours in 1-mm increments.<br />
of -O.5Bq/m2 near the source at 48.0DN, n.ODE. Logic dictates that deposition<br />
should instead be initially heavier when ICS is boosted, and initially lighter when<br />
ICS is reduced. The deposition change in Figure 23 implies that the immediate result<br />
of boosted ICS is decreased deposition. About lOOkm further north (downstream)<br />
increased deposition is observed. Two questions arise from these observations.<br />
The first question is, "How could deposition change amounts be opposite in sign<br />
if scavenging efficiency is increased only?" The second question is, "Why does it<br />
appear that increased scavenging immediately causes a decrease in deposition?"<br />
38
6-hr In-Cloud Default Accum. [Bq/m"2] 04.26.06z<br />
56N.-------~------~--~~~--~----.<br />
55N<br />
54N<br />
53N<br />
52N<br />
51N<br />
50N<br />
49N<br />
48N<br />
~ ...<br />
JQQ<br />
47N<br />
46N<br />
45NSE<br />
6E 7E 8E 9E 10E11E12E13E14E1SE16E17E18E19E<br />
Figure 22<br />
Six-hour-accumulated Cs-137 deposition [Bq/m2] from in-cloud scavenging<br />
sensitivity control run, 19S6.04.26.06Z<br />
In answer to the first question, the deposition differences from altered model<br />
scavenging can be opposite in sign upstream versus downstream from a point that<br />
could be termed the " scavenging error crossover point," or "SECP." At this unique<br />
point along the plume path, if total scavenging errors remain somewhat constant,<br />
the impact of the scavenging errors on deposition reverses. For example, if one<br />
assumes that modeled net scavenging is always and everywhere over-estimated, there<br />
must be a point in time and space where excess scavenging upstream has depleted<br />
the model plume so much that over-estimated scavenging downstream cannot make<br />
up for the concentration deficit in the plume. The resulting pattern of modeled<br />
deposition concentration would be too heavy upstream from the SECP and too light<br />
downstream from the SECP. Conversely, if net scavenging is always and everywhere<br />
under-estimated, the resulting deposition pattern would be too light near the source,<br />
and too heavy far from the source. The SECP principle applies as well to ICS<br />
errors alone if BCS and dry deposition are held constant as in these sensitivity tests.<br />
39
----------------------------------------------------,<br />
6- hr (In - Cloud + 1 %) - Default 04.26.06z<br />
56N.-----~~--------~--------------_.<br />
55N "- ----<br />
54N<br />
53N<br />
52N<br />
,<br />
51N<br />
50~~<br />
~<br />
49N<br />
4BN<br />
«.~<br />
-0.5<br />
47N<br />
46N<br />
45N+-~~--r-,-_._,--~~~~r_~_r_._1<br />
5E 6E 7E 8E 9E 10E 11 E 12E 13E 14E 15E 16E l7E 18E 19E<br />
Figure 23<br />
Difference in 6-hr-accumulated Cs-137 deposition [Bqjm 2 ]; Deposition<br />
from a test run with a 1% boost of in-cloud scavenging efficiency less<br />
deposition from control run, 86.04.26.06Z<br />
However, in the tests, ICS efficiency changes do not necessarily apply everywhere<br />
and always (e.g., a 1% boost of the default constant ICS efficiency may be an overestimate<br />
in some places and an under-estimate in others), so the SEep principle<br />
cannot be applied to model results blindly. In fact, the SEep principle cannot<br />
explain the answer to the second question. The SEep principle is noted, later, in<br />
Section 4.2.2 in an interpretation of deposition pattern changes from modifying the<br />
modeled cloud base.<br />
In answer to the second question, increased model scavenging appears to immediately<br />
cause decreased model deposition due to grid resolution and interpolation<br />
issues within the model. The feature of interest at 48.5°N, 11.0oE in Figures 23<br />
- 28, just north of the emission source, is not a physical phenomenon, but a computational<br />
one. HySPLIT only uses IeS efficiency (to calculate wet deposition)<br />
at gridpoints where precipitation is present. Since no precipitation is modeled at<br />
40
6-hr (In-Cloud - 1%) - Default 04.26.06z<br />
56N.-----~~--------~----------------_,<br />
55N<br />
54N<br />
53N<br />
52N<br />
51N<br />
50N<br />
49N<br />
48N<br />
....::=,<br />
( "':',>'<br />
\f,'-~?)<br />
('i/<br />
0.5<br />
47N<br />
46N<br />
45NSE<br />
6E 7E BE 9E 10E l1E 12E 13E l4E l5E 16E 17E lEE 19E<br />
Figure 24<br />
Difference in 6-hr-accumulated Cs-137 deposition [Bq/m 2 ]; Deposition<br />
from a test run with a 1% reduction of in-cloud scavenging efficiency<br />
less deposition from control run, 86.04.26.06Z<br />
48.5°N, 1l.ooE, and the only difference between the sensitivity control run and each<br />
test run is the rcs efficiency parameter, there must be a reason for the feature that<br />
is unrelated to rcs efficiency. The scale of the feature, the symmetry of the feature<br />
with the feature just north of it, as well as the persistence of both features in the<br />
nearly identical 12Z runs, leads one to believe that the cause of the phenomenon is<br />
initial deposition grid interpolations within HySPLrT. The impact of the feature on<br />
cloud base modification test results in Section 4.2 is negligible because of its relatively<br />
small magnitude, and is irrelevant because no deposition observations in the<br />
dataset selected for this work are available near <strong>Chernobyl</strong> to diagnose the cause of<br />
the feature. Further investigation of the phenomenon is beyond the scope of this<br />
work. Finally, to answer the second question explicitly, the immediate decrease in<br />
deposition is not caused by increased scavenging, but instead is likely an artifact of<br />
HySPLIT's interpolation of continuous variables using a discrete 60-km grid.<br />
41
56N<br />
55N<br />
54N<br />
53N<br />
6-hr (In-Cloud + 5%) -<br />
_.'<br />
Default 04.26,06z<br />
52N<br />
If)<br />
51N 0<br />
50N<br />
49N<br />
48N -2<br />
'.<br />
47N<br />
46N<br />
45N 5E 6E 7E 8E 9E 10E l1E 12E 13E 14E l5E 16E 17E l8E 19E<br />
Figure 25<br />
Six-hour-accumulated Cs-137 deposition [Bq/m2] difference from control<br />
run with a 5% boost of in-cloud scavenging efficiency, 86.04.26.06Z<br />
6-hr (In-Cloud - 5%) - Default 04.26.06z<br />
56N.-----~~----c---~--------------_,<br />
55N<br />
54N<br />
53N<br />
52N<br />
51N<br />
J<br />
50N<br />
49N<br />
48N<br />
47N<br />
46N<br />
45N+-~-.--.-~~~--.-~~~--~~~~<br />
5E 6E 7E BE 9E 10El1E12E13E14E15E16E17E18E19E<br />
Figure 26<br />
Six-hour-accumulated Cs-137 deposition [Bq/m2] difference from control<br />
run with a 5% reduction of in-cloud scavenging efficiency,<br />
86.04.26.06Z<br />
42
6-hr (I n-Cloud + 10%) -<br />
Defa ult 04.26.06z<br />
56N,-------~--~----~---------------~<br />
55N<br />
54N<br />
53N<br />
52N<br />
51N<br />
SON<br />
49N<br />
48N<br />
47N<br />
46N<br />
q;.<br />
-5<br />
45N+-~~--r_,-_r_,--~~~~r_~_r~~<br />
SE 6E 7E BE 9E IOE11E12E13E14E1SEl6E17E18E19E<br />
Figure 27<br />
Six-hour-accumulated Cs-137 deposition [Bq/m2] difference from control<br />
run with a 10% boost of in-cloud scavenging efficiency, 86.04.26.06Z<br />
6-hr (In-Cloud - 10%) - Default 04.26.06z<br />
56N.---~--~--~----~---------------~<br />
55N<br />
54N<br />
53N<br />
52N<br />
51N<br />
SON<br />
49N<br />
48N<br />
47N<br />
46N<br />
- ,<br />
,<br />
( \.<br />
~,'0<br />
45NSE 6E 7E BE 9E IOE11E12E13E14E15E16E17E18E19E<br />
Figure 28<br />
Six-hour-accumulated Cs-137 deposition [Bq/m2] difference from control<br />
run with a 10% reduction of in-cloud scavenging efficiency,<br />
86.04.26.06Z<br />
43
4.1.2 JCS Sensitivity Over Germany, 86. 04.26. 12Z. Because of the initial<br />
interpolation errors described in Subsection 4.1.1, a sensitivity test beyond the first<br />
six hours is required to estimate the impact of leS on total deposition. A 12Z sensitivity<br />
test, covering the period 86.04.26.06Z - 86.04.26.12Z, is presented here. Figure<br />
29 provides 6-hr-accumulated precipitation from the model to aide interpretation of<br />
the 12Z sensitivity test deposition plots in Figures 30 through 36. As in Subsection<br />
4.1.1, sensitivity control run deposition is presented first, in Figure 30, then six figures<br />
displaying difference plots where sensitivity control run deposition is subtracted<br />
from the deposition from each test run. The 12Z sensitivity control run deposition is<br />
an order of magnitude greater than the 06Z sensitivity control run deposition. Each<br />
difference plot in Figures 31, 32, 33, 34, 35, and 36 exhibits the initial interpolation<br />
error like those in Subsection 4.1.1. However, after the plume travels about 100km,<br />
the deposition pattern is straightforward. As expected, a boost of leS efficiency<br />
results in increased deposition as generally indicated in Figures 31, 33, and 35. A<br />
reduction in leS efficiency results in decreased deposition as generally indicated in<br />
Figures 32, 34, and 36. The response in the deposition pattern is approximately<br />
proportional to changes in the leS efficiency, e.g., the change in deposition caused<br />
by a 5% boost in leS efficiency (Figure 33) is approximately 5 times as much as<br />
the change in deposition from a 1% boost in leS efficiency (Figure 31). It is also<br />
noted that, assuming the sensitivity control run is truth, a plume with a 10% error<br />
in leS efficiency, traveling 600km in precipitation falling at 1mm/hr, does not reach<br />
its SEep. Otherwise, there would be a change in sign of downstream portions of<br />
the deposition difference patterns, at least in Figures 35 and 36.<br />
4.2 Modified-Cloud-Base Height Simulation Results<br />
4.2.1 Modified-Cloud-Base Performance Over Time. Daily deposition output<br />
from HySPLIT runs, produced as specified in Subsection 3.5.3, is presented in<br />
Figures 37 - 59 in order by earliest deposition measurement at each city.<br />
Raw<br />
44
Valid Time (UTe): 86/04/26/12<br />
~---n-...,<br />
TPP6 ( mm) AT HEIGHT: 1.000<br />
II 1.00E+00 • 2.00E+OO • 3.00E+OO<br />
• 4.00E+OO 5.00E+OO; 6.00E+00<br />
Figure 29<br />
Reanalyzed ECMWF model precipitation in [mm] accumulated from<br />
86.04.26.06Z to 86.04.26.12Z for comparison to in-cloud scavenging sensitivity<br />
run. Contours in I-mm increments.<br />
45
6-hr In-Cloud Default Accum. [8q/m"2] 04.26.12z<br />
56N.-------~------~--~~~--~----,<br />
55N<br />
.--:-<br />
54N<br />
53N<br />
52N<br />
51N<br />
50N<br />
49N<br />
48N<br />
47N<br />
46N<br />
45N5E 6E 7E BE 9E 10El1E12E13E14E15E16E17E1BE19E<br />
Figure 30<br />
Six-hour-accumulated Cs-137 deposition [Bq/m2] from in-cloud scavenging<br />
sensitivity control run, 1986.04.26.12Z<br />
56N<br />
6-hr (In-Cloud + 1%) -<br />
Default 04.26.12z<br />
55N -<br />
54N<br />
53N<br />
52N<br />
51N<br />
50N<br />
49N<br />
48N<br />
47N<br />
46N<br />
45N<br />
5E<br />
6E 7E BE 9E 10El1E12E13E14E15E16E17E18E19E<br />
Figure 31<br />
Six-hour-accumulated Cs-137 deposition [Bq/m2] difference from control<br />
run with a 1% boost of in-cloud scavenging efficiency, 86.04.26.12Z<br />
46
----------------------------------<br />
6-hr (In-Cloud - 1 %) - Default 04.26.12z<br />
56N~---~---~---------,<br />
55N<br />
54N<br />
53N<br />
52N<br />
51N<br />
50N<br />
49N<br />
48N<br />
47N<br />
46N<br />
45N5E<br />
6E 7E 8E 9E 10El1E12E13E14E15E16E17E1BE19E<br />
Figure 32<br />
Six-hour-accumulated Cs-137 deposition [Bqjm 2 ] difference from control<br />
run with a 1 % reduction of in-cloud scavenging efficiency,<br />
86.04.26.12Z<br />
6-hr (In-Cloud + 5%) -<br />
Default 04.26.12z<br />
56N.--~-~---~---------.<br />
55N .
------------------------------------------<br />
6-hr (In-Cloud - 5%) - Default 04.26.12z<br />
56N.--~-~-----~-------_.<br />
55N<br />
54N<br />
53N<br />
52N<br />
51N<br />
50N<br />
49N<br />
48N<br />
47N<br />
46N<br />
45NSE<br />
6E 7E BE 9E 10E l1E 12E 13E 14E lSE 16E 17E l8E 19E<br />
Figure 34<br />
Six-hour-accumulated Cs-137 deposition [Bq/m2] difference from control<br />
run with a 5% reduction of in-cloud scavenging efficiency,<br />
86.04.26.12Z<br />
6-hr (In-Cloud + 10%) -<br />
56N ,<br />
Default 04.26.12z<br />
55N<br />
54N<br />
53N<br />
52N<br />
51N<br />
50N<br />
49N<br />
48N<br />
47N<br />
46N<br />
45N SE<br />
6E 7E BE 9E 10El1E12E13E14E15E16E17E18E19E<br />
Figure 35<br />
Six-hour-accumulated Cs-137 deposition [Bq/m2] difference from control<br />
run with a 10% boost of in-cloud scavenging efficiency, 86.04.26.12Z<br />
48
-----------------------------------------------<br />
6-hr (In-Cloud - 10%) - Default 04.26.12z<br />
56N '.<br />
55N<br />
54N<br />
53N<br />
52N<br />
J<br />
51N<br />
50N<br />
49N<br />
48N<br />
47N<br />
46N<br />
45NSE<br />
6E 7E 8E 9E 10El1E12E13E14E1SE16E17E18E19E<br />
Figure 36<br />
Six-hour-accumulated Cs-137 deposition [Bq/m2] difference from control<br />
run with a 10% reduction of in-cloud scavenging efficiency,<br />
86.04.26.12Z<br />
measurement data (bars with shade gradient) for all 23 figures are from the REM<br />
databank at JRC - Ispra, Italy, and are based on total daily Cs-137 deposition<br />
measurements (DeCort 90). At many cities, gross features of modeled deposition<br />
distribution are in good agreement with those of measured deposition distribution.<br />
For instance, bimodal distributions are often indicated in both the measurement data<br />
and the modeled data with peaks synchronized to within about one day. The Pearson<br />
correlation coefficient between the entire daily deposition control run dataset and<br />
the daily deposition measurements is 0.5037. The Pearson correlation coefficient between<br />
the daily deposition modified-cloud-base run dataset and the daily deposition<br />
measurements is 0.5050. The formula for correlation follows, where n is the number<br />
of datapoints, and the variables X and Yare either daily measurement data and<br />
daily control run data, or daily measurement data and daily modified-cloud-base run<br />
49
data.<br />
Often, the daily deposition control run and daily deposition modified-cloudbase<br />
runs produce identical deposition, even days after emissions begin. Three<br />
different scenarios could produce identical daily deposition control and modifiedcloud-base<br />
amounts of deposition for a given day and site. If deposition for the<br />
whole day took place without the benefit of precipitation, no cloud base is calculated<br />
in HySPLIT, eliminating any cloud base modification effect. The second scenario<br />
finds precipitation on site, but no plume present between 75%-RH and 80%-RH levels<br />
above the deposition site, i.e., control run and modified-cloud-base run both find the<br />
plume either entirely beneath the cloud base, or entirely above the cloud base. The<br />
third scenario is a vertical resolution issue arising when the 75%-RH and 80%-RH<br />
are at effectively the same level. Such a discontinuity is possible in the model since<br />
the wet deposition algorithm uses discrete layers of humidity data. Even in some<br />
meteorological situations, such as a warm, moist air mass over-running a cold, dry<br />
air mass, a discontinuous relative humidity vertical profile is not an unreasonable<br />
approximation.<br />
50
-----------------~-<br />
Helsinki (Konala) (60.13,25.0) Daily Cs-137 Deposition [Bq/m A2]<br />
1 986 A P r2 8 - 1 986 May 1 4<br />
10000<br />
1000<br />
100<br />
10<br />
1<br />
91<br />
1901<br />
8<br />
f---<br />
30 32<br />
1 19 20 19 18 20<br />
14<br />
-<br />
e-- - - I- e-- 1- - -<br />
1<br />
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16<br />
Dmeasured 18 911 1901 87 30 32 19 20 19 18 14 55 301 20 111 27 8.4<br />
Oconlrol<br />
_modified<br />
38.61762325 292 1.24 1.01 1 .01 1 .01 1 .01 1 .01 1.13 4 .6553.1 15.1 1.2 1 1<br />
651 1145377 207 1.22 1.01 1.01 1 .01 1 .01 1.02 1 .1 3 4 .7253.1 15.1 1.2 1 1<br />
301<br />
111<br />
27<br />
8.4<br />
~<br />
17<br />
Figure 37<br />
Summary of Helsinki (Lat,Lon in decimal degrees) daily total Cs-137 deposition<br />
1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
B ra tis la v a (48.17,17.17) Daily C s -1 37 De po s itio n [B q 1m A 2] 1986 A p r2 8<br />
-1986May14<br />
10000<br />
1000<br />
100<br />
10<br />
1<br />
om e a su re d<br />
Dcontrol<br />
.m odified<br />
34<br />
15<br />
11<br />
9<br />
I-- -<br />
2<br />
117<br />
79<br />
~5<br />
41<br />
3<br />
f-- - .<br />
1 1 1 1 1 1<br />
rL<br />
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17<br />
1 117 349 157 26 91 65 32 79 46 117 5 1 1 1 1 1<br />
1 1371 1012 693 133 213 27_9 65.2 30.2 1067 149 1.21 1.01 1 1 1 1<br />
1 1371 1022 687 135 208 29.4 64 29.8 1067 149 11.21 1.01 1 1 1 1<br />
----------------------<br />
Figure 38<br />
Summary of Bratislava (Lat,Lon in decimal degrees) daily total Cs-<br />
137 deposition 1986Apr27 - 1986May14; measured, control run, and<br />
modified cloud base run.<br />
51
Mora v 5 k Y K ru m 10 V (49.08,16.33) Daily C 5·137 De p 0 5 ilio n [8 q 1m A 2)<br />
1 986 A P r2 8 • 1 986 May 1 4<br />
10000<br />
1000<br />
100<br />
80<br />
17<br />
76<br />
~ .~ J<br />
2 24<br />
391<br />
10<br />
1<br />
om easured<br />
Dcontrol<br />
.m odlfled<br />
- -<br />
1 1 1 1 1 1<br />
1 2 3 4 5 6 7 8 9 10 11 12<br />
-<br />
13 14 15 16 17<br />
1 24 801 171 53 76 59 53 24 52 391 58.8 1 1 1 1 1<br />
1 50122231 2601 119 53.1 46.6 41.3 21 673 29.9 1.2 1 1 1 1 1<br />
1 50122268 2578 107 52.1 47 40.3 20.8 673 29.9 1.2 1 1 1 1 1<br />
Figure 39<br />
Summary of Moravsky Krumlov (Lat,Lon in decimal degrees) daily total<br />
Cs-137 deposition 19S6Apr27 - 19S6May14; measured, control run, and<br />
modified cloud base run.<br />
H 0 f (50.31,11.93) Daily C 5·137 De p 0 5 itio n [8 q 1m A 2)<br />
1986 A P r2 8 • 1986 May 14<br />
10000<br />
1000<br />
100<br />
751<br />
14<br />
40<<br />
t------ I-<br />
9<br />
-<br />
3<br />
238<br />
10<br />
r--- - j--- -- r-<br />
1<br />
om e a su re d<br />
Dcontrol<br />
.m 0 d ifie d<br />
1 1 1 1 1111 1 1<br />
1 2 3 4 5 6 7 8 9 10 11 12 13<br />
1 751 146 30 92 462 238 1 1 1 1 1 1<br />
1 504 6634 389 156 50.5 218 29 18.7 25.5 2.35 1 1<br />
1 504 6634 406 145 59.4 205 28.4 18.9 25.1 2.34 1 1<br />
1 1 1 1<br />
14 15 16 17<br />
1 1 1 1<br />
1 1 1 1<br />
1 1 1 1<br />
~----------------------------------------------------------------------~<br />
Figure 40<br />
Summary of Hof (Lat,Lon in decimal degrees) daily total Cs-137 deposition<br />
19S6Apr27 - 19S6May14; measured, control run, and modified<br />
cloud base run.<br />
52
~<br />
r---<br />
~<br />
r---<br />
Passau (48.58,13.47) Daily Cs·137 Deposition [Bq/mA2]<br />
1 986 A P r2 8 • 1 986 May 1 4<br />
10000<br />
1000<br />
84<br />
43<br />
100<br />
r---<br />
75<br />
10<br />
1<br />
om easured<br />
Dcontrol<br />
.m odified<br />
r---<br />
- - - -<br />
9.8<br />
6.52<br />
1 1 1 1 1 1 1 1 1<br />
~.~ ~<br />
1 2 3 4 5 6 7 8 9 10 11 12 13 14<br />
1 844 439 1 1 1 1 1 1 75 9.8 6.52 1 1<br />
1 1989 62912317 832 19.3 103 20.7 16.1 35.3 4.07 1 1 1<br />
1 1989 62982364 778 23.2 99.4 20.1 16.2 34.9 4.07 1 1 1<br />
1 1 1<br />
15 16 17<br />
1 1 1<br />
1 1 1<br />
1 1 1<br />
Figure 41<br />
Summary of Passau (Lat,Lon in decimal degrees) daily total Cs-137 deposition<br />
1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
Schwandorf (49.33,12.11) Daily Cs·137 Deposition [Bq/mA2]<br />
1986Apr28 ·1986M ay14<br />
10000<br />
1 000<br />
417<br />
814<br />
1 00<br />
- -<br />
12<br />
22<br />
78<br />
31 33<br />
1<br />
10 -<br />
3.2<br />
2.3 2.3<br />
1<br />
rt.1<br />
1 1<br />
1<br />
n n<br />
1 2 3 4 5 6 7 8 9 1 0 11 12 1 3 14 15 16 17<br />
om easured 1 417 479 123 814 22 47 78 31 16 3.2 1 2.3 33 2.3 1 1<br />
Dcontrol 1 129568631713 91 0 30.1 1 1 0 1 2.2 1 3.3 27.92.08 1 1 1 1 1 1<br />
.m odified 1 129568641761 863 32.9 1 06 1 2.1 1 3.5 27.52.08 1 1 1 1 1 1<br />
Figure 42<br />
Summary of Schwandorf (Lat,Lon in decimal degrees) daily total Cs-<br />
137 deposition 1986Apr27 - 1986May14; measured, control run, and<br />
modified cloud base run.<br />
53
Hradec Kralov (SO.21,1S.83)Daily Cs-137 Deposition [Bq/mI\2]<br />
1 986 A P r2 8 - 1 986 May 1 4<br />
10000<br />
3101<br />
1000<br />
I--<br />
241 251<br />
100<br />
10<br />
1<br />
Dm e a su re d<br />
oco"l.ol<br />
.modified<br />
8<br />
- - - _IU -<br />
17.6<br />
1<br />
- - I-- -<br />
1 1 1 1 1 1 1 1 1 1<br />
J ......<br />
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17<br />
1 1 3101 87 241 251 61 1 17.6 13 1 1 1 1 1 1 1<br />
1 39983011195413.7 3.88 67 .9 24.4 12.5 1711 154 1.21 1 1 1 1 1<br />
1 39983038193010.6 6.65 65 .8 23.6 12.5 1711 154 1.21 1 1 1 1 1<br />
Figure 43<br />
Summary of Hradec Kralov (Lat,Lon in decimal degrees) daily total<br />
Os-137 deposition 19S6Apr27 - 19S6May14; measured, control run, and<br />
modified cloud base run.<br />
Kosice (48.73,21.2S)Daily Cs-137 Deposition [Bq/mI\2]<br />
1 986 A P r2 8 - 1 986 May 1 4<br />
100000<br />
10000<br />
1000<br />
0<br />
100<br />
-<br />
.il<br />
32<br />
87 9<br />
10<br />
1<br />
omeasured<br />
Dcontrol<br />
.modifled<br />
rr<br />
nI -<br />
I--<br />
1 1 1·~1 lr. 1<br />
...<br />
1 2 3 4 5 6 7 8 9 10 11 12 13 14<br />
lrJ<br />
1 1 405 55 32 4.2 1 41 41 87 91 1001 1 1<br />
1 926 954 482 7.64 15.7 1.61 1.83 9.27 56.5 1243178761.4 3.66<br />
1 926 954 482 7.55 15.3 1.55 1.82 9.27 56.5 1243178761.4 3.66<br />
1<br />
.--<br />
1 1<br />
15 16 17<br />
1 1 1<br />
1.33 1 1<br />
1.33 1 1<br />
Figure 44<br />
Summary of Kosice (Lat,Lon in decimal degrees) daily total Os-137 deposition<br />
19S6Apr27 - 19S6May14; measured, control run, and modified<br />
cloud base run.<br />
54
-----------------------------------------<br />
Budapest (47.5,19.1) Daily Cs-137 Deposition [Bq/mA2]<br />
1986Apr28 - 1986M ay14<br />
10000<br />
1000<br />
100<br />
10<br />
1<br />
301<br />
239<br />
- f---s-:<br />
863<br />
1 32 124<br />
92<br />
---<br />
2 27.4<br />
22<br />
3<br />
f-- -<br />
1 1<br />
1'1 -.<br />
1 2 3 4 5 6 7 8 9 10 11 12 13 14<br />
1 1 1<br />
15 16 17<br />
omeasured 1 1 301 547 239 61 92 132 28 13 863 124 27 22 1 1 1<br />
Dcontrol 1 331 185 262 72 168 9.3 24 59 275 915 1.9 2.4 1.2 1 1 1<br />
.m odified 1 331 185 263 78 162 10 23 58 275 915 1.9 2.4 1.2 1 1 1<br />
Figure 45<br />
Summary of Budapest (Lat,Lon in decimal degrees) daily total Cs-137<br />
deposition 1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
Mol (51.18,5.12) Daily Cs-137 Deposition [Bq/mA2]<br />
1986Apr28 - 1986M ay14<br />
10000<br />
1000<br />
1191<br />
371 341<br />
100<br />
.0<br />
f--<br />
108<br />
.0<br />
10<br />
- f-- f--<br />
16<br />
10.8<br />
1<br />
1<br />
1<br />
1 1<br />
1 f1 ~ 11<br />
1<br />
1 1 1 1<br />
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17<br />
Dmeasured 1 1 1 58 119 371 108 341 16 1 58 11 1 1 1 1 1<br />
Dcontrol 1 1 1 3.5 39 57 1.8 4.4 2.3 1.1 1 1 1 1 1 1 1<br />
.m odlfied 1 1 1 3.5 40 57 2.1 I 4 2.3 1.1 1 1 1 1 1 1 1<br />
Figure 46<br />
Summary of Mol (Lat,Lon in decimal degrees) daily total Cs-137 deposition<br />
1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
55
1<br />
Harwell (Chilton) (51.61,-1.3) Daily Cs-137 Deposition [Bq/mA2)<br />
1986Apr28 - 1986M ay14<br />
1 000<br />
1 00<br />
9<br />
10 -<br />
0<br />
4<br />
1<br />
2.<br />
2<br />
1.34 1.4<br />
1 1 1 1 1 1 1<br />
1<br />
1 1<br />
r n n D<br />
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17<br />
omeasured 1 1 1 1 1 2.1 6 1 1 4 2 1.3 1 1.4 1 1 9<br />
Dcontrol 1 1 1 1 361 47 1 1 1 1 1 1 1 1 1 1 1<br />
.m odified 1 1 1 1 361 47 1 1 1 1 1 1 1 1 1 1 1<br />
~<br />
Figure 47<br />
Summary of Harwell (Lat,Lon in decimal degrees) daily total Cs-137 deposition<br />
1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
Aachen (50.76,6.1) Daily Cs-137 Deposition [Bq/mA2)<br />
1986Apr28 - 1986M ay14<br />
1 000<br />
1 7 9 1 99<br />
1 00<br />
-<br />
10<br />
1<br />
r---<br />
20<br />
8<br />
2<br />
1 1 1 If) 1 1 1<br />
1 1 1<br />
'---<br />
~ -- n<br />
1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14 1 5 16 1 7<br />
om easured 1 1 1 1 1 1 1 79 1 99 7 20 8 7 1 2 1 1 1<br />
Dcontrol<br />
.m odified<br />
I<br />
1 1 1 3.3 82 87 1 2 11 3.8 1 .2 1<br />
---<br />
1 1 1 1 1 1<br />
1 1 1 3.3 84 86 1 2 1 0 3.8 1 .2 1 1 1 1 1 1<br />
Figure 48<br />
Summary of Aachen (Lat,Lon in decimal degrees) daily total Cs-137 deposition<br />
1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
56
,------------------------------------------------,<br />
Emden (53.35,7.21) Daily Cs-137 Deposition [8q/mA2)1986Apr28·<br />
1986M ay14<br />
1000 ,------- 714 -<br />
------<br />
250<br />
1 1 5<br />
1 00 78 79<br />
47<br />
32<br />
1 2<br />
1 0 - - - ------ f-- f---<br />
1<br />
1 1 1 1 If<br />
1 1 1 1<br />
1<br />
1 2 3 4 5 6 7 8 9 1 0 11 12 1 3 14 1 5 16 17<br />
Dm easured 1 1 1 1 1 1 250 714 1 1 1 5 78 47 79 1 1 12 32<br />
Dcontrol 1 1 1 1 1 .7 1 62 1 66 53 11 4.6 1 1 1 1 1 1 1<br />
.m od ifie d 1 1 1 1 14 1 68 1 52 49 11 4.5 1 1 1 1 1 1 1<br />
Figure 49<br />
Summary of Emden (Lat,Lon in decimal degrees) daily total Cs-137 deposition<br />
1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
Koblenz (50.35,7.6) Daily Cs-137 Deposition [8q/mA2)<br />
1986Apr29 - 1986M ay15<br />
1 000<br />
481<br />
253<br />
1 00<br />
- I--<br />
1 22.9<br />
1 0<br />
1<br />
Om easured<br />
Dcontrol<br />
.m odified<br />
1<br />
27.4<br />
,~<br />
-<br />
1 1 1 1 1 1<br />
--<br />
~~<br />
1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14<br />
1 1 1 1 1 1 48 1 2 53 1 123 27 1 1 1<br />
1 1 22 7 2 5 5 1 25 28 1 0 7.1 2 .5 1 1 1 1<br />
1 1 22 8 .1 2 62 1 1 8 26 9 .7 7.1 2 .5 1 1 1 1<br />
1 1 1<br />
15 1 6 17<br />
1 1 1<br />
1 1 1<br />
1 1 1<br />
Figure 50<br />
Summary of Koblenz (Lat,Lon in decimal degrees) daily total Cs-137 deposition<br />
1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
57
Schleswig (54.45.9.53) Daily Cs-137 Deposition [Bq/mA2]<br />
1986Apr28 - 1986M ay14<br />
1 000<br />
8 54<br />
o • ,<br />
--<br />
1 00<br />
1 0<br />
1<br />
om easured<br />
Dcontrol<br />
.m odlfled<br />
40<br />
23<br />
26<br />
t<br />
17<br />
- - I---<br />
1 1 1 1 1 1 1 1<br />
1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14<br />
1 1 1 1 1 1 8 54 1 583 40 23 26 1 7 1<br />
1 1 1 1 1 47 2 1 1 79 1 9 14 1 .6 1 1 1<br />
1 1 1 1 9 68 1 86 75 1 9 14 1 .6 1 1 1<br />
8<br />
-<br />
1<br />
15 1 6 17<br />
8 1 69<br />
1 1 1<br />
1 1 1<br />
Figure 51<br />
Summary of Schleswig (Lat,Lon in decimal degrees) daily total Cs-137<br />
deposition 1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run_<br />
Bilthoven (52.11.5.18) Daily Cs-137 Deposition [Bq/mA2]<br />
1986Apr28 - 1986M ay14<br />
1 000<br />
22 1<br />
1 00<br />
S--7<br />
95<br />
46<br />
22<br />
1 0<br />
-<br />
5<br />
3<br />
1 1 1 1 1~ 1<br />
1<br />
1 -.1 1 1 1<br />
~ 0<br />
.or<br />
1 2 3 4 5 6 7 8 9 1 2 13 14 1 5 1 6 17<br />
Dm easured 1 1 1 1 1 1 67 9 5 221 22 1 1 1 1 46 5 3<br />
Dcontrol 1 1 1 1 2.7 94 2 1 1 5 2.8 1 .4 1 1 1 1 1 1 1<br />
_modified 1 1 1 1 4.1 94 2 1 14 2.8 1 .3 1 1 1 1 1 1 1<br />
Figure 52<br />
Summary of Bilthoven (Lat,Lon in decimal degrees) daily total Cs-137<br />
deposition 1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run_<br />
58
,---------------------------------------------------------------------------~<br />
Offenbach (50.2,8.65) Daily Cs-137 Deposition [Bq/m A 2)<br />
1986Apr28 -1986M ay14<br />
1 000<br />
1 00<br />
f-- ----6-6<br />
62--<br />
33<br />
35<br />
1 0<br />
-- f-- f-- f--<br />
13<br />
11<br />
r--<br />
1<br />
Dm easured<br />
Dcontrol<br />
.m odified<br />
1 1 1 1 1 1 1 ~1 1 1<br />
1 2 3 4 5 6 7 8 9 1 0 11 12 1 3 14<br />
1 1 1 1 1 1 1 66 13 11 1 33 1 1<br />
1 1 1 58 46 260 1 57 76 16 11 4 1 1 1 1<br />
1 1 1 58 49 272 1 4 9 70 1 5 11 3.9 1 1 1 1<br />
1<br />
1 5 16 1 7<br />
62 35 1<br />
1 1 1<br />
1 1 1<br />
Figure 53<br />
Summary of Offenbach (Lat,Lon in decimal degrees) daily total Cs-<br />
137 deposition 1986Apr27 - 1986May14; measured, control run, and<br />
modified cloud base run_<br />
Glasgow (56.0,-4.49) Daily Cs-137 Deposition [Bq/mA2)<br />
1986Apr28 - 1986M ay14<br />
1 0000<br />
1 71 1<br />
1 000<br />
1 00<br />
1 3 1<br />
r-- r---<br />
20 1 20 1<br />
1 6<br />
1 0<br />
-- 1- 1-- 1- --<br />
1<br />
1 1 1 1 1 1 1<br />
1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14<br />
r:..<br />
1 1 1 1 1<br />
1 5 1 6 17<br />
om easured 1 1 1 1 1 1 1 1 31 171 201 1 6 201 1 1 1 1 1<br />
Dcontrol 1 1 1 1 1 31 56 1 3 1.2 1 1.1 1 1 1 1 1 1 1<br />
.m ad ifie d 1 1 1 1 1 31 56 13 1.2 1 1.1 1 1<br />
I<br />
1 1 1 1 1<br />
Figure 54<br />
Summary of Glasgow (Lat,Lon in decimal degrees) daily total Cs-137<br />
deposition 1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
59
Berlin (52.53,13.42) Daily C s-137 De positio n [B q/m A 2)<br />
1986Apr28 - 1986M ay14<br />
1 00 00<br />
1 000<br />
1 057<br />
285<br />
1 00<br />
--- --<br />
45<br />
1 0<br />
1<br />
om easured<br />
Dcontrol<br />
.m odlfled<br />
-<br />
1 1 1 1<br />
1 .1r<br />
1 1 1 1 1<br />
1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 1 4<br />
1 1 1 1 1 1 1 1 105 285 45 1 1 1<br />
1 1 1 1 1 4.1 1 29 44 19 1 1 3 29 1 1 1<br />
1 1 1 1 2.5 24 1 09 42 19 1 1 3 29 1 1 1<br />
1 1 1<br />
15 16 1 7<br />
1 1 1<br />
1 1 1<br />
1 1 1<br />
Figure 55<br />
Summary of Berlin (Lat,Lon in decimal degrees) daily total Cs-137 deposition<br />
1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
Giessen (50.58,8.67) Daily Cs-137 Deposition [Bq/mA2)<br />
1986Apr28 - 1986M ay14<br />
1 000<br />
743<br />
1 00<br />
-<br />
1 0<br />
5.3<br />
- -<br />
f--<br />
1 1 1r. 1 • 1 1 1 1 1 1 1 1 1 1 1<br />
1<br />
r<br />
1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14 1 5 16 17<br />
Dm easured 1 1 1 1 1 1 1 1 743 5.3 1 1 1 1 1 1 1<br />
Dcontrol 1 1 1 .3 1 66 230 1 73 39 1 8 5.5 1 1 1 1 1 1 1<br />
.m 0 d ifie d 1 1 1 .3 1 .3 88 225 1 58 38 1 8 5.4 1 1 1 1 1 1 1<br />
Figure 56<br />
Summary of Giessen (Lat,Lon in decimal degrees) daily total Cs-137 deposition<br />
1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
60
-----------------------------------------------------------------------<br />
M uenchen (48.13,11.5) Daily Cs-137 Deposition [Bq/mA2]<br />
1986Apr28 -1986M ay14<br />
1 0000<br />
1 397<br />
1 000<br />
- -<br />
1 00<br />
- - -<br />
72<br />
10<br />
- - - - I-- I--<br />
r--<br />
1<br />
Dm easured<br />
Dcontrol<br />
.m odified<br />
1 1 1 1 1 1 1 1 1 1[1 1 1 1<br />
1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14<br />
1 1 1 1 1 1 1 1 1 39 1 1 1 1 1<br />
1 199 1 85 1 52 1 1 9 34 32 1 2 14 3 1 1.7 1 1 1<br />
1 199 1 85 1 57 1 13 33 33 1 2 14 30 1.7 1 1 1<br />
1 1<br />
15 16 1 7<br />
1 1 72<br />
1 1 1<br />
1 1 1<br />
Figure 57<br />
Summary of Muenchen (Lat,Lon in decimal degrees) daily total Cs-<br />
137 deposition 1986Apr27 - 1986May14; measured, control run, and<br />
modified cloud base run.<br />
Berkeley (51.69,-2.42) Daily Cs-137 Deposition [Bq/mA2]<br />
1986Apr28 - 1986M ay14<br />
1 000<br />
1 00<br />
1 0<br />
~<br />
1 1 1 1 1 1 1<br />
'" n 1 ~ 1 1 1<br />
---- ~--<br />
1 2 3 4 5 6 7 8 9 1 0 11 12 13 14<br />
1 1 1<br />
1<br />
1 5 1 6 17<br />
Dm easured 1 1 1 1 1 1 1 1 1.4 1 6.1 1 1 1 1 1 1<br />
Dcontrol 1 1 1 1 305 27 1 1 1 1 1 1 1 1 1 1 1<br />
.m odified 1 1 1 1 305 27 1 1 1 1 1 1 1 1 1 1 1<br />
Figure 58<br />
Summary of Berkeley (Lat,Lon in decimal degrees) daily total Cs-137<br />
deposition 1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
61
-------------------------------------------------<br />
Risoe (55.7,12.07) Daily Cs-137 Deposition [8q/mA2]<br />
1986Apr28 - 1986M ay14<br />
1 000<br />
487<br />
_.-<br />
1 00<br />
1 09<br />
74<br />
2 3<br />
1 0<br />
1<br />
1<br />
1<br />
Dm easured 1<br />
Dcontrol 1<br />
.m odified 1<br />
I-- - - ---<br />
5<br />
2 3 4 5 6 7 8 9 1 0 11 12 1 3 14<br />
1 1 1 1 1 11 1 1<br />
~<br />
1 1 1 1 1 1 1 1 487 1 09 74 23 5<br />
1 1 1 1 1 23 37 12 1 1 0 75 1 1 1<br />
1 1 1 1 5.3 19 37 12 1 1 0 75 1 1 1<br />
1 1 1<br />
1 5 16 1 7<br />
1 1 1<br />
1 1 1<br />
1 1 1<br />
Figure 59<br />
Summary of Risoe (Lat,Lon in decimal degrees) daily total Cs-137 deposition<br />
1986Apr27 - 1986May14; measured, control run, and modified<br />
cloud base run.<br />
4.2.2 Modified-Claud-Base Performance in April Over Germany/Austria.<br />
Herein are described the results of two 5-day <strong>Chernobyl</strong> plume simulations and their<br />
deposition in Germany/Austria. The first simulation is an April deposition control<br />
run using default <strong>Chernobyl</strong> settings as described in Section 3.5.4. The second<br />
simulation is an April deposition cloud base modification run identical to the April<br />
deposition control run, except with 75%-RH cloud bases over land, instead of the<br />
default 80%. Deposition contours from the April deposition control run over the<br />
selected region are depicted in Figure 60. Total deposition from the April deposition<br />
cloud base modification run is shown in Figure 61. The modified-cloud-base deposition<br />
pattern exhibits generally greater deposition than the control run. Also, the<br />
contours are less smooth, indicating higher variability. The difference between the<br />
two fields, displayed in Figure 62, confirms the general increase in deposition with<br />
a lowered cloud base. A quantitative analysis is presented next, then qualitative<br />
analysis of major features of Figure 62.<br />
62
Cu Cs-137 to 1 May Phases 2+3<br />
Control<br />
56N~------~--~--~------~----~~~<br />
55N<br />
53N<br />
52N<br />
50N<br />
49N<br />
48N<br />
47N<br />
45N+-~~--~~~--~-r~~~~~j~·.-.-~--~<br />
5E 6E 7E BE 9E WE i 1 E 12E 13E 14E 15E i 6E 17E 18E<br />
Figure 60 April deposition control run. Modeled <strong>Chernobyl</strong> Cs-137 deposition<br />
concentration [Bq/m2] accumulated over 86.04.26.00Z - 86.05.01.00Z.<br />
63
Cu Cs-137 to 1 May Phases 2+3<br />
75% RH cloud base<br />
56N?-------~----~~-------------~~~<br />
55N<br />
54N<br />
5.3N<br />
52N<br />
51N<br />
50N<br />
49N<br />
48N<br />
47N<br />
46N<br />
Figure 61 April deposition modified cloud base run. Modeled <strong>Chernobyl</strong> Cs-<br />
137 deposition concentration [Bq/m2] accumulated over 86.04.26.00Z<br />
- 86.05.01.00Z.<br />
64
Difference contours<br />
175% Runl - lControl Run~<br />
56N~------~--~~~~--~~~~~~<br />
55N<br />
54N<br />
5.3N<br />
52N<br />
51N<br />
50N<br />
49N<br />
48N<br />
47N<br />
46N<br />
Figure 62<br />
Difference in modeled <strong>Chernobyl</strong> Cs-137 deposition concentration<br />
[Bqjm 2 ] between cloud-base-modified run April deposition and control<br />
run April deposition (modified run deposition minus control run deposition),<br />
accumulated over 86.04.26.00Z - 86.05.01.00Z.<br />
65
Concentrations from each run are extracted at points corresponding to the<br />
395 April-cumulative measurement sites (Subsection 3.5.4) and are compared to the<br />
measurements. As may be suspected from earlier explanations, the cloud base modification<br />
is a relatively subtle adjustment and its impact in a <strong>Chernobyl</strong> simulation<br />
is masked by large errors in other parameters. For display and statistical purposes,<br />
a value of 1 was added to avoid taking the logarithm of zero, then the base-lO logarithm<br />
was taken at each data point, measured and modeled. Adding a constant to<br />
a distribution does not at all affect its correlation with another variable. A normal<br />
probability plot of log-transformed measured deposition values, Figure 63, strongly<br />
confirms the assertion by Rodriguez et al. that the distribution of surface concentration<br />
is log-normal (Rodriguez 95:811). Measurements were of total deposition,<br />
including any Cs-137 deposits prior to the <strong>Chernobyl</strong> accident. To adjust modeled<br />
quantities for pre-<strong>Chernobyl</strong> deposition the lesser of the corresponding measurement<br />
value or 2500Bq was added to each modeled data point. No null measurements are<br />
found in the REM database, so the data is log-transformed without adding a value<br />
of 1 to each data point. Correlation is 0.6059 for a point by point comparison of<br />
the log transformed 5-day Cs-137 measurements in Germany/Austria to log transformed<br />
deposition from the April deposition control run. Correlation is 0.5843 for<br />
the same comparison to the modified-cloud-base run. So, a slightly lower correlation<br />
to measurements is observed using modeled cloud bases lowered to the 75% humidity<br />
threshold. While cloud base parameterization improvement has not been shown<br />
for 5-day <strong>Chernobyl</strong> deposition in Germany and Austria, qualitative analysis of the<br />
results does unearth some clues to the possible role of wet deposition mechanisms at<br />
work in the <strong>Chernobyl</strong> case.<br />
66
Normal Probability Plot<br />
7<br />
-<br />
6 -c-<br />
•<br />
t/)<br />
m 5 +-<br />
-CI)<br />
E<br />
4 -c-<br />
0 3 ,,-<br />
"l"""<br />
C) 2 +-<br />
0<br />
1 -1--<br />
0 I I I I<br />
0 20 40 60 80 100<br />
Sam pie Percentile<br />
Figure 63<br />
Normal Probability Plot of Log Transformed April Cumulative Cs-137<br />
Deposition Measurements in Germany and Austria<br />
The change from 80% to a 75% RH continental cloud base threshold in effect<br />
lowers the cloud base over land without changing the cloud top, the plume height,<br />
or the horizontal pattern of precipitation. In the sensitivity studies in Subsection<br />
4.1.1, the BCS rate was left undisturbed. With the cloud base lower, as long as<br />
the plume is near the cloud base, more of the pollutant will be within the cloud<br />
and less will be below it. So, for modified-cloud-base simulation runs, BCS applies<br />
to less of the plume, and ICS applies to more of the plume. BCS is dependent<br />
on rain duration which, in the model, is always 6hrs. ICS is dependent on rain<br />
amount which varies with each gridpoint, therefore, more variability appears in the<br />
modified-cloud-base run deposition pattern because more ICS is occurring relative<br />
to April deposition control run ICS. Since ICS always counts more in the cloud base<br />
modification run relative to the control run, the sign of the change in deposition<br />
67
depends on which scavenging provided more deposition, lCS or BCS. The fact that<br />
an overall increase in deposition is observed for a lowered cloud base indicates that<br />
modeled lCS is removing more Cs-137 overall than modeled BCS over Germany in<br />
April, 1986. While lCS is proportional to total amount of rain, BCS is proportional<br />
to the amount of time it rains. lCS tends to dominate the wet deposition from<br />
a ten-minute downpour, and BCS is the predominate deposition mechanism for a<br />
3-day drizzle.<br />
The general precipitation pattern over the sampled domain for late April is<br />
heaviest over Austria and gradually diminishing northward (Appendix C). The best<br />
opportunity for BCS to dominate lCS is where the model precipitation is lightest,<br />
north in this case. The patch of negative deposition difference values over the north<br />
half of the Germany/Poland border in Figure 62 is consistent with this thinking.<br />
Since this area is the most likely place for lCS to apply to more of the plume at<br />
the expense of the more dominant BCS. The other negative anomaly, over the<br />
heart of Czechoslovakia, is more difficult to assess. Since, between April deposition<br />
control and April deposition modified-cloud-base runs, no changes in scavenging<br />
occur above the 80%-RH level or below the 75 %-RH level, one only needs to assess<br />
what happens in the layer between those levels. Perhaps, during one or more<br />
precipitation events in the April deposition control run, BCS only slightly reduces<br />
pollutant concentration in the 75%-80% layer, leaving plenty of pollutant in the<br />
layer for deposition in a downstream location. Then, in a parallel April deposition<br />
modified-cloud-base run, during the same precipitation events, lCS depletes the 75%-<br />
80%-RH layer completely. So, the downstream location will have less pollutant<br />
available for scavenging, therefore, deposition will amount to less than that in the<br />
April deposition control run, as in the Czechoslovakia negative feature in Figure 62.<br />
If the April deposition modified-cloud-base run were truth, and the April deposition<br />
control run included the exact scavenging error, the SECP would be somewhere<br />
upstream from the negative feature.<br />
68
V. Conclusions<br />
5.1 Sensitivity Runs<br />
It has been observed that for a 6-hr time step and a polar stereographic computational<br />
grid as coarse as 60km in resolution, a HySPLIT transport and deposition<br />
simulation requires over 6hr s (one time step) to produce a realistic deposition pattern.<br />
Anomalous modeled deposition occurs within 100km of the source and is<br />
probably an artifact of interpolations used to initialize the pollutant plume. It has<br />
been shown that relative changes in deposition due to altering the ICS efficiency<br />
parameter in HySPLIT are nearly proportional to the ICS efficiency alteration. In<br />
other words, when the efficiency of ICS is doubled, a doubling of rain-out deposition<br />
is the result as long as pollutant concentrations are not significantly reduced. The<br />
acronym, SECP (Scavenging Error Crossover Point), has been coined describing the<br />
location at which the effect of scavenging errors on plume concentration bottoms<br />
out and begins to have the opposite effect. For example, over-scavenging initially<br />
produces excess deposition, but at the SECP no excess deposition occurs because<br />
the plume concentration has dropped enough. Downstream from the SEep, deposition<br />
is underestimated as the plume concentration continues to drop too fast.<br />
ICS efficiency sensitivity test simulations were run out to 12hrs and 600km in light<br />
rain (approximately Imm per 6hrs). At these limits, no SECP was apparent in the<br />
deposition pattern. Future studies may help refine the understanding of the SECP<br />
in general and of its possible range of influence on <strong>Chernobyl</strong> simulations.<br />
5.2 Cloud Base Modification Runs<br />
The predictive ability of HySPLIT appears to improve very slightly when modified<br />
to model continental cloud bases at 75% RH instead of its default 80%. The<br />
slight improvement is based on a higher correlation with measured data (0.5050)<br />
of the total bulk results of a modified-cloud-base daily deposition run than that<br />
69
(0.5037) of the total bulk results of the control daily deposition run. Given the<br />
combined severe uncertainties of <strong>Chernobyl</strong> emissions, of model precipitation and<br />
wind fields, and of the measurements themselves, and given that a 0.50 Pearson<br />
correlation is equivalent to raw guessing, the results in this format are inconclusive.<br />
However, an interesting clue arises from a city-by-city analysis. When correlation<br />
coefficients are calculated for each city, a trend of degraded predictive ability with<br />
time emerges. Figures 64 and 65 summarize the correlation of both control run and<br />
modified-cloud-base runs with measured data. The cities are in order by plume arrival,<br />
i.e., by earliest non-zero measurement recorded at each city. The cities where<br />
the modified cloud base has improved HySPLIT's predictive ability were Helsinki,<br />
Hof, Schwandorf, Hradec Kralov, Budapest, and Mol. There is virtually no change<br />
in predictive ability at Bratislava, Moravsky Krumlov, Passau, or Kosice (map of<br />
daily deposition cities in Section 3.5.3, Figure 12). So, the modified cloud base has<br />
performed slightly better than the control at the cities nearest <strong>Chernobyl</strong> (Figure<br />
64, except for Harwell) and the same as, or worse than, the control run at the cities<br />
furthest from <strong>Chernobyl</strong> (Figure 65, and Harwell). This trend could be an indication<br />
that the modified cloud base has induced an actual predictive ability that degrades,<br />
as expected, with distance from the source.<br />
Comparisons of HySPLIT control and cloud-base-modified runs of deposition<br />
in Germany and Austria in April, 1986 indicate a decrease in the predicitive ability<br />
of a HySPLIT <strong>Chernobyl</strong> simulation. Evidence of a SECP just upstream from<br />
Germany was examined in Section 4.2.2. These results are not inconsistent with the<br />
daily deposition run results since the modified-cloud-base run decreased accuracy at<br />
three of five German cities.<br />
By means of a HySPLIT cloud base parameterization revision (75%-RH continental<br />
cloud bases), very small improvement has been demonstrated in the accuracy<br />
of a HySPLIT <strong>Chernobyl</strong> Cs-137 daily deposition simulation, while a decrease in accuracy<br />
has resulted from a 5-day-cumulative <strong>Chernobyl</strong> deposition simulation over<br />
70
Correlation of Control Run & Cloud-Base-Modified Run Daily Deposition to<br />
Measured Daily Deposition<br />
j<br />
1.000<br />
0.800 r--<br />
-<br />
-<br />
0.600 r- - r- -rr-<br />
f----<br />
-<br />
I!! D.400 - r-- r- - - r-- r- r--=<br />
~<br />
0.200 - -<br />
f--- f--- - f--- I- f---<br />
0.000 L.- L- '- L- r-. L- L-<br />
-<br />
L-<br />
-<br />
Bratislav Moravsk Schwan Hradec Budapes<br />
Helsinki Hot Passau Kosice Mol Harwell<br />
a y dorf Kralov t<br />
I 0 control run 0.640 0.800 0.895 0.490 0.091 0.685 0.614 0.578 0.848 0.604 0.383<br />
I_ modified run 0.668 0.800 0.894 0.494 0.090 0.686 0.618 0.578 0.850 0.617 0.383<br />
Figure 64<br />
Set 1. Pearson correlation coefficients of modeled <strong>Chernobyl</strong> Cs-137<br />
daily deposition (control run and modified-cloud-base run) against measured<br />
<strong>Chernobyl</strong> Cs-137 daily deposition, by city. In order by earliest<br />
deposition measurement at each city.<br />
Correlation of Control Run & Cloud-Base-Modified Run Daily Deposition to<br />
Measured Daily Deposition<br />
1.000<br />
0.800<br />
c 0.600<br />
~ r- r-<br />
..!!! 0.400 r-- t-- t--<br />
-<br />
~<br />
u<br />
0 0.200 '---<br />
t-- I-- -<br />
0.000 --<br />
~ '----<br />
-0.200<br />
r-<br />
KI<br />
~ LII<br />
Aache Emde Schle Koble Biltho Offen Glasg Giess Muen Berkel<br />
Berlin<br />
Risoe<br />
n n swig nz yen bach ow en chen ey<br />
10 control run 0.585 0.577 0.507 0.541 0.770 -0.054 -0.039 0.615 0.340 -0.105 -0.098 0.600<br />
I- modified run 0.583 0.516 0.498 0.509 0.749 -0.059 -0.041 0.573 0.332 -0.105 -0.098 0.581<br />
Figure 65 Set 2. Pearson correlation coefficients of modeled <strong>Chernobyl</strong> Cs-137<br />
daily deposition (control run and modified-cloud-base run) against measured<br />
<strong>Chernobyl</strong> Cs-137 daily deposition, by city. In order by earliest<br />
deposition measurement at each city.<br />
71
Germany and Austria. The slightly improved HySPLIT performance in the daily<br />
deposition case does not prove indisputably that a 75%-RH continental cloud base<br />
is more true to the <strong>Chernobyl</strong> environmental conditions. A credible source term<br />
construction could be artificially devised such that 80%-RH continental cloud bases<br />
would produce the more "accurate" daily deposition. The 50% margin of error<br />
in <strong>Chernobyl</strong>'s estimated daily emissions provides wide latitude to do so. Simulations<br />
of atmospheric plume transport and deposition are vulnerable to limitations<br />
in the representation of several aspects of the problem. These aspects include<br />
the particle size distribution and its space and time variations, turbulence, atmospheric<br />
instability, and other dry transport processes, the solubility of the particles,<br />
cloud formation and dissipation processes, fog deposition, and particle resuspension<br />
(Knap 88:48). Cumulative measurements representing all or several days of the<br />
<strong>Chernobyl</strong> deposition are likely to have a large positive bias since Cs-137 deposition<br />
measurement locations would tend to be where the highest radioactivity levels had<br />
been detected. The slightly positive results from the daily deposition simulation<br />
should in no way, then, be taken as proof of a general improvement in modeled<br />
cloud base. One can expect much larger gains in the accuracy of wet deposition<br />
modeling by improving the location, timing, and amount of precipitation inputs. It<br />
appears, in fact, that verifiable improvements to wet deposition parameterizations<br />
must wait both for increased resolution and accuracy of precipitation modeling, and<br />
for an experimental wet deposition case with more precise emission specifications<br />
and more homogeneous, higher resolution deposition measurements. Although the<br />
specific wet deposition parameterization test yielded no conclusive evidence of either<br />
better or worse performance, the exercise constitutes a meaningful starting point for<br />
a researcher interested in either refining the <strong>Chernobyl</strong> source term, or using <strong>Chernobyl</strong><br />
data for validating transport or deposition mechanisms in a model where wet<br />
deposition is a factor, or learning how to use the HySPLIT model and becoming<br />
familiar with its capabilities and limitations.<br />
72
5.3 Future Research Opportunities<br />
HySPLIT lacks the modeling of fog deposition. No clouds (fog) are diagnosed<br />
in the surface layer in HySPLIT, so only dry deposition occurs within the surface<br />
layer in the model. Modification of HySPLIT to include accurate fog modeling<br />
and the increased surface layer deposition that results, especially in up-slope wind<br />
instances, and investigation into its impact on <strong>Chernobyl</strong> could bring model results<br />
more in line with <strong>Chernobyl</strong> deposition measurements. Fog parameterization is an<br />
even larger challenge than cloud parameterization, so unless approached carefully,<br />
adding fog deposition to a model could easily hurt the accuracy of modeled deposition<br />
more than it helps. One could probably make the same argument about cloud<br />
parameterization. Until liquid and ice cloud water content variables are available<br />
routinely from meteorological models, transport model cloud parameterization in<br />
general would still benefit from more accurate cloud diagnosis. The MRF model<br />
run by the National Weather Service (NWS) in addition to treating clouds differently<br />
over land and sea, makes finer cloudiness distinctions by relative humidity in<br />
predefined latitude regions and in four predefined vertical layers based on Real Time<br />
Nephanalysis (RTNEPH) data from USAF (NWS 01). Slingo (Slingo 87) poses and<br />
validates a more complex diagnostic cloud parameterization scheme accounting for<br />
relative humidity, vertical velocity and static stability (specifically, potential temperature<br />
change in the vertical). His approach holds promise for improving regional<br />
deposition distinctions between cumuliform and stratiform precipitation events.<br />
Further model comparisons to <strong>Chernobyl</strong> deposition should include as much<br />
measurement data as possible, increasing the span and resolution of observations in<br />
space and in time to strengthen confidence in results. Since surface-based observations<br />
near the source are scarce, aerial gamma spectrometry measurements taken<br />
over Russia could serve that purpose (DeCort 98). Even though there was about<br />
2000 - 3000Bq/m 2 of Cs-137 from weapons fallout on the ground before the accident,<br />
these readings could help improve model representation of the <strong>Chernobyl</strong> plume early<br />
73
in the accident because millions of Bqjm 2 of Cs-137 were deposited on Ukraine and<br />
Belarus (DeCort 98). The difficulties of source term uncertainty are not unique to<br />
<strong>Chernobyl</strong>. Transport analysts responding to any urgent, short-notice call for an<br />
emission simulation normally have only a crude estimation of the source term. Since<br />
the effective release height is highly dependent on static stability, one could investigate<br />
the modeled vertical profiles at <strong>Chernobyl</strong> and the nearest observed atmospheric<br />
soundings. Principles developed in this project could be applicable and valuable to<br />
operational simulations. The future of transport modeling, like the future of general<br />
meteorological forecasting, may look like ensemble forecasting. Motivated by a<br />
statistical view on stochastic processes like weather, an ensemble forecast is a set of<br />
simulations made up of a best-guess control run and a set of perturbation runs, each<br />
with a slightly different reasonable departure from the control run. An ensemble of<br />
forecasted patterns should provide helpful information about the spectrum of possible<br />
outcomes and about the confidence of any particular run (Draxler OOb). This<br />
method also provides an ongoing opportunity to generate further clues about which<br />
variables are important under specific synoptic regimes.<br />
74
--------------------------------------------- ._--<br />
Appendix A. Glossary of Acronyms<br />
• AFTAC - Air Force Technical Applications Center<br />
• AGL - Above Ground Level<br />
• ARL - Air Resources Laboratory<br />
• ATMES - Atmospheric Thansport Model Evaluation Study<br />
• BCS - Below-Cloud Scavenging<br />
• CEC - Commission of the European Communities (or just EC, European Commission)<br />
• EC - European Commission (see CEC)<br />
• ECMWF - European Centre for Medium-range Weather Forecasts<br />
• ETA - Not an acronym; weather model named after coordinate system with<br />
vertical coordinate, eta (the Greek letter, rt)<br />
• GRADS - GRidded Analysis Display System<br />
• GRlB - GRldded Binary format<br />
• HTML - HyperText Markup Language<br />
• HySPLIT - Hybrid Single-Particle Lagrangian Integrated Thajectories<br />
• ICS - In-Cloud Scavenging<br />
• JRC - Joint Research Centre<br />
• NEA - <strong>Nuclear</strong> Energy Agency<br />
• NOAA - National Oceanographic and Atmospheric Administration<br />
• NCAR - National Center for Atmospheric Research<br />
• REM - Radioactivity Environmental Monitoring<br />
• RH - Relative Humidity<br />
75
• RTNEPH - Real Time Nephanalysis<br />
• SECP - Scavenging Error Crossover Point<br />
• UCAR - University Corporation for Atmospheric Research<br />
• USAEDS - United States Atomic Energy Detection System<br />
• USAF - United States Air Force<br />
76
Appendix B. Radioactivity Primer<br />
B.l Ionizing Radiation<br />
A radioactive atom is an unstable isotope characterized by the high-energy<br />
radiation its nucleus emits upon spontaneous decay (or disintegration) to a more<br />
stable state. This "ionizing radiation" packs enough energy to strip electrons from<br />
materials it hits. Ionizing radiation from a radioactive atom can take the form of:<br />
1. an alpha particle (2 protons with 2 neutrons, i.e., an electron-stripped<br />
helium nucleus)<br />
- can be shielded by a few inches of air<br />
2. a beta particle (stripped electron)<br />
- can be shielded by several inches of plastic<br />
3. gamma ray or x-ray (high-frequency electromagnetic wave)<br />
- can penetrate lead<br />
4. a neutron (stripped)<br />
- can penetrate thick lead shields<br />
Figure 66 (UIC 00) illustrates typical shielding requirements for each of the<br />
four types of ionizing radiation.<br />
77
Figure 66<br />
Typical Shielding Requirements for Different Ionizing Radiation Types<br />
from UIC, 00<br />
There are (at least) three ways to measure ionizing radiation.<br />
• Radiation Activity, a measure of the number of atomic disintegrations per unit<br />
time [e.g., in Bq = S-l]<br />
• Radiation Exposure, a measure of the amount of gamma or x-rays present [e.g.,<br />
in coulombs/kg]<br />
• Radiation Dose, a measure of the amount of radiation absorbed by a subject<br />
[e.g., in sieverts]<br />
- See http://physics.nist.gov/cuu/Units/SIdiagram.html for an extensive summary<br />
of SI (International System) units.<br />
B.2 Cesium-13?<br />
The becquerel is a common unit of Cs-137 deposition radioactivity. One becquerel<br />
of Cs-137 is the amount of Cs-137 substance in which 1 unstable cesium atom<br />
per second undergoes atomic disintegration (emitting a beta particle and gamma<br />
radiation) (MSE 00). The average radiation dose in 1998 from 1kBq/m 2 of Cs-137<br />
deposited in 1986 is about 1 to 2f-LSv. Where soils are more conducive to human<br />
exposure, the average dose is closer to 20f-LSv (DeCort 98:22). Radioactive xenon<br />
78
gas, an abundant product of nuclear fission, decays to Cs-137 which then readily<br />
condenses onto particles present (Glasstone 77:389). Cs-137 decays to barium-137<br />
(Ba-137) with an ionizing radiation (beta particles and gamma rays) of O.662M e V<br />
per decay (Serway 92). The effective dose is highly dependent on the pathway<br />
(respiratory system, skin, digestive system). Although the health effects of exposure<br />
to <strong>Chernobyl</strong>'s fallout should not be trivialized, to date the only clear evidence<br />
for a confirmed correlation between <strong>Chernobyl</strong> fallout dose and illness is thyroid<br />
cancer in children induced by exposure to the iodine isotope, 1-131. Because the<br />
detrimental health effects of Cs-137 are not sudden, and because of deficient human<br />
health records before the accident, it is difficult to isolate the effects of exposure to<br />
<strong>Chernobyl</strong> accident radiation from the existing widespread decline in the Russian<br />
population's general health.<br />
For emergency planning purposes (one application of atmospheric transport<br />
modeling), the uncertain concentration effects of local land use, runoff, and population<br />
habits (DeCort 98:22), combined with the uncertain health effects of radionuclide<br />
exposure/ingestion introduce enough uncertainty to cloud the relative importance<br />
of the magnitude of operational concentration estimates and the location and<br />
timing of radionuclide deposition. To illustrate, consider the evacuation planner<br />
who may be much more interested in which side of a mountain (and when) a radioactive<br />
plume may settle than in exactly how much fallout will land in a certain<br />
neighborhood.<br />
79
Appendix C. Reanalyzed Precipitation Fields from the ECMWF<br />
Model<br />
To facilitate informal diagnosis of wet deposition in simulations within this thesis,<br />
6-hrly model precipitation contours are furnished on the following pages. The valid<br />
time for each plot is the end of the six-hour accumulation. Graphics are produced<br />
with display.exe utility included with HySPLIT software. 86/04/31/00 UTC implies<br />
86/05/01/00 UTC.<br />
80
Table 1<br />
Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr25.<br />
1986042506 missing,<br />
but not needed for comparison<br />
to <strong>Chernobyl</strong> deposition<br />
TPP6 ( mm) AT HEIGHT: 1.000<br />
• 1.00E+OO .2.00E+OO • 3.00E+OO • 4.00E+OO<br />
• 5.00E+OO t. 6.00E+OO O.OOE+OO O.ODE+OD<br />
TPP6 ( mm) AT HEIGHT: 1,000<br />
• 1,OOE+OO • 2,OOE+OO • 3,OOE+OO III 4,OOE+OO<br />
• 5,OOE+OO r:;'" 6,OOE+OO O,OOE+OO O,OOE+OO<br />
TPP6 ( mm) AT HEIGHT: 1.000<br />
• 1.00E+OO • 2.00E+DO • 3.DOE+OO III 4.00E+OO<br />
• 5.00E+OO '6.00E+OO O,ODE+OO O.OOE+OO<br />
81
Table 2<br />
Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr26.<br />
TPP6 ( mm) AT HEIGHT: 1.000<br />
III 1.00E+OO III 2.00E+OO III 3.00E+OO III 4.00E+OO<br />
m 5.00E+OO r" 6.00E+OO O.OOE+OO O.OOE+OO<br />
III 1.00E+OO III 2.00E+OO<br />
m 5.00E+OO ~X! 6.00E+OO<br />
III 4.00E+OO<br />
O.OOE+OO<br />
TPP6 ( mm) AT HEIGHT: 1.000<br />
III 1.00E+OO III 2.00E+OO III 3.00E+OO III 4.00E+OO<br />
ij 5.00E+OO JZ.:. 6.00E+OO O.OOE+OO O.OOE+OO<br />
III 1.00E+OO III 2.00E+OO<br />
r'! 5.00E+OO v:' 6.00E+OO<br />
III 4.00E+OO<br />
O.OOE+OO<br />
82
Table 3<br />
Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr27.<br />
mm)<br />
.1.00E+OO<br />
• 5.00E+OO r;;;£ 6.00E+OO<br />
.4.00E+OO<br />
O.OOE+OO<br />
• 1.00E+OO • 2.00E+OO<br />
• 5.00E+OO ~Y' 6.00E+OO<br />
.4.00E+OO<br />
O.OOE+OO<br />
TPP6 ( mm) AT HEIGHT: 1.000<br />
• 1.00E+OO • 2.00E+OO • 3.00E+OO III 4.00E+OO<br />
• 5.00E+OO !hl 2 [ 6.00E+OO O.OOE+OO O.OOE+OO<br />
TPPS ( mm) AT HEIGHT: 1.000<br />
• 1.00E+OO • 2.00E+OO • 3.00E+OO III 4.00E+OO<br />
• 5.00E+OO ~;;2 6.00E+OO O.OOE+OO O.OOE+OO<br />
83
Table 4<br />
Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr28.<br />
mm)<br />
• 1.ooE+OO • 2.00E+OO<br />
• 5.00E+OO 5tJ~ 6.00E+OO<br />
.4.00E+OO<br />
O.OOE+OO<br />
• 1.00E+oO • 2.ooE+oo<br />
• 5.00E+OO ~:i! 6.00E+OO<br />
.4.00E+OO<br />
O.OOE+OO<br />
NOAA AIR RESOURCES LABORATORY<br />
TPP6 ( mm) AT HEIGHT: 1.000<br />
• 1.00E+OO • 2.00E+OO • 3.00E+OO .4.00E+OO<br />
• 5.00E+OO ~l;;~ 6.00E+OO O.OOE+OO O.OOE+OO<br />
• 1.00E+OO • 2.00E+OO<br />
iL~ 5.00E+OO ~::;; 6.00E+OO<br />
.4.00E+OO<br />
O.OOE+OO<br />
84
Table 5<br />
Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr29.<br />
TPPS ( mm) AT HEIGHT: 1.000<br />
• 1.00E+OO • 2.00E+OO • 3.00E+OO • 4.00E+OO<br />
• 5.00E+OO ~F 6.00E+OO O.OOE+OO O.OOE+OO<br />
TPPS ( mm) AT HEIGHT: 1.000<br />
• 1.00E+OO • 2.00E+OO .3.00E+OO • 4.00E+OO<br />
• 5.00E+OO fi;i1! 6.00E+OO O.OOE+OO O.OOE+OO<br />
( mm)<br />
• 1.00E+OO • 2.00E+OO<br />
• 5.00E+OO I:~: 6.00E+OO<br />
iii! 4.00E+OO<br />
O.OOE+OO<br />
( mm) AT ,"~",o.rr<br />
• 1.00E+OO • 2.00E+OO<br />
• 5.00E+OO ~"I( 6.00E+OO<br />
iii! 4.00E+OO<br />
O.OOE+OO<br />
85
Table 6<br />
Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr30.<br />
TPP6 ( mm) AT HEIGHT: 1.000<br />
• 1.00E+OO • 2.00E+OO • 3.00E+OO • 4.00E+OO<br />
• 5.00E+OO r+:~: 6.00E+OO O.OOE+OO O.OOE+OO<br />
mm)<br />
• 1.00E+OO • 2.00E+OO<br />
• 5.00E+OO fjj';j 6.00E+OO<br />
.4.00E+OO<br />
O.OOE+OO<br />
NOAA AIR RESOURCES LABORATORY<br />
TPPG ( mm) AT HEIGHT: 1.000<br />
• 1.00E+OO • 2.00E+OO • 3.00E+OO • 4.00E+OO<br />
• 5.00E+OO I~~ 6.00E+OO O.OOE+OO O.OOE+OO<br />
TPP6 ( mm) AT HEIGHT: 1.000<br />
• 1.00E+OO • 2.00E+OO • 3.00E+OO • 4.00E+OO<br />
r;;; 5.00E+OO f:.' 6.00E+OO O.OOE+OO O.OOE+OO<br />
86
Appendix D. HySPLIT Settings<br />
This appendix provides exact HySPLIT model settings for all representative simulations<br />
run for this thesis. An attempt has been made to provide enough supplementary<br />
comments to enable the reader to reproduce the simulations in this thesis with<br />
a functioning version of HySPLIT. Australian Meteorological Magazine carried an<br />
article covering the general capabilities of HySPLIT Version 4, the transport modeling<br />
software used for this thesis (Draxler 98b). A complete description of the model<br />
including dispersion equations is available in NOAA Technical Memorandum ERL<br />
ARL-224 (Draxler 98a). The HySPLIT executable program and documentation is<br />
available for download at the following website:<br />
http://www.arl.noaa.gov/ss/models/gethysplit.html<br />
Initial settings for HySPLIT runs are adapted from settings used for <strong>Chernobyl</strong><br />
simulation by Air Resources Laboratory. Some details are available online at:<br />
http://www.arl.noaa.gov/ss/transport/chernobyl.html<br />
D.l Release Height Sensitivity Runs<br />
Following this paragraph is a line-by-line breakdown of the 'Control' file settings<br />
for the HySPLIT diagnostic run from Section 2.5 with a 1500-m point source.<br />
HySPLIT 'Control' file (an ordinary text file) format requires that each numbered<br />
item appears (without the number) on a new line in the 'Control' file. Each numbered<br />
item below is followed by its description. Zeroes in line 27 would trigger<br />
calculations of gaseous emissions. The nominal 1.0 values in line 27 signal to the<br />
model that the pollutant is in particle form. A specified non-zero deposition velocity<br />
in line 28 eliminates the need for the model to calculate fall speed from particle<br />
attributes in line 27. See the HySPLIT User's Guide (Draxler 99) for more details<br />
87
on the 'Control' file settings. The only changes for the 3000-m, 4000-m, and 5000-<br />
m release height sensitivity simulations are the release heights in line 3 and unique<br />
output grid names in line 20.<br />
1. 86 4 25 21<br />
Simulation Starting Time [yy mm dd hh]<br />
2. 1<br />
Number of Emission Starting Locations (double it for uniform vertical<br />
line sources)<br />
3. 51.38 30.1 1500.0<br />
Emission Latitude [decimal degrees] Longitude and Emission Height<br />
[mAGL]<br />
4. 123<br />
Total Simulation Run Time [hours]<br />
5. 0<br />
Vertical Coordinate Type for Simulation Run (0 defaults to met. model's)<br />
6. 10000.0<br />
Ceiling, or Top of Model [m AGL]<br />
7. 1<br />
Number of Setup Meteorology Files<br />
8. D:/HySPLIT /hysplit4/metdata/ChernMet/<br />
Path to Meteorology File<br />
9. analysisp. bin<br />
Filename of Meteorology File<br />
10. 1<br />
88
Number of Pollutants<br />
11. Cs<br />
Pollutant Identification<br />
12. 6.65E+ 14<br />
Pollutant Emission Rate [hr-l] (Concentration [m- 3 ] or Deposition [m- 2 ]<br />
output will be in these units)<br />
13. 24<br />
Hours of Emission<br />
14. 86 4 25 21 23<br />
Release Start Time [yy ma dd hh mn]<br />
15. 1<br />
Number of Output Grids to Generate<br />
16. 48.0 13.0<br />
Center of Output Grid Latitude [decimal degrees] Longitude<br />
17. 0.5 0.5<br />
Spacing (Resolution) of Output Grid Latitude [decimal degrees] Longitude<br />
18. 26.0 36.0<br />
Span (Length and Width) of Output Grid Latitude [decimal degrees]<br />
Longitude<br />
19 .• j<br />
Path Specification for Output Grid File<br />
20. 1500 m<br />
Output Grid Filename<br />
89
21. 1<br />
Number of Vertical Levels in Output Grid<br />
22. 0<br />
Height of Level [m AGL] (0 Triggers Deposition Calculation)<br />
23. 864 26 0 0<br />
Output Grid Sampling Start [yy ma dd hh mn]<br />
24. 86 5 100<br />
Output Grid Sampling End [yy ma dd hh mn]<br />
25. 0 120 0<br />
Output Grid Concentration Type (O=Average or l=Snapshot) and Interval<br />
[hh mn]<br />
26. 1<br />
Number of Deposition Setups (Must Match Number of Pollutants)<br />
27. 1.0 1.0 1.0<br />
Particle Diameter [11m], Density [g/ ee] and Shape Factor<br />
28. 0.0001 0.0 0.0 0.0 0.0<br />
Deposition Velocity [m/ s], Molecular Weight [g], A-Ratio, D-Ratio,<br />
and Effective Henry's Constant<br />
29. 0.0 3.2E+05 5.0E-05<br />
Actual Henry's Constant [M/atm] , In-cloud Scavenging Efficiency Ratio<br />
[L/ L], and Below-cloud Scavenging Rate [S-l]<br />
30. 10976.0<br />
Pollutant Radioactive Decay Half-life [days] (Airborne and Deposited)<br />
31. 0.0<br />
90
ReSllspension Factor [m -1]<br />
D.2 Comparison to ARL Chemobyl Simulation<br />
The following HySPLIT control file contents correspond to those used by ARL<br />
(ARL OOa) and discussed in Section 3.3.<br />
86 4 25 21, 2, 51.38 30.1 750.0, 51.38 30.1 1500.0, 123, 0, 10000, 1, I:j, forecast.bin,<br />
1, C137, 1.00E+15, 24, 86 4 25 21 0, 1, 50.0, 10.0, 0.5 0.5, 30.0 40.0, .j,<br />
dup, 1, 0, 864 27 ° 0, 99 12 31 24 60, ° 84 0, 1, 1.0 1.0 1.0, 0.0001 0.0 0.0 0.0 0.0,<br />
0.0 3.2E+05, 5.0E-05, 10976, °<br />
D.3 In-Cloud Scavenging Sensitivity Control Run<br />
The following HySPLIT control file contents are used for the ICS sensitivity<br />
control run described in Section 3.4.<br />
86 4 25 21, 2, 48.0 11.0 1250.0, 48.0 11.0 1750.0, 147, 0, 10000.0, 2, E:j,<br />
apr86.bin, E:j, may86.bin, 1, cs, 6.65E+14, 7.0, 864 25 21 0, 2, 50.5 12.0, 0.5 0.5,<br />
11.0 14.0, .j, s3dn6, 1,0,86426 00,86427 ° 0, 060, 50.5 12.0, 0.5 0.5, 11.0 14.0,<br />
.j, s3dn24, 1, 0, 86427 ° 0, 99 12 31 240, ° 24 0, 1, 1.0 1.0 1.0, 0.0001, 0.00.00.0<br />
0.0, 0.0 3.2E+05 5.0E-05, 10976.0, 0.0<br />
D.4 Daily Deposition Control Run<br />
The following HySPLIT control file contents are used for the cloud base modification<br />
control run described in Section 3.5.3.<br />
86 4 25 21, 2, 51.38 30.1 1250.0, 51.38 30.1 1750.0, 483, 0, 10000.0, 2, E:j,<br />
apr86.bin, E:j, may86.bin, 1, Cs, 6.65E+14, 7.0,86425 21 0, 1,54.0 10.0, 1.0, 1.0,<br />
14.032.0, .j, dlyO 1. hyc , 1, 0, 86427 ° 0, 99 1231 00, ° 240, 1, 1.0 1.0 1.0, 0.0001<br />
0.0 0.0 0.0 0.0, 0.0 3.2E+05 5.0E-05, 10976.0, 0.0<br />
91
D.5 <strong>Chernobyl</strong> Control Run - Cumulative Deposition on Germany and Austria to<br />
OOZ, 1986MayOl<br />
The following HySPLIT control file contents are used for the cloud base modification<br />
control run described in Section 3.5.4. Control run resolution increased to<br />
0.05 deg lat & Ion for compatibility with display software, GRADS.<br />
Phase II Germany I Austria Control Run HySPLIT Settings<br />
864264,2,51.3830.1350.0,51.3830.1850.0,116,0,10000.0, 1, E:/Chernmet/,<br />
analysisp.bin, 1, Cs, 8.8E+14, 20.0, 8642640,1,45.518.0,0.050.05,21.026.0, .1,<br />
cuall2, 1, 0, 86 4 26 4 0, 86 5 1 0 0, 0 116 0, 1, 1.0 1.0 1.0, 0.0001 0.0 0.0 0.0 0.0, 0.0<br />
3.2E+05 5.0E-05, 10976.0, 0.0<br />
Phase III Germany I Austria Control Run HySPLIT Settings<br />
864270,2,51.3830.1350.0,51.3830.1850.0,96,0,10000.0, 1, E:/Chernmet/,<br />
analysisp.bin, 1, Cs, 2.92E+ 14, 24.0, 86 4 27 0 0, 1, 45.5 18.0, 0.05 0.05, 21.0 26.0,<br />
.1, cua1l3, 1, 0, 864 270 0, 86 5 1 00, 0 96 0, 1, 1.0 1.0 1.0, 0.0001 0.00.00.00.0,<br />
0.0 3.2E+05 5.0E-05, 10976.0, 0.0<br />
D.6 Greece Diagnostic Run HySPLIT Settings - Emission 10m to 1750m<br />
The following HySPLIT control file contents, and minor variations on it, are<br />
used for the April Greece omission diagnostic runs in Appendix G.<br />
8642521,2,51.3830.110.0,51.3830.11750.0,123,0,10000, 1, E:/Chernmet/,<br />
analysisp.bin, 1, Cs, 7.66E+13, 123,8642521 0, 1,45.5 18.0, 0.50.5,21.026.0, .1,<br />
t1750bl0, 1, 0, 86 4 25 21 0, 86 5 1 0 0, 0 123 0, 1, 1.0 1.0 1.0, 0.0001 0.0 0.0 0.0 0.0,<br />
0.0 3.2E+05 5.0E-05, 10976, 0<br />
92
Appendix E. Political Map of Europe<br />
Faroe<br />
Isl~<br />
"<br />
Figure 67<br />
Current Political Map of Europe<br />
93
Appendix F. HySPLIT Source Code Modification<br />
HySPLIT is configured to handle up to eleven land use types, i.e. one water type<br />
(type number 7), and ten terrestrial types.<br />
HySPLIT source code was provided<br />
for this thesis courtesy of Roland Draxler at ARL. Modifying the HySPLIT source<br />
code for the cloud base split parameterization requires a simple code change. The<br />
following is an excerpt from the HySPLIT subroutine, 'depelm.f' with all necessary<br />
modification.<br />
Additions to the original code include all full lines preceded with<br />
'ccc' and one 'ccc' at the start of the line following '0 L D COD E.'<br />
C test for wet removal processes<br />
IF(DIRT(KT)%DOWET.AND.RAIN.GT.O.O)THEN<br />
C determine bottom and top of the precip layer (80% to 60%)<br />
CCC *******************************************************<br />
CCC with modification: (BASE AT 75% OVER LAND, 80% OVER SEA)<br />
CCC *******************************************************<br />
KBOT=O<br />
KTOP=NLVL<br />
DO K=1,NLVL<br />
KRH=QQ(K)*100.0+0.5<br />
CCC *******************************************************<br />
CCC CLOUD BASE MODIFICATION PROPOSED BY aaron@gimail.af.mil<br />
CCC SEE: http://nimbo.wrh.noaa.gov/wrhq/96TAs/TA9629/ta96-29.html<br />
CCC WESTERN REGION TECHNICAL ATTACHMENT<br />
CCC NO. 96-29<br />
CCC NOVEMBER 19, 1996<br />
94
CCC THE EXPLICIT CLOUD PREDICTION SCHEME IN<br />
CCC THE MESO ETA MODEL<br />
CCC Mike Staudenmaier, Jr. - WRH-SSD/NWSFO SLC<br />
CCC * * * * * * 0 L D COD E * * * * * *<br />
CCC IF(KBOT.EQ.O.AND.KRH.GE.80)KBOT=K<br />
CCC * * * * * * NEW COD E * * * * * *<br />
IF(LAND.EQ. 7.AND.KBOT .EQ.O.AND.KRH.GE.80)KBOT=K<br />
IF(LAND.NE. 7.AND.KBOT.EQ.O.AND.KRH.GE. 75)KBOT=K<br />
CCC *******************************************************<br />
IF(KBOT.NE.O.AND.KTOP.EQ.NLVL.AND.KRH.LE.60)KTOP=K<br />
END DO<br />
95
Appendix G. Investigation of Greece Exclusion from Modeled April<br />
Cs-131 Deposition<br />
Since several cumulative Cs-137 concentration measurements exceeding 10 5 Bq/m2<br />
were recorded in Greece on 19S6MayOl, the conspicuous omission of Greece from<br />
April deposition patterns in this thesis bears investigation. No other simultaneous<br />
sources of atmospheric Cs-137 are documented, and Cs-137 arrives in Greece in<br />
the model early in May, so it is safe to assume that the measured deposition came<br />
from <strong>Chernobyl</strong>. This investigation considers three other possible explanations for<br />
the Greece exclusion from the cumulative control run. The first possibility is an<br />
over-estimated deposition velocity, i.e. the modeled dry fall speed is too fast. The<br />
distribution of particle sizes is not known nor is the change of the distribution in<br />
space and time, yet the modeled constant deposition velocity implies a uniform modeled<br />
size distribution. It's conceivable that actual particles of smaller aerodynamic<br />
diameter could have been carried further than the modeled plume before depositing.<br />
Another possible reason the model failed to show transport to Greece is trajectory<br />
looping of the actual plume, i.e. pollutant could leave the model domain and return<br />
(HySPLIT ignores particles that leave the domain of the meteorological model).<br />
Figure 17 suggests this is a strong possibility since the air parcels 'disappear' from<br />
the simulation (i.e, HySPLIT omits particles from any further calculations) when<br />
they cross the meteorological grid boundary at about 40.5 degrees longitude. The<br />
third possible reason for a modeled Cs-137-free Greece is a combination of modeled<br />
wind direction error and modeled wind speed error. Direction errors (especially<br />
near the source) or speed errors (especially in the vertical) could prevent the model<br />
from transporting pollutant to Greece by OZ, May 1. Since no compelling evidence<br />
for control run adjustments has come to light, settings for the simulation should<br />
not be tuned solely to provide for April deposition in Greece, and the data is ignored.<br />
Modeled emission release heights of up to 2990m and 9000m in diagnostic<br />
96
simulations yielded deposition nearer to Greece, but not in Greece. The first (wet)<br />
part checks different source term profiles in the vertical to isolate control run release<br />
heights that could result in Cs-137 deposited in Greece; none do. The second (dry)<br />
part employs the ultimate source term profile in the vertical to isolate control run<br />
plume layers that could result in Cs-137 deposited in Greece; none do. Reference<br />
Appendix E for a map of Europe with political boundaries and designations.<br />
G.l Evaluation of Source Term Height - Wei<br />
Since the source term height is uncertain, it is prudent to see if an error in the<br />
modeled source term height could be the cause of the Grecian deposition omission.<br />
Three diagnostic simulations are presented. Figure 68 is the result of a run set up<br />
the same as the 5-day cumulative control run, but with the emission in a uniform<br />
vertical line from 10m to 1750m. If the reason for Greece's omission was a control<br />
run plume base estimate that was too high, then a plume base low enough would<br />
lead to model deposition in Greece. Figure 69 checks the effects of centering, for the<br />
run's duration, the uniform vertical line source at 1500m, the recommended center<br />
of mass for the first six hours of emission (Klug 92:358). Figure 70 checks the effects<br />
of a source term extending to 9000m. None of these diagnostic simulations deposit<br />
pollutant in Greece.<br />
The contents of the HySPLIT Control file for the simulation in Figure 68 are<br />
recorded in Section D.6. Control file settings for the other two diagnostic simulations<br />
simply reflect the revised top of emission in line 4, and output grid filename in line<br />
21. For a summary of all primary settings in this thesis see Appendix D. For full<br />
details on settings see the HySPLIT User's Guide (Draxler 99).<br />
G.2 Evaluation of Source Term Height - Dry<br />
The exclusion of the Greece region from the model's control run output deposition<br />
fields is not a result of the model over-scavenging the plume by precipitation<br />
97
NOAA AIR RESOURCES LABORATORY<br />
Deposition from 21 z 25 Apr to OOz 01 May (UTe)<br />
OOZ 01 May 86 ECMF INITIAL DATA<br />
.1.0E+05<br />
1.6E+0Cl MAXIMUM AT SQUARE<br />
1.0E-01<br />
Figure 68<br />
Greece Diagnostic Run, Cumulative April Deposition [Bq/m2], Release<br />
Height Profile from 10m to 1750m<br />
98
NOAA AIR RESOURCES LABORATORY<br />
Deposition from 21 z 25 Apr to ODz 01 May (UTe)<br />
aoz 01 May 86 ECMF INITIAL DATA<br />
• 1.0E+05 • 1.0E+03 1.0E+01 1.0E-01<br />
1.5Et05 MAXIMUM AT SQUARE<br />
Figure 69<br />
Greece Diagnostic Run, Cumulative April Deposition [Bq/m2], Release<br />
Height Profile from 10m to 2990m<br />
99
NOAA AIR RESOURCES LABORATORY<br />
Deposition from 21 z 25 Apr to OOz 01 May (UTe)<br />
ooZ 01 May 86 ECMF INITIAL DATA<br />
\' ~.<br />
( 1M2)<br />
• 1.0E+03 • 1 .OE+01 1.0E·01 1.0E-03<br />
6.0E ... 03 MAXIMUM AT SOUARE<br />
Figure 70<br />
Greece Diagnostic Run, Cumulative April Deposition [Bq/m2], Release<br />
Height Profile from 10m to 9000m<br />
100
washout upstream. Dry model runs (i.e., with precipitation fields absent from the<br />
meteorological input file) indicate that the modeled pollutant plume (even with a<br />
vertical line source from 10 to 14000m) did not reach Greece. Figures 71 through<br />
76 display slices of the total 5-day-averaged dry plume at several vertical levels.<br />
The output grids of this simulation imply that no particle emitted from <strong>Chernobyl</strong><br />
at any height between 10m and 14000m at any time between 21Z1986Apr25 and<br />
24Z1986Apr30 floated over Greece. This result is consistent with the null Greece<br />
deposition of the 'wet' runs, because the meteorological model contains significant<br />
precipitation Greece-wide, especially late in the simulation (see precipitation fields<br />
in Appendix C).<br />
The dry runs were identical to the moist with these exceptions:<br />
• The input meteorological model lacked precipitation fields (identical otherwise).<br />
• All runs used the same source term layer (10m to 14000m).<br />
• The transport model top (ceiling) was set at 20000m instead of at 10000m (the<br />
default).<br />
• Each run recorded an average concentration field at a different level (one at<br />
Om, i.e., deposition, as in the wet runs).<br />
• 2000 particles were tracked in each simulation instead of the default 500 (set<br />
in file 'setup.cfg').<br />
101
NOAA AIR RESOURCES LA130RATOIW<br />
Depos itio n from 21 z 25 Apr to 21 z 30 Apr (UTe)<br />
"1 Z 30 Ai>' 86 ECMF 1~ITIAl DATA<br />
40 (")<br />
'"<br />
:0<br />
!l!-<br />
iD<br />
m<br />
iD<br />
fn<br />
iii<br />
4<br />
'iD<br />
Q.<br />
W.<br />
~<br />
N<br />
~<br />
»<br />
~<br />
S<br />
g<br />
~ 5E-Qt \IAXtMUM AT 5QtJARE<br />
Figure 71<br />
Five-day accumulated deposition due to Phase I emissions modeled as<br />
vertical line source from 10m to 14000m. Precipitation turned off.<br />
NOAA AIR RESOURCES LABORATORY<br />
Average from 21z 25 Apr to 21z 30 Apr (UTe)<br />
21Z 30 Apt" 86 ECMF INITIAL DATA<br />
Figure 72<br />
Five-day average concentration at 1000m due to Phase I emissions modeled<br />
as vertical line source from 10m to 14000m. Precipitation turned<br />
off.<br />
102
---------------------------------------------------<br />
NOM AIR RESOURCES LABORATORY<br />
Average from 21z 25 Apr to 21z 30 Apr (UTC)<br />
21 Z 30 Ap' 86 ECMF 1~ITIAl DATA<br />
u..J<br />
o<br />
g<br />
(')<br />
'"<br />
• 1.0E+OO ,. 1_0E-02 c- 1.0E-04 1.0E-06<br />
1.SE-ao \lAXII.r.J~' f
NOAA AIR RESOURCES LABORATORY<br />
Average from 21z 25 Apr to 21z 30 Apr (UTC)<br />
21 Z 30 AV 65 EC MF IN ITIAl DATA<br />
'. 7E-ao \IAl
Bibliography<br />
ARL OOa. ARL. The <strong>Chernobyl</strong> Reactor <strong>Accident</strong>.<br />
http) /www.arl.noaa.gov/ss/transport/chernobyl.html, October<br />
2000. Air Resources Laboratory, Transport Modeling and Assessment<br />
- Silver Spring, MD.<br />
ARL OOb. ARL. ETEX Simulations Using HYSPLIT4.<br />
http)/www.arl.noaa.gov/ss/transport/etex.html, October 2000.<br />
Air Resources Laboratory.<br />
DeCort 90.<br />
DeCort 98.<br />
M. DeCort, G. Graziani, F. Raes, D. Stanners, G. Grippa & I. Ricapito.<br />
Radioactivity Measurements in Europe after the <strong>Chernobyl</strong> <strong>Accident</strong>,<br />
Part II, Fallout and Deposition. Rapport technique EUR 12S00 EN,<br />
Commission of the European Communities Joint Research Centre -<br />
Ispra Site, 1990.<br />
M. DeCort, G. Dubois, Sh D. Fridman & M. G. Germenchuk et al.<br />
Atlas of Caesium Deposition on Europe After the <strong>Chernobyl</strong> <strong>Accident</strong>.<br />
Luxembourg, Office for the Official Publications of the European Communities,<br />
1998. 176 pp.<br />
Draxler 9Sa. Roland R. Draxler & G. D. Hess. Description of the HYSPLIT4 modeling<br />
system. Rapport technique NOAA Tech. Memo. ERL ARL-224,<br />
National Oceanographic and Atmospheric Administration, December<br />
1998.<br />
Draxler 98b. Roland R. Draxler & G.D. Hess. An overview of the HYSPLIT4 modelling<br />
system for trajectories, dispersion and deposition. Aust. Met.<br />
Mag., vol. 47, pages 295-308, 1998.<br />
Draxler 99.<br />
Roland R. Draxler. Hybrid single-particle Lagrangian integrated trajectories<br />
(HY-SPLIT): Version 3.0 - User's guide and model description.<br />
Rapport technique NOAA Tech. Memo. ERL ARL-195, National<br />
Oceanographic and Atmospheric Administration, 5285 Port Royal Rd,<br />
Springfield, VA 22161, 1999. Available from National Technical Information<br />
Service, 5285 Port Royal Road, Springfield, VA 2216l.<br />
Draxler OOa. Roland Draxler, 2000. Electronic Message, 19Sep2000, ARL, Silver<br />
Spring, Maryland.<br />
Draxler OOb. Roland Draxler, 2000. Personal Communication, 23March2000, Air<br />
Force Technical Applications Center, Patrick AFB, Florida.<br />
Glasstone 77. Samuel Glasstone & Philip J. Dolan. The Effects of <strong>Nuclear</strong> Weapons,<br />
3rd edition. United States <strong>Department</strong> of <strong>Defense</strong> and United States<br />
<strong>Department</strong> of Energy, 1977. 653 pp.<br />
105
Graziani 97. Giovanni Graziani. European Tracer Experiment.<br />
http://www.ei.jrc.it/etex/. August 1997.<br />
Graziani 00. Giovanni Graziani, 2000. Electronic Message, 18sep2000, JRC, Ispra,<br />
Italy.<br />
Hagans 00.<br />
Hicks 86.<br />
IGES 01.<br />
Klug 92.<br />
Knap 88.<br />
Fred Hagans. http://www.aftac.gov/mission.htm. 2000.<br />
Bruce B. Hicks. Differences in wet and dry particle deposition parameters<br />
between North America and Europe. Aerosols: Research,<br />
Risk Assessment, and Control Strategies, pages 973-982, 1986. Atmospheric<br />
Turbulence and Diffusion Division, Air Resources Laboratory,<br />
ERL/NOAA, Oak Ridge, TN - Lewis Publishers, Chelsea, MI.<br />
IGES. GRid Analysis and<br />
http://grads.iges.org/grads/, January 2001.<br />
Environment and Society.<br />
Display System.<br />
Institute of Global<br />
W. Klug, G. Graziani, G. Grippa, G. D. Pierce & C. Tassone. ATMES<br />
Report, Evaluation of long range atmospheric transport models using<br />
environmental radioactivity data from the <strong>Chernobyl</strong> accident. Elsevier<br />
Science Publishers, England, 1992. 366 pp.<br />
Anthony H. Knap. The Long-Range Atmospheric Transport of Natural<br />
and Contaminant Substances. Kluwer Academic Publishers, 1988.<br />
NATO Advanced Science Institutes Series C: Mathematical and Physical<br />
Sciences - Vol. 297, 321 pp.<br />
Lauritzen 99. B. Lauritzen & T. Mikkelson. A probabilistic dispersion model applied<br />
to the long range transport of radionucleides from the <strong>Chernobyl</strong> accident.<br />
Atm. Environ., vol. 33, pages 3271-3279, 1999. Riso National<br />
Laboratory, 4000 Roskilde, Denmark.<br />
Metivier 95. Henri Metivier. <strong>Chernobyl</strong> Ten Years On, Radiological and Health<br />
Impact. http://www.nea.fr/htmljrp/chernobyljallchernobyl.html,<br />
November 1995. OECD <strong>Nuclear</strong> Energy Agency.<br />
MSE 00.<br />
NWS 01.<br />
Persson 87.<br />
Curie (measurement) Microsoft Encarta Online Encyclopedia 2000.<br />
http) /encarta.msn.com 1997-2000 Microsoft Corporation. All Rights<br />
Reserved, 2000.<br />
The Medium-Range Forecast Model status as of January 09, 2001.<br />
http://sgi62.wwb.noaa.gov:8080/research/mrf.html, 2001. National<br />
Weather Service.<br />
C. Persson, H. Rodhe & L. E. De Geer. The <strong>Chernobyl</strong> accident - a<br />
meteorological analysis of how radionuclides reached and were deposited<br />
in Sweden. Ambio, vol. 16, pages 20-31, 1987.<br />
106
P5llanen 97. Roy P5llanen, Ilkka Valkama & Harri Toivonen. Transport of radioactive<br />
particles from the <strong>Chernobyl</strong> accident. Atm. Environ., vol. 31, pages<br />
3575-3590, 1997. Finnish Centre for Radiation and <strong>Nuclear</strong> Safety<br />
(STUK), P.O. Box 14, 00881 Helsinki, Finland; and Finnish Meteorological<br />
Institute (FMI), P.O. Box 503, 00101 Helsinki, Finland.<br />
Rodriguez 95. D. Rodriguez, H. Walker, N. Klepikova, A. Kostrikov & Y. Zhuk.<br />
Evaluation of two pollutant disperson models over continental scales.<br />
Atm. Environ., vol. 29, no. 7, pages 799-812, 1995. University of California,<br />
Livermore, CA 94550, U.S.A.; and Institute of Experimental<br />
Meteorology, Obninsk, Kaluga Region 249020, Russia.<br />
Seinfeld 86. J. H. Seinfeld. Atmospheric Chemistry and Physics of Air Pollution.<br />
John Wiley and Sons, New York, 1986. 324 pp.<br />
Serway 92.<br />
Slingo 87.<br />
Raymond A. Serway. Physics for Scientists and Engineers with Modern<br />
Physics, 3rd edition. Saunders College Publishing, 1992. 1444 pp.<br />
J. M. Slingo. The development and verification of a cloud prediction<br />
scheme for the ECMWF model. Quart. J. Roy. Meteor. Soc., vol. 113,<br />
pages 899-927, 1987. European Centre for Medium Range Weather<br />
Forecasts, Reading, England.<br />
Staudenmaier 96. Mike Staudenmaier Jr. The explicit cloud prediction scheme in<br />
the Meso ETA model. Rapport technique Western Region Technical<br />
Attachment 96-29, National Weather Service Western Region Headquarters,<br />
Salt Lake City, 1996.<br />
Tiedke 93.<br />
UIC 00.<br />
M. Tiedke. Representation of Clouds in Large-Scale Models. Mon. Wea.<br />
Rev., vol. 121, pages 3040-3061, 1993. European Centre for Medium<br />
Range Weather Forecasts, Reading, Berkshire, England.<br />
UIC. Radiation and Life. http://www.uic.com.au/ral.htm. 2000. URA<br />
NIUM INFORMATION CENTRE, Ltd., A.C.N. 005 503 828.<br />
107
Vita<br />
Captain Aaron M. Kinser was born in Kettering, Ohio in 1966, and was raised<br />
in nearby Centerville. Centerville High School granted him a diploma in May,<br />
1984. In 1989 he enlisted in the US Air Force as a computer operator and served<br />
one tour assigned to the 354th Communications Squadron, Eielson AFB, Alaska.<br />
While assigned there he was selected for the Airmen Education and Commissioning<br />
Program (AECP) administered by the Air Force Institute of Technology (AFIT),<br />
Wright Patterson AFB, Ohio. Per AECP, Captain Kinser earned a Bachelors Degree<br />
in Atmospheric Sciences in 1996 from the University of Arizona in Tucson, Arizona<br />
and graduated from Officer Training School at Maxwell AFB, Alabama on December<br />
6, 1996. He then served as wing weather officer for the 57th Operational Support<br />
Squadron at Nellis AFB, Nevada until he was selected to earn a Masters Degree in<br />
Meteorology at AFIT starting in the Fall of 1999. Following graduation, Captain<br />
Kinser will be assigned to the Air Force Technical Applications Center (AFTAC),<br />
Patrick AFB, Florida. Captain Kinser can be contacted for the foreseeable future<br />
by text-only email ataaron@gimail.af.mil.<br />
108
REPORT DOCUMENTATION PAGE<br />
Form Approved<br />
OMB No. 0704-0188<br />
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering an<br />
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,<br />
including suggestions for reducing this burden to Washington Headquarters Services, Directorate for Information Operations and <strong>Reports</strong>, 1215 Jefferson Davis Highway. Suite 1204, Arlington,<br />
VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project 10704-01 BB1, Washington, DC 20503.<br />
PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.<br />
1. REPORT DATE (OO-MM-YYYY) 12. REPORT TYPE 3. DATES COVERED<br />
03-2001 Master's Thesis Aug 2000 - Mar 2001<br />
4. TITLE AND SUBTITLE Sa. CONTRACT NUMBER<br />
SIMULATING WET DEPOSITION OF RADIOCESIUM FROM THE<br />
CHERNOBYLACCIDENT<br />
5b. GRANT NUMBER<br />
5c. PROGRAM ELEMENT NUMBER<br />
6. AUTHOR(S) 5d. PROJECT NUMBER<br />
Kinser, Aaron, M., Captain, USAF<br />
5e. TASK NUMBER<br />
Sf. WORK UNIT NUMBER<br />
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) S. PERFORMING ORGANIZATION<br />
REPORT NUMBER<br />
Air Force Institute of Technology (AFIT/EN)<br />
Bldg 640 Rm 100<br />
2950 PSt<br />
Wright Patterson AFB, OH 45433-7765<br />
AFIT/GMIENP/01M-05<br />
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR'S ACRONYM(S)<br />
Air Force Technical Applications Center<br />
AFTAC/TMAR<br />
Attn: Mr. Craig Sloan<br />
11. SPONSOR/MONITOR'S REPORT<br />
1030 S. Highway AlA<br />
NUMBER(S)<br />
Patrick AFB, FL 32925-3002 DSN: 854-7781<br />
12. DISTRIBUTION/AILABILITY STATEMENT<br />
APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED.<br />
13. SUPPLEMENTARY NOTES<br />
14. ABSTRACT<br />
In response to the <strong>Chernobyl</strong> nuclearle0wer plant accident of 1986, a cesium-137 de!i0sition dataset was assembled. Most of the<br />
airborne <strong>Chernobyl</strong> cesium was wet eEosited, either via interception blr falling rain rops or via abs0r.ption into cloud dn~tets<br />
destined to become raindrops. The Hy rid Single-Particle Lagrangian ntCfated Transport (HYSPLI ) model, develope at Air<br />
Resources Laborat02" is used to simulate the transport and deposition of ernobrl cesium-137. A cloud base barameterization<br />
modification is teste and ap~ears to slightly imfarove the accuracy of one HYSPL T simulation of dail Cherno (I cesium-137<br />
deposition over the course 0 the accident at iso ated European sites, and degrades the accuracy of anot h er HYSP IT simulation of<br />
deposition in Germany and Austria accumulated in the month of April 1986. Large uncertaintIes in the emission specifications,<br />
model precipitation fields, and deJ'0sition measurements prevent designating the results as conclusive, but most eVidence points to<br />
improved performance within 50 kilometers of the emission source. Trial and error lessons learned from hundreds of preliminary<br />
model runs are documented, and the exact HYSPLIT settings of successful and meaningful simulations are appended.<br />
15. SUBJECT TERMS<br />
Hybrid Single-Particle Lagrangian Integrated Transport (HYSPLIT) model, atmospheric transport modeling, wet deposition,<br />
cesium-I37 deposition, cloud base parameterization, <strong>Chernobyl</strong><br />
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF<br />
a. REPORT b. ABSTRACT c. THIS PAGE ABSTRACT<br />
U U U UU<br />
18. NUMBEF<br />
OF<br />
PAGES<br />
123<br />
19a. NAME OF RESPONSIBLE PERSON<br />
Lt Col Michael K. Walters, AFITIENP<br />
19b. TELEPHONE NUMBER (/nc/ude area code)<br />
(937)237-7166<br />
Standard Form 298 (Rev. 8/98)<br />
Prescribed by ANSI Std. Z39.18
Radiation Injuries after the <strong>Chernobyl</strong><br />
<strong>Accident</strong>: Management, Outcome, and<br />
Lessons Learned<br />
Scientific Medical Effects of Ionizing Radiation<br />
(MEIR) Course<br />
July 30, 2008<br />
Alla Shapiro, M.D., PhD<br />
Medical Officer,<br />
US Food and Drug Administration<br />
Center for Drug Evaluation and Research<br />
Office of Counter-Terrorism and<br />
Emergency Coordination
Presentation Outline<br />
I. What could have been done at <strong>Chernobyl</strong> to lessen<br />
the effects of radiation damage?<br />
II. Health consequences<br />
• Thyroid gland and radiation exposure<br />
• Acute Radiation Syndrome (ARS) and its outcome<br />
• Neuro-psychological impact<br />
• <strong>Chernobyl</strong> experience of Cutaneous Radiation Syndrome<br />
(CRS)<br />
• Other medical problems caused by the accident<br />
III. Conclusions
What if….
Basic Information on the Radionuclide Releases<br />
and the Types of Exposure at <strong>Chernobyl</strong><br />
• 100% of gaseous fraction of the noble gases and nuclides<br />
may have escaped from the plant<br />
• Cesium, Iodine and Tellurium isotopes accounted for up to<br />
10-20% of the nuclides inventory<br />
• Transuranic elements (Plutonium, Curium and Americium)<br />
were found only in the lungs<br />
• Neutron irradiation was not significant<br />
• ARS was caused by - and gamma-irradiation of the whole<br />
body and by beta-irradiation of the skin surface<br />
Ref: International Atomic Agency. Summary Report on the Post-<strong>Accident</strong> Review<br />
Meeting on the <strong>Chernobyl</strong> <strong>Accident</strong>, Vienna, 1986.
Sheltering<br />
• Sheltering is an effective preventive action in the<br />
area within a radius of 3-10 km from the point of<br />
the accident even in the case of absence of<br />
confirming radiation measurements<br />
• At Pripyat information about the need for<br />
sheltering was delayed<br />
• For other populations including Kyiv<br />
recommendations on sheltering were distributed<br />
on May 10, 1986 after the spread of radioiodine<br />
• The efficiency of sheltering could not be<br />
assessed
Relocation from the site<br />
The preliminary decision to evacuate the town of Pripyat,<br />
which is located less than 3 km from the ChNPP, was<br />
taken on the afternoon of 26 April 1986, when the dose<br />
rate in some parts of the town reached several mSv/hour<br />
By 9 pm on 26 April 1986, 1,350 buses, 2 railway trains<br />
and 3 motor ships were brought into the <strong>Chernobyl</strong><br />
district (12 km from the town of Pripyat)<br />
At 10 pm the USSR Ministry of Public Health decided that<br />
the emergency evacuation of the town was necessary
Relocation from the site<br />
(continued)<br />
The organized evacuation of the town of Pripyat<br />
(49,360 including about 17,000 children and 80 bedbound<br />
patients), was carried out on 27 April 1986,<br />
between 2 pm and 5 pm
Iodine Prophylaxis<br />
Official information from 1986<br />
Total: It was administered to about 5 million people, including<br />
1.6 million children<br />
Pripyat town: It was administered to about 70% of the total<br />
population, including 60% on April 26<br />
Kiev Oblast: <strong>Department</strong> of the Ministry of Health made a<br />
decision on iodine prophylaxis on May 6, 10 days after the<br />
accident<br />
The Russian Federation: It was administered to 71,930 people,<br />
including 25,060 children, from June to the middle of August<br />
1986
Number of thyroid cancer cases in children and adolescents of Ukraine<br />
(aged 0-180<br />
yrs) at the time of the <strong>Chernobyl</strong> accident<br />
400<br />
350<br />
300<br />
250<br />
200<br />
150<br />
100<br />
50<br />
0<br />
19<br />
11<br />
8<br />
14<br />
11<br />
15-18<br />
0-14<br />
25 22<br />
118<br />
35 46<br />
62 69 20 23<br />
22 37 83 83<br />
42 46<br />
13<br />
11<br />
11<br />
24<br />
147<br />
118 129 43<br />
104<br />
251<br />
67<br />
192<br />
183<br />
81<br />
197 66<br />
117<br />
86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04<br />
54<br />
138<br />
61<br />
136<br />
170<br />
284<br />
217<br />
249<br />
69<br />
180<br />
359<br />
85<br />
274<br />
350<br />
331<br />
88<br />
243<br />
105<br />
245<br />
370<br />
92<br />
278<br />
1
Distribution of thyroid cancer cases depending on<br />
patients’ age at the time of the accident<br />
50%<br />
45%<br />
40%<br />
35%<br />
30%<br />
1986 - 1989<br />
1990 - 1995<br />
1996 - 2001<br />
2002 - 2004<br />
25%<br />
20%<br />
15%<br />
10%<br />
5%<br />
0%<br />
0 - 4 5 - 9 10 - 14 15 - 18<br />
4
Papillary Carcinoma<br />
11
Chronic Thyroiditis<br />
350<br />
Controlled territories residents<br />
300<br />
Evacuees<br />
Recovery operation workers<br />
274.1<br />
306.1<br />
250<br />
225<br />
200<br />
150<br />
100<br />
50<br />
0<br />
182.1<br />
169<br />
147<br />
138.4<br />
126.2<br />
108.3<br />
115<br />
96<br />
86.4 82.8 84.7<br />
75<br />
57.7<br />
34.2<br />
9.7 13.5 16.9 18<br />
19<br />
21<br />
25<br />
6.5<br />
187.4<br />
27.3<br />
1992 1993 1994 1995 1996 1997 1998 1999 2000<br />
Data of Ministry of Health of Ukraine
Leukemia<br />
• Consensus exists on the absence of leukemia<br />
excess in inhabitants of the contaminated<br />
territories (French-German Initiative study)<br />
• There is a controversy in data on the leukemia<br />
incidence in children exposed in utero<br />
• Preliminary data from 2003 - 2005 demonstrate<br />
dose-effect relationship in operation recovery<br />
workers irradiated over 100 mGy (US Natl. Cancer<br />
Institute-RCRM joint study of leukemia among clean-up<br />
workers of <strong>Chernobyl</strong> in Ukraine)
Mutation AML1 in ARS Survivors with<br />
Myelodisplastic Syndrome<br />
During L1 gene sequenation in ARS<br />
survivor, who had suffered MDS, it was<br />
revealed punctuated mutation as<br />
repeating of 6 nucleotides<br />
Patient No 24 / AML1 wt<br />
Patient No 263 / appel with ins<br />
1502 (CGGCAT)<br />
T S G I G I G I G M S A M<br />
consecution with mutation
Conclusions<br />
Studies in Ukraine have shown:<br />
• an excess of thyroid cancer and non-cancer<br />
thyroid disease in children & other exposed<br />
groups (recovery operation workers,evacuees,<br />
adult population)<br />
• a controversy in data on the leukemia<br />
incidence in children exposed in utero<br />
• a dose-effect effect relationship on the leukemia<br />
incidence in recovery operation workers<br />
exposed to over 100 mSv
Conclusions (continued)<br />
• an increase in the breast cancer<br />
incidence rate in females participating<br />
in recovery operation works in 1986/87<br />
and female subpopulation still living in<br />
the most contaminated areas<br />
• an increase in all forms of cancer<br />
incidence rate only among recovery<br />
operation workers of 1986-1987<br />
1987 in<br />
comparison with national level
Liquidators in Action
<strong>Chernobyl</strong>, April 26, 1986: Sequence of the<br />
Initial Intervention<br />
Time Intervention Treatment/Disposition<br />
30 min –<br />
3-4 hours<br />
Initial treatment on the site<br />
• evacuation from the site,<br />
antiemetics, sedative,<br />
cardiotonic<br />
4 hours -<br />
12 hours<br />
12 hours -<br />
36 hours<br />
Evaluation and treatment at<br />
the nuclear plant medical<br />
facility<br />
Specialized team arrived<br />
• discharged if condition is OK<br />
• remained hospitalized<br />
• Assessment, blood tests,<br />
administration of KI, priority for<br />
hospitalization established<br />
Guskova et al, 1986. Acute radiation effects in exposed persons at the <strong>Chernobyl</strong><br />
Atomic Power Station <strong>Accident</strong>. Medical Radiology, (477), pp.3-18
Sequence of the Initial Intervention (continued)<br />
• Within the first three days, 299 persons were sent to the<br />
specialized treatment center in Moscow and to hospitals in<br />
Kiev<br />
• Over the subsequent days hundreds of additional persons<br />
were admitted for examination<br />
• Criteria for hospitalization included for patients with the<br />
suspected ARS<br />
» Presence, time of onset and intensity of nausea and vomiting<br />
» Primary erythema of the skin<br />
» Decrease of the lymphocyte count in the peripheral blood<br />
Primary Diagnostic Criteria of ARS:<br />
Diagnostic Coefficient (DC)<br />
Assessment of irreversible myelosuppression according to DC in cases of ARS<br />
Time to the onset of vomiting Hours<br />
0-0.4<br />
0.41-0.8<br />
0.81- 1.2<br />
1.21 – 1.6<br />
>2.0<br />
Diagnostic Score<br />
+8<br />
+4<br />
+2<br />
-2<br />
-10<br />
Lymphocyte count 10 9 x1-1 Diagnostic Score<br />
Lymphocyte count on Day 2 0-0.2<br />
+6<br />
Lymphocyte count on Day 2 0.61- 0.8<br />
-15<br />
Lymphocyte count Days 4 – 7 0.01<br />
+5<br />
Lymphocyte count Days 4 – 7 >0.15<br />
-15<br />
A sum of +10 is the basis for the prognosis of irreversible myelosuppression; a<br />
sum of -10 is a prognosis for NO irreversible myelosuppression.
The Severity and Outcome of ARS in<br />
<strong>Chernobyl</strong> Victims<br />
ARS<br />
Dose<br />
Number of Patients<br />
Grade (Gy) Total Alive Died (days to death)<br />
I 0.8 -2.1 31 31 0<br />
II 2.0 - 4.0 43 42 1 96<br />
III 4.2 - 6.3 21 14 7 16 - 48<br />
IV 6.0 –16.0 20 1 19 14 - 91<br />
TOTAL 115 88 27<br />
Baranov et al, Antibiotics and Chemotherapy, 1989, 34, 7, 555-558; Guskova et al,<br />
“Acute Radiation Effects in Exposed Persons at the <strong>Chernobyl</strong> Atomic Power<br />
Station <strong>Accident</strong>” Medical Radiology, 1986, pp. 3-18.
The Bone Marrow Syndrome and its Treatment<br />
in <strong>Chernobyl</strong> Victims (1)<br />
• Antiseptic regimen<br />
• Supportive therapy<br />
» Isolation<br />
» Air sterilization<br />
» Changes of underclothing for patients at least once/day<br />
» Maintaining the micro-organism population at less than 500/mm3<br />
in the room air<br />
» Antimicrobial decontamination of the intestine<br />
» Administration of systemic antibiotics<br />
» Acyclovir<br />
» Transfusions of blood cells (e.g. fresh donor platelets and RBC)<br />
Ref: UNSCEAR Report, 1986; Robert Gale et al, The Lancet, April 23,1988. Guskova A. et al, 1986.
Supportive Therapy for Neutropenia<br />
• Oral quinolones, fluconazole, acyclovir<br />
prophylactically<br />
• Standard care for hemopoietic failure<br />
• All blood products irradiated at 25 Gy<br />
• Other supportive measures ad libitum
The New Concept of the ARS<br />
6-8 Gy<br />
( MODS)<br />
( MOF)<br />
1 Gy<br />
4 Gy<br />
SUBCLINICAL<br />
BONE MARROW<br />
(SOF)<br />
Reversible if<br />
heterogenous irradiation<br />
30 Gy<br />
50 Gy<br />
INCREASING DOSE
Bone Marrow Syndrome and its Treatment in<br />
<strong>Chernobyl</strong> Victims (2)<br />
• HLA-matched unrelated bone marrow donors from<br />
large HLA-typed volunteer donor pools – 13 patients<br />
• Fetal liver cells – 6 patients<br />
• Bone marrow syndrome combined with other Injuries<br />
• Skin<br />
• GI<br />
• Oropharyngeal<br />
• Radiation pneumonitis
Number of Deaths from Direct Radiation<br />
Effects in first 3 months<br />
Number of patients<br />
died (TOTAL = 27)<br />
Days of death after<br />
the exposure<br />
22 14 - 34<br />
Comments<br />
In 20/22 patients -burns<br />
were the main cause of<br />
death<br />
5 48 – 99*<br />
Died after the bone marrow<br />
recovery stage<br />
* Patient on Day #96 died<br />
from ischemic stroke<br />
Baranov et al, Antibiotics and Chemotherapy, 1989, 34, 7, 555-558; Guskova et al,<br />
“Acute Radiation Effects in Exposed Persons at the <strong>Chernobyl</strong> Atomic Power<br />
Station <strong>Accident</strong>” Medical Radiology, 1986, pp. 3-18.
Indications for an Allogenic BMT or an<br />
Embryonic Live Cell Transplantation<br />
• Whole body<br />
-irradiation dose 6.0 Gy -16.0 Gy<br />
• Irreversible degree of myelosuppression using a<br />
Diagnostic Coefficient (DC)<br />
plus additional criteria<br />
• Vomiting during the first 30 minutes<br />
• Diarrhea during 1-2 hours after the exposure<br />
• Swelling of the parotid glands during the first 24-36 hours<br />
Ref: UNSCEAR 1988 Report
Outcome (Survival or Cause of Death)<br />
in Patients Receiving BMT<br />
Dose<br />
range (Gy)<br />
Bone marrow transplant patients<br />
Number of<br />
patients<br />
Deaths* Deaths** Number of<br />
survivors<br />
< 6.5 4 0 3 1<br />
6.5 – 9.0 3 2 1 1<br />
> 9.0 6 5 0 0<br />
TOTAL 13 7<br />
*skin and GI<br />
injuries<br />
4<br />
**GVHD +<br />
infection<br />
2<br />
Gale et al, 1988; UNSCEAR report, 1988; Baranov et al, 1989; Guskova et al, 1989;
Hemopoetic Stem Cell Transplants ???<br />
• Never an emergency!<br />
• Not if MODS!<br />
• Always consider heterogeneity of irradiation and<br />
possibility of autologous hemopoietic recovery<br />
• HLA typing immediate<br />
• Transplant never before day 14-21<br />
• Low immunosuppression: fludarabine ± ATG<br />
• High cell dose 2x10 6 CD34/kg (peripheral blood),<br />
2x10 8 nucleated cells/kg (bone marrow) and 3x10 7<br />
nucleated cells (cord blood)
Problems that Complicated the Use of<br />
BMT for <strong>Chernobyl</strong> Victims<br />
• Determination of the radiation dose<br />
• Several kinds of irradiation (external -and , and<br />
inhaled and ingested isotopes)<br />
• Partial shielding of body parts by physical structures<br />
• Rapid onset of lymphocytopenia made HLA typing<br />
difficult. Donor-recipient histocompatibility was not<br />
accurately determined<br />
• Most individuals who received a sufficiently high dose<br />
of irradiation had thermal burns as well as injuries to<br />
the GI tract and other tissues
Causes of Death among ARS Survivors<br />
(1986 through 2006)<br />
Cause of death<br />
Oncological and<br />
oncohematological pathology<br />
Grade I<br />
ARS<br />
1<br />
Grade II Grade IIII Total<br />
ARS ARS<br />
2 2 5<br />
Sudden cardiac death 2 2 2 6<br />
Internal organ systems and<br />
neurological diseases<br />
1<br />
3 1 5<br />
Traumas and accidents 2 - - 2<br />
TOTAL 6 7 5 18
Oncological Diseases in ARS Survivors and<br />
non-ARS Patients<br />
No Group Diagnosis First revealed<br />
Outcome<br />
1 non-ARS Sarcoma of hip soft tissues 1992 Died in 1993<br />
2<br />
3<br />
4<br />
5<br />
6<br />
7<br />
8<br />
9<br />
10<br />
11<br />
12<br />
13<br />
non-ARS<br />
non-ARS<br />
non-ARS<br />
non-ARS<br />
non-ARS<br />
non-ARS<br />
non-ARS<br />
non-ARS<br />
ARS 1 d.<br />
ARS 2 d.<br />
ARS 2 d.<br />
ARS 2 d.<br />
Leiomyosarcoma of shin<br />
Cancer of colon<br />
Cancer of colon<br />
Cancer of kidney<br />
Cancer of stomach<br />
Cancer of stomach<br />
Cancer of lung<br />
Cancer of prostate<br />
Cancer of throat<br />
Cancer of colon<br />
Cancer of thyroid gland<br />
Cancer of thyroid gland<br />
Neurinoma of lower jaw<br />
1998<br />
1999<br />
2001<br />
2000<br />
2004<br />
2004<br />
2001<br />
2001<br />
2000<br />
1997<br />
2000<br />
2000<br />
2003<br />
Operated in 1998<br />
Operated in 1999<br />
Died in 2005<br />
Operated in 2001<br />
Died in 2004<br />
Died in 2005<br />
Operated in 2003<br />
Died in 2003<br />
Died in 2001<br />
Operated in 1997<br />
Operated in 2000<br />
Operated in 2001<br />
Died in 2004
Non-bone Marrow Syndromes<br />
Caused by Radiation Exposure<br />
Acute<br />
Radiation<br />
Syndrome<br />
Skin burns<br />
(%)<br />
Oropharyngeal<br />
Syndrome<br />
(%)<br />
Gastrointestinal<br />
(%)<br />
Radiation<br />
Pneumonitis<br />
(%)<br />
115 56<br />
(48.6)<br />
80<br />
(69.5)<br />
17<br />
(14.7)<br />
7<br />
(6.1)<br />
Barabanova A., Vojnosanit Pregl. 2006 May;63(5):477-80<br />
Ministry of Health, Clinical <strong>Department</strong> of the Institute of Biophysics, Moscow,<br />
Russia. abarabanova@rambler.ru
Stages of CRS<br />
Stage Onset Symptoms<br />
Prodromal 24-72<br />
hours<br />
Manifestation Days – 4<br />
weeks<br />
Transient erythema, pruritis<br />
Intense erythema, edema, pruritis, pain,<br />
blisters, erosions, ulcerative necrosis<br />
Subacute 4-6 weeks Erythema, edema, ulcers<br />
Chronic<br />
3 months-<br />
2 years<br />
Keratosis, fibrosis, ulcer, atrophy,<br />
pigment alteration, subcutaneous<br />
vasculitis, ulceration<br />
Late Decades Ulcers, angioma, fibrosis, keratosis,<br />
basal cell carcinoma<br />
Stages of the CRS according to Second Consensus Development Conference on<br />
the Management of Radiation Injuries, Bethesda, MD, 1993
Early and Late Skin Lesions in Radiation-exposed<br />
Patients after the <strong>Chernobyl</strong> <strong>Accident</strong><br />
ARS<br />
(Grade)<br />
Number<br />
of<br />
patients<br />
Body area<br />
Early skin<br />
lesions<br />
(1986)<br />
Late skin<br />
lesions<br />
Basal Cell<br />
Carcinoma<br />
(BCC)<br />
I 5<br />
feet, LE,<br />
trunk, hands<br />
Erythema,<br />
edema<br />
Atrophy, pigment<br />
alteration, xerosis<br />
II 6<br />
LE, UE,<br />
trunk + LE<br />
Erythema,<br />
edema<br />
Atrophy, pigment<br />
alteration, xerosis,<br />
keratosis, ulcers<br />
III 9<br />
Combinations<br />
of the above<br />
Erythema,<br />
edema,<br />
blisters, ulcers<br />
Atrophy, pigment<br />
alteration, fibrosis<br />
keratosis, ulcers<br />
IV 1<br />
trunk +<br />
extremities<br />
Blisters, ulcers<br />
Same as Grade III<br />
plus carcinomas<br />
2 BCC<br />
lesions<br />
Nonconfirmed<br />
group<br />
1<br />
TOTAL 22
Chronic Cutaneous Radiation Syndrome (CRS)<br />
Patient N. In 1986 suffered from severe ARS (3rd degree) and moderatesevere<br />
acute skin damage (2nd - 3rd degree) of right foot. Essential<br />
keratosis and fibrosis. Nail bone of 1 finger was amputated in 1986, the<br />
focus with transplanted skin are well visible.<br />
Courtesy of Drs Belyi and Bebeshko, Kiev, Ukraine
Chronic Cutaneous Radiation Syndrome (CRS)<br />
Patient K. In 1986 suffered from severe ARS (3rd degree) and severe acute<br />
radiation skin damage of both shins (3rd degree). On the frontal surface foci<br />
of hyperpigmentation and telangiectasis are visible (15 years had passed)<br />
Courtesy of Drs Belyi and Bebeshko, Kiev, Ukraine
Chronic Cutaneous Radiation Syndrome (CRS)<br />
Patient N. In 1986 suffered from severe ARS (3rd degree) and moderate<br />
acute skin damage (2nd degree) of right leg. After 15 years following skin<br />
changes dominate: telangiectasis, hyperpigmentation, keratosis, fibrosis.<br />
Courtesy of Drs Belyi and Bebeshko, Kiev, Ukraine
Treatment Experience of Skin Injuries in<br />
<strong>Chernobyl</strong> victims<br />
• Systemic treatment<br />
Hemoperfusion, plasmapheresis, continuous<br />
heparinization and administration of freshly frozen<br />
plasma<br />
• Local treatment<br />
Use of Combutec-2 for local treatment of skin injuries<br />
Aerosol Lioxanol<br />
Solution Balis-2<br />
• Pain management<br />
was challenging and not effective due to an<br />
absence of the local anesthetics in the treatment arsenal<br />
• Necessity of surgical operations at an early stage<br />
Guskova et al, 1988, Baranov et al, 1991, Selezneva, 1990, Barabanova, 2006
Non-radiological Health Effects<br />
• Psychological effects<br />
– Can overwhelm<br />
radiological physical effects<br />
– Symptoms are similar for<br />
different radiation<br />
emergencies (different<br />
scales)<br />
– Need for comprehensive<br />
strategy directed to<br />
different population<br />
groups before/during/after<br />
an emergency
“The largest public health problem<br />
unleashed by the accident is the mental health<br />
impact”<br />
(WHO report of the UN <strong>Chernobyl</strong> expert group, August 2005)<br />
• Stress-related symptoms<br />
• Chronic Fatigue Syndrome<br />
• Effects on the<br />
developing brain<br />
• Organic brain disorders<br />
in highly exposed<br />
clean-up workers<br />
• Suicide
Brain Damage in Clean-up Workers<br />
“Today it is recognized that the Central<br />
Nervous System (CNS) is a radiosensitive<br />
organ whose degree of dysfunction can be<br />
quantified by electrophysiological,<br />
biochemical and/or behavior parameters.<br />
Abnormalities in CNS function defined by<br />
these parameters may occur at a low dose of<br />
whole body radiation”
Impact of Low Level Radiation on<br />
Brain Development<br />
1. Children irradiated in<br />
utero in the first 4-5<br />
months of gestation<br />
have:<br />
• reduced verbal IQ at age<br />
11<br />
• ECG changes in the left<br />
hemisphere<br />
2. Treatment with low dose<br />
radiation in infancy leads<br />
neuropsychological<br />
disorders later in life<br />
120<br />
118<br />
116<br />
114<br />
112<br />
110<br />
108<br />
106<br />
104<br />
102<br />
100<br />
p
Schizophrenia Incidence<br />
Excess in <strong>Chernobyl</strong><br />
Exclusion Zone Personnel<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
<strong>Chernobyl</strong><br />
exclusion zone<br />
Ukraine<br />
Linear<br />
approximation<br />
(Exclusion zone)<br />
0<br />
At Issue: Schizophrenia Spectrum Disorders in Persons Exposed to Ionizing Radiation as a Result of the<br />
<strong>Chernobyl</strong> <strong>Accident</strong> by Konstantin N. Loganovsky and Tatiana K. Loganovskaja Schizophrenia Bulletin,<br />
26(4):751–773, 2000.
Neuro-Psychological Consequences:<br />
Summary<br />
• Genetic predisposition to schizophrenia can be provoked by<br />
environmental stressors including effects of exposure to ionizing<br />
radiation<br />
• Left hemisphere is<br />
the most radiovulnerable<br />
• Neuroimaging abnormalities<br />
are revealed following exposure<br />
to >0.3 Sv<br />
• The CNS effects that could be attributed to exposure to ionizing<br />
radiation are as follows: schizophrenia spectrum disorders;<br />
hronic Fatigue Syndrome; accelerated aging processes and<br />
neurodegeneration; and suicide
‰<br />
40<br />
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
The risk of development of cerebrovascular<br />
diseases is higher in recovery operation workers<br />
with doses of 0.25 Gy and higher as compared to<br />
those with an exposure of less than 0.1 Gy<br />
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003<br />
Years<br />
- < 0,05 Gy; - < 0,25 Gy; - 0,25-0,7 Gy<br />
trend: < 0.25 Gy trend: < 0.05 Gy trend: 0.25 - 0.7 Gy
Lessons Learned: Twenty years of Follow-up after<br />
the <strong>Chernobyl</strong> <strong>Accident</strong> (1)<br />
• Cutaneous component of the ARS had significantly<br />
complicated the clinical prognosis and contributed to or<br />
caused death in patients<br />
• Severe beta-burns of the skin remain an unsolved problem<br />
as a result of their spreading<br />
• The severity of the skin damage could have been avoided<br />
by removing the contaminated clothing<br />
• The prevention of late skin effects depends upon the<br />
effective management of acute lesions
Lessons Learned: Twenty years of Follow-up<br />
after the <strong>Chernobyl</strong> <strong>Accident</strong> (2)<br />
• Communications: who/how to contact, how to verify and<br />
confirm information<br />
• Confidentiality: different understanding of what was<br />
classified and what was not, limited information available<br />
for International professional community<br />
• Public health implications of the radiological accident:<br />
International significance was not as well understood as for<br />
communicable diseases incidents
Lessons Learned: Twenty years of Follow-up<br />
after the <strong>Chernobyl</strong> <strong>Accident</strong> (3)<br />
• The outcomes and late effects of the skin lesions depended<br />
on the depth-dose distribution and on the size of the area<br />
affected<br />
• Radiation-induced fibrosis is a predominant clinical problem<br />
• Appearance of secondary ulcerations presents treatment<br />
challenges<br />
• No malignant melanoma or squamous cell carcinoma have<br />
been detected so far
"An accident has occurred at <strong>Chernobyl</strong><br />
nuclear power station. One of the atomic<br />
reactors has been damaged. Measures are<br />
being taken to eliminate the<br />
consequences of the accident. Aid is<br />
being given to the victims. A government<br />
commission has been set up."
Be clear of what you are trying to say!
Lessons Learned: Twenty years of Follow-up<br />
after the <strong>Chernobyl</strong> <strong>Accident</strong> (4)<br />
• Effective medical care is generally not possible for accident<br />
victims with high-dose TBI<br />
• Most individuals will not receive a sufficiently high dose to make<br />
a bone marrow transplant necessary for hematological recovery<br />
• Only a small number of patients will have bone marrow<br />
syndrome without other life-threatening non-bone marrow<br />
related complication<br />
• Transplants should probably be considered for victims receiving<br />
more than 7 to 8 Gy of external radiation
Lessons Learned: Twenty years of Follow-up<br />
after the <strong>Chernobyl</strong> <strong>Accident</strong> (5)<br />
• Maximize the education of physicians<br />
• Provide medical community with practical tools how to<br />
identify and assess radiation victims<br />
• Explain situation in plain language and avoid conflicting<br />
information<br />
• Stay in touch with collaborating centers in<br />
European and other countries experienced<br />
in managing radiation emergencies
Invisible danger still exists
What was the most unexpected for us?<br />
• Diversity of clinical manifestations of skin lesions<br />
• Unaccustomed course of clinical phases of a<br />
radiation injury to skin<br />
• Significant severity of injuries<br />
• Serious influence of skin burns on the general state<br />
of a patient<br />
• Need for surgical interventions at an early stage
LUCK FAVORS THE PREPARED!<br />
Thank You!
Acknowledgements:<br />
• M.Tronko, T.Bogdanova et al, G. Thomas et al,<br />
Institute of Endocrinology and Metabolism, Acad. Med. Sc., Ukraine; Swansea University,<br />
UK; University, Japan; Institute of Oncology and Radiology, Acad. Med. Sc., Ukraine’’.<br />
• D.A.Bazyka, V.G.Bebeshko, D. Belyi, A.E.Romanenko, V.A.Buzunov, A.E.Prysyazhniuk,<br />
K.M.Loganovsky, M.I.Omelyanets<br />
Research Centre for Radiation Medicine (Kyiv)<br />
• Ihor J. Masnyk, Ph.D., NCI, Epidemiology BranchU.S. Director Chornobyl Research<br />
Projects<br />
• M.N.<br />
• Savkin, L.A. Ilyin, A.K. Guskova<br />
• State Research Center – Institute of Biophysics, Moscow, Russia<br />
• Albert L. Wiley, MD, PhD, Director, Radiation Emergency Assistance Center Training Site<br />
(REAC/TS)<br />
• Patrick Gourmelon, T.M.Fliedner and V. Meineke, Institute for Radiation and <strong>Nuclear</strong><br />
Safety (France), University of ULM (Germany)<br />
• Pierre Flor-Henry, Konstantin Loganovsky, Alberta Hospital Edmonton, Canada, Research<br />
Centre for Radiation Medicine, AMS of Ukraine, Kyiv
Radioprotectants Currently Approved in<br />
Russian Federation<br />
ANTI-RADIATION FIRST- AID KIT CREATED<br />
Includes:<br />
• INDRALIN – neutralizes Cesium and Strontium<br />
• LIOXAZOL (spray) – for early treatment of<br />
radiation sickness and spray for skin burns<br />
• ZASHITA (PROTECTION) – deactivation and<br />
protection of skin
Radioprotectants approved in<br />
Russian Federation (continued)<br />
• DEZOXYNATUM stimulates the proliferation of hemopoetic<br />
cells<br />
• Chemical structure: Sodium salt of DNA extracted from the<br />
milt of sturgeon species<br />
• Mechanism of Action: stimulates<br />
proliferation of hemopoetic cells<br />
• Indication: ARS, hypo-and aplastic<br />
anemia secondary to chemo<br />
or radiation therapy<br />
• Contraindications: none<br />
• Side effects: low grade fever (infrequent)
USING ATMOSPHERIC 137 CS MEASUREMENTS AND HYSPLIT TO CONFIRM<br />
CHERNOBYL AS A SOURCE OF 137 CS IN EUROPE<br />
Erik L. Swanberg 1 and Steven G. Hoffert 2<br />
Veridian Systems 1 , Autometric 2<br />
Sponsored by <strong>Defense</strong> Threat Reduction Agency<br />
Contract No. DTRA01-99-C-0031<br />
ABSTRACT<br />
The <strong>Chernobyl</strong> nuclear reactor accident released considerable amounts of radioactive material into the<br />
environment, including a large amount of 137 Cs. A large fraction of the 137 Cs was deposited on the ground in the<br />
surrounding areas. Two atmospheric monitoring stations that contribute data to the Prototype International Data<br />
Centre (PIDC), one in Stockholm, Sweden, and the other in Helsinki, Finland, routinely measure 137 Cs. It is<br />
believed that the source of this 137 Cs is the ground contaminated by the <strong>Chernobyl</strong> accident. The PIDC<br />
routinely uses HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory) atmospheric modeling<br />
software to determine probable source locations of radionuclides detected during normal operations. In this<br />
paper, HYSPLIT was used in conjunction with the data from the PIDC to more firmly establish the link<br />
between <strong>Chernobyl</strong> and 137 Cs measurements. The results indicate that an air mass containing 137 Cs has a higher<br />
likelihood of having recently been in the <strong>Chernobyl</strong> area than an air mass that does not contain 137 Cs. The<br />
inverse seems true also: an air mass that does not contain 137 Cs is far less likely to have been in the vicinity of<br />
<strong>Chernobyl</strong> in the recent past. These results, while not definitive, are very encouraging. The results also<br />
improve the confidence in HYSPLIT.<br />
KEY WORDS: <strong>Chernobyl</strong>, Cesium, PIDC<br />
OBJECTIVE<br />
In 1986, the <strong>Chernobyl</strong> accident released large amounts of many different radionuclides into the atmosphere.<br />
Starting with its first sample in 1996, the Prototype International Data Centre (PIDC) has measured 137 Cs at<br />
radionuclide monitoring stations in Europe. This paper attempts to link 137 Cs re-suspension in the vicinity of<br />
<strong>Chernobyl</strong> to the 137 Cs regularly measured by monitoring stations. To this end, HSYPLIT (HYbrid Single-<br />
Particle Lagrangian Integrated Trajectory) atmospheric modeling software was used in conjunction with<br />
radionuclide concentration measurements to show that the <strong>Chernobyl</strong> region is likely the source of 137 Cs.<br />
There are about 150,000 km 2 of land contaminated with 137 Cs at greater than 37 kBq/m 2 (UNSCEAR, 2000). It<br />
is not surprising that conditions exist that re-suspend this 137 Cs. One example is a forest fire in the <strong>Chernobyl</strong><br />
area that re-suspended measurable amounts of 137 Cs, that was detected by the PIDC in May, 2000. Once<br />
airborne, the 137 Cs is transported considerable distances in the atmosphere. In particular, the 137 Cs is transported<br />
to European atmospheric monitoring stations before the 137 Cs concentration falls below the minimum detectable<br />
concentration.<br />
There are four PIDC monitoring stations in Europe, all of which regularly measure 137 Cs. Two of these, one in<br />
England and the other in Germany, sample air for seven days. For the purposes of this study, these<br />
measurements are too long to make reasonable predictions about the origin of the 137 Cs re-suspension. Two<br />
other stations, one in Stockholm, Sweden, and the other in Helsinki, Finland, sample air for 24 hours. This<br />
shorter sampling time allows better estimation of an air mass's previous position. This study, therefore, only<br />
uses data from the Swedish and Finnish stations.<br />
To model atmospheric transport, the PIDC runs HYSPLIT daily and generates a Field of Regard (FOR). The<br />
FORs show areas where it is most likely a parcel of air originated during a specified time period, and hence help<br />
864
Report Documentation Page<br />
Form Approved<br />
OMB No. 0704-0188<br />
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and<br />
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,<br />
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and <strong>Reports</strong>, 1215 Jefferson Davis Highway, Suite 1204, Arlington<br />
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it<br />
does not display a currently valid OMB control number.<br />
1. REPORT DATE<br />
OCT 2001<br />
2. REPORT TYPE<br />
3. DATES COVERED<br />
00-00-2001 to 00-00-2001<br />
4. TITLE AND SUBTITLE<br />
Using Atmospheric 137CS Measurements And Hysplit To Confirm<br />
<strong>Chernobyl</strong> As A Source of 137CS In Europe<br />
5a. CONTRACT NUMBER<br />
5b. GRANT NUMBER<br />
5c. PROGRAM ELEMENT NUMBER<br />
6. AUTHOR(S) 5d. PROJECT NUMBER<br />
5e. TASK NUMBER<br />
5f. WORK UNIT NUMBER<br />
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)<br />
Veridian Systems,6 Pecan St.,California,MD,20619<br />
8. PERFORMING ORGANIZATION<br />
REPORT NUMBER<br />
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)<br />
12. DISTRIBUTION/AVAILABILITY STATEMENT<br />
Approved for public release; distribution unlimited<br />
11. SPONSOR/MONITOR’S REPORT<br />
NUMBER(S)<br />
13. SUPPLEMENTARY NOTES<br />
Proceedings of the 23rd Seismic Research Review: Worldwide Monitoring of <strong>Nuclear</strong> Explosions held in<br />
Jackson Hole, WY on 2-5 of October, 2001. U.S. Government or Federal Rights.<br />
14. ABSTRACT<br />
See Report<br />
15. SUBJECT TERMS<br />
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF<br />
ABSTRACT<br />
a. REPORT<br />
unclassified<br />
b. ABSTRACT<br />
unclassified<br />
c. THIS PAGE<br />
unclassified<br />
Same as<br />
Report (SAR)<br />
18. NUMBER<br />
OF PAGES<br />
7<br />
19a. NAME OF<br />
RESPONSIBLE PERSON<br />
Standard Form 298 (Rev. 8-98)<br />
Prescribed by ANSI Std Z39-18
to indicate possible sources of radionuclides detected. These FORs were used to show a link between 137 Cs<br />
measurements and <strong>Chernobyl</strong>. About 18 months of FOR data was used for this study.<br />
RESEARCH ACCOMPLISHED<br />
This section begins by describing the data that was used from the PIDC. It then covers conditions that are<br />
favorable for transporting 137 Cs from <strong>Chernobyl</strong> to the monitoring stations. It also describes how HYSPLIT<br />
was used to generate FOR data. Then the link between 137 Cs measured at the monitoring stations and<br />
<strong>Chernobyl</strong> is established.<br />
PIDC Data<br />
The PIDC receives data from a global system of atmospheric monitoring stations. The Swedish and Finish<br />
systems are two of these stations. The stations send High Purity Germanium (HPGe) gamma-ray spectra to the<br />
PIDC on a regular basis, usually daily. The spectra are analyzed to determine which radionuclides are present<br />
and what their concentrations are. This includes concentrations for 137 Cs. The Swedish and Finish stations have<br />
been providing the PIDC with this data since 1996.<br />
The PIDC uses atmospheric data and atmospheric modeling software to estimate previous locations of air that<br />
was sampled by a monitoring station. The model is run daily. This allows the PIDC to estimate the source of<br />
any unusual radionuclides that are observed at a station. The modeling software results were available from<br />
May 1999 to September 2000, which are the dates included in this study.<br />
137 Cs Transport<br />
In order for 137 Cs to travel from <strong>Chernobyl</strong> to a European monitoring station located over 1000 kilometers away<br />
it must first be re-suspended in the air. Then atmospheric conditions must be such that the 137 Cs is transported<br />
to the monitoring station. Some possible mechanisms for re-suspension include high winds stirring up<br />
contaminated soil (Gillette and Porch, 1978), burning of contaminated vegetation (Garger et al, 1998), or<br />
perhaps a disturbance in the sarcophagus surrounding the reactor. Atmospheric conditions consist mostly of<br />
prevailing winds blowing in the correct direction. This study did not attempt to look at processes that could resuspend<br />
the 137 Cs. Hence discrepancies could result where HYSPLIT indicates air transport in the correct<br />
direction but no 137 Cs was detected. On the other hand, if 137 Cs from the <strong>Chernobyl</strong> region was measured, then<br />
both of the above conditions must have been met.<br />
HYSPLIT<br />
The PIDC uses HYSPLIT atmospheric modeling software to simulate particle trajectories in the atmosphere.<br />
These trajectories are then used to create FORs as follows. Simulated particles were released every hour from a<br />
2 X 2 grid formed from points located at the intersection of even lines of latitude with even lines of longitude.<br />
HYSPLIT used measured atmospheric data to simulate the trajectories real particles would follow if released<br />
under the same conditions. If a particle passed near a station in the simulation, its point of origin was noted as<br />
having a particle reach the detector. The simulations were run for 24, 48, and 72 hours. After a run was<br />
complete, the number of particles that reached the detector was totaled, and the grid point tallies were<br />
normalized so that the sum of all grid point values in an FOR is one.<br />
A few remarks are in order. First, the 2 X 2 grid square is somewhat coarse, resulting in coarse resolution.<br />
Second, the model was only run for 72 hours. The distance between <strong>Chernobyl</strong> and the monitoring stations is<br />
large enough that it could easily take more than 72 hours for an air mass to travel between the two. Therefore,<br />
some 137 Cs measurements were possibly not associated with <strong>Chernobyl</strong> because the air mass took longer than<br />
72 hours to reach the station. Also remember that the atmosphere is a complex system and no model is perfect.<br />
Combining HYSPLIT and 137 Cs Concentrations with Atmospheric Transport<br />
Because FORs are normalized, combining them is accomplished by simply adding them together. The figures<br />
that follow were created with three different objectives in mind. Figure 1 is an average of all FORs that were<br />
965
available for the Swedish station. It serves as a baseline and allows any unusual features to be seen. Figure 2<br />
only includes FORs for days when the concentration of 137 Cs was above 4 Bq/m 3 , which indicates likely<br />
origins of high 137 Cs concentrations. Figure 3 only includes FORs for days when 137 Cs was not measured,<br />
indicating where 137 Cs does not originate. The FORs are drawn as contour plots. Figures 4 through 6 are<br />
corresponding FORs for the Finnish station.<br />
The figures show that when 137 Cs was measured, air was much more likely to have come from <strong>Chernobyl</strong>. In<br />
figures 2 and 5, high likelihood regions are near the stations and <strong>Chernobyl</strong>. There are other regions where the<br />
air might have originated, but they have smaller associated probabilities. Since the stations sample air for 24<br />
hours, it is possible for more than one air mass to pass over a station and be included in a sample. Also, air in<br />
transit for more than 72 hours is not included in the FORs. Data from both stations indicate <strong>Chernobyl</strong> as being<br />
the most likely source for 137 Cs.<br />
It is also interesting to note where the air originates when no 137 Cs is detected (Figures 3 and 6). Air appears to<br />
come from similar places as the average (Figures 1 and 4) but with lower probabilities of originating near the<br />
<strong>Chernobyl</strong> region. This is what would be expected if <strong>Chernobyl</strong> were the source of 137 Cs. Since none of the<br />
other places where air normally reaches the stations from are a source of high 137 Cs concentrations, <strong>Chernobyl</strong><br />
must be the source of the elevated measurements.<br />
Instead of selecting FORs based on whether or not 137 Cs was measured, they can be selected based on having a<br />
high probability of air reaching a radionuclide monitoring station from the <strong>Chernobyl</strong> area. This was done<br />
using four grid points near <strong>Chernobyl</strong>. For Sweden, 137 Cs was seen 11 out of 13 times when one of these four<br />
points had a probability of greater than .05. For Finland the same analysis yields 6 of 9 times. The days with<br />
higher likelihood also include the highest measurements of 137 Cs made at both stations. Table 1 contains a list<br />
of probabilities and 137 Cs concentrations.<br />
For several particularly high concentrations of 137 Cs, HYSPLIT trajectory models were run. Figure 7 shows the<br />
result for one of these runs. It can clearly be seen this day (which was the highest concentration measured to<br />
date at the Swedish station) that the air mass goes almost directly from the <strong>Chernobyl</strong> region to Stockholm.<br />
Table 1. Probabilities of air originating at a grid point near <strong>Chernobyl</strong> and 137 Cs concentrations actually<br />
measured.<br />
Sweden<br />
Finland<br />
Probability Concentration (Bq/m 3 ) Probability Concentration (Bq/m 3 )<br />
0.053 0.59 0.065 Not Detected<br />
0.053 0.58 0.071 0.86<br />
0.055 Not Detected 0.071 0.86<br />
0.065 1.9 0.14 5.7<br />
0.065 1.9 0.17 3.3<br />
0.067 1.8 0.17 3.3<br />
0.069 6.4 0.19 Not Detected<br />
0.069 6.4 0.21 5.7<br />
0.11 45 0.22 Not Detected<br />
0.17 16<br />
0.19 Not Detected<br />
0.32 45<br />
0.39 16<br />
10 66
Swedish<br />
Station<br />
>.01<br />
>.1<br />
>.5<br />
>1<br />
Figure 1. Average of all available FORs for the Swedish station.<br />
<strong>Chernobyl</strong><br />
>.02<br />
>.1<br />
>.2<br />
>.3<br />
<strong>Chernobyl</strong><br />
Swedish<br />
Station<br />
Figure 2. FORs for the Swedish station when the concentration of 137 Cs was greater than 4 Bq/m 3 .<br />
>.01<br />
>.1<br />
>.5<br />
>1<br />
Swedish<br />
Station<br />
Figure 3. FORs for the Swedish station when no 137 Cs was measured.<br />
<strong>Chernobyl</strong><br />
11 67
Figure 4. Average of all available FORs for the Finish station.<br />
<strong>Chernobyl</strong><br />
Finish<br />
Station<br />
>.01<br />
>.1<br />
>.5<br />
>1<br />
>.02<br />
>.1<br />
>.2<br />
>.3<br />
<strong>Chernobyl</strong><br />
Finish<br />
Station<br />
Figure 5. FORs for the Finish station when the concentration of 137 Cs was greater than 4 Bq/m 3 .<br />
>.01<br />
>.1<br />
>.5<br />
>1<br />
Finish<br />
Station<br />
<strong>Chernobyl</strong><br />
Figure 6. FORs for the Finish station when no 137 Cs was measured.<br />
12 68
Swedish<br />
Station<br />
<strong>Chernobyl</strong><br />
Figure 7. Trajectory for the day the highest concentration of 137 Cs was measured in Stockholm.<br />
CONCLUSIONS AND RECOMMENDATIONS<br />
A link was established between 137 Cs re-suspended in the <strong>Chernobyl</strong> region and 137 Cs measured by two<br />
European monitoring stations in several ways. First, it was shown that when 137 Cs was measured, it was more<br />
likely for the air that was sampled to have come from the <strong>Chernobyl</strong> region. Secondly, when no 137 Cs was<br />
measured the sampled air was much less likely to have been in the <strong>Chernobyl</strong> region. Likewise, if air did not<br />
come from the <strong>Chernobyl</strong> region, 137 Cs wasn't seen. It was also shown that if air did move from the <strong>Chernobyl</strong><br />
region to a monitoring station, it was likely to contain detectable concentrations of 137 Cs. While these results<br />
are qualitative in nature, the evidence strongly suggests that a correlation exists.<br />
Another benefit of this study is an increased confidence in the tools used and the manner in which they are used.<br />
Data analysis using two different stations yielded equivalent results. These results agree with what common<br />
sense dictates, namely that re-suspension of 137 Cs in the <strong>Chernobyl</strong> region is the source of 137 Cs in Sweden and<br />
Finland. One can thus conclude that using radionuclide concentrations obtained with currently fielded systems<br />
in conjunction with HYSPLIT produces good results.<br />
If more conclusive evidence of the link between 137 Cs re-suspension near <strong>Chernobyl</strong> and high 137 Cs<br />
concentrations is desired, there are a couple of investigations that could be performed. First, 134 Cs was also<br />
released during the accident. In fact, several of the earlier samples measured by the PIDC contain both 137 Cs<br />
and 134 Cs. Due to its 2-year half life, almost all of the 134 Cs will have decayed. But it might be possible, using a<br />
detection methodology with reduced minimum detectable concentrations, to measure the ratio of 137 Cs to 134 Cs<br />
in both soil in the <strong>Chernobyl</strong> region and in the samples from the monitoring stations. If the ratios match, this<br />
would be strong evidence for one causing the other. Also, determining conditions good for the re-suspension of<br />
137 Cs into the atmosphere could help to improve the data used in this study.<br />
13 69
REFERENCES<br />
Gillette, Dale A. and William M. Porch (1978), The Role of Fluctuations of Vertical and Horizontal Wind and<br />
Particle Concentration in the Deposition of Dust Suspended by Wind, J. of Geophysical Research,<br />
January 20, 1978, 409-414<br />
Garger, E. K., V. Kashpur, H. G. Paretzke, and J. Tschiersch (1998) Measurement of Resuspended Aerosol in<br />
the <strong>Chernobyl</strong> Area Part II. Size Distribution of Radioactive Particles, Radiation and Environmental<br />
Biophysics Vol 36 Number 4, 275-283<br />
HYSPLIT4 (HYbrid Single-Particle Lagrangian Integrated Trajectory) Model (1997). Web address:<br />
http://www.arl.noaa.gov/ready/hysplit4.html, NOAA Air Resources Laboratory, Silver Spring, MD.<br />
United Nations Scientific Committee on the Effects of Atomic Radiation, UNSCEAR 2000, Sources and Effects<br />
of Ionizing Radiation, Volume II: Effects, Report to the General Assembly, with Scientific Annexes.<br />
14 70