Chernobyl Nuclear Accident Defense Department Reports

CHERNOBYL NUCLEAR ACCIDENT DOCUMENTS 

DEPARTMENT OF DEFENSE DOCUMENTS 

BACM RESEARCH 

WWW.PAPERLESSARCHIVES.COM

About BACM Research – PaperlessArchives.com 

BACM Research/PaperlessArchives.com publishes documentary historical research collections. 

Materials cover Presidencies, Historical Figures, Historical Events, Celebrities, Organized Crime, Politics, 

Military Operations, Famous Crimes, Intelligence Gathering, Espionage, Civil Rights, World War I, World 

War II, Korean War, Vietnam War, and more. 

Source material from Federal Bureau of Investigation (FBI), Central Intelligence Agency (CIA), National 

Security Agency (NSA), Defense Intelligence Agency (DIA), Secret Service, National Security Council, 

Department of Defense, Joint Chiefs of Staff, Department of Justice, National Archive Records and 

Administration, and Presidential Libraries. 

http://www.paperlessarchives.com

CHERNOBYL NUCLEAR ACCIDENT DOCUMENTS 

http://www.paperlessarchives.com/chernobyl_nuclear_accident_doc.html

Chernobyl Nuclear Power Plant Accident CIA, Department 

of Defense, Department of Energy, Congressional, GAO, 

and Foreign Press Monitoring Files 

4,010 pages of CIA, Department of Defense, Department of Energy, 

Congressional, GAO, and foreign press monitoring files related to the 

Chernobyl Nuclear Accident. 

On Sunday April 26, 1986, at the Chernobyl Nuclear Power Plant near 

Pripyat, Ukraine, reactor #4 exploded. For the 25 years from 1986 to 

2011, this incident has been referred to as the world's worst nuclear 

power plant accident. 

THE ACCIDENT 

According to reports filed with International Atomic Energy Agency (IAEA) 

on April 25, 1986, technicians at the Chernobyl plant launched a poorly 

executed experiment to test the emergency electricity supply to one of 

its Soviet RBMK type design reactors. The test was meant to measure a 

turbogenerator's ability to provide in-house emergency power after 

shutting off its steam supply. During the experiment the technicians 

violated several rules in place for operating the reactor. 

During the experiment, the emergency shutdown system was turned off. The 

reactor was being operated with too many control rods withdrawn. These 

human errors, coupled with a design flaw that allowed reactor power to 

surge when uncontrolled steam generation began in the core, set up the 

conditions for the accident. 

A chain of events lasting 40 seconds occurred at 1:23 AM on April 26. 

The technicians operating the reactor put the reactor in an unstable 

condition, so reactor power increased rapidly when the experiment began. 

Subsequent analysis of the Soviet data by U.S. experts at the Department 

of Energy, suggests the power surge may have accelerated when the 

operators tried an emergency shutdown of the reactor. According to Soviet 

data, the energy released was, for a fraction of a second, 350 times the 

rated capacity of the reactor. This burst of energy resulted in an 

instantaneous and violent surge of heat and pressure, rupturing fuel 

channels and releasing steam that disrupted large portions of the core. 

The surge destroyed the core of reactor unit four, containing 

approximately 200 tons of nuclear fuel. Some of the shattered core 

material was propelled through the roof of the reactor building. The hot 

core material of reactor 4 started about 30 separate fires in the unit 4 

reactor hall and turbine building, as well as on the roof of the 

adjoining unit 3. All but the main fire in the graphite moderator 

material still inside unit 4 were extinguished in a few hours. 

It was a day and a half before the people living in Pripyat were ordered 

to evacuate. The residents were told they would only be gone for several

days, so they left nearly everything behind. They never returned. Soviet 

authorities made the decision not to cancel May 1, May Day, outdoor 

parades in the region four days later. 

The graphite fire continued to burn for nearly two weeks carrying 

radioactivity high into the atmosphere, until it was smothered by sand, 

lead, dolomite, and boron dropped from helicopters. Despite the wide 

spread of radiation, Soviet officials at first said very little publicly 

about what happened at Chernobyl. It was not until alarms from radiation 

detectors in other countries, many hundreds of miles away, forced the 

Soviets to admit to the Chernobyl accident. 

Radioactive material was dispersed over 60,000 square miles of Ukraine, 

Belarus, and Russia. Smaller amounts of radioactive material were 

detected over Eastern and Western Europe, Scandinavia and even the United 

States. The accident has left some nearby towns uninhabitable to this 

day. 

Radioactivity forced Soviet officials to create a 30-kilometer-wide nohabitation 

zone around Chernobyl, sealing off Pripyat. Still, the power 

plant continued to generate electricity until it was finally shut down in 

December, 2000. 

During the first year after the accident, about 25,000 people, mainly 

Soviet Army troops, were dispatched to the site to clean up the accident. 

Thousands of workers, called liquidators, were employed during the 

following years of the cleanup. 

Around October, 1986 the construction of a 21 story high metal and 

concrete shelter was completed, enclosing the reactor and the radioactive 

material that remained. Almost 200 tonnes of melted nuclear fuel rods 

remain within the damaged reactor. This containment shelter was not 

intended to be a permanent solution for containing the radioactive 

material. Over time, the shelter has weakened; rain entering through 

holes and cracks has caused corroding. 

By 2006 the plans for a new shelter was about 7 years behind schedule, 

with a completion target date of no sooner than 2012. In February of 2011 

it was reported that construction of the shelter may have to be halted, 

due to a $1 billion dollar short fall in the funds needed to complete the 

structure. 

A United Nations report released in February 2011 estimates the disaster 

caused thyroid cancer in 7,000 children in the affected area. The report 

said despite the high rate of cancer, only 15 fatalities in these 7,000 

cases have occurred. 

THE DOCUMENTS 

CIA FILES

215 pages of CIA files dating from 1971 to 1991.The files cover the 

Soviet Union's atomic energy program; The effect of the Chernobyl 

accident on the Soviet nuclear power program; and the social and 

political ramifications of the accident in the Soviet Union. 

A 1981 report covers the less publicized Soviet nuclear "accident" near 

Kyshtym in 1957-58. 

Media reporting of a nuclear accident near Kyshtym first appeared in 

1958. It was not until 1976, when the writings of Soviet dissent Dr. 

Zhores Medvedev began to appear, that wider attention was given to this 

subject. Medvedev, an exiled Soviet geneticist, claimed in several 

articles and books that a "disaster" occurred near Kyshtym in 1957/58. He 

alleged that thousands of casualties and widespread, long-term 

radioactive contamination occurred as the result of an explosion 

involving nuclear waste stored in underground shelters. 

The general consensus today is that a combination of events, rather than 

a single isolated incident at Kyshtym nuclear energy complex caused the 

radioactive contamination in the area. A study of the claims by Medvedev 

can be found in the Department of Energy section, in the 1982 report "An 

Analysis of the Alleged Kyshtym Disaster" 

U.S. GOVERNMENT FOREIGN PRESS MONITORING 

900 pages of foreign media monitoring reports from 1986 to 1992, produced 

by the U.S. government's National Technical Information Service's U.S. 

Joint Publication Research Service. They contain information primarily 

from Russian and Eastern Block news agency transmissions and broadcasts, 

newspapers, periodicals, television, radio and books. Materials from non- 

English language sources are translated into English. 

The reporting includes firsthand accounts of experiences during all 

points of the Chernobyl disaster. Topics covering the accident and its 

aftermath including domestic and international politics, sociological 

affairs, nuclear plant fire, evacuations, sealing the reactor, 

cleanup mobilization, health implications, and people returning to 

region. 

DEPARTMENT OF ENGERY REPORTS 

1,244 pages of reports dating from 1982 to 2009 produced or commissioned 

by the Department of Energy. 

The agencies and institutions contributing to these reports include Los 

Alamos National Laboratory, United States Nuclear Regulatory Commission, 

Lawrence Livermore National Laboratory, Savannah River Nuclear Solutions, 

Oak Ridge National Laboratory, Brookhaven National Laboratory, Argonne 

National Laboratory, and the Pacific Northwest Laboratory. 

Highlights include:

The 1986 Report of the U.S. Department of Energy's Team Analyses of the 

Chernobyl-4 Atomic Energy Station Accident Sequence DOE/NE-0076. 

The U.S. Department of Energy (DOE) formed a team of experts from Argonne 

National Laboratory, Brookhaven National Laboratory, Oak Ridge National 

Laboratory, and Pacific Northwest Laboratory. The DOE team provided the 

analytical support to the U.S. delegation for the August, 1986 meeting of 

the International Atomic Energy Agency (IAEA), and to subsequent 

international meetings. The DOE team analyzed the accident in detail, 

assessed the plausibility and completeness of the information provided by 

the Soviets, and performed studies relevant to understanding the 

accident. 

The 1987 report Radioactive Fallout from the Chernobyl Nuclear Reactor 

Accident 

The Lawrence Livermore National Laboratory performed a variety of 

measurements to determine the level of the radioactive fallout on the 

western United States. The laboratory used gamma-spectroscopy to analyze 

air filters from the areas around Lawrence Livermore National Laboratory 

in California. Filters were also analyzed from Barrow and Fairbanks, 

Alaska. Milk from California and imported vegetables were also analyzed 

for radioactivity. 

Other report titles include: An Analysis of the Alleged Kyshtym Disaster; 

Workshop on Short-term Health Effects of Reactor Accidents; Preliminary 

Dose Assessment of the Chernobyl Accident; Internally Deposited Fallout 

from the Chernobyl Reactor Accident; Report on the Accident at the 

Chernobyl Nuclear Power Station; Radioactive Fallout from the Chernobyl 

Nuclear Reactor Accident; Radioactivity in Persons Exposed to Fallout 

from the Chernobyl Reactor Accident' Radioactive Fallout in Livermore, CA 

and Central Northern Alaska from the Chernobyl Nuclear Reactor Accident; 

Projected Global Health Impacts from Severe Nuclear Accidents - 

Conversion of Projected Doses to Risks on a Global Scale - Experience 

From Chernobyl Releases; The Chernobyl Accident - Causes and 

Consequences; Chernobyl Lessons Learned Review of N Reactor; 

Reconstruction of Thyroid Doses for the Population of Belarus Following 

the Chernobyl Accident; The characterization and risk assessment of the 

Red Forest radioactive waste burial site at Chernobyl Nuclear Power 

Plant; Estimated Long Term Health Effects of the Chernobyl Accident; and 

Environmental Problems Associated With Decommissioning the Chernobyl 

Nuclear Power Plant Cooling Pond. 

DEPARTMENT OF DEFENSE REPORTS 

816 pages of reports dating from 1990 to 2010 produced or commissioned by 

the Department of Defense. 

The reports include: Chernobyl Accident Fatalities and Causes; Biomedical 

Lessons from the Chernobyl Nuclear Power Plant Accident; Nuclear

Accidents in the Former Soviet Union Kyshtym, Chelyabinsk and Chernobyl; 

Retrospective Reconstruction of Radiation Doses of Chernobyl Liquidators 

by Electron Paramagnetic Resonance; Neurocognitive and Physical Abilities 

Assessments Twelve Years After the Chernobyl Nuclear Accident; Simulating 

Wet Deposition of Radiocesium from the Chernobyl Accident; and Radiation 

Injuries After the Chernobyl Accident Management, Outcome, and Lessons 

Learned. 

GAO REPORTS 

184 pages of reports from the United States General Accounting Office, 

whose name was later changed to the Government Accountability Office. The 

four reports are Comparison of DOE's Hanford N-Reactor with the Chernobyl 

Reactor (1986); Nuclear Power Safety International Measures in Response 

to Chernobyl Accident (1988); Nuclear Power Safety Chernobyl Accident 

Prompted Worldwide Actions but Further Efforts Needed (1991); and 

Construction of the Protective Shelter for the Chernobyl Nuclear Reactor 

Faces Schedule Delays, Potential Cost Increases, and Technical 

Uncertainties (2007). 

UNITED STATES CONGRESSIONAL HEARINGS 

634 pages of transcripts from three Congressional hearings: The Chernobyl 

Accident Hearing before the Committee on Energy and Natural Resources, 

Ninety-ninth Congress, 2nd session on the Chernobyl accident and 

implications for the domestic nuclear industry, June 19, 1986; The 

Effects of the accident at the Chernobyl nuclear power plant hearing 

before the Subcommittee on Nuclear Regulation, United States Senate, One 

Hundred Second Congress, second session, July 22, 1992; and The legacy of 

Chernobyl, 1986 to 1996 and beyond hearing before the Commission on 

Security and Cooperation in Europe, One Hundred Fourth Congress, second 

session, April 23, 1996.

~TI 

FLE CY 

N 

Defense Nuclear Agency 

Alexandria, VA 22310-3398 

SWES% Ot 

DNA-TR-89-45 

Chernobyl Accident Fatalities and Causes 

A. Laupa 

G. H. Anno 

Pacific-Sierra Research Corporation 

12340 Santa Monica Boulevard 

Los Angeles, CA 90025-2587 

June 1990 

Technical Report 

CONTRACT No. DNA 001-87-C-0104' 

Approved for public release; - 

distribution is unlimited. 

___ __ ___ ___ _A 

___ __ 

0'.

Destroy this report when it is no longer needed. Do not 

return to sender. 

PLEASE NOTIFY THE DEFENSE NUCLEAR AGENCY, 

ATTN: CSTI, 6801 TELEGRAPH ROAD, ALEXANDRIA, VA 

22310-3398, IF YOUR ADDRESS IS INCORRECT, IF YOU 

WISH IT DELETED FROM THE DISTRIBUTION LIST, OR 

IF THE ADDRESSEE IS NO LONGER EMPLOYED BY YOUR 

ORGANIZATION. 

.N 4

DISTRIBUTION LIST UPDATE 

I 

This mailer is provided to enable DNA to maintain current distribution lists for reports. We would 

appreciate your providing the requested information. 

NAME: 

ORGANIZATION: 

OLD ADDRESS 

CURRENT ADDRESS 

I TELEPHONE NUMBER: ( ) 

z5: 

SUBJECT AREA(s) OF INTEREST: 

DNA OR OTHER GOVERNMENT CONTRACT NUMBER: 

CERTIFICATION OF NEED-TO-KNOW BY GOVERNMENT SPONSOR (if other than DNA): 

I SPONSORING ORGANIZATION: 

CONTRACTING OFFICER OR REPRESENTATIVE: 

El Add the individual listed to your distribution list. Please return OTE:- 

Plaertr the mailing label from 

SD e te cd o n the document so that any additions, 

El Delete the cited organization/individual, changes, corrections or deletions 

S Cacan be made more easily. 

0 E Change of address. 

________________ 

SIGNATURE.-

Director 


ATTN: TITL 

Washington, DC 20305-1000 

Director 


ATTN: TITL 

Washington, DC 20305 1000

REPO NTATON T D CUM AGEForm 

Aproved 

CE R DOMB No. 0704-0188 

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 

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 

colleclon of Information Including suggestions for reduc:ing this burden to Washington Headquarters Servces Directorate for Information Operatlons and ReportIs 1215 Jefferson 

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 

1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED 

1 900601 Technical 870729 - 881231 

4 TITLE AND SUBTITLE 5. FUNDING NUMBERS 

C - DNA 001-87-C-0104 

Chernobyl Accident Fatalities and Causes PE - 62715H 

PR - RM 

6 AUTHOR(S) TA -RH 

A. Laupa: G. H. Anno WU - DH026130 

7 PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8 PERFORMING ORGANIZATION 

REPORT NUMBER 

Pacific-Sierrra Research Corporation PSR Report 1865 


Los Angeles. CA 90025-2587 

9 SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10 SPONSORING/MONITORING 

l)efense Nuclear Agency 

AGENCY REPORT NUMBER 

6801 Telegraph Road DNA-TR-89-45 

Alexandria. VA 22310-3398 

RARP/Young 

11 SUPPLEMENTARY NOTES 

This work was sponsored by the Defense Nuclear Agency under RDT&E RMC Code 

B4662D RM RI-I 00038 STRP 3500A 25904D. 

12a DISTRIBUTION/AVAILABILITY STATEMENT 

12b DISTRIBUTION CODE 

Appro\ed for public release: distribution is Unlimited. 

13 ABSTRACT (Maximum 200 words) 

-Based on aailable sources of information, an assessment is made of the Chernob\l accident 

fatalities and causes of death that resulted from acute injury effects. Accident victims grouped 

according to whole body radiation lose and biologic response were examined and reconciled 

based on comparing various sources of information. Fatalities are identified with the occurrence 

of acute radiation sickness (ARS) syndromes. A maximum likelihood regression analysis 

was performed on tWe available fatalit and close data based on five different statistical models 

to estimate the L[ i'; LDi.. and LDO values. The estimated LD 50 is about a factor or two 

higher than previously published values which is attributed to post accident emergency medical 

and clinical care. . 

14 SUBJECT TERMS 15 NUMBER OF PAGES 

- ('hernobyl Accident Radiation Injury - Bone Marrow 34 

Clinical Classification, Lethality Causes, Skin Damage, 16 PRICE CODE 

Radiation Pulmonitis_ Radiation Doses. Fatalities _ _-_ _ 

17 SECURITY CLASSIFICATION 18 SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20 LIMITATION OF ABSTRACT 

OF REPORT OF THIS PAGE OF ABSTRACT 

UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED SAR 

NSN 7540-280-5500 Standard Form 298 (Rev 2-891 

I 

Prec ltw1 by ANSI qla 21PI A 

?.qA 102

SECURITY CL.ASSIFICATION OF THIS PAGE 

CLASSIFIED BY 

N/A ice Unclissified 

DECLASSIFY ON: 

N/A since Uniclassified 

14. SUBJECT TERMS (CONTINUED) 

Intestimal Syndromle 

Acute Effects 

Accession For 

NTIS C-PA&I 

DTIC TAB~ 

SECURITY CLASSIFICATION OF THIS PAGE 

LIN('I.ASSII 11.1)

CONVERSION TABLE 

Conversion factors for U.S. customary 

to metric (SI) units of measurement 

To Convert From To Multiply By 

angstrom meters (nI I. 000 000 X F- I0 

atmo.phere kilo pascal (kPa) 1.013 25 X [+2 

bar kilo pascal (kPa I.000 000 X F+2 

barn meter 2 (iM 2 ) I.000 00(0 X F'-28 

British Thermal unit (thermochemical) joule (.1) 1.054 350 X E+3 

calorie (thermochemicall joule (09) 4. 184 000 

cal (thermochemical ). cm 2 mega Joule/m 2 (M. l 'm 2 ) 4. 184 (000 X [-2 

curie giga becqucrel (GBq)* 3.700 OttO X I+ I 

degree (angle) radian (rad) 1.745 329 X F-2 

degree Fahrenheit degree kelvin (K) 'rK=(I9f + 459.67) 1.8 

electron volt joule (.11 1.602 19 X [-19 

erg joule (.1) 1.0( 0( X 1[-7 

ergisecond watt (\V) 1.000 GOO) X [-7 

foot meter (m) 3.048 0009(1 X F- I 

fool-pound-force joule (.J) 1.355 818 

gallon (U.S. liquid) meter 3 (nm 3 ) 3.785 412 X F-3 

inch meter (t) 2.54(0 00) X [-2 

jerk joule (.J) 1. ( 000 X F+9 

joule kilogram (.1 Kg) (radiation dose 

absorbed) (irav ((i ' 1.000 OtO 

kilotons terajoules 4.183 

kip ((900 hlb newton (N) 4.448 222 X F+3 

kip'inch 2 (kqi) kilo pascal (kPa) 6.894 757 X F+3 

kiap newton-second/rn 2 (N-sm 2 ) I.00t) 0(00 X F+2 

micron meter (n) 1. ((9)0 00)9 X F-6 

mil meter (m) 2.540 (t9(0 X F-5 

mile (international) meter (m) 1.6(09 344 X [+3 

ounce kilogram (kg) 2.834 952 X I-2 

pound-force (119 avoirdupois) newton (N) 4.448 222 

pound-lorce Inch newton-Ometer (N.m) 1. 129 848 X F-I 

pound-lorce, inch nc,%, ,.'nleter (N' m 1.751 268 X [+2 

pound-force/foo 2 kilo pascal (kPa) 4.788 (26 X E-2 

pound-force/inch 2 (PSI) kilo pascal (kPa) 6.894 757 

pound-mass (ibn avoirdupois) kilogram (kg) 4.535 924 X F-I 

pound-mass-foot 2 (moment of inertia) kilogram-meter2 (kg. m 2 ) 4.214 (O1 X [-2 

pound-mass/foot 3 kilogram 'meer3 (kg.'in 3 ) 1.61 846 X F+ I 

rad (radiation dose absorbed) (;ray (Gy) I* ((100 O0t0 X F-2 

roenlgen coulomb..kilogram (C'kg) 2.579 7i99 X F-4 

shake second (q) I,000 000 X [-8 

qlug kilogram (kg) 1.459 390 X F+ I 

torr (mam I1g, 00(I kilo pascal (kPa 1 1.333 22 X F-I 

The tecquerel (0q) is the SI unit of radioaciviH: lip = I eventis. 

S1he (;ra, 

((;y) is the SI unit of absorbed radiation. 

Iil

TABLE OF CONTENTS 

Section 

Page 

CONVERSION TABLE ..................................... 

iii 

FIGURES.............................................. 

TABLES ................................................. 

vi 

1 INTRODUCTION .......................................... 1I 

DATA SOURCES ON CHERNOBYL VICTIMS ............... 

3 CHERNOBYL FATALITIES ................................. 6 

4 UNCONFIRMED CHERNOBYL FATALITIES .................... 8 

5 CAUSES OF FATAL OUTCOMES ............................ 9 

6 FATALITY INCIDENCE VERSUS RADIATION DOSE............14 

7 LIST OF REFERENCES................................... 24

FIGURES 

Figures 

Page 

I 

Chernobyl accident fatalities: normal distribution. 238 individuals. 

90% confidence interval ....................................... 18 

2 Chernobyl accident fatalities; log-normal distribution. 238 individuals. 


3 Chernobyl accident fatalities: Weibull distribution, 238 individuals, 


4 Chernobyl accident fatalities: logistic distribution. 238 individuals. 


5 Chernobyl accident fatalities: log-logistic distribution. 238 individuals. 

90% confidence interval ....................................... 22

TABLES 

Table 

Page 

1 Chernobyl victim folloXw-up reported by State Committee 

119 8 6 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 

2 Chernobyl victim follow-up reported by Guskova 119871 ..... 4 

3 Chernobyl victim follow-up reported by Fry 119871 .......... 5 

4 C hernobyl fatalities .................................... 7 

5 ARS fatalities at Hospital No. 6 in Moscow ................ 10 

6 ARS syndromes identified as causes of fatalities ............ 10 

7 Chernobyl accident fatalities and causes, Hospital No. 6 in 

M o scow .............................................. 12 

8 Radiation doses of individual patients ..................... 15 

9 Radiation doses of individuals and groups ................. 16 

10 Regression m odels ..................................... 17 

11 Fatality percentile doses ................................ 18 

vi

SECTION 1 

INTRODUCTION 

This report reviews current available Soviet syndromes, each of which could have been 

data sources on Chernobyl accident victims an independent cause of a fatal outcome. 

hospitalized for acute radiation sickness All patient groups identified in the data 

(ARS). Patients are grouped by the whole- sources are correlated in terms of the 

body ganmma radiation close indicating the severity of bone marrow syndrome and 

severity of bone marrow syndrome and ascribed causes of fatal outcome. Based on 

biologic response criteria used to place the available data, all but two of the 

patients in the different dose groups. We Chernobyl fatalities can be accounted for 

compare and reconcile the different data according to patient category, close level. 

sources according to the fatalities reported and lethal cause considerations. 

as a result of the Chernobyl accident. 

Included is a brief review of information on 

additional unconfirmed casualties attrib- A maximunm likelihood analysis of the 

uted to the nuclear accident. 

available data on the Chernobyl accident 

survivors and fatalities was also performed. 

We further review the Soviet data sources Five different statistical models were 

in terms of the reported causes of fatal applied to develop estimates of the incioutcomes 

and the occurrence of ARS dence of lethality with total body close. The 

synldromes identified as causes of fatalities, marked effect in survival owing to medical 

In many cases, fatal outcomes are ascribed attention afforded the accident victims is 

to the effect of several severe ARS dernonstrated.

SECTION 2 

DATA SOURCES ON CHERNOBYL VICTIMS 

There are three sources of original data oil Dr. S. A. Fry reports on presentations by 

Chernobyl patients with acute radiation Dr. A. Barabanova, a Soviet physician 

sickness: 

associated with Dr. A. Guskova at the 

Moscow Hospital No. 6. and by Dr. ). P. 

" USSR State Committee on the Utiliza- Orsanov. an internationally know\n biophystion 

of Atornic Energy. (hereinafter icist who went to Chernobyl two to three 

referred to as State Committee). The days after the accident to Sulper\ise 

Accident at tie Chernobvl Nuclear Power environmental monitoring activities. 

Plant And Its Consequences. August 1986 

" Guskova. Earlh' Acute Effects Among the 

Victims of the Accident at the Chernobvl The main clinical response and medical 

Nuclear Power Plunt, April 1987 

outcome data on ARS patients. as reported 

* Fry. Foreign Trip Report. October 1987 

in the three references, are summarized in 

Tables 1. 2. and 3. 

2

co I ri- 

CZ (U 

4)4 

00 

E 

4-- 

" 

E G- 

C 

-= 

L~)U 

0. 

-U) 

0 

U 

_0C 

C> Cl I5 

cz0 

" C 

O~ ~ E c- -0 

C 

0 O di-- I0 

.. 

U 

U 

C7) 

Z 

-0 

0 C> ) 

CC 

CO-z 

O~a~ja.' 

.; 

U 

CZ 0 

.0 - 

0U 

w

-) C) 

4- o.r 

"7Z0 

o o cL; -r; 

Uri I C , I 1- 0() . 

cz 

-l - - I rV 

a 

"C,6 C i 'r 0 C 

00 - 

> UC. 

- I 

I- 

CL- 

ON~l 

ON 

'00 

E ~ ' - 0 

Co 

C z U C 

C> ' cu 

C Z 

0 

0z , 0 

00 

U E 0 

.0 0 

C3 

0 ON L 

0 - C) 

C 

o . c . U 

CU 

rM 

CV) 0 C 

0 co 0 0 

C_ 

-I 

f -r l a u 0 v 

z

Table 3. Chernobyl victim follow-up reported by Fry 119871. 

Chernobyl Fatalities 

Description 

No. of 

Victims 

Initial (on-site and in Pripyat) 2 

Died as result of ARS 

In Moscow 27 

In Kiev 1 

Died as result of CVA 1 

Total deaths 31 

Radiation-induced Skin Injuries. Moscow Hospital No. 6 

Description 

No. of 

Victins 

Number of ARS patients 115 

Radiation-inducel skin injuries 56" 

Skin injuries incompatible with life 19 

Waves of Erythema 

Wave Onset Remarks 

1st 36 h No obvious skin injury for 36 h. 

First wave lasted up to 24 h. 

2d 5-10 days Erythema more widespread. 

later 

Involved skin areas covered by 

clothing. Coincided with latent 

period of ARS. 

3d 2-3 months Included 28-30 patients. 

8-10 had no previous manifestation 

of skin injury. 

"T\\o of 56 patients also had thermal burns. 

5

SECTION 3 

CHERNOBYL FATALITIES 

The number of fatalities from the Cherno- Guskova did not include the CVA fatality 

byl accident totals 31. Table 4 shows the among the fatalities reported. Only one 

reported nLumbers of death and reconciles fatality is listed for Kiev in Table 2. whereas 

the different data sources. There were two two fatalities are listed in Table 1. 

initial fatalities: one victim died at his 

workplace. his body was never recovered: 

another victim died from severe burns at 

6:00 a.m. in the morning of the accident in 

Dr. Barabanova further reported that the 

CVA victim was among the fatalities at 

the Pripyat hospital. State Committee Hospital No. 6 in Moscow lFri. 19871. This 

119861 reports 28fatalities as occurring piece of information appears to be in error. 

within the first 50 days (Table 1). Three State Committee 119861 reports that two 

people subsequently died from ARS on patients died in Kiev on days four and ten 

Days 77. 96. and 91 (Table 2). resulting in after the accident. Guskova 119871 shows 

a total of 31 fatalities, 

only one death in Kiev. a patient who died 

on the 10th day from combined heat and 

Dr. Barahanova reported that one transfer radiation injuries. The transfer patient who 

patient from Chernobyl. with signs and died shortly after hospital admission from 

symptorns of acute radiation sickness, CVA must have been the Kiev fatality on 

suffered a cerebrovascular accident (CVA) day four. The early death in Kiev has been 

enroute and (lied from that cause shortly confirmed by 'Deputy USSR Minister of 

after admission to the hospital IFry. 19871. Health, Ye. Vorobyev. who told a press 

Hence. the Chernobyl fatalities include 2 conference on 9May1986 that a third 

initial fatalities. 28 ARS fatalities (27 died person, the first victim of radiation 

in Moscow. and I in Kiev), and 1CVA exposure. died three days after the accident 

fatality (Table 3). In Table 4. notice that in a Kiev hospital IGoure. 19871. 

6

- - 0 

LL v 

0 

4 

AC. 

o0 

- I C 

CC3 

0 

EE 

- U 

(U 

-'~c 

-Co0 

-C13 

1-0 

r 

C 

QU( 

0~ 

o~ o M 

7.

SECTION 4 

UNCONFIRMED CHERNOBYL FATALITIES 

According to Dr. Barabanova's presenta- there (ten or so meters away) is roughly 

tion. Fry lOctober 19871 reports the 1000 roentgens. ' ' Later, in referring to Y. 

following: "A journalist/photographer who Velikhov, Vice President of the USSR 

spent some time at the accident site shortly Academy of Sciences, seen walking in front 

after the accident is reported to have died of the camera, the reporter adds: "A little 

about 9 months later: the cause of his death bit earlier in the film. we were told why the 

is controversial according to Dr. Barabano- fam,,ous scientists had to risk their lives. 

va who did not elaborate further." 

What about the invisible cameraman? The 

other mnembers of the filming crew?" 

The Moscow weekly Nedeh*'a cited Noioye 

Russko*ve Sloro of 31 May 1987 on the death Three Soviet sources appear to agree on at 

of the motion picture director Vladimir least one additional Chernobyl fatality. The 

Shevchenko two months earlier frorn Moscow News highlights the radiation risk 

radiation sickness IGoure. 1987a1. Shev- to the camera crew. Ncdelva reports the 

chenko produced the documentary. "Cher- death of the movie director and the hospinobyl. 

Chronicle of Difficult Days." which talization of two cameramen. Dr. Barabawvas 

filmed in May 1986 at the Chernobyl nova mentions the reported but controverpow\er 

station. Nedelha also reported that sial death of a journalist/photographer. 

two cameramen received large radiation 

doses and are Itheni currently hospitalized. Non-Soviet sources report additional fatalities 

and radiation injuries among the 

Tie documentary filn on the Chernobyl cleanup crews according to Radio Free 

accident has been described in the Moscow Europe "Rad Background Report." 14 

News No. 11. 22-29 March 1987 IGoure, October 1986 IGoure, 1987bl. At the same 

1987al. Readers are left with the irnpres- time, official Soviet spokesmen deny that 

sion that a camera and its crew are taken more people have died from radiation 

on a tour through the damaged reactor, overdoses at the Chernobyl atomic power 

often close to high levels of radiation by the plant since the accident there last year 

plant's chief engineer, who says, "That pipe killed 31 people [Los Angeles Times. 19871. 

I. Prohably refers to a dose rate of I(100 R/h. 

8

SECTION 5 

CAUSES OF FATAL OUTCOMES 

All three Soviet data sources discuss the 

importance of radiation damage to skin and 

basically repeat the same message: 

dromes, each of which could have been an 

independent cause of a fatal outcome. 

These two groups total 19 fatalities, which 

is the number of cases with severe skin 

injury commonly cited as being incompat- 

ible with life. Apparently, the remaining 

eight fatalities were caused by combina- 

tions of various syndrornes without radi- 

ation skin damage being an independent 

"In 19 cases, the deaths of patients with 

third- and fourth degree acute radiation 

disease occurred only as a result of 

severe damage, which was incompatible 

with survival, to 50-90% of the surface 

of their bodies." IState Committee, cause of the fatal outcomes. 

19861 

-"n the cases of at least 19 of the 56 Various ARS syndromes reported by 

patients 

paint ses ith aturst the ns we Guskova 119871 as apparent independent 

with burns, the burns were causes of fatal 

unquestionably 

outcomes 

fatal." 

(or being 

IGuskova. 

capable 

1987j of producing a fatal outcome) are listed in 

"Of the 27 patients who died of Table 6. As mentioned before, radiation 

radiation injuries at H~ospital No. 6, 19 skin damage was the exclusive or contributhad 

radiation induced skin injuries over ing cause in 19 fatal outomes. Lethal 

90-100% Isicj of their body surface. It intestinal syndrome, indicated by the 

wvas the Soviets's opinion that in these appearance of diarrhea from day four to 

cases, the skin injuries were so severe clay eight, was noted in 10 patients. These 

as to be considered incompatible with patients were exposed to 10 Gy or higher 

life. even in the absence of bone gamma radiation- they all died within the 

marrow danmage." IFry, 19871 first three weeks after exposure. 

All these statements appear to be true but Acute radiation pulmonitis was observed in 

incomplete. Guskova 119871 discusses the seven patients with third- and fourth-degree 

interrelationships between individual caus- ARS. It was typified by a rapidly intenes 

of lethal outcomes in somewhat more sifying difficulty in breathing, by ventiladetail, 

summarized in Table 5. Although all tion failure, and by the onset of lethal 

patients with third- and fourth-degree hone- outcomes from hypoxernic coma. "Autopsy 

marrow syndrome had severe radiation revealed large blue lungs with marked 

burns (Table 2), only five lethal outcomes interstitial edema."[Guskova. 19871 

could be ascribed exclusively to radiation 

dan,age to vast areas of skin. These five Six cases with lethal outcomes \\ere 

cases did not involve radiation enteritis or ascribed to radiation damage causing 

irreversible myelodepression. and their irreversible hematopoietic aplasia or to 

whole-body (loses did not exceed 6 Gy. In complications caused by bone marrow 

14 other cases. radiation damage to skin transplants. Hemophilia achieved thanato- 

\\as combined with other severe syn- genetic significance only in one case. 

9

Table 5. ARS fatalities at Hospital No. 6 in Moscow. 

Causes of Fatality 

Number of 

Fatalities 

Radiation skin damage exclusively 5 

Skin damage combined with other 

ARS syndromes 14 

Combinations of ARS syndromes other 

than skin damage 8 

Total fatalities 27 

Table 6. ARS syndromes identified as causes' of fatalities. 

ARS Syndrome 

Fatalities 

with Syndrome 

Radiation skin damage 19 

Intestinal syndrome (10 Gy or more) 10 

Acute radiation pulmonitis 7 

Irreversible hematopoietic aplasia or 6 

bone-marrow transplant complications 

Iemophilia 1 

Thermal burns/internal contamination 2 

Radiation induced vascular damage 

I 

10

In Moscow. 56 of the 115 ARS patients -Patient injury or medical treatment 

suffered from radiation induced skin category. 

injuries. Two of the 56 patients formed a 

unique subgroup: they also had severe Patient categories include one cerebrovasthermal 

skin burns produced by steam and cular accident, a patient in the medium 

incurred a significant internal radionuclide gamma dose group who died from lethal 

close estimated as 1.5 to 4 sieverts on the radiation clanage to the Vascular system. 

basis of postmortem radiometry. The radio- Five fatal outcomes are ascribed solely to 

nuclide (loses of all other patients did not radiation damage to large skin areas. These 

exceed 1 to 3 percent of their external close, patients fall in the severe gannia dose 

The gamma dose of these two patients was group: their estimated whole-body closes 

on the order of 4 to 5 Gy. The patients died did not exceed 6 Gy. 

on days 23 and 18, respectively, from their Further. patient groups ilude bone 

cornbined injuries jFry. 19871 . marrow transplants and fetal liver cell 

transplants. These patients wvere selected 

One patient in NIoscoWv suffered a fatal on the basis of a whole-body close of 6 (\ 

CVA 2 . probably associated with general- or greater, estimated from the lym plocyte 

ized radiation induced vascular damage count in peripheral blood and from cyto- 

I Fry. 19871. The patient received an genetic analvsis of chromosome aberraaverage 

bone marrow close of 3 Gy and tions. This close level \\as Understood to 

suffered severe radiation induced skin represent irreversible or extremely proinjuries. 

manifesting as three waves of tracted and deep myelodepression. 

erythema. with the third wave developing 

almost three months after the accident. The bone marrow transplants include 13 

Accordingly. this patient (case 3 presented patients grouped in two subcategories: 

b ini the Dr. second-cegre Barabanova) 

e 

must 

ARS 

be 

group 

the one fatality 

\vlo died a. Six patients had radiation damage to 

on day 96 after the accident (Table 2). 

skin and intestines at a level that was 

deemed not incompatible \\ith life. 

Four patients died between days 27 

Based on currently available Soviet data 

and 29 after the transplants. Two 

sources reviewed. Table 7 provides a 

sumnmary of our effort to correlate the 

patients survived their bone marrowv 

Chernobvl accident fatalities and causes transplants; they had gamma expo- 

terms of: 

sures sin of 5.8 and 9.0 Gy. Gamnia 

grouped ndoses 

for the four patients who died 

ranged from 4.3 to 10.7 Gy, placing 

* Severity of the bone marrow syndrone one fatality in the severe and three 

(the whole-body gamma radiation fatalities in the extremnelv severe 

doses). 

bone marrow syndrome groups. They 

(lied from "mixed virus-bacterial 

* ARS syndrome identified as exclusive infections." also called as "cornplior 

contributing cause of fatal outcomes. cations caused by bone marrowv 

2. \ote thal thi ('\' A ltalit v occurred in \1l co\%: a different ARS patient suffered a (''A enroutle 

to the ho,,pilal and died on cla\ ltour in Kiev. 

11

0 C 

C: 

- I I I 

co 

m 

C 

0 , 

o -- 

-) 0 : r-% 

CO2v 2C 5W 

C6- 

C. I C I 

4,) I I I 

rs X-

transplants" jGuskova, 19871. State "unspecified" category. The totals of 

Committee 119861 reports that all six individual lethal causes have been taken 

bone marrow transplants in this fron Table 6. Adding the numbers of 

group had similar disorders that may patients in relevant patient categories, the 

have been caused by graft-ver- corresponding totals are not always identisus-host 

(GVH) disease. Two of tihe cal with those in Table 6. but are still in 

four fatalities ill this group occurred general agreement. For example. we would 

prior to the puiblicationi of the State have 13 rather than 10 cases of lethal 

Committee 119861 report. It is stated radiation damage to intestines. In cases 

that the GVH reactions may have \\here the fatal outcomes associated with 

been contributing causes of these two a patient category are ascribed to multiple 

deaths. contributing ARS syndronmes ("yes" in 

Table 7), all the syndromes may not 

b. Seven patients had radiation damage necessarily be present for each and every 

to their skin. intestines, and lungs at fatality. 

a level that was deemed incompatible 

with life. They died from acute In other cases. the relevant patient cateradiation 

clamage to the skin. gut. gories identify fewer fatalities than the 

and lungs on days2 to 19 after the totals of Table 6. For example, only four 

translplants (days 15 to 25 after rather than six total cases of lethal inexrposure). 

fections and GV1I- diseases are identified 

directly. In these cases. the question marks 

In the category of fetal liver cell trans- in Table 7 suggest other possible patient 

plants, all six patients died from radiation categories associated with a particular 

darnage to the skin and intestines IGusko- cause of death. 

a. 19871. Patients for the transplantation 

of human embryo liver cells were selected The individual gamma closes of the accion 

the basis of extrernely severe clarnae dent victims are generally unavailable. 

to the skin and intestines and extremely Exceptions are the cerebrovascular acciunfavorable 

prognosis IState Committee dent victim with anl estimated close of 3 G\. 

19861. the four bone marrow transplant fatalities 

who died from mixed Virus and bacterial 

Two patients in the thermal burns and infections with estimated closes of 4.3. 

internal contamination category died from 5.0-7.9. 5.8-6.0. and 10.7 Gy, respectively. 

their combined injuries. including bone and the two thermal burn and internal 

marrow damage and racliation damage to contamination patients with their combined 

skin. closes shown in Table 7. Guskova 119871 

also states that the patients with lethal 

In Table 7 all but two of the 27 fatalities intestinal syndromes received a short-term 

in Moscow can be traced through patient whole-body gamma radiation dose on tihe 

categories, thus leaving only two in the order of 10 Gy or higher. 

13

SECTION 6 

FATALITY INCIDENCE VERSUS RADIATION DOSE 

A total of 238 Chernobyl accident victims 7). Four indiViduals in the fourth close 

received estimated whole-body garnma group have specific close estimates. but no 

radiation closes in excess of 0.8 Gy. They specific values were available for the 

were diagnosed as suffering from acute remaining five individuals. 

radiation sickness (ARS). Based on the 

severity of ARS. the patients were grouped Based on the method of analyzing binary 

in four close range categories. The radiation data given by Cox (1983). a maximum 

data published to date include the dose likelihood analysis of the data given in 

ranges and the number of patients and Table 9 was performed to estimate the 

deaths in each (lose range group. 

incidence of Chernobyl fatalities with close. 

Along with the explicit close estimates given 

Although individual case histories have not in Table 9 for both surVivors and fatalities. 

been published, individual dose estimates it 'vas assumned that the remaining indihave 

been reported for some cases. includ- vidual doses were uniformly distributed 

ing II survivors and 7 fatalities. Table 8 over the corresponding lose range. For 

lists individual loses and reference fol- example, there are five fatalities in the close 

lowed by a number or letter of patient range of 4.2 to 6.3 Gy listed under the 

identification. The numbers associated with "3rd- close group in Table 9, thus. for this 

Guskova 119871 refer to the original patient analysis. closes of 4.41, 4.83. 5.25. 5.67. 

identification system: the letter "x" refers and 6.09 Gy each \\ere assumed for the five 

to patients mentioned but not otherwise fatalities. 

identified in reports. Some of the individual For purposes of comparison, five different 

closes reported by Guskova 

been updated in a 

119871 

later report 

have 

(UNSCEAR. 

model 

moefrswrealidtanyzth 

forns were applied to analyze the 

been updatled i oa ate pted NSC . survivor/fatality binary data: normal. log- 

1987): Table 8 contains the updated values, normal. Weibull. logistic, and log-logistic. 

Table 9 sunmarizes the available data on Table 10 gives the model forms along with 

radiation closes for both individuals and parameter estimates obtained. Based upon 

groups of individuals. In most cases, the the X 2 goodness-of-fit statistic, a ranking 

patients in a dose group divide into two would fall between the Weibull and logsubgroups: 

a few patients with individual normal models, with Weibull the best fit. 

dose estimates. and the remainder distrib- However, all models fit the data well 

uted over the given (lose range. The (p

Table 8. Radiation (loses of individual patients. 

Dose Number of Survivors Fatalitie 

Dose 

Range 

Group (G%) Patients Deaths Dose Patient Reference Dose Patient Relerence 

Ist 0.8-2. I 140 0.9 Guskova 119871 - 97 

1.4 - 48 

1.9 Geiger 119861 - 6 

2nd 2.0 -4. 55 I 3.3 Guskova I19871 - 39 3 FrY 119871 - 3 

3.9 - 21 

3.3 State Committee 119861 - D 

2.25 Fr\ 119871 - 2 

3.0 Telvatniko% 119871 

3rd 4.2-6.3 21 7 5.6 Guskoxa 119871 - x 5.2 (uskoka 119871- 6 

4.9 Geiger 119861 - 2 4.4 - 5 

4th 6-16 22 21 8.7 Guskova 119871 - x 6.4 c;uskoa 1 19871- 28 

1.2 - 16 

9 -8 

7 Fry 119871 - x 7 

Tomals 238 29 II patients 7 patients 

15

C13l 

I)" 

- 0- 

Cu 

oo 

o 

C 

IIV 

0~0 

0 0l 

I-c 

U. c 

01

C) 00 00 kfs v-s 

(- I 1r) ON C0 

\C r- r "0 r 

00 

ON 

C 

C 

E -0 - \- 

C0ON0 

CCO 

5-j 

co "0 

5- 

oO -a u 

E 

+ 

, 6 

" 

'1)z 

a 7) 

cu 11 

Vt re- 

E' r 

00 

17-

Table 11. Fatality percentile doses. 

Percentile Dose (cGy) 

slope* 

Model LD 1 () LD 50 LD 9 (? (percent/cGy) 

Normial 409.9 616.4 822.9 0.194 

Log-normnal 391.4 592.0 895.3 0.159 

WeibUtl 401.0 631.4 843.3 0.181 

Logistic 422.8 606.0 789.3 0.218 

Log-logistic 403.3 590.1 863.5 0.174 

*slope L 90 -1 D) (percent/c~y) 

99.9 

Prior non-Chernobyl estimates 

99.0- (See text) 

95.0-Z 

90.0- 

S70.0- 

.. . . . . . 

5~ 0.0 - ...... 

10.0 -...... 

0... .... 

1.0...... 

.... 

0.1..... 200.00..0.80.100 

... Dose...... 

Figre1.Clerobl ccdet atliie: nrml isriuton.28 nd.iuas.90 

1.08

99.9 


99.0- (See text) 

95.0- 

90.0- 

70.0- 

50.0 

V30.0- 

3000 402507010 

Dose (cGy) 

Figure 21. Chernobyl accident fatalities, log-normial distribution, 238 individuals, 90% 

confidence interval. 

19

99.9 

99.0 Prior non-Chernobyl estimates - 

90.0 

10.0 

CD 5.0 - 

70.0- 

50.0- 

-~30.0- 

1.0- 

01200 300 400 500 700 1000 

Dose (cGy) 

Figure 3. Chernobyl accident fatalities: Weibull distribution. 238 indlividuals, 90%7c 


20

99.9 


_70.0- 

99.0- 

95.0- 

90.0- 

S50.0- 

o30.0- 

m ...... 

5.0 

0.1 200 400 600 800 1000 

Dose (cGy) 

Figu~re 4. Chernobyl accident fatalities: logistic distribution, 238 individuals. 90% 


21

99.9 


(See text) 

99.0- 

95.0- 

90.0- 

= 70.0- 

a 50.0- 

v30.0- 

10.0 

5.0- 

1.0 ' 

200 300 400 500 700 1000 

Dose (cGy) 

Figure 5. Chernobyl accident fatalities: log-logistic distribution, 238 individuals, 90% 


22

five different models used to analyze the the LD 5 () based on prior estimates with that 

data: the 90% confidence bounds are given estimated here from the Chernobyl data. 

by the two dashed lines. In order to That is on average. prior estimates of the 

graphically illustrate the data from Table LD 50 wou*d be about 290 cGy whereas the 

9 used in fitting the various models, the value is about 607 cGy according to our 

data were grouped according to the analysis of the Chernobyl data: Ilis 

horizontal lines shown in the plots. The amounts to about a factor of two reflected 

data used in the maximum likelihood by medical care. 

procedure is treated in binary form. and the 

asterisks on the horizontal line simply Also, as is apparent from tile plots of 

indicate the dose range midpoints of the fatality incidence versus close, in general 

data groups. Accordingly, while the log- there is a significant difference in the slope 

normal (Fig. 2) and log-logistic (Fig. 5) magnitude betwveen the Chernobyl dose 

models appear by inspection to fit the data response curves and that according to prior 

better than the other models, they actually estimates of lethality. As indicated in Table 

have the lowest X 2 values (see Table 10) 11. depending upon the model, slope 

although as indicated above, all five of the estimates based oil analysis of the Chernomodels 

fit the data well. byl data range from about 0.16 to 0.22 

percent/cGy: the corresponding slope value 

A range of incidence for radiation lethality from prior lethality estimates is about 0.44 

based on some prior estimates is also percent/cGy. This difference in the steepindicated 

in Figs. I through 5 by the broad ness of the close response curves could also 

shaded band. The left-most boundary is be Clue in part to medical attention. Albased 

oil an LD 5 () midline tissue close of though more than likely, it is due to factors 

256 cGy and a slope estimated from close other than strictly ARS in combination \\ith 

response curves for large animals (swine. medical attention such as injuries to skin. 

sheep, goats. and dogs) which tend to be which as mentioned above \was a ver\ 

parallel (Bond and Robertson 1957, Cron- significant factor. For example. in the abkite 

and Bond 1958. and Cronkite 1982). sence of medical attention, a larger propor- 

The right-most boundary is based on an tion of lethalities Would be expected at 

1I) 5 ) value of 325 cGy. suggested by lower doses. i.e.. in the second and third 

Lushbaugh (1969) for healthy young close groups. This certainly Would tend to 

adults, with the slope also inferred frorn steepen the dose response curves. lipowevlarge 

animals. 

er, because of the imprecision of data 

regarding burn injuries and other non-ARS 

One means of assessing the efficacy of effects, we have not attempted to perform 

medical care including the various treat- any isolated causal analysis with regard to 

ment modalities of ARS afforded the fatality incidence utilizing truncated sub- 

Chernoby I accident victims is by comparing sets of' data. 

23

SECTION 7 

LIST OF REFERENCES 

Bond, V. P. an(l J. S. Robertson, Ann. Rev. Science Applications International Corp. 

Nuc. Sci., Vol. 7, pp. 135-162. 1957. La Jolla, California, October 1987a. 

Cox, D. R.. Analv'sis of Binary Data, --.- "Medical," The Chernobvl Accident 

Chapman and Hall Ltd., New York, NY, Data Base, Sec. VIII, Science Applica- 

1983. tions International Corp. McLean, Virginia, 

October 1987b. 

Cronkite. E. P., "The Impact of Estimates 

of Human Radiation Tolerance upon Guskova, A. K., Eary Acute Effects among 

Radiation Emergency Management," the Victims of the Accident at the Chernob\l 

Proceedings of a Symposium on the Control Nuclear Power Plant. Ministry of Health 

of Exposure of the Public to Ionizing of the USSR, LN-818-87, April 1987. 

Radiation in The Event of Accident or Los Angeles Times. "The World," Part I. p. 

Attack. National Council on Radiation 2. December 10, 1987. 

Protection and Measurement, Bethesda, 

Maryland, May 1982, pp. 21-27. 

Lushbaugh. C. C.. "Reflections on Some 

Recent Progress in Humnan Radliobiolo- 

Cronkite. E. P. and V. P. Bond. "Acute Re Advaes i n Rad o.ol. 

radiationgy" Advances in Radiation iology. Vol. 

Frceosyr J.. Vol. 9.19. A d 3. Academic Press, New York, 1969. pp. 

Forces Med. J.. Vol. 9. 1958. pp. 277-315. 

313-324. 

Telyatnikov. L. P.. Personal interview with 

Fry. S. A., Trip Reort: International Atomic the Fire Brigade Commander, Novemn- 

Energy Agency (IAEA) Workshop on the ber 1987. 

Aledical Handling of Skin Lesions Follotiing 

High Level Accidental Irradiation, UNSCEAR (United Nations Scientific 

Institute Curie, Paris, France, Center for Committee on the Effects of Atomic 

Epidemiologic Research, Medical and Radiation), A/AC.82/R. 473, Annex H, 

Health Services Division, Oak Ridge Early Effects in Man of High Doses of 

Associated Universities, September 28- Radiation; Appendix 1, "Acute Radi- 

October 2, 1987. 

ation Effects in Victims of the Chernobyl 

Nuclear Power Plant Accident." A. G. 

Geiger. H. Jack. The Accident at Chernobvl Guskova, ed., 10 November 1987. 

and the Medical Response, Journal of 

American Medical Association, Vol. USSR State Committee on the Utilization 

256. No. 5. pp. 609- 612. August 1, of Atomic Energy. "The Accident at the 

1986. Chernobyl Nuclear Power Plant And Its 

Consequences: Annex 7, Medical and 

Gotre, Leon, "Radiation Situation at the Biological Problems," trans. by IAEA. 

AES and in the Danger Zone," The IAEA Experts Meeting, Vienna, Austria, 

Chernobyl Accident Data Base, Sec. V, August 25-29, 1986. 

24

DISTRIBUTION LIST 

DNA-TR-89-45 

DEPARTMENT OF DEFENSE 

ARMED FORCES RADIOBIOLOGY RSCH INST 

ATTN: AFFRI-DEPT OF RAD BIOCHEM 

ATTN: BHS 

ATTN: DIRECTOR 

EXH 

ATTN: MRA 

ATTN: PHY 

3 CYS ATTN: FCPRT 

ATTN: NUC SECURITY 

INTELLIGENCE CENTER, PACIFIC 

ATTN: COMIPAC 

INTERSERVICE NUCLEAR WEAPONS SCHOOL 

ATTN: TTV 

2 CYS ATTN: TTV 3416TH TTSQ 

ATTN: RSD 

JOINT DATA SYSTEM SUPPORT CTR 

ATTN: SCIENTIFIC DIRECTOR ATTN: C-332 

ATTN: TECHNICAL LIBRARY 

ATTN: JNSV 

ASSISTANT SEC OF DEF (C3) 

ATTN: DIR (S&TNFC3) 

ASSISTANT SECRETARY OF DEFENSE 

INTERNATIONAL SECURITY POLICY 

ATTN: NUC FOR & ARMS CNTRL PLCY 

ASSISTANT TO THE SECRETARY OF DEFENSE 

ATTN: EXECUTIVE ASSISTANT 

ATTN: MIL APPL 

ATTN: MIL APPL,C FIELD 

NATIONAL DEFENSE UNIVERSITY 

ATTN: ICAF, TECH LIB 

ATTN: NWCLB-CR 

ATTN: LIBRARY 

ATTN: STRAT CONCEPTS DIV CTR 

NATIONAL SECURITY AGENCY 

ATTN: CHIEF A4213 

NET ASSESSMENT 

ATTN: DOCUMENT CONTROL 

DEFENSE INTELLIGENCE AGENCY 

OFFICE OF THE SEC OF DEFENSE 

ATTN: DB 

ATTN: NAVAL FORCES 

5 CYS ATTN: DB-, RSCH RESOURCES DIV ATTN: STRATEGIC PROG & TNF 

ATTN: DB-5C 

ATTN: DB-6B 

STRATEGIC AND THEATER NUCLEAR FORCES 

ATTN: DIA/VPA-2 (FED RES DIV) 

ATTN: DR E SEVIN 

ATTN: DN 

ATTN: DR SCHNEITER 

ATTN: DT 

ATTN: OFFICE OF SECURITY 

STRATEGIC TARGET PLANNING 

ATTN: OS 

ATTN: JKC (ATTN: DNA REP) 

ATTN: RTS-2B 

ATTN: JLT 

ATTN: JP 

DEFENSE INTELLIGENCE COLLEGE 

ATTN: DIC/RTS-2 

ATTN: JPTP 

AITN: DIC/2C 

THE JOINT STAFF 

DEFENSE LOGISTICS AGENCY 

ATTN: ED30 

ATTN: J-3 SPECIAL OPERATIONS 

ATTN: COMMAND SECURITY OFC 

ATTN: J-3 NUC OPERATIONS BR 

DEFENSE NUCLEAR AGENCY 

ATTN: CD 

ATTN: J-5 NUC & CHEMICAL DIV 

ATTN: JAD/SFD 

ATTN: DFRA 

ATTN: JAD/SSD 

ATTN: DFSPG ULLRICH 

ATTN: JSOA 

ATTN: NANF 

ATTN: SAGA 

ATTN: NASF 

U S EUROPEAN COMMAND/ECJ-3 

ATTN: OPNA 

ATTN: ECJ-3 

ATTN: OPNS 

ATTN: RAEE 

U S EUROPEAN COMMAND/ECJ-5NPG 

20 CYS ATTN: RARP ATTN: ECJ-5 

4 CYS ATTN: T;TL 

U S EUROPEAN COMMAND/ECJ-6-OJ 


ATTN: ECJ-6 

ATTN: TDNM 

2 CYS ATTN: TDTT W SUMMA U S EUROPEAN COMMAND/ECJ2-T 

ATTN: ECJ2-T (TGTS DIV) 

DEFENSE TECHNICAL INFORMATION CENTER 

2 CYS ATTN: DTIC/FDAB U S EUROPEAN COMMAND/ECJ4-LW 

ATTN: ECJ4-LW 

FIELD COMMAND DEFENSE NUCLEAR AGENCY 

ATTN: FCPR 

Dist- 1

DNA-TR-89-45 (DL CONTINUED) 

U S EUROPEAN COMMAND/ECJ5-N 

ATTN: ECJ5-N (NUC BRANCH) 

U S EUROPEAN COMMAND/ECJ7/LW 

ATTN: ECJ-7 LW 

UNDER SEC OF DEFENSE FOR POLICY 

ATTN: DUSP/P 

ATTN: USD/P 

U S ARMY CONCEPTS ANALYSIS AGENCY 


U S ARMY FIELD ARTILLERY SCHOOL 

ATTN: ATSF-CD 

U S ARMY FORCES COMMAND 

ATTN: AF-OPTS 

U S ARMY HUMAN ENGINEERING LAB 


ATTN: DR D HODGE 

UNDER SECRETARY OF DEFENSE 

ATTN: CHAIRMAN PSEAG 

ATTN: CHAIRMAN, DEF SCIENCE BD 

2 CYS ATTN: C31 U S ARMY INFANTRY CENTER 

ATTN: ATSH-CD-CSO 

DEPARTMENT OF THE ARMY 

U S ARMY ITAC 

ARMY RESEARCH INSTITUTE 

ATTN: COMMANDER 

COMBAT MATERIAL EVAL ELEMENT 

ATTN: SECURITY ANALYST 

ATTN: IAX-Z 

U S ARMY MATERIEL COMMAND 

ATTN: DRCDE-D 

ATTN: DRCSS 

DEP CH OF STAFF FOR OPS & PLANS 

U S ARMY NUCLEAR & CHEMICAL AGENCY 

ATTN: DAMO-FDQ 

ATTN: MONA-NU 

ATTN: DAMO-NCN ATTN: MONA-ZB 

ATTN: DAMO-NCN(NUC CHEM DIR) 

U S ARMY TRAINING AND DOCTRINE COMD 

ATTN: DAMO-ZXAATN 

ATTN: ATCD-AO CDO 


ATTN: ATCD-FA 

ATTN: DAMO-NCZ 

ATTN: ATCD-N 

ATTN: DAMO-OD 

ATTN: ATOO-NCO 

ATTN: DAMO-ODSO 

ATTN: DAPE-ODRE 

U S ARMY WAR COLLEGE 

ATTN: DAPE-HREATNLIRY 

ATTN: SARD-ZCS (MAJ ADAMS) 

ATTN: LIBRARY STRATEGIC STUDIES 

HARRY DIAMOND LABORATORIES 

ATTN: SLCIS-IM-TL (TECH LIB) 

JOINT SPECIAL OPERATIONS COMMAND 

ATTN: J-2 

ATTN: J-5 

U S ARMY AIR DEFENSE SCHOOL 

ATTN: COMMANDANT 

U S ARMY ARMAMENT RSCH DEV & ENGR CTR 

ATTN: DRDAR-LCN-F 

U S ARMY ARMOR SCHOOL 

ATTN: ATSB-CTD 

ATTN: TECH LIBRARY 

US ARMY MATERIEL SYS ANALYSIS ACTVY 

ATTN: DRXSY-DS 

USA MILITARY ACADEMY 

ATTN: DEPT OF BEHAV SCI & LDRSHP 

ATTN: DIR NATL SECURITY STUDIES 

USA SURVIVABILITY MANAGMENT OFFICE 

ATTN: SLCSM-SE J BRAND 

USACACDA 

ATTN: ATZL-CAD-N 

DEPARTMENT OF THE NAVY 

FLEET INTELLIGENCE CENTER 

U S ARMY BALLISTIC RESEARCH LAB ATTN: FICPAC, CODE 21 

ATTN: AMXBR-VLD-R (DR RAINIS) 

MARINE CORPS 

ATTN: AMXBR-VLD, DR KLOPCIC 

ATTN: CODE PPO 

ATTN: DRDAR-BLA-S, TECH LIB 

ATTN: PSI G/RASP 

ATTN: DRDAR-BLT 

ATTN: SLCBR-D 

MARINE CORPS DEV & EDUCATION COMMAND 

U S ARMY CHEMICAL SCHOOL 

ATTN: LIBRARY 

ATTN: ATZN-CM-M 

NAVAL OCEAN SYSTEMS CENTER 

U S ARMY COMD & GENERAL STAFF COLLEGE ATTN: CODE 9642-B 

ATTN: ATZL-SWJ-CA 

NAVAL PERSONNEL RES & DEV CENTER 

ATTN: ATZL-SWT-A 

ATTN: CODE P302 

Dist-2


NAVAL POSTGRADUATE SCHOOL 

ATTN: CODE 1424 LIBRARY 

NAVAL RESEARCH LABORATORY 

ATTN: CODE 1240 

ATTN: CODE 2627 (TECH LIB) 

NAVAL SEA SYSTEMS COMMAND 

ATTN: PMS-423 

ATTN: SEA-09G53 (LIB) 

ATTN: SEA-6631 

NAVAL SURFACE WARFARE CENTER 

ATTN: CODE F-31 

ATTN: G GOO 

NAVAL TECHNICAL INTELLIGENCE CTR 

ATTN: NTIC-DA30 

NAVAL WAR COLLEGE 

ATTN: CODE E-11 TECH SVC 

ATTN: CTR FOR NAV WARFARE STDS 


ATTN: LIBRARY 

STRATEGY DEPT 

NAVAL WEAPONS EVALUATION FACILITY 

ATTN: CLASSIFIED LIBRARY 

AIR UNIVERSITY LIBRARY 

ATTN: AIJL-LSE 

ATTN: LIBRARY 

ASSISTANT CHIEF OF STAFF 

2 CYS ATTN: AF/SAMI 

ASSISTANT CHIEF OF THE AIR FORCE 

ATTN: SAF/ALR 

DEPUTY CHIEF OF STAFF/XOX 

ATTN: AFXOOSS 

ATTN: AFXOXFM 

FOREIGN TECHNOLOGY DIVISION, AFSC 

ATTN: CCN 

ATTN: SDA 

SPACE DIVISION/YH 

ATTN: YH 

STRATEGIC AIR COMMAND/ADWN 

ATTN: ADWN 

STRATEGIC AIR COMMAND/SPD 

ATTN: 

SPD 

STRATEGIC AIR COMMAND/STIC 

ATTN: 544 SIW/DI (STIC) 

NUCLEAR WEAPONS TNG GROUP, ATLANTIC 

STRATEGIC AIR COMMAND/XOXO 

ATTN: CODE 222 AIR COM / 


ATTN: XOXO 

NUCLEAR WEAPONS TNG GROUP PACIFIC 

STRATEGIC AIR COMMAND/XPX 

ATTN: CODE 32 

ATTN: XPZ 

ATTN: DOCUMtNT ;ONTROL STRATEGIC AIR COMMAND/XRFS 

OFFICE OF CHIEF ir r .,VAL OPERATIONS ATTN: XRFS 

ATTN: NIS-22 

TACTICAL AIR COMMAND/XPSC 

ATTN: NOP06D 

ATTN: TAC/DOA 

2 CYS AT TN: N0P403 

ATTN: NOP 50 

U S AIR FORCE ACADEMY 

ATTN: NOP60 

ATTN: LIBRARY 

ATTN: NOP60D 

ATTN: USAFA/SP 

ATTN: NOP 603 

ATTN: NOP 91 

USAF SCHOOL OF AEROSPACE MEDICINE 

ATTN: OP 654 

ATTN: RADIATION SCIENCES DIV 

OPERATIONAL TEST & EVALUATION FORCE 


PLANS, POLICY & OPERATIONS 

ATTN: CODE-P 

ATTN: CODE-POC-30 

TACTICAL TRAINING GROUP, PACIFIC 


DEPARTMENT OF THE AIR FORCE 

AFIA/INIS 

AFIS/INT 

ATTN: 

ATTN: AFIA/INKD 

INT 

AIR UNIVERSITY 

ATTN: STRATEGIC STUDIES 

WEAPONS LABORATORY 

ATTN: WL/SUL 

DEPARTMENT OF ENERGY 

LAWRENCE LIVERMORE NATIONAL LAB 

ATTN: PAUL THERMAN 

ATTN: Z DIVISION LIBRARY 

LOS ALAMOS NATIONAL LABORATORY 

ATTN: REPORT LIBRARY 

ATTN: T DOWLER 

SANDIA NATIONAL LABORATORIES 

ATTN: DIV 9013, W LING 

ATTN: TECH LIB 

Dist-3


OTHER GOVERNMENT 

KAMAN SCIENCES CORPORATION 

ATTN: DASIAC 

CENTRAL INTELLIGENCE AGENCY 

ATTN: COUNTER-TERRORIST GP 

PACIFIC-SIERRA RESEARCH CORP 

ATTN: DIRECTOR OF SECURITY 

2 CYS ATTN: A LAUPA 

ATTN: MEDICAL SERVICES 

2 CYS ATTN: G ANNO 

ATTN: NIO-T 

ATTN: H BRODE 

ATTN: N10 R & D - STRATEGIC SUBCOMMITTEE SYS 


ATTN: RC &LDISBCOM E ATTN: D GORMLEY 

ATTN: TECH LIBRARY 2 CYS ATTN: G MCCLELLAN 

FEDERAL BUREAU OF INVEST ACADEMY 

ATTN: BEHAVIORAL RSCH UNIT 

ATTN: LIBRARY 

FEDERAL EMERGENCY MANAGEMENT AGENCY 

ATTN: ASST ASSOC DIR FOR RSCHTN 

ATTN: CIVIL SECURITY DIVISION 

ATTN: G ORRELL NP-CP 

ATTN: OFC OF CIVIL DEFENSE 

R & D ASSOCIATES 

ATTN: C MCDONALD 

2 CYS ATTN: DOCUMENT CONTROL 

J THOMPATN 

ATTN: J THOMPSON 

SCIENCE APPLICATIONS INTL CORP 


U S DEPARTMENT OF STATE ATTN: E SWICK, MS 33 

ATTN: PM/STM 

ATTN: J MARTIN 

ATTN: J WARNER 

U S N CLEAR REGULATORY COMMISSION 

ATTN: L HUNT 

ATTN: DIR DIV OF SAFEGUARDS 

ATTN: R J BEYSTER 

ATTN: SYANIV 


DEPARTMENT OF DEFENSE CONTRACTORS 

ADVANCED RESEARCH & APPLICATIONS CORP 

ATTN: SECURITY OFFICE 

AEROSPACE CORP 

ATTN: LIBRARY ACQUISITION 

BDM INTERNATIONAL INC 

ATTN: C SOMERS 

ATTN: J BRADDOCK 

DATA MEMORY SYSTEMS, INC 

ATTN: T DUPUY 

KAMAN SCIENCES CORP 

ATTN: DASIAC 

ATTN: E CONRAD 

ATTN: B BENNETT 


ATTN: J FOSTER 

ATTN: J PETERS 

ATTN: LGOURE 

ATTN: M FINEBURG 

ATTN: W LAYSON 


ATTN: R CRAVER 


ATTN: JOHN A SHANNON 

SYSTEMS RESEARCH & APPLICATION CORP 

ATTN: R STEELE 

ATTN: S GREENSTEIN 

Dist-4

""n" ""e*""UMENTATION PAGE Form Approved 

1 

filltiae onr reprint)e o vrgeIhu inludg tmefor eewing ristucilons srn~l.g 741""g 

0 MB No. 0704 .0188 

7s 

o,'o 

6. AUTHOR(S) EPR TP AD AESCVEE 

Worm Uniten No. 

Browne, D. 

N/A 

7. PERFORMING ORGANIZATIC NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION 

Armed Forces Radiobiology Research Institute 

REPORT NUMBER 

Defense Nuclear Agency S9- 

Bethesda, ?4D 20889-5145 

9. SPONSORING/ MONITORING AGENCY NAME(S) AND ADORE!P(E5 10. SPONSORING/ MONITORING 


Defense Nuclear Agency - 

Washington, DC 20305 

11. SUPPLEMENTARY NOTES 

1 2a. DISTRIBUTION/AVAILABILITY STATEMENT 1 2b. DISTRIB3UTION CODE 

Approved for public release; distribution unlimited. 

13. ABSTRACT (Maximum 200 words) ~~S 

fewl 

JUitiwOW 

6v 

!A By. 

A, 14 SUBJECT TERMS I15 NUMBER OF PAGES 

5 

________________________________________________________IS._ 

16 PRICE CODE 

17. SECURITY CLASSIFICATION I S. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF 

OF REPORT OF THIS PAGE OF AB3STRACT ABSTRACT 

UNCLASSIFIED 

UNCLASSIFIED 

StnadFrI9 Rv 9 

NSN 7S40-01 -2604500 StanardO Form 298 e -9 

2,14102

Biomedical Lessons From the 

Chernobyl Nuclear Power Plant Accident 

A4PIPID ;i)RLE$ AAOIOLOGY 

0 AESEARCM .%STITUT5 

SCIAiNTIPIC ReSpO" 

SR91-2 

Lt Col Doris Browne, MC 

The Chernobyl nuclear accident afforded the treating physicians a chance to observe clinical ARS in man, defining the degree 

of severity according to average radiation dose exposure, and to mahe prognoses for the individual patients on the course 

of the ARS, based on biological criteria. The author gives a detailed account of the clinical cause of the disease including 

all available laboratory values. She provides valuable data that can be utilized in handling similar accidents in the future. 

The world's worst radiation accident posited on the skin and mucous mem 

occurred at the Chernobyl nuclear branes from the molten steam and Ā Lt Operations Col Dori Division, Browne, Military MC, is Chief, 

Requirements Medical 

power plant in the USSR during the dust. Wet clothing contaminated by and Applications Department IMRAI, Armed 

early hours of April 26, 1986. This ac- the steam and dust provided another Forces Radiobiology Research Institute (AFRRI). 

cident unleashed megacuries of radio- source of contamination. She is a licensed physician in hematologyactive 

contamination into the atmos- Within 15 minutes of the acident, oncology. Dr. Browne attended medical school 

phere, generated an explosive blast first aid was provided by middle level at Georgetown University, and completed an 

internship and residency in internal medicine 

that knocked the thousand-ton lid off medical personnel and emergency at Walter Reed Army Medical Center. She 

the top of the reactor, and sent burn- team members. Individuals with acute was a fellow in hematology and oncology 

ing graphite and heat in a plume about symptoms were transported to the 

three miles high. The aftermath and hospital in Pripyat, where the initial 

at Walter Reed Army Medical Center. She 

joined the staff of William Beaumont Army 

significance of this disaster are still screening took place; others in satis- Medical Center, El Paso, Texas as the Assistant 

Director of the Hematology-Oncology 

being realized in the USSR and neigh- factory condition were instructed to Clinic, Department of Medicine. She joined 

boring countries, go to the hospital for examination. The the staff of the AFRRI as Chief of the Med- 

Details of the accident were re- initial care consisted of antiemetics, ical Operations Division and is responsible 

ported at the International Atomic symptomatic medication, and stable for the Medical Effects of the Nuclear Weapons 

Energy Agency (IAEA) meeting held saturated potassium iodide. The spa- (MENW) Course and all technology transfer 

activities at the institute. She is the Officerin 

Vienna, Austria, in August 1986, cialized emergency team of radiation in-Charge of the Medical Radiobiology Adand 

summarized in IAEA Safety Series accident specialists arrived at the visory Team (MRA T that recently sponsored 

Technic-' Report No. 75.1.2 This report accident site within 12 hours and, the First Consensus Development Conference 

summarizes the basic information on with the on-site medical personnel, on the Treatment of Radiation Injuries. A sumcasualties, 

triage, and treatment, and screened and triaged more than 350 mary report of this conference was published 

in an international journal and the proceedings 

the radionuclides released into the persons within the first 36 hours. Dur- of the conference are in press. She isa member 

atmosphere. The immediate casual- ing the first 24 hours, 132 persons of the American Society of Clinical Oncology, 

ties included only plant personnel, were hospitalized; one individual died American College of Physicians, and the Nefiremen, 

and auxiliary staff present at, from severe thermal burns during the tional Medical Association. 

or in the vicinity of the accident site. first hour, and another worker (a re- 

These casualties were all subject to actor operator) was unaccounted for body, as well as with the intake of adthe 

combined effects of the following: and believed to be buried under the ditional radionuclides through inhalashort-term 

beta/gamma radiation re- collapsed debris. tion. These patients were diagnosed 

b leased in the emission cloud; external The triage officer, a physician, as having ARS resulting from extenbeta/gamma 

radiation from fragments made decisions based on the initial sive beta radiation bums to the skin 

* of the damaged reactor core scat- symptoms and lymphocyte counts. and significant whole-body gamma 

tered through the accident site; in- Persons with severe symptoms were radiation exposure. 

halation of gaseous and aerosolized hospitalized with clinical complaints The diagnostic criteria used to assess 

dust composed primarily of radio- of acute radiation sickness (ARS). the presence of ARS was the presence, 

isotopes of cesium, plutonium, and Three hundred of these patients were intensity, and duration of symptoms 

iodine; and beta/gamma particles de- sent to a specialized treatment center is, nausea, vomiting, and erythema of 

*Chief, Medical Operations Division, Military Re- in Moscow and another 200 were the skin and mucose); time of onset; 

quirements and Applications Department. Armed sent to a hospital in Kiev. The 237 and the peripheral lymphocyte count, 

Forces Rediobiology Research Institute BWthsda, hospitalized individuals received sig- which decreased to less than 10/L 

MD 20814. 

Supported by the Armed Forces Rodioboogy Re- nificant combined radiation effects during the first 24 hours following 

search Institute. Defense Nuclear Agency. Views from the extensive beta/gamma ex- radiation exposure in patients with 

presented in this paper are those of the author; 

no endorsement by the Defense Nuclear Agency posure, which was generally external ARS. During the first 36 hours after 

hes boen given or ahould be inferred. and relatively uniform over the whole the accident, the 237 hospitalized 

Pe a-90-91i0. see emtbiOctaor 990 25 

91 5 C7 054

persons were diagnosed as having a and prognosis of ARS. Hyperamylas- treated with intravenous administraclinical 

pattern consistent with first emia was used as a supplementary tion of triple broad-spectrum antidegree 

through fourth degree ARS. diagnostic tool. biotics, including aminoglycoside, 

After admission to the hospital, they Treatment consisted of supportive cephalosporin and semi-synthetic 

were monitored again for contamina- therapy, which included selective anti- penicillin. If this regimen did not retion 

and, when necessary, decon- microbial intestinal decontamination, duce the fever within 48 hours, three 

taminated with soap, water, and a reverse isolation, empiric systemic anti- or four doses of gamma globulin were 

clothing change. Routine samples of biotic administration, and transfusion administered. An intravenous antiurine 

and blood were drawn for anal- replacement of blood and blood prod- fungal (amphotericin B) was adminysis, 

and thyroid scanning was per- ucts. Definitive treatment of allogeneic istered if the neutropenic fever performed. 

The radiation dose received bone marrow transplantation (BMT) sisted for seven days, along with the 

was estimated by counting the num- and human embryonic liver cell trans- antibiotics and gamma globulin. Paber 

of aberrant chromosomes (dicen- plantation (LCT) was performed on tients with herpes simplex were given 

trics) in cultured lymphocytes (cyto- patients with irreversible myelosup- acyclovir. Approximately one third of 

genetic analysis). The diagnosis of pression. A sterile environment was the patients with third and fourth 

ARS was confirmed during the first maintained through strict observance degree ARS had the herpes virus. 

five days for persons admitted to of hand washing by all attending per- Viral skin lesions were treated with 

the Moscow Hospital. Approximately sonnel upon entering and leaving the topical acyclovir. No deaths were 

seven days after the accident, the room; mandatory use of disposable attributed to bacterial infection alone 

radiation dose was estimated and the gowns, masks, and caps; antiseptic in patients with hematopoietic synpatients 

were categorized into four decontamination of footwear; chang- drome. However, infectious complicagroups 

according to prognosis and ing of patient undergarments daily; tion was the primary cause of death 

severity of hematopoietic syndrome antiseptic washing of walls, floors in patients with ARS complicated by 

(Table I). Twenty-two injured persons and items used in the room; and indi- thermal burns, radiation-induced enwere 

classified as having fourth de- vidually assigned antiseptically treated teritis, or acute graft-versus-host 

gree (extremely severe) ARS; 23 as nursing items. Isolation rooms pro- disease from BMT. The etiology of 

having third degree (severe) ARS; 53 vided air sterilization with ultraviolet terminal septicemia, documented by 

as having second degree (moderate) lamps. The microorganism population surveillance cultures, was most often 

ARS; and 139 as having first degree was maintained at less than 500M 3 from Staphylococcus epidenrdts. 

(mild) ARS. 3 in the room air. Raw fruits and veg- The hematopoietic syndrome was 

Neutrophil count was used to de- etables and canned products were treated with prophylactic and theratermine 

finally the magnitude of radia- eliminated from the patients' diet. 4 peutic fresh random donor platelets 

tion dose. The parameter used was The decision was made early to when the platelet count dropped to 

the time required for the neutrophils perform BMT on patients with third 20 x 10 9 /L or lower, or with the first 

to decrease to 0.5 x 10 9 /L, based on and fourth degree ARS and possible sign of bleeding. Transfusions usudata 

collected over a period of up to irreversible myelosuppression. 5 These ally were required every one to three 

three months in cases that exhibited patients vomited within the first half- days. To inactivate the immunocomtypical 

postirradiation platelet and/or hour, suffered from diarrhea during petent cells from the donor, all blood 

neutrophil counts with distinct de- the first one to two hours, and from components were irradiated with 

pletion and restoration phases. Coin- swelling of the parotid gland during 1,500 cGy of gamma radiation before 

plete blood counts were performed the first 24 to 36 hours of exposure, transfusion. Only one person received 

two to three times per week for two in addition to myelosuppression. single donor platelets. While the mato 

three months. This data was used Infections, manifested by the on- jority of patients showed no evidence 

to definitively confirm the diagnosis set of fever and neutropenia, were of overt bleeding, autopsy results disclosed 

micro-circulatory failure and very 

porous capillaries in several organs. 

Table I. Diagnostic Categories for Acute Radiation Sickness (ARS1. Inroms crorgse 

In some situations, cryo-preserved 

Degree of ARS Dose (cGy) Seventy of ARS Prognosis autologous platelets, as well as allogeneic 

platelets, were used success 

100-200 Mild Very favorable fully. Autologous platelets were taken 

from patients with second and third 

11 200-400 Moderate Relatively favorable dreens onth st d postd 

degree ARS on the first day post- 

III 400-600 Severe Doubtful irradiation. Platelet transfusions pre- 

IV z 600 Extremely severe Poor vented life-threatening bleeding. Three 

to eight transfusions of 250cc per 

26 The Journal of tte US Amiv Medical Deoorye "

person were used to treat patients Table It. Transplantation Cases, Estimated Radiation Dose, and Outcome. 

with second and third degree ARS. 

No evidence of refractoriness de- Degree of ARS Dose IcGy) Treatment Day of Death Cause of Death 

veloped. A considerable number of 

packed red blood cells were trans- IV 920 BMT 15 Skin; pneumonitis 

fused in patients with second and IV 1200 BMT 17 Skin: GI injury 

third degree ARS accompanied by IV 1180 BMT 18 Skin; GI injury 

severe radiation burns. 

Allogeneic BMT taken from 113 Iv 1000 BMT 18 Skin; GI injury 

random related donors was performed I1 550 BMT 21 Hemorrhage 1 

on 13 patients with third and fourth Iv 830 BMT 24 Pneumonitis 

degree ARS. Additionally, six patients 660 BMT 25 ARDS; toxicity 

with fourth degree ARS received 

embryonic LCT, which contained III 440 BMT 2 34 Mixed infection; GVH 

* 

stem cells and few immunocompe- IV 640 BMT 3 48 Mixed infection; GVH 

tent cells to decrease 

tetclst the risk eraeters of 

fIV 750 BMT 3 86 Mixed infection; GVH 

developing acute graft-versus-host 

disease (Table II). Fifty percent (seven IV 1020 BMT 4 91 Mixed infection; GVH 

patients) of the BMT patients died III 560 BMT 5 Alive 

within 17 days of transplantation (15 IV 870 BMT 5 Alive 

to 25 days following radiation exposure) 

from acute radiation injury IV 1110 LCT 14 Skin; GI injury 

to lung, intestine, and/or skin. The IV > 1000 LCT 14 Skin; GI injury 

remaining six patients did not have 

severe skin burns or intestinal injuries IV 1370 LCT 15 Skin; GI injury 

but received a total radiation dose IV 1240 LCT 17 Skin; GI injury 

estimated to be between 440 and IV 1090 LCT 18 Skin; GI injury 

1,020 cGy. Two of the six patients 

survived BMT (having received 560cGy IV 830 LCT 6 30 Toxicity; ARDS 

and 870 cGy doses, respectively) from 

N o t a BMT = bone marrow transplantation; LCT = liver cell transplantation; GI = gastrointestinal ,nurv 

haplo-identical female (sisters) donors. GVH = graft-versus-host: ARDS = acute respiratory distress syndrome irespiratory insufficiency). 

Both experienced transient partial en- 'Hemorrhage from mechanical trauma during catheterization. 

23 graftment of the transplanted mar- BMT M from rmHAietcidnr haplo - 1 identical donor but own myelopoiesis restored. 

38MT from HLA-identical donor. 

4 BMT from haplo-identical donor but own myelopoiesis restored. 

row before rejection 32 and 35 days 

after BMT, respectively, with res- 

5 SMT from haplo-identical donor rejected, own myelopoiesis restored. 

toration of their own myelopoiesis sLCT from male, postmortem evidence of own myelopoisais being restored. 

after 28 days. These two patients 

are still alive at more than three planted marrow engrafted. While re- ity, duration, and recurrence, contriband 

one half years after the acci- covery of autologous myelopoiesis uted significantly to the overall pathodent. 

5 A 62-year-old female patient may occur following large doses of physiology and outcome of the patient. 

who received LCT lived for 30 days; radiation exposures, such as experi- Severe skin injuries were manifested 

postmortem findings showed evidence enced by those patients with third by diffuse hyperemia; secondary eryof 

regeneration of her own myelo- and fourth degree ARS, it is unknown thema; dry and wet desquamation 

poiesis, indicated by female cell karyo- if this occurred due to transient en- with blistering, ulceration, and necrotype. 

She had received a male donor graftment of transplanted stem cells. tic dermatitis; recurrent waves of erytransplant. 

4 6 Radiation-induced skin injuries (beta thema; and after evidence of healing 

The effectiveness of BMT in an radiation burns) were seen only in of the primary lesions, edema, fever, 

emergency situation may be limited to combination with hematopoietic syn- and a worsening of the patient's clinpatients 

receiving less than 900cGy drome radiation injury. Skin doses of ical picture. 4 Topical treatment was 

of gamma radiation with at least 1 % radiation were estimated to be 10 to necessary, with glucocorticoids and 

of marrow stem cells remaining, no 20 times greater than bone marrow analgesia in the more severe cases. 

skin or intestinal radiation injuries, or whole-body doses, confirming the Pain control was relatively ineffecand 

no combined injury. 7 Seven of uncontrolled, nonuniform nature of tive, especially topical anesthesia, 

the 13 BMT patients died of skin and radiation accident exposure. These which seems to be typical for radiaintestinal 

injuries before the trans- skin injuries, according to their sever- tion injuries. 

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 

tients with radiation burns on > 40% ment was indicated for the parotitis, made during the first year after the 

to 100% of body surface area.6 If early which gradually resolved; salivation, accident. Patients usually had dyssecondary 

erythema over > 40% body however, recurred very slowly. trophic and ulcerative skin lesions, 

surface area was present, a clinical Rapidly intense dyspnea with acute some with subcutaneous edema pripicture 

of febrile-toxemia, followed by respiratory insufficiency (adult respi- marily over the knees and feet. Skin 

hepatorenal insufficiency, encephal- ratory distress-like syndrome) was lesions were treated with agents that 

opathy with cerebral edema, coma, seen in seven patients with third and improved local blood circulation and 

and death resulted 14 to 48 days post- fourth degree ARS. This condition tissue trophism. Five patients sufirradiation. 

Plasmapheresis was used to rapidly progressed for two to three fered deep ulcers which required recontrol 

the hepatorenal insufficiency. 1 4 days leading to death. Postmortem peated plastic surgery. 

This treatment prolonged survival examination revealed enlarged blue The immunologic status of patients 

slightly but did not prevent death lungs with interstitial edema but no with second, third, and fourth degree 

from encephalopathic coma. The destruction of mucous membranes of ARS, tested 1 to 1.5 years after the acburns 

may have been the primary the trachea and bronchi. These pa- cident showed a persistent decrease 

cause of death in some cases; how- tients also had severe skin and in- in T-helper lymphocytes with an inever, 

in most cases, the burns were testinal radiation injuries, crease in T-suppressor lymphocyte 

associated with severe hematopoietic Beta radiation caused early damage activity and a significant decrease in 

syndrome and severe acute gastro- to the eye tissues; erythema of the the helper-suppressor ratio. However, 

intestinal syndrome (enteritis). eyelid, with increased vasculature of there was no evidence of a decrease 

In ten patients, the gastrointestinal the lid, and conjunctiva. Cutaneous in the absolute lymphocyte level or in 

syndrome was the life-threatening changes were manifested by waves the T- and B-subpopulations. These 

manifestation of ARS, with severe of erythema, hyperpigmentation, and changes in lymphocyte helper and 

diarrhea suggesting a radiation dose scaling. Partial epilation of the eye- suppressor populations were not seen 

greater than 1,000 cGy. All of these brows was transient, and all patients in patients with first degree ARS. Durpatients 

died within three weeks of ir- retained their eyelashes. (Scalp hair ing the follow-up period, no severe or 

radiation. When the enteritis persisted growth recovered fully.) Other eye life-threatening infections were noted. 

in spite of supportive fluid and elec- changes noted were decreased cor- Immunocorrective therapy was attrolyte 

therapy, death may have been neal sensitivity and superficial radia- tempted using T- and B-activin in 

caused solely by the gastrointestinal tion-induced keratitis, which regressed several cases. Respiratory infections 

syndrome. over one to two months without corneal occurred in three of eight patients 

Large amounts of thick rubber-like opacification. Treatment for the eye with third and fourth degree ARS and 

mucous formed in the oropharyngeal changes included topical ointments to only one of 22 patients with second 

area of about 82 patients and in some the eyelid skin and eyedrops of 20% degree ARS. A competent immune 

cases resulted in respiratory difficulty. albucid, sophradex, and vitamin solu- system remains critical in enhancing 

Initially, some patients showed be- tions into the conjunctival cavity. 4 microbial and viral resistance during 

nign acute radiation-induced inflam- One severely ill patient with fourth convalescence of the irradiated pamation 

of cheeks, tongue, and gums. degree ARS, who survived the acute tient. A plan of long-term follow-up 

Those having third and fourth degree phase, developed angioretinopathy observation remains in effect. 

ARS had, in addition to the rubbery with hemorrhage and plasma dismucous 

plugs, painful erosions and charges about five months post- Conclusion 

ulcers of oral mucosa, which required irradiation. He also had persistent The consequences ot the Chernobyl 

sterile saline irrigation and frequent low diastolic pressure in the central radiation accident provide data on a 

debridement. In a significant number retinal artery. He is one of the two large group of critically ill patients 

of patients this radiation-induced in- surviving BMT patients. No radiation- who received uniform whole-body irflammation 

was complicated by sec- induced lens changes were observed radiation and required treatment of 

ondary bacterical and viral infections, one year postirradiation. ARS in a massive casualty situation. 

In one third of the patients with Convalescence of three to four The event afforded the opportunity to 

severe hematopoietic syndrome, her- months was required for those pa- learn many lessons regarding the 

petic lesions formed massive crusts tients with first and second degree biomedical effects of ionizing radiaon 

the lips and face about three to ARS; a much longer period was tion and to clarify many aspects about 

four days postirradiation. Patients necessary for those having third and the early radiobiclogical effects in 

with fourth degree ARS and herpetic fourth degree ARS. The majority of humans. The accident also provided 

lesions also developed radiation-induced the patients have resumed work but data on severe and extensive beta 

parotitis, inability to salivate, and re- cannot work with radiation sources. radiation skin injuries, which com- 

28 The Journlsl of the US Arrmy Moica DeDa1" ,e"

plicated the course of illness and REFERENCES oflomzng Radiaton, United Nations 

played a significant role in the death 1. USSR State Commission on the Scientific Committee on the Effects 

of 19 of the 31 patients. Combined Utilization of Atomic Energy. The of Atomic Radiation. New York, 

injuries consisting of trauma, thermal accident at the Chernobyl nuclear 1988, pp 613-647. 

burns, and radiation were the cause power plant and its consequences. In- 5. Baranov AE, Gale RP, Guskova AK, 

of death in two patients very early in formation complied for the Post- et at: Bone marrow transplantation 

the course of ARS. Clinical ARS was Accident Review Meeting, part II, after the Chemobyl nuclear accident. 

observed in man, the degrees of se- Annex 7, Vienna, Austria, August NEnglJMed321(4):205-212, 1989. 

verity defined according to average 25-29, 1986. 6. Young RW: Chernobyl in retrospect. 

radiation dose exposure, and prog- 2. International Nuclear Safety Ad- Pharmac Ther 39:27-32, 1988. 

nosis made for the course of ARS visory Group. Summary report on the 7. Browne D, Weiss JF, MacVittie TJ, 

based on biological criteria. Informa- post-accident review meeting on the et at: Conference report: The first 

tion for biological dosimetry was Chernobyl accident. Vienna, Austria: consensus development conference 

obtained from karyotypical analysis, International Atomic Energy Agency on the treatment of radiation injuries. 

lymphocyte counts, and symptoms (IAEA), 1986. (Safety Series No. lntJ]jdwtBio/57(2):437-442, 1990. 

during the early stages of illness; later 75-INSAG-1). 8. Guskova AK, Nadezhina NM, Barathe 

granulocyte count proved to be 3. Bair WJ: Radiological impacts of the banova AV, et al: Acute effects of 

a useful dosimetric tool. While con- Chernobyl accident. Health Physics radiation exposure following the 

tinuous data and follow-up assess- Society Newsletter. February, 1987. Chernobyl accident: Immediate rement 

is necessary, this information 4. Guskova AK, Barabanova AV, suits of radiation sickness and outshould 

prove useful in responding Baranov AE, et al: Acute radiation come of treatment. In Treatment ,.r 

to radiation accidents and providing effects in victims of the Chernobyl Radwhon nrunes. Browne D, Weiss JF, 

effective medical carg to the result- nuclear power plant accident. Ap- Mac Vittie TJ, et al (eds). Plenum, 

ing casualties. pendix to Sources, Effects and Risks New York, 1990. 0 

29

OATE:

$A t#4- 44", ONIb004lt-# Ne Oft0-*.0wM o I 

,*Stl eb,4 W ,11 0' 0Mg . '1 A'f00 I, IP4 1 ** $ et*14d4 *. oa 41A 

-----------.... PAGE Ou OAp .e 

AD A254 669 , e11e,,( sf'1 I144 e f I6m 4OI I .E4 

, AGENC 3Ili REPORT TYPE AND DATES COVERED 

DNA-TR- 

4 ItTLE ANO SUBTITLE 

Nuclear Accidents in the Former Soviet Union: 

5. FUNDING NUMBERS 

Kyshtym, Chelyabinsk and Chernobyl DNA/AFRRI 4020 

*. AUTHORIS) 

Daniel L. Collins, Ph.D. 

Lt Col, USAF 

lE L E T E 

1. PERFORMING ORGANIZATION NAME(S)AND AOORESSIES) GANIZATION 

Defense Nuclear Agency R U 

Armed Forces Radiobiology Research Institute 24 j 1 S12 " 

Bethesda, HD 20889-5145 

9. SPONSORING/MONITORING AGENCY NAMEIS) AN() ADORE SS(E S) 10. SPONSORING/MONTORING 



Armed Forces Radlobiology Research Institute 

Bethesda, MD 20889-5145 

SUPPLEMENTARY NOTES Many of the data contained in this report comes from discussion 

nd brefings. In addition, on 2 December 1991, several scientists from the 

Institute of Biophysics of the USSR Ministry of Health, Chelyabinsk Branch Office, 

ho have studied the Kyshtym and Chelyabinsk nuclear accidents for decades, 

resented human data at AFRRI that have never before been released to the West. 

12e. DISTRIBUTION/AVAILABIUITY STATEMENT 12b. DISTRIBUTION CODE 

Approved for public release: Distribution is unlimited. A 

13. ABSTRACT(Msuimum200 w0 sJ Three nuclear accidents besides Chernobyl have occurred in 

the Former Soviet Union (FSU). The accidents occurred over the geographic area 

around Kyshtym and Chelyabinsk in the Urals between 1949 and 1967 and cntaminated 

over half a million people. The first accident occurred in 1949-1951, the stcond 

on 29 September 1957, and the third in 1967, and involved the air transfer of 

irradiated sand particles. Although these accidents occurred between 25 and 43 

years ago, the first official admission by the FSU was made in June 1989, and it 

was only during late November 1991 that the FSU declared a national disaster 

emergency concerning the affected area. The health ministries are now interested 

in data previously collected from these irradiated populations to examine health 

effects, including cancer, and genetic damage in humans. Data collected from these 

large populations and occupationally exposed workers offer a unique opportunity to 

quantify the adverse health effects of chronic exposure to ission products, 

reactor neutrons and enviromental chemicals. 

94,1 126 

14. SUBJECT TERMS 16. NUMBER OF PAGES 

Kyshtym, Chelyabinsk, Chernobyl, REM, human, psychological 0 

radiation 9 0 Sr, Curies, 13 7 Cs, Neutrons, Isotopes, radiation 

injury, cancer, genetics, enviromental chemicals 

To. PRICECOOt 

17. SECURITY CLASSIFICATION 15. SECURITY CLASSIFICATION 19. SECURITY CLASSIICATION 20. LIMITATION OF 

OF REPORT OF THIS PAGE OF ABSTRACT ABSTRACT 

Unclassified Unclassified Unclassified None

ABSTRACT 

Three nuclear accidents besides Chernobyl have occurred in the Former 

Soviet Union (FSU). 

The accidents occurred over the geographic area around 

Kyshtym and Chelyabinsk in the Urals between 1949 and 1967 and contaminated 

over half a million people. The first accident occurred in 1949-1951, the 

second accident on 29 September 1957, and the third in 1967, and involved 

the air transfer of irradiated sand particles. 

Although these accidents 

occurred between 25 and 43 years ago, the first official admission by the 

FSU was made in June 1989, and it was only during late November 1991 that 

the FSU declared a national disaster emergency concerning the affected area. 

The health ministries are now interested in obtaining data previously 

collected on this irradiated population to examine the health and heredity 

(genome damage, etc.) implications associated with these victims. A 

collaborative involvement by DNA/AFRRI with the Health Ministries would be a 

unique opportunity to obtain previously unavailable human data. 

-il1 

92-23453

Nuclear Accidents in the Former Soviet Unions : 

Kyshtym, Chelyabinsk, and Chernobyl 

Avftf1atj1±t 7 Codeq 

Daniel L. Collins, Ph.D. . .'A/l ad/-or 

Dltt 

Spe3cial 

Lt Col, USAF 

I 

DTMIC QUALIY U11TPECTED 8 

INTRODUCTION 

In September 1991, three weeks after the attempted putsch, I represented 

the Defense Nuclear Agency/Armed Forces Radiobiology Research Institute as 

part of a United States Information Agency (USIA) contingent in Moscow and 

St. Petersburg, Russia. 

I met numerous scientists that were and are 

currently involved in all the nuclear accidents that occurred in their 

country. Much of the data contained in this report comes from discussions 

and briefings. 

In addition, on 2 December 1991, several scientists from the 

Institute of Biophysics of the USSR Ministry of Health, Chelyabinsk Branch 

Office, who have studied the Kyshtym and Chelyabinsk nuclear accidents for 

decades, presented human data that have never before been released to the 

West. 

The radiation situations in the area of Kyshtym and Chelyabinsk are 

unique because over the last 40 years masses of people have been exposed to 

9 0 Sr in the food and water chain. Basic dosimetry investigations focused on 

2

the doses of 9 0 Sr. 

The first estimates of exposure were from nuclide 

measurements taken from the river sediment. These analyses began in 1951, 

one and a half years after the accident occurred. 

ACCIDENTS 

In 1948, the U.S.S.R. began operating a plutonium production plant 

called Mayak in the Kyshtym/Chelyabinsk region. In 1949-1951, an accident 

released 3 million Ci of radiation into the Techa River (Degteva, 1991, see 

Figure 1). 

A second accident occurred in 1957, southeast of Kyshtym, when 

improperly ventilated storage tanks exploded, and 20 million Ci of 

radioactive waste were released into the atmosphere (Akleev, 1991). 

The 

storage complex was located 1.5 Km from the reprocessing plant and consisted 

of 60 underground storage tanks. The radionuclide composition of the 

fallout indicated it was comparatively fresh nuclear waste from 

reprocessing, stored for not more than 6-7 months, from which not only 

uranium and plutonium but also cesium had been removed (Medvedev, 1991). 

Due to the confined nature of the blast, the majority (90%) of the 

nuclear waste dispersed near the tanks in the form of a liquid pulp 

(Burnazyan, 1990). 

However, a plume cloud with an activity of 2 million Ci 

dispersed its contamination over the area shown on the map in Figure 1. The 

contaminates from the plume cloud were confined to the Chelyabinsk and the 

Sverdlovsk provinces. Before the accident, more than 28,000 people lived in 

the 38 villages along the Techa River. Dosages of 90Sr exceeded 0.01 Ci/km 2 

3

and were distributed over 23,000 km 2 (Kosenko, 1991). 

The highest 

concentration of 90Sr was located in an area known as Metlino, located near 

Kyshtym (see Figure 1). The doses in the Metlino area averaged 3 Sv/km 2 . 

The dispersion of radioactivity from the plume cloud is shown by the map in 

Figure 1. 

In the spring of 1967, a third accident occurred resulting in an air 

transfer of radioactive sand particles from the beach of Lake Karachay 

(Akleev, 1991). This lake is located due west of Argayash. A total of 600 

Ci 137Cs and 9 0 Sr was released by the air transfer of sand particles from 

the Lake Karachay beach (Akleev, 1991). 

Radiation from this accident was 

primarily gamma radiation along the Techa River and reached 5R/hour 

(Kosenko, 1991). The average activity along the Techa River was 10 - 5 

Ci/liter of water. A massive three month cleanup between the summer and 

autumn of 1967 (Degteva, 1991), resulted in a ten fold decrease in radiation 

levels. This occurred because the Soviets removed all the water from the 

contaminated reservoir located near the Metlino area (Degteva, 1991). 

(NOTE: We were not told where the water was moved to nor how it was 

disposed of.) 

DISCUSSION 

In the aftermath of the Kyshtym accident, if a village was found to be 

radioactive it was scheduled for evacuation as political circumstances 

dictated (Soyfer, 1991). 

In addition to the area dosimetry, which measured 

4

iver sediment for contamination, human dosimetry began in the summer of 

1951 measuring body excrement and contaminated clothes (Degteva, 1991). It 

should be noted, that during this 1.5 year hiatus between the accident and 

the beginning of dosimetry measurements, the people were neither informed of 

the accident nor of their (internal or external) exposure to any form of 

ionizing radiation (Soyfer, 1991). 

Consequently, all the food and water 

consumed during this time period was contaminated with radiation (Akleev, 

1991). 

In an attempt to further quantify dosimetry from the Kyshtym and 

Chelyabinsk accidents, the bones of deceased persons were exhumed during the 

1960's and resulted in the creation of a data base that reflected their 

*lifetime* exposures to 90Sr (Kosenko, 1991). 

In 1974, the Soviets created 

a data base using live subjects to determine their whole-body doses of 90Sr. 

The methodology used to make these determinations included dosimetric 

examinations of urine samples and frontal lobes (Douplenski, 1991). 

The 

urine samples were obtained from 1,500 residents living beside the Techa 

River (Kosenko, 1991); the 12,500 subjects involved in the frontal lobe 

study also lived along the Techa River (Kosenko, 1991). The teeth of 15,000 

living subjects in this area were also examined for 90Sr doses (Kosenko, 

1991). The location from which these residents were evacuated is known 

today as the "Post Box Chelyabinsk-400 area (Medvedev, 1991). 

The results show that 1,000 people living by the Techa River had greater 

than 1 mCi of 9 0 Sr in their bones (Kosenko, 1991). 

Further analyses showed 

5

that these people born in 1932-1933, and were teenagers during the first 

accident, accumulated three to five times more 90Sr than did those who were 

adults at the time of the first accident (Akleev, 1991). 

The maximum 

reading of 6 mCi per person occurred in those individuals that were 

teenagers during the time of the accidents (Akleev, 1991). 

Measuring the 

metabolism measurements of 9 0 Sr and overlaying them with all three previous 

accidents provided the following results: The exposure levels averaged 0.42 

Sv along the Techa River to the southwest, 0.52 Sv along the middle portion 

of the Techa River, and 2 Sv near Chelyabinsk (Degteva, 1991). 

The three 

accidents affected 437,000 people (Kosenko, 1991). 

Of these, 1,200 people 

obtained 200 rems over a 2 year period (Kosenko, 1991). 

In addition, some 

people received doses of up to 400 rems to their bone cells (Kosenko, 1991). 

RELOCATION 

Of the 39 villages along the Techa River before the accident, only four 

villages are safe to inhabit today (Akleev, 1991). 

An analysis of people 

along the Techa River show decreased leukocyte immune system functioning 

(Akleev, 1991). 

This is attributed to the 3- year period when people daily 

drank the radioactive water and ate contaminated food before their 

relocation (Douplenski, 1991). 

The average time of relocation took 8 years 

to complete (Degteva, 1991). The time range for relocation spanned 3-11 

years for evacuating all residents from the 35 contaminated villages along 

the Techa River (Douplenski, 1991; Medvedev, 1991). 

The total number of 

people moved during this time exceeded 10,500 (Akleev, 1991). 

After the 

6

people were relocated the villages were incinerated to ensure that no human 

habitation would occur in this highly radioactive are?. 

However, the heat 

and smoke created by the incineration process spread the contamination over 

the streams, rivers and lakes further contaminating the food chain for 

animals and humans. 

One exception to the protracted relocation of residents living in the 

contaminated areas occurred out of necessity. Following the Kyshtym 

accident, 1,154 residents of Kasli, near Kyshtym, were removed 7-10 days 

after the accident, due to extremely high 9 0 Sr levels (Akleev, 1991). 

People that were removed during this time period, now have twice the accute 

myeloid leukemias that the control groups have exhibited (Akleev, 1991). 

An historical note, the construction of the radioactive facilities was 

conducted between 1945-1948 by approximately 70,000 inmates from 12 labor 

camps (Douplenski, 1991). The Khyshtym location is N 55-44, E 60-35. The 

Khyshtym restricted area covers 2700 sq/km and contains eight lakes with 

interconnecting watercourses. The Khyshtym atomic plant is situated in a 

tunnel, which extends beneath a river, with only a smoke stack visible from 

the air or ground. During the construction process, one lake was drained, a 

building was built on its lakebed with cement, rubber and lead, then the 

lake was refilled (Medvedev, 1979). 

During the cold war several high 

altitude, reconnaissance aircraft routinely photographed this area. 

In 

1960, Major Francis Gary Powers was shot down by a surface-to-air missile 

while flying over the Khyshtym, Chelyabinsk atomic facilities. 

7

CONTROL GROUPS 

Two control groups were selected for comparison purposes for this 

longitudinal field study of irradiated humans (Degteva, 1991). 

The control 

groups were located just south of this area and were not involved in the 

radiation. The first control group consisted of 34,000 persons of the same 

socioeconomic status as the victims and had no access to the contaminated 

Techa River and were not contaminated by the plume cloud or the other 

accidents (Degteva, 1991). 

The second control group consisted of all people 

in the Chelyabinsk rural province but not living in the contaminated city of 

Chelyabinsk or the irradiated area surrounding Chelyabinsk (Kosenko, 1991). 

The control groups within the nonradioactive portions of the Chelyabinsk 

province consisted of 1.5 million people (Kosenko, 1991). 

This data base 

has been a part of the ongoing research for 40 years (Degteva, 1991). 

The 

information we obtained is a summary compiled from 33 years of data (Akleev, 

1991). [NOTE: There are still 7 years of untapped data because the 

analysis is 7 years behind due to a lack of computer equipment.] 

IMPACT OF IONIZING RADIATIONi ACCIDENTS 

Statistical results showed significantly increased death rates along the 

Techa River over the last 33 years. 

Coefficients were described in terms of 

excessive risk per Gy. 

Findings showed stomach cancers to be two to three 

times greater than in survivors of Nagasaki and Hiroshima, breast cancer two 

8

times greater, and lung and esophageal cancers two to three times greater 

(Degteva, 1991). 

Research is in progress that examines the progeny of 

irradiated mothers for stillborn children, abortions, and miscarriages. 

Analyses thus far indicate a significantly greater number of birthing 

complications in the irradiated mothers than existed among the control 

groups (Degteva, 1991). 

Today, near the Kyshtym reservation, where the town 

of Kasli used to be, the soil still contains from 1000 to 2000 Ci/km 2 of 

9 0 Sr (Medvedev, 1991). It was from this area that 1,154 previous residents 

of Kasli were rapidly evacuated in a 7-10 day period following the accident. 

Several types of military training maneuvers are conducted in these 

contaminated areas today (Douplenski, 1991). 

In addition to the ionizing radiation doses the victims of these three 

radiation accidents received, no humanitarian support existed, the people 

lived on a less than a well-balanced diet, and there was a pervasive lack of 

medical attention and equipment (Degteva, 1991). 

There were and still are 

only 50 beds to care for the 500,000 irradiated people from the three 

accidents (Kosenko, 1991). 

Russian Ministry of Health officials in the Chelyabinsk Branch Office 

wou I like AFRRI as a research partner to perform some collaborative 

research (Kosenko, 1991). 

They perceive AFRRI as being able to assist them 

in increasing the accuracy of their dosimetry and in the creation of an 

interactive computer register. An increased knowledge regarding the late 

effects of ionizing radiation would result from this collaborative effort. 

9

Specifically, this would assist in determining the: 

1) long-term, 2) lowdose, 

3) immunologic, and 4) genetic alterations associated with exposure to 

ionizing radiations (Akleev, 1991). 

It would also provide insight regarding 

the effects that exposure to ionizing radiation has on subsequent 

generations (child birth, miscarriages, abortions, etc.). 

PSYCHOLOGICAL ASPECTS OF NUCLEAR ACCIDENTS 

It is noteworthy that people living in the Kyshtym/Chelyabinsk area are 

very antinuclear as evidenced by the voluntary shutdown of two nuclear 

electrical generation reactors during June-July 1989 (Akleev, 1991; 

Medvedev, 1991). 

Also, although billions of rubles (equivalent to millions 

of U.S. dollars) had been spent on tht 

construction of a breeder reactor, it 

too was shut down approximately 3 years ago, an aftermath of the Che.nobyl 

accident in 1986 (Douplenski, 1991; Medvedev, 1991). 

The reason given for 

shutting down both of these "nuclear" facilities after years of use and 

construction was the psychological animosity that existed in the people 

living in this contaminated region (Douplenski, 1991). 

A similar 

psychologically induced result occurred in Mcscow, where a new, ready to be 

used nuclear power plant was prevented from opening due to public outcry 

(Douplenski, 1991). 

In addition, a nuclear power plant approximately 40 km 

south of St. Petersburg was closed to appease the psycholo_, tally upset 

populace (Karpov, 1991). 

The impact that individual perceptions of ionizing 

radiation have on society is now being realized (IAEA Report, 1991). 

10

In other parts of the FSU, specifically around the contaminated 

Chernobyl region, 35,000 people in the Belorus city of Gomel went on strike 

on 26 April 1991 to protest the fourth anniversary of Chernobyl. 

Similarly, 

60,000 people waving "nationalist" flags packed the square in front of the 

Sofia Cathedral in Kiev, the capital of the Ukraine, and demanued punishment 

for those responsible for the world's worst nuclear accident (Bohlen, 1987, 

Reuters, 1990). 

The Rukh press agency reported similar antinuclear 

demonstrations in Kiev, the western Ukranian city of Lvov, and in the 

Belorus city of Minsk, as well as demonstrations elsewhere in the two 

republics that were the worst hit by fallout from the accident (Reuters, 

1990). 

Reasons for the psychological outcry among the Chernobyl ,ictims are 

numerous. 

The delays caused by scientific and political discussions finally 

resulted in the evacutation of 40,000 residents from an area where the 

contamination was 40 Ci/km 2 . Furthermore, 10,000 people receiving 15-20 

Ci/km 2 and 60,000 people receiving 5-15 Ci/km 2 were not allowed to relocate 

(Kedrovsky, 1991). 

In addition, as part of the Chernobyl cleanup effort, 

numerous people are employed in 2 week (on/off) shifts inside the 6 km "Hot 

Zone" (Shishchits, 1991). 

It is expected that the rate of thyroid cancer is 

5 to 10 times the rate expected for 1.5 million Soviet citizens, leukemia 

rates among children in some areas of the Ukraine are 2 to 4 times normal 

levels, and the death rate for peopie working in the Chernobyl plant since 

the accident is 10 times what it was before the accident (Barringer, 1990). 

Furthermore, the scientific director of the zone surrounding the damaged 

11

Chernobyl power station estimated that the disaster has currently claimed 

over 7,000 lives (Wise, 1991). 

Perhaps the worst aspect of these nuclear accidents is the omnipresent 

invisible threat and the continuing fear that the future is marred by 

irreversible cancer or genetic defects. This may have increased since the 

accidents, when radioactive fallout contaminated the environment, animals, 

and people. An undeniable and continual reminder for the residents is that 

all the timber in the affected areas is radioactively contaminated and 

cannot be used for furniture, for construction, or even for firewood 

(Matukousky, 1990). 

In addition to the forests, the waters of the Pripyat, Sozh, Nevsich, 

Iput, Besyoad, Braginka, Kolpita, and Pokot Rivers are carrying radioactive 

silt into the Dnepr River. The entire grid of power stations on the Dnepr 

River down to the Black Sea is threatened with 60 million tons of 

radioactive silt, as are the 60 million people in these regions. 

The anticipatory stress of these people is readily apparent and can be 

easily understood. Interestingly, according to an assessment by 200 

scientists from 25 countries and 7 multinational organizations done for the 

United Nations International Atomic Energy Agency; stress-related illnesses 

are caused by lack of public information about the disaster and the mass 

evacuations that follow (I.A.E.A., 1991). 

In sum, the psychological stress 

that followed in the surrounding areas outside the radioactive hot zones was 

12

*wholly disproportionate to the biological significance of the radioactive 

contamination" (I.A.E.A., 1991). 

Even when no radiation is released from a nuclear accident, but only 

threatens, as was the case at Three Mile Island, the Kemeny Commission and 

other documents (Collins, 1984; Collins, 1991; Davidson, 1982, I.A.E.A., 

1991) concluded that mental stress would be the main effect. The 

psychological findings could be criticized if used alone, for their 

potential self-serving function. To avoid this perception, neurochemical 

analyses that measured individual stress values were employed (Collins, 

1983; Collins, 1991; Davidson, 1983). By using this multidisciplinary 

approach, we further clarified the adverse effects that exposure, or 

potential exposure to ionizing radiation has on humans. Consequently, since 

increased stress and associated behavioral alterations occur from 

anticipation as occurred at TMI; when actual exposure and deaths occur from 

radiation, the psychological and behavioral actions of the FSU residents are 

easily understood. 

Consequently, the situation in the FSU is likely to become of major 

concern in future years as the future of 60 million people is adversely 

affected by the Chernobyl accident, and another million people (+/-) are 

adversely affected by the three accidents in the Kyshtym and Chelyabinsk 

areas. The psychological, physiological and epidemiological implications of 

these disasters require further study. 

13

2:149-166, 1983. 

Collins, D. L. Persistence differences between Three Mile Island residents and 

a control group. AD-A145567, National Technical Information Service, 

Springfield, VA, 1984. 

Davidson, L. M., Baum, A., and Collins, D. L. Stress and control related 

problems sat Three Mile Island. Journal of Applied Psychology 

12:349-359, 1982. 

Pegteva, M. 0. Vice Director, Candidate of Technical Sciences, Institute 

of Biophysics of the USSR, Ministry of Health, Chelyabinsk Branch Office. 

Personal Conversation, November 1991. 

Douplenski, N. First Secretary, Commission of the USSR for UNESCO, 

Ministry of Foreign Affairs of USSR. Personal Conversation, September 1991. 

(He grew up in this region and still has family living in the area.) 

International Chernobyl Project, The, Technical Report. 

ISBN 92-0-129191-4, Int 

ernational Atomic Energy Agency, pp 277-413, 1991. 

Karpov, V. I. Director of Research, Academy of Sciences of the USSR, Center for 

International, Environmental Cooperation (INENCO), St. Petersburg, Russia. 

Personal Conversation, September 1991. 

15

Kedrovsky, 0. L. Professor, Director of the Institute, All-Union designing 

and research institute of production engineering. 

Personal Conversation, 

September 1991. 

Kosenko, M. M. Head of Clinical Department, Candidate of Medical Sciences, 

Institute of Biophysics of the USSR, Ministry of Health, Chelyabinsk Branch 

Office. Personal Conversation, November 1991. 

Lee, G. A. Chernobyl trial opens; accused officials blame design faults. 

The Washington Post, July 8, 1987, p. Al. 

Matukovsky, N. The Lessons of Chernobyl. Izvestia, Moscow, March 26, 1990, 

p. 3. 

Medvedev, Z. Bringing the skeleton out of the closet. 

Nuclear Engineering 

International 35(436):26-32, 1991. 

Nuclear Regulatory Commission: 

Investigation into the March 1979 Three Mile 

Island-2 Accident by the Office of Inspection and Enforcement, NUREG-0600. 

Nuclear Regulatory Commission, Washington, DC, 1979. 

Reuters News Service. Chernobyl rally attended by thousands. The New York 

Times, New York, April 27, 1990, p. A6. 

Shishchits, I. Leading Researcher, Scientific Secretary of Section of 

16

Scientific-Technical Council, USSR Nuclear Society, All-Union Design and 

Research Institute of Industrial Technology. 

Personal Conversation, 

September 1991 

Soyfer, V. N. Chairman Laboratory of Molecular Genetics, George Mason 

University, Fairfax, VA, Personal Conversation, November 1991, 

Tsaturor, Y. S. Duputy chairman, USSR State committee for Hydrometerology. 

Personal Conversation, September 1991. 

Wise, M. Z. U.N. report blames stress, not radiation, for Chernobyl illnesses. 

The Washington Post, Washington, DC, May 22, 1991, p. A25. 

17

ADDENUM 1 

During the course of these conversations at AFRRI, Professor Valery N. 

Soyfer, Chairman of the Laboratory of Molecular Genetics at George Mason 

University, Fairfax, VA, and a recent defector from the Soviet Union, said, 

"The Soviet Union was the first nation to develop the hydrogen bomb." 

then looked at me and said, "I bet you didn't know that, but we were." 

He 

So I 

said, "Well can you tell me about that?" 

and he said he couldn't say any 

more on that topic. 

ADDENDUM 2 

Dr. Alexander V. Akleev, Institute of Biophysics, U.S.S.R. Ministry of 

Health, Chelyabinsk Branch Office, summarized their research areas. 

They 

have been looking for a radioprotective drug for the last 15 years, but have 

not found anything that has worked. This unsuccessful effort involved 

looking at spleen extracts, compounds in herbs with low molecular weights, 

and peptides to defuse through the cell membrane. Reactor operators have 

been irradiated since they began their research back in the 1948-1950. The 

Institute is interested in late effects of low and medium doses; low doses 

are defined as 20 rads or less, and medium doses are 20 rads or greater. 

18

NORTH 

FIGURE 

Kfis/dn A'ccadett 

*Szire6&sk A 19z 

LEGENDri 

inMie

AD-A284 746 



DNA-TR-92-37-V1 

Chernobyl Doses 

Volume 1-Analysis of Forest Canopy Radiation Response 

from Multispectral Imagery and the Relationship to Doses 

DTIC 

Gene E. McClellan 

George H. Anno 

A ELECTE 

F. Ward Whicker 

Pacific-Sierra Research Corp. SEP 2 01994 

2901 28th Street 

Santa Monica, CA 90405-2938 F: 

September 1994 


VOriginal oontains oolor 

plates; All DTIC reproduot,. 

ions wuil be in bleaok an/ 

whites 

CONTRACT No. DNA 001-87-C-0104 

94-30186 Approved for public release; 

~II ~ 

unlimited. 

II 

!111 l! t Iill Idistribution Is 

94 W'%= D 

I2








ORGANIZATION. 

.0


This mailer is provided to enable DNA to maintain current distribution lists for reports• (We would 

appreciate your providing the requested information.) 

NOTE: 

1l Add the individual listed to your distribution list. Please return the mailing label from the 

document so that any additions, changes, 

F1 Delete the cited organization /individual, 

corrections or deletions can be made easily. 

For distribution cancellation or more 

El Change of address. information call DNA/IMAS (703) 325-1036. 

NAME: 

ORGANIZATION: 

OLD ADDRESS 


z 

I- 

Z 


ATTN: IMAS 

6801 TELEGRAPH ROAD 

ALEXANDRIA, VA 22310-3398 


ATTN: IMAS 


ALEXANDRIA, VA 22310-3398

DISCLAIMER NOTICE 

THIS DOCUMENT IS BEST 

QUALITY AVAILABLE. THE COPY 

FURNISHED TO DTIC CONTAINED 

A SIGNIFICANT NUMBER OF 

COLOR PAGES WHICH DO NOT 

REPRODUCE LFGIBLY ON BLACK 

AND WHITE MICROFICHE,.

REPOT DCUMNTATON AGEFor mApproved 

REPO UME T DO TATIN PA E OMB Noý 0704-0188 

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, 

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 

collection of information, including suggestions for reducing this burden, to Washington headquarters Services Directorate for informaton Operations and Reports. 1215 Jefferson 

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 


940901 Technical 870929- 930930 

4. TITLE AND SUBTITLE 5. FUNDING NUMBERS 


Volume 1-Analysis of Forest Canopy Radiation Response from Multispectral C - DNA 001-87-C-0104 

Imagery and the Relationship to Doses PE -62715H 

"6. AUTHOR(S) 

PR - RM 

TA - RH 

Gene E. McClellan, George H. Anno, and F. Ward Whicker 

WU - DH026130 

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION 

REPORT NUMBER 

Pacific-Sierra Research Corp. 

2901 28th Street PSR Report 2251 

Santa Monica, CA 90405-2938 

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING 



6801 Telegraph Road DNA-TR-92-37-Vl 


RAEM/Kehlet 


This work was sponsored by the Defense Nuclear Agency under RDT&E RMC Code B4662D RM RH 00038 

STRP 3500A 25904D. 

12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE 

Approved for public release; distribution is unlimited. 

13. ABSTRACT (Maximum 200 words) 

This volume of the report Chernobyl Doses presents details of a new, quantitative method for remotely sensing 

ionizing radiation dose to vegetation. Analysis of Landsat imagery of the area within a few kilometers of the 

Chernobyl nuclear reactor station provides maps of radiation dose to pine forest canopy resulting from the accident 

of April 26, 1986. Detection of the first date of significant, persistent deviation from normal of the spectral 

reflectance signature of pine foliage produces contours of radiation dose in the 20 to 80 Gy range extending up to 

4 km from the site of the reactor explosion. The effective duration of exposure for the pine foliage is about 3 

weeks. For this exposure time, the LD 50 of Pinus sylvestris (Scotch pine) is about 23 Gy. The practical lower 

dose limit for the remote detection of radiation dose to pine foliage with the Landsat Thematic Mapper is about 5 

Gy or 1/4 of the LD 5 0. 

'DTic QU~ALI77 ',,:-L: - ,., 


Chernobyl Dose 226 

Forest Damage Conifer Stress Fallout 16. PRICE CODE 

Change Detection Ionizing Radiation Multispectral Imagery 

17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19, SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT 



NSN 7540-280-5500 

Standard Form 298 (Rev.2-89) 

Prescribed by ANSI Sta 239-18 

290-102

UNCLASSIFIED 

SECURi'Y CLASSIFICAnION OF THiS PAGE 

CLASSIFIED BY: 

N/A since Unclassified. 



Accesion For 

NTIS CRA&i 

D TIG TAj3 

Unan n-ouliced 0 

JUSltifca~ioin 

............... 

* 

By 

Ava 1 Idbiiily Cod'es 

Dist Aviat & ,ndl/or 

Special 

MII 

SECURITY CLA SSIVICATIO'N OF THIS PACG 

UNCLASSIFIED

EXECUTIVE SUMMARY 

This volume of the report Chernobyl Doses presents details of a new, quantitative method for 

remotely sensing ionizing radiation dose to vegetation. 

The method uses a time series of 

multispectral images taken from an orbital or airborne platform to reveal changes in spectral 

reflectance of foliage that has been exposed to radiation. The threshold of detection is about 1/4 of 

the median lethal dose (LD 5 0 ) of the dominant plant species. 

Figure S-1 illustrates the sequence of events leading to detection of radiation dose. The 

effective duration of the vegetation exposure is determined by the radioactive decay rate of the 

fallout deposited on the foliage and the rate at which fallout is removed from the foliage by 

weathering. The accumulated dose during this exposure time causes biological damage at the 

cellular level in plant tissues. At high doses (several times the LD 5 0 ), radiation damage to multiple 

tissues kills vegetation in a short time, a matter of one or two weeks for pine trees. At lower 

doses, radiation damage to foliage is significant only for growth tissue at the tips of branches. 

Such damage takes more time to change the appearance of foliage. The dose dependence of the 

delay until the onset of observable foliage response provides the basis for remote detection of 

radiation dose. 

Figure S-2 shows a map of the foliage doses received by the pine forest canopy near the site of 

the Chernobyl nuclear power station as derived from analysis of a time series of eleven Landsat 

Thematic Mapper images spanning a period from one year before to two years after the explosion 

of the Unit 4 reactor on April 26, 1986. Table S- 1 shows the doses for the three contours drawn 

on the map as well as the dose range between contours and the total area within each contour. 

Individual pixels, representing 25 m squares of pine forest, are color-coded according to the 

legend on the map, which shows the image number of first-observed, persistent deviation from 

normal of the spectral signature of the pixel. Table S-2 provides a conversion of this image 

number to a radiation dose range. The dose is quoted as a range since the first observable response 

may have occurred at any time during the interval between the date of the image showing first 

response and the date of the previous image. Black areas on the map are either not pine forest or 

were cleared of pine forest before showing a radiation response. 

Gray pixels on the map 

correspond to pine forest that appears normal at the end of the two year observation period. 

Although not indicated on the map, some of these pixels showed a transient radiation response 

corresponding to doses less than about 20 Gy. 

The methodology developed during this effort and the resulting data contribute to an improved 

understanding of the effects of high levels of fallout radioactivity on vegetation and, especially, on 

the remote observation of radiation-induced foliage response and the extraction of dose estimates 

from those ohservations. The results aid in the understanding of the consequences to personnel of 

ini

Falu 

Deposition 

ProtractedFalu 

_ Radiation 

Exposure 

Removal by 

of Foliage 

Efet at the 

Early, Systemic Continuum Late, 

Effects at .,K- -_ -- > Growth-Related 

Higher Doses of Responses Effects at Lower 

Doses 

Remote Sensing 

Figure S-1. 

Sequence of events that enables remote detection of the exposure of 

vegetation to ionizing radiation and calculation of the dose from the 

time of resulting foliage changes relative to the start of exposure. 

iv

Figure S-2. Chernobyl radiation dose contours derived from foliage changes in pine forest canopy 

observed with the Landsat Thematic Mapper multispectrai imaging device; 1 kan grid lines 

originating at the Unit 4 reactor site. See Table S- I for dose values. 

V

in an area contaminated by radioactive fallout and the impact of vegetation on that 

Table S-1. Description of contours drawn in Figure S-2. 

Pine foliage response Dose range Cumulative 

time at contour Dose at contour within contour enclosed area 

ontour (days) (Gy) (Gy) (kin 2 ) 

inner (blue) 35 54 54 - 80 0.8 

Middle (green) 172 30 30- 54 2.9 

Outer (yellow) 499 20 20 30 14.5 

The final section of this report lists recommendations for improving and extending the results 

presented herein. The recommendation of primary importance is the establishment of a cooperative 

effort with scientists of the former Soviet Union to compare the satellite data with ground studies 

made at specific locations within a few kilometers of the Chernobyl power plant. The comparison 

of the image analysis with other data in Section 8 is encouraging but only qualitative because of the 

unavailability of data with the temporal and spatial resolution of the satellite images. We believe 

better data exists (Gamache, 1993); furthermore, some of the affected forest is probably still 

standing. Histological examination of exposed pine trees observed in the imagery would be of 

great value both for interpreting the imagery and interpreting the histological data. 

Table S-2. 

Pine foliage doses corresponding to first detected response at the times of 

the 9 postaccident images (Images 1 and 2 are preaccident). 

Time postaccident Dose 

Image number (days) (Gy) 

3 3 >133 

4 12 80- 133 

5 28 59-80 

6 35 54-59 

7 172 30-54 

8 220 28-30 

9 380 23-28 

10 499 21-23 

11 763 18-21 

vi

PREFACE 

This volume is the first of three volumes composing the final report to the Defense Nuclear 

Agency (DNA) for contract DNAOOI-87-C-0104, Chernobyl Doses. In addition to summarizing 

investigations carried out by Pacif~c-Sierra Research Corporation (PSR) under that contract, this 

volume presents the analytical work that connects satellite multispectral observations of pine forests 

around Chernobyl to the nuclear radiation dose received by the trees as a censequence of the 

reactor accident of 26 April 1986. Volume 2, Conifer Stress Near Cherno'?yl Derived from 

Landsat Imagery, describes the acquisition and processing of Landsat imagery of the area 

containing the Chernobyl Nuclear Reactor Station and presents the exploratory analysis of the 

imagery using PSR's proprietary HyperscoutTl change detection algorithm. Volume 3, Habitat 

and Vegetation Near the Chernobyl Nuclear Reactor Station presents a detailed exposition on the 

soils, climate, and vegetation of the Poles'ye region of the Ukraine and Belorussia with emphasis 

on the area around Chernobyl. This data provides background for interpretation of the satellite 

imagery. 

The authors wish to acknowledge Frank Thomas of PSR who pointed out the possibility of 

observing radiation effects on vegetation near Chernobyl with multispectral imagery. They also 

wish to recognize the considerable computational expertise of Leigh Matheson of PSR who 

implemented the image processing and analysis described in this report and the skillful manuscript 

preparation by Kathy Howell. Dr. Gerald Gamache was helpful in providing data he obtained in 

Ukraine. 

The authors wish to acknowledge the technical monitor of this project, Robert W. Young of 

DNA's Radiation Policy Division, for his support and encouragement during this work. 

Dr. Young was assisted first by Major Bruce West and then by Major Robert Kehlet. The authors 

also wish to acknowledge Dr. Marvin Atkins and Dr. David Auton of DNA whose interest made 

this work possible. 

rMHyperscout is a trademark of Pacific-Sierra Research Corporation. 

vii


Conversion factors for U.S. customary to metric (SI) units of measurement 

To Convert From To Multiply 

angstrom meters (m) 1.000 000 X E-I0 

atmosphere (normal) kilo pascal (kPa) 1.013 25 X E.2 

bar kilo pascal (kPa) 1.000 000 X E÷2 

barn meter 2 (M 2 ) 1 000 000 X E-28 

British Thermal unit (thermochemical) joule (J) 1.054 350 X E+3 

calorie (thermochemical) joule (J) 4.184 000 

cal (thermochemical)/cm 2 mega joule/m 2 (MJ/m 2 ) 4.184 000 X E-2 

curie giga becquerel (GBq)l 3.700 000 X E+I 

degree (angle) radian (rad) 1.745 329 X E-2 

degree rahrenheit degree kelvin (K) tK=(t0f + 459.67)/1 .8 

electron volt joule (j) 1 .02 19 X L-19 

erg joule (i) 1.000 000 X E-7 

erg/second watt (W) 1.000 000 X E-"' 

foot meter (m) 3.048 000 X E-I 

foot-pound-force joule (J) 1.355 818 

gallon (U.S. liquid) meter 3 (, 3 ) 3.785 412 X E-3 

inch meter (m) 2.540 000 X E-2 

jerk joule (J) 1.000 000 X E+9 

joule/kilogram (J/Kg) (radiation dose 

absorbed) Gray (Gy) 1.000 000 


kip (1000 lbf) newton (N) 4.448 222 X E+3 

kip/inch 2 (ksi) kilo pascal (kPa) 6.894 757 X E+3 

ktap newton-second/mr 2 (N-s/m 2 ) 1.000 000 X E+2 

micron meter (m) 1.000 000 X E-6 

rail meter (M) 2.540 000 X E-5 

mile (international) meter (m) 1.609 344 X E+3 

ounce kilogram (kg) 2.834 952 X E-2 

pound-force (lbf avoirdupois) newton (N) 4.448 222 

pound-force inch o'ton-meter (N'm) 1.129 848 X E-I 

pound-force/inch newton/meter (N/m) 1.751 268 X E+2 

pound-force/foot 2 kilo pascal (kPa) 4.788 026 X E-2 

pound-force/inch 2 (psi) kilo pascal (kPa) 6.894 757 

pound-mass (Ibm avoirdupois) kilogram (kg) 4.535 924 X E-I 

pound-mass-foot 2 (moment of inertia) kilogram-meter 2 (kg-m 2 ) 4.214 011 X E-2 

pound-mass/foot kilogram/metr3 (kg/rn 3 ) 1.601 846 X E+1 

rad (radiation dose absorbed) Gray (Gy)°" 1 .000 000 X E-2 

roentgen coulomb/kilogram (C/kg) 2.579 760 X E-A 

shake second (s) 1.000 000 X E-S 

slug kilogram (kg) 1.459 390 X E+1 

torr (mm Hg, 0 0 C) kilo pascal (kPa) 1.333 22 X E-1 

*The becquerel (Bq) is the SI unit of radioactivity; Bp 

"*The Gray (Gy) is the SI unit of absorbed radiation. 

1 event/s. 

viii

CONTENTS 

Section 

SUM M A R Y ............................................................ ........ 

PREFACE ............................................... 

CONVERSION TABLE ........................................................ 

F IG U R E S ........................................................................ 

TAB LES .................................................................. ... xvi 

Page 

vii 

viii 

ix 

I INTRODUCTION ............................................................... I 

2 REMOTE FALLOUT DETECTION THROUGH IMAGERY ANALYSIS 5 

3 HABITAT AND VEGETATION NEAR THE CHERNOBYL 

NUCLEAR POWER PLANT ............................................... 11 

3.1 Regional Description ................................................ . 11 

3.1.1 G eography ................................................... . 11 

3.1.2 Soils ........................................................... 14 

3.1.3 C lim ate .................................................. ..... 15 

3.1.4 Native Vegetation ........................................... 16 

3.1.5 C rops ......................................................... 19 

3.2 Tasseled Cap Transformation of Landsat Imagery ................ 19 

3.2.1 Description of the Tasseled Cap Transformation ........ 19 

3.2.2 Tasseled Cap Images of the Analysis Area ............... 23 

3.3.2 Unsupervised Clustering of Evergreen Pixels ........... 38 

3.3 Preliminary Classification of Evergreens .......................... 35 

3.3. 1 Selection of Evergreen Pixels .............................. 35 

3.3.2 Unsupervised Clustering of Evergreen Pixels ........... 38 

3.3.3 Identification of Evergreen Classes ............. 41 

3.4 Reference Sites for Pine Forest Classes ............................. 43 

3.4.1 

3.4.2 

Avoidance of Clouds and Haze ........................... 

Representative Spectral Signatures ....................... 

45 

58 

3,. Sites C hosen ................................................. 58 

3.4.4 Other Impacts of Clouds and Haze ....................... 58 

3.5 Preaccident Pine Forest Classification Map ................... 59 

4 RADIONUCLIDE FALLOUT FROM THE CHERNOBYL ACCIDENT 65 

4.1 Release And Deposition ............................................... 65 

4.2 Radionuclide Composition ........................................... 65 

5 RADIOBOTANICAL AND DOS IMETRIC CONSIDERATIONS ......... 67 

5.1 

5.2 

Tree Characteristics Relevant To Radiation Damage .............. 

Beta Radiation Exposure Of Pine Meristems ....................... 

67 

69 

5.3 Beta Versus Gamma Exposure ....................................... 70 

5.3. i Values from the Literature .................................. 70 

ix

CONTENTS (CONTINUED) 

Section 

Page 

5.3.2 Calculations .................................................. 71 

5.5 

Dose Rate Effects ..................................................... 

Dose Rate Scenario for Chernobyl .................................. 

78 

79 

5.6 Sum m ary ............................................................... . 84 

6 DOSE-RESPONSE RELATIONSHIPS FOR PINE TREES ............... 87 

6.1 Literature Review ...................................................... 87 

6.2 Analysis Of Dose Rate Data .......................................... 91 

6.3 Temporal Progression of Damage .................................. 93 

6.4 Dose Versus Time-to-Response for Multispectral Detection ...... 94 

7 DOSE DETERMINATION FOR CHERNOBYL FORESTS ............... 103 

7.1 Spectral Deviation from Class ........................................ 103 

7.1.1 Mahalanobis Distance and the Mahalanobis Spectral 

Deviation Vector ............................................. 104 

7.1.2 Scaled Mahalanobis Distances and Vectors ............... 104 

7.1.3. Spectral Deviations for the Chernobyl Images ........... 108 

7.2 Deviations Caused By Forest Clearing .............................. 113 

7.3 Time-to-response from Imagery ..................................... 115 

7.4 Foliage Dose M aps .................................................... 120 

7.4.1 Dose Map with Hand-Drawn Contours ................... 120 

7.4.2 Iterative Smoothing of the Dose Map ..................... 122 

7.5 S um m ary ................................................................ 129 

8 D ISC U S SIO N ..................................................................... 131 

8.1 Comparison with Measurements of Forest Damage .............. 131 

8.2 Comparison with Aerographic Surveys of Dose Rate ............. 132 

8.2. 1 1-Meter Gamma Dose Rates for the Close-in Area ...... 132 

8.2.2 

8.2.3 

Close-in Foliage and 1-Meter Gamma Doses ............ 

Calculated Ratio of Foliage to 1-Meter Gamma Dose... 

135 

137 

8.3 Comparison with Human Exposure Data ........................... 139 

8.4 Uncertainties in the Image Analysis ................................. 141 

9 CONCLUSIONS AND RECOMMENDATIONS ............................ 143 

10 LITERATURE CITED .................................... 145 

Appendix 

A FALLOUT DOSE CALCULATIONS FOR PINE FOREST CANOPY... A-1 

B SPECTRAL DEVIATIONS FROM CLASS .................................. B-1 

C CHRONOLOGY OF SELECTED EVENTS .................................. C-1 

D EVERGREEN SPECTRAL SIGNATURES BY CLASS AND DATE .... D-1 

x

FIGURES 

Figure 

S-i 

Sequence of events which enables remote detection of the exposure 

of vegetation to ionizing radiation and calculation of the dose from 

the time of resulting foliage changes relative to the start of exposure ........ 

Page 

iv 

S-2 Chernobyl radiation dose contours derived from pine forest canopy 

responses as observed with the Landsat Thematic Mapper multispectral 

imaging device; i km grid lines originating at the Unit 4 reactor site ........ 

V 

1-1 Immediate vicinity of the Chernobyl nuclear power station with Unit 4 

reactor (circled) still burning .................................................... 3 

2-1 Typical coloration of dying pine foliage. Landscape setting ................. 6 

2-2 Forest stress map for 8 May 1986 as presented in Volume 2. 

Accident-affected area circled. White cross is Unit 4 reactor site ............ 8 

2-3 Method of radiation dose determination from remote multispectral 

sensing of pine forest ................................................................ 9 

3-1 Regional map encompassing the 30 kilometer danger zone around the 

Chernobyl nuclear accident. (Reference ..... ) ................................ 12 

3-2 Landsat image of the confluence of the Dnieper and Pripyat Rivers 

at the upper end of the Kiev Reservoir. The sharply outlined black 

footprint shape is the reactor station cooling pond along the Pripyat 

River. [Thematic Mapper false color presentation (R,G,B) = (7,4,1), 

6 June 1985, geocoded, 1 degree longitude by 0.5 degree latitude.] ....... 13 

3-3 Illustration of the Tasseled Cap spectral transformation for Landsat 

Thematic Mapper images; transformation coefficients are listed 

in T able 3-2 ......................................................................... 2 1 

3-4 Tasseled Cap false color images, (R,G,B) = (Br,Gr,Wt) of 38.4 km 

square area around Chemobyl: a) early summer image one year 

preaccident and b) late winter image 5 weeks preaccident .................... 24 


square area around Chernobyl: a) 3 days postaccident and b) 12 days 

postaccident ........................................................................ . 25 


square area around Chemobyl: a) 4 weeks postaccident and b) 5 weeks 

postaccident ......................................................................... 27 

xi

FIGURES (CONTINUED) 

Figui-e 

Page 


square area around Chernobyl: a) 5.6 months postaccident 

and b) 7.2 months postaccident .................................................. 29 


square area around Chernobyl: a) I year postaccident and b) 1.4 years 

postaccident ......................................................................... 31 

3-9 Tasseled Cap (TC) false color images of 38.4 km square area around 

Chernobyl 2.1 years postaccident: a) (R,G,B) = (Br,Gr,Wt) as in 

Figures 3-4 through 3-8 and b) (R,G,B) = (Hz,TC5,TC6), the last 

3 components of the TC transformation ......................................... 33 

3-10 Evergreen vegetation appears bright green and deciduous and 

annual vegetation appears magenta in this false color presentation 

of COMP21 with (R,G,B) = (5,2,5) ............................................ 37 

3-11 ?rocedure for classifying evergreen vegetation ................................ 39 

3-12 a) Like Figure 3-10 except color reversal (R,G,B) = (2,5,2) displays 

deciduous or annual vegetation in bright green and evergreens in 

magenta, and b) areas passing the evergreen criteria of Figure 3-11 

show n in green ..................................................................... 40 

3-13 Spectral signatures of selected classes from the unsupervised clustering 

of evergreen pixels ................................................................. 42 

3-14 Procedure for selecting one good reference site for each 

coniferous/evergreen class ....................................................... 44 

3-15 Cloud/haze enhancement for Date 1 for the 38.4 km square analysis 

area, 46 (R,G,B) = (Th,Th,Hz). Clouds appear blue, warm areas 

yellow. Thermal gradient appearing as variation in yellow shade 

of the cooling pond shows counterclockwise flow of water around 

central barrier ...................................................................... 46 

3-16 a) Reference (control) site locations for Classes 3, 4, 5, 6, and 8 in 

the 38.4 km analysis area and b) cloud/haze enhancement for Date I 

showing three polygons used for classification merger described in 

Section 3.5. Red dots are reference site locations ............................. 47 

xii


Figure 

Page 

3-17 Cloud/haze enhancements for a) Date 2 and b) Date 3. Only Date 3 has 

clouds. Red dots are reference site locations ................................... 49 


clouds. Red dots are reference site locations .................................. 51 


clouds. Red dots are reference site locations ................................... 53 

3-20 Cloud/haze enhancements for a) Date 8 and b) Date 9. Only Date 9 

has clouds. Reference Sites 5, 6 and 8 have been moved on Date 9 

to avoid a faint jet contrail ......................................................... 55 

3-21 Cloud/haze enhancements for a) Date 10 and b) Date 11. Neither date 

has clouds, but Date 10 has a jet contrail. Red dots are reference 

site locations ........................................................................ 57 

3-22 Procedure for generating final preaccident classification of pine forest ...... 60 

3-23 Final preaccident classification of pine forest for the 12.8 km square 

area analyzed in Volume 2. Class numbers according to Table 3-6 ......... 62 

3-24 Final preaccident classification of pine forest for the full 38.4 km 

square analysis area. Class color code same as in Figure 3-23 .............. 63 

5-1 An example of Pinus sylvestris. A six-foot specimen cut and 

photographed in December in Maryland ......................................... 68 

5-2 Illustration of fallout distribution described by a foliar interception 

fraction with neglect of winds ................................................... 72 

5-3 Beta to gamma dose ratio versus contact source density relative 

to unit areal density of incident fallout. Curves are for various 

radii of foliage elements located at the top of the canopy ...................... 76 

5-4 The beta to gamma dose (dose rate) ratio at the center of cylindrical 

foliage elements of various radii at a) the top of the canopy and 

b) the middle of the canopy ....................................................... 77 

xiii


Figure 

Page 

5-5 Calculated dose rates (normalized to 1.0 at t = I day) from gamma 

radiation emanating from fallout on the soil and from beta activity 

on the foliage ............................................................ ......... 82 

5-6 Integral of the beta dose rate curve in Figure 5-4 and a constant rate 

curve that should produce a similar biological response ...................... 85 

6-1 Relationship of LD 5 0 doses at various constant rate (CR) exposure 

times in hours to those for 16-hour CR exposures ............................ 88 

6-2 The relationship of total dose to duration of constant rate exposure 

for various endpoints of damage to pines ....................................... 92 

6-3 The time required to reach the LD100 or GR50 versus short-term 

(8-30 day) dose. Circles represent data for the LD 100 ....................... 95 

6-4 Detectability of the radiation response of pine trees using multispectral 

im agery ............................................................................. . 98 

6-5 Regression line with 68% confidence band for the relationship between 

dose and time of earliest detection relative to the start of exposure for 

two- to four-week, springtime exposures ....................................... 101 

7-1 Two-dimensional illustration of the cluster of points formed by the 

pixel intensity vectors of a forest reference site. See Appendix B for 

the normalization procedure used to express deviations of pixels from 

the cluster center in standard units ............................................... 105 

7-2 The cluster of scaled Mahalanobis vectors for a pine forest class 

reference site using the first three features of the Tasseled Cap 

spectral transform ation ............................................................ 107 

7-3 Color graphic presentation of the scaled Mahalanobis deviation vectors 

for pine forest on Dates I to 5 (194 by 177 pixel area) ........................ 109 

7-4 Color graphic presentation of the scaled Mahalanobis deviation 

vectors for pine forest on Dates 6 to I 1 (194 by 177 pixel area) ............. 

IIl 

7-5 Forest cleared (bulldozed) after the Chemobyl accident is color coded 

by the date number of the first observed clearing .............................. 114 

xiv


Figure 

Page 

7-6 Algorithm for determination of time-to-response for radiation-damaged 

foliage on a pixel-by-pixel basis .................................................. 117 

7-7 Date of onset of persistent deviation from normal of pine forest 

pixels according to algorithm of Figure 7-6 ..................................... 119 

7-8 Time to first observed response with 1 km grid and three hand-drawn 

dose contours. See Table 7-2 for dose ranges .................................. 121 

7-9 Series of intermediate stages leading to a smoothed map of date 

of first observed response by numerical relaxation as described 

in the text ............................................................................ 123 

7-10 Smoothed dose contours from pine foliage response using numerical 

relaxation of dose values from individual pine forest pixels inside the 

w hite border ........................................................................ 125 

7-11 Smoothed dose contours from Figure 7-10 displayed with a gray-scale 

background of the western end of the Chernobyl nuclear power station .... 127 

8-1 Dose rate measurements for close-in fallout contaminated areas for two 

different dates ....................................................................... 133 

8-2 Change in gamma dose rate from radioactive materials in the close-in 

zone as a function of time based on aerographic surveys (data from 

A sm olov et al., 1987) ............................................................. 135 

8-3 Accumulated close in fallout dose (1-m above ground) in I and 

4 km 2 areas ......................................................................... 136 

8-4 Ratio versus time of the accumulated dose (b and g) at the center of 

cylindrical foliage elements at the top of pine forest canopy to the 

accumulated 1-m gamma dose a) under the canopy and b) in an open 

field for the same total fallout deposition ........................................ 138 

xv

TABLES 

Table 

Page 

3-1 Original Landsat-5 TM Tasseled Cap coefficients (Christ et al., 1986) ..... 20 

3-2 Modified Tasseled Cap coefficients used in the present work to place 

all feature intensities in the range 0 to 25......................... 22 

3-3 Landsat scenes analyzed and time of scene relative to reactor explosion .... 23 

3-4 Coordinates (UTM, Zone 36) of the upper left comers of the comer 

pixels of the 1536 x 1536 pixel area analyzed in this report .................. 34 

3-5 Band structure for the winter/summer composite image, COMP21, 

generated from Dates 2 and 1 ..................................................... 36 

3-6 Pine forest classes ................................................................. 61 

5-1 Ratio of beta to gamma doses to foliage calculated from the results 

of Appendix A for fallout retained in the canopy ............................... 73 

5-2 Dose rate per unit source for various sources and dose points. 

Reference canopy volume source density is taken equal to the unit 

ground source density spread over the assumed .............................. 75 

5-3 Calculation of normalized beta dose rate (BDR) versus time assuming 

all dose from beta particles on foliage, all fallout on t = I day ................ 80 

5-4 Calculation of normalized gamma dose rate (GDR) versus time 

assuming all dose from gamma emitters on the ground and all 

fallout on t = I day ................................................................ 83 

6-1 Estimated total doses to produce three endpoints of damage to pines 

for an exposure period of three weeks (assumed equivalent to the 

effective exposure period at Chernobyl) ........................................ 93 

6-2 Detectability by multispectral remote sensing of radiation damage to 

pine trees for spring exposures as a function of time since start of 

exposure and dose ................................................................. 96 

6-3 Dose estimates according to Equation 6.1 for a first detected response 

corresponding to the times of the 9 postaccident images presented 

in this report ........................................................................ 100 

xvi

TABLES (CONTINUED) 

Table 

Page 

7-1 Area of pine forest near the Chernobyl nuclear reactor station cleared 

by the listed image date. Clearing of other vegetation not included ......... 115 

7-2 Description of contours drawn in Figure 7-8 ................................... 120 

7-3 Dose bands and affected areas for the smoothed radiation contour maps 

of Figures 7-10 and 7-11 .......................................................... 128 

8-1 Results of forest damage by radiation for an area with edge about 1 km 

from the site of the Unit 4 reactor explosion (Sobdovych et al., 1992, 

courtesy of Gam ache, 1993) ..................................................... 131 

8-2 Estimation of initial dose rates for contours enclosing specified areas ....... 134 

8-3 The ratio of the pine canopy foliage dose to the gamma dose I m off the 

ground using satellite image and aerographic survey data ................... 137 

8-4 Summary of typical b/g dose ratios for four groups of people who 

suffered radiation lesions in the skin at Chernobyl (Barabanova and 

O sanov, 1990) ...................................................................... 140 

xvii

SECTION 1 

INTRODUCTION 

During an extreme radiation accident such as occurred at the Chernobyl nuclear power station 

on April 26, 1986, explosion and fire may cause lofting of radioactive aerosols and vapors from a 

few hundred meters to more than a kilometer into the air. Winds then transport the material away 

from the accident site. The aerosols settle to the ground, larger, heavier particles nearby and 

smaller, lighter ones farther away. The resulting terrestrial deposition of fallout particles occurs in 

a more or less continuous but irregular fashion, following local and regional weather patterns. The 

finest aerosols enter the global atmospheric circulation. 

Much attention has been given to the regional and global patterns of fallout from the Chernobyl 

accident because of worldwide concerns for human radiation exposure and the entry of radio 

nuclides into the food chain. The primary health concern is the induction of cancers in the human 

population from the resulting low level radiation exposures. The doses at regional and global 

distances are far below the levels required to induce acute radiation sickness or cause visible 

changes in vegetation. 

Locally, however, at the site of the reactor explosion and more than a kilometer downwind 

from the site, radiation doses from radio nuclide fallout were above lethal levels for humans. 

Symptoms of radiation sickness occurred in some accident victims within the first hour of 

exposure (Young, 1988). Early fatalities and causes of death have been reviewed as part of this 

project and described in an earlier report (Laupa and Anno, 1989). 

The circumstance of one victim is of particular interest in the context of the analysis in this 

report. According to Barabanova and Osanov (1990), this person was 1.0 km downwind from the 

reactor at the time of the explosion, remained there for about an hour, and was exposed to both the 

radioactive plume and particles of fallout. According to Barabanova and Osanov, the individual 

was covered with black dust and received an estimated total gamma radiation dose of 12.7 Gy. 

The estimated beta radiation dose to the skin of his scalp, neck and upper body was 250-360 Gy at 

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 

exposure. 

As detailed in Section 6 of this report, the gamma dose alone for this victim is comparable to 

the median lethal dose (LD 50 ) for pine trees, which are only slightly less radiosensitive than 

humans. Beta dose will contribute further to vegetation damage depending on the depth of sensitive 

plant tissues. In addition, vegetation remains in the contaminated area and accumulates dose over a 

much longer exposure time than would mobile human beings who leave the area and are 

decontaminated.

Figure 1-1 shows the Chernobyl nuclear power station and immediate vicinity. The burning 

Unit 4 reactor (dark red dot) is circled and the approximate wind direction at the time of the reactor 

explosion is indicated by the arrow. The image was produced by merging data from a Landsatl 

Thematic Mapper image from April 29, 1986 with data from a SPOT 2 panchromatic image from 

May 1, 1986. The merged image combines the 10 m spatial resolution of the SPOT image and the 

wider spectral coverage of the 25 rn resolution Landsat image. The area shown in the image is a 

square with 5 km sides oriented along the Landsat orbital path. The edge of the forest at the tail 

end of the arrow in Figure 1-1 is only a little over 1 km from the burning reactor, so the victim 

described by Barabanova and Osanov was near the forest. Exposure of the victim is presumably a 

lower limit to the exposure of the nearby forest. 

Volume 3 of this report (Painter and Whicker, 1993) presents a description of the geography 

and vegetation of the Poles'ye region of Ukraine and Belarus where the Chernobyl nuclear power 

station is located. As is true in many other regions of the world, pines trees are the most 

radiosensitive (see, for example, Whicker and Fraley, 1974) of the large, widely occurring species 

of vegetation in this region. Thus, substantial effects on pine forest near the Chernobyl accident 

site are to be expected. Because the radiosensitivity of pines is comparable to that of humans, the 

study of such pine tree responses in a fallout radiation field, especially through remote sensing, is 

relevant to human operations in a radiation environment. 

The area around the Chernobyl nuclear power station after the accident of 26 April 1986 

provides a unique opportunity to observe the response of large scale plant communities to fallout 

radiation. In a review article, Whicker and Fraley (1974) quote only two studies involving realistic 

fallout exposures of plants. One (Murphy and McCormick, 1971) involved spreading feldspar 

particles coated with 90 Y on small plant communities on granite outcrops in the southeastern 

United States. The other (Rhoads and Platt, 1971) observed damage to desert vegetation receiving 

fallout from two small cratering explosions at the Nevada Test Site. Neither of these studies 

encompassed the size and variety of plant communities present at Chernobyl. No other studies 

involve the appropriate mix of beta and gamma exposure to foliage. 

1Landsat is the United States' civil land remote sensing satellite system. Data is obtained from the Earth 

Observation Satellite (EOSAT) Company of Lanham, Maryland. 

2 The Syst•me Probatoire d'Observation de la Terre (SPOT) satellite is operated by the French space agency 

Centre National d'Etudes Spatiales. Data is obtained from the SPOT Image Corporation, a U.S. subsidiary of the 

French company SPOT IMAGE. 

2

Figure 1-1. Immediate vicinity of the Chernobyl nuclear power station with Unit 4 reactor (circled) 

still burning; arrow shows approximate wind direction when the reactor exploded. 

3

Time plays several essential roles in the radiation response of a plant community (Whicker and 

Fraley, 1974), including: 

1) duration of exposure (or dose rate), 

2) season of exposure, 

3) time for manifestation of injury, 

4) repair and recovery times, and 

5) time for development of secondary effects such as community succession. 

Control areas are of primary importance in analyzing these time factors and in separating radiationinduced 

changes from naturally occurring ones. There should be multiple observations over at 

least one annual cycle of both the control and exposed vegetation. In addition, preirradiation 

observations of the experimental area are required to establish the initial condition of the exposed 

vegetation. 

Fortunately, satellite multispectral imagery provides observations of both control areas and 

preirradiation observations of the area affected by the Chernobyl reactor explosion. Control areas 

are available within individual Landsat scenes outside the highly irradiated areas and are an integral 

feature of our analysis. Furthermore, both summer and winter preaccident scenes of the irradiated 

area are included. Finally, the 16-day revisit time for acquisition of Landsat scenes from the same 

satellite on the same path offers ample opportunity to obtain multiple cloud-free images over an 

annual cycle. The analysis reported here uses 7 images during the first year postaccident and 2 

more during the following year. This time series of remote observations is a unique record of the 

spectral response of pine foliage to fallout radiation exposure. Section 2 outlines the procedure 

used in the remainder of the report to estimate dose to pine foliage from these images. 

4

SECTION 2 

REMOTE FALLOUT DETECTION THROUGH IMAGERY 

ANALYSIS 

A substantial amount of forest within a few kilometers of the Chernobyl nuclear power station 

was heavily contaminated with radionuclides by the April 26, 1986 explosion of the Unit 4 reactor 

and the ensuing fire. Radiation doses to conifers in some areas were sufficient to cause 

discoloration of needles within four to five weeks. Other areas, receiving smaller doses, showed 

foliage changes six months to a year later. Although we do not have color photographs of the trees 

affected by the accident, Figure 2-1 shows typical dying pine foliage from landscape plantings in 

northern Virginia. One photo shows branches with needles dying only at the tips. The other 

shows part of a tree whose needles are completely dead except for a few isolated bunches of green 

needles. Healthy trees are located in the background. Progression from green to yellow-green to 

reddish orange to dry brown is typical of the death of pine foliage from many causes including 

exposure to ionizing radiation. 

Multispectral imagery available from satellite sensors is especially suited for remote monitoring 

of such changes in vegetation since the changes affect both the visible and infrared reflectivity of 

foliage. A series of Landsat Thematic Mapper images spanning two years after the Chernobyl 

accident are analyzed for significant, accident-induced change in a companion document, Volume 2 

(McClellan et al., 1992). Accident-related changes in foliage are apparent three days after the 

accident. Changes due to cleanup activities and the progression of foliage deterioration are present 

in images taken in the following months. Finally, delayed effects of low radiation doses on initially 

undamaged foliage were still appearing more than a year after the accident. 

Pacific-Sierra Research Corporation's Hyperscout" m algorithm provided the analytical basis in 

Volume 2 for demonstrating significant, radiation-induced change in the forests around Chernobyl. 

The Hyperscout algorithm will detect image-to-image change, providing a measure of significance 

of change on a pixel-by-pixel basis. The algorithm analyzes reflectivity and emission changes 

occurring in the detection bands of the imaging device. Spectral signature, spatial distribution, and 

time dependence of the observed changes help identify stresses causing the change. The algorithm 

is especially useful when the induced change has a low signal-to-noise ratio and when general 

image variations such as seasonal illumination and annual vegetation cycles cause change that is not 

of interest. It operates on pairs of spatially registered images and is capable of useful results even 

Hyperscout is a trademark of Pacific-Sierra Research Corporation. 

5

Figure 2-1. Typical coloration of dying pine foliage. Landscape setting; upper photo, branches 

almost completely brown with healthy trees in background; lower photo, only needles at tips of 

branches are brown. 

6

when comparing images from different detectors. The sensitive nature of the Hyperscout 

algorithm enabled delineation of both the earliest changes in heavily exposed forest and the latedeveloping 

deterioration of less exposed forest. 

Figure 2-2 shows an example of one of the Hyperscout stress maps (12 days postaccident) 

from Volume 2, with the area of affected forest nearest the reactor site circled. The Unit 4 reactor 

is marked with a white cross. The site marked is the brightest pixel of the fire visible in the 

Landsat image at 3 days postaccident (see Figure 1-1). 

With guidance from the changes detected by the Hyperscout algorithm, this volume presents 

quantitative estimates of radiation dose to the pine foliage near the Chernobyl nuclear power station 

derived from remote Landsat imagery. The physical basis for detecting foliage stress with Landsat 

imagery is presented in Volume 2 and will not be repeated here. Quantitative dose estimation for 

the exposed conifers requires extracting the time of onset of observable radiation-induced foliage 

damage from the images for each pine forest pixel and combining this data with the relationship 

between dose and time of onset (time-to-response) of conifers for ionizing radiation. 

Figure 2-3 outlines the remote sensing method for dose determination from multispectral 

images of pine forest. Similar considerations apply to other vegetation; however, pine trees are 

among the most radiation sensitive species of widely occurring vegetation. The first step indicated 

in Figure 2-3 is the determination of preaccident vegetation classes in the multispectral imagery, 

particularly, the identification of pixels that consist predominantly of pine trees. Generally, forested 

areas are readily apparent from their color, texture and location in false color images as presented in 

Volume 2. However, to optimize data quantity and quality it is necessary to classify individual 

pixels. Section 3 of this volume describes the procedure for quantitative determination of class 

membership on a pixel-by-pixel basis. Since a primary problem is the discrimination among 

coniferous, deciduous, and mixed coniferous-deciduous pixels, we use preaccident images from 

both summer and winter to achieve a reliable forest classification. 

The extent to which tree foliage excludes a view of the ground from above is called canopy 

closure. If closure is high then satellite images will see mostly foliage. If closure is low then rocks, 

soil, and understory vegetation will dominate the image. We used a preaccident, late winter image 

(21 March 1986) with extensive snow cover to assist selection of areas with a high canopy cover 

of pine trees. The snow is a good discriminate for areas with low canopy cover or areas with a 

substantial number of deciduous trees. By eliminating these areas, we obtain a set of pixels 

containing a high canopy cover of pines that increases confidence in our dose estimates for these 

areas. Also, in this late winter image, the Pripyat River and its tributaries are mostly frozen. The 

resulting bright snow and ice along the river basin eliminates substantial misclassification that 

7

S128 256 

Stress Index 

Figure 2-2. Forest stress map for 8 May 1986 as presented in Volume 2. Accident-affected area 

circled. White cross is Unit 4 reactor site; Zone 36 Universal Transverse Mercator (UTM) 

coordinates about X = 298,275 m and Y = 5,697,175 m. 

8

of PracPreaccident 

e C e aImages 

Landsat 

Reference Site "X Pre- and 

Spectral urignatures Postaccident 

Ve sImage Date Landsat Images 

Noml Pixels i.se , \ ,- . 

[ Map for Pine Forest V ersus radiation 

Figure 2-3. Method of radiation dose determination from remote multispectral sensing of 

pine forest. 

9

occurred in our preliminary analysis (Volume 2) using only the summer image (6 June 1985) for 

determination of pine forest classes. In that analysis, many mixed pixels of dark water and 

deciduous vegetation at the rivers edge mimic the spectral response of pine forest . These 

misidentified pixels lead to the spurious indications of stress seen along the river in Figure 2-2. 

The classification of pine forest is based on the location of a reference site for each class that 

contains on the order of one hundred contiguous pixels belonging to the class. The reference sites 

are defined by polygons at fixed geographic locations. These locations must be far enough from 

the reactor to be unaffected by radiation and must not be obscured by clouds or haze on any image. 

The mean spectral signature and its covariance matrix for each class on each date is determined 

from the pixels in these reference sites. This procedure is described in Section 3. 

For many biological responses to radiation exposure, there is a dose-dependent delay between 

exposure and onset of the response. Such is the case for foliage damage in pine trees. We use the 

deviation of the spectral signal of each pine forest pixel from its class mean to detect significant 

change in foliage and, hence, estimate a time-to-response for areas of forest affected by the 

accident. Section 7 presents our time-to-response results. 

The biological endpoint for these satellite observations is change of spectral reflectivity of the 

foliage. However, the endpoints most commonly reported in plant radiobiology literature are 

100% lethality and 50% reduction in growth rate. There are no published reports, as far as we 

know, that provide controlled measurements of the spectral response of foliage to radiation 

exposure that can be directly correlated with other reported endpoints. Fortunately, though, there 

are sufficient visual observations reported for pine trees and sufficient reports of other vegetation 

stress response- to establish a plausible connection between pine tree response and multispectral 

detection. Section 6 discusses the connection made between reported endpoints and their 

relationship to spectral appearance from remote sensing. A functional relationship between timeto-response 

as determined from visible and infrared imagery and foliage dose is deduced. 

Finally, in Section 7 the time-to-response map from the imagery and the relationship between 

radiation dose and time-to-response are combined to generate a map of estimated radiation dose to 

pine trees in the vicinity of the reactor explosion. 

10

SECTION 3 

HABITAT AND VEGETATION NEAR THE CHERNOBYL 

NUCLEAR POWER PLANT 

This section summarizes the geography, soils, climate, native vegetation, and crops of the 

region surrounding Cherrobyl and provides details on the analysis of Landsat imagery leading to 

the delineation of four classes of pine forest within a 38.4 km square area centered approximately 

on the Chernobyl nuclear power station. 

Figure 3-1 is a regional map of the border area between the former Soviet republics of Ukraine 

and Belorussia (Belarus) stretching from Kiev in Ukraine to Gomel in Belorussia. The city of 

Chernobyl lies on the Pripyat River near the northwest comer of the Kiev Reservoir. The city of 

Pripyat is about 15 km further up the Pripyat River adjacent to the Chernobyl nuclear power 

station. The power station and its cooling pond are indicated by a small black rectangle labeled 

"Site of Chernobyl power station." Figure 3-1 provides the geographical framework for the 

regional description in Section 3.1 below. 

Figure 3-2 is the earliest of the series of Landsat images analyzed in this report. It shows a 

north/south oriented rectangle about 58 km by 72 km. The upper left, or northwest, corner of the 

rectangle is clipped because the desired area was near the edge of the Landsat path on this date. 

The angle of the clipped portion corresponds to the angle of the satellite orbit relative to lines of 

longitude at this location. The Dnieper River and the Pripyat River both empty into the Kiev 

Reservoir at the lower right of Figure 3-2. Only the upper end of the reservoir is visible in the 

image as an indistinctly outlined dark area merging with the meandering river beds. The cooling 

pond of the nuclear power station appears as the large, black, footprint-like shape outlined in white 

along the Pripyat River. The north/south extent of the Landsat image of Figure 3-2 corresponds 

approximately to the north/south extent of the dashed circle on the map in Figure 3-1. This 30 km 

radius circle is the danger zone evacuated in the aftermath of the reactor explos'e.' A 38.4 km 

square subset of Figure 3-2 is selected for the ground cover analysis in Sections 3.2 through 3.5 

below. 

3.1 REGIONAL DESCRIPTION. 

In the following regional description, we use the term Former Soviet Union (FSU) to denote 

the old Union of the Soviet Socialist Republics (USSR) but continue to use references to the names 

of the Soviet Socialist Republics (SSRs). 

3.1.1 Geography. 

The Chernobyl Nuclear Power Plant (NPP) and much of the area within 30 km around it are 

located in the Poles'ye region of the Ukrainian and Belorussian (Byelorussian, White Russian) 

11

RechittaDobrush 

Kilinkovichi 

GOMEL' OBLAST' 

Kopachia e ybc 

% %CCeembyi 

.qi~ esicaI~~ 

Orodyanka. 

Chernbyl nclear ccient 

Poiesitoy %12

Figure 3-2. Landsat image of the confluence of the Dnieper and Pripyat Rivers at the upper end of 

the Kiev Reservoir. The sharply outlined black footprint shape is the reactor station cooling pon4 

along the Pripyat River. [Thematic Mapper false color presentation (RG,B) = (7,4,1), 6 June 

1985, geocoded, 1 degree longitude by 0.5 degree latitude.] 

13

Soviet Socialist Republics. The power plant itself and the cities of Pripyat and Chernobyl are in 

the Ukrainian SSR. It lies at about 51°12'N longitude 30 0 8E latitude. 

The Poles'ye (Polesye, Poles'e, Poles'ya, Polesie, Polessie) lies in the Pripyat (Pripyat', 

`qip'at, Pripiat, Pripet) River basin and part of the Dnepr (Dnieper) River basin and includes the 

area called the Pripyat marshes or bogs. It is a vast lowland on the Russian platform, extending 

south to the Volyno-Podolsk Plateau (Keller 1927; Berg 1950; Fridland 1976; Lysenko and 

Golovina 1982). The relief from the center to the edges of the basin is only 55-100 m (Berg 1950; 

USSR 1987). Drainage is very poor and the ground-water table is usually high (Keller 1927; Berg 

1950; Fridland 1976). The banks of the streams in the Poles'ye are very low (Keller 1927). In the 

spring and after heavy summer rains, streams overflow into areas between neighboring streams 

and water from one stream passes into another. The Poles'ye is often divided into three sections: 

the western, the central (or right bank--on the right bank of the Dnepr River), and the eastern (or 

left bank) (Golovina, et al., 1980; Lysenko and Golovina 1982). In the Ukraine, the central 

section is divided into the Kiev Poles'ye and the Zhitomir Poles'ye. The Kiev Poles'ye lies on the 

middle Dnepr slope, in the interfluve of the Pripyat and Teterev Rivers, extending east to the Dnepr 

River (Golovina, et al., 1980; Lysenko and Golovina 1982), and includes the Chernobyl NPP and 

the immediately area 

3.1.2 Soils. 

The soils in the Poles'ye formed in a humid climate on platform plains from a blanket of 

unconsolidated Quaternary fluvioglacial sand and loamy sand over and underlying a glacial 

moraine (Fridland 1976; Lysenko and Golovina 1982). The moraine itself is mostly loamy sand. 

Clay-loam lake deposits are rare and loesses even rarer (Lysenko and Golovina 1982). The parent 

materials are very boldery and gravely; fine particles are usually washed away (Golovina, et al., 

1980). Glacial waters were active for along time during the formation of the parent material 

(Lysenko and Golovina 1982). The soils generally lack carbonates in the parent materials. The 

sandy soils of the Poles'ye are particularly vulnerable to erosion if plowed (Symons 1972). Some 

areas possess aeolian relief, often in the form of parabolic, west-facing dunes (Berg 1950; Fridland 

1976). This significantly affects soil cover composition. 

Most of the Poles'ye soils are poor in humus (USSR 1986). Soils in the western and central 

Poles'ye often have 1-1.5% humus and can have less than 1% (Krupskiy, et al., 1970). The soils 

generally have a low pH (Golovina, et al., 1980) and are low in available nutrients, including 

boron, zinc, phosphorus, potassium, and nitrate (Golovina, et al., 1980; Oleynik 1981; Lynsenko 

and Golovina 1982). There are large quantities of weakly podzolic sandy soils. Podzols develop 

under coniferous forests (Sukachev 1928; Berg 1950). The distribution of podzols is patchy in 

14

the Poles'ye (Fridland 1976). Medium-podzolic sod loamy sandy soils have formed on moraine 

outcrops in the Kiev Poles'ye (Lysenko and Golovina 1982). 

Much of the area is boggy. Bog soils receive excessive moisture for the greater part of the 

year, are sometimes covered with shallow water, and have poor drainage (Berg 1950). Much of 

the Poles'ye is low lying, with poor drainage, ideal for bogs (Keller 1927; Berg 1950). There are 

also sod-podzolic ("half-bog" or "meadow") soils (Berg 1950). Soils on flood plains may be bog, 

meadow, or podzols. Some parts of the Poles'ye contain "islands" of soils foreign to it (Berg 

1950). One such island occurs on the Ovruch ridge (about 90 km west of Chernobyl). This ridge 

is 320 m above sea level and 60 m above the surrounding lowland. 

Based on the soils map in the Ukrainian SSR Atlas (Anonymous 1962), the site of the 

Chernobyl nuclear power plant appears to be at or near the boundary of: 

Type 1: 

Type 27: 

Soddy-slightly podzolic sands and sandy-clayey soils. 

Sods and meadowlands, gleys, sandy loams, and loams. 

It is likely that the coniferous forests near the site are on type I soils. 

3.1.3 Climate. 

The climate of the Poles'ye is influenced by maritime, continental, and local factors. Moisture 

from the Baltic Sea and the "great valley" region of Poland influences the Poles'ye (Borisov 1965; 

Szafer 1966). There is unrestricted passage of marine winds, unhindered by major land relief, s- 

there is considerable marine influence on the climate. 

A small local maximum of relative humidity, caused by the intensified evaporation of water of 

standing water, is noticed over the marshes of the Poles'ye(Borisov 1965). Bogs influence the 

local microclimate (Szafer 1966). These areas have high humidity, low temperature minima, 

evening and morning mists Close to the soil surface, and frequent frosts. Forests also influence the 

microclimate (Szafer 1966). The mean air temperatures are lower than in the open, the daily 

temperature ranges are smaller, the snow cover is less, and winds are stilled. 

In the Poles'ye, the minimum relative humidity at 1 pm in May is 50-55%. The mean humidity 

for June, July, and August is 56-60% (Borisov 1965). Mean temperatures in the region are -6 to 

-7*C in January, 6 to 70C in April, 190 in July, and 7°C in October (Anonymous 1962). Absolute 

minimum temperatures are -30 to -35'C in January, while summer maximums range up to 40'C 

(Anonymous 1962). The growing season (days with mean temperatures above 5 C) is about 160- 

190 days (Anonymous 1962; Szafer 1966). It begins about April 11 and runs to about October 25 

(Anonymous 1962). Summer precipitation exceeds that in winter by a factor of two (Borisov 

1965). In any season, there is precipitation every 2-3 days. In the warm season (April- 

September), the amount of precipitation in the interior of the European FSU is much greater (350- 

500 mm) than on the coasts (200-300 mm) (Borisov 1966). During the cold half of the year 

15

(November to March) in the central belt of European FSU, the precipitation amounts are 100-300 

mm (Borisov 1965). A precipitation maximum is situated on the Pripyat and the upper reaches of 

the Dnepr and Western Dvina Rivers. The maximum precipitation in the Pripyat basin is 680-695 

mm/year (Berg 1950). Precipitation diminishes eastward from the Pripyat (Borisov 1965). In the 

specific region of interest, mean precipitation is 500-600 mm/year, with around 180 days/year 

recording precipitation (Anonymous 1962). The mean intensity of precipitation amounts to 8- 10 

mm/hr in the central belt o the country (51-59°N) (Borisov 1965). 

At 500N, by the second 10-day period in November snow cover is normally continuous 

(Borisov 1965). In western European FSU, snow cover is usually continuous by the last 10 days 

in October or the first 10 days in November. According to Anonymous (1962), sniow cover in the 

Chernobyl area usually lasts from mid-December to Mod-March. At 55'N 30'E, there are an 

average of 2.5 temporary snow covers before the onset of winter (Borisov 1965). In central 

European FSU, winter temperatures and precipitation are highly variable (Borisov 1965). Mean 

maximum snow depth is 10-30 cm. Continuous snow cover generally lasts about 80 days. Rivers 

and streams are generally frozen for about 100 days/year (Kendrew 1942). 

3.1.4 Native Vegetation. 

Most of Belorussia and Ukraine, including the Poles'ye, is in the Eastern European Vegetation 

Province (Takhtajan 1986). The northern, eastern and southeastern boundaries correspond to the 

distributions of Quercus roburl (English oak), Acer platanoides (Norway maple), and Corylus 

avellana (European filbert). The basic plant community types in the region are forests, woodlands 

or carrs, meadows, fens, and bogs. 2 . Many of these communities can grade into one another. 

Bog, meadow, and forest vegetation may occur on flood plains. Meadows may be transitional or 

ecotonal between bogs and forests. In poor sandy soils, peat bogs occupy the low areas, forests 

the more upland areas. Succession can progress from forest to bog or from bog to forest. Most 

bogs in central Europe are being drained (Walter 1978). The replacement vegetation is usually 

meadow grasses, birch, pine, or spruce. While the vegetation types to be discussed are specific to 

the Poles'ye, the species lists given are those known to occur in that type of vegetation in northern 

Ukraine, southern Belorussia, and/or adjacent areas in Poland. Information on both the vegetation 

IBoth Common and Latin names come from a number of literature sources. In order to eliminate synonyms and 

assure current nomenclature, the names in the Flora of the USSR and the Flora Europaea will be used for 

the final report. 

16

of the Poles'ye and the species lists was synthesized from Keller 1927; Sukachev 1928; Wulff 

1943; Berg 1950; Szafer 1966; Walter 1978; Oleynik 1981; and Takhtajan 1986. 

The sandy riverine sections along river channels receive an annual sand deposit Berg (1950). 

The community lying closest to the stream is often a sedge-horsetail fen. The plat", found in this 

type of habitat include Equisetum arvense (field horsetail), E. variegatum, E. palustre (horsetail or 

scouring rush), Carexfusca, C. canescens, C. stellulata, C. dioica, C. flava (sedges), Eriophorum 

latifolium, and E. augustifolium (cottongrass or cottonsedge). 

Immense tracts of the Poles'ye are occupied by bogs. There are several types, but most are 

Sphagnum or peat bogs. These are dominated by sphagnum mosses, including Sphagnum 

recurrum, S. fuscum, S. cuspitatum, S. medium, S. revellum, S. acutifolium, Hypnum schreberi, 

and H. crista-castrensis. Such bogs also have Eriophorum vaginatum (sheathed cottongrass), 

Carex pauciflora, C. limosa, and C. vaginata, and the carnivorous plant Drosera intermedia 

(roundleaf sundew). The vegetation of reed-bulrush bogs include Phragmites communis (common 

reed), S. lacuster (bulrush), Phalaris arundinacea (canarygrass), Calamagrostis neglecta 

(reedgrass), Typha angustifolia, T. latifolia (cattails), Carex vesicaria, C. gracilis, C. 

pseudocyperus, C. rostrata. Sedge bogs often include Carexfiliformis (large sedge), C. vesicaria, 

C. pseudocyperus, C.rostrata. C. elata, C. gracilis, C. acutiformis, C. riparia, C. lasiocarpa, C. 

diandra, C. aespitosa, and C. omaskiana (omskiana sedge) with Deschampsia caespitosa (tufted 

hairgrass) on hummocks. Sedge bogs are often found where Alnus glutinosa (European alder) has 

been cut. Transitional areas begin with either Sphagnum or sedge bogs and these can grade to 

alder, birch, birch-spruce, pine-birch, pine, and birch-aspen-conifer bogs. 

There are two basic types of meadows, flood-plain (or wet) meadows and upland (or fresh) 

meadows. Flood-plain meadows may be inundated for some time each year, especially in late 

winter and early spring. These are found in river valleys and at the peripheries of shallow lakes. 

They often consist of secondary vegetation resulting from drainage of fens or bogs. The 

vegetation includes Festuca ovina (sheep fescue), F. rubra (red fescue), F. pratensis (fescue), 

Agropyron repens (quack- or conchgrass), Poa pratensis (Kentucky bluegrass), P. annua (annual 

bluegrass), Phleum pratense (timothy), Agrostis stolonifera (redtop), Calamagrostis neglecta, 

Deschampsia caespitosa, Molinia coerulea, Carex panicea, C. caespitosa, Eriophorum neglecta, 

and Trifolium pratense (red clover). 

Fresh meadows generally are found in interstream areas, developing as secondary vegetation 

on cut or burned forest sites. Mowing is often used to prevent forest encroachment. They have a 

2 A fen is a constantly or frequently flooded area through which water flows. For simplification, the term "bog" 

includes marsh and swamp and is used to mean flooded areas having zero or very low rates of water flow. 

17

moderate moisture supply which fluctuates widely but generally does not surface. These meadows 

are usually used as pastures or hay meadows and some of the plants, especially the legumes, may 

have been sown. The vegetation includes Arrhenatherum elatius (tall or false oatgrass), Bromus 

mollis (smooth brnme), Agrostis stolonifera, , P. trivialis (bluegrass), P. annua, Phleumn pratense, 

Lolium perenne (perennial rye grass), Cynosurus cristatus (crested dogtail), Festuca pratensis, F. 

rubra, Agropyron repens, Phalaris arundinacea, Calamagrostis negecta, Alopecurus pratensis 

(meadow foxtail), Trifolium repens (white clover), T. pratense, T. hybridum (Alsike clover), T. 

incarnatum (crimson clover), and Medicago sativa (luceme or alfalfa). 

There are several types of carrs (or woodlands). The plants in willow carrs include Salix 

cinerea (gray willow), S. rosmarinifolia (rosemary willow), S. alba, S. fragilis, S. triandra, S. 

purpurea, S. rossica (willows), Alnus gutinosa, Fraxinus excelsior (ash), Betula laevis (European 

birch), Cornus alba (Siberian or white dogwood), Rosa (rose) sp., and Ribes nigrum (black 

current). The plants in elm carrs include Ulnus pedunculata (Russian elm), U. glabra (Scotch 

elm), Quercus robur, and Acer Platinoides. Alder-ash carrs usually include Alnus glutinosa, 

Fraxinus excelsior, Ulnus pedunculata, U. glabra, Acer platinoides, Carpinus betulus (European 

hornbeam), Festuca gigantea, and Agropyron Caninum (wheatgrass). 

The Kiev Poles'ye lies on the ecotone between the mixed coniferous-deciduous forest and the 

forest steppe. There are two basic mixed coniferous-deciduous forest types: Pine-oak forest and 

spruce-oak forest. The common steppe-forest of the area is oak-hornbeam. There are also some 

pine (bor) woodlands in the Poles'ye. It is unlikely that any virgin forests remain in eastern 

Europe (Walter 1978). 

Quercus robur is the dominant in forest-steppe in the European USSR. It grows best in the 

southwestern part of the European FSU and in the Poles'ye. Quercus robur will not grow on 

strongly podzolic soils. It is commonly found on floodplains and often grows in mixed stands 

with Pinus sylvestris (Scotch, Scots, or common pine), Picea abies (European spruce), or 

Carpinus betulus. Fagus sylvatica (European beech) is sometimes a component in oak-hornbeam 

forests on richer soils. 

Pinus sylvestris is a light-loving species, not tolerant of shade. It is not very exacting in soil 

and moisture requirements. It tolerates relatively poor soils and is often associated with bogs and 

"bor" soils. "Bor" forest is a type of sparse pine forest growing on sandy soil, dunes, etc. 

(Fridland 1976). The sandy forested areas, such as the parabolic, west-facing dunes, of the 

Poles'ye are usually covered with Pinus sylvestris, giving the region a "northern" appearance. The 

understory in pine (bor) woodland is often dune vegetation, including Festuca ovina, F. rubra, Poa 

pratensis, Deschampsia caesrvitosa, Carex arenaria (sand sedge), and Trifolium arvense (clover). 

When these forests are cut, the dune vegetation replaces them until there is a secondary pine 

growth. The understory in pine-oak forests often include Betula laevis, Tilia cordata (linden, 

18

lime). Populus tremula (aspen), and Corylus avellana. Pinus trees in bogs are frequently stunted. 

Based on the maps in Anonymous (1962), common pine (Pinus sylvestris) is the most likely forest 

type in the immediate vicinity of the Chernobyl nuclear power plant. 

Picea abies endures shade well but requires humid, relatively rich soils. Its southern limit runs 

through Kiev Poles'ye. Spruce-oak forests are probably occur mainly on "islands" of richer 

podzolic soils scattered in the Poles'ye. Near the southern limit of Picea's range, the most frequent 

understory species in spruce-oak forest are Tilia cordata, Acer platanoides, Fraxinus excelsior, 

Ulnus pedunculata, Corylus avellana, Betula laevis, and Populus tremula. Isolated Pinus sylvestis 

trees are also occasionally encountered in the spruce-oak type. 

3.1.5 Crops. 

Only about half of the area around Chernobyl is suitable for agriculture and much of it is in hay 

meadows and grazing land (Trifolium spp. and grasses) and only fodder crops, including corn 

(Zea mays) and fodder beets LBeta Vulgaris) (Symons 1972; USSR 1987). Winter wheat 

(Triticum aestivum), winter rye (Secale cereale), millet (Panicum miliaceum), winter and spring 

barley (Hordeum vulgare), oats (Avena sativa), flax (Linum usitatissimumn), hemp (Cannabis 

sativa), sugar beets (Beta vulgaris), buckwheat (Fagopyrum esculentum), and potatoes (Solanum 

tuberosum) are all grown from human use in the southern part of the Byelorussian SSR and 

northern part of the Ukrainian SSR (Sanbur and Kovalenko 1969; Fullard 1972; Symons 1972; 

Dewdney 1982). However, these crops are not grown in the poorly drained, boggy parts of the 

Poles'ye (Dewdney 1982). At any one time, 12-17% of the land is being fallowed (Symons 

1972). 

3.2 TASSELED CAP TRANSFORMATION OF LANDSAT IMAGERY. 

As a first step in the analysis of vegetation in Landsat imagery, it is useful to transform the 

spectral bands of the Thematic Mapper (TM) multispectral imaging sensor to a new spectral 

coordinate system that compresses the useful information into fewer bands and provides a physical 

interpretation of the transformed bands. We use the so-called Tasseled Cap transformation. 

3.2.1 Description of the Tasseled Cap Transformation. 

The Tasseled Cap (TC) transformation for the six reflective spectral bands of the Landsat 

Thematic Mapper imaging multispectral sensor was first described by Crist and Cicone (1984). It 

is a descendant of the Kauth-Thomas transformation developed for agricultural and vegetation 

scenes for the Multispectral Scanner (MSS) imaging device on earlier Landsat missions. The TC 

transformation is linear and preserves the Euclidean relationship of the pixel data contained in the 

six reflective TM bands. It is chosen so that most of the spectral-temporal variance in agricultural 

19

and vegetation scenes is included in the first three TC bands. Furthermore, the linear combinations 

of the original TM band intensities that provide the first three TC band intensities are chosen so that 

the magnitude of each of the three has a physical internretation with respect to vegetation and soils. 

Later Crist and coworkers (Crist et al., 1986), adjusted the fourth, fifth and sixth components of 

the tasseled cap transformation to maximize the correlation of the fourth component with 

atmospheric haze. Table 3-1 lists the resulting transformation coefficients. The coefficients are 

used according to the following example: 

TC1 = 0.2909*TM1 + 0.2493*TM2 + ... + 10.3695, 

where TCi represents the ith Tasseled Cap band intensity and TMi represents the ith Thematic 

Mapper band intensity. 

Table 3-1. Original Landsat-5 TM Tasseled Cap coefficients (Christ et al., 1986). 

Tasseled cap Coefficients Additive 

Band Feature TM1 TM2 TM3 TM4 TM5 TM7 Term 

TCI Brightness .2909 .2493 .4806 .5568 .4438 .1706 10.3695 

TC2 Greenness -.2729 -.2174 -.5508 .7221 .0733 -. 1648 -0.7310 

TC3 Wetness .1446 .1761 .3322 .3396 -.6210 -.4186 -3.3828 

TC4 Haze .8461 -.0731 -.4640 -.0032 -.0492 .0119 0.7879 

TC5 Fifth .0549 -.0232 .0339 -. 1937 .4162 -.7823 -2.4750 

TC6 Sixth .1186 -.8069 .4096 .0571 -.0228 .0220 -0.0336 

Figure 3-3 illustrates the transformation from the Landsat TM images presented in Volume 2 to 

the tasseled cap images analyzed in this volume. The first four TC features have interpretations in 

terms of the physical features brightness, greenness, wetness and haze: 

Brightness (Br). 

Brightness is the overall reflectivity with band-by-band weighting; red (TM3) and near infrared 

(TM4) bands are emphasized; soils are generally brighter than vegetation; pines forest has a 

relatively low brightness compared to other vegetation. 

Greenness (Gr). 

Greenness is a basic vegetation index, dominated by the difference between the near infrared 

(TM4) and red (TM3) band intensities; chlorophyll and other plant pigments absorb red light but 

strongly reflect near infrared light giving foliage a high greenness value. In principle, the difference 

20

Landsat Thematic Mapper Image 

- Geocoded and resampled to 25 m square pixels 

- Pixel vector = Intensity in seven (7) spectral bands 

" Six (6) reflective bands 

Band Intensity Spectral Region Wavelength (microns) 

TM1 Blue 0.45 - 0.52 

TM2 Green 0.52 - 0.60 

TM3 Red 0.63 - 0.69 

TM4 Near Infrared 0.76 - 0.90 

TM5 Short Wave Infrared 1.55 - 1.75 

TM7 Short Wave Infrared 2.08 - 2.35 

"• One (1) emissive band 

TM6 Thermal 10.4- 12.5 

Tasseled Cap 

spectral transformation 

I (Linear transformation 

on the six reflective bands) 

Tasseled Cap Image 

- Tailored for agricultural and vegetation scenes, same pixel size 

- Pixel vector = Intensity in seven (7) features 

" Six (6) reflective features 

Band Intensity Feature Abbreviation 

TC 1 Brightness Br 

TC2 Greenness Gr 

TC3 Wetness Wt 

TC4 Haze Hz 

TC5 (minimal information) - 

TC6 (minimal information) - 

"* One (1) emissive band 

TC7 Thermal (same as TM6) 

Figure 3-3. Illustration of the Tasseled Cap spectral transformation for Landsat Thematic 

Mapper images; transformation coeffcients are listed in Table 3-2. 

21

etween green light intensity and either blue or red provides a similar index. However, the 

contrast is higher between near infrared and red with the added benefit that atmospheric scattering 

introduces less background noise in the infrared than at the shorter visible wavelengths. 

Wetness (Wt). 

Wetness increases with soil and vegetation moisture content; it is predominately determined by 

the difference between the combined red/near infrared reflectivity (TM3 + TM4) and the combined 

short wave infrared reflectivity (TM5 + TM7); it is affected by absorption bands of water and 

structural effects related to foliage hydration; there is some contribution from shadowing which 

relates to vegetation height and stand density. 

Haze (Hz). 

Haze is aerosol scattering index that emphasizes the difference between blue light reflectance 

(TM 1) and red light reflectance (TM3); Rayleigh scattering by small aerosols strongly scatters light 

at shorter (blue) wavelengths providing a large haze signal from clouds, mist, smoke and aircraft 

contrails; on clear days the largest "haze" signals come from urban/industrial surfaces, snow and 

ice. 

Our image processing software (see Volume 2) requires band intensities in the range 0 to 255. 

We modified the tasseled cap transformation slightly to insure that all TC band intensities for the 

Landsat Chernobyl images fall in this range. Table 3-2 shows the resulting transformation 

coefficients. Band 1 requires a multiplicative scale factor of 1/2 applied to the corresponding 

coefficients in Table 3-1 to reduce its dynamic range below 2"5. Tasseled cap bands 2 through 6 

require an additive term to eliminate negative values. These changes do not effect the linearity or 

physical interpretation of the TC transformation. 

Table 3-2. Modified Tasseled Cap coefficients used in the present work to place all feature 

intensities in the range 0 to 255. 

Tasseled cap Coefficients Additive 

Band Feature TMJ TM2 TM3 TM4 TM5 TM7 Term 

TCI Brightness .1455 .1247 .2403 .2784 .2219 .0853 5.2 

TC2 Greenness -.2729 -.2174 -.5508 .7221 .0733 -.1648 160. 

TC3 Wetness .1446 .1761 .3322 .3396 -.6210 -.4186 128. 

TC4 Haze .8461 -.0731 -.4640 -.0032 -.0492 .0119 64. 

TC5 Fifth .0549 -.0232 .0339 -. 1937 .4162 -.7823 160. 

TC6 Sixth .1186 -.8069 .4096 .0571 -.0228 .0220 160. 

22

Finally, we note that the Tasseled Cap transformation is named for the distinctive shape in the 

brightness-greenness plane of a scatter plot of crop and soil pixels from an agricultural scene. 

3.2.2 Tasseled Cap Images of the Analysis Area. 

The change detection analysis in Volume 2 covers a 12.8 km square area consisting of 512 x 

512 pixels centered on the main westward trace of initial local radioactive deposition caused in the 

first hours by the reactor explosion and fire. For the analysis of radiation doses presented in this 

volume, we have extended the dimensions of the analysis area threefold to a 38.4 km square 

consisting of 1536 x 1536 pixels centered at the same spot. Thus, the area analyzed in Volume 2 is 

the central 1/9th of the area analyzed here. 

Table 3-3 lists chronologically the 11 images analyzed and the time of each image relative to the 

day of the reactor explosion. References to date numbers or image numbers in this report are made 

according to this table. 

Table 3-3. Landsat scenes analyzed and time of scene relative to reactor explosion. 

Time relative to reactor 

Imageldate 

explosion, 

number Date Days 

1 6/06/85 -324 (-10.7 months) 

2 3/21/86 -36 (-5.1 weeks) 

Day of Accident 4/26/86 

3 4/29/86 3 

4 5/08/86 12 

5 5/24/86 28 (4.0 weeks) 

6 5/31/86 35 (5.0 weeks) 

7 10/15/86 172 (5.6 months) 

8 12/02/86 220 (7.2 months) 

9 5/11/87 380 (1.04 years) 

10 9/07/87 499 (1.4 years) 

11 5/28/88 763 (2.1 years) 

Figures 3-4 through 3-9 present the 1 TC-transformed images of the 38.4 km square area in 

chronological order. These false color images were produced by displaying the first three tasseled 

cap features, brightness, greenness and wetness in red, green and blue, respectively. For 

shorthand notation, we use (R,G,B) = (Br,GrWt). Table 3-4 gives the geographic coordinates of 

the area included in the images in the Universal Transverse Mercator (UTM) coordinate system. 

As usual, north is up and east is the right in these images. 

23

a) DATE 1: 6 JUNN 85 

b) DATE 2: 21 MAR 66 

Figure 3-4. Tasseled Cap false color images, (RG.B) =(BrGrWt) of 38.4 km square area 

around Chernobyl: a) early summer image one year preaccident and b) late winter image 5 weeks 

preaccident. 

24

a) DATE 3: 29 APR,86 

b) DATE 4: 13 MAY 86 

Figure 3-5. Tasseled Cap false color images, (RG,B) =(BrGr,Wt) of 38.4 )an square area 

around Chernobyl: a) 3 days postaccident and b) 12 days postaccident. 

25/26

a)DATE 2 4 MIAY 86 

b) DATE 6: 31 MAY 86 

Figure 3-6. Tasseled Cap false color images, (RG,B) =(BrGr,Wt) of 38.4 kmn square area 

around Chernobyl: a) 4 weeks postaccident and b) 5 weeks postaccident. 

27/28

A)I-\E~ 15) OCT 86 

b) DATE 8. DEC 86 

Figure 3-7. Tasseled Cap false color images, (R,G,B) (Br,Gr,Wt) of 38.4 km squam, area 

around Chernobyl: a) 5.6 months postaccident and b) 7.2 months postaccident. 

29/30

i). .XTj'E 9:1 \AY 6i7 

b) DATE 10. SEPT 87 

Figure 3-8, Tasseled Cap false color images, (R,G,B) =(Br,GrWt) of 38.4 km square area 

around Chernobyl: a) 1 year postaccident and b) 1.4 years postaccident. 

3 1/32

zi) ['ATE 11: 25 MAY 58 

b) DATE 11: 228 MAY 88 TC 4,5.6 

Figure 3-9. Tasseled Cap (TC) false color images of 38.4 kmn squae area around Chernobyl 2.1 

years postaccident: a) (R,G,B) = (Br,Gr,Wt) as in Figures 3-4 through 3-8 and b) (RG,B)= 

(Hz,TC5,TC6), the last 3 components of the TC transformation. 

33

Table 3-4. Coordinates (UTM, Zone 36) of the upper left comers of the comer pixels of the 

1536 x 1536 pixel area analyzed in this report. 

X 

Corner pixel (meters) (meters) 

Upper left 277,100. 5,714,900. 

Upper right 315,475. 5,714,900. 

Lower left 277,100. 5,676,525. 

Lower right 315,475. 5,676,525. 

Y 

Date 1, shown in Figure 3-4a, is a summer image one year before the nuclear accident. This 

image is used as the reference image for the change detection analysis presented in Volume 2. The 

red patches in Figure 3-4a are mostly soil in agricultural fields. The Chernobyl nuclear power 

station at the upper left of the cooling pond and the city of Pripyat also show as red. These areas 

have high brightness, low greenness, and low wetness in the TC spectral coordinate system. The 

darker, mostly bluish areas to the lower left (southwest) of the cooling pond are areas of 

predominately coniferous forest. Portions of the forest have been cleared for farmland. The 

lighter, more greenish areas sometimes grading into the coniferous forest are predominately 

deciduous vegetation. 

Date 2, shown in Figure 3-4b, is a late winter image taken five weeks before the accident. Ice 

and snow have high brightness, low greenness, and high wetness and appear magenta in the image 

since (R,G,B) = (Br,Gr,Wt). The Pripyat River is frozen over and residual snow cover is 

apparent bordering the areas of coniferous forest, which are now the strongest green areas of the 

image. Clearings in the coniferous forest are snow covered even though farmland is not. This 

observation suggests that the snowfall is not recent. It is likely that the floor of the forest is snow 

covered but the canopy is snowless. The snow cover will then show through from the satellite 

perspective wherever pine trees are thinner. This observation from the image implies that in areas 

of high TC greenness, TC brightness and wetness provide a measure of coniferous canopy 

closure. 

Date 3, shown in Figure 3-5a, is three days after the accident. There are more bare fields on 

this date than on Date 1 since it is earlier in the growing season. Likewise, deciduous areas have 

greened somewhat, but there is only moderate contrast between the evergreen and deciduous areas. 

Dates 4, 5, and 6, shown in Figures 3-5b, 3-6a, and 3-6b, respectively, are range from 12 days to 

5 weeks after the accident. By the end of May (Date 6), the contrast between evergreen and 

deciduous vegetation has returned to the higher level of Date I a year earlier. 

34

Date 7, shown in Figure 3-7a, is apparently after the autumn leaf senescence since the 

greenness of the deciduots vegetation has dropped dramatically. Date 8, shown in Figure 3-7b, is 

in December after the accident. It is similar to Date 2 except that the Pripyat River is not frozen and 

there is no snow cover. 

Dates 9 and 10 are shown in Figure 3-8. They are 1.0 and 1.4 years after the accident, 

respectively. Date 11, shown in Figure 3-9a, is 2.1 years postaccident. The cessation of 

agriculture in the analysis area after the accident is apparent from the absence of any fields with 

bare soil after Date 6 (31 May 1986). Such fields show up as bright red areas because of the high 

brightness of most bare soils. The red fields in the images before Date 6 are fading as cultivated 

fields are overtaken by indigenous vegetation. 

On the other hand, the increasing size of the red area around the reactor station on Dates 9, 10, 

and II shows land, including forests, being cleared as part of the radiation decontamination effort. 

On close examination, Date 7 shows earlier forest clearing along a road that crosses the trace of 

highest fallout deposition about 2 km west of Reactor Unit 4. The analysis of radiation response 

must account for these cleared areas and avoid mistaking cleared forest for radiation damaged trees. 

Figure 3-9b shows the other three bands of the Tasseled Cap image for Date 11 presented as 

(R,G,B) = (Hz,5,6). The lack of contrast in this image relative to the first three bands of the same 

image presented in Figure 3-9a illustrates how the Tasseled Cap transformation includes most of 

the pixel variance for vegetation scenes in the first three bands. The largest contrast in Figure 3-9b 

is between the urban/industrial area and the cooling pond, for which the Tasseled Cap 

transformation was not optimized. 

In summary, the Tasseled Cap images for Dates I through 11 as shown in Figures 3-4 through 

3-9 provide the data for all analysis in this volume. 

3.3 PRELIMINARY CLASSIFICATION OF EVERGREENS. 

This subsection describes the preliminary classification of vegetation in the 38.4 km square 

analysis area aimed at the identification of pixels in the image dominated by pine forest. We first 

define areas of evergreen vegetation and then use an unsupervised clustering algorithm (see, for 

example, Duda and Hart, 1973) in TC space to group the pixels into spectrally-related classes. 

3.3.1 Selection of Evergreen Pixels. 

In order to provide good discrimination between coniferous and deciduous forest, we construct 

a composite image from the preaccident images of 21 March 1986 (late winter) and 6 June 1985 

(early summer). Table 3-5 shows the band structure of the composite image. We refer to this 

winter/summer composite image as COMP21 since it is generated from Dates 2 and 1. 

35

Table 3-5. 

Band structure for the winter/summer composite image, COMP2 1, generated from 

Dates 2 and 1. 

Tasseled cap Tasseled cap 

COMP21 Band band(date number) feature name 

1 Br(2) Winter brightness 

2 Gr(2) Winter greenness 

3 Wt(2) Winter wetness 

4 Br(l) Summer brightness 

5 Gr(1) Summer greenness 

6 Wt(1) Summer wetness 

7 Th(1) Summer thermal 

An obvious difference between coniferous and deciduous forest is that coniferous forest has 

moderate greenness in both winter and summer while deciduous forest has high greenness in 

summer and low greenness in the winter. Figure 3-10 exploits this contrast by displaying winter 

greenness and summer greenness as the complementary colors green and magenta, respectively, in 

the same image. Evergreen vegetation, including coniferous forest, appears in varying shades of 

bright green. Deciduous vegetation appears magenta. 

Because the presentation colors in Figure 3-10 are complementary, pixels that have equal 

relative greenness in summer and winter appear as a shade of gray. Very dark shades of green are 

associated with the cooling pond, the power station, and some cultivated fields. Other cultivated 

fields appear nearly white. The Pripyat River and its meandering stream bed are nicely defined 

across the diagonal of the image. The river flows from upper left to lower right. The smaller Uzh 

River flows to the right across the lower portion of the image, dipping temporarily off the lower 

edge before emptying into the Pripyat River at the lower right corner of the image. Although it is 

not readily visible in Figure 3-10, the city of Chernobyl lies on the right bank of the Pripyat River 

along the north side of the Uzh River. 

The false color presentation of COMP21 in Figure 3-10, interpreted in the context of regional 

information presented in Section 3. 1, clearly indicates the presence of the large contiguous areas of 

pine forest in the vicinity of the Chernobyl nuclear power station, especially to the west in the 

direction of the main trace of early fallout deposition. To maximize the amount of data, we want to 

find all pixels (25 m by 25 m each) whose ground cover is dominated by pine trees. The method 

of classifying individual pixels based on their spectral signature is discussed in Volume 2. In 

short, we find a representative set of pixels, called a reference site, which defines the mean vector 

36

JJ 

Abb 

ji 

Figure 3-10. Evergreen vegetation appears bright green and deciduous and annual vegetation 

appears magenta in this false color presentation of COMP21 with (R,G,B) = (5,2,5). 

37

and the covariance matrix for the spectral signature of each vegetation class of interest. Maximum 

likelihood is then used to assign pixels to the various classes. 

Figure 3-11 outlines the procedure for defining classes of evergreen vegetation occurring in the 

38.4 km square analysis area. The procedure described in Figure 3-11 is a preliminary 

classification and need not account for all pixels in the image. In fact, to limit the complexity of the 

task, it is best to preselect pixels of the general type needed for the final analysis. Therefore, we 

limit our preliminary classification to pixels of evergreen vegetation. 

The selection of evergreen pixels is based on ranges of brightness, greenness, and wetness 

from the winter scene of Image 2. The ranges are listed in Figure 3-11. These ranges were chosen 

after examination of histograms of values of brightness, greenness and wetness for identifiable 

areas in Image 2. 

The ranges were adjtisted to cut out areas of water, ice, snow, soil, 

urban/industrial areas, and deciduous vegetation while keeping all pixels with a moderate to high 

value of wintertime greenness. 

As mentioned in Section 3.2.2, because of snow cover on the forest floor, TC brightness and 

wetness provide a measure of canopy closure in areas of coniferous forest. By inspection, we 

found that the wetness intensity value increased from about 146 to about 184 as the canopy closure 

varied from maximum (presumably near 100%) to zero. The wetness value for snow and ice was 

distributed mostly between 184 and 216, with a tail extending to 255. 

Figure 3-12b shows the set of pixels that pass the brightness, greenness, and wetness cuts 

defining evergreen pixels as shown in Figure 3-11. These evergreen pixels are marked in green 

over a gray-scale background image. This figure should be compared with Fig :e 3-10 to judge the 

effectiveness of the evergreen cuts. Notice that very few pixels in the agricultural fields or in the 

bottom land along the river passed the evergreen criteria. Most importantly, the obvious magenta 

areas in Figure 3-10 corresponding to deciduous or annual vegetation do not appear in the set of 

evergreen pixels. 

For the purpose of contrast, Figure 3-12a shows the same presentation as Figure 3-10 except 

that the roles of winter and summer greenness are reversed. The resulting false color image 

displays deciduous areas in bright green and evergreen areas in magenta. 

3.3.2 Unsupervised Clustering of Evergreen Pixels. 

As a first step toward identifying pine forest pixels in the analysis area, we use the first six 

bands of the winter/summer composite image COMP21 to divide the evergreen pixels into a 

manageable number of spectrally-related classes. Manual examination of the shape and structure of 

the evergreen pixel distribution in the six-dimensional spectral hyperspace is difficult, so we use an 

unsupervised clustering algorithm to divide the structure into ten neighborhoods or clusters. 

38

Transformed Images 

Preaccident Winter/ 

for All Dates ,_ Summer Composite 

(Tasseled Cap Bands) A Image - COMP2I 

(See Table 3-5) 

Hayo°n Evergreens 125 _< G(2") _< 255 

Image• 

No Haze 

145 5Wt(2) 5 180 

(See Figure 3-12b) 

Composite Image 

(COMP21) 

Classified According to 

Coniferigouf/Evergreen 

Ten Cluster Signatures 

Conpstentag 

Figure 3-11. 

Procedure for classifying evergreen vegetation. 

39

a)Deeidtiou-z, ill gr~een 

b) Evergreen areas 

Figure 3-12. a) Like Figure 3-10 except color reversal (RG,GB) =(2,5,2) displays deciduous or 

annual vegetation in bright green and evergreens in magenta, and b) areas passing the evergreen 

ciiteria of Figure 3- 11 shown in green. 

40

The avoidance of pixels affected by clouds and haze is discussed in Section 3.4 below. Since 

Image 1 and, hence, COMP21 contains significant clouds in the lower third of the image, we use 

the haze mask from Section 3.4 to avoid affected pixels in the subset used for unsupervised 

clustering. 

For unsupervised clustering, we use the ISODATA algorithm, which stands for "Iterative Self- 

Organizing Data Analysis Technique" (Tou and Gonzalez, 1974). The multipass algorithm is 

based only on the spectral signature of each pixel and not on that of its spatial neighbors in the 

image. The algorithm begins with a selected number (ten in this case) of class mean intensity 

vectors spaced evenly along a straight line through the region occupied by the evergreen pixels. On 

the first pass of the algorithm, all pixels are assigned to the class with the nearest mean. Each class 

mean is then moved to the mean value of the pixels assigned to it. On the next pass, class 

assignments of all pixels are reevaluated and adjusted based on distances to the new means. The 

process is repeated until subsequent iterations produce changes in pixel assignments less than a 

specified percentage. The method is reasonably nonparametric and is not biased toward any spatial 

location on the image. Results are also reasonably independent of the initial placement of means. 

The mean pixel vector and the covariance matrix of the resulting ten classes from the 

unsupervised clustering of evergreen pixels were used to classify all pixels of the COMP21 image. 

Examination of the spectral signatures of the ten classes and the spatial distribution of pixels 

assigned to each class resulted in the elimination of four of the ten as either only marginally 

cvergreen or lacking in contiguous areas of significant size. The spectral signatures of the 

remaining six classes, numbered 3 through 8 are displayed in Figure 3-13 along with the spectral 

signatures of the two forest classes used in the preliminary analysis of Volume 2. Detailed listings 

of the spectral signatures appear in Appendix D. 

3.3.3 Identification of Evergreen Classes. 

The spectral signatures of the six classes retained from the unsupervised clustering are best 

distinguished from one another in Figure 3-13 by the plot of Band 5 versus Band 4 (CHAN 5 

versus CHAN 4) of the COMP21 image. This is a plot of summer greenness versus summer 

brightness. The spectral signatures organize into two branches emanating from Class 5. The 4/3 

branch extends to higher brightness and lower greenness and the 6/8 branch extends to both higher 

brightness and greenness. 

Similarity to the Forest 1 and Forest2 signatures from the analysis of Volume 2 shows that the 

4/3 branch corresponds to the large areas of coniferous forest evident in Figure 3-10. Class 5 is 

pine forest with the highest canopy closure and the least component of deciduous trees since it has 

the lowest brightness of all the classes in both summer and winter and the highest greenness in the 

winter. 

41

U) 

LO 

Class 3 has lower greenness and higher brightness than Class 5 in both summer and winter; it 

apparently consists of similar pine trees with lower canopy closure (less dense foliage) but no 

significant component of deciduous trees which would increase the greenness feature in summer. 

The signature of Class 4 lies between Classes 5 and 3 in both summer and winter, indicating that it 

consists of the same pine trees with intermediate canopy closure. Thus, the 5/4/3 class sequence is 

apparently due to varying canopy closure of similar trees. 

Class 6 is adjacent to Class 5, the high density pine forest. In the winter it has essentially the 

same spectral signature as Class 4 and so may be inferred to contain about the same density of pine 

trees as Class 4. On the other hand, Class 6 has a higher greenness in summer than Class 6, 

indicating that the pine trees in Class 6 are interspersed with deciduous trees. 

Class 8 is most removed from Class 5 along the 6/8 branch in the summer 

brightness/greenness plane. 

The very high summer greenness value indicates a dominant 

component of deciduous or annual vegetation. On the other hand, the winter greenness value still 

indicates a significant evergreen component. Three possibilities are sparse pine trees in deciduous 

forest, evergreen shrubs or vines as understory in deciduous forest, and evergreen shrubs with 

other deciduous or annual vegetation. In exploratory calculations, we found clear examples of 

interspersed pixels of Class 6 and Class 8 where Class 6 showed definite radiation response and 

Class 8 did not. Also, by association with water patterns in the image, Class 8 apparently tends 

toward lower lying areas. We assume for further analysis that Class 8 contains few or no pine 

trees. A reference site for Class 8 is defined and Class 8 is retained in the pixel classification 

procedure, however, to minimize the number of Class 8 pixels that might otherwise be mistaken 

for Class 6. 

Class 7 lies near the 6/8 branch of the summer spectral signatures between Class 6 and Class 8. 

With respect to of its spatial distribution, Class 7 tends to occur at the edges of Class 6 and to have 

no significantly sized areas of its own. We assume that it consists of mixed pixels rather than a 

primary forest type. We do not include it in further analyses because of the lack of a spatially 

homogeneous reference site. 

3.4 REFERENCE SITES FOR PINE FOREST CLASSES. 

Figure 3-14 outlines the procedure for selecting a single reference site for each of the retained 

classes of evergreen vegetation. The starting point is the winter/summer composite image with 

pixels classified according to the signatures of the classes from the unsupervised clustering 

described in Section 3.3.2. Using an interactive mode of the image processing software, polygons 

are drawn around several representative areas of contiguous pixels for each of the retained 

evergreen classes (3,4,5,6, and 8). 

reference sites. 

These homogenous polygons are candidates for class 

43


Cloud/Haze 

(COMP21) 

Enhancements 

Classified According to (Figures 3-16 

Unsupervised Clusters to 3-21) 

Forslcatig oer g 

HomogneousHaze 

Composite 

Mask 

6, 56and 8 

Si4for Classes 43, 

Rersnaivue 3-4 rcdRefrslcigoeodreference Raiatioe onor 

Spectral Signtur ~o fo ac Soiteros/eren 

oiferos/everrenc foclass. class. oi az/lod az/Cous 

Cls4nal4ae

The reference sites will be used to define the normal spectral signature of each pine forest class on 

each image date analyzed. As such, the reference sites must not be obscured by clouds or haze on 

any date and must not be significantly affected by radiation exposure on any date. Guidance from 

the forest response and radiation contours reported in Volume 2 is used to avoid areas of forest 

shcwing radiation damage. Images enhanced to show clouds and haze are used to insure a clear 

view of each reference site on each image. 

3.4.1 Avoidance of Clouds and Haze. 

Figure 3-15 shows a cloud and haze enhanced image for Date 1. The presentation shows 

clouds and hazy areas in blue and all other areas in shades of yellow and gray. The use of 

complementary colors (here, blue and yellow) follows the same principle used to contrast summer 

and winter greenness in Figure 3.10. The cloud and haze enhancement contra.sts the Tasseled Cap 

haze band with the thermal band from the same date. The TC haze band, as discussed in Section 

3.2, provides a high signal from the aerosols in clouds and haze and a generally lower signal from 

land and water surfaces. On the other hand, the thermal band provides a high signal from land 

surfaces warmed by the sun and a lower signal from clouds and haze, which are at the usually 

cooler temperature of the atmosphere. Display of the haze signal in blue and the thermal signal in 

yellow results in good definition of clouds and any nonuniform haze as shown in Figure 3-15. 

Figures 3-16 through 3-21 show the cloud and haze enhanced images for each date. Only 

Dates 1, 3, 5, 6, 9, and 10 have significant clouds or nonuniform haze. As indicated in the 

procedural diagram of Figure 3-14, the cloud/haze enhancements are used in two ways. First, we 

construct a composite haze mask for the full set of images. A threshold value for the TC haze 

feature is determined for each of the six images with clouds or haze such that pixels in obvious 

areas of clouds and haze always exceed the threshold. With these thresholds, we create a single 

numerical mask that marks each pixel of the analysis area that is affected by clouds or haze on one 

or more dates. This haze mask is used during the selection of candidate reference sites to limit 

consideration to areas that are clear of clouds or haze on all dates. 

Second, the individual cloud/haze enhancements are used in the final choice of a single 

reference site for each class to double check that the chosen site is always free of clouds and haze. 

This step is accomplished by plotting the polygons of the candidate reference sites directly on the 

cloud/haze-enhanced images. 

45

Figure 3-15. Cloud/haze enhancement for Date 1 for the 38.4 km square analysis area, 

(R,G,B) = (Th,Th,Hz). Clouds appear blue, warm areas yellow. Thermal gradient appearing as 

variation in yellow shade of the cooling pond shows counterclockwise flow of water around 

central barrier. 

46

a) Trair•ing •ite location• 

--- u 

•!1 •]"• 

•' 

Su[] 

- 

• .I] 

SrJm 

• - 

rJ 

b) Date 1: 6 JUN 85 

Figure 3-16. a) Reference (control) site locations for Classes 3, 4, 5, 6, and 8 in the 38.4 km 

analysis area and b) cloud/haze enhancement for Dam 1 showing three polygons used for 

classification merger described in Section 3.5. Red dots are reference site locations. 

47148 

g J • II III

t 

a1 Date 2.: 211 MAR 86 

b) Date 3: 29 APR 86 

Figure 3-17. Cloud/haze enhancements for a) Date 2 and b) Date 3. Only Date 3 has clouds. Red 

dots are reference site locations. 

49/50

it) Date 4: 

8 MAY 8U 

b) Date 5: 24 MAY 86 



5 1/52

a) Datet 6- .31 MAY 86 

41 

INVb) 

Date 7: 15 OfT 86 



53154

aD Date 8: 2 DE" 86 

b) Date 9: 11 MAY 87 

Figure 3-20. Cloud/haze enhancements for a) Date 8 and b) Date 9. Only Date 9 has clouds. 

Reference Sites 5, 6 and 8 have been moved on Date 9 to avoid a faint jet contrail. 

55/56

a) Date 10: 7 SEP 87 

b) Date 11: 28 MAY 88 

Figure 3-21. Cloud/haze enhancements for a) Date 10 and b) Date 11. Neither date has clouds, 

but Date 10 has a jet contrail. Red dots are reference site locations. 

57

3.4.2 Representative Spectral Signatures. 

During the final selection of reference sites, the spectral signature of each site is examined to 

insure that it is truly representative of its class. This step is accomplished by plotting the spectral 

signatures of each of the candidate sites with the signature of the overall class aý in Figure 3-13. 

3.4.3 Sites Chosen. 

Figure 3-16a shows the final choice of primary reference sites for Classes 3, 4, 5, 6, and 8 

overlaying a gray-scale image of the analysis area. On this image, the actual size and shape of each 

reference site is displayed as a patch inside a circle labeled with the class number of the reference 

site. Included on the cloud/haze enhancements in Figures 3-16 through 3-21 for each of the eleven 

image dates are five red dots marking the location of the reference sites. For better visibility, the 

red dots on the enhanced images are larger than the actual reference sites. 

The primary reference sites shown in Figure 3-16a are used on all images except Date 9. On 

this date, the primary Reference Sites 5 and 8 are slightly obscured by the remnants of a jet 

contrail. Site 5 is moved to a nearby patch in the same forested area and Site 8 is moved across the 

river to a similar site. Finally, primary Reference Site 6 is off the edge of the available image on 

Date 9. It is also moved across the river to a similar site on this date. Reference Sites 3 and 4 are 

unchanged. Site 3 is near, but safely outside, the contrail. Note that a very well defined, 

presumably newer, jet contrail appears in the image on Date 10. Date 5 shows hints of a contrail 

along the same flight path. Date 5 also has a broader diffuse band passing north/south over the 

reactor cooling pond. It may be an old contrail or a thin cloud. 

3.4.4 Other Impacts of Clouds and Haze. 

The enhanced images show that Dates 2, 4, 7, 8, and 11 are entirely free of clouds over the 

analysis area. These images presumably have varying degrees of atmospheric haze from one date 

to another, but the haze is uniform over each image and does not impact the radiation response 

analysis. 

The occurrence of occasional scattered clouds on the other image dates is tolerable in our 

analysis even though the presence of a cloud over a forested area in an image causes an apparent 

deviation from normal of the obscured pixels. We avoid difficulties with the apparent deviation 

due to clouds by requiring persistence from date to date in the detection of radiation-induced 

response. Because the percentage of cloud cover is low and because we have several cloud-free 

images, there is no case where coincidence of cloud cover in successive images mimics damaged 

forest. 

It is interesting to note in passing that the thermal signals from the reactor cooling pond 

represented by the shades of yellow in Figures 3-16 to 3-21 show that after the explosion of Unit 

58

4, the other three reactors were turned off on Dates 3 through 6. At least one reactor was operating 

on Date 7 (October 1986) and thereafter through Date 11. 

3.5 PREACCIDENT PINE FOREST CLASSIFICATION MAP. 

The classification of evergreen pixels in the winter/summer composite image COMP21 requires 

extracting the spectral signature of the reference sites for Classes 3, 4, 5, 6, and 8 from COMP2 1. 

These spectral signatures consist of the mean pixel intensity vector and its covariance matrix for 

each reference site. With this information, the likelihood that a pixel belongs to each class is 

calculated. The pixel is assigned to the class with the highest likelihood as long as a threshold 

value is exceeded. Pixels below the threshold value are not assigned to any class. We used the 

first six bands of COMP21 for the signatures and the maximum likelihood assignment to classes. 

These six bands are the brightness, greenness, and wetness for winter and summer as listed in 

Table 3-5. The maximum likelihood method is discussed further in Section 5.1 of Volume 2. 

Because Image I is missing one corner and is also affected by clouds as illustrated in Figure 3- 

15, the pixel classification from COMP21 is not satisfactory for the whole analysis area. The areas 

of unacceptability are outlined in white by three polygons in Figure 3-16b. Fortunately, Image 4 

from 12 days after the accident is cloud free and can provide a substitute summer component for a 

composite winter/summer image. We have constructed such a composite called COMP24 

according to Table 3-5 by substituting Date 4 for Date 1. Although technically not a preaccident 

composite, COMP24 serves as a reasonable substitute since the areas of unacceptability in 

COMP21 are far enough from the reactor station to show no radiation effects by Date 4. The final 

classification uses COMP21 where acceptable and COMP24 elsewhere. 

The procedure for generating the final pixel preaccident classification map is illustrated in 

Figure 3-22. COMP24 is classified using spectral signatures from its own reference sites in the 

same way as COMP21. Classifications from COMP21 and COMP24 are merged by using the 

COMP21 pixel assignments everywhere except within the three polygons displayed in Figure 3- 

16b where assignments from COMP24 are used. 

Figure 3-23 displays the final pine forest classification on a pixel-by-pixel basis for the central 

512 x 512 pixel section of the analysis area. Classes 3, 4, 5, and 6 are color-coded according to 

the labeled squares in the figure. Unassigned pixels are represented by a gray-scale image taken 

from Date 4 to provide spatial background for the classification map. Figure 3-24 provides the 

same display at a lower magnification for the full analysis area using the same color code. Class 8, 

although used in the maximum likelihood pixel assignments, is not represented in Figures 3-23 and 

3-24 because it is presumed to have negligible pine tree content as discussed in Section 3.3.3. 

59

Preaccident 

Winter/Summer 


COMP21 

Substitute 

Winter/Summer 


COMP24 

Maximum"• Class _ •Maximum--• 

SLikelihood Reference Likelihood 

Casfcto Sites lsicto 

Pixel-by-Pixel ] Pixel-by-Pixel 

Assignments 

Assignments 

CLASS21 

CLASS24 

Preaccident 

Classification Map 

of Pine Forest 

Figure 3-22. Procedure for generating final preaccident classification of pine forest. 

60

Table 3-6 defines the pine forest classes based on the discussion in Section 3.3.3. According 

to Section 3.1, the pine trees are likely to be Pinus sylvestris (ý- -')-h pine). 

Table 3-6. Pine forest classes. 

Class number 

Tree composition 

3 Predominately pines, low canopy closure 

4 Predominately pines, moderate canopy closure 

5 Predominately pines, high canopy closure 

6 Mixed pines and deciduous, pine density like Class 4 

8 Believed to contain very few pine trees 

Figures 3-23 and 3-24 show that the highest density pine trees (Class 5 in dark green) are in 

the large areas of forest southwest of the reactor site. These same areas have large sections of the 

moderate density pine forest (Class 4 in light green). Mixed forest (Class 6 in cyan) tends to 

occupy the borders of these areas. Large areas of mixed forest occur near the southern edge of the 

analysis area in Figure 3-24. The patches of pine forest directly west of the reactor station within a 

few kilometers are dominated by Classes 3, 4, and 6 with essentially no high density pine stands. 

Likewise, across the Pripyat River in tae northeast corner of the analysis area, there is little Class 5 

forest. 

In conclusion, Classes 3, 4, 5, and 6 are used in Section 7 for the analysis of radiation dose 

response ater the explosion of Unit 4 at the Chernobyl nuclear power station. 

61

Figure 3-23. Final preaccident classification of pine forest for the 12.8 km square area analyzed in 

Volume 2. Class numbers according to Table 3-6. 

62

63/4k

SECTION 4 

RADIONUCLIDE FALLOUT FROM THE CHERNOBYL ACCIDENT 

This brief section summarizes the release, deposition, and composition of radioactive fallout 

from the Chernobyl accident. 

,*.1 RELEASE AND DEPOSITION. 

The most reliable description of the daily radionuclide release from Chernobyl is provided in 

Table 4.13 of Annex 4 of USSR (1986). Some 24% of the total release occurred on April 26, 

1986. During the next five days (4/27-5/1), the release dropped considerably, declining each day. 

From 5/2 to 5/5, the release rate increased again, peaking on 5/5 to a level roughly 2/3 that on April 

26. Then, successful mitigative efforts reduced releases on 5/6 and afterward to levels 

insignificant in comparison to the releases over the first 10 days. 

During the initial 10-day release, the primary wind direction changed almost daily. On 

April 26, the wind was toward the west. During the ensuing nine days, it shifted in clockwise 

fashion toward the north, then the east, the south, and southwest (Izrael, Petrov and Severov, 

1987; NEA, 1987). This indicates that fallout deposition to the west of the reactor occurred 

primarily on April 26. The report of Izrael, Petrov and Severov (1987) strongly suggests an 

instantaneous spread of radioactive debris directly to the west immediately after the explosion. 

Continuing releases over the next 24-36 hours appeared to have been spread in directions ranging 

from southwest to northwest. There appears to have been no westward plume during the 10-day 

release after 12 noon on April 27th. This information indicates that the source term relevant to the 

westward-extending plume can be described as an acute (< 1 day) deposition episode occurring 

early in the day on April 26. 

4.2 RADIONUCLIDE COMPOSITION. 

The radionuclide composition of the close-in fallout can be approximated by the discharges 

reported by the Soviet experts (USSR, 1986). The primary radionuclides listed by the Soviets 

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 

of these radionuclides have short-lived radioactive daughters in equilibrium with them, which 

should be add to the list (DOE, 1987). These daughter products include 9 9 mTc, 9 5 Nb, 1 0 6 Rh, 

140 La, 1 44 Pr, and 90 Y. In addition, measurements of airborne radioactivity outside the Soviet 

Union indicated the presence of 1331 in large quantities (Lange, Dickerson and Gudikson, 1987), 

so this radionuclide was also added to the list for our dose rate analysis. It is possible, and is in 

fact likely, that radionuclide fractionation within the debris occurred after the material escaped to 

the atmosphere (DOE, 1987); thus the radionuclide composition of fallout debris could differ 

65

somewhat from that reported by the Soviets. A major effect is the condensation of volatile 

radionuclides such as 1311, 132 Te, and 137 Cs onto refractory particles (DOE, 1987). However, 

this is a kinetic process subject to complexities of time, space, microphysics, and meteorological 

conditions not specifically addressed in this effort. Therefore, in the absence of specific 

information relative to the dominant, westward plume, it is assumed that the relative quantities of 

radionuclides in the fallout that affected the vegetation are the same as those reported by the Soviet 

experts (USSR, 1986), with the addition of radioactive daughter products and 1331, as noted 

above. These radionuclides are used for dose and dose rate calculations in Section 5 and Appendix 

A where tables of relative quantities and radioactive properties are provided. 

66

SECTION 5 

RADIOBOTANICAL AND DOSIMETRIC CONSIDERATIONS 

This section discusses radiobotanical and dosimetric considerations regarding the observation 

and interpretation of radiation damage to pine forest. Tree characteristics such as age, 

morphology, and annual growth cycles play an important role. The interplay of ionizing particle 

range and depth of sensitive tissues in trees determines how the forest responds to the distribution 

of fallout. Detailed calculations are required to judge the relative importance of beta and gamma 

radiation for the induction of observable foliage response. As with most biological systems, the 

response of pine trees to radiation exposure depends not only on dose but on the rate at which the 

dose is delivered. Finally, specific characteristics of the radionuclide mix and how it was released 

during the Chernobyl accident influence the course of radiobotanical response. 

5.1 TREE CHARACTERISTICS RELEVANT TO RADIATION DAMAGE. 

Our satellite imagery analysis and reports such as that of Bohlen (1987) leave little doubt that 

the predominant radiobotanical effects were observed on pine trees. The dominant species of pine 

in the area is assumed (Painter and Whicker, 1993) to be Scotch (or common) pine, 

Pinus sylvestris. Figure 5-1 shows an example of a six-foot Scotch pine grown commercially in 

Maryland that was cut down in December. The date of the Chernobyl accident, 26 April 1986, 

suggests that the trees were in the spring growth phase. In this phase the trees are probably the 

most radiosensitive to acute or short-term irradiation (Woodwell and Sparrow, 1963). On average, 

the growing season begins about 11 April in the region (Anonymous, 1962). 

The radiobiological response of pines is also affected by the age and size of the trees, with 

seedlings being the most radiosensitive and large, healthy trees the least sensitive (Sparrow, 

Schwemrnmer and Bottino, 1971; Sparrow, Rogers and Schwemmer, 1968; McCormick, 1967). 

Tree dimensions and leaf density are also important in the context of understanding the fallout 

exposure scenario. Higher and denser tree canopies initially retain more of the fallout particles on 

foliage (Chamberlain, 1970), also, the lesser is the gamma radiation exposure from fallout on the 

ground surface (Beck and de Planque, 1968). 

It is most likely that irradiation of pine apical and lateral meristems at Chernobyl from fallout on 

the foliage was due primarily to beta particles rather than gamma photons based on a variety of 

dosimetric studies in the past (Aleksakhin, Tikhomirov and Kulikov, 1970; Broido and Teresi, 

1961; Mackin, Brown and Lane, 1971; Kantz, 1971; Rhoads et al., 1969). Even though the 

relative biological effect (RBE) of beta and gamma radiation is similar (~1), the distinction is 

important because the beta radiation dose rate from fallout on foliage is expected to differ 

significantly from that of gamma radiation due to fallout on the ground; and, it is well known that 

67

dose rate is a very important modifier of the radiobiological response (Sparrow, Schwemmer and 

Bottino, 1971). The dose rate versus time for beta and gamma exposures differ because 1) as time 

passes, fallout particles are lost by weathering from foliage (and concurrently build up on the 

ground) and 2) the decay schemes depend on radionuclide species and radiation energy spectra. 

h~20 

Figure 5-1. An example of Pinus sylvestris. A six-foot specimen cut and photographed in 

December in Maryland. 

No precise information was available for this study on the average height, foliar density, or age 

of the radiation-affected pines near Chernobyl. However, a forest is not usually referred to as 

"forest" unless most of the trees are relatively mature, and several published accounts do refer to 

"forest" (Asmolov et al., 1987; Bohlen, 1987). In addition, photo-pictorial documents of the 

accident's environs reveal mature trees in the background. Therefore, in this analysis we assume 

that most affected trees were mature and averaged 12 to 15 m in height. This assumption is 

consistent with a study (Painter and Whicker, 1993) of the area surrounding the Chernobyl 

68

Nuclear Plant where the pines can be 20-30 m in height on solid soil but can be stunted to only 6 to 

8 m high in the bogs that exist in the vicinity. 

We also assume that the canopy was sufficiently dense to intercept initially 50% or more of the 

fallout particles. This assumption is consistent with a calculated mean green biomass (dry weight) 

of at least 2 kg/im 2 for boreal forests (Rodin, Bazilevich and Rozov, 1975). Using the 

Chamberlain (1970) filtration model and a conservative foliar interception constant of 

0.4 m 2 /kg (Whicker and Kirchner, 1987), the predicted foliar interception fraction is 

1-exp(-0.4 x 2.0) = 55%. 

5.2 BETA RAMIATION EXPOSURE OF PINE MERISTEMS. 

To address the question of whether beta exposures could be significant for pines, we consider 

the geometry of pine needles and meristems. If there were a sufficient thickness of nondividing 

tissue to protect the meristem by absorbing the beta energy, then the dominant exposure mode 

would be gamma radiation. This issue involves a review of the histology of pines, as well as an 

assessment of the beta particle energies of the radionuclides composing the fallout. According to 

Biatobok and Zelawski (1976), the shoot apices (within which reside the dividing cells of the 

meristematic tissue) vary considerably in shape and dimensions depending on the stage of 

ontogenetic and annual development. However, the larger apices in Pinus ponderosa (which has 

similar apex morphology to Pinus sylvestris), are about 500 gim in diameter and 120 gim in height. 

These apices are usually surrounded by scales some 200-500 gim thick. While some airborne 

particles might find their way underneath these scales and lie in direct contact with the apex, most 

particles would settle on the outer surfaces of the scales, and their beta particles would have to 

traverse this layer to reach the meristem. 

From the descriptions and photomicrographs of longitudinal sections of Pinus sylvestris apices 

in Biatobok and Zelawski (1976: pp. 207-209), we estimate that the mean distance through 

nondividing tissue that a beta particle would have to traverse in order to reach the meristem would 

be about 400 im, with a lower limit of 100 gm and an upper limit of about 1200 gim. If the 

tissues were fully hydrated (as they should be in late April), the tissue density would be about 

1,0 g/cm 3 . 

The range of a beta particle having an initial kinetic energy of 0.2 Mev is 0.04 g/cm 2 (Public 

Health Service, 1960) or 400 g in water. Thus, beta particles having energies exceeding 0.2 Mev 

should, on average, be able to penetrate into the meristem of Scotch pines if the radionuclide is in 

contact with the scales directly over the apex. All of the radionuclides under consideration in this 

report have beta transformation energies in excess of 0.2 Mev; they range from 0.41 to 4.83 Mev 

(Public Health Service, 1960). Indeed, many of these beta particles exceed 0.5 Mev and would be 

able to traverse well over 1200 jim of unit density material, which would likely correspond to the 

69

upper limit of protection afforded by the nondividing tissue. We conclude from the probable 

dimensions of the Scotch pine apices and the energies anticipated of beta particles from the 

predominant fallout radionuclides, that beta particle exposure must be considered in the total doses 

received by the pine trees around Chernobyl. Radiation transport calculations are presented in 

Appendix A for dose depths relevant to both needles and meristematic tissue. 

5.3 BETA VERSUS GAMMA EXPOSURE. 

Evidence from published reports is cited in this subsection to implicate beta particles as the 

dominant exposure mode for the apical meristem of pine trees. Estimates of the ratio of the betadose 

to the gamma-dose components for various parts of the upper canopy appropriate for pine 

forests near the Chernobyl nuclear power plant, however, require models and calculations 

specifically designed for that purpose. Such calculations are presented in Appendix A and 

discussed in Section 5.3.2 below. 

5.3.1 Values from the Literature. 

Theoretical calculations by Osanov, Tissen and Radzievsky (196% show that beta depth doses 

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 , 

depending on time after fission (which affects the radionuclide mix and hence the beta energy 

distribution). These values are 0.13 to 0.38 of the beta exposure at a depth of only 0.005 g/cm 2 

(50 gtm). 

Other theoretical considerations by Broido and Teresi (1961), show that for equal fallout 

depositions on skin and the soil surface, "the 13-dose at the surface of a contaminated individual 

would be approximately forty times the y-dose measured 1 m above the contaminated surface" 

(assumed to be of infinite extent). Elevating the individual to a height 12 to 15 m above the surface 

(as is the case for Fine meristems of the upper canopy), would decrease the gamma exposure rate 

due to the ground surface fallout to roughly one half that at I m (Beck and de Planque, 1968). The 

y-dose component to the skin (or foliage) due to the fallout on the skin would be approximately the 

same as that I m above the ground. 13-radiation from ground surface fallout does not contribute to 

exposure in the upper canopy since the air between the ground and 10 m up is equivalent to about 

1 cm of water and further shielding is provided by the lower branches. Finally, the estimated 13- 

to-y exposure ratio is about 40:1.5 or 27 for equal deposition on the foliage and on the ground. 

Actual measured doses to vegetation from close-in fallout debris at the Nevada Test Site 

showed that beta doses exceeded gamma doses by more than an order of magnitude (Kantz, 1971). 

In this case, the vegetation comprised shrubs about 25 cm above the soil. In the same setting at the 

Nevada Test Site, Rhoads et al. (1969), demuistrated that mortality of desert vegetation was 

caused by beta particles rather than gamria radiation. 

70

Finally, the world-renowned Russian radioecologist, R. M. Aleksakhin (Aleksakhin, 

Tikhomirov, and Kulikov, 1970) has the following to say about beta radiation damage to trees of 

the coniferous forest: 

"On the basis of data obtained in a study of global fallout, it has been established that 

the coefficient of primary retention of the most important fission fragments in middle-aged 

plantings is not less than 40% - in dense conifer plantings radioactive substances may be 

completely retained." 

"-- the duration of the period of half-purification (weathering half-time) may fluctuate 

from two weeks in plantings purged well by the wind and washed by precipitation to three 

or four months in dense coniferous plantings." 

"--the highest radiation doses (from fallout) will be obtained by the crowns of woody 

plants in the topmost layer" 

"--the needles and buds are comparable in size with the run length of 13 particles -- and 

all cells of these tissues prove ko be accessible to 03 radiation --- a considerable portion of 

the 13 energy will be absorbed in meristematic tissue" 

"-- with radioactive contamination of the crowns, the main contribution to the radiation 

dose of meristematic tissues will be made by 13 radiation." 

"-- radiation on the crowns in the topmost layer -- will exceed by ten or more times the 

radiation dose of mammals -- under the forest canopy." 

5.3.2 Calculations. 

An estimate of the dose of radiation from fallout must account for 1) the spectrum of radiation 

emitted by the radionuclide mix, 2) the transport of radiation from source to exposed foliage, and 

3) the relative exposure rates of beta and gamma particles within the pine tissue. Appendix A 

presents detailed calculations for specific geometrical arrangements of fallout source and foliage 

accounting for these factors. Additionally, quant-tative estimation of the relative importance of beta 

and gamma ray exposures of pine tree tissues requires a model for the distribution of fallout in the 

forest. This distribution involves the initial arrangement of fallout radioactivity at the end of the 

deposition episode and the time dependent redistribution caused by weathering. 

First, we consider the distribution of fallout at the end of the deposition episode. Figure 5-2 

illustrates a simple model for visualizing the distribution of fallout particles in the forest. We will 

not treat horizontal gradients of fahlout concentratior, in our calculations so horizontal movement 

caused by wind is neglected. Fallout particles stick (aadiere) with some probability when they 

encounter a surface within the canopy. This probability varies with particle size, shape, 

composition and the nature of the encountered surface. The overall fraction of radioactivity that is 

71

etained in the canopy at the end of the deposition episode is the foliar interception fraction, f, 

discussed in Section 5.1. 

Fallout 

Canopy> 

77.7• 

Foliar interception fraction f 

'. 

Ground deposition fraction (1 f) 

Figure 5-2. Illustration of fallout distribution described by a foliar interception fraction with 

neglect of winds. 

The fraction (1-f) of the incident fallout radioactivity not intercepted by foliage is assumed to 

fall uniformly on to the floor of the forest. Thus, the two sources of radioactivity are the fallout 

assumed to be uniformly distributed in the foliage and the fallout deposited on the ground. We 

refer to these two sources as the canopy source and the ground source, respectively. We assume 

that the canopy and ground sources have the same radionuclide composition so that both emit the 

same spectra of gamma and beta radiation. 

In Appendix A, the canopy source is further divided into two contributions to improve 

calculational accuracy. The first consists of fallout particles in direct contact (deposited on) the 

surface of the foliage element under consideration. The second is the fallout deposited on all other 

foliage throughout the canopy volume. Calculations for the first contribution, called the foliage 

contact source, use an estimated surface density of fallout on the foliage element. Calculations for 

the second contribution, referred to as the canopy volume source, use a mean volume density of 

72

fallout distributed uniformly throughout the canopy. For the transport of radiation from the ground 

source and from the canopy volume source, the canopy mass (exclusive of tree trunks) is 

approximated by its mean density over the volume of the horizontal layer occupied by the canopy. 

Table 5-1 summarizes the ratio of beta radiation doses to gamma radiation doses for the foliage 

contact source and the canopy volume source. The dose due to the foliage contact source is 

overwhelmingly due to beta radiation for all the dimensions of foliage elements chosen. The 

homogenized canopy volume source produces beta dose rates that are two or three times larger than 

the gamma dose rates at the same location. 

Table 5-1. 

Ratio of beta to gamma doses to foliage calculated from the results of Appendix A 

for fallout retained in the canopy. 

Foliage Contact Source 

Dimensions offoliage element 

Radius (p) Length (21) P3/ ydose tatio at center of element 

(cm) 

(cm) 

0.04 4.5 58.8 

0.04 9.0 59.1 

0.0625 9.0 46.3 

0.10 4.5 34.7 

0.15 9.0 25. 

0.40 2.0 8.5 

Canopy Volume Source 

Dose point at: middle of canopy 3.2 

top of canopy 2.0 

In order to estimate the beta to gamma dose ratio for the total absorbed dose to a foliage 

element, doses from all radiation components are summed including the ground source. A model 

for relative densities oi the fallout sources is specified and results from Appendix A are utilized 

calculate total reference dose rates or doses for foliage elements. 

The complications from the wind-driven, horizontal component of fallout motion and the finite 

size of forested areas are neglected. The calculated dose rates to the foliage are referenced to the 

same amounts of radioactive fallout distributed throughout the canopy mass and on the ground 

surface below at the end of the deposition episode. Accordingly, on the basis of a unit source 

density, reference dose rates to the canopy are calculated for I P3-particle or y-ray per cm--sec on 

the ground surface and 10-3 3-particles or y-rays per cm 3 -sec within the canopy volume, based on 

73

vertical thickness of 10 m for the canopy. For summing dose contributions at a given dose point, 

the canopy volume source density is scaled in proportion to the foliar intercept fraction f. The 

ground fallout source is scaled in proportion to (4-f). 

An appropriate model for the areal source density for fallout material in surface contact with 

foliar canopy components is less straightforward to formulate than for ground surface fallout 

underneath. Complicating factors include the geometry of canopy components, the aerodynamic 

behavior of particle trajectories in the canopy, the effective adherence of particles that make contact 

with canopy component surfaces, and weathering influences. 

In the absence of definitive information regarding the contact source density, we examine the 

effect of varying the density between the unit incident fallout density and one tenth of that value. 

The value of the initial foliar intercept fraction, f = 0.6, is taken from Kerr et al. (1971) to be 

consistent with the average foliage density used for transport calculations in Appendix A. Also, 

this value is close to that of 0.55 estimated in Section 5.1. Table 5-2 summarizes the dose rate per 

unit source for the various sources and dose points from Appendix A. Summing the doses over 

beta and gamma sources scaled appropriately for each dose point gives the dependence of the beta 

to gamma dose ratio as a function of the contact source density in the assumed range for each of the 

foliage elements. 

Figure 5-3 shows the variation of the beta to gamma dose ratio with contact source density for 

selected foliage elements at the top of the canopy. Since the dose from the contact source at the 

center of the thinner foliage elements is dominated by beta radiation, the beta to gamma ratio for 

these elements is strongly affected by the assumed contact source density factor. Thicker elements 

are less influenced by the contact source strength because the beta penetration to their centers is 

suppressed. 

It is likely that the adherence probability and, hence, the contact source density varies from one 

part of the foliage to another and is surely close to one for areas of the pine branches that are sticky 

to the touch, which are quite common. Smoother surfaces such as the needles are likely have a 

lower adherence probability. Also, there is some reduction in fallout flux as it filters through to 

lower canopy levels. We assume that a value of 0.5 for the contact source density factor is 

reasonably representative and use it for all further calculations. 

With the assumption of a contact source density factor of 0.5 and an initial foliar intercept 

fraction of 0.6, Figure 5-4 shows how the beta to gamma dose ratio changes with weathering. 

Just after deposition, the ground fraction is 0.4; it increases as weathering effects move fallout 

particles from the canopy to the ground. The values plotted in Figure 5-4 for foliage elements of 

various radii located at the top and middle of the canopy are calculated with the assumption that 

74

Table 5-2. 

Dose rate per unit source for various sources and dose points. Reference canopy 

volume source density is taken equal to the unit ground source density spread over 

the assumed 10 m canopy height. 

Beta dose rate Gamma dose rate 

Dose point 

(cGy/h x 10 7 )/unit 

source 

(cGy/h x 10 7 )/unit 

source 

Unit Ground Source 12 m 10.1 

I(f or y)/cm 2 - sec 7 m- 16.9 

1 m 185a 29.4 

Reference Canopy 12 m 33.4 16.5 

Volume Source 7 m 66.9 20.6 

10- 3 (p3 or ))/cm 3 -sec I m 16.5b 

Radius of foliage 

element (cm) 

Unit Contact 0.04 302. 5.13 

Source, Cylindrical 0.0625 236. 5.10 

Geometry 0.10 173. 4.98 

1(P or /cm 2 -sec 0.15 124. 4.95 

0.40 32.0 3.78 

aBased on J / dose ratio of 6.3 according to Barabanova and Osanov (1990). 

bAssumed equal to the value at the 12 m dose point. 

weathering lowers the canopy retention for both the foliage contact source and the canopy volume 

source in unison and that the ground fraction increases accordingly. Weathering of the ground 

fallout into the earth is neglected. 

As discussed in Section 5.2, the most sensitive growth tissues of the pine are located in the 

apical meristems at the tips of branches (and roots) and are typically less than 0. 10 cm below the 

surface of the foliage elements. Figure 5-4 shows that the calculated b-kta to gamma ratio for these 

tissues is greater than six in both the middle and upper canopy before the effects of weathering. 

Even when half the intercepted fallout has moved to the ground (ground fraction = 0.7), the ratio is 

still above four. These calculations confirm the assertions in the literature as reported above that 

beta doses dominate for the most sensitive growth tissues of the pine. 

75

14 

Dose point at canopy top 

12 Foliage element 

radius = 0.04 cm 

o 0.0625 

10..1 

8e 8 

0 

6 - 

4 

2 

,- - 0.40 

0 I I 

0.0 0.2 0.4 0.6 0.8 1.0 

Contact source density factor 

Figure 5-3. Beta to gamma dose ratio versus contact source density relative to unit areal density 

of incident fallout. Curves are for various radii of foliage elements located at the 

top of the canopy. 

76

0 

12 I 

10 

S8 

S80.0625 

0.04 cm 

a) Canopy top 

S6 - 0 10_O~ O_ 

:• 4 0 .15 - -- - -- . _- . 

o 

2 ~- 

0 

0.4 0.5 0.6 0.7 0.8 0.9 1.0 

12 1 1 1 

b) Canopy middle 

10 

84 

80 0.041cm 

S 0.0625 

S6 

2 0.40 

0 , I 

0.4 0.5 0.6 0.7 0.8 0.9 1.0 

Ground source fraction 

Figure 5-4, 

The beta to gamma dose (dose rate) ratio at the center of cylindrical foliage elements 

of various radii at a) the top of the canopy and b) the middle of the canopy. 

77

Lateral meristematic tissue, which is located in the cambium and provides for growth in the 

diameter of branches, may be better protected from beta radiation than the apical meristems. 

However, Figure 5-4 shows that tissues as deep as 0.4 cm still have beta to gamma ratios larger 

than two until substantial weathering has taken place. Figure 5-3 shows that lowering the assumed 

contact source density factor from 0.5 to 0.1 has only a weak influence on the beta to gamma ratio 

for such deep tissues. 

We conclude that beta doses are much more important than gamma doses for the sensitive 

growth tissue in the apical meristems of pine trees. At doses below the LD 50 , the dominant effect 

of radiation is on the apical meristems. Established growth is outwardly unaffected. Thus, visible 

affects may not show up until the growing season when the lack of viable apical meristems inhibits 

new growth. We expect that late developing effects from these relatively low doses are likely to be 

associated with high beta to gamma ratios for the causative dose. 

For doses much higher than the LD 5 0 , however, trees can be completely dead in a matter of a 

few weeks. Radiation response in this short time requires more than just sterilization of the growth 

tissue. The broader systemic effects associated with acute mortality (involving irradiation and 

response of all canopy foliar components) may be more influenced by penetrating gamma 

radiation. It is likely that early death of a tree from large doses is associated with lower beta to 

gamma ratios (but still not lower than one or two as discussed above) than for later occurring 

foliage deterioration at lower doses. 

5.4 DOSE RATE EFFECTS. 

It has long been recognized that the radiobiological response to a given dose depends on the 

time over which the dose is delivered. This dependency is sometimes caused by subcellular repair 

which is more effective if the exposure is protracted. In other cases, if exposure times are short 

compared to the cell division cycle time, high dose rates are clearly more damaging than low dose 

rates because of reduced recovery time during irradiation. 

This is particularly true in 

physiologically active plants which may have greater recovery potential than dormant plants. In the 

case of chronic or long-term exposures however, somewhat different patterns may emerge. For 

example, dormant plants might show a greater effect from a long-term irradiation exposure than do 

actively-growing individuals; the dormant cells can accumulate a higher total dose and repair 

mechanisms may be less efficient (Whicker and Fraley, 1974). 

Most of the research on the effects of radiation on pines has involved either acute (< 1 day) or 

chronic (> 1 year) exposures, with constant dose rates applied over the specified period. 

Unfortunately, neither of these dose rate regimes match the exposure of the pines at Chernobyl. 

The likely exposure conditions and mix of radionuclides at Chernobyl would have produced a 

78

declining dose rate with time. The rate of decline was too slow to allow the exposure to be 

considered acute, but was too rapid to approximate a chronic exposure. 

We have dealt with this dilemma by estimating a credible dose rate versus time curve for the 

westward plume at Chernobyl. From the integral of such a curve, one can estimate the length of a 

comparable constant-rate exposure. Knowing the length of a constant-rate exposure that would be 

comparable to Chernobyl, one can examine the more relevant literature to estimate the doseresponse 

relationship. A few studies on short-term (8-30 day) constant rate exposures of pines 

have been conducted (e.g., Monk, 1966; McCormick, 1967; Pedigo, 1963; Miller, 1968; and 

Platt, 1963). 

These studies involved pine species other than P. sylvestris; however, the 

chromosome characteristics of P. sylvestris are similar to those of pines in general and there is not 

a great deal of variation among the pines in radiosensitivity (Sparrow, Rogers and Schwemmer, 

1968). In addition, dose rate effects per se have been studied (e.g., Sparrow, Schwemmer and 

Bottino, 1971; Amiro, 1986) and these data will also be considered in the development of the doseresponse 

algorithm. 

5.5 DOSE RATE SCENARIO FOR CHERNOBYL. 

Because the larger part of the dose to pine meristems is assumed to have been delivered by beta 

particles, a normalized beta dose rate function is estimated from the list of radionuclides and their 

decay rates, the relative quantities released, the total beta transformation energies, and an assumed 

rate of weathering from the foliage. The relative beta dose rate (BDR) at time t (in days) is: 

where: 

BDR = Y' Ri Ei exp(-ki t) (5.1) 

i 

Ri = Estimated abundance proportion of nuclide i, MCi (USSR, 1986) 

Ei = Energy of beta transformation of nuclide i (Public Health Service, 1960) 

ki = Effective loss rate constant of nuclide i, ki = A + Xw, where Xp = physical decay 

constant and Xw = weathering rate constant, 0.0495/day (Hoffman and Boes, 1979) 

The normalized beta dose rate (NBDR) is calculated from: 

=t BDR(t) 

NBDR(t) = BDR(t = 1) (5.2) 

79

C') Ný Ný ',- Cii N Ni- N 

J CD CD t w 0 C N al C-4 

1) It 

i-C N) N' CM LO N 0' a) 

0 '-0 Cl) 00 00C0 000 0 l0 

cU0 lr N O4 N () 1, U) 1, () 00 LD U) N U) 

cm 0) 00 * 0 mt 0i it (D ONC - U)0 

o 0 ~ \ ~ \ ' 

-V- o I I 

)C 

ZO 000 00-0 01-00 0-0 0 In 0 

it O D - in U) '-T ( CO) NY N NM W) CV) 0 0 

o b0 00(0 CD') CO) D It N4 CO) N4 W N Uf) 0 to U) 

(0 CD 0 qr V) 0 C)000 U) 0)LO t CM NO-)C 

66666 C;ci i 6;C ci 6 c~i 6 uC; 

ý0 CD N - M NON vt CD 0 CD) 0 CD N C) 0 0q 

0 CD0 C 0) CD OD N`ý L I 0) C14 N In M -N N 0' N 

T- ODC) Cf(D UN C\) CD U) 0~d NO) - C1 D 0) CJ C 0) qc 

00 ) C ( D U) CD ON N D N -0 0' 00 

C ) 0-D CD N- LOOU CD ) U) ON NO CO C 0) ' 

0) W) MC i-D W C N N W M 4 NO 10MN(0 

'1 o Mi Ni C M C') Mt 

M( 0i 66o 100 C 

ON~~~ 0-0))CD0)iN))0 N 

0 0 

Nu0 NC CD 010 IN N O CD M' N' U) '-N 

13C' 0 ON 0)- M4 C i-CD U) M '4 )q CD CD 

0 U) C') C' CD a 0 

04' 04 CD 1000 00 

(- 0' 0 i0 CY ')D ') 14 ' N U) OU 0 00) LO N O 

-), 00-ID CO'q ) N D U C DD I -o N c) 6D6* - Nr -a 

C~~~~~~~l~~ 0 0 0 0 C) 

0 

~ 

0 t- ) T- 10 OCD tn0N (0(0 ) CU ON U) (D '4- M4 

u - 

o)'-N O O N 0 0 W 0 0 00 t 

C') COD 

UL U)CýC 

*00 

0 

N 

-J 0 

~) U) C ) + + Cu (0)i + + 

4) WN~~ (1 ) (00 cc '41) (D ~ + 06-n9 

~ ' ' e ' 

LC) 0)1 0 wZ a ) ' 

0- 

80

The values for Ri, Ei, ki and BDR(t) for each radionuclide (see Section 4) is summarized in 

Table 5-3. The normalized beta dose r,:! s (NBDR) are plotted through 60 days in Figure 5-5. 

For comparison, there are also shown a normalized gamma dose rate curve based on aerial surveys 

(Asmolov et al., 1987) and a normalized gamma exposure rate curve that is calculated for the list of 

radionuclides in Table 5-3. In the latter calculation, the relative gamma dose rate (GDR) for 

radionuclides on the ground is estimated from: 

GDR = 7, Ri Gi exp(-Xpi t) (5.3) 

where: 

R i = is the same quantity as in Equation 5.1 

Gi = gamma dose factor for nuclide i in tR/hr per mCi/km 2 (Beck, 1980) 

Xpi = physical decay constant for nuclide i 

As with the BDR, the GDR values were normalized to 1.0 at t = 1 day. Table 5-4 summarizes the 

calculations for the normalized gamma exposure rate curve. 

Inspection of Figure 5-5 reveals that the estimated dose rate curve for the beta component is 

very similar to the aerially-measured gamma component for the first two weeks, after which the 

beta curve declines more rapidly. The shapes of the calculated and measured gamma exposure rate 

curves are similar after the first two weeks. It is not clear exactly how the aerial measurements 

were made. If they were taken just above the forest canopy and if a large fraction of the fallout 

was initially retained by the canopy, then the aerially-measured curve would include a beta 

component and should resemble the calculated beta curve for the first couple of weeks, since both 

would be affected by foliar weathering (Xw) as well as the mix of rqdionuclides. The more likely 

situation is that the aerial measurements were taken sufficiently above the canopy to avoid turbulent 

propeller down wash that might disturb measurement . If this were the case, the measurements 

would be largely of gamma radiation. As the fallout material weathered from the foliage and 

accumulated on the ground, the aerial measurements would be expected to more closely approach 

the calculated gamma component arising from the soil, and this seems to be the case. The 

somewhat widening increase with time between the measured curve (dashed) and that from gamma 

emitters "on soil" (solid) could reflect weathering, both from the foliage to the ground and from 

migration of fallout into the soil over time. Based on this discussion, it appears that the relative 

shapes of the three curves in Figure 5-4 are compatible with the essential facts and with plausible 

assumptions we have made. 

81

1.0 

From gamma emitters 

- •" on soil (GDR) 

"0.1 

U _Aerial survey Z. 

Z _(Asmolov et al. 1987) 

From beta activity 

on foliage (NBDR) 

0.01 

0 10 20 30 40 50 60 70 

Time (days after 4/26/86) 

Figure 5-5. 

Calculated dose rates (normalized to 1.0 at t = I day) from gamma radiation 

emanating from fallout on the soil and from beta activity on the foliage. 

82

~ 3)TV ~ V C E CV) CE)ý 

W C W111 1 w w w w w II 

M~ to CE) CV) to UD CE) CO) N4 () It t. C. 

L WWWWWbWWWIII LI b L L L L 

W CD 

N~ 100 0 CE) N- cr -U) IN CE) 0: 0 U- N, Ný 

C') co mE t- coE) VDC) in c) NO 0 t q N d 

o NU)10ws 4 40; NýCýi-N 0) 

4 ~ ~ N It ~C) CM) 

ItN CV) NVN O CY V) It Lc 

b W WW WWWW W W WWWII W cr) 

'R 10 ME W- W- 0 ON CD l 0 'I W N 

U D CE) 10'U N 'It "I Pý r- c0i 

00 - 0) (D Cl ( 0 Ci -NO'( T- N1 N - cm U 

N c4 ) C') CE) CV) N NC' ) N CE) 'r q V 

S0 '0 M NCDN 0 t W -tCE)q N 1 

nE i-C CE) 00 N ) M M0) N 

. . . . 

cz 

0m 

NN 

-q- NO N 0- m o m N R 

00 aE 0 It NO 0 r- N) ' 0 

N C) CE CE CE N N CE) i .t 0- C) 6V 

> m 0 o0 N 0 0) N 0 0 m) 100 CN E) 

0 0 0C CO) N 0)1 (D Itt CY) 0 0 Y N D 

N E) C) CE) N N CE) i-CE CE P, 6 

00D ro 10- Nb NL C) 11-0 Ni a,1.11.1O NL ) 

CM 0O C\ N\ 0' 

0 1 .-LWUJWWWWWWUWW 0) 

"M CON C C'l ON 00' CE ) N 10N 

0) Nl (p i-i '- N c Ný Ný q ! 

00 

CE) CE RN CE) CE ( CE) N R C CE) N) C 

M M It N O 

N 0 M V 

&CE~e'Ow 00 

00 

W2 

u Ei 0 0 

o) ,- - 

0Y) + + + C 

0 L)0 a:cN c Z 

n cV)C) CE) CE) 0)LO000 lqT It mi E ' 

'- -' '-)0)' i i ~ i N NV) itl 1 if 

83

None of the three dose rate curves in Figure 5-5 would produce an "acute" exposure. Of the 

three, the beta dose rate curve declines the most rapidly and hence provides the shortest effective 

duration of exposure. The beta curve is considered the most plausible representation of the actual 

dose rate curve experienced by meristems of the damaged pines; its time-integral is plotted in 

(Figure 5-6). An exposure following this curve produces 50% of the total exposure in nine days 

and some 80% of the exposure in 21 days. A subjective estimate of the exposure time at a constant 

rate that would produce a similar biological response is about 21 days. Note that from day 2 to day 

12, when nearly 50% of the dose has been delivered, the dose rates of the constant rate and 

calculated beta rate curves are quite similar. We assume that the lower dose rates for the constantrate 

exposure for the first few days would be roughly compensated for by the higher dose rates 

after day 12. After 20-30 days, the additional cumulative exposure from fallout is probably not 

very significant in terms of biological response. 

From this analysis, we surmise that experiments in which constant-rate exposures are delivered 

to mature pines for periods ranging from roughly 2 to 4 weeks during the early growing season 

should produce dose-response relationships comparable to those for the Chernobyl forest. 

5.6 Summary. 

Important results from Section 5 regarding the radiobotanical response of the pine forests near 

the Chernobyl acrdent include: 

1) beta radiation doses are expected to dominate gamma doses by at least a factor of six for 

late foliage responses (after months) induced by doses less than the LD 50 , 

2) beta radiation dose contributions are expected to be at least one or two times the gamma 

contribution for large doses that cause tree mortality within a matter of weeks or days, 

3) the effective exposure time for foliage doses from beta radiation near the power station 

determined by the decay of radionuclides and weathering of fallout from the foliage is 

estimated to be about 3 weeks, and 

4) results of constant dose rate experiments with pines exposed during their growing seasons 

for periods of 2 to 4 weeks may be used to interpret the observed responses at Chernobyl. 

84

0 

z 

cc$ 

_0 -0 

cu 

co 

S -4 

CLD 

xN 

o 

a c 

0o 

ci) 0 

aso 

013ils 

asopeleqAII~nwE 

0-.0'0 

485/8

SECTION 6 

DOSE-RESPONSE RELATIONSHIPS FOR PINE TREES 

Based on a review of radiobotanical literature for pine tree response to ionizing radiation 

exposure, this Section examines the dependence of pine tree response on dose and dose rate and 

the relationship between time-to-response and dose for short term exposures relevant to the 

Chernobyl accident. A quantitative relationship is derived relating the time of earliest detection of 

response in multispectral imagery to the dose received by pine trees. 

6.1 LITERATURE REVIEW. 

Stress to foliage changes its spectral reflectance in various wavelength bands. Temperature, 

water content and leaf pigments largely determine the observed spectrum. Reflectance changes 

observed in the tree canopy in response to radiation stress are probably primarily due, initially, to a 

reduction in photosynthesis and an increase in dark respiration (Ursino, Moss, and Stimac, 1974). 

These effects are caused by damage to shoot apical meristems which retards growth and function 

of photosynthetic tissues (Bostrack and Sparrow, 1969). Eventual death of trees likely results 

from starvation brought about by a critical reduction in the amount of photosynthetic tissue 

(Bostrack and Sparrow, 1970). Thus, dying and dead trees should reveal a progressive decline in 

the amount of green tissue and water content of the foliage. As the foliage dries and falls from the 

branches, reflectance should become progressively similar to open fields or even to bare ground if 

the understory vegetation is sparse. The rate of progression of these changes is expected to depend 

on dose and season of the year. Because of the nature of these changes, it is appropriate to focus 

on mortality and growth reduction as the damage endpoints likely to be revealed by satellite 

images. 

Based upon measured chromosome characteristics, Sparrow, Schwemmer, and Bottino (1971) 

estimated the acute LD50 exposures for 82 woody plants, including 13 species of pine. The 

predicted LD 50 exposure for Pinus sylvestris was 620 R. The comparable values for the other 

pines ranged from 410 to 770 R, with an overall mean of 615 ± 97 (1 s.d.) R. These values 

were all for a 16 hour constant rate exposure. Sparrow, Schwemmer, and Bottino (1971) also 

provide some data on the total dose required to produce an LD 50 for a given constant rate exposure 

time. The data were not developed from pines, but pines are expected to behave similarly. The 

relevant data are plotted in Figure 6-1 relative to the LD 5 0 for a 16-hour constant rate exposure. 

Unfortunately, the data do not reflect exposure times over 36 hours. A linear extrapolation to 

exposure times of 2-4 weeks, however risky, predicts LD 5 0 values for Scotch pines of 

1300-1500 R. 

87

00 

_ _ _ _ _ _ 

_ _ _ _ _ _ _ _ _ _) 

E 

oo0 

-CI 

t 

00 

03 u0 

CL 0 

0)nsdx q 9L -'-10C- 

88L

Sparrow, Schwemmer, and Bottino (1971) also indicate that a fallout decay simulation (FDS) 

treatment is comparable to an 8-hour constant rate exposure. 

However, this is based on 

experiments in which the decay rate followed the t -1.2 relationship with I hour as the initial 

reference time. In this case, the exposure rate at 1 hour would be 16 times that at 10 hours. There 

is no indication, to our knowledge, that the initial dose rates at Chernobyl were that much higher 

than the 10-hour dose rates. This would require a very acute deposition of fresh fission products 

(as from a weapon explosion) with no prior inventory buildup of longer-lived fission products as 

would be expected with used reactor fuel. 

The study by Miller (1968) appears quite relevant because it involved a 29-day constant rate 

gamma exposure to Pinus palustris trees during the month of April. Sixty percent of the 

population survived an exposure of 2,100 R. Complete mortality occurred within three months at 

8,700 R or greater. We estimate the LD 5 0 to be about 2,600 R from a plot of survival versus total 

exposure. Noticeable (- 12%) mortality occurred at an exposure of 700 R. A 50% reduction in 

growth was observed at approximately 500 R. In Miller's study, 29-day irradiations of nearby 

plots were also conducted in summer, fall, and winter. The LD 5 0 for the summer and fall 

irradiations was only slightly higher than that of the spring irradiation. In the case of growth 

reduction, the spring, summer and fall irradiations gave nearly identical results. In winter, the 

pines were considerably more resistant. 

Monk (1966) also studied the response of Pinus palustris to a constant rate gamma exposure. 

In this study, the trees (which were 5 years old) were exposed for 16 days in mid-May. A 50% 

reduction in growth was observed at 400 R, a value similar to that reported by Miller (1968). The 

exposure required to kill all the trees by the end of the growing season was about 9,300 R, again 

similar to the Miller (1968) study. Data on partial mortality were not reported. 

A third study of Pinus palustris was carried out by McCormick (1967). In this case, the forest 

was exposed at a constant rate for 8.3 days during August. All these pines < 5 years of age 

receiving 800 R or more died within 4 months. Older trees (up to 12 year old) required 2,800 R 

for complete mortality. This forest also contained some Pinus ellottii. All individuals of this 

species receiving > 300 R died and a clear relationship between plant size and radiosensitivity was 

observed. In all cases, the larger trees were more resistant. Microenvironmental changes were 

also observed at various exposure levels. In areas receiving > 2,000 R, temperature gradients 

were more like those of open fields than forests. 

The effects of mixed neutron-gamma radiation in air emanating from the Lockheed Aircraft 

Corporation reactor in northern Georgia on Pinus taeda and P. rigida were studied by Platt (1963) 

and Pedigo (1963). Most of the exposures were delivered in a 2-week period in June 1959 and a 

3-week period in August 1960, so the dose rate regime should be relevant to Chernobyl. After the 

June irradiation, pines receiving more than 7,500 rads turned reddish-brown within a few days and 

89

were dead within a few weeks. Those receiving about -',000-6,000 rads took much longer (up to 

8-10 months) to dic. Doses of about 1,000-3,000 rads caused death of terminal buds, inhibited 

reproduction and growth, and reduced photosynthesis, but caused little mortality. A complication 

in this Lockheed reactor study was the presence of neutrons, which may have an RBE greater than 

1, but we have no data to assess this effect for pine mortality or growth. The gamma/neutron dose 

ratios reported in Platt (1963) ranged from 0.5:1 to 3.0:1, depending on location. If the neutron 

RBE is greater than 1 for these effects, the doses of pure gamma radiation needed to produce the 

same effects would be higher. 

Donini (1967) studied the histological response of five year old Pinus pinea and P. halepensis 

to various total doses and dose rates. From such data, the relationship of dose rate to total 

exposure time to produce plant death was estimated. The exposure times ranged from 25 to 380 

days for P. halepensis. Both curves plotted as a straight line on a log-log scale and were thus 

extrapolated back to a 20-day exposure period. This yielded a lethal total exposure of 860 R for P. 

halepensis and 3,000 R for P. pinea. These species have nuclear volumes of 1,000-1,100 4tm 3 

(Donini, 1967), from which one would predict acute (16 hr) LD 5 0 values of roughly 700 R 

(Sparrow, Rogers, and Schwemmer, 1968). The large discrepancy in radiosensitivity between the 

two species is puzzling in view of their similar nuclear volumes. Nevertheless, it is evident that to 

produce an LD 50 a 2-3 week irradiation requires a somewhat to much larger dose than does a 

16-hour exposure. 

A nine-day gamma exposure in autumn of a pine birch stand in the Soviet Union resulted in an 

LD 1 00 for pine of 5,000-6,000 rads (Karaban et al., 1978). The threshold dose for obvious 

damage to the buds and needles was about 800 rads. 

Two excellent chronic gamma irradiation studies on pine forests were carried out by Woodwell 

and Rebuck (1967) and by Amiro and Dugle (1985). Unfortunately, it is not possible to infer from 

their data lethal doses to pines subjected to 2 to 3-week exposures. Nevertheless, both of these 

studies support the concept that prolonged exposures are less effective than acute ones. 

Approximately 80% mortality to Pinus rigida was observed after 7,000 R was delivered over an 

11 -month period (Woodwell and Rebuck, 1967). A total dose of 8,000 rad over a period of about 

2 years produced significant (- 80%) mortality in Pinus banksiana (Amiro and Dugle. 1985). 

One year old Pinus sylvestris seedlings were exposed to various constant exposure rates for 

150 days in Canada (Sheppard, Thibault, and Guthrie, 1982). Needle growth at 17 rad/day 

essentially terminated after 60 days with a total dose of about 1,000 rad. Relationships in 

Klechkovskii, Polikarpov, and Aleksakhin (1973) indicate that this dose would need to be about 

1,700 rad to cause severe growth inhibition and result in total lethality of the population. The ratio 

in radiosensitivity for seedlings relative to adult trees is reported by McCormick (1967) to be about 

90 

MUM

3.5 for P. palustris, so that a dose of about 6,000 rad delivered over 60 days should kill all adult 

Scotch pines if P. sylvestris shows the same ratio. 

A study by Amiro (1986) on 2 year old Pinus banksiana seedlings examined the effects of 

both dose rate and total dose on the relative growth of the plants. Converting his multiple 

regression predictive equation to the units used in this report: 

RG = 1.072 - (2.303 x 10-4) D - (2.369 x 10-3) D/t 

where: 

RG = growth rate relative to controls 

D = total dose in raf 

t = time of the exposure in days 

The third term in the equation above is the dose rate factor. For an exposure time of 21 days and a 

measurable growth response of RG = 0.8, this equation predicts a dose of - 800 rad. A severe 

growth inhibition (RG = 0.2) is predicted for a dose of 2,500 rad. If this dose corresponds to 

0.6 of the LD100 (Klechkovskii, Polikarpov, and Aleksakhin, 1973), the LD 100 would be about 

4,200 rad for a 21-day exposure. 

6.2 ANALYSIS OF DOSE RATE DATA. 

We have plotted the relevant data from the literature review discussed above such that all 

estimates can be given some weight in developing a dose rate or exposure duration relationship for 

assessing the pine tree damage in the Chernobyl forest. The total dose required to produce three 

endpoints of damage is plotted against the duration of the constant-rate exposure in Figure 6-2. 

The LD 100 represents the minimum dose required to produce 100% mortality to the pines, while 

the LD 5 0 represents the lethal dose to 50% of the population. The GR 5 0 represents a 50% 

reduction in growth rate, relative to controls. Generally, lethal effects were scored within the 

current growing season for exposure periods < 60 days. In the case of the longer exposure 

periods, the effects were generally scored within 1-3 years after the start of irradiation. Growth 

rate effects were usually scored within the current growing season for exposure periods < 60 days. 

The lines in Figure 6-2 are drawn by eye. 

The point estimates of the dose required to produce the three endpoints of damage for a 21-day 

exposure period as read from Figure 6-2, along with uncertainty bounds, are listed in 

Table 6-1. The uncertainty bounds are subjective, but they include a 1-week uncertainty in the 

effective exposure period at Chernobyl, as well as the scatter in the data. The suggested 

uncertainty bounds are within a factor of 1.8 of the best point estimate. Because the data are based 

91

H i l l I I 1 111 1il 1 1 

0~ 

ElE 

00 

CIOI 

00 

00 

ot 

G 11 

-l If3 I.2 

cc -00~ 

0 0 orFR 

O~ Z 

-0 

0.0 ~ 0 

0 0" 

0 

C~j 

(E~o)asopje-o 

92

on short-term exposures (< 60 days) received in the early growing season, the endpoints in Table 

6-1 should be reached by the end of the growing season in late September or early October. 

Table 6-1. Estimated total doses to produce three endpoints of damage to pines for an exposure 

period of three weeks (assumed equivalent to the effective exposure period at 

Chernobyl). The upper and lower bounds consider the scatter in the data from the 

literature, as well as a I-week uncertainty in the effective equivalent exposure period. 

Required Dose (rad) 

Endpointa Best Estimate Lower Bound Upper Bound 

LD 100 (complete mortality) 6,000 3,300 10,800 

LD50 (50 percent mortality) 2,300 1,300 4,100 

GR50 (50 percent growth reduction) 800 440 1,400 

"aEndpoint observed by the end of the growing season. 

We judge that the maximum stress indicated by satellite image processing would correspond 

roughly to the LD 100 . Stress at the GR 50 level is probably not detectable. At the GR 5 0, the 

vegetation will be hydrated, green, and functional, even though growth rate and reproduction are 

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 

higher than the GR 50 . The LD 50 will be detectable by satellite image processing certainly by the 

time half the trees are dead. 

6.3 TEMPORAL PROGRESSION OF DAMAGE. 

Most of the relevant studies on the effects of short-term radiation exposure on pines failed to 

score the vegetation frequently enough to document the rate of progression of the damage. In most 

cases, the damage was simply scored at the end of the growing season, and sometimes the 

following season. Fundamental radiobiological considerations would suggest that the rate of 

progression of damage should increase with total dose and dose rate, and also with the rate of 

mitotic activity. 

The study by Platt (1963) is one of the more helpful with regard to time-progression of 

damage. Quoting from this author: 

"Within one week after the June irradiation, pine trees receiving doses of 7,500 rads or 

more began to turn a brilliant orange-red and died within a few weeks. Those receiving 

about 4,000 rads took much longer to die." 

93

Multiple regression analyses of data from this study indicated that needle production in pines was a 

complex function of time after exposure, dose, and time-dose interactions. The work by Miller 

(1968), also one of the more relevant studies, showed that 100% mortality occurred within three 

months in pines receiving > 8,700 R. Unfortunately the time to death for pines receiving even 

higher exposures was not reported. 

Pedigo (1963) reported noticeable effects within a few days in pines receiving a 2-week 

exposure of 8,000 rads or more. Significant mortality was recorded within one week after the 

exposure for pines receiving > 12,000 rads, but not until about 100 days after exposure for pines 

receiving 9,000 rads. 

In the case of McCormick's (1967) study (an 8-day exposure in August), the older pines 

receiving 2,800 R took about four months to die. No additional trees died during the subsequent 

two years of the study. 

Using the limited database, Figure 6-3 shows the approximate time required to reach the LD 1 oo 

or the GR 5 0 of pines for short-term (8 to 30-day) exposure periods plotted against total dose. The 

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 

in Figure 6-2. Considering the scatter in the data and other neglected variables such as growth 

stage, temperature, and moisture availability, we estimate that the uncertainty in the time to reach 

the endpoint is roughly a factor of 2. The curve would predict that very high doses (> 12,000 rad) 

would be required for prompt (

cause growth impairment within a few days. 

6.4 DOSE VERSUS TIME-TO-RESPONSE FOR MULTISPECTRAL 

DETECTION. 

The reports of Monk (1966), Pedigo (1963), Platt (1963), and Miller (1968) provide visual 

descriptions and other data on pine tree response to radiation that allow qualitative estimation of the 

detectability of radiation response using Landsat Thematic Mapper images. These four reports are 

applicable for trees at least five years old and for springtime exposures lasting from two to four 

weeks. Table 6-2 lists data extracted from these reports. 

For various combinations of absorbed dose and time since start of exposure, Table 6-2 repeats 

with minimal paraphrasing the description of damage found in thc indicated Reference. The 

qualitative estimate of multispectral detectability for each dose/time combination in Table 6-2 is 

based on the descriptions of damage and the authors' experience with vegetation analysis using 

multispectral imagery. Factors considered include: 

1) leaf senescence in deciduous trees is easily detected, so dead pine trees with brown needles 

will be easily detected, 

94

10o 3 2 

102 0D 

04-. 

C 

40 

16 

E 

10 

GR 5 o 

LD 1 0o 

1 I Iiiiul I I I ill 

102 103 104 40,000 

Short-term dose (cGy) 

Figure 6-3. 

The time required to reach the LD 1 00 or GR 50 versus short-term (8-30 day) 

dose. Circles represent data for the LDIoO; squares are data for the GR 50 . 

Numbers represent the authors: 1 = Pedigo (1963); 2 = Platt (1963); 3 = Miller 

(1968); 4 = McCormick (1967); 5 = Sheppard, Thibault and Guthrie (1982); 

and 6 = Amiro (1986). 

2) new growth needles and old growth needles on pine trees have a different color to the eye 

and have infrared spectral reflectivities that differ by 50% to 80% through the growing 

season (Wolfe and Zissis, 1978), and 

3) minor variations in needle color or number will be masked by pixel-to-pixel fluctuations in 

canopy cover and residual misregistration of pixels. 

95

Table 6-2. Detectability by multispectral remote sensing of radiation damage to pine trees for 

spring exposures as a function of time since start of exposure and dose. Exposure 

duration of two to four weeks. 

T.ie since 

start of Estimate of 

Dose exposure multispectral 

(Gy) (days) detectabilitya Description of damageb Referencec 

3. 330 X No visible evidence of damage, terminal and 3 

lateral growth are 60-65% of normal 

3.5 90 X All trees alive, 85% survival of terminal 4 

buds, terminal growth 70% of normal 

5. 142 X Terminal buds alive, stem elongation 

reduced 

11.6±1.7 142 X Terminal buds dead, proliferation of lateral 

bud formation with subsequent growth 

20. 20 x No unusual needle fall 2 

(16- 23) 

7. 90 0 12% survival of terminal buds, terminal 4 

growth 15% of normal, 50% survival of 

lateral buds, 10% of trees dead (may not be 

distinguishable from normal variations in 

canopy cover) 

25. 280 [] First effects 2 

30. 199 0 First effects 2 

45. 20 0 Trees shed needles produced during first 2 

(16- 23) flush 

80. 7 El First signs of color, trees began to turn 2 

reddish brown 

20. 390 U Pines that had received more than 2000 rads 3 

were markedly affected 

21. 90 U 0% survival of terminal buds, no terminal 4 

growth, 40 % of trees dead 

32 ± 6 142 U Trees alive but terminals dead, no lateral I 

development 

40. 270 U Dead 8 to 10 months following an exposure 3 

greater than 4000 rads 

42. 90 U 90% of trees are dead 4 

45. 390 U 13 months following irradiation, pines that 3 

had received more than 4000 to 5000 rads 

were dead 

75. 28 U Began to turn a brilliant orange-red and died 3 

within a few weeks 

96

Table 6-2. Detectability by multispectral remote sensing of radiation damage to pine trees for 

spring exposures as a function of time since start of exposure and dose. Exposure 

duration of two to four weeks. (Continued) 

Time since 

start of Estimate of 

Dose exposure multispectral 

(Gy) (days) detectabilirya Description of damnageb Referencec 

87. 90 U All trees dead at this dose and higher 4 

93. 142 U Trees dead, minimum lethal exposure (at this 

time) was 9261 R. 

120. 16 U Brilliant red-brown, coloring completed 2 

aDetectability with Landsat Thematic Mapper (judgment of authors): 

x no, unlikely. 

o maybe, uncertain, 

[ yes, very likely, 

b"Dead" means no green needles whatsoever; terminal and lateral refer, respectively, to the ends 

and sides of branches; terminal and lateral buds both contain apical meristem. 

cReferences: 

1. Monk (1966). 17 day exposure to 137 Cs y-radiation; starting 11 May 1965, about one 

month after pine shoot elongation had begun; five-year-old longleaf pine (Pinus palustris). 

2. Pedigo (1963): 14 day irregular exposure to output of Lockheed air-shielded reactor. n:7 

ranging from 1:1.7 to more than 3:1; dominant erposure during second week starting about 

June 14, 1959: loblolly pine (Pinus taeda) forest containing trees with trunks up to at least 

12" in diameter. 

3. Platt (1963): Report based on same experiment as Reference 2 (Pedigo, 1961). 

4. Miller (1968): 29 day exposure to 137 Cs y-radiation; starting April 6, 1966, "at the very 

beginning of the growing season;" 8-year-old longleaf pine (Pinus palustris). 

Figure 6-4 provides a graphical presentation of the detectability estimates from Table 6-2 with 

different symbols representing the three categories of detectability. Both dose and time since stadt 

of exposure should be considered as independent variables in Figure 6-4. The detectability data 

divides the dose/time plane into three regions corresponding to the three categories of detectability. 

97

000 -- f- -*t 

- i~rzz 

+-4 

TI 

----- j~ ~ 

00 

Foes Dee __ -r 

--- 

(98

Although the data is somewhat sparse, the general outlines of the three regions are apparent in 

Figure 6-4. 

Figure 6-4 includes the LD 10 0 line from Figure 6-3 for reference. Because dead in the 

literature is defined as having no green needles, it is clear that any patch of forest lying along or to 

the right of the LD 1 00 line will be detectable with the multispectral images of the Thematic Mapper. 

The other straight line in Figure 6-4 is hand-selected to approximate the boundary between the 

likely and uncertain estimates of detectability from Table 6-2. We assume that this line is a 

reasonable estimate of the relationship between dose an(.' the time of earliest detectable response. 

This choice is conservative on the high side. 

A controlled experiment might reveal that 

multispectral analysis will detect responses well into the uncertain region of the dose/time plane. It 

seems unlikely that the line of earliest detectability would be much closer to the LD 100 line. 

We assume for the present analysis that the relationship between dose and time of first 

detectable response is linear on a log-log plot, that is, the relationship follows a power law. The 

factor likely to cause the biggest deviation from the assumed linearity is the seasonal variation in 

detectability arising from the spectral contrast between new and old needles. Consider a patch of 

pine trees that has received a dose at the beginning of the growing season large enough to kill most 

buds but not enough to significantly affect old growth. According to data presented by Miller 

(1968) such a dose would lie in the range of 7 to 15 Gy for a 4 week exposure in April. During 

the growing season, the appearance of unexposed trees is dominated by new growth, which would 

be absent from the exposed trees. Consequently, in multispectral imagery, unexposed and 

exposed trees would differ according to the spectral contrast between new and old needles. This 

contrast would fade after the growing season with the onset of cold weather as new needles on the 

unexposed trees assume the spectral reflectivity of mature growth. The contrast presumably 

returns the following growing season as radiation damage again is manifested by inhibited growtl 

and mortality of some trees. Thus, in this intermediate dose range, the contrast between exposed 

and unexposed trees and, hence, the detectability of response, probably goes through a maximum 

toward the end of the first growing season and a minimum during the following winter. 

This potential fluctuation in the detectability of pine tree radiation response in the intermediate 

dose range is represented by the s-shaped, dotted curve in Figure 6-4. Detectability occurs at a 

dose minimum 3 to 5 months after the start of exposure (during August or September) and then 

moves back to higher doses 7 to 9 months after the start of exposure (December to March). 

.•',though this fluctuation in detectability may influence the interpretation of our results, we 

nevertheless approximate the earliest detectability with a straight line in Figure 6-4 since the 

detectability data is sparse and we do not have late summer satellite observations to interpret. 

The times of the 9 postaccident images relative to the day of the Chernobyl accident are 

indicated by the circles along the left edge of Figure 6-4. 

99

Figure 6-5 presents a regression analysis intended to quantitatively estimate the boundary 

between the likely and uncertain estimates of detectability from Table 6-2. The data for the 

regression is selected from Figure 6-4 by taking the left-most occurrences (lowest dose for a given 

time interval) of the likely category and all of the points in the uncertain category. The data at 90 

days has been excluded in both cases since we do not have an image of Chernobyl at that time and 

because of the probable temporary increase in detectability at the end of the growing season 

discussed above. Finally, the data at 142 days does not have a point in the uncertain category, so 

one has been interpolated halfway between the likely and unlikely points. Time of observation is 

taken as the independent variable and the dose as the dependent variable for the regression 

analysis. The regression line with 68% confidence band is plotted in Figure 6-5. The regression 

line is given by 

D = 198 t- 0 . 365 (6.1) 

where t is the time of first detectable response in days since start of exposure and D is the estimated 

dose in Gray. This relationship is our best estimate for detectability with remotely sensed 

multispectral imagery. 

It does not apply to close visual inspection of trees nor to growth 

measurement, both of which would reveal response at lower doses and earlier times. 

Table 6-3 provides evaluations of Equation 6.1 for the postaccident image dates analyzed in 

this report. Equation 6.1 is based on data ranging from only 1 week to 1 year. Values for times 

outside this range are extrapolations as noted in Table 6-3 and are subject to additional uncertainty. 

Table 6-3. Dose estimates according to Equation 6.1 for a first detected response corresponding to 

the times of the 9 postaccident images presented in this report. 

Time postaccident Estimated dose 

Image number (Days) (Gy) 

3 3 133. (extrapolation) 

4 12 80. 

5 28 59. 

6 35 54. 

7 172 30.2 

8 220 27.6 

9 380 22.6 

10 499 20.5 (extrapolation) 

11 763 17.6 (extrapolation) 

100

1000 __ 

-- -- 

2 Symbols as 

if 

ii -___ ___ 

nKgure 6 / 4- 

__________ - - t 

_____ I Jl 

100 

CD 

(/3 

0 

cz 

10 

100 Li 

first detection, d 

Figure 6-5. 

Regression line with 68% confidence band for the relationship between dose and 

time of earliest detection relative to ihe start of exposure for two- to four-week, 

springtime exposures. 

101

In conclusion, Equation 6.1 represents the relationship between total absorbed dose and the 

time of first detected response relative to the start of exposure. This relationship is used for the 

dose estimates presented in this report. It is based mostly on gamma dose. 

102

SECTION 7 

DOSE DETERMINATION FOR CHERNOBYL FORESTS 

This section maps estimated radiation doses to pine forest near the Chernobyl nuclear power 

station through an analysis of a time series of Landsat Thematic Mapper (TM) images for two years 

following the Chernobyl accidental nuclear reactor explosion. The first subsection describes a 

method for defining spectral deviation from normal for a pixel known to belong to a given class 

and presents such deviations based on the 11 Tasseled Cap images of Chernobyl and the 4 pine 

forest classes with reference (or control) sites for each class as given in Section 3. The second 

subsection discusses the progression of forest clearing during postaccident cleanup operations. 

The third subsection describes the calculation of time-to-response on a pixel-by-pixel basis for pine 

forest showing radiation response and discusses the correlation with radiation dose to foliage. The 

fourth subsection presents two alternative methods for generating dose contours from the satellite 

data. The final subsection discusses the resulting dose maps and summarizes results from this 

section. 

7.1 SPECTRAL DEVIATION FROM CLASS. 

Section 3 describes the procedures used to map four preaccident classes of pine forest based on 

pixel-by-pixel classification of spectral reflectivity. Each class of pine forest is represented by a 

reference site containing on the order of 100 contiguous pixels. Each pixel encompasses a 25 m by 

25 m square of forest and is represented by a vector of spectral intensity values transformed to the 

Tasseled Cap coordinate system. The vector components measure the average spectral features of 

all trees and other surfaces within the pixel that are visible from above. In addition, each pixel 

vector has noise components due mainly to atmospheric scattering of light into the line of sight of 

the pixel. 

In this subsection, we establish the bagis for detecting the average radiation response of the 

pine trees in a pixel by quantifying the deviation of each pixel vector from normal, defined as the 

mean vector of the reference (or control) site for the class to which the pixel belongs. The 

deviation for each pixel is expressed as a Mahalanobis vector. The deviations for the pine forest 

within a few kilometers of the Chernobyl nuclear reactor station are presented in a color format 

designed to show deviations in Tasseled Cap brightness, greenness, and wetness for each date 

analyzed. 

103

7.1.1 Mahalanobis Distance and the Mahalanobis Spectral Deviation Vector. 

Figure 7-1 illustrates a cluster of pixels from a class reference site using two axes of the 

Tasseled Cap coordinate system. Each pixel is represented by a point plotted according to the 

components of its spectral intensity vector x. For calculational purposes, we assume that each 

cluster of pixels may be reasonably approximated by a multivariate normal distribution. The 

resulting hyperelliptical distribution in general has its principle coordinate axes tilted with respect to 

the Tasseled Cap axes and has unequal variance along the principle coordinate axes. 

The spectral deviation of a pixel x from its class is expressed by the deviation vector d = x - pt, 

where p. is the mean vector of the class reference site. In order to have a consist meaning for the 

spectral deviation from class to class, the deviation vector must be expressed in standard units. 

The method that we use is explained in Appendix B. For each class of pine forest, a normalizing 

matrix N is used to convert the deviation vector d to a Mahalanobis vector m by the transformation 

m = Nd. (7.1) 

The matrix N is calculated from the covariance matrix of the class reference site as described in 

Appendix B. 

The normalization of each pixel deviation according to Equation 7. 1, transforms the set of 

deviations for the reference site to a distribution with unit variance in all directions. By definition, 

the deviations have a mean of zero. The magnitude of the Mahalanobis vector, called the 

Mahalanobis distance (Duda and Hart, 1973), measures the distance of a pixel from its class mean 

in standard units. It is the multivariate equivalent of the normal deviate z customarily used with a 

univariate standard normal distribution. 

The Mahalanobis distance is used to determine the significance of the deviation of a pixel from 

its class mean. To define significant deviation, a threshold is set such that any pixel with 

Mahalanobis distance greater than the threshold is judged to have moved significantly away from it 

class through the action of some factor not affecting the class in general. Since the Mahalanobis 

distance is expressed in standard units, we can use the same threshold for deviations with respect 

to all classes. 

7.1.2 Scaled Mahalanobis Distances and Vectors. 

The image processing software used for the present analysis stores only 8 bit intensity values, 

that is, integers ranging from 0 to 255. Calculations are done with floating point arithmetic, but 8 

bit numbers are used to store results in image format and for color display of images. For 

104 

0i

V) 

Q) CSt Reference Deviation 

Vector d 

Site 

Cluster 

Q- 

-o.... - .. .- 

S~PrincipOe 

0 Coordinate 

Axes of 

Cluster 

Tasseled Cap Greenness 

Figure 7-1. 

Two-dimensional illustration of the cluster of points formed by the pixel 

intensity vectors of a forest reference site. See Appendix B for the 

normalization procedure used to express deviations of pixels from the cluster 

center in standard units. 

105

convenience, a scaled Mahalanobis vector v and a scaled Mahalanobis distance s are defined 

according to 

v = 4m+128i (7.2) 

and 

s = 4 Iml, (7.3) 

where i is a vector with all components are equal to one. The dimensionality of the scaled 

Mahalanobis vector v is equal to the number of spectral bands being utilized in the analysis and can 

range from one up to the original number of bands in the multispectral image. The scale factor of 4 

in both equations scales the values such that in one dimension, a standard deviation becomes four 

units. In other words, the scaled Mahalanobis distance s is expressed in integer multiples of 0.25 

standard deviations. This resolution is sufficient since deviations of one or two standard 

deviations, expressed here as 4 or 8 units, are below the level of significance. A full scale 

deviation of 255 corresponds to 255/4 = 63.75 standard deviations, a dynamic range that is quite 

sufficient for present purposes. 

The scaled Mahalanobis vector v has an additive term of 128 on each of its components to shift 

the mean value from zero to 128, the center of the 8 bit scale. In this way, the cluster of scaled 

deviations for a reference site forms a spherical distribution at the center of the hypercube of 

possible deviation vectors with 8 bit components. The cluster has a standard deviation of 4 along 

each axis. Again, the only purpose of the scaling defined by Equations 7.2 and 7.3 is to 

accommodate the 8 bit data scale. 

The analysis of vector deviations in this report uses only three components from the Tasseled 

Cap transformed images, namely, brightness, greenness, and wetness. Figure 7-2 illustrates a 

scaled cluster for a class reference site using these three dimensions. The value of the scaled 

Mahalanobis distance s from Equation 7.3 corresponds to the radial displacement of a pixel from 

the center of the cluster in Figure 7-2. 

It is used to judge the significance of a deviation 

independently of the combination of feature changes causing the deviation. A threshold is chosen 

such that any deviation greater than the threshold is flagged as significant. Following the usual 

procedures of statistical analysis, the threshold must be set high enough that the random occurrence 

of deviations above threshold on the tail of the multivariate distribution for normal forest as 

illustrated in Figure 7-2 does not interfere with the detection of the spatial pattern of deviations 

caused by radiation exposure. 

106

0 50 

L0n 

1001 

o v fn 

o 

o• Ln, 

Figure 7-2. 

The cluster of scaled Mahalanobis vectors for a pine forest class reference site using 

the first three features of the Tasseled Cap spectral transformation. (Cluster 

enlarged for clarity.) 

107

The scaled vector deviation v in Equation 7.2 specifies the direction of the deviation in the 

three-dimensional space, that is, it tells how much each of the Tasseled Cap features deviates from 

normal. It may be used to distinguish deviations with specific spectral signatures. It particular, the 

Mahalanobis vector is used in this report to detect pine forest that was cleared as part of the 

decontamination effort after the Chernobyl accident. 

7.1.3. Spectral Deviations for the Chernobyl Images. 

We have calculated both vector and scalar spectral deviations according to Equations 7.1-3 and 

Appendix B for all pixels in the 38.4 km square analysis area presented in Section 3 belonging to 

the four pine forest classes defined in Figure 3-24. Figures 7-3 and 7-4 provide a color graphic 

presentation of the vector deviations for a 4.4 km by 4.8 km area near the reactor station. The 

upper left panel in Figure 7-3 shows the four pine forest classes color-coded as in Figure 3-23 and 

superimposed on a gray-scale background image of the area for spatial orientation. The other five 

panels of the same size show the vector deviations for the pine forest pixels on the first five inmage 

dates; for clarity, no background image is include. The five panels are numbered according to the 

image/date numbering scheme of Table 3-3. Figure 7-4 shows the vector deviations in the same 

area for the last six image dates. 

The color graphic presentation of vector deviations in Figures 7-3 and 7-4 is generated by 

displaying the components of the scaled Mahalanobis vector (Equation 6.2) for Tasseled Cap 

brightness, greenness, and wetness in the red, green, and blue channels, respectively, of the color 

images. The process can be visualized in the color space of the display images with the aid of 

Figure 7-2 by replacing each Tasseled Cap label on the axes with the corresponding red, green, or 

blue display intensity. The scale of gray levels (equal intensities of red, green, and blue) from 

black to white stretch along the diagonal of the color space from the origin (0,0,0) to 

(255,255,255). 

The cluster corresponding to a class reference site is centered at midscale (128) on each of the 

color channels. The resulting display is gray at the middle of the intensity scale, halfway between 

black and white. A patch of forest with negligible deviations from the class mean would appear as 

nearly uniform gray in Figure 7-3. Since the pixels of normal forest have random deviations from 

the class mean as illustrated :n Figure 7-2, normal forest will appear to be speckled with pastel 

colors lying near the midlevel gray point. 

The three reference bars along the right edge of Figures 7-3 and 7-4 show the appearance of 

normal distributions like that of Figure 7-2 with standard deviations of 4, 6, and 8 units along each 

axis as indicated by the number in each bar. Note that only the intensity of color fluctuations and 

not their spatial scale is different in the three reference bars. By definition, reference sites for the 

108

Figure 7-3. Color graphic presentation of the scaled Mahalanobis deviation vectors for pine forest 

on Dates I to 5 (194 by 177 pixel area). Upper left panel shows forest classes (see Figure 3-23). 

Three reference bars along right edge illustrate appearance of normal forest 

109/110

Figure 7-4. Cclor graphic presentation of the scaled Mahalanobis deviation vectors for pine forest 

on Dates 6 to 11 (194 by 177 pixel area). Three reference bars along right edge llustrate 

appearance of normal forest. 

III

pine forest classes will appear like the upper bar with a stmtidlrd d-iiaticn of 4 units along each 

color axis. The reference bars with standard deviations cf 6 , 8 are included since 

geographically separated patches of norrnal forest may havc aciditicnal random variance not present 

in the corresponding reference site. 

Examination of the Date 1 spectral deviattons as represented by the speckled gray patterns of 

the color graphic display in Figure 7-3 shows co signifikant diffeerence from the reference bars as 

expected since Date 1 was before the reactor accidtnt and is used to define the forest classes. Date 

2, the preaccident winter image, is also reasonably close to the reference bars as it should be. 

The strongest deviations in Date 3 of Figure 7-3 are the magenta patches caused by clouds 

which have high brightness and wetness with low greenness. The resulting mixture of red and 

blue with negligible green gives the magenta color. 

Figure 7-4 clearly shows the progression of increasingly significant disturbance of the pine 

forest as a result of the reactor accident. By Date 9, strong red patches show forest that has been 

cleared for operational purposes and for radioactive decontamination These areas are expanded by 

Date 10. Some of the ground cover on these cleared areas has changed by Date 11 as evidenced by 

the shift from red to magenta. The red line through the large forest patch in the lower third of this 

image area is probably the disturbance caused by the construction of a ground water barrier as 

discussed in Appendix C. It first appears in the images on Date 7 (October 1986) with an 

additional segment on Date 9 (May 1987). 

Not much forest clearing had been done by Date 7. There were no clouds on this date. The 

spectral deviations of the pine forest show a strong radiation response as the brown and orange 

swath across the middle of the left half of the image area. This swath corresponds to the westward 

trace of initial fallout deposition from the reactor explosion. Additionally, tbc. irregular patch of 

forest at the top of the image area on Date 7 shows a radiation response with a somewhat different 

spectral signature as evidenced by the magenta hue in Figure 7-4. This response of the irregular 

patch has diminished on the winter image of Date 8 but returns the following year on Dates 9 and 

10. 

According to the data and discussion in Section 6.4, this fluctuation in detcctability of the 

radiation response of the irregular patch suggests a short term total dose to tome canopy in the 10 to 

20 Gy range. The magenta hue on Date 7 indicates a reduction in greenness relative to normal 

pines, corresponding to a lack of new growth during the sumnmer after the accident. As the new 

growth on unexposed pines matures, the winter image of Date 8 shows almost no difference 

between the exposed patch and normal forest. The shift to an orange hue by Date 10 indicates that 

many pines have died by that time. 

112 

i m Bm *. =now

Although it may not be apparent because of the resolution and fidelity of color reproduction of 

this report, the forest along the westward trace of fallout shows a widening response on Dates 4 

through 6. This widening is demonstrated by the change detection analysis in Volume 2 of this 

report (McClellan et al., 1992). It is also apparent when the deviation images of Figures 7-3 and 

7-4 are displayed on a high resolution color monitor of an image processing system. Because of 

limitations of color reproduction, further analysis of radiation response in this report is based on 

numerical analysis of the scaled Mahalanobis deviation vectors and distances. 

7.2 DEVIATIONS CAUSED BY FOREST CLEARING. 

In the detection of radiation-induced spectral deviations, clearing of forest must not be 

interpreted as severe radiation response. Figure 7-5 shows a map of pine forest that was present 

postaccident through May 1986 but appears to have been cleared (bulldozed) by the date of the 

image number indicated in the figure. The figure includes. I km grid lines originating at the reactor 

(Unit 4). Clearing activity west of the reactor station seems to be mainly for decontamination. The 

earliest clearing is near the elbow in the main road about 1.5 km west and 0.5 km south of the 

reactor site where the road crosses the main trace of fallout deposition. Appendix C includes an 

eyewitness description by a man who walked along this road on the morning of the accident. He 

encountered a band of graphite flakes crossing the road and reported that the forest later "came 

under the ax." 

The large area in green in Figure 7-5 located 1.5 krn east and 2 km south of the reactor site was 

cleared between December 1986 and May 1987. This area shows no apparent radiation response 

before being cleared and so is presumed to have been cleared for operational reasons. The spectral 

signature of this area on Date 9 is used to define cleared forest on all dates. For this purpose, the 

scaled Mahalanobis vector is used as described in Section 7.1.2. 

A representative scaled 

Mahalanobis vector for cleared forest is calculated on Date 9 from a sample of pixels from the large 

cleared area. The vector is then normalized to provide a unit vector pointing in the direction of 

spectral deviation of cleared forest relative to normal forest. A forest pixel on any date is 

designated as cleared if the dot product of its scaled Mahalanobis deviation vector with this unit 

vector exceeds a threshold. The threshold is set high enough to eliminate spurious indications of 

clearing for isolated pixels and to avoid classifying radiation-damaged forest as being cleared. 

Table 7-1 lists the areas of newly cleared pine forest on the various dates in Figure 7-5. These 

areas are lower limits to the amount of vegetation actually cleared since Figure 7-5 includes only 

pixels belonging to one of the four classes of pine forest defined in Section 3. The listed areas 

include the pixels along the path cleared through the forest 3 km south of the reactor for the 

construction of a ground water barrier as described in Appendix C. 

113

Figure 7-5. Forest cleared (bulldozed) after the Chernobyl acident is color coded by the date 

number of the first observed clearing. Grid lines are spaced by I kn,, with an intersection at the 

Unit 4 reactor. 

114

Table 7-1. Area of pine forest near the Chernobyl nuclear reactor station cleared by the listed image 

date. Clearing of other vegetation not included. 

Image Area Cumulative Area 

Number Date Number of Pixels Cleared (ha) (ha) 

7 15 Oct 86 5 0.3 0.3 

8 2 Dec 87 0 0.0 0.3 

9 11 May 87 1115 69.7 70.0 

10 7 Sep 87 493 30.8 100.8 

11 28 May 88 410 25.6 126.4 

7.3 TIME-TO-RESPONSE FROM IMAGERY. 

The time-to-response for the foliage comprising a single pixel is derived from the onset of 

significant, persistent deviation of the spectral signature of the pixel from normal the spectral 

signature of the class to which the pixel belonged before the accident. The time-to-response for the 

foliage in the pixel is the time difference between this response and the time of the accident, that is, 

time-to-response is measured relative to the start of radiation exposure. 

Significant deviation from normal on a given date is judged by the scaled Mahalanobis distance 

of a pixel relative to its class mean as described in Section 7.1.2. A threshold distance is set for 

each date so that any pine forest pixel with scaled Mahalanobis distance exceeding threshold is 

flagged as having a significant deviation on that date. The procedure is equivalent to using a 

multivariate z score to judge the significance of an observed deviation. The Mahalanobis distance 

is calculated in 3 dimensions using the intensity values for Tasseled Cap brightness, greenness, 

and wetness. Thresholds of 20 to 24 are low enough to give good detection of the radiationinduced 

spatial patterns of change described in Volume 2 of this report and high enough to 

minimize the random occurrence of isolated pixels exceeding threshold at large distances from the 

reactor site. 

In addition, a neighborhood-dependent algorithm for threshold comparison is used to improve 

the signal-to-noise ratio for detecting radiation response. In this algorithm, the Mahalanobis 

distance for each pixel is replaced by the magnitude of the average Mahalanobis vector of the pixel 

and any of its eight neighboring pixels that are also pine forest. The vector average reduces the 

influence of slight registration errors from image to image as well as noise from other sources. To 

take advantage of this noise reduction, the threshold for each pixel is scaled by the inverse of the 

square root of the number of pixels being averaged. That is, when a pine pixel at the center of a 

115

3 by 3 square patch of pixels has N - I neighbors in the patch that are also pine forest, the pixel is 

flagged as having a significant deviation if 

N 

Xvi 

T 

> N 1/2 (7.4) 

N 

where vi is the scaled Mahalanobis distance of the i th pine forest pixel in the 3 by 3 patch and T is 

the threshold value for a single pixel. A form of Equation 7.4 that is more efficient for calculations 

is 

N 122(75 

vi > NT 2 (7.5) 

i= I 

Figure 7-6 illustrates the algorithm for the determination of time-to-response given that 

significant deviations have been flagged in each image and lists the threshold T used for each 

image. The algorithm is based on finding the first significant deviation that persists for the same 

pixel on all subsequent image dates. The deviation must precede any indication that the forest in 

the pixel has been cleared. Any pixel that does not contain pine forest, or is cleared of forest before 

any indication of radiation response, yields no data regarding radiation exposure. Such pixels are 

assigned a code of zero and will become black background in response maps presented as images 

in this report. Any pixel that appears normal (does not exceed threshold) on Date 11, the last 

image of the series, has no persistent response by definition and is coded as normal forest with no 

radiation response. Such forest pixels yield a null response and are coded to appear gray in 

response or dose images. These areas show where fallout contamination was too low to produce 

lasting visible damage to the foliage of pine trees. A pixel that shows a first persistent deviation 

from normal on Date n is coded with the value n to indicate the date of first observed response. Of 

course, the first "observable" response will have actually occurred sometime between this date and 

the previous image date. 

The requirement of a persistent response over two years has the disadvantage of eliminating 

responses to low doses where the trees recover normal appearance within two years. On the other 

hand, the requirement has two major advantages. The first, and of most practical importance, is 

the elimination of spurious responses caused by clouds, jet contrails, and local haze patterns 

occurring on a single date. The second advantage is further reduction of noise signals that are 

uncorrelated from date to date. These advantages derive from the fact that several of the images in 

116

Start Processing 

Thresholds for significant deviation 

Image: 1 2 3 45 67 8 91011 

T: 40 55 40 40 60 40 40 40 40 40 45 

(Scaled Mahalanobis Distance) 

belong to a pine 

forest class? 

Yes 

NDoes the pixel show 

•"Yes 

a persistent 

deviation from 

normal? N 

No Data 

(Coded black) 

Did an indication of 

forest clearing occur 

persistent as early deviation as the 

Ye from normal? No 

Null 

Response, 

Forest normal 

(Coded gray) 

No Data 

(Coded black) 

Response Time 

Detected 

(Color-coded by 

date) 

Figure 7-6. Algorithm for determination of time-to-response for radiation-damaged foliage on a 

pixel-by-pixel basis. Parenthetical remarks refer to presentations in response time 

figures. 

117

the series are cloud free and none of the images have extensive cloud cover, so the probability of a 

persistent response at a given location due to a combination of clouds and random noise is nil. 

Since the last date (Date 11) has no subsequent image for establishing persistence, the threshold 

used for Date 11 is set higher than that of the other images to reduce spurious responses on that 

date. 

Figure 7-7 presents results for the date of first observed radiation response according to the 

algorithm of Figure 7-6. The date number of response for each pixel is color-coded according to 

the legend in the image. Table 6-3 lists the date numbers and corresponding estimated foliage 

doses according to the analysis of published data presented in Section 6. No persistent responses 

exceed threshold on the preaccident Dates 1 and 2 or on postaccident Date 3. Only two pixels in 

the central blue area of Figure 7-7 have responses on Date 4. The main early response in 

Figure 7-7 is for Dates 5 and 6, four to five weeks after the accident date. Since the responses 

could have started immediately after Date 4 at 12 days, the dose for the blue area is estimated to be 

between 54 and 80 Gy according to Table 6-3. Surrounding the blue area and extending further to 

the west is a dark green area showing first observed response on Date 7 (October 1986). The 

estimated dose for this area is 30 to 54 Gy. The yellow-green and yellow areas have first observed 

response between Dates 8 and 10 with estimated doses in the 20 to 30 Gy range. 

Note that the irregular patch of forest in the upper middle of Figure 7-7 shows only a few 

pixels with persistent responses, beginning on Dates 9 and 10 with estimated dose in the 20 to 28 

Gy range. The irregular patch consists mostly of gray pixels, indicating a dose less than about 18 

Gy. As pointed out in Section 7.1.3, this area shows a transient response maximizing on Date 7 

(October 1986) and disappearing over the winter. As discussed in conjunction with Figure 6-4, 

this transient response is likely to be associated with doses in the range of 7 to 15 Gy. The 

response of this irregular patch shows that the multispectral detection technique for pine tree 

radiation response could be extended to doses below 20 Gy if the analysis accounted for transient 

as well as persistent responses. 

A small, vertically oriented, rectangular group of about 20 pixels just to the right of the 

irregular patch shows a solid response on Date 7 for an estimated dose of 30 to 54 Gy. Figure 7-5 

shows that this same group of pixels was cleared between Dates 9 and 10. The group is 1.2 km 

west and 1.3 km north of the reactor site. Figure 7-5 shows that the irregular patch of forest 

discussed above is located 2 km west and 1.5 km north of the reactor site and had not been cleared 

as of Date 11 (May 1988). The patch is at the eastern edge of Pripyat. If these trees are still 

standing, they would provide a good retrospective sample on which to perform histological studies 

to correlate with the responses and estimated doses derived from satellite images. 

The path cut through the largest patch of gray, normal forest in the lower part of Figure 7-7 is 

associated with a ground water barrier several kilometers long and 30 - 35 m deep constructed to 

118

Figure 7-7. Date of onset of persistent deviation from normal of pine forest pixels according to 

algorithm of Figure 7-6; 6.4 km square area. See Table 6-3 for corresponding dose estimates. No 

responses occur for the preaccident Dates I and 2 or for Date 3 three days postaccident. 

119

impede the migration of radionuclides in the !round. The presence of the ground wdter barrier 

apparently affects the pine trees adjacent to the path, especially on the southwest side. The trees 

show significant deviations from normal extend:-rg further from the barrier as time progresses 

between 1 and 2 years after the accident. These indications of stress are probably not due to 

radiation damage to the foliage. They may indicate a ground water effect on the health of the root 

,s stems of the trees not involving any radiation damage. 

7.4 FOLIAGE DOSE MAPS. 

Two procedures have been used to construct dose contours from the data on time to first 

response for pine trees shown in Figure 7-7. The first method uses shape of gamma dose rate 

contours measured aerographically as guidance for the hand-drawn contours wherever satel iite 

image data is sparse or nonexistent. The second method uses a numerical relaxation techniqte, to 

generate smoothed contours from the satellite image data alone in an area with sufficient data for 

analysis. 

7.4.1 Dose Map with Hand-Drawn Contours. 

Figure 7-8 shows the time-to-response data from the satellite image analysis with three handdrawn 

radiation dose contours. Table 7-2 lists the total area enclosed by each contour and the dose 

range of the area inside each contour. 

Table 7-2. Description of contours drawn in Figure 7-8. 

Time of observed Dose within Cumulative 

response contour enclosed area 

Contour (image number) (Gy) (ha) (ki2) 

Inner (blue) 5 and 6 54 to 80 a 82. 0.8 

.liddle (green) 7 30 to 54 287. 2.9 

:ter (yellow) 8, 9 and 10 20 to 30 1450. 14.5 

'r limit set by the sparsity of response on Date 4. 

tours in Figure 7-8 are hand-drawn starting in the westward trace of highest fallout 

'here there is sufficient data from pine forest response to draw the contours as 

;tween areas with different dates of first observed response. The intent is to draw 

.ooth contours along the average boundary between different dates of response. 

-)urse, scattered pixels with response dates that do not match the contours. These 

120

Figure 7-8. Time to first observed response with 1 km grid and three hand-drawn dose contours. 

See Table 7-2 for dose ranges. 

121

variations presumably are caused by nonuniformity in fallout deposition and by noise in the 

detection process. The contour segments drawn where there is image data are extended around the 

reactor where there is no image data following the shape of published gamma dose rate contours 

measured a year after the reactor explosion (Asmolov et al., 1987). These Soviet-supplied 

contours are described in Volume 2 of this report (McClellan, 1992). Note the contours in Volume 

2 and the outer contour in Figure 7-8 have been extended across the reactor cooling pond to 

approximate the situation in the absence of the body of water. 

Since the contours in Figure 7-8 are derived from observations of pine tree response, they 

represent an estimate of the dose at any location that would be received by a patch of pine forest at 

that location. The relationship of this dose to the dose that would be accumulated by some other 

detector can be estimated from the results of the radiation transport calculations of Appendix A. In 

particular, Section 8 discusses the relation between the pine tree dose and the gamma dose 1 m 

above the ground under the canopy. 

The extrapolation of the contours around the reactor in Figure 7-8, based on gamma dose rate 

contours, is uncertain because of the difference between in situ pine forest response and that of the 

gamma detectors one year after the accident. As discussed in Section 5, the pine forest canopy 

responds mainly to the dose accumulated during the first few weeks after the reactor explosion 

with a beta contribution at least as large as the gamma contribution. On the other hand, the curves 

of Asmolov et al. (1987) show the gamma activity a year after the explosion affected by weathering 

effects and cleanup activities. Because of this difference, it is of interest to consider a method of 

generating contours from the pine tree response alone. 

7.4.2 Iterative Smoothing of the Dose Map. 

Figure 7-9 illustrates an iterative numerical relaxation method for smoothing of the dose map. 

About 3 km 2 is cut from the area of highest radiation response in Figure 7-8. The upper left panel 

in Figure 7-9 shows the date of first observed response for this area following the color code of 

Figure 7-8. Gray pixels indicating forest that is normal on Date 11 are assigned date value 12. 

The lower right panel shows the result of the relaxation method with date numbers of first 

observed response used as the numerical value assigned to each pixel. 

On each iteration, the response date number for each pixel not containing pine forest is replaced 

by the average date number (taken as a continuous variable) of its four nearest neighbor pixels. 

Values for pine forest pixels in the initial data set are held fixed at their initial values to provide 

boundary conditions for the relaxation solution. Relaxation of values at the border of the cut out 

area use a neighborhood average excluding pixels outside the border. This choice lets the value at 

122

Figure 7-9. Series of intermediate stages leading to a smoothed map of date of first observed 

response by numerical relaxation as described in the text. 

123

the border float according to the influence of the nearest pine forest pixels in the initial data set 

without requiring a definite value at the border. Pixels with no value (displayed as black) in the 

initial data set are initialized with date value zero. Iteration of the map until the value in every pixel 

changes by less than 0.0001 from the previous iteration gives the lower right panel in Figure 7-9. 

Stopping the iteration when all changes are below 0.5 and other intermediate values indicated in the 

figure gives the full series shown in Figure 7-9. For practical purposes, stopping the iteration 

when changes are less than 0.001 is good enough. 

This numerical relaxation technique converges to a solution of Laplace's equation, 

V 2 n = 0 (7.6) 

where n is the date of first observed response treated as a continuous variable. In electrostatics, 

Laplace's equation describes the spatial variation of the electric potential in a region of uniform 

electrical conductivity. The pixels in the initial data set used as fixed boundary conditions 

correspond to areas held at a fixed voltage by an external source of electrical potential. There is no 

physical reason why the radioactivity level in a wind-driven fallout field should satisfy Laplace's 

equation. The numerical relaxation technique is simply a convenient tool for smoothing the dose 

map. 

Figure 7-9 shows substantial small scale variation caused by isolated pixels in the initial data 

set. Furthermore, Figure 7-9 gives a map of date of first response rather than dose. Figure 7-10 

addresses both of these issues. Based on the success of the relaxation solution in Figure 7-9, 

Figure 7-10 uses a larger area of data. To reduce small scale variations, isolated pixels in the initial 

data set are ignored as boundary conditions and allowed to vary in the relaxation calculation. Next, 

before relaxation, the date numbers in the initial data set are replaced by radiation doses according 

to Table 6-3. Iteration proceeds until fractional changes in all dose values drop below 0.001, then 

the pixels are assigned to dose bands and colored according to Table 7-3. 

Figure 7-11 displays the smoothed contours from Figure 7-10 with a gray-scale image of the 

area. Both Figures 7-10 and 7-11 include one kilometer grid lines originating from the site of the 

Unit 4 reactor. The site of the reactor is chosen as the brightest pixel from the reactor fire visible 

on 26 April 1986 (Date 3), three days after the explosion. The Zone 36 UTM coordinates of the 

upper left hand comer of this pixel are X = 298,275 m and Y = 5,697,175 m as determined by the 

geocoding of our Landsat image data (see Volume 2). 

124

Figure 7-10. Smoothed dose contours from pine foliage response using numerical relaxation of 

dose values from individual pine forest pixels inside the white border; I km grid lines. 

See Table 7-3 for estimated dose bands. 

125/126

Figure 7-11. Smoothed dose contours from Figure 7-10 displayed with a gray-scale background 

of the western end of the Chernobyl nuclear power station; I km grid lines. See Table 7-3 for 

estimated dose bands. 

127

Table 7-3. Dose bands and affected areas for the smoothed radiation contour maps of Figures 7-10 

and 7-11. Areas are within the relaxation area (white boundary of Figure 7-10) only. 

Dose band Band area Cumulative area 

Color (Gy) (ha) (ha) 

Red >94 0.1 0.1 

Yellow 54 to 94 9.1 9.2 

Green 35 to 54 16.8 26.0 

Cyan 23 to 35 86.1 112.1 

Blue 13 to 23 353.8 465.9 

The smoothed contours produced by the relaxation method appear too irregular and sometimes 

discontinuous in the central area. For example, the yellow areas in Figure 7-11 with a dose range 

from 54 to 94 Gy originate mainly from the observed responses on Dates 5 and 6. The appearance 

of this data in Figure 7-8 (within the inner contour), suggests that all of the area between the 

yellow patches in Figure 7-11 probably received at least as much dose as the yellow patches 

themselves. However, because of the wide spacing of the three main areas of response on Dates 5 

and 6, the relaxation method allows the influence of more distant pixels to lower the interpolated 

dose between the yellow patches. Although hot spots do occur in fallout fields, the spacing of the 

yellow patches in Figures 7-10 and 7-11 would not be expected in this situation and is clearly an 

artifact of the spacing of the patches of pine forest in the initial data set. 

Except for the exaggerated irregularities induced by the patchiness of pine forest in the initial 

data set, the smoothed contours of Figure 7-11 seem reasonable. For example, the western outline 

of the outer most dose band (blue, 14 -26 Gy) in Figure 7-11 has the same general shape as the 

outer most, hand-drawn contour (yellow, 20 - 30 Gy) in Figure 7-8. 

The relaxation method has the advantage of providing smoothed radiation dose contours 

without reference to other data and free of human bias. It has the disadvantage that it does not 

account for the physical transport properties of wind blown fallout patterns and tends to generate 

dose contours having the same spottiness as the pine forest spatial distribution. With further 

study, it might be possible to modify the relaxation method to better approximate true fallout 

patterns. 

128

7.5 SUMMARY. 

Both methods presented in Section 7.4 for generating maps with dose contours or bands from 

the pixel-by-pixel pine tree response data have advantages and disadvantages. Both provide a 

qualitative description of the likely initial fallout deposition pattern, but both must be used with care 

if quantitative extrapolations of dose outside the existing areas of pine forest are required. The most 

reliable results are the actual pixel-by-pixel dose estimates obtained through interpretation of the 

date of first observed response (Figure 7-7 ur 7-8) with the dose and time-to-response relationship 

from Figure 6-5 or Table 6-3, keeping in mind that response observed on a certain date means that 

the actual first observable response may have occurred any time in the interval bracketed by the 

date of first observed response and the previous image date. The dose estimate for the pixel lies in 

the interval defined by the doses for the two dates from Table 6-3. Accordingly, the dose bands in 

Table 7-3 were obtained using the geometric mean of the dose for the first observed response and 

the dose for the preceding date. 

The following list summarizes findings from this section: 

1) Pixel spectral deviations for pine forest canopy relative to the average spectral signature of a 

pine forest class may be used on a pixel-by-pixel basis to detect foliage radiation response 

in multispectral imagery (Figures 7-3 and 7-4). 

2) Time to first observed response for persistent spectral deviations over an annual cycle may 

be used to map pine canopy doses (Figure 7-7) near and above the LD 50 , about 23 Gy for 

a three week exposure. 

3) Mapping of dose responses significantly below the LD 5 0 requires interpretation of 

temporal variation in detectability of spectral deviations caused by the annual growth cycle. 

It is likely that transient responses may be detected down to 10 or even 5 Gy. 

4) Dose estimates at distances of 1 to 2 km downwind of the reactor site and from 30 to 54 Gy 

at 2 to 3 km downwind (Figure 7-8) from the Landsat multispectral imagery for pine forest 

canopy along the main westward trace of initial fallout deposition from the Chernobyl 

reactor explosion range from 54 Gy to 80 Gy. 

These doses estimates are based on the 

equivalent gamma dose that produces the same foliage spectral response as the actual mixed 

beta and gamma doses received by the foliage. 

129/130

SECTION 8 

DISCUSSION 

This section discusses the pine forest foliage doses derived from the satellite imagery in relation 

to data from the former Soviet Union regarding three types of measurements: 1) doses and 

responses of trees near the Chernobyl nuclear power plant, 2) aerographic and ground surveys of 

gamma dose rates following the accident, and 3) exposures of accident victims. A final subsection 

discusses uncertainties in the derivation of foliage doses from the satellite imagery. 

8.1 COMPARISON WITH MEASUREMENTS OF FOREST DAMAGE. 

The only quantitative exposure and response data we have for trees within a few kilometers of 

the Chernobyl nuclear power station has been obtained by Gamache (1993) through visits with 

Ukrainian scientists (Sobdovych et al. 1992). Table 8-1 lists this data, which refers to forest with 

edge about 1 km from the Unit 4 reactor site. This distance corresponds to the boundary of the 

nuclear power station on the west side. We are fortunate to have this information although some 

uncertainty remains regarding the data. For example, it is tempting to assume that the subject 

forest is west of the reactor site along the main trace of fallout deposition, but there are also forest 

patches northwest, southwest, and south of the reactor site that are not much more than 1 km 

away. The forest reportedly consisted of pine trees but Table 8-1 implies an LD 50 of 49 Gy, above 

the upper bound in Table 6-1 for pine trees suffering a few week exposure. Figure 6-2 indicates 

that an effective exposure time of about four months would be required to increase the LD 5 0 for 

pines to 49 Gy; however, our image analysis shows significant response along the main trace after 

only one month. 

Table 8-1. Results of forest damage by radiation for an area with edge about 1 km from the site of 

the Unit 4 reactor explosion (Sobdovych et al., 1992, courtesy of Gamache, 1993). 

Distance from 

edge offorest 

Calculated 

absorbed dose 

Tree crown 

damage 

Recovery 

of trees 

(M) (Gy) (%) Degree of harm (%) 

0 100 100 Completely dry 0 

(Edge of forest 

wood 

nearest reactor) 

35 65 50 Very strong damage 10- 15 

90 49 20-30 Medium 50 

350 5

In addition, Table 8-1 lacks an indication of the time at which the listed endpoints were 

observed. Presumably, the percent of crown damage must have been observed well before the 

percent recovery of trees since 50% lethality of trees would be inconsistent with only 20 - 30 % 

crown damage if both endpoints were observed at the same time. Also, the nature of the calculated 

absorbed dose is not known to us. In particular, does it include dose from beta radiation and to 

what tissue and depth does it correspond? 

The precipitous drop reported for the dose 350 m from the edge of the forest also needs 

explanation. Given that the plume from the initial explosion and fire rose more than a kilometer 

(Appendix C) and that the forest location is only 1 to 1.5 km from the reactor site, it would be 

surprising to have such a sharp gradient in dose within 250 to 350 meters along the main trace of 

initial deposition. Such a drop would be more likely, however, in moving off the trace laterally. 

Another possibility is that the reported forest is not on the main trace but was contaminated later in 

the 10 day release sequence by a plume that remained near the ground and was rapidly absorbed as 

it moved through the forest canopy. In the absence of accurate position information, the data in 

Table 8-1 cannot be directly compared with the satellite data. 

In spite of the unknown factors, it is encouraging that the doses listed in Table 8-1 are of the 

same order of magnitude as the doses from the analysis of the satellite image data listed in Table 7- 

2 and 7-3. The attempt at comparison strongly emphasizes the need for well documented ground 

measurements at specific times and with locational accuracy and resolution comparable to the 25 m 

spatial resolution of the imagery. 

8.2 COMPARISON WITH AEROGRAPHIC SURVEYS OF DOSE RATE. 

Following the Chernobyl accident, frequent aerographic and surface measurements were made 

by the Soviets to determine and monitor the gamma field dose rates surrounding the nuclear reactor 

station. These measurements provide spatial contours and decay rates of the gamma dose rate 

within the area of our satellite observations. Utilizing the Soviet data, this subsection calculates 

integrated gamma ray doses for the three week period following the accident and compares these to 

the pine forest doses extracted from the satellite image data. The ratio of the pine foliage dose to 

the gamma dose one meter off the ground is compared to the same ratio derived from the 

calculations of Appendix A. 

8.2.1 1-Meter Gamma Dose Rates for the Close-in Area. 

Figure 8-1 shows plots of gamma dose rate versus contaminated area based on measurements 

referenced to two dates, May 29, 1986 and May 1, 1987. The area on the abscissa is that enclosed 

by a given isodose-rate contour. Accordingly, dose rates within a given area are larger than that at 

the contour bounding it. The Soviet dose rate data are given only for close-in areas greater than 

132

about 2 km 2 . Therefore, estimates down to 0.8 km 2 are based on extrapolation indicated by the 

dashed part of the curves in Figure 8-1. 

104 

103 

E6 

May 29,1986 

D102% 

(D 

0 

May 1, 1987 

10 

1 10 102 103 

Contaminated fallout area (kin 2 ) 

Figure 8-1. 

Dose rate measurements for close-in fallout contaminated areas for two different 

dates: May 29, 1986 (Izrael, Petrov and Severov, 1987) and May 1, 1987 

(Asmolov et al., 1987). 

133

Table 8-2 provides, in the second and fourth columns, values for selected contour dose rates 

read from Figure 8-1 for the two reference dates. 

Table 8-2. Estimation of initial dose rates for contours enclosing specified areas. 

Measurements of: May 29, 19 8 6 a May 1, 19 8 7 b 

Contaminated area Contour Dose ratec Contour Dose ratec 

within contour dose rate on 4126/86 dose rate on 4126/86 

(kin 2 ) (mR/h) (Rid) (mR/h) (Rid) 

14.5 450 68 

4.0 1150 173 38 158 

2.9 1450 218 

1.0 2500 376 100 416 

0.8 3000 450 

aMeasurements from Izrael, Petrov and Severov (1987). 

bMeaurements from Asmolov et al. (1987). 

CExtrapolated back to date of accident according to Figure 8-2. 

In order to obtain doses accumulated over the first three weeks after the accident, it is necessary 

know the time dependence of the dose rate. Figure 8-2 gives a plot of Soviet data for the change 

with time of the gamma dose rate in the close-in zone due to the decay and weathering of 

radioactive material. We have approximated the data with a smooth curve by fitting it with an 

empirical relationship, 

F (t) = exp{- [in (1 + t)] 2 / a), (8.1) 

where, a = 6.7784 and postaccident time t is measured in days. Figure 8-2 plots the gamma dose 

rate in relative units normalized to the day of the accident (t = 0 on April 26, 1986); note that the 

abscissa is (1 + t). The dose rate values on the day of the accident are estimated with Equation 8.1 

and shown in columns 3 and 5 of Table 8-2. The cumulative gamma dose for a postaccident 

exposure time t is determined for any initial dose rate R 0 by integrating Equation 8.1 over time, 

i.e., 

D (t) = Ro F (t')dt'. (8.2) 

134

2 

0.1 

a: 

0.01 

1 10 100 500 

Time (days) 

Figure 8-2. Change in gamma dose rate from radioactive materials in the close-in zone as a 

function of time based on aerographic surveys (data from Asmolov et al., 1987). 

Figure 8-3 compares the 1-m gamma dose accumulated at the 1 km 2 and 4 km 2 contours as a 

function of postaccident time. The a and b curves reflect the Soviet measurements reported on 

May 1, 1987 and May 29, 1986, respectively. The middle curve, labeled 1 - 4 km 2 , of Figure 8-3 

plots the geometric mean dose for the 1 and 4 km 2 curves. Both the a and b curve pairs give 

essentially the same geometric mean value for the area between the 1 and 4 km 2 contours. The a 

and b curves are in good agreement considering the 11-month time difference of the data. We will 

use the earlier set for comparison with the satellite data since it is taken closer to the time of 

important exposure for the pine forest. 

8.2.2 Close-in Foliage and 1-Meter Gamma Doses. 

The analysis in Section 5 indicates an effective exposure time of about 3 weeks for the pine 

foliage around Chernobyl. This duration is based on the assumption that fallout initially retained in 

135

the forest canopy makes an important dose contribution, especially the beta component. The 

effective duration of exposure is determined by the decay of radionuclides and the weathering of 

fallout particles from the foliage to the ground. Accordingly, radiobotanical data for exposure 

times of two to four weeks is used to derive radiation doses from the satellite imagery. With this 

point of view, the pine forest doses extracted from the imagery represent total doses received in 

about the first 21 days after the initial deposition. 

104 a 2 

10b 1 km 2 

1 -4 km 2 

4 km 2 

0 

102 

1 10 102 103 

Post-accident time (day) 

Figure 8-3. 

Accumulated close in fallout dose (1-m above ground) in 1 and 4 km 2 areas; a and b 

based on gamma field measurements (USSR, 1987) on 5/1/87 and 5/29/86 

respectively; the middle curve is the geometric mean between 1 and 4 km 2 . 

Table 8-3 lists the pine forest foliage doses represented by the hand-drawn contours of Figure 

7-8. For comparison, Table 8-3 lists the accumulated three week gamma dose one meter off the 

ground according to the survey data of May 29, 1986 (Izrael et al. 1987). These gamma doses are 

calculated with Equation 8.2 and the initial dose rates listed in Table 8-2. For each of the three 

136

contours, the foliage dose exceeds the 1-m gamma dose. The ratios, also listed in Table 8-3, 

decrease from 2.9 at the lowest dose to 1.2 at the highest dose. 

Table 8-3. The ratio of the pine canopy foliage dose to the gamma dose 1 m off the ground 

using satellite image and aerographic survey data. 

Foliage dose at 1-m r-dose (21 days) 

Dose rato 

Contour area contour (Table 7-2) at contour 

(km 2 ) (Gy) (Gy) (foliage/I-m )I 

14.5 20. 6.9 2.9 

2.9 30. 22. 1.4 

0.8 54. 38. 1.2 

8.2.3 Calculated Ratio of Foliage to 1-Meter Gamma Dose. 

Section 5.3.2 presents calculations of the expected beta to gamma dose ratio for cylindr;.cal pine 

foliage elements of various radii using the radiation transport calculations of Appendix A. The 

same assumptions and transport calculations provide an estimate of the ratio of the accumulated 

dose to foliage from beta and gamma rays to the accumulated 1-m gamma dose either under the 

canopy or in an open field. Figure 8-4 shows the resulting dose ratios as a function of 

accumulation time for foliage elements in the upper canopy. The physical parameters assumed for 

the canopy are given in Appendix A. We assume that a calculation of the gamma dose in an open 

field is the best analog for the Soviet survey data. 

The dose ratios in Figure 8-4 are based on an assumed initial foliar intercept fraction of 0.60 

and a weathering rate of 0.0495 per day as discussed in Section 5.5. Fallout decay rates are 

neglected. The measured ratios listed in Table 8-3 may be compared with the calculated ratios at 21 

days postaccident from Figure 8-4b. The comparison shows that the two higher dose contours 

have measured dose ratios corresponding to foliage elements of about 0.4 cm in radius. The ratio 

for the lower dose contour corresponds to a foliage element of 0.10 cm radius. 

The measured ratios are similar to the calculated ratios, although the indicated radii of foliage 

elements are somewhat larger than expected from the morphological discussion in Section 5. The 

apical meristematic tissue is estimated to be at a depth ranging from 0.01 to 0.12 cm in depth with a 

typical value of 0.04 cm. It is possible that other sensitive tissues at greater depths may contribute 

to the radiation response. For example, the new growth extension behind the apical meristem has a 

radius of about 0.15 cm and may be relatively sensitive to exposure early in the growing season. 

137

10 " 1I 

o 9 a) 1-my dose 

under canopy 

8 

I 70.047 - cm = foliage element radius 

)D"-0.10 

C 4 

0.0625 

S6 

b4 0. 15 

2 0.40 - - SZ4 

' - .. . .. . 

0 

O 1 I 

I I . . . I i , I 

0 10 20 30 40 50 60 

7 

M i•b) 1-mydose 

S6 04cmin open field 

0.04 cm 

5 

4 

-4 

" 0.0625 

~4 '-o.io 

.0.15 

blj -l 

--- 

0 

44 0.40 - 

40O. 

0 I I I I I I , 

0 10 20 30 40 50 60 

Time postaccident, d 

Figure 8-4. Ratio versus time of the accumulated dose (P and y) at the center of cylindrical foliage 

elements at the top of pine forest canopy to the accumulated 1 -m gamma dose a) 

under the canopy and b) in an open field for the same total fallout deposition. 

138

It is noteworthy that the trend of the data in Table 8-3 is toward a higher ratio and an implied 

smaller depth of sensitive tissue for lower doses. This trend supports the conclusion in Section 5 

that responses of pine trees to lower doses of radiation are likely to be more sensitive to beta 

exposure of apical meristems than are the responses at higher doses, which may be dominated by 

the systemic effects of more penetrating gamma rays. 

8.3 COMPARISON WITH HUMAN EXPOSURE DATA. 

Skin lesions from P-irradiation were an integral feature of the acute radiation syndrome 

suffered by the victims of the accident at the Chernobyl nuclear power plant (Barabanova and 

Osanov, 1990). About half of 115 patients who were exposed in and around the plant had 

radiation-induced lesions in addition to injury to the hematopoietic system. This significant 

contribution of .B-dose to human injury in a fallout field parallels the importance of the 0-dose to 

vegetation in a fallout field as emphasized in Section 5. 

Table 8-4 lists typical 5/y dose ratios for four groups of accident victims at the Chernobyl. The 

patients are grouped according to mode of exposure as described in the footnotes to Table 8-4. 

Barabanova and Osanov (1990) describe the variations in occurrence and severity of skin lesions 

within these groups. The two depths at which dose ratios are presented in Table 8-4 correspond to 

0.007 cm and 0.15 cm of unit density material. The smaller depth is essentially at the skin surface 

with little attenuation of beta dose. At 0.15 cm depth, however, beta dose is substantially 

attenuated as evidenced by the third column of Table 8-4. Even so, the P/t dose ratio at 0.15 cm 

depth is still well above 1, especially when contact sources contribute or when only a few feet of 

air shields the skin from the beta source. 

According to calculations presented in Section 5, the 0/y dose ratio at the center of a cylindrical 

element of pine foliage with radius 0.15 cm varies from about 5 to about I as the ground source 

fraction increases from 40% to 90% and the contact source fraction decreases correspondingly. 

This range agrees qualitatively with that in Table 8-4 at 150 mg/cm 2 for the more nearly planar 

geometry of human skin. Quantitative comparison would require accounting for any differences in 

assumed radionuclide mix and source distribution and the difference between cylindrical and planar 

geometry. 

Cylindrical geometry increases the 0/1y ratio relative to the same depth in planar 

geometry. 

The last column of Table 8-4 lists the range of gamma doses for the individuals in each patient 

group. Generally, these doses cannot be compared to the foliage doses derived from the satellite 

imagery because the victims were located in or very close to the Unit 4 reactor building rather than 

near the forests. However, the most severe case in Group II was an individual located 1.0 km 

downwind from the reactor site for about 1 hour following the reactor explosion. As reported by 

Barabanova and Osanov, his y dose was 12.7 Gy and his 0 dose was about 30 Gy at a depth of 

139

150 mg/cm 2 . These doses should be somewhat comparable to those received by any nearby trees 

over the same exposure period. 

Table 8-4. Summary of typical P/y dose ratios for four groups of people who suffered 

radiation lesions in the skin at Chernobyl (Barabanova and Osanov, 1990). 

13/r dose ratio #-dose fl'ydose ratio Range of 7 doses 

Groupa (at 7 mg/cm 2 ) attenuationb (at 150 mg/cm 2 ) (GY) 

1 3 head(face) 3 1 2- 5.8 

5 shoulder/chest 2 

20 feet 7 

11 20-30 10 2-3 4-12.7 

III 20 4 5 9-14 

IV (> 13 )c 3 (>4) 2-11.5 

aGrouped according to characteristic patterns of irradiation: 

Group I. 

Group II. 

Distant f-exposure (high energy); 15 patients exposed in and around the plant 

commencing 3 to 5 hours after explosion, little contact dose. 

Deposition of thin source (relatively low energy); 6 patients exposed downwind by 

plume and by contact with fallout immediately after the explosion. 

Group 111. Exposure in cloud (fireman); 6 patients exposed on roof of Unit 4 for 30 to 40 

minutes, clothing protected somewhat from 1-rays. 

Group IV. Deposition of thick source (various energies); 29 patients who were plant operators 

working in Unit 4 at the time of the explosion, wet clothing impregnated with 

radionuclides, various exposure modes. 

bAttenuation factor from 7 mg/cm 2 to 150 mg/cm 2 skin tissue depth. 

clnferred from Table 1 of Barabanova and Osanov. 

The y dose of 12.7 Gy in about one hour implies a dose rate of about 30,000 R/d. This rate is 

one hundred times larger than the rates calculated from the aerographic surveys and listed in 

Table 8-2. If applied to a stationary pine tree for three weeks according to Equation 8.2, this 

initial dose rate would result in a dose about one hundred times higher than actually observed either 

from the satellite data or from the aerographic surveys. It seems likely then that the dose received 

by this individual was dominated by the airborne cloud dose for both beta and gamma rays as the 

140

adioactive plume passed over and around him and not by fallout deposition. Since the cloud dose 

would affect the forest for only a limited time as well, this assumption avoids inconsistency in the 

dose calculations for the pine forest. It indicates, however, that the cloud dose may be a 

substantial contributor to the dose received by the pine forest since the doses received by this 

individual are as much as 20% or 30% of the doses estimated for the nearby pine forest. Occurring 

over a shorter exposure time, the cloud dose would also be weighted more strongly than the fallout 

dose in the induction of biological responses in the foliage. 

8.4 UNCERTAINTIES IN THE IMAGE ANALYSIS. 

The following uncertainties provide important caveats for the interpretation of the results 

presented in this report and equally important guides for future research: 

1) Detectability offoliage radiation response. 

There is little quantitative information available on radiation-induced spectral changes in pine 

foliage. Our assumptions regarding the detectability from orbit of morphological and biological 

changes based on visual descriptions in the literature are subject to uncertainty. 

2) Seasonal variation offoliage response. 

The seasonal variation in the detectability of radiation-induced damage to foliage needs better 

definition, especially for intermediate doses in the sublethal to mid-lethal range. For these 

intermediate doses, meristem damage may be obvious during the growing season but not cause 

observable spectral changes outside the growing season. 

3) Relative biological effect (RBE). 

Most of the published data for the time dependence of spectral changes of foliage are for 

gamma rays or mixed exposure to gamma rays and fission neutrons. Fallout involves mixed 

exposures of gamma and beta rays. Although beta and gamma rays have similar microdosimetry, 

differences in depth distribution of dose may influence observable spectral changes. Neutron 

exposure causes ionization tracks at higher average linear energy transfer (LET) than either beta or 

gamma rays. In mammals, the RBE of neutrons is sometimes unity but may vary upwards or 

downwards by a factor of two or more depending on the biological endpoint under consideration. 

We are not aware of published data on RBE for spectral changes in foliage. 

141

4) Radiation transport in foliage. 

We have not accounted for the full three dimensional, heterogeneous nature of the fallout 

deposition and radiation transport problem. This neglect will influence the relative importance of 

the exposure contribution from beta and gamma exposure. 

5) Radionuclide release and fractionation. 

Uncertainty in the radionuclide mix released from the reactor and fractionation during transport 

and deposition contributes to the uncertainty in the beta to gamma dose ratio for the sensitive 

tissues of the foliage at different distances from the reactor. 

6) Cloud dose to the foliage. 

We have little information regarding the contribution of the radioactive cloud to direct exposure 

of the foliage as it drifted over and about the trees within a few kilometers of the reactor. 

7) Weathering of the fallout. 

Weathering of fallout particles from the foliage to the ground and subsequent migration into the 

ground are important factors in the time dependence of the beta to gamma dose ratio and the 

effective exposure time for the foliage. 

8) Physiological basis of pine tree response. 

Analysis of the spectral response of pine tree foliage to irradiation, especially the time 

dependence of that response, requires a careful study of the physiology of pines, including the 

behavior of important radiosensitive tissues and their depth distribution. The change in the 

sequence of physiological events leading to mortality as the dose is increased above the LD 5 0 is of 

particular importance and needs further elucidation. 

9) Influence of decontamination and containment activities. 

We know that helicopters sprayed polymers around the reactor site to immobilize fallout 

particles, but we do not know the timing, or location of such spraying and whether it influenced 

the satellite spectral observations for pine forest. 

142

SECTION 9 

CONCLUSION AND RECOMMENDATIONS 

Our analysis of Landsat imagery of the area within a few kilometers of the Chernobyl nuclear 

reactor station provides maps of radiation dose to pine forest canopy resulting from the accident of 

April 26, 1986. Detection of the first date of significant, persistent deviation from normal of the 

spectral reflectance signature of pine foliage produces contours of radiation dose in the 20 to 80 Gy 

range extending up to 4 km from the site of the reactor explosion. 

According to arguments presented in Section 5, the effective duration of the dominant exposure 

of the pine foliage is about 3 weeks. For this exposure time, the L 5 0 of Pinus sylvestris (Scotch 

pine) is about 23 Gy. At this dose level, the onset of persistent deviation from normal for the 

spectral signature is delayed until about one year after exposure. Around twice the L 50 , persistent 

deviation begins as soon as 4 to 5 weeks after the start of exposure. In a limited area, response of 

pine foliage was observed as soon as 12 days after the start of exposure, corresponding to a dose 

above 80 Gy. 

A patch of forest about 2 km west and 1.5 km north of the Unit 4 reactor site showed a 

transient deviation from normal during the late growing season of 1986. This deviation is likely 

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 

transient deviations, the practical threshold for remote detection of radiation dose to pine foliage 

with the Landsat Thematic Mapper is probably about 1/4 of the L 50 . 

These conclusions, stated relative to the L 50 , should remain valid even if the effective exposure 

time for the pine forest at Chernobyl is found to be different than three weeks and our dose 

estimates are adjusted accordingly. The threshold of detectability by remote observation of foliage 

response relative to the L 50 is likely to apply to other evergreen plant species, as well. 

We beliel/e that the results reported here contribute to an improved understanding of the effects 

of high levels of fallout radioactivity on vegetation and, especially, on the remote observation of 

radiation-induced foliage response and the extraction of dose estimates from those observations. 

The results may be used to gain an improved understanding of the radiation exposure 

consequences to personnel operating in an area contaminated by radioactive fallout and the impact 

of vegetation on that exposure. 

The following recommendations build on the results of the present research: 

1) Establish a cooperative effort with scientists of the former Soviet Union to better compare 

satellite data with ground studies made at specific locations within a few kilometers of the 

Chernobyl power plant. 

143

2) Reexamine the question of the relative importance of specific pine tree tissues in causing 

outward manifestations of radiation injury, especially considering variations during the 

annual cycle. 

3) Determine the contribution and impact of the airborne cloud exposure to the total dose 

received by the pine forest. 

4) Perform calculations of the beta and gamma exposures of sensitive tissues in the pine tree 

using improved geometry and improved data or tCie appropriate radionuclide mix. 

5) Analyze the satellite imagery to account for variations in seasonal observability of sublethal 

to midlethal damage. 

6) Extract characteristic, time-dependent spectral signatures for the pine forest as a function of 

dose to the foliage as a guide for future observations. 

7) Extend the methodology for dose estimation to include exposures during seasons other than 

spring. 

144

SECTION 10 

LITERATURE CITED 

Aleksakhin, R.M., F.A. Tikhomirov and N.Y. Kulikov, 1970, Status and Problems of Forest 

Radioecology, Soviet J. Ecology, 1:19-27. 

Amiro, B.D. and J.R. Dugle, 1985, Temporal Change in Boreal Forest Tree Canopy Cover Along 

A Gradient of Gamma Radiation, Canadian J. Botany, 63:15-20. 

Anno, G. H. and A. Laupa, 1989, Chernobyl Accident Fatalities and Causes, Technical Report 

DNA-TR-89-275, Defense Nuclear Agency, Washington, DC, February 1989. 

Anonymous, 1962, Atlas of the Ukrainian SSR and Maldavian SSR, Central Administration of 

Geodesy and Cartography, Moscow (In Russian). 

Asmolov, V.G., et. al., 1987, The Accident at the Chernobyl Nuclear Power Plant: One Year 

After, International Conference on Nuclear Power Performance and Safety, Vienna, 

28 September-2 October 1987, IAEA-CN-48/63, International Atomic Energy Agency, Vienna 

(Translated from Russian). 

Barabanova, A. and D. P. Osanov, 1990, The Dependence Of Skin Lesions On The Depth-Dose 

Distribution From Beta-irradiation Of People In The Chernobyl Nuclear Power Plant Accident, 

Int. J. Radiation. Biol., Vol. 57, No. 4, 775-782. 

Beck, H.L., 1980, Exposure Rate Conversion Factors for Radionuclides Deposited on the 

Ground, EML-378, Environmental Measurements Laboratory, U.S. Department of Energy. 

New York, NY. 

Beck, H. and G. De Planque, 1968, The Radiation Field in Air Due to Distributed Gamma-Ray 

Sources in the Ground, HASL-195, Environmental Measurements Laboratory, U.S. 

Department of Energy, New York, NY. 

Berg, L.S. 1950, Natural Regions of the USSR (O.A. Titelbaum, Translator), MacMillan Co., 

New York. 

Biatobok, S. and W. Zelawski (eds.), 1976, Outline of the Physiology of Scots (Scotch) Pine, 

Foreign Scientific Publications, Department of the National Center for Scientific, Technical and 

Economic Information, Warsaw. 

Borisov, A.A., 1965, Climates of the USSR (R. A. Ledward, translator), Aldine Publishing Co., 

Chicago, 225pp 

Bohlen, C., 1987, Chernobyl's Slow Recovery, Washington Post, July 21, 1987:A 1, A25. 

Bostrack, J.M. and A.H. Sparrow, 1969, Effects of Chronic Gamma Irradiation on the Anatomy 

of Vegetative Tissues of Pinus Rigida Mill, Radiation Botany 9:367-374. 

Bostrack, J.M. and A.H. Sparrow, 1970, The Radiosensitivity of Gymnosperms, i1, One the 

Nature of Radiation Injury and Cause of Death of Pinus Rigida and P. Strobus after Chronic 

Gamma Irradiation, Radiation Botany 10:131-143. 

145

Broido, A. and J.D. Teresi, 1961, Analysis of the Hazards Associated with the Radioactive Fallout 

Material, I, Estimation of A- and f3-doses, Health Physics 5:63-69. 

Chamberlain, A.C., 1970, Interception and Retention of Radioactive Aerosols by Vegetation, 

Atmos. Env. 4:57-78. 

Crist, E. P. and R. C. Cicone, 1984, A Physically-Based Transformation of Thematic Mapper 

Data -- The TM Tasseled Cap, IEEE Transactions on Geoscience and Remote Sensing, Vol. 

GE-22, No. 3, May 1984. 

Crist, E. P., R. Laurin, and R. C. Cicone, 1986, Vegetation and Soils Information Contained in 

Transformed Thematic Mapper Data, Proceedings of IGARSS 1986 Symposium, Zurich, 8-11 

September 1986, pp. 1465-1470. 

DOE, 1987, Health and Environmental Consequences of the Chernobyl Nuclear Power Plant 

Accident, DOE/ER-0332, UC-41 and 48, National Technical Information Service, Springfield, 

VA 22161. 

Dewdney, J.C., 1982, USSR in Maps, Holmes & Meier Publishers, Inc., New York. 

Donini, B., 1967, Effects of Chronic Gamma-Irradiation on Pinus Pinea and Pinus Halepensis, 

Radiation Botany, 7:183-192. 

Duda, R. 0., and Hart, P. E., 1973, Pattern Classification and Scene Analysis, John Wiley & 

Sons. 

Fridland, V.M., 1976, Patterns of Soil Cover, Israel Program for Scientific Translation, 

Jerusalem, 291 pp. 

Fullard, H., 1972, Soviet Union in Maps, George Philip & Son Ltd., London. 

Gamache, G., 1993, Private communication. Report in preparation for the Defense Nuclear 

Agency. 

Golovina, L. P., M. N. Lysenko and T. I. Kisel., 1980, Content and Distribution of Zinc in the 

Soil of the Ukrainian Poles'ye, Soviet Soil Science 12(1):73-80 

Hoffman, F.O. and C.F. Baes, III, 1979, A Statistical Analysis of Selected Parameters for 

Predicting Food Chain Transport and Internal Dose of Radionuclides, NUREG/CR- 

1004/ORNL/NUREGHM-282, Oak Ridge National Laboratory, Oak Ridge, TN 37830. 

International Advisory Committee, 1991, The International Chernobyl Project: An Overview, 

International Atomic Energy Agency, Vienna. 

Izrael, Y.A., V.N. Petrov and D.A. Severov, 1987, Modeling Radioactive Fallout Near the 

Chernobyl Nuclear Power Plant Accident, Meteorologiya i Gidrologiya 7:5-12. 

IZVESTIYA, 1989, Confession of a Veteran of Chernobyl. Written From a Hospital Bed, 

18001257 Moscow IZVESTIYA in Russian, 19 June 1989, Morning Edition, page 4. 

Reprinted in English in National Affairs, Suffering From Chernobyl-Related Illness Continues, 

FBIS-SOV-89-124, 29 June 1989, page 77-79. 

146

Kantz, A.D., 1971, Measurement of Beta Dose to Vegetation From Close-in Fallout, pp. 56-70, In 

D.W. Benson and A.H. Sparrow (eds.), Survival of Food Crops and Livestock in the Event of 

Nuclear War, CONF-700909, National Technical Information Service, Springfield, VA. 

Karaban, R.T., N.N. Mishenkov, B.S. Prister, R.M. Aleksakhin, F.N. Tikhomirov, G.N. 

Romanov, and M.A. Naryshkin, 1978, Radiation Effects on Arboreal Plants During the First 

Year After Acute Gamma-Irradiation of a Forest, Lesovedenii, 1:39-44. 

Keller, B. A., 1927, Distribution of Vegetation on the Plains of European Russia, Journal of 

Ecology 15:189-233. 

Kendrew, M. A., 1942, The Climates of the Continents, Oxford University Press, New York, 

473 pp. 

Klechkovskii, V.M., G.G. Polikarpov, and R.M. Aleksakhin, 1973, Radioecology (Translated 

from Russian by H. Kaner and H. Mills; translation edited by D. Greenberg, Israel Program 

for Scientific Translations, London), John Wiley and Sons, NY. 

Krupskiy, N. K., V. P. Kuz'michev, and R. G. Derevyanko., 1970, Humus Content in 

Ukrainian Soils, Soviet Soil Science 2:278-288. 

Lange, R., M.H. Dickerson and P.H. Gudiksen, 1987, Dose Estimates from the Chernobyl 

Accident, UCRL-96934 (preprint), Lawrence Livermore National Laboratory, Livermore, CA 

94550. 

Lysenko, M. N. and L. P. Golovina., 1982, Boron Content and Distribution in the Soils of the 

Ukrainian Poles'ye, Soviet Soil Science 14(1):89-97. 

Mackin, J., S. Brown and W. Lane, 1971, Measurement and Computational Techniques in Beta 

Dosimetry, pp. 51-55. In D.W. Benson and A.H. Sparrow (eds.), Survival of Food Crops 

and Livestock in the Event of Nuclear War. CONF-700909. National Technical Information 

Service, Springfield, VA. 

McClellan et al., 1992, Chernobyl Doses: Volume 2 -- Conifer Stress Near Chernobyl Derived 

from Landsat Imagery, Technical Report DNA-TR-92-37-V2, Defense Nuclear Agency, 

Washington, DC, December 1992. 

McCormick, J.F., 1967, Effects of Ionizing Radiation on a Pine Forest, pp. 78-87. In D.J. 

Nelson and F.C. Evans, (eds.), Symposium of Radioecology, CONF-670503, U.S. Atomic 

Energy Comm., Div. Tech. Inform. Ext., Oak Ridge, TN. 

Miller, G. L., 1968, The Influence of Season on the Radiation Sensitivity of an Old Field 

Community, Ph.D. Thesis, University of North Carolina, Chapel Hill, NC. 

Monk, C.D., 1966, Effects of Short-Term Gamma Irradiation on an Old Field, Radiation Botany 

6:329-335. 

NEA, 1987, The Radiological Impact of the Chernobyl Accident in OECD Countries, Nuclear 

Energy Agency, Organisation for Economic Co-Operation and Development, Paris Cedex, 

France. 

Oleynik, V. S., 1981, Genetic Characteristics of Peat Soils in the Ukrainian Western Poles'ye, 

Soviet Soil Science 13(4):32-37. 

147

Osanov, D.P., J.J. Tissen, and G.B. Radzievsky, 1969, Dose Distribution of 13 Radiation of 

Fission Products in the Tissue-Equivalent Material, Health Physics 17:489-495. 

Painter. E.L., F.W. Whicker, 1993, Chernobyl Doses: Habitat and Vegetation Near the 

Chernobyl Nuclear Reactor Station, Vol. 3, Defense Nuclear Agency, DNA-TR-92-37-V3. 

Pedigo, R.A., 1963, Effects of Ionizing Radiation on Pinus Taeda L, pp. 295-299. In V. Schultz 

and A.W. Klement, Jr., (eds.), Radioecology, Van Nostrand-Reinhold, Princeton, NJ. 

Platt, R.B., 1963, Ionizing Radiation and Homeostasis of Ecosystems, pp. 39-60. In 

G.M. Woodwell (ed.), Ecological Effects of Nuclear War, BNL-917 (C-43), Brookhaven 

national Laboratory, Upton, NY. 

Public Health Service, 1960, Radiological Health Handbook, Division of Radiological Health, 

Public Health Service, U.S. Dept. of Health, Education, and Welfare, Washington, DC. 

Rhoads, W.A., R.B. Platt, R.A. Harvey, and E.M. Romney, 1969, Ecological and Environmental 

Effects from Local Fallout from Cabriolet, 1, Radiation Doses and Short-Term Effects on the 

Vegetation from close-in Fallout, USAEC Rept, PNE-956, EG&G, Inc., Goleta, CA. 

Rodin, L.E., N.I. Bazilevich, and N.N. Rozov, 1975, Productivity of the World's Main 

Ecosystems, pp. 13-26. In Productivity of World Ecosystems, National Academy of 

Sciences, Washington, DC. 

Sanbur, G. N. and I. I. Kovalenko., 1969, Improvement and Rational Utilization of Lowland 

Saline Soils of the Southern Poles'ye and Northern Forest-Steppe of the Ukrainian SSR, 

Soviet Soil Science 1:1401-1408. 

Sheppard, M.I., D.H. Thibault, and J.E. Guthre, 1982, Radiosensitivity and the Role of Pinus 

Sylvestris L. In the Revegetation of Radioactive Disposal Areas, Environmental and 

Experimental Botany 22(2):193-198. 

Sobdovych, E. V., V.P. Dudkin, S.N. Kireyev, N.N. Kaletnik, I.N. Los, V.V. Gudzenko, D.A. 

Bobyleva, and L.Y. Tabachny, 1992, Chernobyl Catastrophe: Reasons and Consequences for 

Kiev, Minsk, and Moscow. Ukrainian Academy of Sciences: Kiev. 

Sparrow, A.H., A.F. Rogers, and S.S. Schwemmer, 1968, Radiosensitivity Studies with Woody 

Plants - I. Acute Gamma Irradiation Survival for 28 Species and Predictions for 190 species, 

Radiation Botany 8:149-186. 

Sparrow, A.H., S.S. Schwemmer, and P.J. Bottino, 1971, The Effects of External Gamma 

Radiation from Radioactive Fallout on Plants with Special Reference to Crop Production, 

Radiation Botany 11:85-118. 

Sukachev, V. N., 1928, Principles of Classification of the Spruce Communities of European 

Russia, Journal of Ecology, 16:1-18. 

Symons, L., 1972, Russian Agriculture: A Geographic Survey, John Wiley and Sons, 

New York, 348 pp. 

Szafer, W., 1966, The Vegetation of Poland, Pergamon Press, Oxford, 728 pp. 

148

Takhtajan, A., 1986, Floristic Regions of the World, (T. J. Crovello, translator), University of 

California Press, Berkeley. 

Tou, J. T. and R, C. Gonzalez, 1974, Pattern Recognition Principles, Addison-Wesley Publishing 

Company, Reading, Massachusetts, 1974 

Ursino, D.J., A. Moss, and J. Stimac, 1974, Changes in the Rates of Apparent Photosynthesis in 

21% and 1% Oxygen and of Dark Respiration Following a Single Exposure of Three-year-old 

Pinus Strobus L. Plants to Gamma Radiation, Radiation Botany 14:117-125. 

USSR, 1986, USSR State Committee on the Utilization of Atomic Energy: The Accident at the 

Chernobyl Nuclear Power Plant and its Consequences, Information compiled for the IAEA 

experts' meeting, 25-29 August 1986, Vienna. 

USSR, 1987, The Accident at the Chernobyl Nuclear Power Plant: One Year After, International 

Conference on Nuclear Power Performance and Safety, Vienna, Austria, 

28 September-2 October 1987, IAEA-CN-48/63, International Atomic Energy Agency, 

Vienna. 

Walter, H., 1978, Vegetation of the Earth and Ecological Systems of the Geobiospher, (J. Wieser, 

translator), Springer-Verlag, New York. 

Whicker, F.W. and L. Fraley, Jr., 1974, Effects of Ionizing Radiation on Terrestrial Plant 

Communities, pp. 317-366. In Advances in Radiation Biology, Vol. 4, Academic Press, Inc. 

Whicker, F.W. and T.B. Kirchner, 1987, PATHWAY. A Dynamic Food Chain Model to Predict 

Radionuclide Ingestion After Fallout Deposition, Health Physics 52(6):717-737. 

Wolfe, W. L. and G. J. Zissis, eds., 1978, The Infrared Handbook (Environmental Research 

Institute of Michigan), p. 3-140. 

Woodwell, G.M. and A.L. Rebuck, 1967, Effects of Chronic Gamma Radiation on the Structure 

and Diversity of an Oak-Pine Forest, Ecological Monographs, 37:53-69. 

Woodwell, G.M. and A.H. Sparrow, 1963, Effects of Ionizing Radiation on Ecological Systems, 

pp. 20-38. In G.M. Woodwell (ed.), Ecological Effects of Nuclear War, BNL-917 (C-43). 

Brookhaven National Laboratory, Upton, NY. 

Wulff, E. V., 1943, An Introduction to Historical Plant Geography, (E. Brissenden, translator), 

Chronica Botanica Co., Walham, Massachusetts, 223 pp. 

Young, R. W., 1988, Chernobyl in Retrospect, in Pharmac. Ther. Vol. 00, Eds. J. F. Weiss and 

M. G. Simic, Pergamon Press, Great Britain, 1988 

149

APPENDIX A 

FALLOUT DOSE CALCULATIONS FOR PINE FOREST CANOPY 

This appendix describes fallout beta and gamma radiation dose calculations in pine 

forest canopy for three dose points located 1, 7, and 12 meters above the ground, as 

indicated in Figure A-1. The 1-meter dose point is in air; the 7-meter point is in the 

middle of the canopy; and the 12-meter point is at the top of the canopy. Although we 

refer to dose calculations, the results are given in terms of dose rate, in cGy/h per unit 

source intensity. 

Fallout radiation sources are assumed to be homogeneously distributed both in the 

canopy mass (leaves and twigs) and on a flat surface of the ground below. 

homogenized canopy mass density of 0.0054 gm/cm 3 for medium density pine forest in 

mid-European latitudes, based on Kerr et al. (1971), is assumed to be of infinite extent 

laterally and finite vertically as indicated in Figure A-1. The dose calculations are all 

given in terms of the dose rate cGy/h per unit area or volume intensity, depending upon 

An 

Top canopy dose pt. 

Zu77 

Tree,. 32f Mid-canopy dose pt. 

• !•i:liiii!!i!!~ i !iiiii!~~iiii• . S.. T re e 32 ft 3• • 

CaCanopy 

.:+il!:iiii+ • i~~iii~:::+. height 40 ftIJ (10 m)* 

] 

....... ii~ (12 I s ,, 

.: !!N .~iii 

i'.... 

~ii!iii~ii..........iil S.... 

Air 8ft (2 m) 

"Dimensions assumed 

777777771'/77/7777•7////// 

7m 

Figure A-1. Pine forest canopy model for dose calculations based on Kerr et al., 1971. 

Dose points are 1, 7, and 12 meters ground surface. 

A-1

whether the fallout source is distributed over a surface (such as the ground) or within a 

space (such as the canopy), respectively. 

The beta particle dose calculations are not explicitly performed for each beta 

decay spectral component, but rather are based on the mean and maximum beta energies 

according to a semiempirical beta dose relationship (discussed later in this Appendix). 

Accordingly, the convention we utilize to designate the unit source intensity for beta 

particles is fp/cm 2 -sec or P/cm 3 -sec for radionuclide decay where "Pr" effectively 

represents all the beta decay spectra components from the excited radionuclide states. 

For the gamma dose calculations, the explicit gamma ray energies and 

corresponding fractional yields associated v. ith radionuclide decay 

are utilized. 

Accordingly, the convention we utilize to designate the unit source intensity for gamma 

rays is y/cm 2 -sec or y/cm 3 -sec where "y" represents the frequency weighted effective 

gamma ray decay components. 

GAMMA DOSES. 

Gamma dose calculations for foliage (pine needles or budding meristems) at the 

canopy dose points indicated in Figure A-1 are done in three parts. The ground fallout 

source is treated as one part and the fallout retained in the canopy is divided into two 

parts, that due to fallout in direct contact with (deposited on) the foliage element whose 

dose is being calculated and that due to fallout deposited on the rest of the canopy. 

Parameters for the gamma dose calculations are listed by gamma energy in Table A-1. 

Calculations were first performed for each gamma energy and then the specific 

radionuclide source parameters listed in Table A-2 were applied to calculate the dose for 

the radionuclide source mix in Table A-2 based on weighting by the individual released 

source strength, Q(MCi). 

The doses determined for each gamma energy in Table A-1 provide enough data to 

easily interpolate to find dose values for the radionuclide disintegration energies in 

Table A-2. Accordingly, if Rij(x) represents the gamma dose rate component for the ith 

energy of the jth radionuclide, the total dose rate at a dose point position x is 

Dy(x) = Y fiX Rij(x)Yij (A.1) 

j 

where, fj = Qj / XQj and Yij is disintegration yield in Table A-2. 

J 

The gamma dose is determined by integrating a point source dose function G(r) over 

a surface or volume source region, 9t, 

A-2

-) 

M 

A T-- (N CN 0 W) ot M 

w 

ChW 

l 

N 0 - O 

W)4 (iel 4C (2r 

t- r - It 0 n r- A t t 

(A 0 

U) 

0~ 

to 

E 

0u 

r 

a.tt 

011 

Cuo 

3 7o 

Cu > e 

COz 

rig 

w 

n 

'-4 co 

I~~~ 

W E 

g~ oo o0oo o 0 c 

A-3

00 1,- 

o 

00t 

0 000 

C6 

C6 

N %000 

00 

0 

en Rt00 

a e0% 

en 0 

cc 00 e 

0 0 C)N 

ui 00 N C 0 

A? 65 66 c; 6 cc, 

Q6 z C) C4 

00 n 00 )e 

0n 0 0 000 

0 - 00 W)C n000ýN 

>4 6 65dC 6 6 i6 6 6 66 1U0 

0)0 

E %C - 0% q C, *0' '0 ON -1(' 

00~ 00 q ~ n 0 e 

'0 00 CPO t 0 4N N0 0r' 

6 CC6 6 C5 66 d d 6 d 6 

N N% Q '0 r- -N n 

- ?- M I) 

00000~0 l-4:0'0002 

00 

ff 0~ei Cý 4 ( C4 " %f 46c 

E 

~~-4N.O00 

'00% 

0%cc 

E ~ 4 

o 66666 66 

NN''0y4~ '0 ~U 

0%b 

A-4a

L= KEy(9en/p)SJG(r)d9C 

where, Ey is the gamma energy in MeV, (pen/P) is the energy absorption coefficient in 

cm 2 /gm, S= 1, is the unit gamma source strength per unit area or volume, and K is a dose 

conversion constant, 

1.6xl 10( ergs )X3600(s-e) 

K 1=, MeV 7 =5.76x 10- 5 ( gm -cGy -sec 

10(ergs) 

MeV-h } 

Then an energy dependent dose constant, P.f(E), is 

Py(E) =5.76x 1O 5 E(-ten / p) 

and the dose rate, DL, will have units of (cGy/h)/(y/sec-cm 2 or cm 3 ). 

Canopy Fallout Volume Source/12- and 7-Meter Dose Point. The calculational 

geometry for the canopy fallout source and dose point at the top of the canopy is shown 

below in Figure A-2. 

Co 12 m dose point, Dp 

Canopy top 

_________ 

Ar $- - 

1= 10 m 

;~dA pd4odp 

Figuie A-2. Calculational geometry--canopy fallout volume source, dose point at 12 meters, side 

view upper panel, aerial view lower. 

A-5

As shown in Figure A-2, a flat disk source of radioactive fallout material of radius a 

is imbedded in the homogenized canopy mass a distance z from the top of the canopy. 

The differential element of area is dA=pd4dp. The dose at the top of the canopy from 

disc source having unit gamma source intensity, Sa=1 y/cm 2 -sec, is given by, 

2n a -U 

D= (E)Sa 2 f d B(4r)e" pdp 

o 

o 

- Py(E)Sa 2 Br)2 dp (cGy/h)/(y/cm 2 -sec). (A.2) 

0 

Since, r 2 =p 2 +z 2 , and rdr=pdp, 

DpPy(E)Sa s B(jr)e- d 

2 (A.3) 

z 

Implementing the Berger's form of the dose buildup factor for gamma radiation given by, 

B(p) = 1 + agreb 

,(A.4) 

where a and b are fit constants given in Table A-1, the gamma dose integral becomes, 

D Dp= Py(E)Sa[je 2 dr+aj e-(-b)Prdr . (A.5) 

The first term in the bracket, we call I(z,s) can be written as 

1 1 (z,s) = -du -J--du 

u u 

These integrals are a specific form of the general exponential integral function, 

b 

m -t 

Then I(Ipz, ps) = EI(Mtz) - Ej(ps) 

Integrating the second term in the bracket of Equation A.5 above, the dose, Dp1 

becomes, 

A-6

DP= Y(E)Sa Ejjz IP) ,(-~u - e -(1-b)gs] (A.6) 

Assuming the radioactive fallout to be distributed over a homogenized volume of the 

canopy mass, a unit finite volume source element at a depth z in the canopy is dSv = 

Sadz. Then, dDp 

dDp= 2){ }dz (A.7) 

For purposes of integrating the disc source over the canopy mass volume, the source disc 

may be approximated by a radially infinite source plane were s-'o and therefore, both 

EI(jts) and the exponential term, exp[-(1-b)ps], in Equations (A.6) and (A.7) are zero. 

Accordingly, the canopy mass is approximated as a homogeneous, vertically finite, 

laterally infinite slab source of radioactive material. Integrating Equation (A.7), 

D 

= Py(E)Sv 2 {tzz-z(ab)pzdz} Ej(pz)dz + 1--a e-1b)fd 

Py(E)Sv { a [1_ e-(l-b)W (A.8) 

0 E(d+(1- b)2• 

Utilizing a property of exponential integral functions, 

En(y) 

d-n+1 d E (y) 

the firbt integral in the bracket of Equation (A.8) is, 

t 

• 

fJE 1 (pz)dz= dE 2 (Y) [1- E 2 (W)] 

0 90 I 

Employing the parameters given in Table A-I, the dose at the 12-meter point at the top 

of the canopy, for f = 1000 cm, is then given by, 

2 

DP yE) { 1 - E 2 (Ige) + (a-) e--(I-b)WI 

=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 

given by Equation (A-9) for f = 500 cm, and then doubled. Figure A-3 gives the 

calculated gamma dose rates per unit source, (cGy/h)/(y/cm3-sec) at the top (12 rn dose 

point) and middle (7 rn dose point) of the canopy for gamma energies from 0.1 to 

3.0 MeV. The results are interpolated to obtain the dose rates, Rij(x), for the gamma-ray 

source, disintegration energies of the radionuclides in Table A-2. 

Then the source 

intensity-weighted gamma dose rate due to the homogenized canopy source at the two 

canopy dose points are given below employing Equation (A. 1). 

10-2 

EC.) 

10-3 - 

7i 

CD 

10-4 

0.1 1.0 3.0 

Gamma energy, MeV 

Figure A-3. Gamma dose rate in the canopy from canopy fallout volume source versus gamma energy. 

A-8

Dose Point 

Dose Rate, (cGy/h)/(Y/cm 3 -sec) 

Top of canopy 1.651 x 10-3 

12-meters 

Middle of canopy 2.064 x 10- 3 

7-meters 

Ground Fallout Source/12- and 7-Meter Dose Points. 

The calculational geometry for the ground fallout source for the dose point at the 

middle and top of the canopy is shown below in Figure A-4. In Figure A-4, a flat disc of 

radioactive material represents the ground fallout source a distance f = 7 m below the 

mid canopy dose point. The vertical air distance between the bottom of the canopy and 

the source disc is e 1 = 2 m; this region has an air density, Pi = 1.226 x 10- 3 g/cm 3 . The 

canopy thickness between the dose point and bottom of the canopy is f = 5 m; this 

region has a homogenized canopy density P2 = 0.0054 g/cm 3 . 

Canopy top 

_ _ _ _ __ OA__ _ _ _ _ _ _ __ _ _ _ _ _ _ _ 

12 m dose point 

''S r t-7m 

Canopy 

Air =2m 

77-r~d'~7,7777L1 

p 

dA PdOp 

Figure A-4. Calculational geometry--ground fallout source/dose point at 7 meters. 

A-9

The dose at the middle of the canopy from the disc source of unit source intensity 

(Sa = 1 y/cm 2 -sec) is given by, 

2x 

a 

Dp Py(E)Sa fd*f B(ri) exp(-ET.tri)pdp (A.10) 

0 0 

Since r 2 = p 2 + e 2 , and rdr = pdp, changing the variable of integration, 

P 

f(E)Sa B( ri)-exp(-4.r)rdr . (A.11) 

2 = r22 

Using Berger's form of the buildup factor, given by B(gtiri) 

combining with the exponential term in Equation (A.10), 

1 +alg, rii exp(bip.giri) and 

B(.•liri)exp(-•,lgiri) =exp(-Igiri)+ .•ailairi exp[-,Y.,(-bi)•tiri ] 

i i i i i 

Then since ri = eir/e, some of the summation terms can be simplified for purposes of 

integration, i.e., 

y~ =-gX.ri r =(r , a=19 1 

lagiri =r aigitir = Or, P3=-Yaigite ,and 

i i i 

r y 

F,(l-bj)gjtri=- 

i 

.(1-bjifi =gi= Y= •-Y(1-b)gjtji 

i 

where, p&i = (t/p)ipi and (p/p) 1 is the gamma mass attenuation coefficient for air (Pt = 

1.226 x 10-3, g/cm 3 ); (p,/P) 2 is that for the canopy (p 2 = 0.0054, g/cm 3 ). 

Equation (A. 11) can be rewritten as, 

(ESaj Oc+Jre idr 

D 2 f(e OeT)r (A. 12) 

t 

Integrating of first term in brackets results in the difference between two exponential 

integral functions: EI(ac) - E 1 (cts). Integrating the second term in the brackets, the dose 

is given as, 

Dp= PyP(E)S, FEI(f) --EI(as) + P.(e-Yf -e-15) 

A2 O1 

A-10

Then assuming an infinitely extended ground source plane (s--oo), the dose in the canopy 

from the fallout source on the ground is, 

Dp = 2.88x10-SE,(l.te /P)[Ej(oe)+ke-'ie] (cGy /h) /(/cm 2 -sec) . (A.13) 

Employing Equation (A.13) with the parameters given in Table A-1, doses were 

determined for various gamma energies at the middle of the canopy for t = 700 cm, f, = 

200 cm, and f2 = 500 cm; and at the top of the canopy for t = 1200 cm, tj = 200 cm,. 

and e2 = 1000 cm. 

Figure A-5 gives the calculated gamma dose rates per unit fallout ground source 

(cGy/h)/(y/cm 2 -sec) at the 7-meter and 12-meter dose points as function of gamma 

energy ranging from 0.1 to 3.0 MeV. 

Ground Fallout Dose at 1-Meter Height. 

The same form as given by Equation (A.13) was employed to calculate the ground 

fallout source dose at 1 meter in air above the surface for t = 100 m, a = (p/p) 1 pl, 

P = a(g/p) 1 and y = (1 - b)(gVp) 1 . Figure A-5 also gives the calculated gamma dose rates 

for the 1-meter dose as a function of gamma-ray energy. 

The results of the ground fallout source dose rates at all three dose points, 1 meter, 

7 meters, and 12 meters, were interpolated to obtain the dose rates, Rij(x), for the 

gamma-ray source, disintegration energies of the radionuclides in Table A-2. These 

values were then weighted and summed by the disintegration energy yields and 

radionuclide source intensity according to Equation (A. 1), and given below. 

Dose Point 

Dose Rate (cGy/h)/(y/cm 2 -sec) 

Top of canopy 1.011 X 10-6 

12 meters 

Middle of canopy 1.690 x 10-6 

7 meters 

1-meter height in air 2.936 x 10-6 

Canopy Fallout Contact Source/Dose to Foliage Elements. 

Calculations were performed to estimate the fallout gamma dose at the center of 

cylindrical foliage elements such as a pine needle or a budding meristem due to fallout 

deposited directly on the foliage element in question. It is assumed that fallout material 

A-I1

is homogeneously deposited on the surface of a cylinder of length 21 and radius p as 

shown in Figure A-6. 

10"5 - " .... __ ___ 

10-5 

(A 

E 

10-6 

107 

o / , 

0.1 1.0 3.0 

Gamma energy, MeV 

2t 

Figure A-5. Gamma dose rate from ground fallout source versus gamma energy. 

Dose point - 

I 

SdA 

= pddz 

Figure A-6. Calculational geometry for canopy fallout contact source. 

A-12

A point source gamma dose function at a small element of source area dA is 

integrated over the cylindrical surface to give the gamma dose at the dose point due to 

radionuclide fallout material on the surface. 

2M 

te -Ur 

Dy =Py(E)Sa fdfe 'dodz . (A.14) 

0 0 

Changing the variable of integration z--r, r2 = z 2 + p 2 , dz = rdr/z and z 2 = r 2 - p2, 

D = = Py(E)PSa 2Y 2 O r -"7r dr-- 

2 P r ,qr- -p" 

Py'(E)PSa e Per r (A.15) 

- 2.__''- '-____ 

22- 

Since wr

Calculations were based on an average length of 4.5 cm. For the budding meristem, the 

calculations were based on cylinders 0.04 and 0.0625 cm in diameter to simulate the dose 

depths discussed in Section 5.2. For horizontal orientation of the needle or meristem, the 

results obtained with Equation (A.17) should be doubled since the integration was over 

the half length f. On the other hand, if fallout was only on the top surface of the cylinder 

the results would in turn be halved due to symmetry considerations. We assume that 

these two factors cancel. However, assuming a random directional orientation of the axis 

of the pine needle (or meristem) between --t/2 from the horizontal, the values obtained 

from Equation (A.17) were multiplied by the average value of the cosine, Z (2/R) = 

0.63662 to account approximately for the effective area seen by incident fallout particles. 

Accordingly, the gamma dose values for various foliage element dimensions are given 

below. 

Dose Point 

Dose Rate (cGy/h)y/cm 2 -sec 

Radius (p) cm Length (2f) cm Random Orientation 

0.04 4.5 5.13 x 10-7 

0.04 9.0 5.14 x 10-7 

0.0625 9.0 5.10 x 10-7 

0.10 4.5 4.983 x 10-7 

0.15 9.0 4.95 x 10-7 

0.40 2.0 3.783 x 10-7 

BETA DOSES. 

Beta dose calculations were performed for the two dose points in the canopy 

indicated in Figure A-1 from fallout source radionuclide material dispersed in the canopy 

biomass. Beta dose calculations were also performed to estimate the dose in the foliage 

elements from radioactive fallout deposited on their surfaces. Because of the limited 

range of beta particles, the beta dose to the canopy dose points from fallout material on 

the ground surface was neglected. 

Fallout radionuclide parameters for the beta dose calculations are listed in Table A-3. 

Beta doses were calculated for each radionuclide component and then weighted 

according to each radionuclide source intensity, Q(MCi), and summed to obtain the dose 

for the source mix. If Rj(x) represents the beta dose rate component for the jth 

radionuclide, the dose rate at position x is, 

A-14

00 

* 72 

.00 

en enz 0r ce 0' 4C, t 

a, tn M Cu 

oZ 

aal 

-tr 0 

C 

2~~~ 0o0-u~ 

CA 

6.- 

AU-

D6(x)= fjRj(x) , (A.18) 

where, f 1 = Qj /I Qj . 

j 

The beta dose calculations are based on integrating an energy dependent point source 

dose function over the source geometries of interest. The point source dose function J(r) 

is a semi-empirically derived relationship for beta particles given by Loevinger, Japha, 

and Brownell (1956), 

where 

1.6 x 10- 

k 

Jr=k 2 {c[1 - 

(vr) 

(yr / c -e1vr/c)]+ vrel-vr IA1 

[]f0, r2c/v 

= 1.273 x 10-9p 2 v 3 E[t , 

a =-[3c2(c2-1)eC 

-0.55E. 

c = 3(e") 1 

V = v(E) 18.2 1. 37 (cm 2 ,g)an 

(E 0 - 0.036) 

r =source 

2 

-to -detector distance, g / cm 

p = density, g / cm 3 , 

Ep= average beta energy, MeV, 

Eo= maximum beta energy, MeV, 

v(Eo) = apparent absorption coefficient, cm 2 / g. 

(A.19) 

Canopy Fallout Volume Source/12- and 7-Meter Dose Point. 

The calculation procedure for canopy fallout source and canopy dose points is the 

same as that given for integrating the point source gamma dose function over the canopy 

volume of finite vertical width, and infinite lateral extent. Referring to the calculational 

geometry given in Figure A-2, the dose at the top of the canopy a vertical distance z from 

disc source of infinite radius imbedded in the canopy is, 

A-16

2n 

Dp = fJ(r)dA = f d# J(r)rdr 

V 0 z 

=_2xk ['c/v c[1- (vr /c)e1-(vr/c)] rd vrelw rdr (A.20) 

1 f r 2 fd+ dr.(.0 

V z rzr 

[ ]=O, rZc/v. 

Carrying out the integration of Equation (A.20), we obtain, 

D = pl(E){c[l + n(c / vz) - el-(vz/c) ]+ e-Vz}. 

(A.21) 

Expressing Dp above as a dose rate per unit source, (cGy/h)/(I/cm 2 -sec), Pp(E) = 2.879 x 

10" 5 vfa(Sa - 1). 

In order to obtain the beta dose at the top of the canopy, the infinite disc source 

(Equation A.21) is integrated over the canopy source region of thickness t = 1000 cm 

and of density Pc = 0.0054 g/cm. Furthermore, at that point, the beta dose is calculated 

at the center, x, of a pine needle of density p = 1 g/cm 3 , and thickness 0.2 cm, where 

then, x = 0.1 gm/cm 2 . The surface source of unit intensity S. extends to a unit volume 

source Sv, i.e., dSV = Sadz. The volume integral is then, 

x+1 

DP = fSDp(z)dz 

SC+(A .22) 

= PO(E)S,{fc41 +en(c/vz)-e1-(Pz/c)]+ 9lvzdz} 

For both dose points, 12 m or 7 m, , t = 1000 cm (5.4 g/cm 2 ) and f = 500 cm 

(2.7 g/cm 2 ) respectively, the canopy slab source is of infinite thickness since the 

distances exceed the maximum range, r., of the radionuclide disintegration betas given in 

Table A-3. Then given, I k ro, and integrating Equation (A.22), 

DP(x,*) = 0.5Dp c 2 [3 -e'(vx/c) - ix-( 2 + n c + el-VX } 

(.3 

]=0-x>c/v1 

[ 

A-17

where, Dp is the beta particle dose in the interior of a large (infinite) source as is the case 

of the canopy slab source. The beta dose at the surface of the top of the canopy is Dp/2, 

as indicated in Equation (A.23), assuming negligible back scatter from the air above. 

Expressing Dp as the dose rate per unit source, (cGy/h)/(I,/cm 3 -sec), Dp is, 

= 1.6x10-'6(ergs/g)xfEp(MeV/j3) 

100( ergs ) 0 X .0054( g__ 

[,g -cGy 

c 

= 1.0667 x 10- 2 E19 (cGy / h)/ (D/cm3 -sec) . (A.24) 

The dose point given by Equation (A.23) is actually 0.1 cm above the canopy slab 

surface to account for penetration into the center of a pine needle. Locating the dose 

point 0.1 cm below the canopy slab surface just inside the canopy, the beta dose is 

D , (x inside, o) = Do - Dp(x,o-) (cGy / h)/(P /cm' -sec). (A.25) 

Equation (A.25) is used to determine the beta dose rate at the top of the canopy (12- 

meter dose point). Then because of infinite slab thickness conditions of the canopy mass, 

the beta dose rate in the middle of the canopy (7-meter dose point) is obtained by 

doubling the value obtained from Equation (A.25) for the 12-m dose point. Calculations 

were performed for each radionuclide and weighted according to the source intensity and 

summed to obtain the beta dose rate for the radionuclide fallout mix; they are given 

below. 

Dose Point Beta Dose Rate (cGy/h)/(P/cm 3 -sec) 

Top of canopy 

3345 x 

12 meters 

Middle of canopy 

6.69 x 

7 meters 6.69_×_10-3 

Canopy Fallout Contact Source/Dose to Foliage Elements. 

Calculations were performed to estimate the fallout beta dose at the center of foliage 

elements such as a pine needle or a budding meristem. It is assumed that fallout material 

is homogeneously deposited on the surface of a cylinder of length 2V and radius p as 

shown in Figure A-6. 

A-18

The point source dose function, J(r), at the small element of source area, dA, is 

integrated over the cylindrical surface to give the beta dose at the dose point, 

Dp = SaJ J(r) 2npdz. 

(A.26) 

A 

Changing the variable of integration to r, r 2 = z2 + p 2 dz = rdr/z, and z = 2 , the 

dose becomes 

27pk 2l• f 2 J(r)rdr r2 V r2-p2 

•1 c[i_(vr/c)e - (vr/c)] dr + el-vrdr (A.27) 

Since in Equation (A.27) [ 0, sJt -+p2 > c / v, the integral of the first term in the [I 

brackets is (1/p)cos"1 (p/(c/v)). Then the dose becomes, 

P P [L L(c/S. v ,-.Lpdrl+v j e dr (A.28) 

.i Vr 2 -P 2 P 1 7 7 J 

The two integrals in Equation (A.28) were evaluated numerically. 

Because of the 

conditions indicated above for the [] bracket, the upper limit, UL is subject to the 

conditions given as follows. 

Condition 

c/v>Jp-2+t2 >P 

Fpv +t2 >c/V>p 

UL 

I/p 2 -+e 

c/V 

Jfp-7+ -t2 >p>c/v 0 

Equation (A.28) gives the beta dose rate per unit source intensity (cGy/h)/(p/cm 2 -sec) in 

the center of the foliage element where P3(E) = 2.879 x 10-5 pvE'a. 

A-19

The assumptions made for the calculations regarding dimensions, fallout source, 

surface symmetry, orientation, etc., are all discussed above under the subsection "Canopy 

Fallout Contact Source/Dose to Foliage Elements," which describes fallout gamma ray 

calculations parallel to those described here for beta radiation. Utilizing Equation (A.18) 

with the radionuclide fallout source intensity values in Table A-3, the beta dose rates 

obtained are as follows. 

Dose Point 

Dose Rate (cGy/h)y/cm 2 -sec 

Radius (p) cm Length (21) cm Random Orientation 

0.04 4.5 3.02x 10-5 

0.04 9.0 3.04 x 10-5 

0.0625 9.0 2.36 x 10-5 

0.10 4.5 1.73 x 10-5 

0.15 9.0 1.24 x 10-5 

0.40 2.0 0.32 x 10-5 

SUMMARY. 

The various calculated beta and gamma cose components from fallout radiation are 

summarized in Table A-4. The dose rates are all expressed in terms of cGy/h per unit 

source either in the canopy or on the ground surface and reflect equal radionuclide fallout 

deposition in the canopy mass and on the ground surface below. Dose rates are given at 

three locations (see Figure A-i), two in the canopy and a third at 1 meter above the 

ground surface for reference. In the canopy mass, one dose point is located at the top, 12 

meters above the ground, and another in the middle, 7 meters above the ground. 

The results given for gamma and beta dose rates per unit volume source in the 

canopy reflect the adjustment made for the reference volume density in the canopy to be 

consistent with equal fallout deposition in the canopy mass and on the ground surface. 

That is, the fallout deposited in the canopy mass is assumed to be homogeneously 

distributed over the 1000 cm thickness of the canopy rather than on the plane surface of 

the ground, i.e., Sv(A or y/cm 3 -sec) = Sa(O or y/cm 2 -sec) x10- 3 . 

In addition to the dose rate contributions in the canopy from fallout radionuclide 

sources distributed in canopy mass and ground surface below, Table A-4 gives dose rates 

at these locations in the middle of cylindrical foliage elements from fallout radionuclides 

assumed to be deposited on their surfaces. 

A-20

caa 

a. 4.a 

.4,r4 

2- - - 

2 x x x x * 2 

0%q ON r 

oo 

0 A- 

c 

a- 

m co U. co U 

Q 

U.: 

- 

- - - - 2 

A-21 X

Inspection of Table A-4 indicates that the dominant dose component in canopy mass 

is beta radiation due to fallout deposition on the surface of foliage elements assuming 

that the fallout deposits equally in the canopy mass and on the ground surface below and 

that the surface density of fallout on the foliage elements is the same as that on the 

ground. 

A-22

APPENDIX B 

SPECTRAL DEVIATIONS FROM CLASS 

This appendix presents our method for constructing a deviation vector for a pixel in standard 

units relative to a reference class of pixels. 

Each pixel in a multispectral image is represented by a vector of its band intensities. The 

pixels belonging to a single class form a cluster in the hyperspace of band intensities. We 

assume that the cluster of pixels belonging to a reference site for the class may be approximated 

by a multivariate normal distribution. Let u be the mean vector of the pixels in ihe reference site 

and I be the covariance matrix for the reference site. 

In Bayes decision theory, the distance of a pixel x from the mean of its class is called the 

Mahalanobis distance (Duda and Hart, 1973) when it is scaled relative to the covariance matrix 

Y. The square of the Mahalanobis distance r is given by 

r2 = (x -)T Z-1 (x - A). (B-1) 

For a multivariate distribution, the Mahalanobis distance r is the equivalent of the normal 

deviate z used for a univariate normal distribution. It is a measure of the deviation of a pixel 

from its class mean in standard units. Since r is a scalar quantity, it provides no information 

regarding the direction of the deviation in hyperspace. 

We have generalized the Mahalanobis distance to a Mahalanobis vector m that points in the 

direction of deviation and whose magnitude is the Mahalanobis distance r. For a pixel that 

deviates from its class, the magnitude r indicates the significance of the deviation. For a 

statistically significant deviation, the direction of the vector m carries information regarding the 

likely cause of the deviation. 

Assume that a multispectral image has been transformed to Tasseled Cap space (TC space). 

In general, the cluster for a reference site will be hyperelliptical with its principle axes tilted with 

respect to the TC axes and with unequal variance along the principle axes. The upper panel of 

Figure B-I illustrates a reference site cluster in two dimensions. The eigenvectors of the 

covariance matrix of the cluster lie along the principle component coordinate axes of the cluster. 

The eigenvalues are the variances of the cluster in the direction of these axes. 

B-1

U) 

ou Reference Deviation 

SCluster 

-~Site 

Vco 

0 A •' 

(U 

0) 

Principle 

Coordinote 

Axes of 

Cluster 

Tu,,;:eled Cap Greenness 

"1 Normalization" 

Wetness 

Deviation 

•. t. ;" .. ;= p 

S..-Maholonobis Vector m 

* ..- *- .- 

"- Greenness 

.. "'" ... . Deviation 

Figure B-I. 

The normalization procedure defined in Appendix B transforms the reference site 

cluster into a cluster with unit covariance matrix. 

B-2

Let U be the matrix whose column vectors are the unit eigenvectors of the covariance matrix 

Y- Let x be a pixel vector in TC space and d = x - g be the deviation of the pixel from the 

reference site mean in TC space. The deviation d can be expressed in the principle component 

coordinate system of the reference site cluster by the transformation 

d = UT d. (B-2) 

Likewise, the covariance matrix .p of the cluster in the principle component coordinate system is 

: p = UT yU. (B-3) 

Since U is constructed from the eigenvectors of 1, the matrix I 

is diagonal with its diagonal 

components equal to the variance of the cluster along each of its principle axes. In fact, the usual 

procedure for finding the matrix U is to diagonalize Y- 

The reference site cluster may be scaled to a spherical distribution with unit variance along 

each of its axes in the principle component coordinate system by dividing each component of the 

pixel deviation vectors dp by the standard deviations of the cluster along the corresponding 

principle axis. This scaling is accomplished by multiplication of each pixel deviation vector by a 

normalizing matrix Np: 

mp =- Np dp, 

(B-4) 

The matrix Np is a diagonal matrix given by 

N Np = Ip-1/2 (B-5) 

where the root is taken term-by-term on the right hand side of the equation. In other words, the 

diagonal terms of N are the inverses of the standard deviations of the reference site cluster along 

the principle axes. 

The vector mp given by Equation 4 for a pixel is the desired Mahalanobis vector expressed in 

the principle component coordinate system of the cluster. Since U is a unitary matrix, 

U- 1 = UT. So mnp may be expressed in the Tasseled Cap coordinate system by the 

transformation 

which is the inverse transformation of that in Equation 2. 

m= Ump. (B-6) 

B-3

Combining Equations 6, 4, 2, and the definition of d, we have 

m= UNPUT (x-) (B-7) 

Equation 7 shows that the normalizing matrix N for calculating Mahalanobis vectors in the 

TC coordinate system is 

N = U [UT L U] -1/2 UT. (B-8) 

where the root on the square brackets is taken term-by-term. Finally, the desired Mahalanobis 

vector expressing the deviation vector of a pixel x from its class in standard units in Tasseled 

Cap space is 

m = N (x - R). (B-9) 

This normalization, when applied to each pixel of the reference site cluster, results in a cluster 

with unit covariance matrix as illustrated in the lower panel of Figure B- 1. 

If the cluster is well approximated by a multivariate normal distribution, then the significance 

of the deviation of any given pixel from the mean of the cluster may be judged with a chisquared 

test on the value of r 2 for the pixel. In practice, the usual procedure is to adjust a 

threshold of significance on r 2 by inspection of the resulting spatial and temporal patterns of 

significant deviations. The threshold is raised just enough to eliminated random spatial patterns 

of deviation or to eliminate deviations that occur before the time of a known stimulus. 

B-4

APPENDIX C 

CHRONOLOGY OF SELECTED EVENTS 

SATURDAY, 26 APRIL 1986,01:23 

" two explosions of Unit 4; concrete, graphite, and debris escaped through roof; hole 

exposed graphite core (1) 

"* smoke and fumes with radioactive material rose in a hot plume about 1800 m high (1) 

"* heavier debris and particles fell near site (1) 

"• lighter particles to west and north (1) 

"• winds at 1500 m were 8-10 km/s from SE (1) 

"* plant firemen arrived within minutes (1) 

"* burning graphite on roof of Unit 3 (1) 

"• At 

"° At = 1.5h Unit 3 shutdown (1) 

"• At = 24h 

Units 1 &2shutdown(1) 

26 APRIL 1986,0500 

First person report of Valeriy Fedorovich Zosimov (2) 

"Early in the morning, about 0500, I slipped into Kiev to meet my 

family. We returned in a private car. We reached Kopachy, not far 

from Pripyat, near the station. A captain with a portable radio gave us 

permission to leave the vehicle in Kopachy and from there walk home 

to Pripyat. So, we went.... My 10-year-old daughter, my wife, and me. 

Ahead and behind people were also walking, and from Kopachy the 

destuoyed fourth unit could already be seen. 

Where the power 

transmission line crosses the road there was a long band of graphite 

smoke. 

I shook the light black flakes from my coat, and they 

immediately dispersed. We arrived home at night (later a large part of 

the forest around that path was declared hot and came under the ax-it 

was dangerous there! Just as it was when I walked along it with my 

family?)" 

26 APRIL 1986, DAWN 

° all fires extinguished except burning graphite in core (1) 

C-1

27 APRIL 1986 

1200 announcement of evacuation broadcast in Pripyat (1) 

1400 - 1700 Pripyat evacuated with 1200 buses that had assembled in Chernobyl (1) 

Line several kilometers long (1) 

Some of population of Pripyat had already left, so the number transported was 

less than the 44,600 projected (1) 

Population of Pripyat moved initially to surrounding towns and villages 

2 MAY 1986 

Evacuation of 30 km danger zone begun. 

4 MAY 1986 

High radiation levels force government headquarters from Pripyat to Chernobyl. 

6 MAY 1986 

End of atmospheric release of radioactivity from core (1) 

6 MAY 1986 

Evacuation of the danger zone (30 km radius) completed (1) 

JUNE 1986 

Start construction of hydraulic emergency structures (3) 

Water Protection (1) 

As part of the protection of rivers and the Kiev Reservoir, an effort was made to slow the 

movement of long-lived radionuclides through ground or surface water. Three major 

undertakings were: 

• 140 dams and dikes to limit runoff from the site area into the codirn pond and the Pripyat 

river. 

* existing silt traps at the bottoms of the rivers, the cooling pond, and the Kiev Reservoir 

were scoured. 

* ground water barrier was built around the plant to prevent the flow of radioactive water 

towards the River Dnepr. The barrier was 8 km long and 30-35 m deep, down to the 

impxrmeable clay layer. 

C-2

Hydraulic Emergency Structures (3) 

a filtration-proof wall in the soil along part of the perimeter of the site of the power plant 

and wells to lower the water table 

* a drainage barrier for the cooling pond 

* a drainage cutoff barrier on the right bank of the Pripyat river 

• a drainage interception barrier in the south-west section of the plant 

* drainage water purification facilities 

MID-NOVEMBER 1986 

Completion of sarcophagus (1) 

APRIL 1987 

Completion of work begun in May 1986 for protecting the water system. (1) 

SOURCES 

(1) International Advisory Committee, 1991. 

(2) IZVESTIYA, 1989. 

(3) 

C-3/C-4

APPENDIX D 

EVERGREEN SPECTRAL SIGNATURES BY CLASS AND DATE 

Z1. Spectral Signature of Classes on the Winter/Summer Composite Image CONP21 

Name of class = FOREST1 Number of points in sample = 481 

Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7 

Min 45.00 124.00 150.00 57.00 138.00 111.00 147.00 

Mean 53.53 134.61 160.67 65.30 147.03 128.89 151.16 

Max 66.00 144.00 180.00 81.00 154.00 143.00 157.00 

Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7 

Band 1 21.86 -19.39 24.98 17.52 -11.79 -19.44 6.87 

Band 2 -19.39 19.16 -23.20 -16.96 12.02 19.46 -6.31 

Band 3 24.98 -23.20 31.25 21.48 -14.66 -23.80 7.58 

Band 4 17.52 -16.96 21.48 21.89 -13.55 -25.50 6.35 

Band 5 -11.79 12.02 -14.66 -13.55 10.49 16.15 -4.39 

Band 6 -19.44 19.46 -23.80 -25.50 16.15 35.48 -7.12 

Band 7 6.87 -6.31 7.58 6.35 -4.39 -7.12 5.72 

Name of class = FOREST2 Number of points in sample = 1945 


Min 43.00 122.00 142.00 54.00 140.00 98.00 137.00 

Mean 46.39 145.35 149.59 56.57 155.10 142.73 138.33 

Max 74.00 150.00 179.00 93.00 160.00 149.00 142.00 

Covar Band 1 Band 2 Band 3 Band 4 Band F Band 6 Band 7 

Band 1 6.73 -6.66 7.96 4.67 -1.30 -5.27 1i15 

Band 2 -6.66 8.56 -9.37 -5.18 2.73 6.66 -1.33 

Band 3 7.96 -9.37 12.41 6.27 -2.98 -7.82 1.60 

Band 4 4.67 -5.18 6.27 6.25 -2.68 -7.18 1.03 

Band 5 -1.90 2.73 -2.98 -2.68 3.63 4.37 -. 66 

Band 6 -5.27 6.66 -7.82 -7.18 4.37 11.86 -1.18 

Band 7 1.15 -1.33 1.60 1.03 -. 66 -1.18 1.06 

D-I

Name of class = UNSUP CLASS 3 


Min 42.00 126.00 146.00 58.00 123.00 89.00 125.00 

Mean 55.32 132.61 158.74 71.12 147.73 118.98 152.09 

Max 59.00 149.00 175.00 97.00 174.00 133.00 180.00 


Band 1 11.33 -5.84 9.35 -1.41 -. 32 4.63 -. 18 

Band 2 -5.84 13.56 -6.40 1.01 1.20 1.15 .59 

Band 3 9.35 -6.40 36.26 -9.20 -3.18 10.78 -1.46 

Band 4 -1.41 1.01 -9.20 28.99 3.12 -15.72 -9.25 

Band 5 -. 32 1.20 -3.18 3.12 29.52 6.53 -6.79 

Band 6 4.63 1.15 10.78 -15.72 6.53 34.63 -7.10 

Band 7 -. 18 .59 -1.46 -9.25 -6.79 -7.10 43.78 



Min 42.00 126.00 146.00 18.00 137.00 115.00 125.00 

Mean 51.91 136.68 157.51 62.03 150.74 130.99 146.72 

Max 59.00 151.00 174.00 78.00 163.00 147.00 174.00 


Band 1 10.17 -7.66 9.11 2.67 -. 35 -1.66 -. 85 

Band 2 -7.66 11.74 -9.06 -1.64 1.08 2.31 2.28 

Band 3 9.11 -9.06 18.57 .77 -. 43 -1.17 -2.34 

Band 4 2.67 -1.64 .77 10.25 -. 39 -5.85 -1.02 

Band 5 -. 35 1.08 -. 43 -. 39 8.53 1.75 .19 

Band 6 -1.66 2.31 -1.17 -5.85 1.75 12.17 -3.08 

Band 7 -. 85 2.28 -2.34 -1.02 .19 -3.08 23.89 



---- ............... ----- --------.--........................ 

Min 42.00 126.00 146.00 5.00 113.00 125.00 34.00 

Mean 46.82 142.92 151.21 56.30 154.12 140.47 139.67 

Max 59.00 153.00 167.00 69.00 169.00 159.00 161.00 

Covar Band I Band 2 Band 3 Band 4 Band 5 Band 6 Band 7 

Band 1 4.06 -2.71 2.70 .54 -1.26 -. 32 -1.13 

Band 2 -2.71 6.73 -4.05 -1.05 2.93 3.20 -. 16 

Band 3 2.70 -4.05 5.62 .61 -2.36 -2.30 .03 

Band 4 .54 -1.05 .61 8.11 .89 -3.98 3.05 

Band 5 -1.26 2.93 -2.36 .89 8.21 3.03 .50 

Band 6 -. 32 3.20 -2.30 -3.98 3.03 10.68 -6.14 

Band 7 -1.13 -. 16 .03 3.05 .50 -6.14 17.15 

D-2



Min 42.00 126.00 146.00 37.00 137.00 107.00 104.00 

Mean 49.97 138.04 154.41 61.46 159.59 136.74 139.40 

Max 59.00 151.00 168.00 77.00 176.00 156.00 156.00 


Band 1 7.19 -3.56 3.73 .41 -3.19 1.09 -1.35 

Band 2 -3.56 8.94 -3.20 -. 13 1.46 .59 3.31 

Band 3 3.73 -3.20 10.78 -1.84 -5.09 .05 -. 12 

Band 4 .41 -. 13 -1.84 8.30 2.24 -2.16 -3.38 

Band 5 -3.19 1.46 -5.09 2.24 24.94 4.07 1.72 

Band 6 1.09 .59 .05 -2.16 4.07 10.97 -3.73 

Band 7 -1.35 3.31 -. 12 -3.38 1.72 -3.73 20.67 

Name of class UNSUP CLASS 7 


Min 42.00 126.00 146.00 57.00 147.00 109.00 114.00 

Mean 55.18 131.70 159.84 67.61 163.87 130.24 142.73 

Max 59.00 146.00 175.00 88.00 175.00 143.00 166.00 


Band 1 7.75 -3.97 4.85 1.18 -1.81 -1.14 .08 

Band 2 -3.97 9.72 -4.55 -1.12 .80 1.79 2.22 

Band 3 4.85 -4.55 24.13 -5.80 -1.34 1.59 .43 

Band 4 1.18 -1.12 -5.80 18.08 1.04 -10.21 -. 90 

Band 5 -1.81 .80 -1.34 1.04 21.94 3.48 4.36 

Band 6 -1.14 1.79 1.59 -10.21 3.48 17.77 -6.67 

Band 7 .08 2.22 .43 -. 90 4.36 -6.67 27.50 



Min 42.00 126.00 146.00 61.00 160.00 123.00 85.00 

Mean 54.21 131.43 157.61 70.10 177.29 136.55 137.41 

Max 59.00 146.00 173.00 82.00 193.00 157.00 155.00 


Band 1 11.00 -6.93 7.83 2.89 2.30 -2.29 -. 01 

Band 2 -6.93 12.81 -6.84 -4.95 -2.17 2.12 3.75 

Band 3 7.83 -6.84 24.59 -. 39 6.41 -3.94 4.62 

Band 4 2.89 -4.95 -. 39 11.97 4.68 -2.58 -5.08 

Band 5 2.30 -2.17 6.41 4.68 27.57 3.32 4.10 

Band 6 -2.29 2.12 -3.94 -2.58 3.32 13.36 -7.05 

Band 7 -. 01 3.75 4.62 -5.08 4.10 -7.05 24.94 

D-3

Name of class = CLASS 3 REFERENCE SITE Number of points in sample 70 


Min 50.00 128.00 155.00 64.00 137.00 112.00 151.00 

Mean 55.31 132.07 161.81 69.44 144.97 123.59 154.19 

Max 59.00 137.00 169.00 77.00 153.00 131.00 158.00 


---- --------........ 

----- --------........................- 

Band 1 3.96 -3.29 3.97 2.35 -. 79 -2.21 1.26 

Band 2 -3.29 3.82 -4.13 -2.76 1.92 3.27 -1.45 

Band 3 3.97 -4.13 7.35 .88 -3.35 -. 92 1.02 

Band 4 2.35 -2.76 .88 9.79 -2.40 -9.56 2.70 

Band 5 -. 79 1.92 -3.35 -2.40 8.00 2.05 -. 37 

Band 6 -2.21 3.27 -. 92 -9.56 2.05 13.70 -3.74 

Band 7 1.26 -1.45 1.02 2.70 -. 37 -3.74 2.72 

Name of class CLASS 4 REFERENCE SITE Number of points in sample = 253 


Min 45.00 131.00 151.00 56.00 145.00 126.00 146.00 

Mean 50.11 137.43 156.95 59.35 150.43 131.92 150.58 

Max 56.00 143.00 164.00 63.00 156.00 138.00 157.00 


Band 1 4.94 -4.09 5.64 1.99 -. 13 -2.06 -. 36 

Band 2 -4.09 6.30 -6.03 -2.09 1.09 3.31 -. 51 

Band 3 5.64 -6.03 9.20 2.77 -. 80 -3.59 -. 34 

Band 4 1.99 -2.09 2.77 2.57 -. 94 -3.21 .14 

Band 5 -. 13 1.09 -. 80 -. 94 4.22 2.96 -. 39 

Band 6 -2.06 3.31 -3.59 -3.21 2.96 7.45 -1.46 

Band 7 -. 36 -. 51 -. 34 .14 -. 39 -1.46 6.83 

Name of class = CLASS 5 REFERENCE SITE Number of points in qample = A4 


Min 43.00 141.00 146.00 54.00 152.00 134.00 137.00 

Mean 45.12 146.38 148.48 55.53 155.51 143.44 137.69 

Max 50.00 150.00 155.00 59.00 160.00 147.00 140.00 


Band 1 .66 -. 32 .40 .19 .11 -. 15 .12 

Band 2 -. 32 1.48 -. 80 -. 27 .47 .61 -. 20 

Band 3 .40 -. 80 1.58 .27 -. 37 -. 60 .19 

Band 4 .19 -. 27 .27 .61 -. 13 -. 47 .00 

Band 5 .11 .47 -. 37 -. 13 2.19 1.20 -. 40 

Band 6 -. 15 .61 -. 60 -. 47 1.20 2.57 -. 41 

Band 7 .12 -. 20 .19 .00 -. 40 -. 41 .48 

D-4

Name of class = CLASS 6 REFERENCE SITE Number of points in sample = 467 


Min 46.00 131.00 146.00 57.00 149.00 130.00 132.00 

Mean 48.29 138.08 154.12 61.14 157.37 137.04 136.99 

Max 55.00 143.00 160.00 66.00 169.00 143.00 141.00 


Band 1 1.65 -. 71 1.25 .70 -1.18 -1.55 .81 

Band 2 -. 71 2.89 -2.22 -. 47 -. 43 1.43 -. 51 

Band 3 1.25 -2.22 4.60 .06 -2.18 -2.50 .98 

Band 4 .70 -. 47 .06 2.47 2.56 -. 32 -. 71 

Band 5 -1.18 -. 43 -2.18 2.56 14.12 5.21 -3.63 

Band 6 -1.55 1.43 -2.50 -. 32 5.21 6.40 -2.39 

Band 7 .81 -. 51 .98 -. 71 -3.63 -2.39 3.29 

Name of class = CLASS 8 REFERENCE SITE Number of points in sample = 133 


Min 48.00 126.00 147.00 63.00 169.00 134.00 135.00 

Mean 53.65 131.73 158.92 67.95 177.39 138.28 137.26 

Max 59.00 137.00 169.00 73.00 186.00 142.00 139.00 


Band 1 7.33 -5.52 8.04 3.90 6.26 -. 40 -.01 

Band 2 -5.52 6.96 -7.57 -3.83 -6.27 .29 -.14 

Band 3 8.04 -7.57 12.63 4.63 7.68 -. 53 -.04 

Band 4 3.90 -3.83 4.63 4.42 6.75 -. 32 -. 13 

Band 5 6.26 -6.27 7.68 6.75 14.60 1.02 -.72 

Band 6 -. 40 .29 -. 53 -. 32 1.02 2.27 .03 

Band 7 -. 01 -. 14 -. 04 -. 13 -. 72 .03 .40 

D-5

E2. Spectral Signatures for Five Zvergreen Classes on each Date 

Class 3, Date 1; 6 JUN 85 Number of points in sample = 70 


Min 64.00 137.00 112.00 119.00 152.00 159.00 151.00 

Mean 69.44 144.97 123.59 121.96 156.47 160.17 154.19 

Max 77.00 153.00 131.00 125.00 160.00 162.00 158.00 


Band 1 9.79 -2.40 -9.56 .28 -3.07 .17 2.70 

Band 2 -2.40 8.00 2.05 -. 66 1.66 -. 29 -. 37 

Band 3 -9.56 2.05 13.70 .14 3.17 -. 01 -3.74 

Band 4 .28 -. 66 .14 1.24 .19 .17 -. 10 

Band 5 -3.07 1.66 3.17 .19 2.73 -. 21 -. 90 

Band 6 .17 -. 29 -. 01 .17 -. 21 .49 -. 09 

Band 7 2.70 -. 37 -3.74 -. 10 -. 90 -. 09 2.72 

Class 3, Date 2; 21 MAR 86 Number of points in sample = 72 


Min 50.00 128.00 155.00 121.00 158.00 159.00 89.00 

Mean 55.31 132.07 161.83 127.69 159.75 160.62 89.47 

Max 59.00 137.00 169.00 134.00 162.00 162.00 90.00 


Band 1 3.88 -3.23 3.90 3.66 -. 37 .12 -. 14 

Band 2 -3.23 3.73 -4.08 -3.73 .25 -. 05 -. 03 

Band 3 3.90 -4.08 7.21 4.84 -. 55 .09 .04 

Band 4 3.66 -3.73 4.84 5.97 .03 .06 -. 08 

Band 5 -. 37 .25 -. 55 .03 .79 -. 13 -. 08 

Band 6 .12 -. 05 .09 .06 -. 13 .50 .00 

Band 7 -. 14 -.03 .04 -. 08 -. 08 .00 .25 

Class 3, Date 3; 29 APR 86 Number of points in sample = 72 


Min 58.00 137.00 124.00 120.00 156.00 158.00 131.00 

Mean 61.18 139.53 131.51 123.36 158.92 159.65 132.78 

Max 66.00 143.00 136.00 126.00 161.00 162.00 136.00 


Band 1 3.23 -. 68 -4.04 -. 29 -. 47 .23 .73 

Band 2 -. 68 1.94 .38 -. 16 .23 -. 26 -. 42 

Band 3 -4.04 .38 7.92 .53 .05 -. 28 -1.18 

Band 4 -. 29 -. 16 .53 1.43 -. 01 -. 08 .07 

Band 5 -. 47 .23 .05 -. 01 1.28 -. 16 .07 

Band 6 .23 -. 26 -. 28 -. 08 -. 16 .49 .20 

Band 7 .73 -. 42 -1.18 .07 .07 .20 1.03 

D-6

Class 3, Date 4; 8 MAY 86 Number of points in sample = 72 

Band I Band 2 Band 3 Band 4 Band 5 Eand 6 Band 7 

Min 57.00 139.00 119.00 117.00 155.00 159.00 140.00 

Mean 62.51 143.28 127.58 120.69 158.74 160.43 143.69 

Max 69.00 148.00 134.00 123.00 162.00 162.00 148.," 


Band 1 4.81 -1.96 -5.09 .43 -1.65 -. 15 .53 

Band 2 -1.96 4.01 1.87 -1.24 1.16 .29 -. 39 

Band 3 -5.09 1.87 8.80 .48 1.89 .50 -2.30 

Band 4 .43 -1.24 .48 2.26 .00 .10 - .42 

Band 5 -1.65 1.16 1.89 .00 1.86 .02 -. 34 

Band 6 -. 15 .29 .50 .10 .02 .58 -. 24 

Band 7 .53 -. 39 -2.30 -. 42 -. 34 -. 24 3.81 



Min 63.00 137.00 121.00 118.00 155.00 159.00 147.00 

Mean 66.93 143.11 128.10 123.19 158.24 160.28 150.40 

Max 73.00 150.00 134.00 126.00 161.00 162.00 154.00 


Band 1 4.56 -1.78 -4.53 .64 -1.55 .26 .89 

Band 2 -1.78 5.02 2.63 -1.41 1.43 .02 -. 86 

Band 3 -4.53 2.63 8.51 -. 24 2.03 .07 -1.97 

Band 4 .64 -1.41 -. 24 1.85 -. 44 .14 .22 

Band 5 -1.55 1.43 2.03 -. 44 1.45 .06 -. 31 

Band 6 .26 .02 .07 .14 .06 .59 -. 02 

Band 7 .89 -. 86 -1.97 .22 -. 31 -. 02 2.73 



Min 63.00 136.00 124.00 124.00 154.00 159.00 151.00 

Mean 67.59 140.59 131.67 126.63 158.56 160.57 154.63 

Max 74.00 147.00 139.00 129.00 162.00 162.00 159.00 


Band 1 5.13 -1.37 -6.03 .62 -1.34 .66 1.03 

Band 2 -1.37 3.74 2.94 -. 32 .48 -. 07 -. 13 

Band 3 -6.03 2.94 9.49 -. 74 1.67 - .44 -1.84 

Band 4 .62 -. 32 -. 74 1.21 -. 02 .03 .07 

Band 5 -1.34 .48 1.67 -. 02 1.59 - .13 -. 36 

Band 6 .66 -. 07 -. 44 .03 -. 13 .51 -. 09 

Band 7 1.03 -. 13 -1.84 .07 -. 36 -. 09 3.32 

D-7

Class 3, Date 7; 15 OCT 86 Number of points in sample 72 


Min 30.00 150.00 133.00 97.00 159.00 158.00 101.00 

Mean 33.32 152.24 137.10 99.65 160.65 159.03 101.54 

Max 36.00 155.00 141.00 101.00 163.00 161.00 102.00 


Band 1 1.55 .54 -. 86 -. 10 -. 07 .09 .36 

Band 2 .54 1.40 .25 -. 17 - .07 .05 .19 

Band 3 -. 86 .25 2.54 .37 -. 24 -. 07 -. 12 

Band 4 -.10 -. 17 .37 .76 -. 03 11 -. 02 

Band 5 -. 07 -. 07 -. 24 -. 03 .74 .10 -. 08 

Band 6 .09 .05 -. 07 .11 .10 .53 -. 01 

Band 7 .36 .19 -. 12 -. 02 -. 08 -. 01 .25 

Class 3, Date 8; 2 DEC 86 Number of points in sample = 72 


Min 22.00 146.00 137.00 90.00 158.00 159.00 78.00 

Mean 23.86 147.79 138.87 92.65 160.43 159.63 78.85 

Max 26.00 149.00 141.00 95.00 162.00 161.00 79.00 


Band 1 .59 -. 03 -. 04 .08 .07 .20 .05 

Band 2 -.03 .61 .03 .00 .07 -. 03 -. 03 

Band 3 -. 04 .03 .59 .22 .01 .08 .00 

Band 4 .08 .00 .22 1.09 .06 .09 .02 

Band 5 .07 .07 .01 .06 .76 -. 04 -. 01 

Band 6 .20 -. 03 .08 .09 -. 04 .30 .05 

Band 7 .05 -. 03 .00 .02 -. 01 .05 .13 



Min 57.00 140.00 106.00 115.00 153.00 159.00 127.00 

Mean 61.22 144.14 124.44 117.43 158.15 160.18 129.64 

Max 74.00 147.00 131.00 120.00 162.00 162.00 132.00 


Band 1 9.06 -2.49 -10.47 .60 -2.95 .31 1.36 

Band 2 -2.49 3.02 2.77 -. 10 .72 .09 .16 

Band 3 -10.47 2.77 16.02 .12 3.73 -. 38 -2.24 

Band 4 .60 -. 10 .12 1.25 -. 20 -. 07 .38 

Band 5 -2.95 .72 3.73 -. 20 1.97 -. 17 -. 64 

Band 6 .31 .09 -. 38 -. 07 -. 17 .74 .23 

Band 7 1.36 .16 -2.24 .38 -. 64 .23 1.87 

D-8

Class 3, Date 10; 7 SEP 87 Number of points in sample 72 


Min 50.00 141.00 134.00 114.00 158.00 158.00 124.00 

Mean 52.53 144.90 140.28 116.81 159.93 159.42 126.14 

Max 56.00 150.00 145.00 119.00 163.00 161.00 128.00 


Band 1 1.75 -. 03 -1.84 -. 09 -. 20 .17 .48 

Band 2 -. 03 3.61 1.34 .15 .46 -. 34 -. 39 

Band 3 -1.84 1.34 5.74 .87 .62 -. 06 -.99 

Band 4 -. 09 .15 .87 1.47 .21 -. 03 -. 33 

Band 5 -. 20 .46 .62 .21 1.03 -. 07 -. 09 

Band 6 .17 -. 34 -. 06 -. 03 -. 07 .35 -. 06 

Band 7 .48 -. 39 -. 99 -. 33 -. 09 -. 06 .96 



Min 52.00 146.00 103.00 108.00 150.00 158.00 145.00 

Mean 56.49 149.96 123.54 112.11 156.17 159.93 148.24 

Max 65.00 159.00 132.00 115.00 159.00 161.00 153.00 


Band 1 5.33 -. 70 -7.86 .28 -2.02 .34 1.83 

Band 2 -. 70 7.13 1.28 -. 84 -. 33 -. 86 -. 23 

Band 3 -7.86 1.28 16.69 1.04 3.29 -. 40 -4.47 

Band 4 .28 -. 84 1.04 1.57 .21 .15 -. 47 

Band 5 -2.02 -. 33 3.29 .21 2.42 -. 07 -. 42 

Band 6 .34 -. 86 -. 40 .15 -. 07 .52 .26 

Band 7 1.83 -. 23 -4.47 -. 47 -. 42 .26 4.22 



Min 56.00 145.00 126.00 116.00 156.00 158.00 146.00 

Mean 59.35 150.43 131.92 118.71 159.34 160.18 L50.58 

Max 63.00 156.00 138.00 122.00 162.00 162.00 157.00 

Covar Band 1 Band 2 Band 3 Band 4 Band 5 Band 6 Sand 7 

Band 1 2.57 -. 94 -3.21 -. 02 -. 21 -. 15 .14 

Band 2 -. 94 4.22 2.96 -. 10 .14 -. C4 -. 39 

Band 3 -3.21 2.96 7.45 .25 .41 .21 -1.46 

Band 4 -. 02 -. 10 .25 1.23 .21 .09 -. 29 

Band 5 -. 21 .14 .41 .21 1.10 .02 - .07 

Band 6 -. 15 -. 04 .21 .09 .02 .54 .28 

Band 7 .14 -. 39 -1.46 -. 29 -. 07 .28 6.83 

D-9



Min 45.00 131.00 151.00 118.00 157.00 158.00 88.00 

Mean 50.12 137.42 156.96 123.72 159.64 160.58 89.24 

Max 56.00 143.00 164.00 129.00 162.00 163.00 91.00 


Band 1 4.92 -4.07 5.62 3.72 -. 25 .36 -. 47 

Band 2 -4.07 6.27 -6.03 -4.48 .27 -. 28 .60 

Band 3 5.62 -6.03 9.14 5.18 -. 43 .41 -. 81 

Band 4 3.72 -4.48 5.18 5.15 -. 12 .41 -. 53 

Band 5 -. 25 .27 -. 43 -. 12 .73 .01 .07 

Band 6 .36 -. 28 .41 .41 .01 .70 -. 04 

Band 7 -. 47 .60 -. 81 -. 53 .07 -. 04 .68 



Min 51.00 139.00 132.00 119.00 159.00 158.00 127.00 

Mean 53.91 143.06 137.56 121.47 160.78 160.02 129.60 

Max 57.00 147.00 142.00 124.00 164.00 162.00 135.00 


Band 1 1.29 .09 -1.46 -. 04 .02 .10 .29 

Band 2 .09 1.84 .49 -. 39 .03 -. 18 -. 22 

Band 3 -1.46 .49 4.37 .33 .15 -. 14 -. 54 

Band 4 -. 04 -.39 .33 1.32 .03 .17 -. 13 

Band 5 .02 .03 .15 .03 .82 .04 -. 10 

Band 6 .10 -.18 -. 14 .17 .04 .51 .03 

Band 7 .29 -. 22 -. 54 -. 13 -. 10 .03 2.78 



Min 51.00 144.00 126.00 115.00 157.00 159.00 135.00 

Mean 55.02 148.54 133.40 118.34 160.17 160.18 140.12 

Max 59.00 155.00 140.00 122.00 163.00 162.00 148.00 


Band 1 2.12 -. 36 -2.71 -. 13 -. 18 .18 .61 

Band 2 -. 36 3.08 1.72 -. 28 .03 -. 29 -. 64 

Band 3 -2.71 1.72 6.50 .59 .24 -. 23 -1.02 

Band 4 -. 13 -. 28 .59 1.52 .18 .06 - .16 

Band 5 -. 18 .03 .24 .18 1.08 -.11 -. 16 

Band 6 .18 -. 29 -. 23 .06 -.11 .47 .20 

Band 7 .61 -. 64 -1.02 -. 16 -. 16 .20 5.98 

D-10


Band 1 Band 2 Band 3 Band 4 Bind 5 Band 6 Band 7 

Min 55.00 144.00 126.00 116.00 158.00 158.00 141.00 

Mean 58.52 148.84 134.08 119.60 160.03 160.31 146.03 

Max 63.00 155.00 ±40.00 122.00 163.00 162.00 153.00 


Band 1 2.17 -. 28 -2.41 .02 -. 13 .13 .36 

Band 2 -. 28 4.28 1.73 -. 90 .01 .12 -. 37 

Band 3 -2.41 1.73 5.73 .21 .19 -. 04 -. 88 

Band 4 .02 -. 90 .21 1.44 .00 .10 - .51 

Band 5 -. 13 .01 .19 .00 1.01 -. 02 -. 38 

Band 6 .13 .12 -. 04 .10 -. 02 .60 -.15 

Band 7 .36 -. 37 -. 88 -. 51 -. 38 -. 15 5.63 



Min 57.00 141.00 131.00 122.00 156.00 158.00 143.00 

Mean 60.10 145.17 138.00 125.03 160.02 160.02 147.70 

Max 64.00 152.00 143.00 128.00 163.00 162.00 153.00 


Band 1 1.91 -. 36 -2.14 -. 23 -.11 .00 .59 

Band 2 -. 36 3.50 1.47 .05 .01 -. 22 -. 35 

Band 3 -2.14 1.47 5.31 1.03 .11 .02 -1.54 

Band 4 -. 23 .05 1.03 1.33 .03 .01 -. 45 

Band 5 -.11 .01 .11 .03 1.17 -. 06 -. 27 

Band 6 .00 -.22 .02 .01 -. 06 .75 .09 

Band 7 .59 -. 35 -1.54 -. 45 -. 27 .09 3.19 

Class 4, Date 7; 15 OCT 86 Number of points in sample = 257 


Min 30.00 148.00 131.00 97.00 158.00 157.00 100.00 

Mean 32.05 152.93 138.25 99.37 160.67 159.48 101.35 

Max 36.00 156.00 142.00 102.00 164.00 162.00 103.00 


Band 1 1.28 .39 -. 52 -. 08 -.11 -. 16 .04 

Band 2 .39 2.20 1.26 -. 20 -. 01 -. 07 -. 01 

Band 3 -. 52 1.26 3.01 .15 -. 03 .09 .03 

Band 4 -. 08 -. 20 .15 .90 .02 .08 -. 02 

Band 5 -.11 -. 01 -. 03 .02 1.04 .10 .02 

Band 6 -. 16 -. 07 .09 .08 .10 .69 .00 

Band 7 .04 -. 01 .03 -. 02 .02 .00 .28 

D-11



Min 22.00 146.00 136.00 89.00 159.00 158.00 79.00 

Mean 23.41 148.33 138.48 91.72 160.47 159.38 79.39 

Max 25.00 150.00 141.00 94.00 162.00 161.00 81.00 


Band 1 .44 A08 -. 07 .11 .04 .09 -. 01 

Band 2 .08 .75 -. 25 -. 19 -. 03 -. 10 .01 

Band 3 -. 07 -. 25 .74 .03 .05 .11 -. 04 

Band 4 .11 -. 19 .03 1.09 .00 .07 .06 

Band 5 .04 -. 03 .05 .00 .54 .08 00 

Band 6 .09 -. 10 .11 .07 .08 .38 .00 

Band 7 -. 01 .01 -. 04 .06 .00 .00 .27 



Min 50.00 143.00 126.00 113.00 157.00 158.00 125.00 

Mean 54.30 148.23 132.39 115.82 159.75 159.70 127.62 

Max 59.00 153.00 138.00 119.00 162.00 161.00 131.00 


Band 1 2.54 -. 51 -2.76 -. 05 -. 21 .32 .13 

Band 2 -. 51 2.07 1.59 -. o2 .02 -. 24 -. 21 

Band 3 -2.76 1.59 6.08 .22 .18 -. 19 -. 28 

Band 4 -. 05 -. 02 .22 1.02 .08 -. 01 .05 

Band 5 -. 21 .02 .18 .08 .93 -. 02 -.11 

Band 6 .32 -. 24 -. 19 -. 01 -. 02 .57 .00 

Band 7 .13 -. 21 -. 28 .05 -.11 .00 1.89 

Class 4, Date 10; 7 SEP 87 Number of points in sample = 257 


Min 47.00 142.00 138.00 113.00 158.00 157.00 123.00 

Mean 49.09 145.98 142.65 116.34 160.18 159.33 124.18 

Max 51.00 153.00 147.00 119.00 163.00 161.00 128.00 


Band 1 .81 -. 07 -. 86 .07 .16 .14 .14 

Band 2 -.07 2.33 .86 -. 07 -. 16 -. 18 .12 

Band 3 -. 86 .86 2.76 .34 -. 33 .05 .05 

Band 4 .07 -. 07 .34 1.29 .15 .03 -. 09 

Band 5 .16 -. 16 -. 33 .15 1.00 .08 .01 

Band 6 .14 -. 18 .05 .03 .08 .60 .09 

Band 7 .14 .12 .05 -. 09 .01 .09 1.24 

D-12

Class 4, Date 11; 28 MAY 88 Number of points in sample 257 


Min 47.00 149.00 120.00 108.00 153.00 157.00 139.00 

Mean 50.36 153.25 130.41 110.87 157.78 159.98 142.70 

Max 56.00 161.00 136.00 114.00 162.00 162.00 151.00 


Band 1 1.78 -. 29 -2.01 .54 -. 39 -. 03 .40 

Band 2 -. 29 3.41 1.79 -. 58 .07 -. 07 -. 61 

Band 3 -2.01 1.79 5.96 .00 .61 .19 -1.55 

Band 4 .54 -. 58 .00 1.39 -. 06 .08 .24 

Band 5 -. 39 .07 .61 -. 06 1.34 -. 02 -. 09 

Band 6 -. 03 -. 07 .19 .08 -. 02 .58 .02 

Band 7 .40 -. 61 -1.55 .24 -. 09 .02 4.43 

Class 5, Date 1; 6 JUN 85 Number of points in sample 664 


Min 54.00 152.00 134.00 117.00 156.00 157.00 137.00 

Mean 55.53 155.51 143.44 119.85 159.20 159.55 137.69 

Max 59.00 160.00 147.00 123.00 161.00 162.00 140.00 


Band 1 .61 -. 13 -. 47 .06 .00 .03 .00 

Band 2 -. 13 2.19 1.20 -. 03 -.11 -. 06 -. 40 

Band 3 -. 47 1.20 2.57 .25 -. 09 .08 -. 41 

Band 4 .06 -. 03 .25 1.33 A04 .04 -. 07 

Band 5 .00 -. 11 -. 09 .04 .78 .02 .06 

Band 6 .03 -. 06 .08 .04 .02 .61 -. 03 

Band 7 .00 -. 40 -. 41 -. 07 .06 -. 03 .48 



Mil 43.00 141.00 146.00 114.00 158.00 158.00 87.00 

Mean 45.12 146.38 148.43 117.71 159.95 159.97 89.56 

Max 50.00 150.00 155.00 122.00 163.00 162.00 91.00 


Band 1 .66 -. 32 .40 .48 .05 .08 .07 

Band 2 -. 32 1.47 -.80 -. 72 -. 05 -. 17 -. 08 

Band 3 .40 -. 80 1.59 .65 -. 03 .18 -. 03 

Band 4 .48 -. 72 .65 1.59 .08 .10 -. 01 

Band 5 .05 -. 05 -. 03 .08 .66 -. 02 .02 

Band 6 .08 -. 17 .18 .10 -. 02 .47 -. 03 

Band 7 .07 -. 08 -.03 -. 01 .02 -. 03 .51 

D-13

Class 5, Date 3; 29 APR 86 Number of points in sample 665 


Min 52.00 143.00 138.00 119.00 158.00 158.00 123.00 

Mean 54.00 145.96 144.03 122.73 160.86 159.72 123.80 

Max 56.00 149.00 148.00 126.00 164.00 161.00 126.00 


Band 1 .40 -. 01 -. 25 .09 .01 .07 .08 

Band 2 -. 01 1.02 .23 -. 29 .00 -. 12 -. 08 

Band 3 -. 25 .23 1.38 .23 .04 .00 -. 15 

Band 4 .09 -. 29 .23 1.30 .06 .02 -. 02 

Band 5 .01 .00 .04 .06 .88 .04 -. 02 

Band 6 .07 -. 12 .00 .02 .04 .41 .02 

Band 7 .08 -. 08 -. 15 -. 02 -. 02 .02 .43 



Min 52.00 147.00 136.00 116.00 157.00 158.00 128.00 

Mean 5:,.75 151.02 141.14 119.68 160.28 159.86 129.90 

Max 57.00 155.00 144.00 123.00 163.00 162.00 133.00 


Band 1 .52 -. 15 -. 36 .07 -. 06 .12 .11 

Band 2 -. 15 1.29 .46 -. 18 .07 -. 18 -. 25 

Band 3 -. 36 .46 1.67 .09 .21 .01 -. 31 

Band 4 .07 -. 18 .09 1.17 .04 .03 .05 

Band 5 -. 06 .07 .21 .04 .81 -. 03 -. 05 

Band 6 .12 -. 18 .01 .03 -. 03 .49 .03 

Band 7 .11 -. 25 -. 31 .05 -. 05 .03 .58 



Min 54.00 149.00 136.00 116.00 157.00 158.00 134.00 

Mean 55.49 152.51 142.02 119.99 160.08 159.60 135.21 

Max 59.00 155.00 145.00 124.00 163.00 162.00 138.00 


Band 1 .57 -.11 -. 22 .07 .01 .03 -. 02 

Band 2 -.11 1.43 .57 -. 16 .06 -. 09 - .31 

Band 3 -. 22 .57 1.64 .20 .04 .06 -. 16 

Band 4 .07 -. 16 .20 1.40 .05 .11 -. 02 

Band 5 .01 .06 .04 .05 1.01 .03 -. 07 

Band 6 .03 -. 09 .06 .11 .03 .50 .08 

Band 7 -. 02 -. 31 -. 16 -. 02 -. 07 .08 .95 

D-14



Min 55.00 144.00 138.00 120.00 157.00 158.00 138.00 

Mean 57.00 148.73 145.70 125.28 160.19 159.87 139.94 

Max 60.00 152.00 149.00 129.00 164.00 162.00 142.00 


Band 1 .62 -. 15 -. 04 .37 .00 -. 02 -. 09 

Band 2 -. 15 1.46 .41 -. 48 -. 02 -. 01 -. 02 

Band 3 -. 04 .41 1.78 .46 .01 -. 04 -. 28 

Band 4 .37 -. 48 .46 1.96 .07 .01 -. 19 

Band 5 .00 -. 02 .01 .07 .85 -. 05 -. 02 

Band 6 -. 02 -. 01 -. 04 .01 -. 05 .60 .01 

Band 7 -. 09 -. 02 -. 28 -. 19 -. 02 .01 .72 



Min 30.00 154.00 138.00 96.00 156.00 157.00 101.00 

Mean 31.98 156.68 142.43 98.86 160.25 159.21 102.04 

Max 34.00 160.00 146.00 103.00 163.00 161.00 103.00 


Band 1 .49 .26 -. 02 .04 -. 01 .01 -. 01 

Band 2 .26 1.23 .22 -. 15 -. 10 -. 12 -. 15 

Band 3 -. 02 .22 1.14 .13 -. 10 .07 -. 09 

Band 4 .04 -. 15 .13 .99 -. 03 -. 03 -. 01 

Band 5 -. 01 -. 10 -. 10 -. 03 .91 .00 .06 

Band 6 .01 -. 12 .07 -. 03 .00 .72 .05 

Band 7 -. 01 -. 15 -. 09 -. 01 .06 .05 .50 



Min 21.00 147.00 137.00 89.00 157.00 158.00 78.00 

Mean 23.39 149.57 139.48 91.70 160.22 159.31 78.58 

Max 25.00 153.00 145.00 95.00 162.00 161.00 80.00 


Band 1 .33 .05 -. 05 .10 .04 .04 .00 

Band 2 .05 .79 -. 06 -. 17 -. 07 -. 05 .00 

Band 3 -. 05 -. 06 .76 .07 -. 05 .05 .03 

Band 4 .10 -. 17 .07 .99 .08 -. 03 .01 

Band 5 .04 -. 07 -. 05 .08 .59 -. 01 -. 02 

Band 6 .04 -. 05 .05 -. 03 -. 01 .32 -.01 

Band 7 .00 .00 .03 .01 -. 02 -. 01 .29 

D-15



Min 49.00 150.00 135.00 112.00 157.00 157.00 118.00 

Mean 51.55 153.28 139.53 115.86 159.91 159.23 119.29 

Max 55.00 157.00 143.00 120.00 163.00 161.00 122.00 


Band 1 .58 .01 -. 30 .23 -. 02 .08 .07 

Band 2 .01 1.54 .82 -. 25 -. 06 -. 10 -.29 

Band 3 -. 30 .82 2.23 .18 -. 05 .02 -. 40 

Band 4 .23 -. 25 .18 1.26 .12 .11 .03 

Band 5 -. 02 -. 06 -. 05 .12 .83 .02 .09 

Band 6 .08 -. 10 .02 .11 .02 .42 .02 

Band 7 .07 -. 29 -. 40 .03 .09 .02 .60 


Band I Band 2 Band 3 Band 4 Band 5 Band 6 Band 7 

Min 46.00 148.00 145,00 114.00 148.00 157.00 118.00 

Mean 47.87 151.43 149.96 116.19 159.69 159.18 119.03 

Max 64.00 193.00 169.00 119.00 164.00 162.00 121.00 

Covar Rand 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7 

Band 1 .99 1.69 .72 .18 -. 42 .10 -. 02 

Band 2 1.69 5.11 2.15 -. 25 -1.19 .26 -. 07 

Band 3 .72 2.15 1.78 .13 -. 67 .15 -. 05 

Band 4 .18 -. 25 .13 1.29 .04 .03 -. 05 

Band 5 -. 42 -1.19 -. 67 .04 .94 -.10 -. 01 

Band 6 .10 .26 .15 .03 -. 10 .47 -. 01 

Band 7 -. 02 -. 07 -. 05 -. 05 -. 01 -.01 .24 



Min 46.00 150.00 122.00 108.00 153.00 158.00 131.00 

Mean 48.19 154.80 135.24 110.55 158.38 160.17 132.67 

Max 52.00 i58.00 139.10 115.00 168.00 162.00 135.00 


Band 1 .49 -. 02 -. 41 .16 .10 .07 .02 

Band 2 -. 02 1.19 .44 -. 30 709 -. 14 -. 22 

Band 3 -. 41 .44 2.07 .27 -. 15 .03 -. 20 

Band 4 .16 -. 30 .27 1.66 .06 .17 -. 09 

Band 5 .13 .09 -. 15 .06 1.37 -. 06 -. 03 

Band 6 07 -. 14 .03 .17 -. 06 .35 -. 07 

Band 7 .02 -. 22 -. 20 -. 09 -. 03 -. 07 .41 

D-16

Class 6, Date 1; 6 JUN 85 Number of points in sample 467 


---- ................ ----- --------.-.......................- 

Min 57.00 149.00 130.00 118.00 157.00 158.00 13.-00 

Mean 61.14 157.37 137.04 121.26 159.55 159.66 136.9-' 

Max 66.00 169.00 143.00 125.00 162.00 162.00 141.00 


Band 1 2.47 2.56 -. 32 .62 -. 14 .19 - .71 

Band 2 2.56 14.12 5.21 -. 55 -. 34 .31 -3.63 

Band 3 -. 32 5.21 6.40 .61 -. 54 .25 -2.39 

Band 4 .62 -. 55 .61 1.97 .00 .14 -. 49 

Band 5 -. 14 -. 34 -. 54 .00 .99 -. 06 .17 

Band 6 .19 .31 .25 .14 -. 06 .54 - .17 

Band 7 -. 71 -3.63 -2.39 -. 49 .17 -. 17 3.29 



---- ........................--- ............................. 

Min 46.00 131.00 146.00 120.00 158.00 159.00 86.00 

Mean 48.29 138.10 154.11 123.77 160.13 160.48 87.53 

Max 55.00 143.00 160.00 129.00 163.00 162.00 89.00 


Band 1 1.64 -. 73 1.26 1.02 -. 12 .08 .04 

Band 2 -. 73 2.92 -2.29 -1.60 .16 -. 04 .17 

Band 3 1.26 -2.29 4.63 1.95 -. 32 .02 -. 29 

Band 4 1.02 -1.60 1.95 2.42 -. 06 .12 -. 07 

Band 5 -. 12 .16 -. 32 -. 06 .66 .01 .03 

Band 6 .08 -. 04 .02 .J2 .01 .46 .01 

Band 7 .04 .17 -. 29 -. 07 .03 .01 .39 



---- ........................--- ............................. 

Min 53.00 139.00 131.00 121.00 159.00 158.00 124.00 

Mean 55.68 142.57 140.01 124.55 161.25 159.84 126.26 

Max 60.00 146.00 144.00 128.00 164.00 162.00 130.00 


Band 1 1.02 .12 -. 84 .19 -. 10 -. 03 .39 

Band 2 .12 1.33 .45 -. 26 -. 08 -.13 -. 19 

Band 3 -. 84 .45 3.19 .31 .04 .02 -. 47 

Band 4 .19 -. 26 .31 1.46 .01 .14 .06 

Band 5 -. 10 -. 08 .04 .01 .89 .02 -. 03 

Band 6 -. 03 -. 13 .02 .14 .02 .56 .01 

Band 7 .39 -. 19 -. 47 .06 -.03 .01 1.26 

D-17



Min 52.00 146.00 128.00 11'.00 158.00 158.00 130.00 

Mean 55.11 1J0.67 137.15 120.04 160.66 159.91 133.41 

Max 60.00 161.00 142.00 123.00 164.00 162.00 139.00 


Band 1 1.47 .83 -. 91 -. 13 -. 10 .08 .27 

Band 2 .83 5.07 2.04 -. 16 -. 21 -. 22 -1.50 

Band 3 -. 91 2.04 4.06 .41 -. 22 -. 07 -1.48 

Band 4 -. 13 -. 16 .41 1.30 -. 02 -. 08 -.10 

Band 5 -.10 -. 21 -. 22 -. 02 1.17 -. 07 .02 

Band 6 .08 -. 22 -. 07 -. 08 -. 07 .51 .22 

Band 7 .27 -1.50 -1.48 -. 10 .02 .22 2.18 



Min 56.00 148.00 131.00 117.00 158.00 158.00 134.00 

Mean 58.56 154.63 138.10 121.14 160.41 159.92 138.30 

Max 66.00 169.00 144.00 125.00 163.00 162.00 143.00 


Band 1 2.32 2.18 -. 12 .30 -. 28 .14 -. 06 

Band 2 2.18 11.45 4.34 -. 07 -. 97 -. 22 -3.25 

Band 3 -. 12 4.34 5.40 .44 -. 50 .05 -2.14 

Band 4 .30 -. 07 .44 1.47 -. 04 .15 -. 14 

Band 5 -. 28 -. 97 -. 50 -. 04 .89 -. 03 .26 

Band 6 .14 -. 22 .05 .15 -. 03 .64 -. 02 

Band 7 -. 06 -3.25 -2.14 -. 14 .26 -. 02 3.52 



Min 58.00 143.00 135.00 122.00 157.00 157.00 137.00 

Mean 60.39 149.61 141.88 125.93 160.63 159.46 141.68 

Max 65.00 160.00 147.00 130.00 163.00 162.00 147.00 


Band 1 1.51 1.51 -. 62 -. 03 -. 17 .02 .05 

Band 2 1.51 11.08 3.94 -. 20 -. 36 .23 -4.18 

Band 3 -. 62 3.94 4.46 .30 -. 09 .31 -2.47 

Band 4 -. 03 -. 20 .30 1.62 .18 .02 -. 14 

Band 1 -. 17 -. 36 -. 09 .18 1.12 -. 01 .09 

Bend 6 .02 .23 .31 .02 -. 01 .67 -. 24 

Band 7 .05 -4.18 -2.47 -. 14 .09 -. 24 3.57 

D-18

Class 6, Date 7; 15 OCT 86 Number of points in sample 475 

Band 1 Band 2 Band 3 Bind 4 Band 5 Band 6 Band 7 

Min 29.00 151.00 130.00 95.00 158.00 157.00 97.00 

Mean 32.05 154.14 138.69 99.49 160.63 159.66 99.03 

Max 42.00 159.00 142.00 103.00 164.00 162.00 101.00 


Band 1 2.54 1.06 -1.49 -. 31 -. 16 -. 02 .53 

Band 2 1.06 1.69 -.04 -. 24 - .10 -. 07 .20 

Band 3 -1.49 -. 04 2.57 .33 09 -. 03 -. 44 

Band 4 -. 31 -. 24 .33 1.22 -. 02 .06 -.13 

Band 5 -. 16 - .10 .09 -. 02 1.01 -. 04 -. 02 

Band 6 -. 02 -. 07 -.03 .06 -. 04 .69 -. 04 

Band 7 .53 .20 -. 44 -. 13 -. 02 -. 04 .62 



Min 21.00 143.00 136.00 88.00 158.00 141.00 78.00 

Mean 23.51 148.73 138.-2 91.85 160.43 159.20 78,95 

Max 27.00 152.00 143.00 95.00 163.00 162.00 81.00 


Band 1 .67 .18 -. 19 -. 04 .0: -. 16 .03 

Band 2 .18 1.00 -,32 -. 08 .09 .38 02 

Band 3 -. 19 -. 32 .88 .05 -.03 -. 37 -.05 

Band 4 -. 04 -. 08 .05 1.16 -.11 .22 -.06 

Band 5 .02 .09 -. 03 -.11 .62 .01 .01 

Band 6 -. 16 .38 -. 37 .22 .01 2.38 .00 

Band 7 .03 .02 -. 05 -. 06 .01 .00 .24 



Min 51.00 147.00 123.00 113.00 157.00 158.00 124.00 

Mean 54.75 151.17 130.56 115.37 159.68 159.65 125.47 

Max 61.00 154.00 136.00 118.00 161.00 161.00 131.00 


Band 1 3.00 .09 -2.87 .22 .05 .18 1.16 

Band 2 .09 1.99 .37 .01 .17 -. 43 .21 

Band 3 -2.87 .37 5.95 .12 -. 21 -. 21 - .70 

Band 4 .22 .01 .12 1.23 .06 -. 05 -. 01 

Band 5 .05 .17 -. 21 .06 .75 - .09 - .11 

Band 6 .18 -. 43 -. 21 -. 05 -. 09 .47 .19 

Band 7 1.16 .21 -. 70 -. 01 -.11 .19 3.14 

D- 19



Min 47.00 143.00 143.00 114.00 158.00 157.00 117.00 

Mean 49.52 147.73 146.81 117.94 160.93 159.17 119.07 

Max 54.00 153.00 151.00 121.00 163.00 161.00 121.00 

Covar P.4*d 1 Band 2 Band 3 Band 4 Band 5 Band 6 Band 7 

Band 1 .89 .66 -. 36 -. 07 .04 .16 .14 

Band 2 .66 2.71 .43 -. 29 -. 08 -. 03 -. 34 

Band 3 -.36 .43 1.82 .12 -. 14 .01 -. 15 

Band 4 -.07 -. 29 .12 1.03 .00 .00 -. 03 

Band 5 .04 -. 08 -. 14 .00 .90 -. 05 .03 

Band 6 .16 -. 03 .01 .00 -. 05 .52 .10 

Band 7 .14 -. 34 -. 15 -. 03 .03 .10 .90 



Min 48.00 153.00 126.00 107.00 154.00 158.00 129.00 

Mean 50.45 159.47 133.48 110.87 158.31 159.78 134.19 

Max 56.00 171.00 141.00 116.00 161.00 162.00 140.00 


Band 1 1.54 1.50 -. 49 .20 -. 08 .10 .13 

Band 2 1.50 11.28 4.48 -. 48 -. 18 -. 15 -3.76 

Band 3 -. 49 4.48 4.76 .20 .15 -. 04 -2.16 

Band 4 .20 -. 48 .20 1.73 .08 .05 .19 

Band 5 -. 08 -. 18 .15 .08 1.27 -.01 -.11 

Band 6 .10 -. 15 -. 04 .05 -. 01 .48 .09 

Band 7 .13 -3.76 -2.16 .19 -.11 .09 3.86 



Min 63.00 169.00 134.00 117.00 155.00 158.00 135.00 

Mean 67.95 177.39 138.28 119.17 158.59 159.74 137.26 

Max 73.00 186.00 142.00 122.00 161.00 162.00 139.00 


Band 1 4.42 6.75 -. 32 -. 20 -. 54 .19 -. 13 

Band 2 6.75 14.60 1.02 -. 88 -1.48 .24 -. 72 

Band 3 -- 32 1.02 2.27 .15 -. 32 .15 03 

Band 4 -. 20 -. 88 .15 .93 .09 -. 04 .16 

Band 5 -. 54 -1.48 -. 32 .09 1.27 .00 .18 

Band 6 .19 .24 .15 -. 04 .00 .68 .02 

Band 7 -. 13 -. 72 .03 .16 .18 .02 .40 

D-20



-----------------.--------.--------.--------- -------------- 

Min 49.00 126.00 147.00 120.00 157.00 159.00 88.00 

Mean 53.65 131.71 158.93 126.55 159,76 161.13 89,21 

Max 59.00 137.00 169.00 132.00 162.00 163.00 91.00 


------------.--------.----------------.--------.-------- 

Band 1 7.35 -5.55 8.06 5,49 -. 86 .35 -. 22 

Band 2 -5.55 6.98 -7.57 -5.49 .37 -. 46 .26 

Band 3 8.06 -7.57 12.58 7.25 -1.13 .62 -. 26 

Band 4 5.49 -5.49 7.25 6.79 -. 56 .47 -. 22 

Band 5 -. 86 .37 -1.13 -. 56 .76 -. 03 .04 

Band 6 .35 -. 46 .62 .47 -. 03 .52 -. 03 

Band 7 -. 22 ,26 -. 26 -. 22 .04 -,03 .27 



Min 56.00 143.00 119.00 117.00 158.00 158.00 126.00 

Mean 61.39 147.63 130.09 120.01 160.72 160.01 127.68 

Max 67.00 152.00 139.00 124.00 163.00 161.00 129.00 


------------- -------- -------- -------- -------- -------- -------- 

Band 1 4.82 1.16 -5.01 -. 65 .09 -. 06 .14 

Band 2 1.16 5.25 1.71 -. 89 .41 -. 31 .42 

Band 3 -5.01 1.71 10.11 .77 -. 15 -. 16 .26 

Band 4 -. 65 -. 89 .77 1.32 .02 .18 -. 14 

Band 5 .09 .41 -. 15 .02 1.01 .05 -. 09 

Band 6 -. 06 -. 31 -. 16 .18 .05 .41 -. 14 

Band 7 .14 .42 .26 - .14 -. 09 -. 14 .76 



------------.--------.----------------.--------.-------- 

Min 57.00 151.00 121.00 116.00 156.00 158.00 131.00 

Mean 62.92 160.24 131.91 118.81 159.69 159.92 132.87 

Max 69.00 172.00 141.00 122.00 162,00 162.00 136.00 


Band 1 5.59 7.11 -1.56 -. 51 -. 63 -. 50 .30 

Band 2 7.11 22.16 9.40 -. 30 -. 81 -1.68 -. 69 

Band 3 -1.56 9.40 13.43 .51 -. 18 -. 76 -,96 

Band 4 -. 51 -. 30 .51 1.01 .12 .14 -. 22 

Band 5 -. 63 -. 81 -. 18 .12 1.26 -. 09 -.11 

Band 6 -. 50 -1.68 -. 76 .14 -. 09 .56 .01 

Band 7 .30 -. 69 -. 96 -. 22 -.11 .01 .96 

D-21

Class 8, Date 5; 24 MAY 86 Numbe:- of points in sample 135 


Min 60.00 162.00 135.00 117.00 157.00 158.00 134.00 

Mean 67.19 172.80 139.22 120.01 159.32 159.76 135,54 

Max 73.00 183.00 143.00 123.00 162.00 161.00 137.00 


Band 1 5.46 9.44 .73 .82 -. 65 .37 .18 

Band 2 9.44 20.93 3.54 1.45 -1.74 .62 .59 

Band 3 .73 3.54 2.87 .53 -. 52 .17 .21 

Band 4 .82 1.45 .53 1.61 -. 24 .17 .08 

Band 5 -. 65 -1.74 -. 52 -. 24 .94 -. 03 -.11 

Band 6 .37 .62 .17 .17 -. 03 .50 .04 

Band 7 .18 .59 .21 .08 -.11 .04 .54 



Min 63.00 156.00 139.00 122.00 157.00 158.00 136.00 

Mean 67.46 167.14 142.46 125.16 159.56 160.36 137.97 

Max 72.00 174.00 147.00 129.00 162.00 162.00 143.00 


Band 1 3.83 5.87 -. 22 .49 -. 48 .04 .14 

Band 2 5.87 11.65 .54 .45 -. 90 .26 -.11 

Band 3 -. 22 .54 2.36 .16 -. 30 .20 -. 03 

Band 4 .49 .45 .16 1.63 -. 15 .24 .06 

Band 5 -. 48 -. 90 -. 30 -. 15 3.04 -. 03 .04 

Band 6 .04 .26 .20 .24 -. 33 .68 -. 01 

Band 7 .14 -.11 -. 03 .06 .04 -. 01 .95 



Min 33.00 151.00 126.00 94.00 157.00 159.00 102.00 

Mean 38.20 154.40 134.71 97.36 161.30 160.08 102.42 

Max 45.00 158.00 138.00 101.00 163.00 162.00 103.00 


Band 1 4.72 1.40 -2.28 -1.05 .35 .29 -. 12 

Band 2 1.40 2.07 .29 -. 44 -. 27 -,02 .06 

Band 3 -2.28 .29 4.18 .68 -. 62 -. 21 .25 

Band 4 -1.05 -. 44 .68 1.28 -.19 -. 10 .07 

Band 5 .35 -. 27 -. 62 -. 19 .97 -. 02 -. 08 

Band 6 .29 -. 02 -. 21 -. 10 -.02 .39 -. 02 

Band 7 -. 12 .06 .25 .07 -.08 -. 02 .24 

D-22



Min 23.00 145.00 134.00 90.00 158.00 158.00 77.00 

Mean 24.68 147.41 137.64 92.24 160.64 160.04 77.73 

Max 27.00 150.00 140.00 95.00 162.00 161.00 78.00 


Band 1 .51 .02 -. 37 .06 .02 .06 -. 01 

Band 2 .02 .74 -. 09 -. 06 - .01 -. 08 .09 

Band 3 -. 37 -. 09 1.38 -. 03 .17 .08 .04 

Band 4 .06 -. 06 -. J3 1.20 .03 -. 01 .00 

Band 5 .02 -. 01 A17 .03 .68 .01 -. 05 

Band 6 .06 -. 08 .08 -. 01 .01 .35 .01 

Band 7 -. 01 .09 .04 .00 -. 05 .01 .20 



Min 53.00 145.00 116.00 112.00 156.00 158.00 128.00 

Mean 58.74 150.18 124.34 114.59 159.23 159.90 130.07 

Max 65.00 157.00 132.00 118.00 162.00 161.00 133.00 


Band 1 6.32 2.59 -6.56 .02 -. 71 .20 -. 29 

Band 2 2.59 7.78 2.23 -. 03 .00 -. 19 -. 73 

Band 3 -6.56 2.23 13.48 .22 1.05 -. 09 .25 

Band 4 .02 -. 03 .22 1.05 .01 .02 -. 18 

Band 5 -. 71 .00 1.05 .01 1.06 .00 .12 

Band 6 .20 -. 19 -. 09 .02 .00 .58 .03 

Band 7 -. 29 -. 73 .25 -. 18 .12 .03 1.22 



Min 51.00 151.00 139.00 114.00 158.00 158.00 120.00 

Mean 54.83 156.87 143.07 117.13 160.37 159.62 120.24 

Max 58.00 162.00 145.00 119.00 163.00 161.00 121.00 


Band 1 2.14 2.37 -1.12 -. 25 .11 -.13 .07 

Band 2 2.3- 4.11 -. 84 -. 49 -. 01 -. 19 .01 

Band 3 -1.12 -. 84 2.38 .29 -. 35 .13 -. 01 

Band 4 -. 25 -. 49 .29 1.03 -. 02 .06 -. 06 

Band 5 .11 -. 01 -. 35 -. 02 .75 .05 .00 

Band 6 -. 13 -. 19 .13 .06 .05 .47 .00 

Band 7 .07 .01 -. 01 -.06 .00 .00 .19 

D-23



Min 56.00 170.00 129.00 107.00 148.00 158.00 130.00 

Mean 60.39 177.44 134.47 110.21 156.41 159.75 131.67 

Max 65.00 187.00 149.00 114.00 160.00 161.00 135.00 


Band 1 5.39 8.15 .61 .24 -1.17 .13 .48 

Band 2 8.15 17.00 4.27 -. 64 -2.58 .20 1.20 

Band 3 .61 4.27 7,07 .15 -2.22 .27 .86 

Band 4 .24 -. 64 ,15 1.95 -. 12 .10 -. 13 

Band 6 -1.17 -2.58 -2.22 -. 12 2.95 -. 15 -. 26 

Band 6 .13 .20 .27 .10 -. 15 .49 .19 

Band 7 .48 1.20 .86 -. 13 -. 26 .19 1.49 

D-24



ATTN: DEPT OF RADIATION BIOCHEMISTRY 

ATTN: BHS 


ATTN: EXH 

ATTN: MRA 

ATTN: PHY M WHITNALL 

ATTN: RSD 

ATTN: SCIENTIFIC DIRECTOR 




ATTN: NUC FORCES & ARMS CONTROL PLCY 



ATTN: MIL APPL C FIELD 


DNA-TR-92-37-V1 

OASD 

ATTN: DUSP/P 

ATTN: USD/P 

OFC OF MILITARY PERFORMANCE 

ASSESSMENT TECHNOLOGY 

2 CY ATTN: F HEGGE 

STRATEGIC & SPACE SYSTEMS 


U S EUROPEAN COMMAND/ECJ-6-DT 

ATTN: ECJ-6 

U S EUROPEAN COMMAND/ECJ2-T 

ATTN: ECJ-3 

ATTN: ECJ2-T 

ATTN: ECJ5-N 

ATTN: ECJ5N 

USSSTRATCOMIJ531T 

DEFENSE INTELLIGENCE AGENCY ATTN: J-521 

ATTN: DB 

ATTN: JPEP 

5 CY ATTN: DB-4 RSCH RESOURCES DIV 3416TH TTS INTERSERVICE NUC WPNS SCHOOL 

ATTN: DB-5C 

ATTN: TTV 

ATTN: DB-6B 

2 CY ATTN: TTV 3416TH TTSO 

ATTN: DB-6E 

ATTN: DiW-4 


ATTN: DN 

ATTN: DT 

ARMY RESEARCH LABORATORIES 



ATTN: OS 





ATTN: DIC/2C 


ATTN: DAMO-SWN 


ATTN: DAMO-ZXA 

ATTN: DFRA JOAN MA PIERRE 

ATTN: NANF 

NUCLEAR EFFECTS DIVISION 

ATTN: NASF 

ATTN: STEWS-NE-T 

10CYATTN: RAEM 

2 CY ATTN: TITL TEXCOM 

ATTN: D PACE 

DEFENSE SYSTEMS SUPPORT ORGANIZATION 

ATTN: JNGO 

U S ARMY AIR DEFENSE ARTILLERY SCHOOL 



2 CY ATTN: DTIC/OC U S ARMY ARMOR SCHOOL 



ATTN: FCPR 

ATTN: FCPRT 


ATTN: FCTO 


ATTN: ICAF TECH LIB 


ATTN: LIBRARY 




Dist-1 


U S ARMY BALLISTIC RESEARCH LAB 

ATTN: AMSRL-SL-BS DR KLOPCIC 

ATTN: AMSRL-SL-BS E DAVIS 

ATTN: SLCBR-D 

ATTN: SLCBR-DD-T 

ATTN: SLCBR-TB 

U S ARMY CHEMICAL SCHOOL 

ATTN: CRAL FRAKER 

U S ARMY COMBAT SYSTEMS TEST ACTIVITY 

ATTN: JOHN GERDES 

ATTN: MIKE STANKA

DNA-TR-92-37-V1 (DL CONTINUED) 

U S ARMY COMD & GENERAL STAFF COLLEGE 










U S ARMY FIELD ARTILLERY SCHOOL ATTN: CODE 1240 

ATTN: ATSF-CD ATTN: CODE 5227 


ATTN: AF-OPTS 

U S ARMY FOREIGN SCIENCE & TECH CTR 

ATTN: C WARD 


ATTN: PMS-423 

ATTN: SEA-07R 



U S ARMY INFANTRY CENTER 

ATTN: G RIEL 



U S ARMY ITAC 


ATTN: IAX-Z 


U S ARMY LABORATORY COMMAND ATTN: CODE E-1 1 


ATTN: CTR FOR NAV WARFARE STUDIES 



ATTN: LIBRARY 


ATTN: STRATEGY DEPT 

4 CY ATTN: MONA-NU DR D BASH 

ATTN: MONA-NU MAJ BLISS 

NAWCWPNSDIV DETACHMENT 

U S ARMY TEST & EVALUATION COMMAND 

ATTN: STECS-NE 

U S ARMY VULNERABILITY/LETHALITY 

ATTN: AMSLC-VL-NE DR J FEENEY 


ATTN: LIBRARY 

ATTN: STRATEGIC STUDIES INSTITUTE 

U S MILITARY ACADEMY 

ATTN: DEPT BEHAVORIAL SCI & LEADERSHIP 

ATTN: COL J G CAMPBELL 

ATTN: SCIENCE RESEARCH LAB 


NUCLEAR WEAPONS TNG GROUP ATLANTIC 


ATN: DOCUMENT CONTROL 

NUCLEAR WEAPONS TNG GROUP PACIFIC 

ATTN: CODE 32 

OPERATIONAL TEST & EVALUATION FORCE 


PLANS. POLICY & OPERATIONS 

ATTN: CODE-P 



STRATEGY AND POLICY DIVISION 

ATTN: DRXSY-DS ATTN: N 51 

US ARMY MEDICAL RESEARCH & 

ATTN: NIS-22 

ATTN: NOP 50 

DEVELOPMENT COMMAND ATTN; NOP 603 

ATTN: SGRD-PLE ATTN. NOP 91 

US ARMY MODEL IMPROVEMENT STUDY 

MANAGEMENT AGENCY 

ATTN: E VISCO 

ATTN: NUC AFFAIRS & IN5L NEGOT BR 

ATTN: N455 


WALTER REED ARMY MEDICAL CENTER 

ATTN: G RIPPLE 


MARINE CORPS 


ATTN: PSI G RASP 


ATTN: CODE 9642-B 

AFIS/INT 

ATTN: INT 




ATTN: AUL-LSE 

ATTN: LIBRARY 

ARMSTRONG LABORATORY 

ATTN: AL/CFBE DR F S KNOX 

Dist-2



ATTN: AFSAAISAMI 

ARI 

ATTN: JOHN LOCKHART 

DEPUTY CHIEF OF STAFF FOR PLANS & OPERS 

ATTN: AFXOOSS 

HO ACC/XP-JSG 

ACrTN: ACCXP-JSG 

EAI CORPORATION 

ATTN: MILLARD MERSHON 

ATTN: DENNIS METZ 

ATTN: KRISTIN GAVLINSKI 

HQ USAFA/DFSELD 

FULTON FINDINGS 

ATTN: LIBRARY 

ATTN: WES FULTON 

HORIZONS TECHNOLOGY INC 

NATIONAL AIR INTELLIGENCE CENTER ATTNS F GREY 

ATTN: 

CCN 

ATTN: SDA 

HORIZONS TECHNOLOGY, INC 

ATT'IN: J MARSHALL-MIES 

USAF NUCLEAR CRITERIA GROUP SERETARIAT 

ATTN: AFWL/NTN R SUDER 

INSTITUTE FOR DEFENSE ANALYSES 

ATTN: DOUGLAS 

DEPARTMENT 

SCHULTZ 

OF ENERGY 

ATTN: J ORLANSKY 


ATTN: DR T JONES 

DPEARTMENT OF ENERGY 

ATTN: B SANTORO 

ATTN: G KERR 

ATTN: J WHITE 

ATTN: W RHOADES 


ATTN: A WARSHAWSKY 



ATTN: D STROTTMAN 



ATTN: TECH LIB 3141 



ATTN: COUNTER-TERRORIST GROUP 



ATTN: NIO-T 

ATTN: N10 STRATEGIC SYS 



ATTN: NP-CP 

ATTN: OFC OF CIVIL DEFENSE 

U S DEPARTMENT OF STATE 

ATTN: PM/STM 

U S NUCLEAR REGULATORY COMMISSION 


ATTN: S YANIV 


ARES CORP 

2 CY ATTN: A DEVERILL 

JAYCOR 

ATTN: M FINEBERG 

ATTN: CYRUS P KNOWLES 


ATTN: DASIAC 


AT-IN: R STOHLER 


ATTN: DASIAC 

LOCKHEED MISSILES & SPACE CO, INC 

ATTN: WE-YOUNG WOO 

LOGICON R & D ASSOCIATES 


ATTN: DOUGLAS C YOON 


ATTN: S WOODFORD 

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 

ATTN: DUNCAN C MILLTER 

MICRO ANALYSIS AND DESIGN 

ATTN: N LAVINE 

ATTN: R LAUGHERY 

ATTN: TROTH 


2 CY ATTN: F W WHICKER 

2 CY ATTN: G ANNO 

20 CY ATTN: G E MCCLELLAN 

ATTN: H BRODE 


ATTN: D GORMLEY 

2 CY ATTN: G MCCLELLAN 

PSYCHOLOGICAL RESEARCH CENTER 

ATTN: L GAMACHE 

Dist-3

DNA-TR-92-37-VI (DL CONTINUED) 


ATTN:D KAUL 

ATTN: E 

L 

SWiCK 

HUNT 


AT'ITN: W WOOLSON 


ATTN: JMCGAHAN 


ATTN: P VERSTEEGEN 




TECHNICO SOUTHWEST INC 

ATTN: S LEVIN 

TRW INC 

ATTN: TIC 

TRW OGDEN ENGINEERING OPERATIONS 

ATTN: D C RICH 

TRWSIG 

ATTN- DR BRUCE WILSON 

UNIVERSITY OF CINCINNATI MEDICAL CENTER 

ATTN: E SILBERSTEIN 

DIRECTORY OF OTHER 

UNIV OF ILLINOIS-WILLIARD AIRPORT 

ATTN: H L TAYLOR 

Dist-4



AD-A259 085 

S.... IiilII|IlH~l D.A-TR-92-3,,.v2 


Volume 2-Conifer Stress Near Chernobyl Derived 

from Landsat Imagery 

Gene E. McClellan 

Terrence H. Hemmer 

Ronald N. DeWitt 




December 1992 ST EI 


SELEC f 

S DF'C2 2 1992 II 

E 


Approved for public release; 

distribution ti unlimited. 

92-32436 

I!IIII4lU 9 9 .0 f.j, .• - !








ORGANIZATION. 

-4a Zi

-DISCLA]IIR NOTICE 

THIS DOCUMENT IS BEST 

QUALITY AVAILABLE. THE COPY 

FURNISHED TO DTIC CONTAINED 

A SIGNIFICANT NUMBER OF 

COLOR PAGES WHICH DO NOT 

"REPRODUCE 

LEOIBLY ON BLACK 

AND WHITE MICROFICHE.


This mailer is provided to enable DNA to maintain current distribution lists for reports. (We would 


NOTE: 

o- Add the individual listed to your distribution list. Please return the mailing label from 

the document so that any additions, 

0 Delete the cited organization/individual. changes, corrections or deletions can 

be made easily. 

O3 Change of address. 

NAME- 

ORGANIZATION: 

OLD ADDRESS 


Z 

TELEPHONE NUMBER: ( ) 

I-I 

WU 

w DNA PUBUCATION NUMBER/TITLE CHANGES/DELETIONSIADDITIONS, etc.) 

as 

Z 

,< 

uI 

UJ 

I-I 

(Atafh ShtM if mmr. Space is Required) 

* DNA OR OTHER GOVERNMENT CONTRACT NUMBER: 


SPONSORING ORGANIZATION: 


SIGNATURE:


ATTN: TITL 




ATTtN- TITL 



REPORT DOCUMENTATION PAGE 

OBN.00-I 

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. 

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 

colectlion of formabon. inctuding suggestions for reducing this burden. to Washington HeedQuarters Services Directorate for intormation O oeration5 and Reports. 1215 Jefferson 

Davis Highway, Suite 1204. Arlington. VA 22202-4302. and to the Office at Management and Budget. Paperwort Reduction Protect (0704-0188}. Washington. DC 20503 


1 921201 Technical 870929 - 920228 


Chernobyl Doses C - DNA 001-87-C-0104 

Volume 2- Conifer Stress Near Chernobyl Derived from Landsat Imagery PE - 62715H 

6. AUTHOR(S) PR - RM 

Gene E. McClellan, Terrence H. Hemmer, and Ronald N. DeWitt 

TA -RH 

WU - DH026130 


REPORT NUMBER 


12340 Santa Monica Boulevard PSR Report 2250 





6801 Telegraph Road DNA-TR-92-37-V2 


RARP/Young 



STRP 3500A 25904D. 




This volume presents Landsat Thematic Mapper imagery of the area surrounding the Chernobyl Nuclear Reactor 

Station and derives quantitative estimates of the spatial extent and time progression of stress on coniferous forests 

resulting from the 26 April 1986 reactor explosion and release of radioactive material. Change detection between 

pre- and postaccident images demonstrates convincingly that remote sensing of the spectral reflectance of 

coniferous forests in visible and infrared wavelengths at moderate spatial resolution (30 meters) will detect the 

effects of large radiation doses to the forest canopy. 

This work was initiated at a time when the expectation for direct data from the Soviet Union on local, accidentinduced 

radiation levels was limited and the satellite data provided an alternative source. Although information 

exchange with the former Soviet Union has improved dramatically, the results of this report are important, since 

they prove the feasibility of large-scale, spectral response measurements on radiation-exposed pine trees in a natural 

environment. Volume I presents the derivation of radiation doses from the imagery reviewed in this volume, 

describes changes in spectral reflectivity of the affected trees as a function of dose and time, and discusses the military 

operational implications of these results. 


Chernobyl Forest Damage Landsat 80 

Change Detection Conifer Stress Fallout 16. PRICE CODE 

Ionizing Radiation Multispectral Imagery 

17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19, SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT 



NSN 7540-20-5= 

Standrcd Form 296 (Rev.2-9) 

i Preecnbed by, ANSI SM. 239-18 

255- 02

UNCLASSIFIED 

SECURITY CLASSIFICATION OF THIS PA(rE 





UN•cLASSIFIED

PREFACE 

This volume is the second in a series of three volumes composing the final report to the 

Defense Nuclear Agency (DNA) for contract DNAOO1-87-C-0104, Chernobyl Doses. This 

document was prepared by investigators at Pacific-Sierra Research Corporation (PSR) as a topical 

report for that contract but is being published as a volume of the final report. It decribes the 

acquisition and processing of Landsat imagery of the area containing the Chernobyl Nuclear 

Reactor Station and presents the exploratory analysis of the imagery using PSR's proprietary 

Hyperscoutr' change detection algorithm. Volume 1, Analysis of Forest Canopy Radiation 

Response from Multispectral Imagery and the Relationship to Doses, presents the analytical work 

that connects these multispectral observations of pine forests in the images to the nuclear radiation 

dose received by the treas as a consequence of the reactor accident of 26 April 1986. Volume 3, 

Habitat and Vegetation Near the Chernobyl Nuclear Reactor Station, presents a detailed exposition 

on the soil, climate, and vegetation of the Poles'ye region of the Ukraine and Belorussia with 

emphasis on the area around Chernobyl. 

The authors wish to acknowledge Frank Thomas and George Anno of PSR, who recognized 

the potential for remote sensing of radiation-damaged foliage around Chernobyl; Wayne Hallada 

and the late Quentin Wilkes of PSR, who arranged the necessary equipment and image 

acquisitions; and finally, the skillful manuscript preparation by Kathy Howell and Sunny Wiard. 

The authors wish to acknowledge the technical monitor of this project, Robert W. Young of 

DNA's Radiation Policy Division, for his support and encouragement during this work. 

Dr. Young was assisted first by MAJ Bruce West and then by MAJ Robert Kehlet. The authors 

also wish to acknowledge Dr. Marvin Atkins and Dr. David Auton of DNA whose interest made 

this work possible. 

DTIC QUALITY 1 i-Y'YD 2 

Accesion For 

NTIS CRA&I 

DTIC TAB 

Unannounced [ 

Justification 

By ......................................... 

Distribution I 

Dist 

Availability Codes 

IAvail and/or 

Special 

ThEHypersou is a regisweid trdemark of Pacific-Sierra Reserc Caupmuon. 

SI . ,



To Convert From To MulUply 

angstom meters (m) 1.000 000 X K-10 

atmosphere (normal) kilo pascal (kPa) 1.013 25 X E+2 

bar kilo pascal (kPa) 1.000 000 X E+2 

barn meter 2 (M 2 ) 1.000 000 X E-28 

British Thermal unit ( emicl) joule (JW 1.054 350 X E+3 

calorie (thermochemicsl) joule (JW 4.184 000 

cal (thermoWhe cal)/,ml mep joule//m2(MJ/m2) 4.184 000 X E-2 

curie g"ig becquerel (GBqr 3.700 000 X E+I 


degmre Fahrenheit degree kelvin (K) tK=(tof + 459.67)11.8 

electron volt joule (J) 1.602 19 X E-19 

erg Joule (J) 1.000 000 X E-7 

erg/second watt MW) 1.000 000 X E-7 

foot meter (m) 3.048 000 X E-I 

foot-pound-force joule (J) 1.355 818 

gallon (U.S. liquid) meter 3 (m3) 3.785 412 X E-3 

inch meter (m) 2.540 000 X E-2 

jerk joule (J) 1.000 000 X E+9 

joule/kilogram tJ/Kg) (radiation dose 

absorbed) Gray (Gy) 1.000 000 


kip (1000 Ibf) newton (N) 4.448 222 X E+3 

kip/tnch2 (ksl) kilo pascal (kPa) 6.894 757 X E+3 

ktap newton-second/mr (N-aim2) 1.000 000 X E+2 

micron meter (mW 1.000 000 X E-6 

mil meter (mW 2.540 000 X E-5 

mile (international) meter (m) 1.609 344 X E+3 

ounce kilogram (kg) 2.834 952 X E-2 

pound-force (lbf avoirdupois) newton (N) 4.448 222 

pound-force inch newton-meter (N-m) 1.129 848 X E-I 

pound-force/Inch newton/meter (N/m) 1.751 268 X E+2 

pound-force/foot' kilo pascal (kPa) 4.768 026 X E-2 

pound-force/Inch2 (psi) kilo pascal (kPa) 6.894 757 

pound-mass (Ibm avoirdupois) kilogram (kg) 4.535 924 X E-I 

pound-mass-foot2 (moment of inertia) kilogram-meter' (1g.m') 4.214 01! X E-2 

pound-mass/foot3 kilogram/meter' (kg/lnm) 1.601 846 X E+ I 

red (radiation dose absorbed) Gray (Gy)" 1.000 000 X E-2 

roentgen coulomb/kilogram (C/kg) 2.579 760 X E-4 

shake second Is) 1.000 000 X E-8 

slug kilogram (kg) 1.459 390 X E+1 

torn Imm 14ft OC) kilo pascal (kPa| 1.333 22 X E-I 

"The becquerel (Dq) Is the St unit of radioactivity: Bp a I event/s. 

"•The Gray (Gy) ti the S8 unit of absorbed radiation. 

iv


Section 

Page 

PREFACE ................................................................................ iii 

CONVERSION TABLE ............................................................... iv 

FIGURES ............................................................................... vi 

TABLES .................................................................................. vii 

1 INTRODUCTION ...................................................................... 

1.1 Background .................................................................... 

1 

1 

1.2 Organization of report ....................................................... 2 

2 THE PHYSICAL BASIS FOR THE SPECTRAL DETECTION OF 

VEGETATION STRESS ............................................................. 3 

3 QUANTITATIVE ANALYSIS OF MULTIDATE IMAGERY .................... 6 

3.1 Factors affecting the utility of multidate imagery .......................... 6 

3.2 Selection of an algorithm ..................................................... 7 

4 THEMATIC MAPPER IMAGERY ................................................. 9 

4.1 Introduction to the thematic mapper ..................................... 9 

4.2 Motivation for use of themaic mapper imagery ........................... 11 

4.3 Image selection and procurement ........................................ 11 

4.4 Image preparation ............................................................ 16 

5 IMAGE ANALYSIS ........................................ 19 

5.1 Image partitioning ............................................................ 19 

5.2 Mapping stress ................................................................ 25 

6 GROUND TRUTH ................................................................... 53 

7 DISCUSSION .......................................................................... 58 

8 CONCLUSION ........................................................................ 64 

9 REFERENCES ........................................................................ 66

FIGURES 

Figure 

Page 

4-1 Vicinity of the Chernobyl Nuclear Reactor Station on 8 May 1986 .......... 18 

5-1 Partitioning of 6 June 1985 image into classes derived from training 

sites 3 (yellow) and 4 (green) .................................................... 24 

5-2 Date: 6 June 1985, 1 year preaccident ....................................... 27 

5-3 Date: 21 March 1986, 2 months preaccident ............................... 28 

5-4 Date: 29 April 1986, 3 days postaccident .................................. 29 

5-5 Date: 29 April 1986, 3 days postaccident, forest stress map, colored 

according to indicated scale. Gray-scale background consists of 

nonforested areas ................................................................. 31 

5-6 Date: 8 May 1986, 12 days postaccident ................................... 32 

5-7 Date: 8 May 1986, 12 days postaccident, forest stress map .............. 33 

5-8 Date: 24 May 1986, 4 weeks postaccident ................................. 35 

5-9 Date: 24 May 1986, 4 weeks postaccident, forest stress map ............ 36 

5-10 Date: 31 May 1986, 5 weeks postaccident ................................... 37 

5-11 Date: 31 May 1986, 5 weeks postaccident, forest stress map ............ 38 

5-12 Date: 15 October 1986, 5.6 months postaccident ......................... 39 

5-13 Date: 15 October 1986, 5.6 months postaccident, forest stress map ...... 40 

5-14 Date: 2 December 1986, 7.2 months postaccident ........................ 41 

5-15 Date: 2 December 1986, 7.2 months postaccident, forest stress map..... 42 

5-16 Date: 11 May 1987, 1 year postaccident .................................... 44 

5-17 Soviet-supplied gamma dose rate contours of 1 May 1987 overlaid 

on the I IM ay 1987 ................................................................ 45 

5-18a Date: 11 May 1987, 1 year postaccident, forest stress map .............. 46 

5-18b Date: 11 May 1987, forest stress map with gamma dose rate contours 

of 1 M ay 1987 ..................................................................... 47 

5-19 Date: 7 September 1987, 16 months postaccident ........................... 48 

5-20 Date: 7 September 1986, 16 months postaccident, forest stress map. 49 

5-21 Date: 28 May 1988, 25 months postaccident ............................. 51 

5-22 Date: 28 May 1988, 25 months postaccident, forest stress map ......... 52 

7-1 Date: 11 May 1987, enhanced image ....................................... 61 

7-2 Boundaries of areas classified as similar to training site 4 overlaid on the 

11 May 1987 enhanced image ................................................... 62 

7-3 Date: 6 June 1985, preaccident image with the same image 

enhancement transformation shown in Figure 7-1 ............................. 63 

vi

TABLES 

Table 

Page 

2-1 Estimated short-term radiation exposures required to damage 

various plant communities .......................................................... 3 

4-1 Landsat TM spectral bands ........................................................ 9 

4-2 Landsat scene quality and cloud contamination analysis for scenes 

with less than 30 percent cloud cover ........................................ 12 

4-3 Scene procurement recommendations ........................................... 14 

4-4 Landsat scene acquisitions ...................................................... 15 

4-5 Corner points of the acquired images ........................................ 15 

4-6 Corner points of the geocoded images after resampling ...................... 16 

4-7 Corner points (Zone 36 UTM coordinates) of the 512 x 512 pixel area 

analyzed in this report ............................................................ 17 

5-1 Training site modified transformed divergencies .............................. 23 

6-1 Second order warp coefficients for the gamma dose contours ............... 54 

6-2 Accuracy of the registration of GCPs ........................................... 55 

7-1 Comparison of present work with that of Goldman and coworkers ........... 59 

vii

SECTION 1 

INTRODUCTION 

On 26 April 1986, at 1:24 a.m. local time, the Chernobyl nuclear reactor number four blew up. 

This report describes the research of investigators at Pacific-Sierra Research Corporation (PSR) in 

studying the effects of radioactivity deposited in the immediate area by the accident. Specifically, it 

describes the quantitative analysis of the multispectral (MS) imagery of the area surrounding the 

Chernobyl nuclear reactor. 

1.1 BACKGROUND. 

Before the explosion only theoretical models suggested our ability to monitor remotely the 

effects caused by widely distributed large doses of radiation. The sensitivity of vegetation to large 

doses of radiation had been measured only in the lab and in small-scale field experiments. The 

Chernobyl accident provides an opportunity to observe both short- and long-term effects of high 

radiation dose on plants. The large spatial scale of the affected area allows us to evaluate the utility 

of remotely sensed multispectral imagery in quantifying the extent of damage to foliage and in 

estimating the radiation dose that was deposited by the accident. 

The accident occurred before the changes of Glasnost could be taken for granted. Indeed, 

because the initial tendency of Soviet authorities was to deny the accident, there was little basis for 

anticipating sufficient and accurate information from Soviet sources. Therefore, any independent 

source of information that could augment or verify the information released by Soviet authorities 

seemed desirable. In particular, some method for obtaining an independent assessment of dose 

was regarded to be of primary interest 

The most obvious method was to use remote sensors. Because overflights of Soviet territory 

were deemed impossible, the remote sensor would have to be in a low-altitude orbit. The remote 

measurement of radiation at this distance is prevented by atmospheric attenuation. On the other 

hand, the remote measurement of the effects of the radiation on vegetation was possible with the 

existing satellite sensor mix. Two of these sensors were the Landsat Thematic Mapper (TM) and 

SPOT maintained by the United States and France, respectively. Because both of these are 

commercial systems, data is readily available. In effect, vegetation provides an on-site biological 

dosimeter that can be read remotely with multispectral imagery. 

Such remote sensing of vegetation stress is an important scientific endeavor contributing to 

early detection of the effects of major disasters such as Chernobyl as well as the chronic effects of 

major pollution sources. Analysis of imagery of the Chernobyl accident provides a benchmark for 

this capability.

1.2 ORGANIZATION OF REPORT. 

Following this introduction, Section 2 presents the physical basis for the spectral detection of 

vegetation stress. Section 3 describes the usual approaches to the detection of stress. Section 4 

follows with a description of the sensor, imagery selection criteria, and preprocessing 

requirements. Section 5 presents the imagery, including detected indications of vegetation stress. 

Since limited ground truth (radiation measurements) was published by the Soviets, this ground 

truth can be compared to the detected stress. Section 6 presents this comparison. Section 7 

reiterates salient features and compares this work with other similar work. Section 8 contains 

concluding remarks and makes recommendations for continuation and improvements. 

2

SECTION 2 

THE PHYSICAL BASIS FOR THE SPECTRAL DETECTION OF 

VEGETATION STRESS 

We have asserted that dose estimates for the Chernobyl area can be obtained by monitoring the 

effects of the radiation on indigenous vegetation. In making this assertion, we assume that the 

accident sufficiently dosed the vegetation to invoke a response detectable in the spectral regions 

monitored by satellite sensors. This section provides the justification for these assumptions. 

We simplify the analysis by concentrating on only one type of plant community. Three criteria 

governed the selection of the type of plant community to be monitored. First, our desire to 

determine the radiation dose dictated a plant type relatively sensitive to radiation. Second, the 

selection of a plant species pervading the area of interest allows estimates of the areal extent of the 

radiation effects. Third, because the time scale of the manifestation of stress could not be 

anticipated, we desired a plant type subject to little seasonal variation. 

Coniferous forests meet all three requirements. Table 2-1 shows that coniferous forests are 

very sensitive to radiation. The extensive forests located throughout the area around the Chernobyl 

reactor complex comprise predominantly conifers, with common pine, Pinus sylvestris (see 

Volume Ill of this reportl) the primary species. Finally, conifers show relatively little seasonal 

spectral variation compared to deciduous trees. However, seasonal changes in illumination and 

variations in the canopy closure of intermixed deciduous species, as well as changes caused by the 

maturing of the predominant foliage, vary the observed reflectance spectrum somewhat from 

season to season. 

Table 2-1. Estimated short-term radiation exposures required to damage various plant communities 

(Whicker and Fraley, 1974). 

Exposures to produce (kR) 

Community type Minor effects Intermediate Severe effects 

effects 

Coniferous forest 0.1-1 1-2 >2 

Deciduous forest 1-5 5-10 >10 

Shrub 1-5 5-10 >10 

Tropical rain forest 4-10 10-40 >40 

Grassland 8-10 10-100 >100 

Moss-lichen 10-50 50-500 >500 

3

Before presenting the spectral manifestations of stress, we discuss the spectral characteristics 

of healthy vegetation, with emphasis on those areas of the spectrum that can be monitored by either 

Landsat or SPOT. In the visible region of the electromagnetic spectrum extending from 

wavelengths of 0.4 to 0.7 micrometers (Wim), chlorophyll absorption (Hoffer, 1978) dominates the 

reflectance spectra of healthy vegetation. Chlorophyll absorbs throughout this region, but it 

absorbs less in the green than in the blue or red. This preferential absorption causes the green 

color of healthy vegetation. 

The infrared (IR) region of the spectrum comprises several subregions: the near infrared (NIR); 

short-wave infrared (SWIR); mid-wave IR (MWIR); and thermal or long-wave IR (LWIR). The 

wavelength interval defining each of these spectral regions varies from author to author. This 

report defines the NIR region to extend from about 0.7 jim to 1.3 Aim. In this subregion, the 

spongy mesophyll tissue in the interior of the leaves causes the foliage to reflect strongly, typically 

as much as 45 to 50 percent of incident illumination. The SWIR spectral region, from 

approximately 1.5 jim through 2.5 jim, includes two absorption features of the leaf water content, 

reducing the foliage spectral reflectance. Neither Landsat nor SPOT are sensitive in the MWIR, 

which extends from about 3 to 5 jim. The LWIR includes electromagnetic radiation from 8 jim to 

12 am and beyond. Because at terrestrial temperatures most materials have spectral characteristics 

in the LWIR that tend to be emissive rather than reflective, this band is referred to as the thermal 

MR. Thus, in this region the temperature and the emissivity of the vegetation determines its spectral 

response. Under normal conditions, transpiration and efficient heat exchange keep the 

temperature of foliage close to air temperature (Estes, 1983; Weibelt and Henderson, 1976). 

The spectral manifestations of stress depend strongly on the cause of the stress (Estes, 1983). 

For example, moisture stress initially results in increased reflectance in the SWIR, because there is 

less water to absorb the infrared radiation (Hoffer, 1978). Desiccation of pine needles to 

approximately 48 percent of fresh weight reduces NIR reflectance dramatically, while SWIR 

reflectance continues to rise (Westman and Price, 1988). An expected lack of cooling by 

transpiration would allow stressed foliage to be warmed by sunlight to temperatures higher than 

those of healthy vegetation and would result in relatively higher emissions in the thermal HR. If the 

moisture stress impedes chlorophyll production, the reflectivity in the visible can also be expected 

to increase. The spectral properties of other pigments in the leaf may now become dominant, 

resulting in a yellowish or reddish color. 

Cellular damage results in a marked decrease in the reflectance in the NIR. Little or no change 

in the visible is expected unless the damage reduces chlorophyll content or causes the production or 

destruction of other pigments. 

Unfortunately, the specific spectral manifestations of ionizing radiation damage to foliage are 

not well known. However, one form of radiation-induced stress may be at least partially 

4

predictable. The spectral reflectance of new pine needle growth is significantly higher than that of 

old pine needle growth at all wavelengths (Wolfe and Zissis, 1978). If the radiation stress results 

in either reduced or accelerated growth of new needles (an effect expected to be highly dependent 

on dose), then this deviation may be detectable.

SECTION 3 

QUANTITATIVE ANALYSIS OF MULTIDATE IMAGERY 

The problem of identifying stress in imagery can be approached in two ways. The first 

approach assumes that stressed vegetation has unique spectral signatures, and that these spectral 

signatures are sufficiently different from the spectral signatures of healthy vegetation that a suitable 

spectral transformation can be applied to the image to render stressed vegetation readily apparent to 

the image analyst. For example, Johnson (Johnson, 1989) developed a phenomenologically based 

image-enhancement transformation designed to detect a specific spectral manifestation of stress. 

Section 7 discusses this transformation in more detail. 

The other approach to identifying stress in imagery attempts to exploit the additional 

information content of multidate imagery. Multidate algorithms generally fall into two major 

categories: change detection algorithms and dynamic-system algorithms. Change detection 

algorithms tend to ignore explicit spectral properties and concentrate only on differences from date 

to date, except insofar as the spectral properties are associated with particular kinds of ground 

covers that manifest the change. Spectral properties are occasionally used, but only at a later stage 

to characterize the detected change. Dynamic-system algorithms tend to retain most of the spectral 

information. 

3.1 FACTORS AFFECTING THE UTILITY OF MULTIDATE IMAGERY. 

An accurate assessment of the fallout from the Chernobyl nuclear reactor accident through its 

effect on coniferous forests can be made only if a number of rather stringent requirements are met. 

First, there must be a sensor capable of detecting and tracking the changes in the spectral properties 

of the vegetation. Second, all unrelated factors that affect the spectral properties of vegetation on 

an image-to-image basis must be normalized out so that direct multidate comparisons can be made. 

Third, the data must be very accurately registered. Fourth, data must be available at frequent 

enough intervals. 

Numerous extraneous factors can affect the comparison of two or more images. Some of these 

factors are environmental (e.g., haze, humidity, sun angle, and cloud cover). Others are 

instrumental (e.g., view angle, pointing accuracy, spectral sensitivities, and resolution). Even if 

the same sensor at the same relative position at the same time of day is used, the differences 

between images can be quite large. Either the images must be corrected for these differences, or an 

algorithm that is insensitive to these extraneous differences must be used. 

Because multidate algorithms use the spectral information for the same ground point on two or 

more dates, the images must be registered (geocoded). Geocoding processes the images from all 

6

dates so that the picture element (pixel) at the same position in each date's image corresponds to a 

common area on the ground. Registration accuracy is of fundamental importance in the analysis of 

multidate imagery. Variations in the satellite orbit, pointing accuracy, jitter, and numerous other 

reasons make exact registration impossible, but registration to within a fraction of a pixel width is 

often possible. 

Finally, multidate image analysis requires that images be collected at time intervals that are 

small compared to the time scale of the process being monitored. In the case of radiation stress, 

the local radiation dose determines that time scale and causes it to vary from one place to another 

according to the distribution of the radiation fallout. This variation imposes a requirement for 

frequent images to monitor the development of stress in higlidy contaminated areas, but less 

frequent imaging will suffice in the regions of lesser contamination. Because cloud cover often 

renders satellite data useless, data collections at time intervals smaller than the time scale of the 

stress development process must be attempted. 

3.2 SELECTION OF AN ALGORITHM. 

The approach first envisioned for this project involved tracking the spectral properties of large 

areas of forest as a function of time. This approach falls in the dynamic-system category. Large 

areas of the forest would be delineated by polygons for the analysis. We can fit multivariate 

normal distributions to the measurements taken on each date within each polygon in the following 

manner. For a given ground pixel, there is an intensity value measured in each band, and these 

intensities can be formed into an ordered set called the pixel vector. A polygon's mean ii tensity 

vector can be comlited by averaging the pixel vectors over all pixels within the polygon. Once the 

polygon's mean intensity vector has been determined, there can be computed for each pixel a 

deviation vector (representing a pixel intensity vector's departure from the mean vector) by simply 

subtracting the polygon's mean intensity vector from the pixel vector. The polygon's covariance 

matrix can then be computed by forming a matrix comprising all pairings of the intensities for each 

pixel deviation vector and then averaging these matrices over all pixels in the polygon (Swain, 

1978). The resulting mean intensity vector and intensity covariance matrix suffice to characterize a 

multivariate normal distribution for that particular polygon for that particular day. This process 

would be performed for all polygons for all images. 

Next, a reference image (date) would be chosen, and quantitative methods would be used to 

identify the spectral deviations of each forested area from the spectral properties of the same 

forested area on the reference image. This approach requires the removal from the multidate 

images of any variations that are not directly related to stress. Many of these variations can be 

removed by rescaling the date-of-interest image (the image for which the stress is to be calculated) 

to the reference image. This rescaling can significantly reduce the effects of seasonal variations in 

7

the incidence angle of solar illumination. If this rescaling is done on a band-by-band basis, 

rescaling can also reduce the effects of those variations in the atmospheric condition that are 

uniform over the entire area of interest. Rescaling cannot reduce the effects of variation in cirrus 

cloud density or other local atmospheric conditions. 

To perform the rescaling, we selected several forest "training sites." A forest "training site" is 

a forested area that is assumed to be spectrally identical, before solar and atmospheric variations, 

on the two dates. Statistics were computed for each of the training sites. Unfortunately, the scale 

factors that were computed from the mean vectors were different for different training sites. This 

suggests that there were significant spectral variations across the image that were not related to 

stress and could not be normalized out. Further, rescaling by a multiplicative constant 

concomitantly requires the rescaling of the covariance matrix elements by the square of the 

corresponding multiplicative constant for diagonal elements and by the products of multiplicative 

constants for the off-diagonal elements. The resulting rescaled covariance matrices did not match 

the reference date covariance matrices. Because these difficulties would adversely affect the 

sensitivity, and hence the usefulness of the ensuing analysis, this approach was abandoned. 

Fortunately, in the interim PSR developed in an independent research and development (IR&D) 

project, an extremely sensitive method of stress detection, the HyperscoutT' algorithim. Because 

of its sensitivity, this algorithm was used in the quantitative portion of the study. In addition, 

several procedures for identifying the spectral signatures that correspond to stressed vegetation 

were developed. 

TmHypwcm is a rgisred tradeark of Pacific-Siena Reanwh Cpocradon.

SECTION 4 

THEMATIC MAPPER IMAGERY 

4.1. INTRODUCTION TO THE THEMATIC MAPPER. 

The Landsat Thematic Mapper (TM) sensor collects data in seven bands in 4 regions of the 

electromagnetic spectrum (see Table 4-1): three visible; one near infrared (NIR); two short-wave 

infrared (SWIR); and one thermal infrared (Engel, 1984). That is, seven images were obtained: 

one through each of the seven spectral band filters. Each region of the spectrum supplies 

information on different manifestations and levels of stress as discussed in Section 2. 

A comment on the numbering of the TM bands may help prevent confusion. As discussed 

above and displayed in Table 4-1, the numbering of the bands is not strictly in order of increasing 

wavelength. 1 Specifically, band 6 is out of sequence. In order of increasing wavelength, the 

bands are: 1 through 5, 7, and then 6. 

Table 4-1. Landsat TM spectral bands. 

Band 

number 

Spectral 

region 

Wavelength band IFOV* 

(microns) (meters) Stress sensitivities 

1 Visible 0.45-0.52 30 Pigmentation changes 

(Especially chlorophyll) 

2 Visible 0.52-0.60 30 

3 Visible 0.63-0.69 30 

4 NIR 0.76-0.90 30 Plant structure damage 

5 SWIR 1.55-1.75 30 Moisture stress 

7 SWIR 2.08-2.35 30 

6 Thermal 10.40-12.50 120 Plant heat stress 

*Instantaneous field of view 

1 Thematic Mapper Simulator (TMS) data, on the other band, is sequential (ie., TMS bands 6 and 7 are reversed 

from TM bands 6 and 7). No further reference to TMS data is made in this study. Band numbers will always refer to 

"TIM band natnbers and, theefore, will be out of wavelength sequence. 

9

A TM image is built up by scanning a sensor that has an instantaneous field of view (IFOV) 

that when projected on the ground measures 30 meters (m) in bands 1 through 5 and 7, and 120 m 

in band 6. During processing, all bands are usually resampled to 28.5 m. Geocoded products, 

such as those used in this study, are resampled to 25 m. 

A frequently encountered misconception is that because the smallest resolvable object can be no 

smaller than a pixel, no information on a scale smaller than a pixel can be extracted. In fact, 

information from objects much smaller than the pixel dimensions can sometimes be extracted. One 

dramatic example of this was the identification of natural gas flarings in the NOAA-6 Advanced 

Very High Resolution Radiometer (AVHRR) data (Matson and Dozier, 1981). The pixel size of 

the NOAA-6 AVHRR sensor was 1.1 kilometer (kin) on a side at nadir, while the dimensions of 

natural gas flarings are clearly only a very small fraction of that size. Further, for AVHRR pixels 

only partially contaminated by cloud, it was possible to determine both the fraction of pixel 

occupied by the cloud and the brightness temperature of the cloud. How is this possible? In the 

case of the natural gas flarings, the flaring was many orders of magnitude brighter than the 

background. Thus, the overall brightness of a pixel was affected although the pixel was orders of 

magnitude larger than the flarings. In the case of cloud contamination, the key is that the data are 

multispectral. Because there were two "thermal" bands (one MWIR and one thermal) on the 

AVHRR instrument, two intensity levels were recorded for each pixel. In some cases these 

intensity levels may be used to extract information on a subpixel scale. The more bands collected, 

the more subpixel information that can be extracted. 

The Landsat 4 and 5 satellites are in 705-kmn sun-synchronous orbits (Irons, 1985). The 

descending node equatorial crossing time (local time) is 9:45 a.m. Thus, all Landsat images are 

acquired in the morning at about the same local time. This minimizes the scene-to-scene image 

variations caused by the solar illumination angle. 

The TM sensor collects a 185-kmi wide swath centered at nadir. To completely cover the 

Earth's surface, 233 orbits are required. Because the orbit period is 98.9 minutes, 16 days are 

required to complete the 233 orbits. Thus, neglecting overlap, each area on the ground can be 

imaged no more frequently than once every 16 days. This 16-day repeat cycle of the TM is not 

frequent enough to monitor the areas receiving the highest dose, especially when allowance was 

made for cloud cover. Fortunately, because the image swaths overlap, a TM image of the 

Chernobyl area could be acquired every 7 or 9 days. The very high probability of cloud cover is 

still a problem for which there is no solution. For areas receiving lower doses, the 7- or 9-day 

repeat cycle is probably frequent enough to monitor radiation stress. 

10

4.2 MOTIVATION FOR USE OF THEMATIC MAPPER IMAGERY. 

Early detection and continuous monitoring of the coniferous tree stress was critical to the 

accuracy of the exposure estimates. Thus, the sensor needs to be sensitive to those regions of the 

spectrum that most clearly exhibit the effects of radiation stress. Unfortunately, the spectral 

manifestations of radiation stress were uncertain. This uncertainty was the primary motivation for 

the use of TM data; TM collects data in four regions of the spectrum, while SPOT collects data in 

only two regions of the spectrum. Thus, even though SPOT has a much higher spatial resolution 

(10-m panchromatic, 20-m multispectral), it does not have the spectral range of TM. 

PSR's Hyperscout change detection algorithm is theoretically independent of the sensor, 

provided that any sensor that is used collects data in the appropriate spectral regions. 

4.3 IMAGE SELECTION AND PROCUREMENT. 

To select images for analysis, we first obtained a list of all TM acquisitions from EOSAT 

Corporation. Second, to avoid purchasing a large number of images that were not usable because 

of cloud cover, we identified those images collected in the area around Chernobyl that were cloud 

free. The simplest approach to identifying cloud free images would have been to use the automatic 

cloud cover assessment on the Landsat acquisition listing. Unfortunately, this number represents 

the cloud cover percentage over the entire scene (an area which is 100 nautical miles (nmi) on a 

side) and not the cloud cover over the immediate area of interest. Even if the cloud cover rating for 

the scene is 80 percent or more, it is still possible that the area of interest is cloud free. 

Conversely, a cloud cover rating of 10 percent did not guarantee that the area of Chernobyl was 

cloud free. Thus, a more localized analysis was desirable. For this study, cloud contamination 

was assessed in two ways. First, Multispectral Scanner data, which are recorded simultaneously 

with TM data, were reviewed at the (MSS) microfilm library located at EOSAT Corporation's 

headquarters in Lanham, Maryland, and observations on image quality and cloud cover 

contamination were recorded (see Table 4-2). Next, black and while (B&W) prints were obtained 

as indicated in Table 4-2 for the more interesting dates and studied in detail. Unfortunately, MSS 

data were not recorded in the spring of 1988, and there is a long waiting time for TM prints. For 

these reasons the procurement decisions for the 1988 data were based solely on the automatic cloud 

cover assessments. 

11

Table 4-2. 

Landsat scene quality and cloud contamination analysis for scenes with less 

than 30 percent cloud cover (see notes following table). 

Date Landat PathWrow Scene cloud B&W print Comment 

sensors cover (%) ordered 

06/29/88 TM-4 182/24 20 No Not processed to film 

* 05/28/88 TM-4 182/24 0 No Not processed to film 

01105/88 TM-4 182/24 20 No Not processed to film 

10/18/87 TM & MSS 181/24 0 No Not reviewed 

10/02/87 TM 181/24 10 No Not processed to film 

09/23/87 TM & MS 182/24 30 Yes Not processed to film 

* 09107/87 TM & MSS 182/24 20 Yes Clouds in left quarter, 

good full scene 

08/22/87 IM 182/24 30 No Not processed to fldm 

07/22/87 TM-4 181/24 10 Yes 6.7-in. film available; 

cloudy 

07/21/87 MSS 182/24 20 Yes Cloud firee near nuclear 

plant 

06/28/87 TM & MSS 181/24 10 Yes Clouds to northwest 

* 05/11/87 TM & MSS 181/23 10 Yes Very good; cloud free 

02/04/87 TM & MSS 181/24 0 Yes Snow cover 

01/10/87 MSS 182/24 0 Yes Snow cover; ice free 

cooling pond 

01/03/87 TM 181/24 0 Yes Cirrus wisps and 

shadows; poor 

12/25/86 TM 182/24 0 No No film 

12/18/86 TM 181/24 20 No No film 

12/09/86 TM&MSS 182/24 10 Yes Some cinrus and 

popcorn; poor 

12/02/86 TM & MSS 181/24 10 Yes Good 

10/22I86 TM 182/24 20 No No film 

10/15/86 TM & MSS 181/24 0 Yes Good 

09/28/86 MSS 182/24 10 Yes Popcorn clouds to 

north 

08/111/86 MSS 182/24 20 Yes Some cirrus 

08/03/86 TM & MSS 182/24 30 Yes Clouds to the 

northwest 

07/18/86 TM 182/24 30 No No film 

07/03/86 MSS 181/24 10 Yes Some popcorn clouds 

12

Table 4-2. 

Landsat scene quality and cloud contamination analysis for scenes with less than 

30 percent cloud cover (see notes following table). (Continued) 

Date Landsat Path/row Scene cloud B&W print Comment 

sensors cover (%) ordered 

06/17/86 MSS 181/24 10 Yes Very good 

06/16/86 IM 182/24 20 Yes Heavy popcorn 

* 05/31/86 TM & MSS 182/24 10 Yes Very good full scene 

* 05/24/86 TM & MSS 181/24 10 Yes Ok, but some popcorn 

clouds 

05/16/86 MSS-4 181/24 10 Yes Good 

* 05/08/86 TM & MSS 181/24 0 Yes Very good 

05/07/86 MSS-4 182/24 0 Yes Very good 

04/30/86 MSS-4 181/24 10 Yes Good 

* 04/29/86 TM & MSS 182/24 10 Yes Some popcorn clouds 

* 03/21/86 TM & MSS 181/24 0 Yes Frozen river, ice-free 

cooling pond 

09/01/85 MSS 182/24 20 No Cloud to the south 

08/25/85 MSS 181/24 10 No Cloud free 

* 06/06/85 TM & MSS 181124 0 Yes Cirrus to the south 

10/25/84 TM&MSS 181/23 10 Yes Thin cirrus over reactor 

area 

07/13/84 MSS 181/24 10 No Very good 

05/01/84 MSS 182/24 20 Yes Some cloud but good 

06/25/83 MSS 181/24 10 No Some haze but good 

Notes: 

1. An * to the left of the date indicates that the data were purchased for that date. 

2. Landsat license restrictions apply only to data acquired after 25 September 1985. 

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 

ordered an initial set of imagery (see Table 4-4). This initial set included nine images that spanned 

the period from 6 June 1985 through II May 1987. The two more recent images were ordered 

later (see Table 4-4) after preliminary results (discussed below) indicated that the area showing 

stress was still expanding. 

Table 4-3. Scene procurement recommendations. 

Date Priority Sensor Medium Area 

05/31/86 1 TIM Tape Full Scene 

08/03/86 2 TM Tape Quad 4 

10/15/86 3 TM Tape Quad 3 

10/25/84 4 TM Tape Quad 3 

03/21/86 5 7IM Tape Quad 3 

12/02/86 6 TM Tape Quad 3 

05/11/87 7 TM Tape Quad 3 

06/17/86 8 MSS Tape Full Scene 

02/04/87 9 TIM Tape Quad 3 

05/08/86 10 TM Tape Quad 3 

06/06/85 11 TIM Tape Quad 3 

04/29/86 12 TM Tape Quad 4 



06/28/87 15 TIM Tape Quad 3 

09/07/87 16 TIM Tape Full Scene 


05/08/86 18 MSS Film Full Scene 

14

Table 4-4. Landsat-scene acquisitions. 

Image number Date Scene ID Padh/row/quad 

1 6/06/85 Y5046208185 181/024/3 

2 3121/86 Y5075008144 181/024/3 

3 4/29/86 Y5078908200 182/024/4 

4 5/08/86 Y5079808133 181/024/3 

5 5/24/F6 Y5081408131 181/024/3 

6 5/31/86 Y5082108191 182/024/4 

7 10/15/86 Y5095808082 181/024/3 

8 12/02/86 Y5100608071 181/024/3 

9 5/11/87 Y5116608123 181/024/3 

10 9/07/87 Y5128508213 182/024/4 

11 5/28/88 Y4214308224 182/024/4 

Because only a small area was believed to have received a sufficient radiation dose to result in 

visible stress, only a 1 degree of longitude by 0.5 degree of latitude area was ordered. Table 4-5 

shows the comer points of this area area. 

Table 4-5. Comer points of the acquired images. 

Corner Longitude Latitude 

(:±ddd:mm.ss.s) 

(±dd:nmm'ss.s) 

Upper left 29:45:00.0 51:40:00.0 

Upper right 30:45:00.0 51:40:00.0 

Lower left 29:45:00.0 51:10:00.0 

Lower right 30:45:00.0 51:10:00.0 

15

4.4. IMAGE PREPARATION. 

As discussed above, precise registration of images to one another is crucial for change 

detection. Registration involves three basic processes: ground control point selection, image 

warping, and resampling. Ground control points (GCPs) are spatial features whose location is 

known in both images or in one image and a map. The locations of these GCPs are then used to 

calculate a mathematical mapping (called a warp) from an arbitrary position in one image to the 

corresponding ground position in the other image or map. This mapping can be envisioned as the 

stretching and twisting (warping) of one image so that it can be overlaid on the reference image 

(map). Once overlaid, any position on the ground occurs at the same location in both images. 

After successful warping, pixels in the warped image will not, in general, be coincident with pixels 

in the reference image. Instead, they will be off center and skewed relative to those in the reference 

image. To correct for this problem, the warped image is resampled to a set of pixels congruent 

with those at the reference image. 

The TM data were delivered directly to STX Corporation of Lanham, Maryland, for 

registration (geocoding). Only the image dated 31 May 1986 was directly registered to maps. The 

maps used were 1:250,000 scale Joint Operations Graphics (JOGs) identified as Series 1501, 

Sheet NM 35-3, Edition 2-GSGS and Sheet NM 36-1, Edition 1. The remaining eight images 

were then registered to the map-registered 31 May 1986 image. This allowed the selection of a 

greater number of control points, resulting in a more precise registration among the images than 

could have been attained by registering all of the images to the maps. Two more recently procured 

images (numbers 10 and 11) were registered in like fashion to the 6 June 1985 image. 

Because the maps used to register the first image were Universal Transverse Mercator (UTM) 

projections, Grid Zone 36, all of the subsequent images were registered to this projection. That is, 

any of the images can be overlaid on a UTM map (or visa versa). The comer points of the area 

covered by these images are shown in Table 4-6. 

Table 4-6. Comer points of the geocoded images after resampling. 

Longitude Latitude UTM X UTM Y 

Corner (degrees) (degrees) (meters) (meters) 

Upper left 294246.9574E 513956.7881N 272700. 5729000. 

Upper right 304459.6821E 514124.8298N 344475. 5729000. 

Lower right 304631.7845E 510959.8775N 344475. 5670725. 

Lower left 294501.3475E 510833.4597N 272700. 5670725. 

16

A 12.8-km sub-area (512 x 512 pixels) was identified for intensive analysis. The corner points 

of this area are shown in Table 4-7. Unless otherwise specified, these coordinates apply to all 

color figures shown in this report. 

Table 4-7. Comer points (Zone 36 UTM coordinates) of the 

512 x 512 pixel area analyzed in this report. 

Comer X Y 

(meters) 

(meters) 

Upper left 289900. 5702100. 

Upper right 302675. 5702100. 

Lower left 289900. 5689325. 

Lower right 302675. 5689325. 

Figure 4-1 shows a sample image of this 512 x 512 pixel area. The Pripyat river flows 

through the upper right hand comer of the image. The dark area adjacent to the river is in the 

cooling pond of the Chernobyl Nuclear Reactor Station. The industrial area comprising the reactor 

station itself is at the upper left corner of the cooling pond. The cooling water intake channel (right 

angle bend) and the outlet channel (obtuse angle) are evident. Water circulates counterclockwise in 

the pond. The reactor buildings are located in an east-west row just above the east-west portion of 

the intake channel. Reactor four is leftmost in the row. 

The city of Pripyat is at the upper center of the image, just south of the river and northwest of 

the reactor site. The city of Chernobyl, an old river port, is about 10 km to the southeast, not on 

this image. 

The bright green areas in Figure 4-1 are mostly agricultural lands; the bright pinkish or reddish 

areas are bare fields. The dark green areas are predominantly conifer forests as evidenced by 

seasonal progressions in subsequent figures. 

The standard geometric-correction algorithm applied to TM data (Irons, 1985) results in an 

image that appears to be somewhat blurred. But, unlike Goldman and coworkers (Goldman, 

1987), we made no attempt to sharpen the images. If the images are sharpened, even small 

(subpixel) registration errors may result in large apparent image-to-image changes for 

corresponding pixels. Further, because these errors are systematic, real stress at the single pixel 

level may be entirely obscured by this misregistration-induced apparent stress. 

17

Figure 4-1. 

Vicinity of the Chernobyl Nuclear Reactor Station on 8 May 1986; 7,4,1 false-color 

presentation (12.8 km by 12.8 km, north is up). Sixteen of the indicated polygons 

are areas selected as candidate training sets for the forest classification process. 

18

SECTION 5 

IMAGE ANALYSIS 

This section discusses the techniques used to identify forest pixels in the Landsat imagery and 

reviews the false-color images and forest stress results for each date. 

5.1. IMAGE PARTITIONING. 

Because only forested areas are of interest for our analysis, we applied methods based on 

statistical decision theory to eliminate other areas from consideration. Maximum-likelihood 

classification (Swain, 1978), a parametric method for statistical classification, requires estimates of 

the parameters of the statistical distributions that are assumed to characterize the variation of 

spectral intensities over statistically homogeneous regions of the image. The pixels within selected 

areas of the image (referred to as training areas or training sites) are appropriately analyzed to 

estimate statistical parameters of an assumed statistical distribution for the pixels belonging to each 

class. Each training area yields a set of statistical parameters and a corresponding class distribution 

in the form of an analytic function involving the class parameters. 

The distributions are usually assumed to be multivariate normal in their analytic form. The 

parameters of the multivariate normal distribution law include the mean vector and the covariance 

matrix. The mean vector is a band-by-band average of the intensity taken over every pixel in the 

training site. Thus, if seven bands are used, the mean vector is seven dimensional (i.e., it has 

seven components, one for each band average). The covariance matrix comprises the covariances 

estimated from all pairings of the band intensities and will be a square array of dimension equal to 

the number of bands. 

An analytic distribution derived in this way provides a means for estimating the probability that 

a pixel belonging to the corresponding class will be found to lie at any given point in the 

multispectral feature space. Thus, probabilities that a pixel belongs to each of the classes under 

consideration can be estimated at any given point in the multispectral feature space. Maximumlikelihood 

classification assigns a pixel to the class for which its position in the feature space has 

the greatest probability. In this manner the pixel classification process assigns each pixel of the 

image to one or the other of the classes, resulting in a partitioning of the image into the predefined 

classes. 

Training sites are usually determined from ground truth. Unfortunately, because ground truth 

of the Chernobyl area was not readily available to us at the time of this analysis, we identified 

training sites visually by spectral characteristics and characteristic texture. Potential training sites 

identified in this way are displayed as polygons in Figure 4-1. 

19

Ideally, assignment of a pixel to a given class depends on the joint probability that it would be 

characterized by a pixel vector X (its vector of intensity values) while belonging to class Coi. This 

joint probability is given by the following equation: 

p( Op(G) . exp- .(X - M). T'1: (X - M.)] 

(2xj(1) 

where, 

n = the number of bands (also the dimensionality of X and U, and the order 

of S), 

p(co) = the probability of class wi, 

Mi = the mean vector for class coi, 

X = the pixel vector (n intensity values), and 

Mi = the covariance matrix for class mi. 

Note that this is not a conditional probability. However, it is related to the conditional probabilities 

by 

p(X C, O)= p(Xji) p(Ci) =p(OiIX) p(X) (2) 

The pixel is then assigned to class wji if and only if 

pXoI) po)>pXcj) p( j) (3) 

for all j = 1,2,..., m, where m is the number of classes. Because in this study the class 

probabilities p(coi) were unknown, they were assumed to equal one another, causing the class 

assignments to be determined by the conditioned probability of the pixel vector's occurrence, given 

the hypothetical class. 

For this study we required the training sites to be minimal in number and near the reactor, but 

not subject to radiation stress. We chose the smallest possible number of training sites for the 

analysis because maximum-likelihood classification is computationally expensive. Ideally, one 

training site would be selected for each distinguishable forest type. The distinctions among forest 

types might be based on ground truth data such as differences in the mixture of coniferous and 

deciduous trees, age of the stand, densities or crown closures, or species in the area. Because little 

ground truth was available at the time of the study, the selection of training sites had to be based 

entirely on their remotely sensed multispectral characteristics. It was found that two training sites 

20

provide a satisfactory classification of the pixels in the image, a classification that includes nearly 

all of the pixels in regions believed to be forested while including few in regions that were clearly 

not forested. In the process of selecting these two training sites a number of potential training sites 

were selected in the area near the Chernobyl nuclear reactor, these are shown on Figure 4-1. Note 

that only 16 of these are forest. Five were chosen to include only water, and one was believed to 

be sand. These nonforest sites were chosen in anticipation of a need to normalize the various 

images using features whose spectral response could be assumed stable over time. Because the 

method finally chosen to detect stress did not require such normalization, the nonforest training 

sites were not used. 

Partitioning the image assigned every pixel in the image to one or the other of the predefined 

classes. Because we were interested only in forests, we selected only forest training sites and 

included only forest classes in the set of predefined classes. After the image was partitioned in this 

way, pixels that obviously contained water, bare soil, and all manner of other things were 

assigned, along with the forest pixels, to the closest of the forest classes. As yet no consideration 

had been given to the possibility that a pixel's spectral features might have been far more likely to 

have arisen among pixels belonging to nonforest classes such as water, bare soil, etc; we had yet to 

define what was and was not a forest. 

Nonforested areas could have been removed from consideration by maximum-likelihood 

classification in the same manner as was used for the forest classes. However, this would have 

entailed developing training areas or sites for every class of landcover that occurred in the image, 

and it would have required that the a probability be calculated for every pixel for every such class. 

A computationally more efficient process was to apply a threshold to the probability computed for a 

pixel's membership in the forest classes. If the pixel's multispectral features occurred in a 

particular forest class with a probability less than a certain threshold value, the pixel was assumed 

to belong to a class other than that forest class. In practice, we adjusted the probability threshold 

while monitoring the outcome of the classification process, raising the threshold if too many stray 

pixels in the image seemed to be assigned to the desired classes and lowering the threshold if too 

many of the pixels thought to belong to the desired classes were being rejected. The resulting 

statistical decision process may be looked upon as a casual implementation of the Neyman-Pearson 

decision rules; the threshold for pixels belonging to the forest class was made as low as possible 

without accepting too many pixels that lay outside the apparent areas of forest. 

When the image was partitioned using only one training site (e.g., training site 4 on 

Figure 4-1), many of the forested areas were misclassified'as nonforested areas (errors of 

omission) when a reasonable threshold value was selected. This threshold value was determined 

qualitatively by viewing the effect of a given threshold value on the classification. Lowering the 

21

threshold sufficiently would include all suspected forested areas, but at the cost of including too 

many other areas that were perceived to be nonforested (errors of commission). 

Finding that no adjustment of the threshold would lead to an entirely satisfactory result, we 

concluded that the use of only one training site was not sufficient and considered using two. The 

question then became "which two?" As discussed above, the selection should be based on 

statistical arguments. To this end, PSR developed a modified form of the transformed divergence 

(Swain, 1978) specifically for use in this study. The divergence between two training sites was 

defined by 

4I _ ) _ I 

•)= ( 

l1]+l 

The transformed divergence is given by 

DT= 

ij I 8(5) 

Unfortunately, the transformed divergence saturated (yielded 100) for even small differences in 

cluster statistics if a large number (>4) of bands was used. In addition, the results were strongly 

dependent on the number of bands used. To reduce the effects of these problems, we defined the 

modified transformed divergence by 

D MT = 100 1 xp(- 

ii L\ 2n/ (6) 

The modified transformed divergence was relatively independent of the number of bands used 

in the analysis provided that the bands used were not highly correlated. Table 5-1 presents the 

modified transformed divergence between every pair of potential forest training sites shown in 

Figure 4-1. These results led us to choose sites 3 and 4 as training sites to partition the 6 June 

1985 image. Figure 5-1 shows the results of this classification. Areas colored green were 

assigned to the class developed from training site 4, and areas colored yellow, to the class 

developed from training site 3. This classification map (recall that the image is registered, so this 

really is a map) was then overlaid on a gray-scale TM band 4 image of the area. It is not important 

for the present analysis to know the difference between the types of forest in these training sites. 

We need only to account for their statistically distinct multispectral characteristics. 

22

cn eno 

00k0t- t Oo %0 roo~~ M C 4e 

oo0% a 00 00-c m0% a m 

4000 Q 00 N*AC 0 4% 1 

Ml V_ t*- r- n 0 t- N 00 Aý t~-00O 

C.) eq 0%~ N~~'O - (nl* 4 in~ 00*4 - t 00 

00e cC C : *M ýQ oc* o ioc-t 

oo a6f~ 0 %Vcý0 ~ 4 % %0O~ 

0 0%Ot-00 w' w we4 m0 M 

So -00 C -%n W 0o0 in t- cc cc O 

%Q~ 0000 00 ) ) o ý - Q 

co % 000 

0%* n0 

cn n o r- @000 r- V % 00 0% %C 

w 00 

23 %t

Ar 

Figure 5-1. 

Partitioning of 6 June 1985 image into classes derived from training sites 3 (yellow) 

and 4 (green). Gray-scale background consists of nonforested areas. 

24

To minimize the effect of local atmospheric variations, we wished to choose training sites as 

close to the reactor as possible. On the other hand, if training sites too close to the reactor were 

chosen, they could have been subject to radiation stress, especially if the image used to identify 

forested areas was acquired after the accident. This study used the image dated 6 June 1985 

(before the accident) to partition the image. Nevertheless, the training sites were selected to be 

outside the area in which radiation stress was expected to appear in postaccident images. 

5.2. MAPPING STRESS. 

In this section, each of the selected images is reviewed in chronological order with important 

features identified. These require some explanation. Standard false-color prints of TM data 

display band 4 as red (R), band 3 as green (G), and band 2 as blue (B). 

This presentation is referred to as a 4,3,2 RGB false color. In this report, the color ordering 

will always be RGB unless otherwise stated). As discussed in Section 2 above, vegetation should 

appear red in this standard presentation. The shadowing of trees by one another causes forests to 

appear highly textured and darker. 

There is another standard false-color display gaining acceptance in the community (Johnson): 

the 7,4,2 false color. This results in a more appealing and less arcane image because it appears 

much like a normal (3,2,1 RGB), true-color image but has better contrast information content; the 

7,4,2 false-color image contains information from three quite different regions of the spectrum, 

rather than being restricted to the visible region as is a 3,2,1 image. In the 7,4,2 presentation bare 

soils tend to appear red, with drier soils appearing a brighter red color. Vegetation appears green, 

and shallow or turbid water tends to appear blue. 

The color prints presented here use a nonstandard 7,4,1 false-color display. This combination 

captures most of the advantages of the standard 7,4,2 false-color display, but with the added 

advantage that it makes cirrus clouds more obvious. Because cirrus clouds can (and invariably do) 

interfere with analysis, knowledge of the location of even light cirrus clouds can be important 

For images acquired on or after 29 April 1986, stress maps have been produced using PSR's 

Hyperscout change detection algorithm with the 6 June 1985 image as reference. The stress 

algorithm assigns a single number, called the stress index, to each forested pixel in the image, and 

these are displayed in the stress map. Again, the resulting stress map, like the images from which 

it was derived, is registered to the UTM projection, Grid Zone 36. The stress map could be 

displayed as a gray scale, but the resulting image would be difficult to discern in areas manifesting 

low stress. In this report, the stress gray scale is converted to a color scale. A legend showing 

this color scale appears with the stress maps shown below. Values of the highest stress index are 

colored red. As the stress index decreases, the color shifts continuously through orange, yellow, 

green, and blue to magenta. Thus, magenta-colored areas are the least stressed. Nonforested 

25

areas, which could not be analyzed for stress, are filled in with a gray-scale TM band 4 image to 

establish a spatial context for the interpretation. The accompanying text discusses significant 

features of these stress maps, as well as of the 7,4,1 false-color images. 

The stress maps derive from an analysis of only three of the 7 bands, namely, 3, 4, and 7. 

Band 6 (the thermal band) was not used because the images frequently included clouds. The low 

temperature of clouds, coupled with the 120-m resolution of band 6, affects an area larger than that 

where clouds are evident. Even light cirrus clouds would mislead a stress calculation that includes 

band 6 because of its high sensitivity to the temperature of the cloud cover. Bands 1 and 2 (blue 

and green) were not used because they also are relatively sensitive to cirrus clouds and atmospheric 

scattering as compared to the red or infrared. This left only bands 3, 4, 5, and 7. Bands 5 and 7 

are very highly correlated, so either may be used. We arbitrarily choose band 7 over band 5. 

Figure 5-2 shows the reference image from 6 June 1985, 1 year before the accident. Recall 

that this is also the image used to classify forest types. A significant feature of this image is that 

the forested areas due west of the reactor appear darker and may even have a brownish tint. This is 

also the type of coloring (spectral reflectance) that might be expected from radiation-stressed 

coniferous forests. Because it is unlikely that the coloring present on 6 June 1985 was caused by 

radiation stress (certainly not that caused by the release on 26 April 1986), this image indiates that 

qualitative (visual) identification of stress from the false-color images can be misleading. 

Figure 5-3 shows the 21 March 1986 image also taken before the accident. Some snow cover 

was evident (blue color), and the rivers (but not the warm cooling pond) were ice covered. At 

training site 3 very little snow cover was visible through the canopy. On the other hand. some 

snow appeared at training site 4. This observation can be interpreted in at least two ways. First, 

training site 4 may have more deciduous trees, or, second, the forest may not be as dense in 

training site 4 as it is in training site 3. Which, if either, of these interpretations is correct has not 

been determined. 

A study, similar in intent to this study, was recently preformed by Goldman and coworkers 

(Goldman, 1987). They used enhanced Landsat TM data to identify stress visually. They 

concluded that stress could not be detected visually on images taken before 16 June 1986 (i.e., 

until more dian 7 weeks after the accident). 

Three days after the accident, Landsat acquired an image of the reactor area. This image, taken 

29 April 1986, shows some cloud cover as seen in Figure 5-4. The reactor, still extremely hot, 

was readily apparent as a deep red area due north of the reactor water inlet pooL Another reddish 

white feature appeared 1.1 km to the west of the reactor and was mistakenly reported to be another 

burning reactor. There was no visibly stressed vegetation. 

26

Figure 5-2. 

Date: 6 June 1985, 1 year preaccident; 7,4,1 false color. This date was used as 

a reference for change detection. 

27

Figure 5-3. Date: 21 March 1986, 2 months preaccident; 7,4,1 false color. Snow cover is 

evident (bright blue) around the forest patches. 

28

Figure 5-4. 

Date: 29 April 1986, 3 days postaccident; 7,4,1 false color. The deep red pixel is 

the thermal emission from the hot reactor core. 

29

Figure 5-5 presents the stress map for this date. The area of high stress index that appears 4 

km due west of the reactor was cleared before the accident. Nearly all of the rest of the apparent 

stress was caused by clouds. The centers of cumulus clouds and their shadows usually appear 

deep red in the stress index maps, while at their fringes there appears a sharp, rainbowlike 

progression of colors from deep red to magenta. Cirrus clouds are much more insidious; their 

effects vary in color and extent. In some cases cirrus clouds show no effect at all, while in others, 

they produce a pattern of apparent stress similar to that expected of real stress. In Figure 5-5, there 

is a small area approximately 0.9 km west by southwest of the reactor that has a sufficiently high 

stress index to be displayed in the green to yellow color range. On the basis of this single image, it 

is debatable whether this is real stress, cloud effects, smoke, debris covering the foliage, or 

something else. A qualitative measure of the statistical nature of the noise can be obtained by 

looking at areas that were not stressed (e.g., training site 3). These areas are colored mostly blue 

and magenta except near the edges of the forests. From this observation, we conclude that areas in 

the stress map speckled blue and magenta are probably not really stressed but only the effect of 

random noise-level detections of change. In addition, any apparent stress that appears only at the 

edges of forests is suspect, because of possible registration errors. 

The 8 May 1986 image, presented in Figure 5-6, was collected 12 days after the accident. 

There were no clouds in the image, nor was there stressed foliage visible in the 7,4,1 false-color 

presentation. 

Figure 5-7 presents the stress map made from the 8 May 1986 image. In it appears a very well 

defined 0.9-km long strip of forest, beginning 2 km west by southwest of the reactor and running 

to the west, where stress was apparent. This area also had a higher than normal stress index on 

29 April 1986, but the area was not as clearly deft-ed. The shape of this stressed area was 

consistent with that of a directed explosion or wind-deposited fallout. Because by this time the fire 

at the reactor was reported to have been extinguished, the detected change seems unlikely to be 

attributable to smoke. Thus, in this stress map we can very clearly see some indications of 

accident-related stress or change that appeared within 12 days after the accident. Located 6 km 

south of the reactor there appears an apparent stress or change feature that we have interpreted to be 

an artifact of the change analysis algorithm. The exact cause is unknown. This artifact appears on 

almost all of the images that follow; its stress index changes very little in time. 

30

0 128 256 

St•es Index 

Figure 5-5. 

Date: 29 April 1986, 3 days postaccident, forest stress map. colored according to 

indicated scale. Gray-scale background consists of nonforested areas. 

31

Figure 5-6. Date: 8 May 1986, 12 days postaccident; 7,4,1 false color. 

32

0 128 256 

Sbss Index 

Figure 5-7. 

Date: 8 May 1986, 12 days postaccident, forest stress map. 

33

Figure 5-8 shows the 24 May 1986 image collected 4 weeks after the accident. Although it is 

difficult to see on the 7,4,1 false-color image, there was a large cirrus cloud in the area of the 

reactor. There were also some "popcorn" clouds in the area. The obscuration of the cirrus cloud 

and the natural color of the area make positive visual detection of stress uncertain in the false-color 

image. On the other hand, an area of change becomes very obvious and spatially well defined 

when presented on the corresponding stress map (Figure 5-9). Unfortunately, so does the large 

cirrus cloud. The cirrus cloud appears as a blue to green swath on the stress map, affecting the 

bulk of the eastern side of the image. The "popcorn" clouds and their shadows show up as deep 

red. By 24 May 1986 aerial spraying to prevent the wind from redistributing radioactive dust, 

along with the dust itself, may have contributed to the changes that were detected. 

The image taken on 31 May 1986 (Figure 5-10) also included some clouds. Positive visual 

identification of stress remains uncertain, but the stress map shows the stressed area very clearly 

(Figure 5-11). Little change appears in the spatial distribution of stress between 24 and 31 May, 

except that a small new area just south of the previously stressed area began to indicate a very high 

stress index. Because this new area was adjacent to a road, it probably indicates changes caused 

by human activities other than radiation release. 

After 31 May 1986, the next image selected for analysis was that collected 15 October 1986. 

An interim image collected 16 June 1986 was not purchased because of cloud cover exceeding our 

standards. 

By 15 October 1986 a band of stressed forests that appeared to the west of the reactor can be 

readily discerned in the false-color image (Figure 5-12). Also visible is evidence of extensive 

human activity. The Soviets had started to clear some of the contaminated forests near the road that 

passes to the west of the reactor. There was evidence of extensive diking operations in the 

northwest and northeast comers of the image. Indications appeared of a cleared swath passing 

through the coniferous forest to the southwest of the reactor. Because this image was collected in 

the fall and there were deciduous trees in the area, the spectral definition of stress becomes affected 

by the natural changes of the deciduous foliage and the seasonal sun angle changes. As a result, 

the apparent stress is lower than before. Examination of the stress map in Figure 5-13 confirms 

this hypothesis. The stressed area remains obvious and fairly well defined, but the stress index is 

lower. 

The situation regarding deciduous foliage is similar on 2 December 1986 (Figure 5-14). In 

addition, the low sun elevation causes a further loss of sensitivity. Although the stresses 

(Figure 5-15) appear lower than before, the areal extent of the stress does not appear to have 

changed significantly. 

34

Figure 5-8. Date: 24 May 1986,4 weeks postaccident; 7,4,1 false color. 

35

I I I 

0 128 256 

Stress Index 

Figure 5-9. Date: 24 May 1986,4 weeks postaccident, forest stress map. 

36

Figure 5-10. Date: 31 May 1986, 5 weeks postaccident; 7,4,1 false color. 

37

I 

II 

0 128 256 

Stess Index 

Figure 5-11. Date: 31 May 1986, 5 weeks postaccident, forest stress map. 

38

Figure 5-12. Date: 15 October 1986, 5.6 months postaccident; 7,4,1 false color. 

39

0 128 256 

Stress Index 

Figure 5-13. Date: 15 October 1986, 5.6 months postaccident, forest stress map. 

40

Figure 5-14. Date: 2 December 1986, 7.2 months postaccident; 7,4,1 false color. 

41

.......... 

0 128 256 

Stress Index 

Figure 5-15. Date: 2 December 1986, 7.2 months postaccident. forest stress map. 

42

The Soviet mitigation efforts were much more extensive by 11 May 1987 (Figure 5-16). A 

large area is cleared to the west of the reactor complex, as is an even larger area south by southeast 

of the reactor. Much of the diking appears to have been completed, and the areas behind the dikes 

appear flooded. Again, there is a rather large cirrus cloud partially obscuring most of the forests in 

the western part of the image. 

Goldman and coworkers (Goldman, 1987) state that no new stressed areas had appeared by 11 

May 1987. Indeed, a visual inspection of the false-color image provides no evidence to the 

contrary. However, the stress map indicates significant new areas beginning to show stress. 

Before looking at the stress map, the reader may find it enlightening to attempt to identify these 

areas visually in the false-color image. As an aid to the reader, Figure 5-17 shows Sovietsupplied 

gamma dose rate contours dated 1 May 1987 overlaid on the TM image. This ground 

truth will be discussed in detail in Section 6 below. The innermost contour is for 100 mR/hr, with 

contours for 50, 10, 5, 2, 1, 0.5, and 0.1 mR/hr appearing progressively outward. If these 

contours are correct, the bulk of the coniferous forest to the south was still receiving a dose of at 

least 5 mR/hr, and all areas between the coniferous forest and the cooling pond to the east and the 

reactor complex to the north were receiving a dose of at least IOmR/hr. Because the areas 

receiving doses higher than this had already shown an elevated stress index, these areas were the 

most likely next candidates. 

Turning to the stress map shown in Figure 5-18a, we see that these areas were indeed 

beginning to show symptoms of stress. These newly stressed areas included much of the 

coniferous forest to the south of the reactor and nearly all of the forested areas between the 

coniferous forest and the cooling pond to the east and the reactor complex to the north. For ease of 

comparison, Figure 5-18b shows the dose rate contours overlaid on the stress map from 

Figure 5-18a. 

No images that were free of clouds and haze (see Table 4-2) were collected in 1987 after 

22 July 1987. However, an image taken 7 September 1987 (Figure 5-19) was procured to help 

confirm the indications of continued stress implied by the 11 May 1987 stress map. Unfortunately 

for the comparison, the Soviets had cleared much of the suspect area lying within the 10 mR/hr 

dose contour. The northeastern edge of the coniferous forest continued to appear heavily stressed 

(Figure 5-20) but not as heavily stressed as before. The apparent reduction in stress may have 

been due to the heavy haze. Indications of new stress appear in the 7 September 1987 stress map 

in an area just north of the area first showing stress. 

43

Figure 5-16. Date: 11 May 1987, 1 year postaccident; 7,4,1 false color. 

44

Figure 5-17. 

Soviet-supplied gamma dose rate contours of 1 May 1987 overlaid on the 

11 May 1987 7,4,1 false-color image. The innermost contour is 100 mR/h, the 

next innermost 50 mR/h, then 10, 5, 2, 1, 0.5 and 0.1 mR/h, respectively. 

45

I I I 

0 128 256 

Stress Index 

Figure 5-18a. Date: 11 May 1987, 1 year postaccident, forest stress map. 

46

I 

II 

0 128 256 

Stress Index 

Figure 5-18b. 

Date: 11 May 1987, forest stress map with gamma dose rate contours of 

1 May 1987 (see Figure 5-17 for contour values). 

47

Figure 5-19. Date: 7 September 1987, 16 months postaccident; 7,4,1 false color. 

48

I I I 

0 128 

Sb'ess Index 

256 

Figure 5-20. Date: 7 September 1986, 16 months postaccident, forest stress map. 

49

Because newly stressed areas were still being found and the 7 September 1987 image was of 

low quality, a more recent image was procured. This image, dated 28 May 1988, is presented on 

Figure 5-21. Again, Soviet mitigation efforts had cleared some of the newly suspect arrias, so 

continued monitoring of those areas was not possible. The clearing of the newly suspect areas 

suggests that these armas had been affected enough that they needed to be decontaminated, although 

no mitigation efforts were applied to the coniferous forest to the south of the reactor, at which 

signs of stress persisted (Figure 5-22). 

5o

Figure 5-2 1. Date: 28 May 1988, 25 months postaccident; 7,4,1 false color. 

51

I I I 

0 128 256 

S&ess Index 

Figure 5-22. Date: 28 May 1988, 25 months postaccident, forest stress map. 

52

SECTION 6 

GROUND TRUTH 

Some ground truth became available (Asmolov, 1987) in the form of gamma radiation dose 

contours for 1 May 1987 (i.e. approximately I year after the accident). Unfortunately, there are a 

number of difficulties associated with the use of these data. First, the copy of the dose contours 

that was available to us is very poor; the contours appear to be hand drawn, and in some places the 

contours are not closed, making digitization difficult. Also, the contour map has no grid lines and 

no accurate ground control points (GCPs). In fact, the only usable GCPs are river bends. 

Further, the indicated courses of these rivers on the contour map do not match the maps that were 

used to register the images. Because they match the images fairly well, registration directly to the 

images is a viable alternative. Two cities are shown as circles, but the location of one of them, 

Chernobyl, is apparently in error by several kilometers. The course of the river Uzh also appears 

to be in error near this city. The location of the reactor itself is not marked on the map. 

In spite of these difficulties, 36 GCPs were identified. Using these GCPs, the coefficients of a 

second-order warp were calculated by means of a least squares algorithm. When this warp was 

applied to the digitized contours and the results overlaid on an image, the 100 mR/h contour was 

centered on reactor 3 instead of reactor 4. Use of a third-order warp resulted in unacceptable (i.e., 

unlikely) distortions of the contours near the edges of the map. Attempts to generate a fourth-order 

warp were unsuccessful. Because the ability to weigh GCPs preferentially is inherent to the 

algorithm used, another GCP, the best guess of the location of the reactor based on the shape of 

the contours, was added to the set and given a relative weight of 100. The coefficients for a 

second-order warp were then recalculated using the least squares algorithm. The equations of this 

warp are shown in equations 7 and 8: 

and 

xref=al+a 2 x+a 3 -y+a 4 - x2+a 5 * x- y+a 6 *y 2 (7) 

yrf= bi +b 2 . x+b 3 . y+b 4 . x2+b • x- y+b 6 .y 2 (8) 

53

where, 

ai = the coefficients of the warp for the x-coordinate of the image, 

bi = the coefficients of the warp for the y-coordinate of the image, 

x - the x-coordinate of a point on the dose contour map, 

xref = the corresponding column (x-coordinate) on the image, which is related to UTM 

map coordinates, 

y - the y-coordinate of a point on the dose contour map, and 

yref - the corresponding row (y-coordinate) on the image, which is related to UTM 

map coordinates. 

The x- and y-axes of the contour plots are arbitrary because no grid lines or map projection was 

given for the dose contour map. Still, the relative magnitudes of these coefficients are of some 

interest; they are presented in Table 6-1. A measure of the accuracy of the registration can be 

obtained by warping the location of the GCPs on the contour map through Equations (7) and (8); 

then comparing these warped coordinates to the true location of those GCPs in the image. These 

results are presented in Table 6-2. The contour lines were warped and overlaid on an image 

(Figure 5-17). In order starting with the innermost, the contours are for doses of 100, 50, 10, 5, 

2, 1, 0.5 and 0.1 mR/h. 

Table 6-1. Second order warp coefficients for the 

gamma dose contours. 

i ai bi 

1 -0.2760500E+04 0.4388116E+04 

2 0.1866979E+00 0.1391407E+00 

3 -0.4149658E+00 -0.4166605E+00 

4 0.8310569E-05 -0.3711088E-05 

5 0.1337978E-04 -0.1319325E-05 

6 0.1379850E-04 -0.5189849E-05 

54

Table 6-2. Accuracy of the registration of GCPs. 

x Xref (pixels) xfit (pixels) Error (pixels) 

14864.79 1024.00 1024.99 0.99 

13347.76 249.00 271.18 22.18 

13376.74 271.00 282.24 11.24 

13525.32 365.00 357.72 -7.28 

13434.05 297.00 301.57 4.57 

13533.28 364.00 352.89 -11.11 

13620.08 397.00 387.50 -9.50 

13750.20 470.00 459.64 -10.36 

13847.89 505.00 504.52 -0.48 

13866.87 509.00 511.20 2.20 

14116.60 653.00 648.12 -4.88 

14083.66 635.00 619.49 -15.51 

14224.67 712.00 700.87 -11.13 

14381.22 794.00 775.78 -18.22 

14688.25 939.00 936.75 -2.25 

15065.93 1148.00 1140.12 -7.88 

15145.92 1204.00 1184.85 -19.15 

15427.47 1334.00 1320.82 -13.18 

15801.45 1519.00 1518.99 -0.01 

15829.66 1530.00 1532.57 2.57 

15960.41 1608.00 1606.15 -1.85 

16016.58 1637.00 1633.56 -3.44 

16197.20 1732.00 1734.57 2.57 

15469.94 1337.00 1334.33 -2.67 

15438.92 1319.00 1316.72 -2.28 

15338.93 1262.00 1262.07 0.07 

15215.50 1196.00 1194.90 -1.10 

15246.26 1218.00 1211.36 -6.64 

15326.76 1251.00 1254.99 3.99 

14944.11 1049.00 1049.89 0.89 

14803.87 976.00 975.37 -0.63 

14782.98 971.00 965.25 -5.75 

14692.60 917.00 917.88 0.88 

14563.63 849.00 848.41 -0.59 

55

Table 6-2. Accuracy of the registration of GCPs (Continued). 

X Xref (pixels) xfu (pixels) Error (pixels) 

14174.82 636.00 642.65 6.65 

14255.32 685.00 684.05 -0.95 

13806.38 449.00 448.57 -0.43 

Y Yref(pixels) yfu (pixels) Error (pixels) 

9029.81 1274.00 1273.78 -0.22 

10839.32 260.00 267.20 7.20 

10758.94 331.00 311.86 -19.14 

10647.04 365.00 376.65 11.65 

10548.19 434.00 428.17 -5.83 

10486.92 475.00 463.98 -11.02 

10235.79 607.00 602.26 -4.74 

10244.13 592.00 600.88 8.88 

10062.87 689.00 701.09 12.09 

9982.62 745.00 744.78 -0.22 

9959.43 756.00 762.79 6.79 

9729.84 888.00 885.48 -2.52 

9808.04 834.00 846.50 12.50 

9536.02 993.00 995.44 2.44 

9302.09 1116.00 1126.06 10.06 

9117.25 1237.00 1230.63 -6.37 

9116.23 1239.00 1232.37 -6.63 

8452.58 1591.00 1586.75 -4.25 

7977.77 1843.00 1839.49 -3.51 

7837.40 1915.00 1912.75 -2.25 

7895.74 1868.00 1883.86 15.86 

7595.00 2046.00 2040.27 -5.73 

7642.69 2014.00 2017.33 3.33 

7861.99 1892.00 1895.45 3.45 

7782.38 1938.00 1936.27 -1.73 

7783.66 1936.00 1934.14 -1.86 

7515.21 2069.00 2070.78 1.78 

7574.82 2037.00 2040.58 3.58 

7613.80 2019.00 2021.74 2.74 

7408.67 2132.00 2120.83 -11.17 

56

Table 6-2. Accuracy of the registration of GCPs (Continued). 

y yref(pixels) Yfit (pixels) Error (pixels) 

7390.46 2123.00 2127.52 4.52 

7320.72 2162.00 2162.85 0.85 

7291.87 2183.00 2175.80 -7.20 

7373.52 2124.00 2131.29 7.29 

7468.49 2079.00 2073.78 -5.22 

7507.46 2050.00 2055.69 5.69 

7593.20 2000.00 2000.42 0.42 

57

SECTION 7 

DISCUSSION 

Because only very limited ground truth is available for comparison with the results of this 

study, it is difficult to verify the accuracy of the stress maps that were produced. The 1987 gamma 

dose rate contours seem to correlate well with the later stress maps (Figure 5-18b), but the 

correlation is not perfect. 

There are several important points to be made concerning these contours. First, the contours 

correspond to the gamma dose rate at a time (1 May 1987) approximately 1 year after the accident 

and not to the accumulated dose. Second, these contours show only the gamma dose rate, not the 

alpha or beta particle dose rates. Third, as discussed above, the position or these contours is 

probably not very accurate. In fact, for these reasons, the forest response can be considered to 

provide a more accurate indication of the integrated dose to foliage than do the 1 May 1987 dose 

contours. 

There is one final noteworthy point concerning the position of the contours. It is possible that 

Soviet mitigation efforts were more effective over the damaged reactor than over the neighboring 

undamaged reactor. In that case, the center of the 100 mR/hr dose contour might really belong 

over the undamaged reactor. Mitigation efforts might also help to explain the southerly 

displacement of the dose contours relative to the areas of forest .showing the highest stress 

(Figure 5-18). Much of the area that first showed stress, and that was later removed, lies outside 

the 100 mR/hr dose contour. 

In lieu of ground truth, other methods for checking the results of this study can be applied to 

increase confidence in the results. Are the stress maps consistent with Soviet mitigation efforts in 

the area? Are the results self-consistent over time? Do the results of this study agree with those of 

other similar studies? 

The first of these questions was addressed in considerable detail in Section 5. Basically, most 

Soviet mitigation efforts in forested areas appear to have been consistent with the stress map. The 

only notable exception was the northern edge of the coniferous forest located to the south of the 

reactor. This area consistently showed stress in the stress maps, but no signs of mitigation were 

evident. Some possible explanations are (1) that the Soviets may believe that this area will recover, 

(2) the area may have low priority for current operations, or (3) the afflicted area is too large for 

mitigation to be practicable. 

With the exception of the winter months (when stress seems less detectable in the present 

analysis), the stress maps appear to be consistent with one another. The discolored area visible on 

the 29 April 1986 stress map is the same area that appeared stressed on the 8 May 1986 stress map. 

58

Similarly, the stressed area on 24 May 1986 includes all the stressed areas of 8 May 1986. The 

24 May 1986 and the 31 May 1986 stress maps show little difference. The stressed area of 

11 May 1987 subsumes that in all the previous images. In addition, apart from that for the winter 

months, the effects measured by the stress index increased fairly uniformly with time. Thus, the 

stress maps are consistent with one another. 

The last question is somewhat more difficult to address because of the lack of similar studies. 

Although the scope and even the data used for the study by Goldman and coworkers (Goldman, 

et al., 1987) were similar to that of the present work, the approaches and the results obtained were 

quite different. A comparison of the results obtained by these two studies is shown in Table 7-1. 

Most of the aspects of this comparison were discussed either above or in Section 5. In their study, 

a group of analysts (albeit experienced) attempted to identify stress visually by looking at 

photographic prints. Such a process is obviously subjective. Photographic prints are far from an 

ideal media, for the very production of the prints tends to involve variables usually controlled by 

subjective judgments (e.g., exposure and color balance). In addition, the dynamic range that can 

be achieved with the photographic process is not very large, resulting in a loss of information. In 

contrast, the results of this work were obtained largely by quantitative means. In this respect, we 

believe the methods and results of this study speak for themselves. 

Johnson (1989) has developed a phenomena-based image-enhancement transformation that is 

gaining popularity in the intelligence community. Designed to detect a specific spectral 

manifestation of stress (the decrease in reflectance in TM band 4 and concomitant increase in 

reflectance in band 5), the transformation squares each pixel's intensity in band 4 and divides that 

by the pixel's intensity in band 5 (i.e., 42/5). The transformed data are displayed as a color 

composite image (42/5 as red, band 4 as green, and band 5 as blue). 

Table 7-1. Comparison of present work with that of Goldman and coworkers (1987). 

Technique/result Present work Goldman and coworkers 

Algorithm type Change detection Normalized difference vegetation 

index 

Data preparation Image registration Spatially filtering image sharpening 

Method of injury assessment Quantitative (algorithmic) Qualitative (visual) 

First detectable change 8 May 1986 (12 days) 16 June 1986 (51 days) 

Latest significant change 11 May 1987 (1 year) 15 October 1986 (5.6 months) 

59

Care must be exercised in the interpretation of such single-date composite images. For 

example, one interpretation of a 42/5,4,5 composite of the 11 May 1987 image of the Chernobyl 

area has vast new areas of forest beginning to show stress. Figure 7-1 shows this composite 

image. The area of stressed foliage indicated by our analysis (Figure 5-18) shows as a dark gray 

band in Figure 7-1. Johnson interprets the forest patches along the middle and upper left edge of 

the image as also being stressed. These purportedly newly stressed forested areas appear darker in 

the composite image than do other "unstressed" forest areas. 

This interpretation is based in part on the assumption that this pattern did not appear on any 

imagery dated before 11 May 1987 and that the darkened forest should appear the same as the 

undarkened. However, comparing the forest classification map (Figure 5-1), which was derived 

from the 6 June 1985 image, with the color composite in Figure 7-1 suggests that the pattern 

actually did exist a year before the accident. To make this comparison easier, the boundaries of 

forested areas classified as being similar to training site 4 are colored white; all other forested 

regions are classified as being similar to training site 3. Figure 7-2, showing this outline overlaid 

on the composite, indicates that these supposedly newly stressed areas seem to correspond to areas 

covered by a different forest type rather than to the radiation dose contours published by the 

Soviets. In other words, the apparent stress is likely an artifact of the analysis rather than an effect 

of the radiation release. Indeed, a similar composite image shown in Figure 7-3 derived from the 

preaccident image of 6 June 1985 shows the same darkening pattern as the 11 May 1987 

composite; therefore, the slightly dark appearing forest cannot be a new manifestation of radiation 

induced stress. 

Dose estimates based on our analysis are not presented in this interim report. Such estimates 

will depend on being able to relate the radiation dose either to the spectral manifestations of stress 

or to the times at which these manifestations first became detectable. Dose rate may complicate the 

relationship by affecting the spectral characteristics of injury as well as of the dose-response time. 

Also, the question of the relative proportion of beta dose versus gamma dose requires careful 

examination. These issues will be addressed in the final report for this project. 

60

Figure 7-1. Date: 11 May 1987, enhanced image; 42/5,4,5 false color. 

61

Figure 7-2. 

Boundaries of areas classified as similar to training site 4 overlaid on the 

11 May 1987 enhanced image; 42/5,4,5 false color. 

62

Figure 7-3. Date: 6 June 1985, preaccident image with the same image enhancement 

transformation shown in Figure 7-1. Darker forested areas in this enhancement 

correspond to a different forest type rather than to radiation damaged areas. 

63

SECTION 8 

CONCLUSION 

Perhaps the most significant conclusion that can be drawn from this work is that remotely 

collected, low-resolution multispectral data can be used to identify radiation stressed foliage and to 

monitor it quantitatively. Furthermore, this capability is aided by the availability of a sensitive 

change detection algorithm, Hyperscout, that can detect this stress very early and reliably. Early 

detection is critical for accurate dose estimates. 

When the data are in final form, a number of important results are anticipated. The first of 

these are estimates of the doses received by indigenous plant life. Secondly, the time history of the 

multispectral manifestations of various levels of radiation stress can be obtained, thus providing the 

first large-scale spectral measurements of the effects of a full range of radiation doses on conifers. 

This database would facilitate monitoring future accidents or even determining whether an accident 

had occurred. Such a database is not readily available from any other source. 

Much work remains to be done. The most important task remaining is to estimate radiation 

dose from the foliage response. Once the stress maps are finalized, the spectral changes 

corresponding to various levels of stress will be determined. These spectral histories, in turn, will 

be used to estimate the health of the conifers. Finally, these health histories will be used to infer 

dose estimates. 

Once dose estimates have been obtained, the process can be inverted. That is, the temporal 

aspects of the spectral manifestations of various levels of radiation stress can be extracted from the 

data. 

Continued monitoring of the area is recommended for at least two reasons. First, the 

identification of new areas showing stress would expand the area for which dose estimates are 

available. This new dose information might be incorporated to refine the estimates of dose levels 

received by human populations. Second, but not unrelated, lightly stressed forests should be 

monitored for evidence of recovery. Again, the primary goal is the refinement of dose estimates. 

Of equal importance is the collection of spectral information not available from other sources. 

Obtaining accurate dose estimates is the primary goal, and accurate determination of the dates 

of various stages of injury to the conifer community is required to obtain these estimates. There 

are available no high quality TM data for the period extending from 8 May 1986 to 24 May 1986, 

nor from 31 May 1986 to 15 October 1986. Yet, most of the spectral manifestations of the stress 

appear to have occurred during these two time intervals, while there seems to have been little 

change between 24 May 1986 and 31 May 1986. Because of the lack of suitable TM images, the 

question arises as to whether any other sources of multispectral data can be used to fill in these two 

64

critically important gaps. The algorithm used in this work can detect and map stress using data 

from any sensor, provided the data are sufficiently complete. Thus, both SPOT and MSS data 

should be considered potentially useful for filling in these gaps, provided that the analysis requires 

neither SWIR nor LWIR data. 

Another potentially useful approach to determining stress is to use TM data that was acquired at 

night. Night data should be much more sensitive to thermal stress. Thermal stress may be 

detectable much earlier than other forms of stress. If so, it could prove extremely valuable for 

detecting and monitoring the early stages of stress and, thus, for estimating dose. 

There are also two modifications in processing that should be considered. First, instead of 

using the same reference image for all analyses, as was done in this study, the reference date 

should be chosen to correspond to the season of the image being analyzed. This may significantly 

improve the accuracy of the resulting stress map. For example, the 15 October and 2 December 

1986 dates would be referenced to the 21 March 1986 date instead of the 6 June 1985 date. 

Unfortunately, snow cover present in the 21 March 1986 image may have a negative impact. 

The second processing modification involves using the first three components of the Tasselled 

Cap transformation (Crist, Laurin, and Cicone, 1986) instead of bands 7, 4, and 3. The fourth 

component, usually referred to as "Haze," might also allow the automatic elimination of distracting 

areas of apparent change attributable to cloud cover or haze. 

Extensions or alternatives to current algorithms should be investigated in the hope that even 

more sensitive algorithms may be found. One alternative is the use of neural networks to identify 

stress areas. Another promising area to explore is the simultaneous investigation of several 

(instead of just two) dates of imagery. This multitemporal processing might significantly improve 

the accuracy of the stress maps by removing the effects of noise and less than perfect registration. 

The final kinds of potential improvement are aesthetic. Some such kinds of improvement 

include the removal of areas contaminated by cloud, cloud-shadow, or haze from consideration, 

because these areas spuriously produce high stress values and confuse the presentation. In like 

manner, measures could be developed to the remove edge effects and to reduce noise by mode or 

other filtering. 

65

SECTION 9 

REFERENCES 

Asmolov, V.G., et. al., 1987 

"The Accident at the Chemobyl Nuclear Power Plant: One Year After," International Atomic 

Energy Agency, IAEA-CN-48/63. 

Crist, E.P., R. Laurin, and R.C. Cicone, 1986 

"Vegetation and Soils Information Contained in Transformed Thematic Mapper Data," 

Proceedings of IGARSS 1986 Symposium, Zurich, 8-11 September 1986, pp.1465-1470. 

Engel, J., 1984 

"Thematic Mapper (TM) Instrument Description," LANDSAT-4 Science Investigations 

Summary, Vol. 1, NASA Conference Publication 2326, pp. 41-61. 

Estes, J. E., Ed., 1983 

Manual of Remote Sensing, Vol. II (American Society of Photogrammetry), pp. 1511-1513 

and 2136-2187. 

Goldman, M., et aL, 1987 

"Radiation Exposure Near Chernobyl Based on Analysis of Conifer Injury Using Thematic 

Mapper Satellite Images," University of California, Davis, California 95616 (unpublished). 

Hoffer, R. M., 1978 

"Biological and Physical Considerations in Applying Computer-Aided Analysis Techniques to 

Remote Sensor Data, Remote Sensing: The Quantitative Approach, edited by P. H. Swain and 

S. M. Davis (McGraw-Hill,.New York) pp. 227-289. 

Irons, J., 1985 

"An Overview of LANDSAT-4 and the Thematic Mapper," LANDSAT-4 Science 

Characterization Early Results, Vol. II, NASA Conference Publication 2355, pp. 15-46. 

Johnson, A. J., 1989 

Central Intelligence Agency, Washington, D.C. 20505, private communication. 

Matson, M., and J. Dozier, 1981 

"Identification of Subresolution High Temperature Sources Using a Thermal IR Sensor," 

Photogrammetric Engineering and Remote Sensing, 47(9):1311-1318. 

Swain, P.H., 1978 

"Fundamentals of Pattern Recognition in Remote Sensing," Remote Sensing: The Quantitative 

Approach, edited by P. H. Swain and S. M. Davis (McGraw-Hill, New York) pp. 136-187. 

Westman, W.E. and C.V. Price, 1988 

"Spectral Changes in Conifers Subjected to Air Pollution and Water Stress: Experimental 

Studies," IEEE Transactions on Geoscience and Remote Sensing, 26(1): 11-21. 

Whicker, F.W., and L. Fraley, Jr., 1974 

Effects of Ionizing Radiation on Terrestrial Plant Communities, Vol. 4 of Advances in 

Radiation Biology, edited by J.T. Lett, H. Adler, and M. Zelle (Academic Press, New 

York). 

66

Wiebelt, J.A., and J.B. Henderson, 1976 

Techniques and Analysis of Thermal Infrared Camouflage in Foliated Backgrounds (U.S. 

Army Mobility Equipment Research and Development Command), AD A038186. 

Wolfe, W. L., and G. J. Zissis, eds., 1978 

The Infrared Handbook (Environmental Research Institute of Michigan), p. 3-140. 

67


DNA-TR-92-37-V2 



ATTN: TTV 


2 CYS ATTN: TTV 3416TH TTSO 


ATTN: BHS 


ATTN: DIRECTOR ATTN: C-332 

ATTN: EXH 

ATTN: JNSV 

ATTN: MRA 

ATTN: PHY 


ATTN: RSD 

ATTN: ICAF TECH LIB 




ATTN: LIBRARY 





ATTN: NUC FORCES & ARMS CONTROL 







ATTN: F HEGGE 


PROGRAM ANALYSIS & EVALUATION 

ATTN: DB 

ATTN: NAVAL FORCES 

5 CYS ATTN: DB-4 RSCH RESOURCES DIV ATTN: STRATEGIC PROGRAMS & TNF 

ATTN: DB-5C 


ATTN: DB-6B 


ATTN: DB-6E 


ATTN: DIAs'VPA-2 

ATTN: DIW-4 


ATTN: DN 

ATTN: JKAC 

ATTN: DT 

ATTN:JLT 


ATTN: JPEP 

ATTN: OS 



DEFENSE INEIGCE CATTN: 

ED30 J-3 STRATEGIC OPS 

ATTN: 

DIV 

DIC/RTS-2 

ATTN: J-3 SPECIAL OPERATIONS 

ATTN: DIC/2C 

ATTN: J-8 S CONLIN 


ATTN: JAD/SFD 


ATTN: JSOA 


U S EUROPEAN COMMAND/ECJ-6-DT 

ATTN: CID 

ATTN: ECJ-6 



ATTN: NANF 

ATTN: ECJ2-T 

ATTN: NASF 

ATTN: OPNA 

U S EUROPEAN COMMAND/ECJ3-CCD 

ATTN: OPNS 

ATTN: RAEE 

ATrN: ECJ-3 

20 CYS ATTN: RARP U S EUROPEAN COMMAND/ECJ4-LW 

2 CYS ATTN: TITL ATTN: ECJ4-LW 



2 CYS ATTN: DTICIFDAS ATTN: ECJ5-N 


ATTN: FCPR 

ATTN: FCPRT 



ATrN: FCNM 

U S EUROPEAN COMMAND/ECJ7/LW 

ATiN: ECJ-7 LW 


ATTN: DUSP/P 

ATTN: USD/P 

Dist-1









ATrN: SLCH-NW-RS JOSIP SOLN 

ATTN: SLCIS-IM-TL 


ATTN: J-2 

ATTN: J-5 

NUCLEAR EFFECTS DIVISION 









U S ARMY MATERIAL COMMAND 

ATTN: DRCDE-D 


ATTN: MONA-NU 

4 CYS ATTN: MONA-NU DR D BASH 


ATTN: STECS-NE 


ATTN: LIBRARY 



ATTN: BEHAVORIAL SCI & LEADERSHIP 







USACACDA 





ATTN: SLCBR-D 

ATTN: PMS-423 


ATTN: SEA-06GN 


ATTN: SLCBR-VL-I DR KLOPCIC 

MARINE CORPS 

ATTN: SLCBR-VL-V 


ATTN: PSIG RASP 


ATTN: JCOHN GERDES 


ATTN: MIKE STANKA ATTN: CODE 9642-B 











U S ARMY FIELD ARTILLERY SCHOOL ATTN: CODE 1240 

ATTN: ATSF-CD ATIN: CODE 2627 



ATTN: AF-OPTS ATTN: CODE F-31 

ATTN: G RIEL 


ATTN: C WARD 

U S ARMY INFANTRY CENTER 


U S ARMY ITAC 

ATTN: IAX-Z 

U S ARMY LABORATORY COMMAND 






ATTN: CODE E-1 1 



ATTN: LIBRARY 




Dist-2





NUCLEAR WEAPONS TNG GROUP, PACIFIC 

ATTN: CODE 32 




ATTN: D STROTTMAN 


OFFICE OF CHIEF OF NAVAL OPERATIONS 

MARTIN MARIETTA ENERGY SYSTEMS INC 

ATTN: NIS-22 

ATTN: B SANTORO 

ATTN: NOP 06D 

ATTN: G KERR 

ATTN: NOP 403 

ATITN: J WHITE 

ATTN: NOP 50 


ATTN: NOP 60 

ATTN: NOP 60D 


ATTN: NOP 603 ATTN: TECH LIB 3141 

ATTN: NOP 91 

ATTN: OP 654 


ATTN: PMS/PMA-423 


OPERATIONAL TEiT & EVALUATION FORCE 


PLANS, POLICY & OPERATIONS 

ATTN: CODE-P 





ACADEMY LIBRARY DFSELD 

ATTN: LIBRARY 

AFIS/INT 

ATTN: INT 




ATTN: AUL-LSE 

ATTN: LIBRARY 

ATTN: COUNTER TERRORIST GROUP 


ATITN: MEDICAL SERVICES 

ATTN: NIO-T 




ATTN: G ORRELL NP-CP 

ATTN: R SANDS 


ATTN: PM/STM 



ATTN: S YANIV 


ARES CORP 

ATTN: A DEVERILL 


ATTN: F GREY 


2 CYS ATTN: AF/SAMI HORIZONS TECHNOLOGY, INC 

ATTN: J MARSHALL-MIES 


ATTN: SAF/AQA 

DEPUTY CHIEF OF STAFF PLANS & OPERS 

ATTN: AFXOOSS 

FOREIGN TECHNOLOGY DIVISION 

ATTN: CCN 

ATTN: SDA 


ATTN: DASIAC 


ATTN: R STOHLER 


ATTN: DASIAC 

TACTICAL AIR COMMAND/XPSC 

ATTN: TAC/DOA 


ATTN: RADIATION SCIENCES DIV 



ATTN: DR T JONES 





ATTN: DOUGLAS C YOON 



Dist-3



ATTN: R LAUGHERY 

MISSION RESEARCH CORP 

ATTN: DR NEIL GOLDMAN 


ATTN: B BENNETT 

ATTN: D BAREIS 

ATTN: FOSTER 

ATTN: J MCGAHAN 



2 CYS ATTN: G ANNO 

12 CYS ATTN: G E MCCLELLAN SCIENCE APPLICATIONS INTL CORP 

ATTN: H BRODE 


2 CYS ATTN: R N DEWITT 

2 CYS ATTN: T H HEMMER SCIENCE APPLICATIONS INTL CORP 



ATTN: D GORMLEY 


2 CYS ATTN: G MCCLELLAN ATTN: S LEVIN 


ATTN: D KAUL 

ATTN: E SWICK 

ATTN: L HUNT 


ATTN: W WOOLSON 



TRW SPACE & DEFENSE SECTOR 

ATTN: DR BRUCE WILSON 



Dist-4

AD-A260 167 

A lexandria, VA 22310-3398 l,* 


Alexandria, VA 22310-.3398 

DNA-TR-92-37-V3 

Chernobyl Doses, 

Volume 3-Habitat and Vegetation Near the Chernobyl 

Nuclear Reactor Station 

DTIC 

~ ELECTF. 

Elizabeth L. Painter 

i IN•9 199EIF 

F. Ward Whicker JAN % 93f 

Pacific-Sierra Research Corporation I U 

12340 Santa Monica Boulevard C 


January 1993 



I 

Approved for public release; 

distribution Is unlimited. 

93-01259








ORGANIZATION. 

*N4o


This mailer is provided to enable DNA to maintain current distribution lists for reports. (We would 


o Add the individual listed to your distribution list 

o Delete the cited organization/individual, 

o Change of address. 

NOTE: 

Please return the mailing label from 

Fthe document so that any additions, 

changes, corrections or deletions can 

be made easily. 

NAME: 

ORGANIZATION: 

OLD ADDRESS 


Z 

TELEPHONE NUMBER: ( ) 

cc DNA PUBLICATION NUMBER/TITLE CHANGES/DELETIONS/ADDITIONS, etc.) 

0' (Attach Sheot f more Space is fequftd) 

Z 

UJ 

WI 

I'- 

• : DNA OR OTHER GOVERNMENT CONTRACT NUMBER: 




SIGNATURE:


ATTN: TITL 




ATTN: TITL 




Form ppoved• 

I OMB No. 0704-0188 

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. 

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 

l Cbibimaion. .c.ud.lig t sggeanime r• g V1 lulden W i hlin Hedqua•rer Soei Diectorate fr mnborrmabon Operalns arid Reports. 1215 Jefterson 

Datlighway. Suite 1204. Arbingtorn. VA 220-40. gird W Vie Office of Managementl 

and Budget. Paperwor Reduction Proiect (0704-oiU). Washingtion. DC 20603 


1 930101 Technical 870929- 920228 


Chernobyl Doses C - DNA 001-87-C-0104 

Volume 3-Habitat and Vegetation Near the Chernobyl Nuclear Reactor Station PE - 62715H 

6. AUTHOR(S) PR -RM 

TA -RH 

Elizabeth L Painter and F. Ward Whicker 

WU - DH026130 


REPORT NUMBER 


12340 Santa Monica Boulevard PSR Report 2251 





6801 Telegraph Road DNA-TR-92-37-V3 


RARP/Young 



STRP 3500A 25904D. 

12a. DISTRIBUTION/AVAILABIUTY STATEMENT 12b. DISTRIBUTION CODE 



This volume presents a detailed exposition on the soils, climate, and vegetation of the Poles'ye region of 

Ukraine and Belorussia with emphasis on the area around the Chernobyl Nuclear Reactor Station. This data 

provides background for interpretation of multispectral satellite imagery of the area. Volume I uses these 

images and the information of this report to analyze the radiation response of the canopy of the coniferous forests 

in the immediate vicinity of the reactor station after the accident of 26 April 1986. 


Chemobyl Forest Damage Landsat 90 

Change Detection Conifer Stress Fallout 16. PRICE CODE 

Ionizing Radiation Multispectral Imagery 

17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. UMITATION OF ABSTRACT 



NSN 7S0-220-6600 

Form 296 (Rv.2-B19) 

U NISN byI A Va. 

290-102

UNCLASSIFIED 

SECVWV CA3S W•TE *% OF ThIS MOE 





i fl•, OAWIFUAl OF TM MW 

UNCLASSIFIED

PREFACE 

This volume is the third in a series of three volumes composing the final report to the Defense 

Nuclear Agency (DNA) for contract DNA001-87-C-0104, Chernobyl Doses. It was prepared at 

Colorado State University from locally available literature under a consulting agreement with 

Pacific-Sierra Research Corporation (PSR). It was generated as a topical report for that contract, 

but is being published as a volume of the final report. It provides a detailed exposition on the 

soils, climate, and vegetation of the Poles'ye region of the Ukraine and Belorussia with emphasis 

on the area around the Chernobyl Nuclear Reactor Station. This data provides background for the 

interpretation of satellite imagery of the area. Volume 2, Conifer Stress Near Chernobyl Derived 

from Landsat Imagery, describes the acquisition and processing of satellite multispectral imagery 

of the area containing the Chernobyl Nuclear Reactor Station and presents the exploratory analysis 

of the imagery using PSR's proprietary Hyperscout m change detection algorithm. Volume 1, 

Analysis of Forest Canopy Radiation Response from Multispectral Imagery and the Relationship 

to Doses, presents the analytical work that connects these multispectral observations of pine forests 

in the images to the nuclear radiation dose received by the trees as a consequence of the reactor 

accident of 26 April 1986 and summarizes other work conducted for the contract. 

ST ! S 

6t&Ltt 

Wt4 flit 0 

Uzstuaoji* JUD__Strtdit on/ jaw/ 

Availability Codes 

[Avail sad/or 

rDist 

Special



TO Convert Proma TO multiply 

angstroin meters 1m) 1.000 000 X E-10 

atmosphere (normal) kilo pascal (kPa) 1.013 25 X E+2 

bar kilo pascal (kPa) 1.000 000 X E+2 

barn meter' (in2) 1.000 000 X E-28 

British Thermal unit (tbrobnia) joule WJ 1.054 350 X E+3 

calori (thermochemical) joule (J) 4.184 000 

cal (tbermowmchemcl)/cm2 mega jDUle/M2(Mj/m2) 4.184 000 X E-2 

curse gi4p becquerel (G~qr- 3.700 000 X E+I 


degree Fahrenheit degree kelvin (K) t 3 1 u(tf.+ 459.67)/ 1.8 

electron volt joule(JW 1.602 19 XE-19 

asg joule (J) 1.0O00O00xE-" 

erg/second watt MW 1.000 000 X "- 

foot meter (M) 3.046000 X E-1 

foot-pound-force joule (J) 1.3558616 

gallon (U.S. liquid) meter 3 EM3) 3.765 412 X E-3 

inch meter 1m) 2.540 000 X E-2 

jerk joule (J) 1.000 000 X E+9 

joule/kilogrm (J/Kg) (radiation dose 

absorbed) Gray 10y) 1.000 000 


kip (10001b1) newton (N) 4.448 222XEZ+3 

hip/Inch 2 (kal) kilo pascal (kPa) 6.694 757 X E.3 

ktap newton-second/rn' (N-s/rn') 1.000 000 X E+2 

Micron meter (m) 1.000 000 X E-6 

mil meter (m) 2.540 000 X "- 

mile (International) mete 4m) 1.609 344 X +3. 

ounce kilogram (kg) 2.634 952 X 3-2 

pound-force (bf avoirdupois) newton (N) 4.448 222 

pound-force inch newton-meter Min) 1.129 648 XE-1 

pound-frc/inch newton/meter Motr) 1.751 266 X E.2 

pound-force/foot' kilo pascal (kPa) 4.766 026 X E-2 

pound-forc/inch' (psi) kilo paacal (kPa) 6.694 757 

pound-mass MIm avoirdupois) kilogram (kg) 4.535 924 X E-1 

pound-mass-lbot2 (xmomet of inertial kilogram-eter' (kgrn'3 4.214011l X E-2 

pound-rnaslhst kilogram/meter' (hg/rn') 1.6018464 X E+1 

rdaflt " w don absorbed) Gray EGyr' 1.000 000 X E-2 

roenie.coulomb/kiogram (C/hg) 2.579 760 X E-4 

sakeh second (s) 1.000 000 X E-8 

48kilogram (kg) 1.459 390 X E3.1 

eory (mmo HS, 0C) kilo pascal (kPal 1.333 22 X E- I 

*Ybe b soquere (Dq) Is tde 91 unit of radloac~tvty Sp a 1 event/s. 

I-Tbe Grey (Qy) in the 81 unit of absorbed radiation. 

iv


Section 

Page 

PREFACE ................................................................................. iim 

CONVERSION TABLE ............................................................... iv 

1 INTRODUCTION ....................................................................... 1 

2 GEOGRAPHY ........................................................................... 2 

3 SOILS ..................................................................................... 4 

4 CLIMATE ................................................................................. 7 

5 VEGErATION ........................................................................... 9 

6 SUCCESSION ......................................................................... 19 

7 BOG DRAINAGE AND RECLAMATION ....................................... 21 

8 CROPS .................................................................................. 22 

9 REFERENCES ........................................................................... 23 

Appendices 

A TREES AND SHRUBS ............................................................... A-1 

B 

HERBS .................................................................................... B-i 

C CROPPLANTS .......................................................................... C-1 

v

SECTION 1 

INTRODUCTION 

This report was prepared to support the use of satellite multispectral imagery to detect changes 

in the vegetation near the Chernobyl Nuclear Reactor Station that were caused by the release of 

radioactive material during the accident of 26 April 26 1986. It is based on a review of literature 

available locally at Colorado State University. 

The review covered information on the Poles'ye region of the Ukraine and Belorussia, 

republics of the former Soviet Union. This area contains the 30 kilometer danger zone that was 

evacuated in the aftermath of the accident. The information is intended to aid in the interpretation 

of features and vegetation types in the imagery of the area within the danger zone. 

Section briefly describes the geography, Section 3 the soils, and Section 4 the climate of the 

Poles'ye, with emphasis on the area near the reactor station. Section 5 presents an extensive 

discussion of vegetation of the region. Sections 6, 7, and 8 describe succession, bog drainage, 

and typical crops, respectively. Appendixes A. B, and C provide descriptions of the more 

important species of trees and shrubs, herbs, and crop plants, respectively.

SECTION 2 

GEOGRAPHY 1 

The Chernobyl Nuclear Reactor Station (NRS) and much of the area within 30 kilometers (kin) 

around it are located in the Poles'ye 2 region of the Ukrainian and Belorussian 3 Soviet Socialist 

Republics. The power plant itself and the cities of Pripyat 4 and Chernobyl are in northeastern 

Kievskaya Oblast, in the interfluve of the Pripyat and Teterev Rivers, in the Ukrainian SSR. It lies 

at about 51 0 12'N longitude 30 0 8'E latitude. 

The Poles'ye lies in the Pripyat River basin and in part of the Dnepr 5 River basin and includes 

the area called the Pripyat marshes or bogs. It is a vast lowland (more than 13 million hectares) on 

a large, morainal outwash and outwash-alluvial plain on the Russian platform. It is bounded on 

the north, east, and south by greater rises in elevation than can be found within the region itself. 

The area has a very nonhomogeneous geological structure. From west to east, it occupies the 

various geostructural regions of the Russian platform: the Galician-Volyn lowland, the Ukrainian 

crystalline shield, the Dnepr-Donestsk depression, with the east approaching the Voronezh 

crystalline shield. 

The terrain is almost uniformly low, broken only by low moraine hills, sandy knolls or 

mounds (usually old dunes), and slightly elevated plains. The relief from the center to the edges of 

the basin is only 50 to 150 meters (m). Some areas possess eolian relief, often in the form of 

parabolic, west-facing dunes, but maximum heights within the Poles'ye are only 200 m. The 

interfluve of the Pripyat and Teterev Rivers has an unusual morainal-hilly relief with some 

relatively deep erosion. Poorly drained sandy plains are located between the moraine hills. 

The Poles'ye is often divided into three sections: the western, the central (or right bank-on 

the right bank of the Dnepr River), and the eastern (or left bank). In the Ukraine, the central 

section is divided into the Kiev Poles'ye and the Zhitomir Poles'ye. The Kiev Poles'ye lies on the 

1 Geological information has been synthesized from Keller 1927, Berg 1950, Akademiya Nauk SSSR 1963, Lydolph 

1964, Fridland 1976, Golovina et al. 1980, Lysenko and Golovina 1982, USSR 1987. 

2 polesye, Poles'e, Poles'ya, Polesie, Polessie, Polesiye, Pollessic lowlands, Polyesye (Where there are multiple 

spellings of a Russian name in English. the one most frequently encountered in Soviet literature has been used. All 

others are listed in a foomote.) 

3 Byelorussian or White Russian. 

4 pripyan. Prip'az. Pripiat, Pripet. 

2

middle Dnepr slope, in the interfluve of the Pripyat and Teterev Rivers, extending east to the Dnepr 

River, and includes the Chernobyl NRS and the area immediately around it. The Zhitomir is west 

of the Kiev Poles'ye. 

3

SECTION 3 

SOILS 6 

Glaciofluvial plains are formed by deposition from thaw waters of glaciers. They consist of 

outwash and broken-up moraine materials. The soils in the Poles'ye formed in a humid climate on 

glaciofluvial plains from a blanket of unconsolidated Quaternary outwash deposits of sand and 

loamy sand over- and underlying a terminal glacial moraine. Glaciofluvial sediments are sandy or 

gravelly-pebbly stratified sediments of flowing glacial waters. Outwash plains are relatively 

smooth sandy surfaces deposited near the edges of continental ice. The thickness of these deposits 

range from 1.5 m in the southeast to 150 m in the northwest Moraines are deposits of continental 

ice, consisting mostly of unstratified, unsorted, nonuniform materials of varying mechanical 

composition. The Poles'ye terminal moraine is mostly loamy sand. Medium and clay-loam lake 

deposits are rare in the Poles'ye and loesses, even rarer. 

The parent materials are course-textured, very bouldery, and gravelly; fine particles are usually 

leached or washed away. In the Kievskaya Oblast, glacial waters were active for a long time 

during the formation of the parent material. The soils generally lack carbonates in the parent 

materials. Most of the soils are at least periodically waterlogged (with spring and autumn floods, 

and sometimes after prolonged summer rains). 

Most of the Poles'ye soils are poor in humus; soils in the western and central Poles'ye often 

have I to 1.5 percent humus and can have less than 1 perceat. The soils are mostly noncalcareous. 

They generally are acid, with a pH of 4.5 to 5.5, and are low in available nutrients, including 

boron, zinc, phosphorus, potassium, and nitrate. 

The soils in the Poles'ye are podzolic, 7 peaty bog, and peaty-gley meadow soils. More than 

75 percent of the soils are meadow or bog soils. Soils on floodplains may be bog, meadow, or 

podzols. Parent materials around the Chernobyl NRS is from the Ukrainian crystal shield and is 

mainly sands or sandy-loams. Many of the soils in the northeastern part of the Kiev Poles'ye, in 

the interfluve of the Pripyat and Teterev Rivers, are weakly podzolic (frequently gleyed) sods. 

6Descriptions, distributions, and definitions of soils have been synthesized from Sukachev 1928, Berg 1950, 

Vilenskii 1960, Lydolph 196%, Fridland 1965, Symons 1972, Brady 1974, Krupskiy et al. 1970, Walter 1978, 

Smith 1980, Golovina et al. 1980, Oleynik 1981, Lysenko and Golovina 1982, USSR 1986, USSR 1987. 

7 The soil classification system found in the Soviet literature surveyed for this report is not the system currently 

being used in much of the western world. Podzols have been placed in a group called "spodosols," most bog soils in 

"histosols," and most gley soils in "inceptisols" (Brady 1974, Smith 1980). This report continues to use the 

classification used in the USSR. Many of the old soil names come originally from Soviet soil literature. 'Podzol" 

is a Russian word, and "gley" is Ukrainian. 

4

The soils of the area frequently are flooded or contain excess soil water, at least part of the 

year. There are three sources of excess soil water in the Poles'ye: (1) in the spring and often in 

the autumn from stream overflow; (2) from a groundwater table that is very close to the soil 

surface; and (3) from high amounts of atmospheric precipitation. The banks of the streams and 

floodplain terraces are both very low. In the spring and autumn, and after heavy summer rains, 

streams overflow onto floodplain terraces and into interfluvial moraine plains between neighboring 

streams, and water from one stream sometimes passes into another. 

Much of the area around the Chernobyl NRS is poorly drained and contains a large block of 

low-lying wetlands with bog soils. Because bog soils receive excessive moisture for the greater 

part of the year, are sometimes covered with shallow water, and have poor drainage, they become 

peaty. 

Podzols have a very thin organic layer on top of a gray, leached eluvial layer, which in turn 

overlies a dark brown horizon. Podzols develop in cool, moist climates under coniferous forest. 

Podzolization occurs when rainfall in forests is sufficient to carry away many elements, especially 

calcium, magnesium, potassium, iron, and aluminum. Becuase the needles of coniferous trees are 

acid, conifers generally return insufficient bases to the surface soil, which becomes acid. The 

distribution of podzols is patchy in the Poles'ye. 

There are large quantities of weakly podzolic sandy soils ("bor sands"). ("Bor" is a pine 

woodland on poor, sandy soil.) Dry bor sands are found on eolian and moraine hills in the 

Poles'ye. 

Meadow soils are derived from bog or podzolic soils and are also called "halfbog," "turfpodzol," 

or "sod-podzol" soils. "Half-bog" soils develop when herbaceous plants and/or 

deciduous trees grow on drained bogs. These are often found under wet meadows. Deciduous 

trees growing on podzols add nitrogen and other minerals. An herbaceous understory can 

develop, transforming the upper horizons of podzols into "sod-" or "turf-podzols." Such soils are 

found under fresh meadows. Soddy, moderately podzolic soils have formed on moraine outcrops 

in the Kiev Poles'ye. 

When forests are cleared but not adequately drained, the podzols deteriorate. Deterioration is 

marked by formation of gley horizons. Gleyed soils include gley, gley-podzol, and sod-gley soils. 

Gleization occurs when the iron in soils where water is at or near the surface is usually reduced to 

ferrous compounds, having a gray or bluish color. These soils therefore have a more or less dense 

but not unusually viscous layer of loamy or clayey material of a gray color. Gley are also found in 

seasonally waterlogged sites. Sod-gley soils are found in meadows that have become boggy; areas 

with gleyed soils often become sphagnum bogs. 

Some parts of the Poles'ye contain "islands" of soils foreign to it. The are small loess islands 

in both the Zhitomir Poles'ye to the west of the Kiev Poles'ye and the left bank to the east. One 

5

such island occurs on the Ovruch ridge (about 90 km west of Chernobyl). This ridge is 320 m 

above sea level and 60 m above the surrounding lowland. 

6

SECTION 4 

CLIMATE 8 

The Poles'ye has a humid temperate-continental climate, influenced by maritime, continental, 

and local factors. 

Bogs, which are common around the Chernobyl NRS, influence the microclimate. These areas 

have high humidity, low temperature minima, evening and morning mists close to the soil surface, 

and frequent frosts. 

Forests, which are common, also influence the microclimate. The air temperatures are lower 

than in the open; the daily temperature ranges are smaller, the snow cover is less thick; and winds 

are stilled. 

The moisture from the Baltic Sea and the "great valley" region of Poland influences the 

Poles'ye. Because there is unrestricted passage of marine winds, unhindered by major land relief, 

there is considerable marine influence on the climate. 

A small local maximum of relative humidity, caused by the intensified evaporation of water 

from wetlands, is noticed over the Poles'ye wetlands. The minimum relative humidity at 1 p.m. in 

May is 50 to 55 percent. The mean humidity for June, July, and August is 56 to 60 percent. 

The Poles'ye has hot summers and relatively mild winters. Mean temperatures in the region 

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 

minimum temperatures are -30 to -35 °C in January, while summer maximums range up to 40 0C. 

The growing season (days with mean temperatures above 5 °C) is about 160 days. The 

growing season in bogs begins 8 to 10 days later than in the surrounding nonbog habitat and ends 

10 to 12 days earlier. 

Summer precipitation exceeds winter by a factor of 2. In any season, there is precipitation 

every 2 to 3 days. In the warm season (April to September), the amount of precipitation in the 

interior of the European USSR is much greater (350 to 500 millimeters [mm]) than on the coasts 

(200 to 300 mm). During the cold half of the year (November to March) in the central belt of 

European USSR, the precipitation amounts are 100 to 300 mm. 

A precipitation maximum is situated on the Pripyat and the upper reaches of the Dnepr and 

Western Dvina Rivers. The maximum precipitation in the Pripyat basin is 680 to 695 mm/year. 

Precipitation diminishes eastward from the Pripyat. In the specific region of interest, mean 

precipitation is 500 to 600 mm/yr, with around 180 days/year recording precipitation. Two-thirds 

81nfmrmation about the climate has been synthesized from Kendrew 1942, Berg 1950, Anonymous 1962, Borisov 

1965, Szafer 1966, Skoropmnov 1968, Lysenko and Golsvina 1982, USSR 1987. 

7

of the precipitation is received in the warm season (April to September). The mean intensity of 

precipitation amounts to 8 to 10 mm/hr in the central belt of the country (51 to 59 0 N). 

At 50°N by the second 10-day period in November snow cover is continuous. In western 

European USSR, snow cover is usually continuous by the last 10 days in October or the first 10 

days in November. At 55ON 300E, there is an average of 2.5 temporary snow covers before the 

onset of winter. In central European USSR, winter temperatures and precipitation are highly 

variable. Mean maximum snow depth is 10 to 30 centimeters (cm). Continuous snow cover 

generally lasts about 80 days. Rivers and streams are generally frozen for about 100 days/year. 

8

.. SECTION 5 

VEGETATION 

On a global scale, the Poles'ye is in the Eastern European vegetation province, whose 

northern, eastern, and southeastern boundaries correspond to the distributions of Quercus robur, 

Acerplatanoides, and Corylus avelkana (Takhtajan 1986). In the Flora of the USSR, the Poles'ye 

is in the Upper Dnepr floral region (Komarov 1968). According to Lydolph (1964), Poles'ye 

means "woodland" and much of the natural vegetation was once pine-dominated woodlands. 

About 30 percent is still forested. 

Habitats in the region are classified by how wet they are and whether the vegetation includes 

trees. The basic plant community types in the region are wetlands, meadows, carts, woodlands, 

and forests. 9 ,1 0 , 11 These may grade into one another. In the Chernobyl district, natural 

vegetation (mostly forest, woodlands, and wetlands) covers nearly 50 percent of the terrain (USSR 

1987). This does not include the meadows that have been created from draining mires or felling 

forests, in which the natural vegetation is often augmented with pasture grasses and legumes. In 

all, only one-third of the landscape is croplands. 

Much of the Kiev Poles'ye is floodplain terraces. Floodplains are usually divided into the area 

adjacent to the streambed (the streamside zone), a central zone, and an "upper" zone. The 

streamside zone has the thickest silt deposit and relatively high mesorelief (depressions and 

ridges1 2 ). Ridges in this zone are usually 1 to 10 m high and depressions awe quite shallow. The 

central zone is usually undulating, with low mesorelief and shallow groundwater. Ridges in the 

central zone usually are not flooded and may even experience some drought. Depressions are 

9 While the vegetation types are specific to the Poles'ye, the species lists are those known to occur in that type of 

vegetation in northen Ukraine, southern Belorussia, andlor adjacent areas in Poland. 

10 [)finitions and descriptions of habitat, vegetation types, succession, bog drainage, and reclamation have been 

synthesized from Keller 1927, Berg 1950, Vilenskii 1960, Akademiya Nauk SSSR 1963, Viktorov et al. 1964, 

P'yavchemko 1965, Szafer 1966, Skmpanov 1968, Stankevich and Rubin 1968, Remezov and Pogrebnyak 1969, 

Ivanov 1972, Fridland 1976, Walter 1978, Komamv 1980, Smith 1980, Moore 1984, USSR 1987. 

1 llnformation on the flora of vegetation types and descriptions on the species lists was synthesized from Keller 

1927. Sukachev 1928, Wulff 1943, Berg 1950, Vilenskii 1960, Larin 1962a"b, 1964, Akademiya Nauk SSSR 1963, 

Komarov 1963, 1964, 1967, 1968, 1970, 1971a,b, Tutin 1964, 1968, 1972, P'yavachenko 1965. Szafer 1966, 

Reme and Pogrebnyak 1969, Ivanov 1972, Bialobok and Zelawski 1976, Leathart 1977, Phillips 1978, Walter 

1978, Hora 1981, Oieynik 1981, Moore 1984, Takhtajan 1986. USSR 1987. 

12 MorW e idgsa anl depressions ue gomocphic in origin and ae not synonymous with hummocks and hollows 

which ae formed by the vegetation and are much smaller in scale than ridges and depressin. There are some 

differnces in flood meakow species disribution elated to meorelief. Teme cam be found with the individual species 

in the species •ist that follow. 

9

flood-prone. The central zone can contain areas in which flooding is prolonged (greater than 40 

days), moderate (20 to 40 days), and short (fewer than 20 days), and areas that do not experience 

annual flooding; the vegetation in them varies accordingly. The "upper" zone is the lowest in 

absolute elevation and is usually flat and excessively wet. Fen, bog, meadow, and woodland 

vegetation may all occur on floodplains. Fens and carrs are usually found close to the stream. 

Transitional bogs are usually found in the central zone. Annually flooded meadows are found both 

close to the stream and in the central zone. Fresh meadows and moist forests, habitats that do not 

experience annual are usually in the central zone. Floodplain bogs, boggy wet forests, and wet 

meadows are found in the upper zone. 

There are also extra- (or supra-) floodplain areas. The interfluve of the Pripyat and Teterev 

Rivers is unusually morainal-hilly. There are poorly drained sandy plains are located between the 

moraine hills. Dry woodlands and forests are found on the moraine hills. Sphagnum bogs and 

wet forests occur on the poorly drained moraine plains and fresh meadows in better drained areas. 

Immense tracts of the Poles'ye are occupied by wetlands. The Russian word "botolo" is a 

general term for wetlands. It is translated in a number of ways, but the best may be mire: a 

peatland, 13 a wetland with considerable water retained by an accumulation of partially decayed 

organic matter, including wetlands with or without flowing water. There are a number of English 

terms used to describe different types of mires. Marshes are wetlands with emergent vegetation 

(i.e., plants with their leaves above water but their roots in soil that is covered [part or all of the 

time] with water). Marshes include both fens, swamps, and bogs. Fens are mires fed by moving 

water, either surface or ground water. Swamps are fens with only surface water influence. If 

woody vegetation is added to a fen, these communities are called carrs. Fens are sometimes 

called low moors or low moor bogs. The community closest to a stream is usually a fen or carr. 

Fens are also frequently found at the edges of lakes. Bogs are mires where water is supplied 

exclusively by atmospheric precipitation. The terms "high moor" bog and "raised" bog are 

sometimes used to distinguish true bogs from fens. Transitional bogs are intermediate between 

fens and bogs. The term "marsh" is sometimes reserved for transitional bogs. Most bogs are not 

flat but consist of hummocks, hollows, and intermediate sites.1 4 

131B geologcal terms, a mire must have a layer of peat at least 20 So 30 an deep (Walter 1978). "Moom" means 

mir in Gamm; a Ein SU it is used for tranutional bogs 

14 1ouauom Abou which Vpeda grow on hummocfs hollows, and intmeediaw sites can be found with the plant 

4in 

dt species IeL 

10

The dominant vegetation in sedge fens (swamp) is usually sedges, horsetails, and/or 

cottongrass. Plants 15 include Equisetuu arvenSe, 16 E. pahistre, E pratense, E. variegatum, Carex 

acutiformis, C. appropin quata, C. aquatilis, C. caespitosa, C. canescens, C. chordorrhiza, 

C. diandra, C. dioica, C. gracilis, C. heleonastes, C. lasiocarpa, C. limosa, C. panicea, 

C. paniculata, C. psuedocyperus, C. ripania. C. stellulata, C. vesicaria, C. vulpina, Cyperus 

Jlavescens, C. pal wnicus, Eriophorum angustifolium, E~ gracile, E. latifolium, Scirpus Lacustris, 

silvauicus, Junciss compressus, J. inflexus, J. tenageia, Alopecurus arwulinaceus, A. geniculatus, 

Glyceria aquasica, Leersia oryzoides, Phragmites commwuns, Poa palustris, Setaria glauca, and 

S. viridis. In colored aerial photographs, 17 fens appear smooth dark green (the darker, the wetter 

the mire). In black and white photos, they appear dark, either without outlines or with an 

elongated outline. Soils beneath sedge fens become peaty. 

Reed-bulrush fens (swamps) also have peat soil. The vegetation includes Phragmites 

communis, Scirpus acicudaris, S. eupalustris, S. holoschoenus, S. lacuster, S. mamillatus, 

S. ovatus, S. pauciflorus, S. sylvaticus, Beckmannia eruciformis, Calamagrostis neglecta, 

Glyceria aquauica, G. fruitans, G. hemoralis, Leersia 0' yzides, Phalaris arwzdinacea, Typha 

angustifolia, T. latifilia, Carex ekuta, C. omskiana, C. panicea, C. pseudo-cyperus, C. vesicaria, 

Juncus compressus, and J. tenageia. These usually sit in 50 to 100 cm of water, with plants 

anchored in a silty bottom. 

Soils in transitional bogs are also peaty. Vegetation includes Carex acutiformis, 

C. appropin quata, C. aquatilis, C. caespitosa, C. canescens, C. chordorrhiza, C. diandra, 

C. dioica, C. disticta, C. disperma, C. elata, C. fihiformis, C. gracilis, C. heleonastes, 

C. junceila, C. lauiocarpa, C. limosa, C. muricata, C. omskiana, C. pauciflora, C. pseudocyperus, 

C. riparia, C. vesicaria, C. vulpina, Eriopho rum angustifoblium, E. gracile, 

E. latifoblium, K. vaginatum, Scirpus lacustris, Juncus compressus, JI effusus, JI lampocarpus, 

I. eeruji, J. squarrosus, Alopecurus arundinaceus, A. geniculatus, Beckmannia eruciformis, 

Calamagrostis neglecta, Deschampsia caespitosa, Molirsia coerulea, Equisetum heleocharis, 

E. palustre, and L. variegatum. Sphagnum apiculatum, Sph. compactum, Sph. cuspidatum, 

Sph. fuscum, Sph. papillosum, Sph. recurvum, Sph. rubellum, and Sph. squarrosum are 

common mosses. Transitional bogs are often found in shallowwater (less than 1 in), where Alnus 

1 5 Commuumty type we gmueraL They are often subdvided int two to may mate detailed communities. Not all 

pheat musd far a comumity wfi neoeussay be found together, rahe they an found in one or more of the 

eommities of the type. Mit nondcuakum species order d=e no nesammily reflect umportace. 

1C mm 

for indvidua ueciet cm be found in the specie li& 

1 7 pbrdw iForo About the upince of vegetatio in saera photographs con be found with the descriptions in 

he Species lsL

glutinosa has been cut. Transitional bogs often contain shrubs, including Betulda humilus, Salix 

aurita, S. cinerea, S. lapponum, S. myrtilloides, S. phylicifolia, S. rosmarinifolia, Ledum 

palutre, and Chamaedaphne calyculata. Trees sometimes found in or at the margin of transitional 

bogs include A/nus incana, Betula pubescens, Pinus sylvestris, and Populus tremula. From the 

air, transitional bogs appear as a patchwork of dark green (wet peat), green (hollow), and yellow 

(hummock) patches, without sharp boundaries. Water appears as black pools. Tree crowns are 

visible in aerial photographs. 

Most of the bogs in the Chemobyl region are Sphagnum peat bogs, with or without trees. The 

common shrubs in sphagnum bogs include Andromeda polifolia, Betula humiUs, Calluna vulgaris, 

Chamaedaphne calyculata, Empetrum nigrum, Ledum palhssre, Rubus chamaemus, Vaccinium 

oxycoccus, V. uliginosum, and V. vitis-idaea. Common herbs include Carex limosa, 

C. chordorrhiza, C. heleonastes, Eriophorum vaginatum, Juncus bulbosus, J. squarrosus, 

Equisetum variegatum, and the carnivorous plant Drosera intermedia. The mosses include 

Sphagnum cuspidatum, Sph. fuscum, Sph.medium, Sph. recurvum, Sph. rubellum, Sph. 

russovii, Pohlia nutans, Polytrichum commune, and P. stricum. In aerial photographs, sphagnum 

bogs have parallelnarrow, dark zigzag lines on a lighter, brownish (sphagnum) background. In 

black and white photos, they are light to dark gray (with light dominant), or they appear to have 

sinuous light belts with dark intervals (hummocks and hollows). 

Sphagnum-pine bogs are the most common type of wooded sphagnum bog. Trees are not a 

principal vegetation component of the community. The species are a mixture of those found in 

sphagnum bogs and pine-sphagnum forests. Pines in bogs only reach 8 m in height. From the 

air, sphagnum-pine bogs have the same appearance as raised bogs; however, the crowns of trees 

are visible. 

Meadows are usually (1) wet meadows, (2) flood meadows, or (3) fresh meadows. 

Meadows may be transitional or ecotonal between wetlands and moist forest. In black and white 

aerial photographs, meadows appear uniformly light colored and usually have very definite 

boundaries. 

(1) Wet meadows are inundated for some time each year, especially late winter and early 

spring. The groundwater table fluctuates frequently but never drops very low. They often have 

peaty soils. Wet meadows often are secondary vegetation resulting from drainage of fens or bogs 

or when moist forest is removed. The vegetation includes Agrosuis alba, A. canina, A. stolonifera, 

Alopecurus arundinaceus, A. geniculatus, A. pratensis, Beckmannia eruciformis, Calamagrostis 

neglecma, C. phragmitoides, Catabrosa aquatica, Deschampsia caespitosa, Festuca arundinacea, 

F. ovina, F. pratensis, F. rubra, Glyceria aquatica, G. fruitans, G. plicata, Hierochloe odorata, 

Molinia oedruea, Nardw mnictv , Phajaris armudinacea, Poa palustris, Triseium sibiricum, Carex 

aquatiuis, C. caespimosa, C. canescens, C diandra, C dioica, C. disticta, C. gracilis, C. juncella, 

12

C. lasiocarpa, C. limosa, C. muricata, C. panicea, C. riparia, C. vesicaria, C. vulpina, 

Eriophorum angustifolium, Scirpus compressus, S. eupalustris, S. holoschoenus, S. ovatus, 

S. pauciforus, S. silvaticus, Juncus ambiguus, J. compressus, J. effusus, J. tenageia, Equisetum 

arvense, E. hekocharis, and E. palustre. 

(2) Flood meadows are usually inundated for at least short periods but are reasonably well 

drained, have a lower water table than wet meadows during most the year, and have moist soils for 

the rest of the growing season. Plants include Agropyron repens, Agrostis alba, A. canina, A. 

stolonifera, Alopecurus arundinaceus, A. pratensis, A. tenuis, Anthoxanthum odoratum, 

Arrhenatherum elatius, Beckmannia eruciformis, Bromus inermis, Calamagrostis epigeios, 

Deschampsia caespitosa, Festuca arundinacea, F. pratensis, Glyceria aquatica, G. fruitans, G. 

plicata, Hierochloe odorata, Koeleria delavignii, Phalaris arundinacea, Phleum pratense, Poa 

palustris, P. pratensis, P. trivialis, Trifolium hybridum, T. pratense, T. repens, Carex aquatilis, 

C. gracilis, C. riparia, C. vesicaria, Scirpus eupalustris, Juncus ambiguus, J. compressus, J. 

tenageia, Equisetum arvense, E. heleocharis, and E palustre. Species restricted to flood meadows 

with short periods of inundation include Festuca ovina, Arrhenatherum elatius, Medicagofalcata, 

Trifolium pratense, and T. repens. 

(3) Fresh ("true" or "dry") meadows are often secondary vegetation on cut or burned forest 

sites. Mowing is often used to prevent forest encroachment. They have a moderate moisture 

supply that fluctuates widely but generally does not surface. They sometimes become quite dry in 

summer. These meadows are usually used as pastures or haymeadows and some of the plants, 

especially the legumes, may have been sown. The vegetation includes Agropyron repens, 

Alopecurus pratensis, A. tenuis, Anthoxanthum odoratum, Arrhenatherum elaius, Briza media, 

Bromus inermis, Cynosurus cristatus, Dacrylis glomerata, Festuca rubra, F. pratensis, Holcus 

lanatus, H. mollis, Lolium perenne, Nardus stricta, Phleum pratense, Poa pratensis, P. trivialis, 

P. annua, Medicago falcata, M. lupulina, M. sativa, Melilomus alba, M. officinalis, Trifolium 

repens, T. hybridum, T. pratense, Carex leporina, and Equisetum arvense. Beckmannia 

erucibrmis, Phak/ris arwndinacea, and Poa palustrisare are occasionally found there. 

Herbaceous plants also commonly dominate forest and woodland margins and openings 

(glades). Many openings are man-made, the result of burning, felling, or thinning forests or 

woodlands. Herbs found there include Agropyron repens, Beckmannia pinnatum, Briza media, 

Bronu inermis, Calamagrostiu epigeios, Cynosurus criuratus, Dactylis glomerata, Melica nutans, 

Carex leporina, and Equisesum sylvaticum. Prunus spinosa, Rosa acicularis, R. mollis, 

R. tomentosa, Rubhts chamaemus, Salix caprea, S. livida, S. nigricans, S. phylicifolia, and 

Populus tremala are often found at the margin of forest openings. Mixed forest-meadow 

vegetation is divided into three types: (1) scattered, with herbaceous vegetation evenly spread 

throughout a thinned tree stand (in oak and linden groves there are about 300 to 800 trees 

13

hectare (ha) and in birch woodlands there are about 200 to 600 trees/ha); (2) flower-bed, with 

alternating openings and dense forests; and (3) windbreak, with forest strips about 25 to 30 m 

wide alternating with rectangular openings about 70 to 80 m wide. 

Relict Rhododendron luteum thickets are found at the margins of pine- and spruce-deciduous 

forests on peaty soils in the Poles'ye. Ledum palustre, Vaccinium uliginosum, and V. viuis-idaea 

are common dwarf shrubs there. 

Heaths (or heathlands) are ericoid shrub fens in poor, acid, podzolized soils, in meadow-type 

sites, often where pine forest has been cut. Calluna vulgaris is the dominant dwarf shrub. Other 

dwarf shrubs include Arctostaphylos uva-ursi, Empetrum nigrum, Ledum palustre, Vaccinium 

vids-daea, and V. uliginosur. The common herbs include Nardus stricta and Uncus squarrosus. 

Sphagnum fuscum is a common moss. 

More than 60 percent of the forests are dominated by Pinus sylvestris. The deciduous forests 

are dominated by oak, hornbeam, birch, and alder. It is unlikely that any virgin forests remain in 

eastern Europe (Walter 1979). 

There are a number of Russian terms used to describe forests and woodlands. (Woodlands 

are low density forests.) "Bor" is a pine woodland on poor, sandy soil. Bor woodlands are 

common on the poor glaciofluvial sands of the Poles'ye. Pine forests can be found in sites that 

climatically should be deciduous or coniferous-deciduous forest. "Bor oak" is occasionally used 

for oak growing in sandy soils, and "bor birch" for birch in similar soils. Pine-birch woodlands 

on poor soils are frequently also called "bor." "Subor" is a mixed deciduous-pine forest, 

dominated by pine, with a strong lower stratum of deciduous trees, usially oak but also birch or 

beech. Pine-oak forests are "subor," as are pine-birch forests on better soils. "Ramen" is a more 

closed stand, usually dominated by spruce. "Suramen" is a mixed "ramen." "Dubrava" is an oak 

forest or woodland (grove) and includes oak hornbeam forests, and "sudubrava" is a woodland 

dominated by deciduous trees usually considered subordinant in oak forests, such as birch 

woodlands and linden and aspen groves. 

Carr is an English term for a wooded wetland. There are several types of carrs in the 

Poles'ye. 

Willow (or sallow) camrs occur on either reed or sedge peat. They are usually found in the 

stream zone of floodplains. Trees may include Salix alba, S. caprea, S. rossica, Alnus glutinosa, 

A. incana, and Betula pubescens. The shrubs include Sa/ix acuufolia, S. cinerea, S. dasyclados, 

S. nigricans, S. pentandra, S. purpurea, S. riandra, S. xerophila, and Cornus a/ba. Dwarf shrubs 

include Salix aurita, S. livida, S. myrtlloides, S. rosmarinifolia, Rosa mollis, R. omenaosa, and 

Ribes nigrnm 

Osier thickets, with osier willows such as Salix purpurea and S. viminalis, are often found on 

shore dunes. 

14

Willow-poplar carts occur on recent alluvial soils in the stream and central zones of 

floodplains. Soils are s~turated during floods but waters flow through, and they can dry out 

during nonflood periods. The trees include Populus nigra, P. alba, Salix alba, S. caprea, and 

sometimes Alnus glutinosa and Fraxinus excelsior. Shrubs include Salix cinerea, S. livida, 

S. pentandra, S. purpurea, S. rosmarinifolia, S. triandra, and S. viminalis. 

Wet alderwoods occur on reed fen peats. The are usually located in the excessively wet areas 

in the stream and central zones of floodplains. Trees in alderwoods include Alnus glutinosa, 

A. incana, Fraxinus excelsior, and Betula pubescens. Shrub willows include Salix cinerea and 

dwarf shrub willows include Salix myrtilloides and S. rosmarinifolia and may also include others 

found in willow carts (above). Ribes nigrum, Betula humilus, Vaccinium myrtillus, and V. vitisidaea 

are also common dwarf shrubs. Herbs include Agropyron caninum, Deschampsia 

caespitosa, Festuca gigantea, Glyceria lithuanica, Milium effusum, Molinia coerulea, Phragmites 

communis, Carex argyroglochin, C. caespitosa, C. elongata, and C. inumbrata. 

Alder-ash and elm carts are small communities with damp to wet, weakly acid to neutral soils. 

The plants in elm cams include Ulmus iaevis, U. glabra, Quercus robur, and Acer platanoides. 

Elm carms are found on sites that are innundated only during major floods. 

The dominant trees in alder-ash carms include Alnus glutinosa, A. incana, and Fraxinus 

excelsior, mixed with a few Acer plantinoides and Carpinus betulus. Ribes nigrum is a common 

shrub. Carex remota and Equisetum silvaticum are important herbs. Alder-ash cams occur on 

humic-gley soils. 

Aspen groves (dominated by Populus tremula), linden groves (dominated by T7Iia cordata), 

and birch woodlands (dominated by Betula pendula in slightly moist soils and Betula pubescens in 

wetter soils, with a mixture in moderately moist soils) are all "sudubrava." They are all early 

successional woodlands, and they all have all have strongly herbaceous understories. Occasional 

Quercus trees may be intermixed. Alopecurus tenuis, Calamagrostis arundinacea, C. epigeios, 

Melica nwtans, and Eriophorum vaginatum are important herbs. Birch woodlands often develop in 

fire clearings in moist soils, with Arctostaphylos uva-ursi as an understory shrub. 

The Kiev Poles'ye lies on the ecotone between the mixed coniferous-deciduous forest and the 

forest steppe. There are two basic mixed coniferous-deciduous forest types: pine-oak forest and 

spruce-oak forest. The common steppe-forest of the area is oak-hornbeam. 

Quercus robur is a codominant in each of them and is the dominant tree in forest-steppe in the 

European USSR. It grows best in the southwestern part of the European USSR and in the 

Poles'ye. Quercus robur will not grow on strongly podzolic soils. It is usually found on 

floodplains. It usually grows in mixed stands with Pinbs sylvestris, Picea abies, or Carpinus 

beudus. 

15

In addition to Q. robur and C. betulus, the trees found in oak-hornbeam forests ("dubrava") 

include Acer platanoides, A. pseudoplatanus, Betula pendula, Fraxinus excelsior, Tilia cordata, 

Ulmus laevis, and U. glabra. On richer soils, Fagus sylvatica is sometimes a component on 

moderately moist soils. Cornus alba and Corylus avellana are common shrubs. Agropyron 

caninum, Koeleria delavignii, Melica nutans, and Millium effusum are important herbs. Carex 

pilosa is important on drier soils, while Festuca gigantea is important on wetter ones. Oakhornbeam 

forests are not found in stagnant water. The groundwater depth is usually greater than 

1 m. They are most common on older alluvial soils and on frontal and ground moraines. They 

are often found between wet alderwoods and poorer pine-oak oak forests. From the air, oakhornbeam 

forests appear bright light gray (in black and white photographs). The surface of the 

canopy appears uneven, and trees appear bunched in groups of three to five. 

Acidiphilous pine-oak forests C'subor") are found in poor habitats with weakly to moderately 

podzolized, sandy or sandy-loam soils. In addition to Pinus sylvestris and Q. robur, the trees 

include Betula pendula and/or Betula pubescens, Populus tremula, and Tilia cordata. Corylus 

avellana is the dominant shrub, and Vaccinium myrtillus is the most important dwarf shrub. 

Acidiphilous beechwoods (also "subor") are found on moist, moderately fertile soils. The 

dominant trees are Fagus sylvestris and Pinus sylvestris. The nondominant tree species are those 

found in acidiphilous pine-oak forests. 

Most natural pine stands ("bor") are in poor habitats. There are many "bor" woodlands in the 

Poles'ye, dominated by Pinus sylvestris. Pine-dominated stands types include (1) dry to freshon 

well-drained soils with marked podzolization, (2) damp or moist--on poorly drained soils with 

attenuated podzolization (semi-bog soils), (3) wet-on poorly drained bog soils. 

Dry "bor" woodlands are found on the old dunes and morainal hills. The parabolic, westfacing 

dunes of the Poles'ye are usually covered with Pinus sylvestris giving the region a 

"northern" appearance. These woodlands rarely have understory shrubs. The herbaceous 

understory vegetation can include Festuca ovina, F. polesica, Koeleria glauca, and Poa pratensis. 

These sites have light, dry sandy soils, and the lichens are common. Lichens include Cladonia 

alectoria, C. alpestris, C rangiferina, C. silvatica, C. uncialis, and Cetraria islandica. C alpestris 

often dominates in late succession. About 100 years is required for pine on dry soil to mature 

enough to make good timber. Trees are unequal in height, crowns are wide, and branches 

pendant Betuda is not found in dry "bot" woodlands. 

Moist "bor" forests are dominated by Pinus with either Polytrichum mosses or evergreen 

ericoid shrubs (e.g., Andromeda polifolia, Arctostaphylos uva-ursi, Chamaedaphne calyculata, 

Calluna vulgaris, Vaccinium myrtillus, and V. vitis-daea) dominating the understory. 

Brachypodium silvancum, Festuca ovina, F. rubra, Molinia coerulea, and Nardus stricta are 

16

important herbs. Trees are well proportioned, with narrow crowns. Pines in moist habitats are 

longer lived than in dry or wet habitats and are the best quality timber. 

Wet "bor" woodlands can have either flowing or stagnant water. In the former Phragmites 

communis (growing in hollows) is the dominant understory species, with Nardus stricta as an 

important herb. In the latter, Sphagnum mosses dominate the understory. Both are dominated by 

Pinus. In "subor" woodlands, Betula pubescens is strongly subdominant, but woodlands in 

which Bzga is a lesser component are often referred to as "bor." Wet pine-sphagnum woodlands 

are found in poorly drained depressions on terraces above floodplains. After 100 years, pines on 

wet soils may be only 6 m tall and reach a total heighz of only 8 m. 

There are several subtypes of pine-sphagnum "bor" woodlands. (1) Betula pubescens is an 

important tree component. Carex lasiocarpa, Equisetum heleocharis, and E. vaginatum are 

important herbaceous components. The mosses include Sph. centrale and Sph. warnstorfii, 

Sph. girgensohnii, Pleurosium schreberi, and Aulacomium palustre. This type is often found in 

depressions with gentle slopes, with peaty-gley or peaty, intensely boggy soils. (2) Salix caprea is 

an important tree, Salix pendrandra is a common shrub, and Vaccinium vitis-idaea, V. myrtillus, 

V. oxycoccus, Ledum palustre, Empetrum nigrum, Chamaedaphne calyculata are common dwarf 

shrubs. Equisetum silvaticum, Calamagrostis epigeios, Carex lasiocarpa, Eriophorum 

angustifolium, E. gracile, and E vaginatum are important herbaceous species. The mosses include 

Sph. magellanicum, Sph. compactum, Sph. fuscum, Sph. russovii. Sph. apiculatum, and 

Pleurosium schreberi. This type occurs in slight depressions with impeded runoff. (3) Betula is 

an important tree component; Pinus decreases as Betula increases in this type. Salix aurita, 

Vaccinium myrtillus, V. vitis-idaea, and Ledum palustre are the important dwarf shrubs; V. 

uliginosum is unusual. The mosses include Sph. acutifolium, Sph. warnstorfii, Polytrichum 

commune, Hylocomium proliferum, Pleurozium schreberi, and Aulacomium palustre. This type is 

transitional between slightly flowing and stagnant wet woodlands and is found in slightly hillocky 

shallow depressions. (4) Betula pubescens is an important tree component. There are also 

infrequent Alnus. Vaccinium vitis-idaea, V. myrtillus, Andromeda polifolia, Calluna vulgaris, 

Chamaedaphne calyculata, Ledum palustre, Empetrum nigrum, and Betula humilus are important 

dwarf shrubs. Carex lasiocarpa and Calamagrostis lanceolata are important herbs. The moss is 

mostly Polytrichum commune with some Sphagnum mosses, including Sph. fuscum. This type is 

found in shallow, usually closed depressions between drier pine stands. (5) Pinus is sparse. The 

common subshrubs are Rubus chamaemus, Vaccinium viis-idaea, and V. uliginosum; V. myrtillus 

and V. oxycoccus are infrequent. Equisetum variegatum is a common herb. The mosses are 

Sph. angust'folium, Sph. acutifolium, Sph. fuscum, Sph. magellanicum, and Sph. russovii, with 

Sph pulchrum and Sph. pahwstre in hollows. This type is transitional between woodland and bog. 

17

Soils are highly acidic. (6) Pinus is sparse. Shrub and moss species are the same as subtype 5. 

Eriophorwn vaginatum is the common herb. 

Picea abies endures shade well but requires humid, relatively rich soils. Its southern limit runs 

through Kiev Poles'ye. Spruce-oak forests probably occur mainly on "islands" of richer podzolic 

soils scattered in the Poles'ye. Understory trees include Acer platanoides, Betulda pendula, 

Carpinus betulus, Fraxinus excelsior, Populus tremula, Tilia cordata and Ulmus laevis. Shrubs 

include Corylus avellana and Rubus chamaemus. Herbs include Carex argyroglochin, C. loliacea, 

Scirpus sylvaticus, Agropyron caninun, Calamagrostis phragmitoides, Deschampsia caespitosa, 

Glyceria lithuanica, Melica nutans, and Millium effusum. Mosses include Sphagnum girgensohnii 

and Sph. squarrosum. 

18

SECTION 6 

SUCCESSION 

In poor sandy soils, peat bogs occupy the low areas, forests the upper ones. Succession can 

progress from forest to bog or bog to forest. 

With increased moisture, succession progresses from Pinus-Polytrichum moist forests to 

Pinus-Sphagnum wet forests, then to sphagnum-pine bog. 

The first sere in secondary succession after an moderately dry oak-hornbeam "dubrava" has 

been felled is a pine-birch "bor," followed by a pine-oak-birch "subor," and then finally a return to 

"dubrava." 

When moist pine forests are overly thinned or cleared for meadows, without adequate drainage 

and with grazing, boggy conditions develop and podzols deteriorate. With time these meadows 

become sphagnum bogs. Without grazing, meadows that do not become bogged are colonized by 

birch, aspen, alder. Pine eventually returns under these, and, with time, they return to pine 

forests. 

If birches are left when pines are felled in pine-birch woodlands, birch woodlands with grassy 

understories often develop within 3 to 5 years. Because Calamagrostis epigeios inhibits pine 

regeneration, it can retard secondary succession. 

In deforested dry pine woodlands on sandy hills or dunes, trees and lichens are often replaced 

by herbaceous dune vegetation, much of which is already found in the understory, including 

Festuca ovina, F. polesica, Koekria glauca, K. gracilis, and Poa pratensis. 

Aspen groves, linden groves, and birch woodlands are all early successional woodlands on 

moderately moist soils. These can become oak-hornbeam forest. 

Willow-poplar carrs usually begin as osier thickets. These are replaced with alder-ash carts, 

which can become oak-hornbeam forest. 

As succession proceeds in a fen, it first becomes a willow carr or wet alderwood, which is 

replaced by alder-ash can'. If the site dries, this is replaced by an oak-hornbeam forest. Further 

drying leads to a pine-oak woodland. 

If a fen is grazed, the first step is a wet meadow. If the groundwater table drops, it will 

become a fresh meadow. If grazing is abandoned, the site will become an oak-hornbeam forest. 

Further drying again results in a pine-oak woodland. 

Flood meadows are created when fens are drained. In time these will become forest. 

If flowing water decreases in a fen but the site remains wet, peat decomposition will decrease, 

and pines and sphagnum will increase. 

19

With time, a wet pine-sphagnum woodland will become increasingly bog and eventually will 

become a sphagnum-pine bog. As the pines become less and less robust, the tree component can 

be lost, or nearly so. 

20

SECTION 7 

BOG DRAINAGE AND RECLAMATION 

Most bogs in central Europe are being drained. The replacement vegetation is usually meadow 

grasses, birch, pine, or spruce. 

Many of the wetlands in the Poles'ye have been drained. Between 1872 to 1898, 400,000 ha 

of wetlands were converted to meadows, 130,000 ha to croplands, and 1,300,000 ha to accessible 

forests. In 1950, it was estimated that 1.3 million ha in Belorussia and 1 million ha in the Ukraine 

"needed" reclamation. Between 1950 to 1970, an extensive drainage program was caried out in 

the Poles'ye in both republics. By 1972, the Poles'ye had ceased to be a primarily waterlogged. 

Continuous extensive drainage of the Pripyat wetlands may disturb the groundwater balance 

both in the Poles'ye and in the drier regions to the south. Drainage can lead to excessive drying of 

rivers and to destruction of hydrophylic shrubs on berry grounds. 

Because the sandy soils of this region are particularly vulnerable to wind erosion when 

plowed, conversion to croplands has not always been successful. Drainage has made some old 

farmlands worthless. In some areas, the sandy soils have dried out and become blown sand; some 

peat soils need irrigation to grow wheat or beets. Since 1968, the formerly excessively wet 

Poles'ye has experienced dust storms. There is a joke that, in the Poles'ye, those attempting to 

"reclaim" wetlands "turn all swamps into deserts" (Komarov 1980). Because of the soils beneath 

the bogs, it might have been better if they had been converted to meadows instead of croplands; 

however the plans of the agricultural bureaucrats specified plowed croplands. 

Pines in some tree farms have died from viral diseases; they needed the acid soils formed from 

stagnant water to do well. These diseased nursery trees also infected nearby stressed older trees. 

21

SECTION 8 

CROPS 1 s 

Only about half of the area around Chernobyl is suitable for agriculture. Much of it is in hay 

meadows and grazing land (Trifolium spp. and grasses), 19 and only about 25 percent of the area is 

actually in croplands. 

Triticum aestivum, Secale cereale, Panicum miliaceum, Hordeum vulgare, Avena sativa, 

Fagopyrum esculentum, Beta vulgaris, Solanum tuberosum, Linum usitatissimum, Cannabis 

satii, Humulus lupulus, Daucus carota, Brassica oleracea var. capitata, B. oleracea var. acephala, 

B. rapa, and Cucumis sativa are all grown in the southern part of the Belornssian SSR and 

northern part of the Ukrainian SSR. However, most can not be grown on poorly drained, boggy 

sites of the Poles'ye. As a result, wheat and other mesic soil crops are less important here than in 

other parts of Belorussia and the Ukraine. Cereal grains are grown on about half the crop lands, 

fodder crops on 35 to 40 percent, potatoes on about 8 percent, and flax on up to 5 percent. At any 

one time, 12 to 17 percent of the land is being fallowed. 

On drained bogs with moderate peat beds, crops are often replaced, after a few years, with 

perennial grasses and legumes, creating an artificial meadow for hay and grazing. Many of the 

dwarf shrubs growing in bogs produce edible fruits and these are harvested as "crops." In some 

places, they are cultivated in the bogs. 

18 Information regvding cops has been synmeuized from Lydolph 1964, Skoropanov 1968, Sambur and Kovalenko 

1969, Fuflard 1972, Synau 1972, Komamov 1980, Dewdocy 1982, USSR 1987. 

19 See meadows, under Sectin 5, Vegeaio. 

22

SECTION 9 

REFERENCES 

Akademiya Nauk SSSR. 1963. The Increase of Productivity of Swamped Forests. Israel 

Program for Scientific Translation, Jerusalem. 291 pp. 

Anonymous. 1962. Atlas of the Ukrainian SSR and Moldavian SSR. Central Administration of 

Geodesy and Cartography, Moscow. (In Russian) 

Berg, L. S. 1950. Natural Regions of the U.S.S.R. Translated by 0. A. Titelbaum. MacMillan 

Co., New York. 

Bialobok, S., and W. Zelawski, eds. 1976. Outline of Physiology of Scots Pine. Foreign 

Scientific Publications, Department of the National Center for Scientific, Technical and 

Economic Information, Warsaw. 

Borisov, A. A. 1965. Climates of the U.S.S.R. 

Publishing Co., Chicago. 

Translated by R. A. Ledward. Aldine 

Brady, N. C. 1974. The Nature and Properties of Soils. MacMillan Publishing Co., Inc., New 

Yor0. 

Dewdney, J. C. 1982. U.S.S.R. in Maps. Holmes & Meier Publishers, Inc., New York. 

Fridland, V. M. 1976. Patterns of Soil Cover. Israel Program for Scientific Translation, 

Jerusalem. 

Fullard, H. 1972. Soviet Union in Maps. George Philip & Son Ltd., London. 

Golovina, L. P., M. N. Lysenko, and T. I. Kisel. 1980. "Content and Distribution of Zinc in the 

Soil of the Ukrainian Poles'ye." Soviet Soil Science 12(l):73-80. 

Hora, B., ed. 1981. Oxford Encyclopedia of Trees of the World. Oxford University Press, 

Oxford, England. 

Ivanov, K. E., ed. 1972. The Hydrology of Marshlands. Israel Program for Scientific 

Translations, Jerusalem. 

Keller, B. A. 1927. Distribution of Vegetation on the Plains of European Russia. Journal of 

Ecology 15:189-233. 

Kendrew, M. A. 1942. The Climates of the Continents. Oxford University Press, New York. 

Komarov, B. 1980. The Destruction of Nature in the Soviet Union.. Translated by M. Vale & 

J. Hollander. M. E. Sharpe, Inc., New York. 150 pp. 

Komarov, V. L. 1963. Flora of the USSR. Volume 2. Israel Program for Scientific Translations, 

Jerusalem. 

23

-. 1964. Flora of the USSR.. Volume 3. Israel Program for Scientific Translations, 

Jerusalem. 

• 1967. Flora of the USSR. Volume 18. Israel Program for Scientific Translations, 

Jerusalem. 

- 1968. Flora of the USSR. Volume 1. Israel Program for Scientific Translations, 

Jerusalem. 

- 1970. Flora of the USSR. Volume 5. Israel Program for Scientific Translations, 

Jerusalem. 

. 1971La. Flora of the USSR. Volume 10. Israel Program for Scientific Translations, 

Jerusalem. 

1971b. Flora of the USSR. Volume 11. Israel Program for Scientific Translations, 

Jerusalem. 

Krupskiy, N. K., V. P. Kuz'michev, and R. G. Derevyanko. 1070. "Humus Content in 

Ukrainian Soils." Soviet Soil Science 2:278-288. 

Larin, I. V. 1962. Pasture Economy and Meadow Cultivation. Israel Program for Scientific 

Translations, Jerusalem. 

1962. Pasture Rotation. Israel Program for Scientific Translations, Jerusalem. 

1., ed. 1965. Advances in Pasture and Hay-meadow Management. Israel Program for 

Scientific Translations, Jerusalem. 

Leathart, S. 1977. Trees of the world. Hamilyn, London. 

Lydolph, P. E. 1964. Geography of the USSR. John Wiley & Sons, Inc., New York. 

Lysenko, M. N., and L. P. Golovina. 1982. "Boron Content and Distribution in the Soils of the 

Ukrainian Poles'ye." Soviet Soil Science 14(1):89-97. 

Moore, P. D., ed. 1984. European Mires. Academic Press, London. 

Oleynik, V. S. 1981. "Genetic characteristics of peat soils in the Ukrainian western Poles'ye." 

Soviet Soil Science 13(4):32-37. 

Phillips, R. 1978. Trees of North American and Europe. Random House, New York. 

P'yavchenko, N.I., ed. 1965. Improvement in Forest Growth on Peat-Bog Soils of the USSR 

Forest Zone and Tundra. Israel Program for Scientific Translations, Jerusalem. 

Remezov, N. P., and P. S. Pogrebnyak. 


1969. Forest Soil Science. Israel Program for 

24

Sambur, G. N., and I. L Kovalenko. .1969. "Improvement and Rational Utilization of Lowland 

Saline Soils of the Southern Poles'ye and Northern Forest-Steppe of the Ukrainian SSR." 

Soviet Soil Science 1:1401-1408. 

Skoropanov, S. G. 1968. Reclamation and Cultivation of Peat-bog Soils. Israel Program for 


Smith, R. L 1980. Ecology and Field Biology. Harper & Row, Publishers, New York. 

Stankevich, V. S., and P. R. Rubin. 1968. Drainage and Reclamation of Bogs and Bogged 

Areas. Israel Program for Scientific Translations, Jerusalem. 

Sukachev, V. N. 1928. "Principles of Classification of the Spruce Communities of European 

Russia." Journal of Ecology 16:1-18. 

Symons, L. 1972. Russian Agriculture: A Geographic Survey. John Wiley and Sons, New 

York. 

Szafer, W. 1966. The Vegetation of Poland. Pergamon Press, Oxford. 

Takhtajan, A. 1986. Floristic Regions of the World. Translated by T. J. Crovello. University of 

California Press, Berkeley. 

Tutin, T. G., V. KL Heywood, N. A. Burges, D. H. Valentine, S. M. Walters, and D. Webb. 

1964. Flora Europaea. Volume 1. Cambridge University Press, London. 

Tutin, T. G., V. H. Heywood, N. A. Burges, D. M. Moore, D. H. Valentine, S. M. Walters, 

and D. A. Webb. 1968. Flora Europaea. Volume 2. Cambridge University Press, London. 

1972. Flora Europaea. Volume 3. Cambridge University Press, London. 

USSR. 1986. The Accident at the Chernobyl Nuclear Power Pant and its Consequences. 

Information compiled by the USSR State Committee on the Utilization of Atomic Energy for 

the IAEA Experts' Meeting, 25 to 29 August 1986, Vienna (IAEA Translation). 

USSR. 1987. The Accident at the Chernobyl Nuclear Power Plant: One Year After. International 

Conference on Nuclear Power Performance and Safety, 28 September to 2 October 1987, 

Vienna (IAEA wanslation). 

Viktorov, S. V. Ye. A. Vostokova, and D. D. Vyshivkin. 1964. Short Guide to Geobotanical 

Surveying. MacMillan Co., New York. 

Vilenskii, D. C. 1960. Soil Science. Israel Program for Scientific Translations, Jerusalem. 

Walter, H. 1978. Vegetation of the Earth and Ecological Systems of the Geobiosphere. 

Translated by J. Wieser translator. Springer-Verlag, New York. 

Wulff, E. V. 1943. An Introduction to Historical Plant Geography. Translated by E. Brissenden. 

Chronia Botanica Co., Waltham, Massachusetts. 

25

APPENDIX A 

TREES AND SHRUBS 

List of Species 2 0 ' 2 1 

Noncultivated 2 2 

"Mmeciuse more than one valid Latin nam was sometimes found for species, I chose the conservatie name 

(choosig thernam which stmasse simibiaites or jlumper's name). Lotn names of dicoyledons follow the Flora 

Ewvaerop those of mnuo olqledoms are those used in the curren liateature. The authorof Flora of she USSR tended 

to emphasize differences (i~e., was what is called a splitter elevating subgeneric names to genera, subspecific 

qilthas to specific), his names often do nom reflect those most frequently encountered tn the literature. 

2 1Reggreaces nchide those used for specie and habitats in Section 3. Vegetation 

Z 2 lchuifg species seeded int meadows, exclding species growing sa crop in plowed fields. 

A-1

GYMNOSPERMS 

PINUS SYLVESTRIS 23 

Scots or Scotch pine24 

common pine, wild pine 

forest pine 

P. sylvestrs is the dominant tree in the Poles'ye and is the most abundant in much of Europe, 

where it is often called simply "common pine" (also its common name in the USSR). It can live 

for up to 400 years. It is an erect evergreen tree and can be 20 to 30 m tall (occasionally up to 

40 m). In bogs, trees are quite stunted; after 100 years trees often are only 6 m tail, and they 

rarely get to be more than 8 m tall. Its crown is round, high, broad, and (in aerial views) appears 

smooth and dark (although lighter than spruce canopies). Intervals between crowns appear 

uniform and are not deeply shaded. The trunk is often crooked, with regularly whorled branches, 

diverging at acute angles. The bark is deeply fissured, thick, and dark brown on the lower trunk 

but smooth, thin, and red or red-brown on the upper trunk. 

There are two types of vegetative apices. Long shoot are elongated stems with scaled leaves in 

the axils. These produce dwarf shoots, lateral long shoots, and strobili (cones). Short shoot are 

deciduous shoots, which do not elongate and which produce the leaves. Buds are acute and more 

or less resinous. Twigs are glabrous and are initially yellow green, becoming gray-brown with 

age. 

Primary leaves are always single and are found only on seedlings. Within 2 to 3 years, they 

are completely replaced by secondary (permanent) leaves. Secondary leaves (needles) are paired, 

glaucous, and blue-green in color. Needles are elongate, semicylindrical, tapering toward the 

apex. The internal surface is flat and either slightly concave or slightly convex. Stomata are found 

on both sides of the needle distributed in rows along its entire length. The external surface is 

semicircular. They are twisted, 2 to 7 cm long, and about 2 mm in wide. Needle length varies 

with year of growth; they may continue to grow for more than one season. Needles on trees from 

eastern Europe are 5 to 7 cm long; the longest needles found have been on pines from the Poles'ye 

and Volhynia. Shoots with female cones usually have longer needles than shoots with male cones. 

Individual leaves can survive 5 or more years. 

Degree of stomatal opening decreases from a morning maximum to almost complete closure by 

evening. Regardless of soil moisture, there are considerably fewer stomata open in the afternoon. 

23 Someimes spelled sve . 

24Most raulasmrs give the English common me far a species. However, Ameaic and other European as weln 

as lcal, common ames wer sometimes encountezed. I have listed all common ames I found for each species. 

A-2

Reproduction from pollination to mature seed takes 3 years. Seeds form 1 year after 

pollination. Cones remain small during the first year after pollination. There is rapid growth and a 

change in color (from gray-brown to rich green) in the second summer. Mature seed cones are 3 to 

7 cm long and 2 to 3.5 cm wide. They are yellow-brown when they are mature. Free-standing 

trees can bear seeds when they are about 15 years old; trees in a closed canopy do not begin to bear 

seeds until they are about 20 years old. Abundant seed crops are produced only every 3 to 4 years, 

but at least a few trees yield seeds every year. Self-pollination produces poorer quality seed than 

does cross pollination. 

P. sylestris usually forms mycorrhizal associations in all soil types, although there are 

qualitative and quantitative differences among soils. Pines in bog soils are very mycorrhizal, 

probably because of the low nutrient status of the soil. More than 40 species of fungi have been 

found associated with P. sylvestris in Europe. Annual radial growth in P. sylvestris is dependent 

on temperature and nutrient depletion. There can be as many as seven flushes, with more than one 

growth ring per year. Annual apical growth can be completed as early as mid-June. Maximum 

needle elongation may occur as -arly as early May. 

There is quite a bit of difference among trees in different habitats. In dry sites, trees are 

unequal in height, crowns are wide, branches pendant, and needles are quite blue-gray. Growth 

begins earlier, and the growth increment is more clearly marked than in moister habitats. The size 

of the increment depends on the amount of summer rainfall. In moist habitats, trees are well 

proportioned, crowns are narrower, and leaves are greener. Trees live longer (usually more than 

200 years). The size of the growth increment depends on the amount of summer rainfall. The best 

quality timber comes from moist forests. In wet habitats, trees have shorter lives, usually 100 to 

200 years. Growth begins late, after the groundwater table subsides. Sphagnum mosses inhibit 

regeneration from seed. Trees growing in very wet habitats are susceptible to windthrow (or 

blowdown). 

It is codominant with Quercus robur in pine-oak mixed forests and with Fagus silvatica in 

acidiphilous beechwoods. It is dominant in "bor" forests. It is found in sphagnum-pine bogs and 

wooded transitional bogs. 

P. sylvestris is a light-loving species, not tolerant of shade. It is xerophytic, 25 not very 

exacting in its moisture requirements and can be found growing under almost any moisture regime 

from dry to swampy. On wet soils, it is susceptible to windthrow. It is an oligotroihic 26 

25C=win 

dry sad. 

A-3

acidophyte. 27 It can tolerate relatively poor soils and is often associated with bogs and sandy 

soils. It has moderate transpiration rates. 

PICFA ABIES2 

Norway spruce 

European spruce 

Common spruce 

P. abies is an important evergreen tree in much of eastern Europe. Its southern limit runs 

through Kiev Poles'ye, but it apparently is not common there. It is usually 30 to 50 m tall but 

occasionally reaches 60 m. It is full grown in 30 to 50 years, but it can live for more than 300 

years. 

Its crown is pointed, pyramidal, and nearly rests on the ground. In aerial views it appears very 

uneven and darker than that of P. sylvestris. 

Its branches are whorled, and it lacks short shoots. Needles are spirally arranged, 1 to 2 cm 

long, and green. Mature cones are 10 to 14 cm long. 

P. abies endures shade well but requires humid, relatively rich, loamy soils. Because poor 

soils predominate in the Poles'ye, P. abies is probably limited to widely scattered "islands" of 

richer podzolic soils. 

It is mesotrophic. 29 It grows best on moderately moist soils. On wet soils, it is susceptible to 

windthrow. 

27 Toume acidic wiLt 

2 9P. mete (fiqpemm, foud in ok" fkeme)- P. abies. 

29 kRqoiu at lo• uodme W&,l of mnuaims. 

A-4

DICOTYLEDONS 

QUERCUS ROBUR 

English oak 

British oak 

European oak 

pedunculate oak 

Q. robur is the most important deciduous tree in the mixed forest of the region. It is found 

with Pinus sylvestris or Picea abies in mixed forests and with Carpinus betulus in oak-hornbeam 

forests. It is found in elm cam and is occasional in linden and aspen groves and birch woodlands. 

Its crown appears blunt and is very large. Oak-hornbeam forests appear bright light-gray in 

aerial photographs (especially in summer). Because of the mixture of other species, the surface of 

the canopy appears somewhat uneven and trees may appear bunched. 

It is mesotrophic, has high transpiration rates, and is a mesoxerophyte. 30 It grows best on 

moderately moist soils. On wet soils, it is susceptible to windthrow. 

It grows to be 30 to 50 m tall and lives up to 800 years. 

ACER PLATANOIDES 

Norway maple 

sharp-leafed maple 

It is a deciduous tree, usually 18 to 20 m (occasionally 30 m) tall. The crown is densely 

leafed, and it has a spreading canopy. It is eutrophic, 3 1 has high transpiration rates, and is a 

xeromesophyte. It grows best in slightly moist soils. It is found in alder-ash carrs, oak-hornbeam 

forests, and spruce-oak forests. 

ACER PSEUDOPLA7TNUS 

sycamore 

sycamore maple 

great maple 

It is a deciduous tree, up to 30 m tall. It is eutrophic. It is found in oak-hornbeam forests. 

Dense, domed crown; spreading canopy. 

30 rowfum moduuely dry d~L 

3 IRelhu~ vrbvely high nutrent levels. 

A-5

ALNUS GLUTINOSA 

European alder, black alder 

common alder 

It is a deciduous tree, usually up to 20 m tall (rarely up to 35 m). The crown is pyramidal. Its 

bark is dark brown and fissured. It is mesotrophic, has moderate transpiration rates, and is a 

hydrophyte. 32 It always grows in moist soils, especially cleared forest sites and drained bogs. It 

is unable to grow in dry ones. In boggy soils, grows well only on the peaty layer or on peat, 

which is at times dry and whose upper layers are strongly decomposed; it is usually found on 

hummocks. It is codominant in alder-ash carts and wet alderwoods. It is found in sallow 

(willow) cans. It is sometimes found in willow-poplar carrs. 

ALNUS INCANA 

gray alder, white alder 

speckled alder 

European alder 

It is a small deciduous tree (sometimes a shrub), 5 to 20 m tall. It grows on moderately welldrained 

soils. Its bark is smooth. It is mesotrophic and is a mesohydrophyte. 33 In boggy soils, it 

usually grows on hummocks. It is found in alder-ash camrs, sallow (willow) carts, wet 

alderwoods, wooded transitional bogs, and at forest margins. It is an early successional species in 

cleared moist and wet forest sites. ALNUS GLUTINOSA x A. INCANA hybrids are not 

uncommon where parents grow together. 

BETULA PENIDULA3 

European white birch 

silver birch, common birch 

It is a deciduous tree, 10 to 30 m tall. The bark is smooth and white. Its crown is lower, more 

elongate, lighter colored, and grows closer to the ground than pine. Spaces between crowns are 

small and uniform. (In a pine-birch mixture, it is difficult to distinguish the two, and the mixture 

can be difficult to distinguish from pure stands of either.) 

32 Toklrm exaemive flooding (more than 40 days), grows only in damp arean, often in standing water. 

33 .rokrms mod flooding (20.40 days). 

34Bau pendal . ILwrucosa of ealy iteram. 

A-6

It is oligotrophic, an acidophyte, and a mesophyte. 35 It grows best in slightly moist soils, 

especially on burns and other clearings. In wet soils, it is susceptible to windthrow. Sphagnum 

mosses inhibit regeneration from seed. It does not grow well on dry soils. In bog soils, it grows 

well only on the peaty layer or on peat, which is at times dry and whose upper layers are strongly 

decomposed; it is usually found on hummocks. 

It can be dominant in early successional birch woodlands on moist grounds. It is found in 

acidiphilous beechwoods, pine-oak woodlands, and spruce-oak forests. It often grows with 

B. pubescens. 

BETULA PUBESCENS 

downy birch, red birch 

pubescent birch 

It is a low to medium-height deciduous tree, 10 to 25 m tall. Its crown is lower, more 

elongate, lighter colored, and grows closer to the ground than pine. Spaces between crowns are 

small and uniform. (In a pine-birch mixture, it is difficult to distinguish the two, and the mixture 

can be difficult to distinguish from pure stands of either.) 

It is oligotrophic, an acidophyte, and a mesohydrophyte. It grows best on moderately moist 

soils, especially on burns and other clearings. On wet soils, it is susceptible to windthrow. 

Sphagnum mosses inhibit regeneration from seed. It does not grow well on dry soils. In bog 

soils, it grows well only on the peaty layer or on peat, which is at times dry and whose upper 

layers are strongly decomposed; it is usually found on hummocks. 

It is often locally dominant in poor, acid (peaty) woodlands. It can be dominant in early 

successional birch woodlands on moist grounds. It is found in wet pine-sphagnum forests 

(subtype 4), sallow (willow) carrs, wet alderwoods, pine-oak woodlands, and wooded transitional 

bogs. It is often found with B. pendula. 

CARPINUS BETULUS 

common hornbeam 

European hornbeam 

It is a deciduous tree that is usually 15 to 25 m tall (occasionally up to 30 m). The trunk is 

fluted. Its bark is smooth and gray. The crown is rounded. It is mesotrophic, an acidophyte, and 

a mesophyte. It grows best in slightly moist soils and cannot tolerate flooding. It is codominant in 

oak-hornbeam forests. It is found in alder-ash carts. 

35 

equim mods aly mot sl. 

A-7

FAGUS SIL VA TICA 

European beech 

common beech 

It is a deciduous tree that is usually 30 m tall (occasionally up to 50 m in dense beechwoods). 

The bark is smooth and gray. It is frequently cultivated on well-drained soils. It grows best on 

slightly moist soils and cannot tolerate flooding. It is codominant in acidiphilous beechwoods 

(with Pinus sylvestris). It is sometimes found in oak-hornbeam forests. 

FRAXINUS EXCELSIOR 

European ash 

common ash 

It is a deciduous tree, usually up to 30 m tall (occasionally up to 45 m). The bark is gray. It is 

eutrophic, has high transpiration rates, and is a mesophyte. It grows best in slightly moist soils. 

In boggy soils, it usually grows on hummocks or intermediate sites. 3 6 It is found in wet 

alderwoods, alder-ash carrs, oak-hornbeam, and spruce forests. It is sometimes found in willowpoplar 

carts. 

POPULUS ALBA 

white poplar 

It is a deciduous tree. It grows in spring-inundated soils. It is found in willow-poplar carts. 

POPULUS NIGRA 

black-barked poplar 

common black poplar 

water poplar 

It a deciduous tree, up to 30 m tall. The bark is dark gray. Leaves are deltoid. It grows in 

spring-inundated soils. It is found in willow-poplar cams. It usually grows mixed with Salix and 

Alnus. 

36 &Mae Wpm between hummoc md Wolows in bop. 

A-8

POPULUS TREMULA 

aspen, asp 

European aspen 

trembling aspen 

trembling poplar 

It is a deciduous tree that is usually 15 to 20 m tall (occasionally up to 50 m). It is a clonal tree. 

Individual shoots are shortlived. Bark of young shoots is smooth, greenish-white but is brown 

and ridged at base of older shoots. 

It is mesotrophic. It grows in spring-inundated soils and is a mesohydrophyte. In very wet 

habitats, it is susceptible to windthrow. In boggy soils, it is usually found on hummocks. Aspen 

canopies are much lighter than pine or birch and appear flat. When aspen grows mixed with pine 

or birch, the woodland may have a "flower-bed" appearance because of the clonal growth form of 

aspen. 

It is dominant in early successional aspen groves on moist soils. It is found in acidiphilous 

beechwoods, pine-oak woodlands, spruce-oak forests, and at the margins of transitional bogs and 

forest openings. 

SALIX ALBA 

white willow 

European white willow 

It is a tree willow, 10 to 25 m tall. The bark is gray. It is a mesohydrophyte and grows best 

on moderately moist soils. It is found in willow-poplar carrs. 

SALIX CAPREA 

goat willow, pussy willow 

goat sallow, great sallow 

It is a tree willow, 6 to 10 m tall. It is mesotrophic and a mesohydrophyte. It grows best on 

moderately moist soils. In boggy soils, it is usually found on hummocks. It is found in willowpoplar 

and sallow (willow) carrs, at forest margins, and in wet pine-sphagnum forests (subtype 2). 

SALIX PENTANDRA 

bay willow 

bay-leaf willow 

laurel willow 

laurel-leaf willow 

It is a deciduous tree in damp woods, 7 to 10 m tall. Twigs are glabrous and shiny. It is 

mesotrophic. It is found in sallow (willow) and willow-poplar carms. 

A-9

SALIX ROSSICA 

It is a deciduous tree, 8 to 20 m tall. It is found in sallow (willow) carrs. 

T7LJA CORDATA 

small-leafed lime 

small-leafed linden 

It is a deciduous tree, up to 35 m tall. The crown appears pale, large, and spreading. It 

appears almost white in late-spring (flowering) photos. It is mesotrophic and a mesophyte. It is 

dominant in early successional linden woodlands on moist soils. It is found in oak-hornbeam 

forests, acidiphilous beechwoods, pine-oak woodlands, and spruce-oak forests. 

ULMUS GLABRA 

Scotch elm, Wych elm 

scabrus elm, mountain elm 

It is a deciduous tree, up to 40 m tall. It is eutrophic, has high transpiration rates, and is a 

mesophyte. It grows best in moderately moist soils. It is codominant in elm carrs. It is found in 

oak-hornbeam forests on flood-plains and flat interfluvial plains. 

ULMU$ LAEVIS37 

European white elm 

European elm, Russian elm 

fluttering elm 

It is a deciduous tree, 20 to 35 m tall. It is eutrophic, has high transpiration rates, and is a 

mesohydrophyte. It grows best in moderately moist soils. It is codominant in elm camf. It is 

found in oak-hornbcam forests and spruce-oak forests. 

37 Uw L/aevis - U. pedwucwata of early iitumwe. 

A-1O

SHRUBS (OVER 2 M) 

CORNUS ALBA 

Siberian dogwood 

It is a deciduous shrub, up to 3 m tall. Twigs are dark red. It is found in sallow (willow) carrs 

and oak-hornbeam forests. 

CORYLUS AVELLANA 

European filbert, cobnut 

European hazelnut 

common hazelnut 

wild hazelnut 

It is 4 to 6 m tall. The bark is smooth and brown. It is eutrophic and a mesophyte. It is found 

in oak-hornbeam forests and acidiphilous beechwoods, pine-oak woodlands, and spruce-oak 

forests. 

PRUNUS SPINOSA 

blackthorn, sloe berry 

It is a deciduous shrub, 4 to 8 m tall. It is found in at the margins of forests. 

RHODODENDRON LUTEUM 

pontic azalea 

It is an evergreen ericoid shrub, 1 to 6 m tall. It is a mesotrophic acidophyte. In the USSR, 

this species is principally found in western TransCaucasus. In the Poles'ye, it is assumed to be a 

Tertiary relict. It is found at the margins of pine- and spruce-deciduous forests on peaty soils. 

SAM1X ACUTIFOLJA 

sharp-leafed willow 

It is a tall deciduous shrub (rarely a tree), 10 to 12 m tall. It is a xerophyte. It is found in 

sallow (willow) cars. 

A-i1

SALIX CINEREA 

gray willow 

gray sallow 

It is a deciduous shrub, 5 to 10 m tall. It is a hydrophyte. It is found in sallow (willow) and 

willow-poplar carts, wet alderwoods, and shrubby transitional bogs. 

SALIX DASYCLADOS 

It is a deciduous shrub, 5 to 8 m tall. It is found in sallow (willow) carts. 

SALIX NIGRICANS 

dark-leafed willow 

black-leafed willow 

It is a deciduous shrub, 1 to 8 m tall. It is found in sallow (willow) and willow-poplar camrs 

and at forest margins. 

SALIX PENTANDRA 

bay willow 

bay-leaf willow 

laurel willow 

laurel-leaf willow 

It is deciduous and is a shrub (3 to 5 m tall) in peat soils. In boggy soils, it is usually found on 

intermediate sites. Twigs are glabrous and shiny. It is found in wet pine-sphagnum forests 

(subtype 2), transitional bogs, and in wet alderwoods. 

SALIX PHYLUCIFOLIA 

tea-leafed willow 

It is a deciduous shrub, up to 3.5 cm tall. It is found in sallow (willow) camts, transitional 

bogs, forest openings and at margins. 

SAMIX PURPUREA 

purple osier 

purple willow 

It is a deciduous shrub, usually 2 to 5 m tall (rarely up to 10 m). It is found in sallow (willow) 

and winow-poplar cans. 

A-12

SALIX T7JANDRA 

almond-leafed willow 

French willow 

It is deciduous and is usually a shrub (sometimes small tree), 4 to 10 m tall. It is found in 

sallow (willow) and willow-poplar carts. 

SALIX VIMINAUS 

common osier 

osier willow 

It is a deciduous shrub, 5 to 10 m tall. It is found in willow-poplar carts (earlier succession). 

SALMX XEROPHILA 

It is a deciduous shrub, up to 6 m tall. It grows in sallow (willow) carrs. 

A-13

DWARF SHRUBS (UNDER 2 METERS) 

ANDROMEDA POLUFOLUA 

andromeda, bog rosemary 

bog-rosemary andromeda 

It is an evergreen ericoid dwarf shrub, 15 to 40 cm tall. It is an oligotrophic acidophyte. In 

boggy soils, it is usually found on hummocks or tree boles. It is found in wet pine-sphagnum 

forests (subtype 4), sphagnum bogs, and moist pine woodlands. 

ARCTOSTAPHYLOS UVA-URSI 

bearberry 

It is an evergreen, mat-forming, ericoid dwarf shrub, 25 to 130 cm tall, with long, prostrate 

branches. It is an oligotrophic acidophyte. In boggy soils, it is usually found on hummocks. It is 

found in heaths, thin birch woodlands, and moist pine woodlands. 

BETULA NANA 

dwarf birch 

It is a deciduous dwarf shrub, 20 to 70 cm tall (rarely more than 2 m), with spreading or 

procumbant branches. It is oligotrophic. In boggy soils, it is commonly found on hummocks. 

While this is primarily a tundra-taiga species, it is sometimes found in the forest zone. Its habitats 

are the same as those of B. humilus. 

BETULA HUMILUS 

low birch 

dwarf birch 

It is a small, much-branched deciduous shrub, 1 to 2 m tall. It is oligotrophic. 

characteristic of continuously wet peat. In boggy soils, it is commonly found on hummocks or 

tree boles. It is found in wet pine-sphagnum forests (subtype 4), shrubby transitional bogs, 

alderwoods, and sphagnum bogs. 

It is 

A-14

CHAMAEDAPHNE CALYCULATA 

leatherleaf 

ground laurel 


boggy soils, it is usually found on hummocks or tree boles. It is found in moist pine forests, in 

wet pine-sphagnum forests (subtype 1, 2,4), in shrubby transitional bogs, and in sphagnum bogs. 

CALLUNA VULGARIS heather 


boggy soils, it is usually found on hummocks or tree boles. It is the dominant shrub in heaths. It 

is found in moist pine-sphagnum forests, in wet pine-sphagnum forests (subtype 4), and (in late 

successional stages) in sphagnum bogs. 

EMPETRUM NIGRUM 

black crowberry 

It is an evergreen ericoid dwarf shrub, up to 120 cm tall. It is an oligotrophic acidophyte. In 

boggy soils, it is usually found on hummocks or tree boles. It is found in pine-sphagnum 

woodlands (subtypes 2, 4, 5, 6), heaths, and sphagnum bogs. 

LF_•UM PALUSTRE 

ledum, marshtea 

crystaltea ledum 

wild rosemary 


boggy soils, it is usually found on hummocks or tree boles. It is found in moist pine forests, wet 

pine-sphagnum forests (subtypes 1, 2, 3, 4), heaths, rhododendron thickets, sphagnum bogs, and 

shrubby transitional bogs. 

RIDES NIGRUM 

black current 

It is a deciduous dwarf shrub, 1 to 2 m tall. In boggy soils, it is usually found on hummocks. 

It is found in wet alderwoods. 

A-15

ROSA ACICULARIS 

dog rose 

It is a deciduous dwarf shrub, up to 1 m tall. It is found at forest margins. 

ROSA MOLWS 

hair rose 

It is a deciduous dwarf shrub, 50 to 200 cm tall. It is found in sallow (willow) carts. 

ROSA TOMENTOSA 

downy rose 

It is a deciduous dwarf shrub, 50 to 150 cm tall. It is found in sallow (willow) carts and at 

forest margins. 

RUBUS CHAMAEMUS 

raspberry, cloudberry 

It is a deciduous dwarf shrub. In boggy soils, it is usually found on hummocks. It is found in 

wet pine-sphagnum forests (subtypes 5, 6), shrubby sphagnum bogs, and spruce-oak forests. 

SALIX A URITA 

roundear willow 

roundear sallow 

eared sallow 

It is a deciduous dwarf shrub, 1 to 2 m (rarely 3 m) tall, with creeping stems. It is oligotrophic 

and is a hydrophyte. In boggy soils, it is usually found on hummocks or intermediate sites. It is 

found in wet pine-sphagnum forests (subtype 3), sallow (willow) carrs, wet alderwoods, and 

shrubby transitional bogs. 

SAMX CINEREA 

gray willow, gray sallow 

It is a deciduous shrub, dwarfed in peat-bogs, with creeping stem. It is found in shrubby 

sphagnum and transitional bogs and in sallow (willow) carrs. 

A-16

SALIX LAPPONUM 

burr willow 

It is a deciduous dwarf shrub, 1 to 1.5 m tall. It is found in shrubby transitional bogs. 

SALIX LIVIDA 

livid willow 

It is a deciduous dwarf shrub, about 50 cm tall. It grows at woodland margins and in sallow 

(willow) willow-poplar car. 

SALIX MYRTILLOIDES 

myrtle willow 

It is a deciduous dwarf shrub, 30 to 80 cm tall (rarely up to 2 m), with subterranean creeping 

stem. It is eutrophic. In boggy soils, it is usually found on intermediate sites. It is found sallow 

(willow) camrs, wet birch woodlands, and shrubby transitional bogs. 

SAMIX ROSMARINIFOL!A 

rosemary willow 

It is a deciduous dwarf shrub, 75 to 100 cm tall, with subterranean creeping stem. It is an 

oligotrophic acidophyte. In boggy soils, it is usually found on intermediate sites. It is found in 

sallow (willow) and willow-poplar carts and in transitional bogs. 

VACCINIUM MYRTILLUS 

myrtle whortleberry 

myrtle bilberry 

It is a deciduous ericoid dwarf shrub, 15 to 40 cm tall. It is oligotrophic acidophyte. In boggy 

soils, it is usually found on hummocks or tree boles. It is found in wet pine-sphagnum forests 

(subtypes 2, 3, 4), wet alderwoods, acidiphilous beechwoods, and pine-oak woodlands. It is 

infrequent in wet pine-sphagnum forests (subtypes 5, 6). 

A-17

VACCINIUM OXYCOCCUS 

small cranberry 

cranberry 

It is a evergreen ericoid dwarf shrub, up to 80 cm tall. It is an oligotrophic acidophyte. It is 

found in wet pine-sphagnum forests (subtypes 1, 2, 5, 6) and sphagnum bogs. Fruit harvested as 

a "crop." 

VACCINIUM ULIGINOSUM 

bog whortleberry 

bog bilberry 

It is a deciduous ericoid dwarf shrub, 50 to 100 cm tall. It is an oligotrophic acidophyte. In 

boggy soils, it is usually found on hummocks. It is found in moist pine forests, heaths, 

rhododendron thickets, and (in late successional stages) in sphagnum bogs. It is occasionally 

found in wet pine-sphagnum forests (subtypes 3, 4). Fruit harvested as a "crop." 

VACCINIUM VITJS-IDAEA 

cowberry 

cranberry 

mountain cranberry 

foxberry 

It is a dwarf ericoid shrub, up to 30 cm tall, with persistent leaves. It spreads by creeping 

rhizomes. It is an oligotrophic acidophyte. In boggy soils, it is usually found on hummocks and 

tree boles. It is found in wet pine-sphagnum forest (subtypes 1, 2, 3, 5, 6), wet alderwoods, 

heaths, rhododendron thickets, and (in late successional stages) in sphagnum bogs. Fruit 

harvested as a "crop." 

A-18

APPENDIX B 

HERBS 

Nonvascular and Lower Vascular Plants 

B-1

MOSSES 

AULACOMIUM PALUSTRE 

green moss 

In boggy soils, it is usually found in hollows. It is found in wet pine-sphagnum forests 

(subtype 1, 3, 4). 

HYLOCOMIUM PROUFORMIS 

green moss 

It is found in wet pine-sphagnum forest (subtype 3). 

PLEUROZIUM SCHREBERI 

In boggy soils, it is found on hummocks. 

(subtypes 1, 2, 3). 

green moss 

It is found in wet pine-sphagnum forests 

POHLIA NUTANS 

green moss 

It is oligotrophic. In boggy soils, it is usually found in intermediate sites. It is found in 

sphagnum bogs and spruce forests. 

POLYIRICHUM COMMUNE 

haircap moss 

It is oligotrophic. In boggy soils, it is usually found on hummocks and intermediate sites. It is 

found in wet pine-sphagnum forests (subtypes 3, 4), sphagnum bogs, and heaths. 

POLYTRICHUM STRICTUM 

haircap moss 

It is found in sphagnum bogs and heaths. 

B-2

SPHAGNUM ACUTIFOLUUM 

It is oligotrophic and an acidophyte. 

(subtypes 3, 5, 6). 

It is found in wet pine-sphagnum forests 

SPHAGNUM APICULATUM 

It is oligotrophic and an acidophyte. In boggy soils, it is usually found in hollows and 

intermediate sites. It is found in wet pine-sphagnum forests (subtype 2) and transitional bogs. 

SPHAGNUM CENTRALE 

It is oligotrophic and an acidophyte. It is found in wet pine-sphagnum forests (subtype 1). 

SPHAGNUM COMPACTUM 

It is oligotrophic and an acidophyte. It is found in wet pine-sphagnum forests (subtype 2) and 

transitional bogs. 

SPHAGNUM CUSPIDATUM 

It is oligotrophic and an acidophyte. In boggy soils, it is usually found in hollows. It is found 

in sphagnum bogs and transitional bogs. 

SPHAGNUM FUSCUM 

It is oligotrophic and an acidophyte. In boggy soils, it is usually found on hummocks. It is 

found in sphagnum transitional bogs, heath, and wet pine-sphagnum forests (subtype 2, 4, 5, 6). 

B-3

SPHAGNUM GIRENSOHNII 

It is oligotrophic and an acidophyte. In boggy soils, it is usually found on hummocks and 

intermediate sites. It is found in wet pine-sphagnum forests (subtype 1). 

SPHAGNUM MAGELLANICUM 

It is oligotrophic and an acidophyte. In boggy soils it is usually found on hummocks. It is 

found in wet pine-sphagnum forests (subtypes 2, 4, 5, 6). 

SPHAGNUM MEDIUM 


found in sphagnum bogs. 

SPHAGNUM PAPILLOSUM 

It is found in transitional bogs. 

SPHAGNUM PULCHRUM 


in wet pine-sphagnum forests (subtype 5, 6). 

SPHAGNUM RECURVUM 


in sphagnum and transitional bogs. 

B-4

SPHAGNUM RUBELLUM 

It is oligotrophic and an acidophyte. In boggy soils, it is usually found on hummocks and 

intermediate sites. It is found in sphagnum and transitional bogs. 

SPHAGNUM RUSSOVII (RUSSOWII) 


found in wet pine-sphagnum forests (subtypes 2, 5, 6) and sphagnum bogs. 

SPHAGNUM SQUARROSUM 

It is mesotrophic. In boggy soils, it is usually found in hollows and intermediate sites. It is 

found in transitional bogs and wet spruce forests. 

SPHAGNUM WARNSTORFI! 

It is oligotrophic and an acidophyte. It is found in wet pine-sphagnum forests (subtypes 1, 3). 

B-5

LICHENS 

CLADONIA ALECTORIA 

shrubby lichen 

It is found in dry pine woodlands. 

CLADONIA ALPESTRIS 


It is found in dry pine woodlands and often dominates in late succession. 

CLADONIA RANGIFERINA 



CLADONIA SILVA TICA 

forest lichen 


CLADONIA UNCIALIS 


CETRARIA ISLANDICA 


B-6

EQUISETACEAE 

EQUISETUM ARVENSE 

field horsetail38 

fens. 

It can grow in deep silt deposits. It is found in wet, flood, and fresh meadows and in sedge 

EQUISETUM HELEOCHARIS 

swamp horsetail 

It can endure prolonged 39 surface flooding and deep deposits of silt. It is found in wet 

pine-sphagnum forest (subtype 1), transitional bogs, and wet and flood meadows. 

EQUISETUM PALUSTRE 

marsh horsetail 

It can grow in deep silt deposits. It is found in sedge fens, transitional bogs, and wet and 

flood meadows. 

EQUISETUM PRA TENSE 

meadow horsetail 

fens. 

In flood meadows, it is found on ridges. It is found in flood and fresh meadows and sedge 

EQUISETUM SILVATICUM 

forest horsetail 

silvan horsetail 

It is found in wet pine-sphagnum forests (subtype 2), alder-ash carts, and forest openings. 

38 Eqiaetium is ais b known 

scouring nbh. 

39 M=e don 40 dayL 

B-7

EQUISETUM VARIEGATUM . variegated horsetail 

It can grow in deep silt deposits. It is found in wet pine-sphagnum forests (subtype 5), 

sphagnum bogs, transitional bogs, and sedge fens. 

B-8

MONOCOTYLEDONS 

Typhaceae 

TYPHA ANGUSTIFOLJA 

narrow-leafed cattail4o 

It is a hydrophyte. It is found in reed-bulrush fens. 

TYPHA LATIFOLIA 

broad-leafed cattail 

It is a hydrophyte. It is found in reed-bulush fens. 

Juncaceae 

JUNCUS AMBIGUUS 

It is found in wet and flood meadows. 

JUNCUS BULBOSA 

bulbous rush 

tuberous rush 

It is found in sphagnum and sphagnum-pine bogs. 

JUNCUS COMPRESSUS 

round-fruited rush 

It is found in reed-bulrush fens, and sedge fens, transitional bogs, and wet and flood 

meadows. 

4&pm ais kam ow ars 

mnaa 

B-9

JUNCUS EFFUSUS 

loose-spreading rush 

It is found in transitional bogs and wet meadows. 

JUNCUS INFLEXUS 

It is found in sedge fens. 

JUNCUS LAMPOCARPUS 


JUNCUS LEERSII 


JUNCUS SQUARROSUS 

squarrose rush 

It is found in heathlands, sphagnum bogs, and transitional bogs. 

JUNCUS TENAGEJA 

It is found in reed-bulrush fens, sedge fens, and wet and flood meadows. 

B-10

Cyperaceae 

CAREX ACU77FORMIS 

sharp sedge 

It is mesotrophic. In boggy soils, it is commonly found in hollows and intermediate sites. It is 

found in sedge fens and transitional bogs. 

CAREX APPROPINQUATA 

It is eutrophic. In boggy soils, it is commonly found on hummocks and intermediate sites. It 

is found in sedge fens and transitional bogs. 

CAREX AQUA T7LIS 

water sedge 

straight-leafed water sedge 

It is oligotrophic. In boggy soils, it is found in hollows and intermediate sites. It is found in 

wet and flood meadows, sedge fens, and transitional bogs. 

CAREX ARGYROGLOCHIN 

silver-arrow sedge 

It is found in alderwoods and spruce forest. 

CAREX CAESPITOSA 

tufted sedge, turfy sedge 

It is mesotrophic and a hydrophyte. In boggy soils, it is usually found in hollows and 

intermediate sites. It is found in wet and flood meadows, sedge fens, transitional bogs, and wet 

alderwoods. 

CAREX CANESCENS 

gray sedge 

hoary sedge 

It is found in wet meadows and transitional bogs. 

B-II

CAREX CHORDORRHIZA 

It is oligotrophic. In boggy soils, it is usually found in hollows and intermediate sites. It is 

found in sphagnum bogs, transitional bogs, and sedge fens. 

CAREX DIANDRA 

diandrous sedge 

It is rhizomatous. It is mesotrophic. In boggy soils, it is usually found in hollows and intermediate 

sites. It is found in sedge fens, transitional bogs, and wet meadows. 

CAREX DIOICA 

dioecious sedge 

creeping separate-headed sedge 

It is mesotrophic. In boggy soils, it is usually found on hummocks and in intermediate sites. 

It is found in sedge fens, transitional bogs, and wet meadows. 

CAREX DISPERMA 

dispermous sedge 

It is mesotrophic. In boggy soils, it is usually found on hummocks and in intermediate sites. 


CAREX DISTJCTA 

distictous sedge 

It is rhizomatous. It is found in transitional bogs and wet meadows. 

CAREX FLAVELLA 

little yellow sedge 

It is found in wet meadows and willow-poplar cars. 

B-12

CAREX GRACILIS 

sheep sedge, marsi, sedge 

slender sedge 

It is mesotrophic and a hydrophyte. It is nonresistant to subsurface flooding but can endure 

prolonged surface flooding. It can grow in moderate silt deposits. In boggy soils, it is usually 

found in hollows and intermediate sites. It is found in flood and wet meadows, sedge fens, and 

transitional bogs. 

CAREX HELEONASTES 

It is mesotrophic. In boggy soils, it is usually found in hollows. It is found in sedge fens, 

transitional bogs, and sphagnum bogs. 

CAREX INUMBRATA 

It is found in wet alderwoods. 

CAREX JUNCELLA 

little-rush sedge 


transitional bogs and wet meadows. 

CAREX LASIOCARPA 

hairy fruited sedge 


wet pine-sphagnum forests (subtypes 1, 2, 4), transitional bogs, sedge fens, and wet meadows. 

CAREX LEPORINA 

hare's-foot sedge 

It is found in fresh meadows and at forest margins. 

B-13

CAREX LIMOSA 

bog sedge 

mud sedge 

It is oligotrophic. In boggy soils, it is usually found in hollows. It is found in sphagnum 

bogs, transitional bogs, sedge fens, and wet meadows. 

CAREX LOLJACEA 

It is mesotrophic. In boggy soils, it is found on hummocks and intermediate sites. It is found 

in spruce forests. 

CAREX MURICATA 

muricate sedge 

greater prickly sedge 

It is found in transitional bogs and wet meadows. 

CAREX OMSKIANA 

Omskian sedge 

It is found in transitional bogs and reed-bulrush fens. 

CAREX PANICEA 

pink-leafed sedge 

It is eutrophic. In boggy soils, it is found in hollows and intermediate sites. It is found in wet 

meadows, and sedge fens. 

CAREX PANICULATA 

paniculate sedge 

great panicled sedge 

It is eutrophic. In boggy soils, it is found in hollows. It is found in sedge fens, reed-bulrush 

fens, and wet meadows. 

B-14

CAREX PAUCIFLORA 

few-flowered sedge 

It is oligotrophic. In boggy soils, it is found on hummocks and intermediate sites. It is found 

in transitional bogs and sphagnum bogs. 

CAREX PILOSA 

hairy sedge 

It is eutrophic. It is found in drier oak-hornbeam forests. 

CAREX PSUEDO-CYPERUS 

false flatsedge 

It is mesotrophic. In boggy soils, it is found in hollows and intermediate sites. It is found in 

sedge fens and transitional bogs. 

CAREX REMOTA 

distant-spiked sedge 

It is found in alder-ash carrs. 

CAREX RIPARIA 

riparian sedge 

great common sedge 

riverine sedge 

It is mesotrophic. In boggy soils, it is usually found in hollows. It is found in wet and flood 

meadows, sedge fens, and transitional bogs. 

CAREX VESICARIA 

bladder sedge 

short-beaked bladder sedge 

It is found in transitional bogs, sedge fens, and wet and flood meadows. 

CAREX VULPINA 

fox sedge 

great sedge 

It is found in wet meadows, sedge fens, and transitional bogs. 

B-15

CYPERUS FLAVESCENS 

yellow flatsedge yellow galingale 


CYPERUS PANONICUS 


ERIOPHORUM ANGUSTJFOLIUM 

narrow-leafed 

cottongrass 4 1 

It is found in sedge fens, transitional bogs, wet meadows, and wet pine-sphagnum forest 

(subtype 2). 

ERIOPHORUM GRACILE 

slender cottongrass 

It is found in sedge fens, transitional bogs, and wet pine-sphagnum forest (subtype 2). 

ERIOPHORUM LATIFOLJUM 

broad-leafed cottongrass 

It is found in sedge fens, transitional bogs, and wet meadows. 

ERIOPHORUM VAGINA TUM 

sheathed cottongrass 

haretail cottongrass 

In boggy soils, it is usually found on hummocks. It is found in wet pine-sphagnum forest 

(subtype 2), wet birch woodlands, sphagnum bogs, and transitional bogs. 

4 1 Erphmm is sao caled axomede and boVwOOL 

B-16

SCIRPUS ACICULARIS needle bulrush 42 

It is found in reed-bulrush fens. 

SCIRPUS COMPRESSUS 

compressed bulrush 

It is found in wet meadows. 

SCIRPUS EUPALUSTRUS 

marsh bulrush, marsh clubrush 

creeping scirpus 

It is found in wet and flood meadows and in reed-bulrush fens. 

SCIRPUS HOLOSCHOENUS 

It is found in reed-bulrush fens and wet meadows. 

SCIRPUS LACUSTRIS 

lake bulrush 

lake clubrush 

It is mesotrophic and a hydrophyte. In boggy soils, it is found in hollows. It is dominant in 

reed-bulrush fens. It is found in wet meadows and sedge fens. 

SCIRPUS MAMILLATUS 

It is found in reed-bulrush fens. 

SCIRPUS OVA TUS 

ovate bulrush 


42 Bulpjh is also spelled bullrus. 

B-17

SCIRPUS PA UCIFLOR US 

few-flowered bulrush 

chocolate-headed clubrush 


SCIRPUS SILVATICUS 

wood scirpus, wood bulrush 

It is mesotrophic. In boggy soils, it is found in hollows and intermediate sites. It is found in 

wet meadows, reed-bulrush fens, and spruce forests. 

B-18

POACEAE- 3 

AGROPYRON CANINUM 

dog wheatgrass 

dog couchgrass 

doggrass 

It is mesotrophic. It is found in oak-hornbeam and spruce forests. 

AGROPYRON REPENS 

quackgrass 

creeping couchgrass 

creeping wheatgrass 

creeping doggrass 

twitchgrass, quitchgrass 

quickgrass, quakegrass 

squitchgrass 

Is a mesohydrophyte. It is nonresistant to subsurface flooding but can endure prolonged 

surface floods and deep deposits of silt. It is often dominant in flood meadows. It is also found in 

fresh meadows and at forest margins and in openings. 

AGROSTIS ALBA 

redtop bentgrass 

white bentgrass 

marsh bentgrass 

redtop florin 

bonnetgrass 

It is mesotrophic. It is a mesophyte and can endure prolonged surface flooding. It can adjust 

to acid soils. It is found in wet and flood meadows. It is recommended for cultivation in flood 

meadows and drained bogs. 

AGROSTIS CANINA 

velvet bentgrass 

dog bentgrass 

brown bentgrass 

finetop bentgrass 

mountain redtop 

It is mesotrophic. It is found in wet and flood meadows. 

B-19

AGROSTIS STOLONIFERA 

carpet bentgrass 

creeping bentgrass 

black twitch 

It is mesotrophic and a mesophyte. It can endure moderate"4 subsoil flooding. It is found in 

wet and flood meadows and on the peripheries of transitional bogs. 

ALOPECURUS ARUNDINA CEUS 

creeping foxtail 

In flood meadows, it is found in interridge depressions. 

meadows, sedge fens, and transitional bogs. 

It is found in wet and flood 

ALOPECURUS GENICULATUS 

marsh foxtail, water foxtail 

floating foxtail 

elbowit-grass 

It is found in wet meadows, sedge fens, and transitional bogs. 

ALOPECURUS PRA TENSIS 

meadow foxtail 

It is a mesohydrophyte. It can endure moderate subsoil flooding and prolonged surface 

flooding. It is found in wet, flood, and fresh meadows. It is recommended for cultivation in fresh 

and flood meadows and drained bogs. 

ALOPECURUS TENUIS 

slender foxtail 

It is found in flood and fresh meadows and birch woodlands. 

ANTHOXANTHUM ORDORATUM 

sweet vernalgrass 

sweet-scented vernalgrass 

It is mesotrophic and a mesophyte. It is found in fresh and flood meadows. 

4420 to 40 days. 

B-20

ARRHENATHERUM ELATIUS 

tall oatgrass 

meadow oatgrass 

false oatgrass 

French ryegrass 

It is a mesophyte. It is nonresistant to subsurface flooding but can endure short4 5 surface 

floods. It is found in fresh meadows and flood meadows with short inundation periods. 

BECKMANNIA ERUCIFORMIS 

common sloughgrass 

It is a mesohydrophyte. It is able to endure prolonged surface flooding. It is found in wet, 

flood, fresh meadows, and reed-bulrush fens. It is recommended for cultivation in drained bogs. 

BRA CHYPODIUM PINNA7TUM 

heath falsebrome 

It is often dominant at forest and woodland margins and in forest openings. 

BRA CHYPODIUM SILVATICUM 

forest falsebrome 

slender flasebrome 

It is found in moist pine woodlands. 

BRIZ4 MEDIA 

common quaking grass 

doddering jockies 

doddering dillies 

earthquakes 

It is found in fresh meadows and forest openings. 

45Ln thma 20 days. 

B-21

BROMUS INERMIS 

awnless bromegrass 

awnless brome 

smooth brome 

Hungarian brome 

It is a mesophyte. It cannot resist subsoil flooding but can endure moderate46 surface flooding 

and deep deposits of silt. In flood meadows, it is found in interridge depressions. It is found in 

flood and fresh meadows and at forest margins and in openings. It is recommended for cultivation 

in fresh meadows and drained bogs. 

CALAMAGROSTIS ARUNDINACEA 

forest reedgrass 

sand reedgrass 

beachgrass 

It is a mesophyte. It inhibits pine regeneration in successional woodlands. Found in linden 

and aspen groves and birch woodlands. 

CALAMAGROSTIS EPIGEIOS 

chee reedgrass 

small reedgrass 

wood smallreed 

It is mesotrophic and a mesophyte. It can grow in deep silt and can endure moderate surface 

flooding. It inhibits pine regeneration in successional woodlands. It is found in flood meadows, 

wet pine-sphagnum forest (subtype 2, 4), linden and aspen groves, and birch woodlands. 

CAL4MAGROSTIS NEGLECTA 

slim-stem reedgrass 

It is mesotrophic. It is found in wet meadows and at the peripheries of transitional bogs. 

CALAMAGROSTIS PHRAGMITOIDES 

tall reedgrass 

It is oligotrophic. It is found in wet meadows and damp spruce forests. 

4620 so 40 dap. 

B-22

CA TABROSA AQUA TI CA 

brookgrass 

water whorigrass 


CYNOSURUS CRISTA TUS 

crested dogtail 

crested dog's-tail grass 

It is a mesophyte. It is found in fresh meadows and moist forest openings. 

DACT"YJS GLOMERATA 

common owhardgrass 

cocksfoot 

rough cocksfoot 

It is a mesophyte and can endure moderate surface flooding. It is not able to grow in deep or 

moderate silt deposits but can grow in thin deposits. It is found in fresh and flood meadows and in 

forest openings. It is recommended for cultivation in fresh meadows. 

DESCHAMPSIA CAESPITOSA 

tufted hairgrass 

It is mesotrophic and a mesohydrophyte. It can endure moderate subsoil flooding. It grows in 

moist to very wet soils and in thin deposits of silt. It is found in wet and flood meadows, in 

alderwoods, and spruce forests. 

FESTUCA ARUNDINACEA 

tall fescue, reed fescue 

alta fescue 


FESTUCA GIGANTEA 

giant fescue 

tall-bearded fescue 

It is found in wetter oak-hornbeam forests. 

B-23

FESTUCA OVINA 

sheep fescue 

sheep's fescue 

It is mesotrophic. It can grow in thin deposits of silt and can e;. lure short periods of flooding. 

In flood meadows, it is found on ridges. It is found in fresh meadows and flood meadows with 

only short floods, as well as dry pine woodlands and on dry sandy hills and dunes. 

FES7TUCA POLESICA 

Polish fescue 

It is found in dry pine woodlands and on dry sandy hills and dunes. 

FESTUCA PRATENSIS 

meadow fescue 

It is mesotrophic. It is a mesophyte but can endure moderate surface flooding. It is not able to 

grow in deep silt deposits but can grow in moderate deposits. It is found in wet, flood, and fresh 

meadows and in moist pine-sphagnum forests. It is recommended for cultivation in flood and 

fresh meadows and drained bogs. 

FESTUCA RUBRA 

red fescue, creeping fescue 

It is mesotrophic and a mesohydrophyte. It can endure moderate subsoil and surfacc flooding. 

It can grow in poor soils and in thin deposits of silt. It is found in wet, flood, and fresh meadows, 

as well as moist pine woodlands. 

GLYCERIA AQUATICA 

reed mannagrass 

reed sweetgrass 

reed meadowgrass 

reed whitegrass 

It is mesotrophic and a hydrophyte. It is found in reed-bulrush fens, sedge fens, and wet and 

flood meadows. 

B-24

GLYCERIA FRUITANS 

floating mannagrass 

It is mesotrophic and a hydrophyte. It is found in reed-bulrush fens and wet and flood 

meadows. 

GLYCERIA HEMORALIS 

It is found in wet alderwoods and reed-bulrush fens. 

GLYCERIA LITHUANICA 

Lithuanian mannagrass 

It is eutrophic. It is found in wet alderwoods and spruce forests. 

GLYCERIA PLJCATA 

plicate mannagrass 


HIEROCHLOE ODORATA 

common sweetgrass 

sweet holygrass 

sweet vanillagrass 

sweet senecagrass 

It is mesotrophic. It is found in wet and flood meadows. 

HOLCUS LANATUS 

common velvetgrass 

meadow softgrass 

Yorkshire fog 

It is found in fresh meadows. 

HOLCUS MOLI.S 

downy velvetgrass 

creeping softgrass 


B-25

KOELERIA DELAVIGNII 

Delavign junegrass 

It is found in flood meadows and oak forests. 

KOELERIA GLAUCA 

glaucus junegrass 

It is found in dry pine woodlands and on dry sandy hills and dunes. 

LEERSIA ORYZOIDES 

rice cutgrass 

rice whitegrass 

false rice 

It is found in sedge and reed-bulrush fens. 

LOLIUM PERENNE 

perennial ryegrass 

English ryegrass 

perennial raygrass 

It is a mesophyte and is nonresistant to subsurface flooding. It has only weak resistance to 

drought It is found in fresh meadows and is recommended for cultivation there. 

MELICA NUTANS 

mountain melic 

It is found in aspen groves, oak-hornbeam forests, and forest openings. 

MILIUM EFFUSSUM 

wood millet 

spreading milletgrass 

It is mesotrophic. It is found in alderwoods, oak-hornbeam forests, birch woodlands, aspen 

and linden groves, and spruce forests. 

B-26

MOLJNIA COERULUA 

purple heathgrass 

purple moorgrass 

purple molinia 

It is mesotrophic and a hydrophyte. In boggy soils, it is usually found on hummocks and 

intermediate sites. It is found in wet meadows, moist pine forests, wet alderwoods, at the 

peripheries of transitional bogs, and in spruce forests. 

NARDUS STRICTA 

matgrass 

It is oligotrophic and mesophytic. It can grow in poor soils and in thin deposits of silt. In 

boggy soils, it is usually found in intermediate sites. It is found in wet, flood, and fresh 

meadows, heathlands, moist pine forests, and wet pine woodlands with flowing water. It grows 

in moist and very moist "bor" and "subor" sites. 

PHALARIS ARUNDINACEA 

reed canarygrass 

reed swordgrass 

It is mesotrophic and a mesohydrophyte. It is able to endure prolonged surface flooding. It 

can grow in deep silt deposits. In boggy soils, it is usually found in intermediate sites. It is found 

in reed-bulrush fens and in wet, flood, and fresh meadows. It is recommended for cultivation in 

fresh and flood meadows and drained bogs. 

PHLEUM PRATENSE 

timothy, common timothy 

meadow timothy 

common cat's-tafilgrass 

meadow cat's-tailgrass 

herd's grass 

It is mesotrophic and a mesophyte and can endure moderate, but not prolonged, surface 

flooding. It is found in flood and fresh meadows. It is recommended for cultivation in flood and 

fresh meadows and drained bogs and is often supplementally sown for forage. 

B-27

PHRAGMITES COMMUNIS 

common reed 

giant reed 

ditch reed 

pool reed 

It is mesotrophic and a hydrophyte. In boggy soils, it is usually found in hollows and 

intermediate sites. It is dominant in reed-bulrush fens. It is found in wet pine forests with flowing 

water, in wet alderwoods, and wet meadows. 

POA ANNUA 

annual meadowgrass 

dwarf meadowgrass 

annual bluegrass 

causeway grass 

Suffolkgrass 

low speargrass 


POA PALUSTRIS 

fowl meadowgrass 

fowl bluegrass 

swamp meadowgrass 

false redtop 

It is mesotrophic and a mesohydrophyte. It can endure prolonged surface flooding and can 

grow in moderate silt deposits. In flood meadows, it is found in interridge depressions. It is 

found in wet, flood, and fresh meadows and sedge fens. 

POA PRATENSIS 

smooth meadowgrass 

smooth-stalked meadowgrass 

Kentucky bluegrass 

junegrass 

speargrass 

greengrass 

It is a mesophyte and can endure moderate surface flooding. It is not able to grow in deep or 

moderate silt deposits but can grow in thin deposits. It is found in flood and fresh meadows. It is 

recommended for cultivation in flood and fresh meadows and drained bogs and is often 

supplementally sown for forage. It is also found in dry pine woodlands and on sandy hills and 

dunes. 

B-28

PAO TRIVIALIS 

rough meadowgrass 

rough-stalked meadowgrass 

rough bluegrass 

rough-stalked bluegrass 

evergreen 

It is mesotrophic and a mesophyte. It can endure moderate subsoil flooding. It is found in 

flood and fresh meadows and is often supplementally sown for forage. 

SETARIA GLA UCA 

glaucus bristlegrass 

glaucus foxtail 

pigeongrass 

bottlegrass 


SETARIA WRIDIS 

green bristlegrass 

green foxtail 

green pigeongrass 


TRISETUM SIBIRICUM 

Siberian trisetum 

Siberian falseoat 

Siberian oatgrass 


B-29

DICOTYLEDONS 

DROSERA ROTUNDIFOUA 

round-leaf sundew 

It is carnivorous, oligotrophic, and an acidophyte. In boggy soils, it is usually found on 

hummocks. It is found in sphagnum bogs and transitional bogs. 

MEDICAGO FALCATA 

yellow lucerne 

yellow alfalfa 

It is a mesophyte and can endure short surface floods. It is not able to grow in deep silt 

deposits. In flood meadows, it is found in interridge depressions. It grows in fresh meadows, 

flood meadows with only short inundations periods, and drained bogs. It is often supplementally 

sown for forage. 

MEDICAGO LUPULINA 

black medic 

hop medic 

nonesuch 

It is a mesophyte, It grows in fresh meadows and is often supplementally sown for forage. 

MEDICAGO SA TIVA 

common lucerne 

common alfalfa 

purple medic 

Spanish trefoil 

It is a mesophyte and can endure short surface floods. It will not grow in excessively moist 

soils or soil with even moderately long flooding. It grows in fresh meadows, flood meadows with 

short inundation periods, and drained bogs and is often supplementally sown for forage. 

MELILOTUS ALBA 

white sweetclover 

white melilot 

Bokara clover 

It is a mesophyte. It is found in fresh meadows and is often supplementally sown for forage. 

B-30

MELULOTUS OFFICINAUS 

yellow sweetclover 

yellow melilot 

It is a mesophyte. It grows in fresh meadows and is often supplementally sown for forage. 

TPJFOLIUM HYBRIDUM 

alsike clover, pink clover 

Swedish clover 

It is a mesophyte. It is not able to grow in deep silt deposits but can grow in moderate 

deposits. It can endure moderately acid soils. It grows in flood and fresh meadows and pastures. 

It is recommended for cultivation in fresh and flood meadows and drained bogs and is often 

supplementally sown for forage. 

TRIFOLJUM PRA TENSE 

red clover, meadow clover 

common clover 

broad-leaf clover 

cowgrass clover 

meadow trefoil 

It is mesotrophic and a mesophyte. It is not able to grow in deep silt deposits but can grow in 

moderate deposits. It can withstand short surface floods and soil acidity. It is found in fresh 

meadows, flood meadows with short inundation periods, and pastures on relatively fertile, moist 

(but well-drained) soils. It is recommended for cultivation in fresh meadows and drained bogs and 

is often supplementally sown for forage (see T. sativum). 

TRIFOLUUM REPENS 

white clover 

Dutch clover 

It is mesotrophic and a mesophyte. It grows best on moderately moist soils and can endure 

short surface floods. It is not able to grow in deep silt deposits but can grow in moderate deposits. 

It is found in fresh meadows and flood meadows with short inundation periods, usually on welldrained 

soils It is recommended for cultivation in fresh and flood meadows and is often 

supplementally sown for forage. 

TRIFOLJUM SATIVUM 

field clover 

This is a name for the cultivated form of T pratense. 

B-31

..APPENDIX C 

CROP PLANTS 

C-I

A VENA SA TIVA 

oats (spring) 

In the Poles'ye, it is cultivated on moderate podzols. 

BETA VULGARIS 

sugar beet, table beet 

forage or fodder beet 

It is grown for forage and silage as well as human consumption. In the Poles'ye, it is 

cultivated on moderate podzols and, with wheat, "dominates" the best soils. 

BRASSICA OLERACEA var. A CEPHALA 

kale 

BRASSICA OLERACEA var. CAPITATA 

cabbage 

BRASSICA RAPA 

turnips 

CANNABIS SATIVA 

hemp 

It is grown for fiber. 

CUCUMIS SA TIVA 

cucumbers 

DAUCUS CAROTA 

table carrots 

fodder carrots 

In the Poles'ye it is grown in drained bogs. 

FAGOPYRUM ESCULENTUM 

buckwheat 

In the Poles'ye, it is cultivated on poor and moderate podzols. 

C-2

HORDEUM VULGARE 

barley (winter and spring) 


HUMULUS LUPULUS 

hops 

In the Poles'ye, it is cultivated on moderate podzols. 

UNUM USITATISSIMUM 

flax 

It is grown for fiber. Fine-quality fiber can be obtained from plants grown on podzolic and 

gley soils with considerable fertilizing. In the Poles'ye, it is cultivated on poor and moderate 

podzols. 

PANICUM MILIACEUM 

millet 


SECALE CEREALE 

rye (winter) 


SOLANUM TUBEROSUM 

potato 


TRITICUM AESTIVUM 

wheat (winter and spring) 

In the Poles'ye, it is grown only on the best soils and, with beets, "dominates" them. 

C-3

ZEA MAYS 

corn 

It is usually grown as a forage or silage plant. In the Poles'ye, it is cultivated on moderate 

podzols. 

C-4




ATTN: BHS 


ATTN: EXH 

ATTN: MRA 

ATTN: PHY 

ATTN: RSD 





ATrN: NUC FORCES & ARMS CONTROL PLCY 





DNA-TR-92-37-V3 


ATTN: C-332 

ATTN: JNSV 


ATrN: ICAF TECH LIB 


ATrN: LIBRARY 






ATTN: F HEGGE 



ATrN: DR SCHNEITER 



ATTN: DB 

ATTN: JLT 

5 CYS ATTN: DB-4 RSCH RESOURCES DIV ATTN: JPEP 

ATTN: DB-SC 

ATTN: DB-6B DB-6E 

U S EUROPEAN ATTN: ECJ-6 COMMAND/ECJ-6-DT 

ATTN: DIANPA-2 

ATTN: DIW-4 

ATTN: DN 

ATrN: DT 


ATTN: OS 



ATTN: DIC/2C 




ATTN: ECJ2-T 

U S EUROPEAN COMMAND/ECJ3-CCD 

ATTN: ECJ-3 

U S EUROPEAN COMMAND/ECJ4-LW 

ATTN: ECJ4-LW 


ATTN: ECJ5-N 

U S EUROPEAN COMMANDiECJ7/LW 

ATrN: ECJ-7-LW 




ATIN: NANF 

ATTN: DUSP/P 

ATTN: NASF 

ATTN: USD/P 

ATTN: OPNA 

ATTN: OPNS 


ATTN: RAEE 

20 CYS ATTN: RARP COMBAT MATERIAL EVAL ELEMENT 

2 CYS ATTN: TITL ATTN: SECURITY ANALYST 



2 CYS ATTN: DTIC/FDAB ATTN: DAMO-SWN 



ATTN: FCPR 


ATTN: FCPRT 

ATTN: SLCH-NW-RS JOSIP SOLN 


ATTN: SLCIS-M-TL 



ATTN: FCNM ATIN: J-2 

ATTN: J-5 


ATTN: ITV NUCLEAR EFFECTS DIVISION 

2CYS ATTN: T'TV3416THTTSO 


Dist-1

DNA-TFI-92-37-V3 (DL CONTINUED) 


ATFN: COMMANDANT 





USACACDA 




ATTN: ATSB.CTD 


MARINE CORPS 



ATTN: PSI GIRASP 

ATTN: SLCBR-D 



ATfN: SLCBR-TB ATTN: CODE 9642-B 

ATTN: 

ATTN: 

SLCBR-VL-I 

SLCBR-VL-V 

DR KLOPCIC 





ATTN: JOHN GERDES ATGRCDE 1424 OL 

ATTN: MIKE STANKA 



NAVAL 

ATTESACO 

RESEARCH LABORATORY 

1240 

ATTN: ATZL-SWJ-CA ATTN: CODE 1240 

ATIN: ATZL-SWT-A ATTN: CODE 2627 


ATIN: TECHNICAL LIBRARY 

U S ARMY FIELD ARTILLERY SCHOOL 

ATTN: ATSF-C:D 


ATTN: AF-OPTS 


ATTN: C WARD AIFRTC 


ATTN: PMS-423 

ATTfN: SEA-06GN 



ATTN: G RIEL 




U S ARMY INFANTRY CENTER ATfTN: CODE E-1 1 




U S ARMY ITAC 

ATTN: LIBRARY 

ATrN: IAX-Z 


U S ARMY LABORATORY COMMAND 






U S ARMY MATERIAL COMMAND ATTN: CODE 222 

ATTN: DRCDE-D 



NUCLEAR WEAPONS TNG GROUP. PACIFIC 

ATTN: MONA-NU ATTN: CODE 32 

4 CYS ATTN: MONA-NU DR D BASH 

OFFICE OF CHIEF OF NAVAL OPERATIONS 


ATTN: NIS-22 

ATTN: STEC$-NE(70) 

ATTN: NOP 06D 

ATTN: NOP 403 

U S ARMY WAR COLLEGE ATTN: NOP 50 

ATTN: LIBRARY ATTN: NOP 60 


ATTN: NOP 60D 

ATTN: MOP 603 

ATTN: NOP 91 


AITN: BEHAVORIAL SCI & LEADERSHIP ATTN: OP 654 


ATTN: PMSOPMA523 


US ARMY MATERIEL SYS ANLYSIS ACTVY 


OPERATIONAL ATrIN: TEST COMMANDER & EVALUATION FORCE 

Dist-2


PLANS. POLICY & OPERATIONS 

ATTN: CODE-P 





ACADEMY LIBRARY DFSELD 

ATTN: LIBRARY 

AFIS/INT 

ATTN: INT 

ATTN: NIO-T 




ATfN: G ORRELL NP-CP 

ATTN: R SANDS 


ATTN: PM/STM 



ATTN: S YANIV 




ATTN: AUL-LSE 

ATTN: LIBRARY 


ARES CORP 

ATTN: A DEVERILL 


ATTN: F GREY 


2 CYS ATTN: AF/SAMI HORIZONS TECHNOLOGY, INC 

ATTN: J MARSHALL-MIES 


ATrN: SAF/AOA 

DEPUTY CHIEF OF STAFF FOR PLANS & OPERS 

ATTN: AFXOOSS 

FOREIGN TECHNOLOGY DIVISION 

ATTN: CCN 

ATTN: SDA 


ATTN: DASIAC 


ATTN: R STOHLER 


ATrN: DASIAC 

TACTICAL AIR COMMANDIXPSC 

ATTN: ACC/XP-JSG 





ATrN: RADIATION SCIENCES DIV 

ATrN: DOCUMENT CONTROL 

ATrN: DOUGLAS C YOON 





ATrN: DR T JONES 


LAWRENCE LVERMORE NATIONAL LAB 

ArTN: R LAUGHERY 

ATrN: Z DIVISION LIBRARY 

MISSION RESEARCH CORP 


ATTN: DR NEIL GOLDMAN 

ArTN: D STROTTMAN 



2 CYS ATrN: E L PAINTER 

MARTIN MARIETTA ENERGY SYSTEMS INC 

2 CYS ATrN: F W WHICKER 

ATTN: SANTORO 2 CYS ArmN: G ANNO 

ATTN: G KERR 

ATrN: H BRODE 

ATTN: J WHITE 



ATrN: D GORMLEY 


2 CYS ATTN: G MCCLELLAN 

ATTN: TECH LIB 3141 


OlHER GOVERNMENT 

ATrN: D KAUL 

ArmN: E SWICK 


ATTN: L HUNT 

ATTN: COUNTER-TERRORIST GROUP 


ArmN: DIRECTOR OF SECURITY 

ATTN: W WOOLSON 


Dist-3

DNA-TR42-37-V3 (DL CONTINUED) 


ATTN: J MCGAHAN 




ATN•: R CRAVER 




ATTN: S LEVIN 



TRW SPACE & DEFENSE SECTOR 

ATTN: DR BRUCE WILSON 



Dist-4

Armed Forces Rad I Research Institute 

Retrospective Reconstruction of 

Radiation Doses of 

Chernobyl Liquidators by 

Electron Paramagnetic Resonance 

A ove, 

reieosU 

Scientific Center of Radiation Medicine 

Academy of Medical Sciences, Ukraine 

19980223 032

! 35b PL • .•,s - 

Retrospective Reconstruction of 

Radiation Doses of 

Chernobyl Liquidators by 

Electron Paramagnetic Resonance 

Authored by 

Scientific Center of Radiation Medicine 

Academy of Medical Sciences, Ukraine 

254050, Kiev-5O, Melnikova 53 

Vadim V. Chumak, Ilia A. Likhtarev, 

Sergey S. Sholom, Larisa F. Pasalskaya, 

and Yuri V. Pavienko 

Published by 


Bethesda, Maryland, USA 

Editor and NIS Initiatives Coordinator 

Glen I. Reeves, M.D.

Cleared for public release: distribution unlimited. 

AFRRI Contract Report 97-2 

Printed December 1997 

Defense Nuclear Agency Contract DNA001 -95-C-001 7 

For information about this publication, write Armed Forces Radiobiology Research 

Institute, 8901 Wisconsin Avenue, Bethesda, MD 20889-5603, USA, or telephone 

011-301-295-0377, or send electronic mail to reeves@mx.afrri.usuhs.mil. Find 

more information about AFRRI on the Internet's World Wide Web at http://www. 

afrri.usuhs.mil. 

This and other AFRRI publications are available to qualified users from the Defense Technical 

Information Center, Attention: OCP, 8725 John J. Kingman Road, Suite 0944, Fort Belvoir, 

VA 22060-6218; telephone 703-767-8274. Others may contact the National Technical 

Information Service, 5285 Port Royal Road, Springfield, VA 22161; telephone 703-487- 

4650. AFRRI publications are also available from university libraries and other libraries 

associated with the U.S. Government's Depository Library System.

Preface 

On April 26, 1986, Reactor #4 at the Chernobyl Nuclear Power Plant 

near Pripyat, Ukraine, exploded, releasing millions of curies of radioactive 

materials into the environment. The reaction was swift, with 

firefighters and medics being mobilized within hours of the accident. 

After initial care of the injured and instituting measures to prevent 

further exposure of the general population in the vicinity, attention 

was turned to cleanup of the damaged reactor and the radioactive 

debris. Hundreds of thousands of workers (called "liquidators") were 

employed in the cleanup. Authorities were aware of the risks of 

immediate and long-term health effects to these people and took 

measures to limit the dose received. 

With the passage of time, the liquidators have developed leukemia, 

solid tissue neoplasms, cardiovascular disease, and other illnesses. 

The question of what relationship these illnesses, which also occur in 

unexposed populations, bear to the radiation exposure received at 

Chernobyl naturally arises. This question is vitally important, not only 

for compensation purposes, but also for advancing our knowledge of 

the effects of protracted radiation exposure on human health and for 

setting or reevaluating safety standards. But the critical first step in 

finding the answer is accurately ascertaining what dose was actually 

received. Physical dosimeters were not always used, and were not 

always used reliably, during the several operations involved in the 

Chernobyl cleanup. It is necessary to employ accurate, reliable biological 

indicators of radiation effects to reconstruct exposure received. 

The implications of using electron paramagnetic resonance (EPR) 

analysis as one such state-of-the-art technique in performing dose 

reconstruction clearly go beyond Chernobyl. There are other areas of 

the world with widespread environmental contamination at dose 

levels sufficient to cause adverse health effects, such as along the Techa 

River in the Southern Urals region of Russia and in the areas surrounding 

the former nuclear weapons test site near Semipalatinsk, 

Kazakstan. There have also been accidents involving small numbers 

111

Retrospective Reconstruction of Radiation Doses by EPR 

of individuals where the actual dose received is, for one reason or 

another, not accurately known. 

Because of the importance of a means of accurately and precisely 

estimating cumulative radiation exposure for epidemiologic studies, 

the Armed Forces Radiobiology Research Institute (AFRRI) elected to 

fund this study. The authors are highly competent investigators who 

also have connections with similarly skilled independent scientists 

who could refine their techniques and improve their results. In addition, 

they have access to the data from the Chernobyl liquidators 

accessible to followup, most of whom are now in Ukraine. Although 

the scope of this study was limited, its results should provide a significant 

step in improvingthe utility of EPR in dose reconstruction as well 

as in getting a clearer picture of the magnitude of the radiation 

exposure actually received at the world's most tragic reactor accident. 

Glen I. Reeves, M.D. 

Editor and NIS Initiatives Coordinator 

AFRRI 

iv

Contents 

Preface ..................................................... 

A b stract .................................................... 1 

Introduction ................................................ 3 

Task I. Development of a Routine High-Performance EPR- 

Dosim etric Technique ................................. 5 

EPR Measurements, Spectra Analysis, and Dose Reconstruction 5 

Sam ple Collection ....................................... 8 

Sam ple Preparation ...................................... 8 

Task 2. Quality Assurance Program .......................... 11 

Stage 1. Intercalibration With Homogenized Samples ....... 11 

Intercomparison Design ......................... 12 

Methods and Results ............................. 12 

Stage 2. Intercalibration With Whole Teeth Irradiated Under 

Laboratory Conditions ......................... 14 


Methods and Results ............................. 15 

Stage 3. Intercomparison of Teeth From Liquidators ........ 17 


M ethods and Results ............................. 18 

D iscussion ............................................. 19 

Task 3. Test of Practical Dose Determination .................. 21 

Dose Reconstruction .................................... 21 

System Developm ent ................................... 27 

D iscussion ................................................. 29 

Effects of M edical X Rays ................................ 29 

Effects of UV Light ...................................... 30 

Nonlinearity of Dose Response Curves .................... 31 

Sum m ary ................................................. 33 

References ................................................ 35 

Appendix-Identification Form for Tooth Sampling ............. 37 

iii 

V

Abstract 

Accurate, reliable dose reconstruction is a critical component in the 

epidemiological followup of liquidators. Dosimetry of teeth by electron 

paramagnetic resonance (EPR) is a state-of-the-art laboratory technique 

that is key to this effort. The Scientific Center of Radiation Medicine 

(SCRM) has developed and refined this technique in order to meet the 

practical demands of large-scale epidemiologic followup of the Chernobyl 

liquidators. Independent analysis using similar technology was 

performed by investigators at the University of Utah and showed good 

correlation with the SCRM results. The lower limit of detection for 

reliable dose reconstruction was 100 mGy. Techniques were applied to 

samples from approximately 135 liquidators involved in cleanup activities 

within the first 2 years after the Chernobyl accident in 1986. 

Mean dose was 287 mGy, geometric mean was 205 mGy, and median 

dose value was 200 mGy. The reconstructed dose values range from 30 

to 2220 mGy. Correlation of results between the two institutions was 

generally within 17%. This report also addresses some confounding 

factors (previous medical x-ray exposures, ultraviolet light effects on 

anterior teeth, nonlinearity of dose response curves below 100 mGy) 

and how to deal with them. 

Keywords: dosimetry, retrospective, EPR, technique, doses, liquidators, 

Chernobyl

Introduction 

Electron paramagnetic resonance (EPR) dosimetry using teeth is generally 

accepted as a highly attractive method for reconstructing individual 

radiation doses long after exposure [1]. However, until recently, 

EPR dosimetry was generally considered a unique state-of-the-art laboratory 

procedure unsuitable for practical dose reconstruction. Moreover, 

the accuracy and consistency of results produced by this technique 

were not proven. 

A tool for dose reconstruction is acutely needed, particularly in the 

Chernobyl situation. Although these dose records are incomplete and 

not all liquidators know their doses, an official dose record is included 

in most of the identification forms of cleanup workers. The objective 

of the effort funded by DNA (contract number DNA001-95-C-0017) 

was to develop reliable retrospective estimates of radiation doses 

received by the Chernobyl liquidators. This goal was approached in 

three stages. 

First, the Scientific Center of Radiation Medicine (SCRM) developed 

the EPR dosimetry technique to provide reliable and efficient reconstruction 

of doses. Each of the basic steps of EPR-dosimetric methodology 

was subjected to rigorous analysis and optimization. Methodological 

research and development of the technique are far beyond the 

scope of this contract, but the technique has been explicitly presented 

elsewhere [2,3] and is widely accepted worldwide. Figure 1 illustrates 

that every step incorporated a number of innovations and specific 

features. This improved version of EPR dosimetric technique was 

employed for both routine dose reconstruction and interlaboratory 

cross-calibration. 

Second, a sophisticated cross-calibration effort was undertaken in 

order to assure the quality of results. This effort included a series of 

internal tests as well as intercomparisons with an experienced counterpart 

in the USA. 

3


Third, after completion of the above tasks, routine reconstruction of 

doses to liquidators began. The doses to 135 individuals were assessed 

in accordance with the technique designed in Task One. These results 

were entered into a database and are available for researchers. 

Sampl 

Acquisition 

Collection of teeth along with extended ID data 

Separate processing of inner and outer parts of tooth to account for possible x-ray 

exposure 

P State-of-the-art purification of enamel in several progressive steps under EPR 

quality control 

ck Spetr rdincof p Subtraction mixture of nonirradiated of the instrument teeth background from young people signal 

-b Use of Mn:MgO standard for adjustment of magnetic field 

"- approim of e spectra after every o s pr a e m t 

stage of purification Deovlto Manipulation of background and measured signals by 

"F Use of Mn:MgO spectrometric of EPR S etra varying amplitude under fixed g-factor and width of signals 

standard 

Radiation-induced signal intensity determined as peak-to- 

". Recording of the signal of empty• peak amplitude 

resonator for use as instrument! 

background "- Monte Carlo procedure for 

x-a Medca dosear iskron determined: of 

of Cumulative 

, Empirical relationship deerindro for rough 

"• Use of additive dose method with 4-6 irradiations Dose assessment of uncertainty 

"• Additive dose values are at the expected level of the 

cumulative dose. % 

"• Laboratory irradiation of up to 20 samples in parallel at 

Evlution 

137Cs source with TILD control 

"• Least squares of thefEPPRtechniqu best fit 

"• Account for natural background and medical exposure components of cumulative dose 

"• Dose from the natural background is a function of the age of patient and type of tooth 

•Medical x-ray dose is determined: 

-precisely with differential method by analysis of inner and outer parts of tooth 

-approximately using empirical values of dose per x-ray examination 

of the 

"Accidental" 

Assessment of 

Uncertainty 

•. 

CUK.D 

Figure 1. Principal steps of EPR dosimetry and innovative features of the SCRM version 

of the EPR technique

Task 1 

Development of a Routine High- 

Performance EPR-Dosimetric Technique 

Extensive scientific and technological investigations were conducted 

in order to make EPR dosimetry usable as a routine tool for dose 

reconstruction. Special attention was paid to ensuring reliable results. 

With respect to the demands of epidemiological followup, optimal EPR 

technique must meet the following requirements: 

"* Sensitivity of the technique and accuracy of the results must be 

adequate to meet the practical needs of post-Chernobyl followup. 

"* The results produced by the technique must be consistent with 

other (independent) dosimetric methods and internal standards. 

"* The technique must be reproducible at different times and in other 

laboratories. 

"* Performance of the technique must be high enough to meet practical 

demands. 

EPR Measurements, Spectra Analysis, and Dose 

Reconstruction 

A brief profile of the EPR measurement procedure is presented in table 

1. The use of the Mn:MgO spectrometric standard allows for objective 

control of stability of the system and ease in accounting for possible 

drifts and deviations. The standard is used for calibration of the 

instrument in terms of both sensitivity and g factor. The empty resonator 

signal is recorded daily and subtracted as instrument background 

noise in each series of measurements. 

5


Table 1. Brief profile of the EPR measurement procedure used at 

SCRM 

Feature 

Characteristics 

1. Instrument BRUKER ECS-106 

2. Laboratory irradiator 

13 7 Cs 

3. Buildup for secondary electrons + 

4. Number of additional irradiations 5 

5. Dose increment/cumulative dose AD = 174 for Dx < 500 mGy 

values for calibration curve 

AD = 348 for Dx -> 500 mGy 

6. Method of best fit of the Linear regression 

calibration curve 

7. Sample preparation + 

8. Parameters of EPR registration: 

Microwave frequency 

9.81 GHz 

Microwave power 

10 mW 

Center field 

348 mT 

Sweep width 

10.0 mT 

Modulation frequency 

100 kHz 

Modulation amplitude 

0.23 mT 

Conversion time 

164 ms 

Time constant 

328 ms 

Number of scans 15 

Measurement time 

60 and 45 minutes 

See item 5 

9. Processing of measured Subtraction of standard backspectra 

and detection of ground signal by selectively 

radiation-induced (RI) varying its intensity and fixing 

signals 

g factor and width 

10. Type of standard Spectrum of the mixture of 

background signal 

nonirradiated teeth from students 

18-25 years old 

11. Use of presumably nonirradiated 

samples as reference 

12. Approach to the error For doses less than 500 mGy, 

propagation 

the error is determined by uncontrolled 

impurities of enamel 

with maximum intensity 

equivalent to RI signal from 50 

to 80 mGy dose; for higher 

doses, error is determined by 

statistical error of RI signal 

determination. 

6

Development of a Routine High-Performance EPR-Dosimetric Technique 

The procedure for mathematical processing was based on investigation 

of the principal background signal variability. We discovered that 

background signals of different teeth differ only in terms of intensity, 

while g factors and width of lines for background signals are constant 

for all teeth. Therefore, we use a standard background signal, which 

was obtained from a mixture of several dozen nonirradiated teeth that 

were extracted from young people (18-25 years old), in order to minimize 

the natural background dose. 

The optimal procedure for subtraction of the principal background 

signal using the BRUKER ECS-106 or similar instrument is as follows: 

1. The standard background spectrum is shifted by the constant 

magnetic field relative to the sample spectrum, using an Mn:MgO 

marker until the g factors of both spectra coincide. 

2. The amplitude of the standard background spectrum is adjusted to 

coincide with the sample spectrum. 

3. The two spectra are subtracted while keeping the Mn:MgO signal 

constant. 

4. The resultant subtracted g factors of the maximum and minimum 

components of the suspected radiation-induced (RI) signal are compared 

to the positions of the relevant points determined for highdose 

signals. 

5. If points coincide, the intensity of the original radiation signal is 

measured and the value is used as the first experimental point in 

the individual calibration curve. If no coincidence occurs, no confident 

dose reconstruction is possible, and the tooth is considered 

to be exposed to a dose of less than 0.1 Gy. 

This rather conservative approach to spectrum interpretation ensures 

against the measurement of artifacts and misleading readouts. 

To account for individual radiosensitivity of teeth, an internal standard 

was used. Each sample was exposed to additional doses under 

controlled laboratory conditions. The 1 3 7 Cs secondary standard irradiator, 

calibrated in terms of absorbed dose in air using an 8 mm plastic 

screen for buildup of secondary electrons, was used for this purpose. 

The additional doses and the results of subsequent measurement 

produce the calibration curve of the individual tooth. The intersection 

of this curve with the abscissa corresponds to the amount of exposure. 

7


Sample Collection 

The following samples were collected for use in both practical dose 

reconstruction and interlaboratory cross-calibration: 

* 69 teeth from liquidators (40 were multiple teeth from 15 individuals) 

* 19 teeth from unexposed young people in Ukraine 

* 6 teeth from unexposed people in the United States 

Teeth from cleanup workers were collected in the course of routine 

dental treatment in the Liquidators' Clinic in Kiev, which is dedicated 

to the medical and dental care of liquidators. All teeth were extracted 

for clinical reasons only; none was collected solely for dosimetric 

purposes. The extraction procedure was performed by a skilled dental 

surgeon who had been appropriately instructed. Every sample was 

accompanied by a Tooth ID Form (see appendix), which contains ID 

information on the patient, a history of occupational and medical 

exposure, activities and length of stay in the restricted zone, and the 

officially recorded dose value. After extraction, the teeth were washed 

with water and dried at room temperature. Each sample was transferred 

to our laboratory in a clearly labeled, individual container. 

Sample Preparation 

A new, elaborate method of sample preparation was one of the most 

efficient innovations introduced into the EPR-dosimetric technique. 

Application of this procedure leads to the substantial reduction of 

background EPR signals that normally are superimposed on the radiation-induced 

signal in tooth enamel, making detection of doses below 

0.5 Gy difficult. 

When samples arrive at the laboratory, they are registered in the log 

and computer database. Then sample preparation begins. Since the 

properties of the original material vary, the degree of purification 

needed to obtain an optimal specimen may vary. Accordingly, purification 

is normally performed in several progressive steps. 

1. Removal of the tooth root. 

2. Splitting of the tooth into its inner and outer parts with a diamond 

saw (this step is taken in order to take possible medical x-ray 

exposures into account). 

8 

3. Fragmentation of the tooth into 1- to 2-mm particles.

Development of a Routine High- Performance EPR-Dosimetric Technique 

4. Chemical treatment of the tooth with KOH alkaline in an ultrasonic 

bath for 4 to 7 days to remove dentine and organic components of 

tooth enamel. 

5. Removal of the remainder of the dentine (especially in the tooth 

parts with the highest curvature) with a hard-alloy dental drill. 

6. Crushing of samples to 0.1 to 0.25 mm-sized grains. 

7. Additional purification of tooth enamel using a heavy liquid, so- 

3 

dium polytungstate, with a specific weight of 2.92-2.94 g/cm 

8. Several washings of samples with distilled water; the last washing 

takes several hours under ultrasonic processing. 

9. If needed, visual control of samples using a binocular microscope 

to remove nonenamel inclusions. 

The failure of any given procedure leads to application of further 

treatment. This algorithm is graphically presented in figure 2. 

STEP 1. 

" Crush bulk tooth into grains of 1-2 mm size. 

" Treat with NaOH under ultrasound and temperature 60 0C over the 24-hour period. 

P, jRecord EPR spectrum. 

NO Is sample ýYES 

STEP 2. 

6 Crush to smaller grains (0.1-0.25 

mm). 

. Treat with NaOH (as in STEP 1) NO"-YES 

over 8- to 12-hour period. Was STEP 2 

- Wash in distilled water under done? 

ultrasound for 2 hours. 

pure enough? 

STEP 3. 

* Separate heavy liquid (sodium NO YES 

polytungstate; liquids of 2.65 g/cm 3 Wa•s STP 3 

and 2.85 g/cm 3 densities are used 

for removal of light and heavy 

inclusions, respectively. 

done? 

STEP 4.i NO Was STEP 4 

. Manually remove nonenamel paramagnetic particles under visual control. done? 

Figure 2. Flow chart of the sample-preparation process 

9


150 

100 

.50 

S0- 

- -50 

A 

-100 

-150 ,..,,,,,,, ...... i .. .i .. •,a.... a.... i...... 

3425 3450 3475 3500 3525 

Gauss (G) 

150- 

100- 

- 50- 

S0- 

B 

-100 

- 5 . . . , . . , . . ..,. .. . . , ..... 

., ,... . . . . . 

3425 3450 3475 3500 3525 

Gauss (G) 

Figure 3. Effect of sample purification. EPR spectra (A) 

before and (B) after purification of the low-dose sample 

Figure 3 shows the EPR spectra of the same low-dose sample recorded 

before and after application of the purification procedure. Clearly, the 

spectrum was dramatically improved, making dose reconstruction 

with the sample possible. 

Another important issue is evaluating the contribution of medical x 

rays to the measured cumulative dose. Oversensitivity of tooth material 

to low-energy photons is well known, causing serious difficulties 

in determining Chernobyl-related doses. As much as a 7 to 1 difference 

in deposited dose in enamel versus that in soft tissue represents a 

threat to the utility of EPR as a dosimeter. The results of a study of this 

problem are presented elsewhere [4], giving a clue to a solution to this 

potentially severe problem. 

10

Task 2 

Quality Assurance Program 

Dr. Ed Haskell of the Division of Radiobiology, College of Medicine, 

University of Utah (UU), agreed to take part in the cross-calibration 

studies. These were performed in three stages, namely: 

Stage 1: Intercalibration using samples of tooth enamel with uniform 

properties that were exposed to known radiation doses under 

laboratory conditions 

Stage 2: Intercalibration using whole teeth exposed in vitro 

Stage 3: Intercomparison using liquidator's teeth accidentally exposed 

in vivo 

Each stage was designed to more closely approximate reality. Thus, the 

first stage dealt with rather ideal samples while the third intercomparison 

involved full-scale dose reconstruction using teeth specifically 

from exposed individuals. Interpretation of the results became increasingly 

difficult with each stage, as the number of uncertainty factors 

increased. However, the overall results appear to have engendered 

confidence in the adequacy of the dose assessments by EPR dosimetry 

with teeth. 

Stage 1. Intercalibration With Homogenized Samples 

For the first stage, the simplest intercalibration was performed with 

unrealistic but extremely uniform samples. This stage could be considered 

as a check for the reproducibility of the technique for practical 

dose reconstruction. The significant advantage of the intercomparison 

design was the possibility of objectively judging the results-no uncer- 

11


tainty factors could lead to a fuzzy interpretation of the dose 

determinations. 

Intercomparison Design 

The main purpose of this intercalibration 1 was to test the results 

produced by different techniques using samples with well-known and 

uniform properties, thus allowing for objective evaluation of results. 

Since sample preparation techniques in different laboratories vary 

significantly, only the minimum necessary treatment was applied to 

tooth samples. 

A large number of nonirradiated human teeth with minimum background 

doses were treated mechanically in order to extract tooth 

enamel in the form of 0.1- to 0.25-mm grains. These were mixed 

together to form homogeneous material. This material was then divided 

into 100 mg portions and forwarded for irradiation to the IAEA 

Laboratory in Siebersdorf, Austria. The samples were irradiated with 

dose levels of about 100, 250, 500, and 1000 mGy. Those receiving doses 

up to the 500 mGy level were irradiated with a 13 7 Cs source at a dose 

rate of 800 Gy/min. Those at the 1000 mGy level were exposed to a 6 0 Co 

source at a dose rate of 200 mGy/min. Sets of samples with five 

different dose levels (unknown to the participants) were shipped to 

SCRM and UU. The participants were invited to use their own EPR 

dosimetric routines (including chemical treatment of samples, EPR 

measurement, and interpretation of spectra) to determine the exposures 

received by the samples. 

Methods and Results 

The intercomparison revealed significant variations in the experimental 

techniques used for dose reconstruction with EPR of tooth enamel. 

The SCRM technique is shown in table 1. The major differences in the 

UU technique were as follows. 

" The background signal was not subtracted. The intensity of the RI 

signal was determined as the difference of intensity at the points 

corresponding to the g factor of the first maximum and first minimum 

of the RI signal. 

"* The samples were measured 10 times each, shaking the tube after 

every measurement. 

12 

This cross-calibration was performed as part of the First International 

Intercalibration of EPR dosimetry with teeth, which was sponsored in part 

by CEC contract COSU-CT93-0051.


Laboratory irradiation was performed for those samples that demonstrated 

the most significant variations. For these samples (8 of 

14), the value of the initial RI signal was determined as the intersection 

of the calibration curve with the ordinate axis. For the 

remainder, the initial intensity of the RI signal was assessed as the 

mean of 10 measurements. At this point, three samples with minimum 

RI signal values were considered as empirical zero. Doses of 

other samples were determined by subtracting the average intensity 

of three low-dose signals from the RI signal intensity. The 

results were divided by the value of the average calibration factor 

derived from the analyses of eight samples. 

* The residual signal, which corresponded to the average dose of 

three nonirradiated samples, was estimated to be 68 mGy. 

The results of the intercalibration are presented in figure 4 and table 

2. Clearly, the precision of dose determination depends greatly on the 

dose value and the technique used for reconstruction. The results 

obtained at SCRM demonstrated an excellent agreement with the 

preset dose values [5]. On the other hand, the UU technique showed 

variability over the whole range of doses, from 0 to 1000 mGy. 

These results were discussed in May 1995 duringthe 4th International 

Symposium on EPR Dosimetry and Applications. The constructive 

discussion of the peculiarities, possible advantages, and shortcomings 

of the techniques used at SCRM and UU led to the harmonization of 

certain approaches used in the two laboratories for subsequent stages 

of the cross-calibration. The UU approach was changed to conform 

more closely with some of the particular procedures under the SCRM 

approach. 

cc ~ 1200- A 1200 . B 

1000 1000- 

S800 800 

">, 600 600 

can) 00 400 

0 400- 

" " 200 200 

cc 0 0 

a) 

-2 0 0 0 ' -2 0 0 L ' 8 8 L a - 

0 200 400 600 800 1000 0 200 400 600 800 1000 

CHU..... CD, 

Nominal dose value, mGy 

Figure 4. Results of intercalibration with homogenized samples. Measured 

doses versus nominal dose values. (A) UU, (B) SCRM 

13


Table 2. Results of intercalibration using homogenized samples 

irradiated in vitro 

Participant Sample Measured dose Nominal dose 

number value, mGy value, mGy 

University 87 20.7±17 0 

of Utah 88 49.8±28 0 

90 -93±9 0 

21 132±15 100 

4 209±39 100 

6 237±26 100 

35 137±58 250 

43 138±68 250 

58 312±31 250 

75 413±67 500 

66 448±34 500 

56 479±49 500 

106 940±60 1000 

100 1128±21 1000 

SCRM 80 0±50 0 

81 10±50 0 

8 80±50 100 

16 140±50 100 

23 150±50 100 

36 240±50 250 

37 260±50 250 

42 290±50 250 

60 440±50 500 

68 460±50 500 

69 520±50 500 

76 970±100 1000 

108 970±100 1000 

Stage 2. Intercalibration With Whole Teeth 

Irradiated Under Laboratory Conditions 

14 

The main purpose of stage 2 was to address the possible influence of 

sample preparation procedures on the results obtained using different 

techniques.



The sample set consisted of six teeth, collected by a local dentist in the 

United States. The prior dose of x rays was unknown. Each tooth was 

split in two at the plane perpendicular to the jaw contour. One half of 

each tooth was irradiated using a 6 °Co source with a dose rate of about 

10 Gy/h. Measurements were considered to be dependent only upon 

the total dose, not the dose rate. Three dose levels between 100 and 500 

mGy were used; dose was unknown to the participants. One half of 

each tooth remained unirradiated in order to provide a postmeasurement 

check for detectable dose due to dental x rays as well as to provide 

a check should anomalies appear in the spectra or the results. 

The participants knew that the pairs of teeth numbered 1 and 4, 2 and 

5, and 3 and 6 were irradiated with equal doses in order to provide 

direct comparison of the results with the equivalent laboratory-added 

dose. 2 Teeth 4, 5, and 6 were shipped to the SCRM; the others remained 

at UU. Shipping was done by express mail specially labeled to avoid 

x-ray inspection and thus minimize transportation dose. Participants 

were instructed to apply their customary EPR dosimetric technique. 


The samples of tooth enamel were prepared as described in Task 1 

above. The tooth size was sufficient to perform separate analyses of the 

outer and inner parts in order to control, and if possible, account for 

unreported x-ray examinations. 

The EPR spectrum of tooth 6 differed significantly from the expected 

shape of the signal of irradiated enamel. Chemical processing with 

NaOH alkali (8 hours in an ultrasonic bath at 60 °C) was used to purify 

the enamel (see figure 2). Although the subsequent EPR analysis revealed 

a noticeable improvement, the shape of the signal was still 

distorted. The specimen was examined under the microscope, and one 

paramagnetic nonenamel particle was located and removed. That 

significantly improved the signal, making the sample appropriate for 

dose reconstruction. 

A special effort was made to control for possible x-ray exposure prior 

to intercalibration. Although the results of this examination do not 

allow for the reliable quantitative assessment of lifetime dose, there 

are strong indications of both a dose gradient (i.e., the surface of the 

tooth nearest the x-ray source receives a higher dose than the surface 

2 The total dose to be measured consisted of two principal components: 

unknown lifetime dose of the tooth donor and the known dose added at the 

laboratory. 

15


opposite the source, which is typical for dental x-ray examinations) and 

non-zero readouts in nonirradiated parts of the teeth. The findings 

(table 3) are not statistically significant and are given only as guidance 

to demonstrate general tendencies. Unfortunately, x-ray doses for all 

the teeth were below the threshold of reliable dose reconstruction with 

the EPR technique, and therefore these figures could not be used to 

correct the results of the intercalibration. 

Table 3. 

(in mGy) 

Dose assessments for different parts of teeth (SCRM) 

Tooth Inner part External part Inner part External part Mean dose of 

number (unexposed) (unexposed) (exposed) (exposed) exposed half 

T4 30 50 230 240 230±50 

T5 30 50 230 270 250±50 

T6 20 50 290 300 300±70 

UU measurements were taken at a microwave power of 2 mW (a 

method chosen after a series of tests as the most promising one to 

minimize noise and, therefore, uncertainty of the dose determination). 

An analysis of expected uncertainties was performed before analysis, 

and the number of spectra to be collected at each power was set at 42 

for the irradiated samples, 12 at the additive dose level of 1 Gy, and 6 

at the 10-Gy additive dose level. Samples were stored for a minimum 

of 12 hours at room temperature following each irradiation in a 6 0 Co 

irradiator at a dose rate of 10 Gy/h (5 mm of aluminum was used for 

electron buildup). Dose increments were 210, 435, 435, 435, 870 mGy. 

Both sets of dose determinations are presented in table 4. As expected, 

the results of the second stage intercalibration were not as clear-cut as 

those of stage 1, and they could not be interpreted definitely. Both 

laboratories demonstrated good agreement (within 17%) with nominal 

Table 4. 

Intercomparison of whole teeth exposed in vitro 

Laboratory- 

Group Sample added dose Laboratory Measured dose 

mGy 

mGy 

1 T1 171 UU 190±50 

T4 SCRM 230±50 

2 T2 256 UU 180±50 

T5 SCRM 250±50 

3 T3 200 UU 190±100 

T6 SCRM 300±70 

16


dose values, although the results produced by the SCRM technique 

tended to overestimate the doses. The latter may be explained by the 

contribution of the lifetime dose (particularly medical x-ray exposure) 

to the total dose to the tooth, which was determined by the dose 

reconstruction. Neither data on x-ray examination nor age of patients 

was available, making assessment of this component of the total dose 

impossible. Since the detected doses corresponding to the preintercalibration 

history of the teeth were found to be below the threshold 

of reliable dose reconstruction, the correction of the results was, 

unfortunately, impossible. 

Stage 3. Intercomparison of Teeth From Liquidators 

The third stage of cross-calibration was designed to test the capability 

of the two techniques to perform dose reconstruction using teeth 

exposed in vivo. Since the actual doses were unknown and, therefore, 

absolute validation of the results is impossible, the expected yield of 

this effort was a consistency check. 


The initial intention was to provide both laboratories with identical 

samples exposed in vivo. Three groups of samples from liquidators 

totaling 34 specimens were shipped to the United States: 

* 13 halves of large teeth (molars)-the remainders were retained at 

SCRM 

* 15 teeth from pairs extracted simultaneously from the same individual 

and therefore presumably having the same doses 

* 6 samples in the form of pieces of mechanically separated tooth 

enamel 

Due to the limited time that could be allocated by UU for examination 

of the samples, the number of dose reconstructions was reduced to 

five. The specimens selected (numbers X23, X24, X25, X26, and X28) 

were represented by granular samples of tooth enamel. For each 

sample, the initial separation of tooth enamel was performed at SCRM 

using a steel dental drill. After the removal of dentine, the pieces of 

enamel were collected and the whole sample was divided into parts of 

about 100 mg each for independent determination of dose by the 

participants. The characteristic size of enamel grains was about 500 

micrometers, although the dimensions of individual particles varied 

from hundreds of microns to several millimeters. 

17



SCRM used basically the same standard technique described above. 

The technique used at UU was somewhat modified. The parameters 

used for the second EPR intercomparison of the liquidators' teeth were 

as follows: 8-mW microwave power, 5-Gauss modulation amplitude, 

20-second conversion time, 35-Gauss sweep width, 168-ms time constant, 

and 105 gain. 

The dose reconstruction was done using the spectra from the samples, 

a baseline enamel sample with negligible dose, and an empty EPR 

tube. The spectrum of the tube was subtracted from each spectrum 

taken of all the samples and the baseline samples. This was done in 

proportion to the number of scans taken in each enamel spectrum, 

that is, an enamel spectrum composed of 12 sweeps was corrected by 

subtracting the spectrum of the empty tube, which itself was normalized 

to 12 sweeps. The spectra of the samples were then precisely 

normalized to the standard of 10 sweeps and 100 mg per spectrum by 

the normalization factor: 

(10 sweeps/# of sweeps taken) * (100 mg sample weight) 

This adjustment was not necessary for the baseline sample as it was 

precisely weighed before measurement. The spectrum of the baseline 

sample was then subtracted from all the spectra of the irradiated and 

unirradiated enamel samples. From these resulting background and 

baseline-free spectra, we did the dose reconstruction on the g-perpendicular 

signal extremes using standard least squares fitting and error 

propagation for the dose estimates and errors, respectively. The additive 

dose technique was employed using only one applied dose of 5 or 

10 Gy (10 Gy if the sample mass was less than 35 mg). The number of 

spectra taken at each dose was 20 to 25 for the zero dose and 12 for the 

one applied dose (25 if the sample mass was less than 35 mg). 

18 

In this intercomparison, an additional check of the purity of the 

samples was performed using the EPR spectra recorded before additional 

irradiation and the spectrum of a milk tooth as the standard of 

the background signal. Two (X24 and X25) of five samples had satisfactory 

purity. Three others were subjected to treatment with an NaOH 

solution. After the chemical treatment, the shape of the signal had 

improved. Two of the samples (X26 and X28) were considered to be 

purified completely. Although the purity of the third one (X23) had 

improved, it still had some distortions in the spectrum. Because of 

significant loss of mass (it had decreased from 70 to 33 mg), we decided 

to refrain from further purification and proceed with EPR 

measurements.


This third stage of cross-calibration had relatively poor results (table 

5). The doses, determined in different laboratories, coincided within 

declared uncertainty ranges for only two individuals out of five. The 

results from the two laboratories differed significantly (up to 60%) for 

some of the samples. The results of this intercomparison are discussed 

elsewhere [6]. 

At the present stage of the intercomparison, it is impossible to determine 

the major reasons for these discrepancies. Adequate interpretation 

of the results requires additional investigation and, possibly, 

conducting the intercomparison with a partially modified design. One 

possibility could be to use teeth from individuals whose doses have 

been assessed by independent methods of retrospective dosimetry 

(such as a FISH test or analytical dose reconstruction). 

Table 5. Results of intercomparison with teeth of liquidators 

exposed in vivo 

Sample SCRM, Gy UU, Gy 

X23 0.36±0.05 0.67±0.10 

X24 1.42±0.14 1.60±0.24 

X25 1.08±0.11 1.56±0.23 

X26 1.50±0.15 1.56±0.23 

X28 0.48±0.05 1.18±0.18 

Discussion 

The cross-calibration performed within Task 2 of the project was the 

first international, full-scale effort to harmonize EPR-dosimetric techniques 

developed in Ukrainian and US laboratories as well as to 

perform quality assurance of this method. The three stages of cross-calibration, 

for the most part, covered all degrees of complexity and 

adequacy of approaches to dose reconstruction. Generally positive, the 

results of the cross-calibration have proven the applicability of EPR 

dosimetry to practical reconstruction of individual doses. 

The reproducibility of the results of the different versions of EPR 

technique that were designed and used on different continents is, at 

worst, within the 60% standard deviation interval. Clearly, even with 

this conservative and potentially improvable uncertainty interval, EPR 

dosimetry could produce more accurate dose assessments than any 

other method of retrospective dose reconstruction to be used in a 

post-Chernobyl epidemiological followup. Moreover, stage 2 of the 

cross-calibration experimentally demonstrated that lifetime diagnostic 

x-ray examinations may lead to overestimation of doses within only 

19


30% limits. This important point, which needs additional investigation, 

could resolve positively the greatest concern presently associated with 

the use of EPR dosimetry for reconstruction of individual doses among 

the liquidators. 

20

Task 3 

Test of Practical Dose Determination 

A system for retrospective reconstruction of doses received by the 

Chernobyl liquidators was tested in Task 3. 

Dose Reconstruction 

The teeth from liquidators were collected in the course of dental 

surgical practice in the Kiev central liquidators' polyclinics. Extracted 

teeth were accompanied by special ID forms (see appendix) reflecting 

the personal data necessary for tracing the individual, information 

about occupational contacts with ionizing radiation, lifetime medical 

x-ray examination of the head, data on location of the extracted tooth, 

and the diagnosis leading to extraction. After extraction, the teeth were 

preserved in formalin in small bottles. Periodically (approximately 

once a month), the teeth were transported to the Laboratory of External 

Exposure Dosimetry for storage, processing, and determination of 

radiation doses. 

Upon arrival, all teeth were subjected to preprocessing, including 

rinsing in distilled water and drying at 80 'C. The tooth root was 

separated and, if necessary, residuals of soft tissues and damaged areas 

of teeth were removed. Then, teeth were placed in intermediate storage 

under room conditions. 

The dose determination cycle began with chemical treatment as discussed 

in Task 1 and illustrated in figure 2. The samples of pure tooth 

enamel were subjected to measurements, including recording of EPRspectra 

(with parameters as indicated in table 1) and laboratory irradiation 

with preset doses. Individual calibration curves were plotted for 

all measured samples, and doses were determined accounting for 

individual radiosensitivity of enamel. It was found that calibration 

21


- A curves may not always 

200 -linear regression 

2polynome of the 

be fitted by linear regression. 

In some cases, 

100 f dose-response curves 

were superlinear (fig- 

0 

050 0 

50 

50 100 150 200 250 

ure5b)orsublinear(figure 

5c). Although the 

S200 B 

Dose, cGy 

total fraction of teeth 

with nonlinear dose- 

150 response curves was 

._ rather small (about 5%), 

S100 

this phenomenon needs 

to be accounted for in 

"order to avoid signifi- 

0 * cant under- or overesti- 

-40 0 40 80 120 160 mation of dose values. 

Dose, cGy 

The sources of this ef- 

300 - C fect need to be studied 

200r 

100 

and localized. 

The results of dose reconstruction 

of 146 

0 0 50 * ' 200 teeth from 135 liquida- 

-25 0 50 100 150 2'00 tors are shown in table 

Dose, cGy 

6. The cumulative doses 

and dose values are not 

Figure 5. Different types of dose-response corrected for lifetime 

(calibration) curves for tooth enamel sam- exposure. Age of the paples. 

(A) linear dose response (95% of sam- tient and type of tooth 

ples), (B) superlinear dose response (4% of give a clue to the 

samples), (C) sublinear dose response (1% amount of the dose due 

of samples). 

to natural background. 

The type of tooth is also 

important for the possible 

contribution of ultraviolet 

(UV) light to 

the generation of paramagnetic centers, which may be more pronounced 

for the front teeth (classes A and B). 

22 

The frequency distribution of individual doses measured with EPR 

dosimetry is presented in figure 6. It may be seen that the shape of 

the distribution is close to lognormal-mean dose is 287 mGy, 

geometric mean is 205 mGy. Median dose value is 200 mGy. The 

reconstructed dose values range from 30 to 2220 mGy. The individual 

with the highest dose is a policeman who performed his guard 

mission outdoors during the first days after the accident. On some 

occasions, several teeth used in the investigation came from the 

same individual.


Table 6. Individual dose values reconstructed in course of EPR 

dosimetric exercise 

Sample Age at Beginning of 

number/ time of tooth X-ray cleanup work Type of Total dose 

ID code extraction examinations (year) tooth* (cGy) 

1 2 3 4 5 6 

177/13862 60 - 86 A 39 

178/16042 53 - 86 C 16 

180/17568 25 - 86 C 5.5 

182/2069 56 + 86 C 13 

183/59 45 - 86 C 24 

185/10709 43 + 86 C 12 

187/15961 61 - 86 C 19 

190/17871 64 - 86 C 14 

C 29 

192/17099 60 - 86 C 25 

194/20457 45 - 86 C 10 

195/13337 58 - 86 A 34 

198/12006 55 + 86 B 15 

199/20491 44 + 86 C 11 

200/7107 46 - 86 C 10 

201/4385 43 - 86 C 17 

202/1304 56 + 86 C 28 

204/1801 47 + 86 C 8 

205/ 40 + 86 C 13 

208/9731 56 - 86 C 12 

301/15518 60 + 86 C 19 

302/16012 58 - 86 C 28 

303/13827 39 86 U 6 

304/13473 55 + 86 C 15 

305/7826 39 86 A 31 

307/4513 48 - 86 C 8 

278/14877 40 - 86 C 4 

28/10058 55 + 86 C 12 

281/14571 57 - 86 C 11 

284/17631 57 - 86 A 31 

A 16 

287/19245 57 - 86 C 18 

298/8822 66 - 86 B 34 

33/483 57 - 86 A 46 

8/4416 44 + 86 U 18 

A 21 

4/13930 53 + 86 C 45 

C 67 

39/15579 62 + 86 B 78 

16/4689 52 + 86 C 38 

9/7821 55 - 86 C 53 

*A - incisors and canines, B - premolars, C - molars, U - unknown 

23


24 

Table 6. Continued 

1 2 3 4 5 6 

10/11338 53 + 86 A 40 

15/13919 59 - 86 C 22 

29/16047 53 + 86 C 19 

24/6902 55 - 86 A 64 

53/1926 54 + 86 C 18 

46/7125 46 - 87 C 13 

97/10075 56 - 86 A 55 

98/199 52 - 86 A 59 

43/17164 55 + 87 C 120 

67/16344 38 + 86 A 66 

138/17320 41 + 87 C 9 

151/3598 64 + 86 C 12 

132/3915 66 - 86 A 23 

64/3978 40 + 86 C 27 

84/2743 54 - 87 C 20 

181/9287 63 + 86 B 14 

6C 53 - 86 C 142 

19C 62 - 86 C 25 

105/11735 44 - 86 C 16 

106/9138 44 - 86 C 20 

108/573 62 + 86 C 30 

C 23 

7/1670 62 + 86 A 30 

71/763 64 - 86 C 18 

72/18249 42 - 86 C 3.5 

78/18414 30 + 86 C 5 

C 8 

79/15694 79 - 86 C 13 

82/7452 49 - 86 C 5 

83/18187 49 + 86 C 13 

86/17174 50 - 87 C 12 

87/15171 55 - 86 C 6 

35/3602 28 + 86 C 20 

41/4068 37 + 86 C 18 

42/3304 55 - 86 C 7 

44/8329 47 - 86 B 18 

61/8329 B 28 

45/8203 60 + 86 B 20 

49/9037 65 + 86 C 19 

56/9737 56 - 86 A 29 

59/7727 45 - 86 C 8 

50/7727 C 6 

65/4299 50 + 86 C 9 

68/10373 50 + 87 C 8 

"A- incisors and canines, B - premolars, C - molars



1 2 3 4 5 6 

69/17092 60 - 86 C 8 

109/17545 45 - 87 C 21 

113/8025 29 - 87 C 6 

123/13820 52 + 86 A 50 

125/10396 55 + 86 C 14 

129/13494 57 + 86 A 24 

130/15240 64 - 86 A 60 

142/5009 68 - 86 A 71 

145/4768 56 - 86 A 35 

148/16267 49 + 87 C 67 

153/13349 60 + 86 C 96 

209/1408 40 - 86 U 47 

212/8456 46 - 86 A 39 

216/8544 45 - 86 C 25 

217/13506 43 - C 13 

218/13953 59 86 U 20 

219/15067 64 + 86 U 13 

223/3467 57 - 86 C 23 

224/18355 71 - 87 C 65 

225/17113 56 - 87 C 14 

227/18318 55 - 86 C 10 

228/8369 53 + 86 C 14 

249/8369 54 C 39 

230/2721 57 - 86 C 12 

231/13381 63 - 86 C 195 

234/16478 63 - 86 C 16 

235/8152 56 + 86 C 23 

236/4325 47 - 86 A 27 

238/9123 62 - 86 A 70 

247/14939 43 - 86 C 18 

250/6863 55 + 86 C 21 

C 15 

251/7583 45 + 86 A 23 

263/8939 64 - 86 A 35 

300/17933 66 + 86 C 40 

81/10298 33 - 89 C 8 

308/19933 51 - 87 C 7 

309/8120 56 - 88 C 13 

312/2043 61 + 86 C 23 

313/16137 68 - 87 C 26 

314/4571 58 + 86 U 64 

317/21873 55 - 86 C 50 

320/7004 48 86 C 6 

321/20577 44 - 87 C 7 

*A- incisors and canines, B - premolars, C - molars, U - unknown 

25



1 2 3 4 5 6 

322/15777 38 - 86 A 36 

B 31 

325/23887 42 + 86 C 8 

326/15707 64 - 86 A 42 

330/17012 61 + 86 C 12 

331/8719 33 + 86 C 10 

332/14533 59 - 86 C 35 

334/3713 57 + 86 C 12 

335/15336 51 - 86 B 3 

337/4722 68 - 86 A 21 

340/13695 52 - 86 C 32 

C 20 

2C 86 U 57 

17C 40 86 U 30 

18C 35 86 U 222 

20C 41 86 U 40 

21C 86 U 30 

22C 37 86 U 15 

*A - incisors and canines, B - premolars, C - molars, U - unknown 

The doses generally depend on the amount of time spent working in 

the 30-km zone. As may be seen from table 6, most of the liquidators 

involved in the current dose reconstruction effort began their work in 

1986. Median dose of this group is 211 mGy, while doses to liquidators 

of 1987 and later years are lower-164 mGy. Maximum doses to liquidators 

from 1986 and 1987 were 2220 mGy and 1200 mGy, respectively. 

This observation is in good agreement with the fact that the most 

dose-intensive activities were performed during the first months after 

50 

_T 40 

0- 

0 

a) 

0 

S20 

E 

z 10 

0'= 6= 1 W I I 

10 30 50 70 90 110 130 150 170 190 210 230 

Dose, cGy 

Figure 6. Frequency distribution of individual doses to liquidators 

determined by EPR dosimetry of teeth. 

26


the accident, when dose rate levels were much higher and most of the 

cleanup work took place. 

System Development 

Practical demand dictates a need for reconstruction of radiation doses 

to the large groups of liquidators included in cohorts studied for 

epidemiological followup. The possibility of long-term storage of tooth 

samples together with increasing performance of the EPR-dosimetric 

technique make this task feasible. 

However, since the teeth used for dose reconstruction are extracted for 

medical reasons only, sampling is a random and relatively infrequent 

event. Besides, the process of natural tooth loss is an important factor, 

reducing the available sampling population over time. 

Therefore, a systematic approach to dose reconstruction from teeth, 

including sample acquisition, is required. For longitudinal epidemiological 

followup of an exposed population, the problem of availability 

of samples may be solved by organizing a widespread network for 

acquisition of teeth extracted from the members of the studied cohort. 

This network should be based on centers with a high density of 

liquidators and other heavily exposed populations. Since the productivity 

of EPR dosimetry is limited and not yet sufficient to process all 

the potential influx of samples in real time, a central bank of bioprobes 

should be established for acquisition, storage, processing, and retrieval 

of tooth material. Potentially, every participant of this studied cohort 

sooner or later would be covered by this effort, yielding tooth samples 

to the bioprobe bank. Teeth from those individuals who were included 

in the study cohort and have died could be received in the course of 

autopsy. Dose values, reconstructed by means of EPR, could be entered 

in the personal dosimetric file of the individual for access by 

radioepidemiologists. 

Development of such an infrastructure for dose reconstruction is 

underway now in Ukraine. The acquisition network (figure 7) would 

be based on special liquidators' hospitals in seven regional centers, 

covering about 45% of the heavily exposed cleanup workers. The samples, 

along with ID forms, would be transferred to the central bioprobe 

bank for long-term storage and subsequent processing. 

The role of the central bioprobe bank is to coordinate activities in 

acquisition of teeth, EPR dosimetry, and data management on a national 

scale. Results of the ongoing EPR dose reconstruction will be 

forwarded to the National Registry and sent in parallel to the local 

health care bodies in order to provide feedback to individuals whose 

teeth had been submitted for examination. Access to the individual 

27


dose records will also be provided to the researchers involved in 

post-Chernobyl followup studies. 

Chernobyl Ministry 

of Ukraine 

lDnipropetrovsk i Donetsk Kharkiv I Kir°vgr 

P •, va Health, •Ministryl F Zo rtzhiya 

, IoUraino of Uov L• z 1] 

Kiev 

Central bioprobe bank 

Consumers of 

Bioprobe section Database section dosimetric information 

"* conservation * registration of incoming probes * epidemiology 

"• storage * tracing of the samples 1..- • medicine 

"* sample preparation * storage of reconstructed doses * social service 

"* access for analysis e retrieval and output of doses o exposed individuals 

t 

t 

methodological EPR-dosimetry lab 1 1EPR-dosimetry lab 2 EPR-dosimetry lab3 

center H 

Figure 7. Infrastructure of the system for retrospective reconstruction 

of doses received by the Chernobyl liquidators 

28

Discussion 

Although performance and capability were tested in a series of crosscalibrations 

and routine dose reconstructions, several issues of key 

importance need to be resolved prior to extensive use of EPR dosimetry 

with teeth for followup studies. These yet unresolved problems may 

threaten the utility of EPR dosimetry. Some of these effects have been 

known for a long time; others were discovered recently. Among these 

are the well-known effect of enhanced sensitivity to low-energy photons 

and the recently reported generation of paramagnetic centers by 

UV light [7]. Nonlinearity of dose-response curves in the dose range 

below 1 Gy was observed by the authors of this report only during the 

dose reconstruction exercise and is yet unpublished. 

Effects of Medical X Rays 

Irradiation of tooth enamel with low-energy photons may lead to 

substantial (up to seven times) overestimation of the tissue-absorbed 

dose. This effect has pronounced energy dependence, with the highest 

oversensitivity at 60 keV. The signals from paramagnetic centers produced 

by high-energy (accidental) and low-energy (medical x-ray) photons 

are identical, making discrimination of these signals by means of 

EPR impossible. As a result, a dose measured by EPR is the sum of an 

accidental component (dose of interest) and a component due to 

medical exposure. The degree of significance of the latter depends on 

the relative value, which is a function of incidence energy, dose per 

examination, and number of examinations. This means that the type 

of x-ray apparatus used in the dental practice is very important, determining, 

after all, the degree of significance of the x-ray component. 

In order to clarify this issue, it is necessary to conduct a systematic 

investigation of the effects connected with x-ray exposure. This inves- 

29


tigation should include both experimental and theoretical evaluation 

of dose responses prompted by different types of x-ray examination, 

including different geometry, x-ray apparatus, and dose per examination. 

The issue of doses deposited on neighboring and opposite teeth 

should be studied also. The contribution to the total tooth dose from 

different types of x-ray examination (e.g., gamma-tomography, panoramic 

diagnostics of the mandible, skull and sinus films) and radiotherapy 

is still unknown and requires special investigation. This work 

will demand the use of both mathematical and physical phantoms for 

the simulation of realistic situations. 

Since x-ray practices are different in Ukraine and the United States, 

special attention in this research should be paid to comparison of these 

two cases and the development of approaches to the solution of this 

problem. Intercomparisons using teeth exposed to x rays or exposed 

to mixed fields would be useful as well. Such studies should either 

conclude that the x-ray contribution to the tooth dose is insignificant 

(and establish the limits of the application of this assumption) or else 

recommend how to mitigate or account for the effect of x-ray 

examination. 

Analysis of different teeth from the same person may be useful for 

understanding of the effects of x-ray examination on different types of 

teeth and other factors. Among the 135 individuals studied in the 

present research, multiple teeth were available from 11 persons. In 

nine cases, teeth were extracted simultaneously, with presumably 

equal accidental doses. In two cases (individuals 190/17871 and 

284/17631), the doses determined using different teeth showed a 

discrepancy above the 40% standard deviation accuracy granted by the 

technique. Since the teeth in both cases were of the same type, and 

x-ray examination was not reported on the ID form, this phenomenon 

could not be explained by the contribution of x-ray exposure or variations 

in the type of tooth. Some as yet unknown effects may be 

responsible for such deviation. Unfortunately, the limited scope of 

research and similarities in lifetime exposure and type of teeth give 

little material for analysis. However, so far, the largest deviation of 

doses determined for similar teeth was 52%, which is not a particularly 

large error, considering errors typical for other methods of retrospective 

dosimetry. 

Effects of UV Light 

30 

Another phenomenon which may affect the reliability of dose reconstruction 

with tooth enamel is the generation of paramagnetic centers 

by UV light. Information about the role and the qualitative and quantitative 

characteristics of this effect is quite contradictory. This effect 

was first discovered and reported by Ivannikov et al. [7] in 1995. The

Discussion 

series of experiments conducted worldwide to study this effect brought 

no clarification. 

According to existing information and our own data, the centers 

generated in tooth enamel have position and shape very similar to 

those of radiation-induced centers. That means that discrimination of 

UV- and radiation-induced signals by spectrometric means may be 

difficult. It is expected that UV irradiation effects are most pronounced 

for front teeth; such factors as time spent outdoors and elevation of the 

living area above sea level may also influence the degree of this effect. 

At present, the problem of UV irradiation needs to be approached in a 

systematic way; this phenomenon must be studied from the point of 

view of its physical, spectrometric, and kinetic (half-life) properties. 

Processes of generation of paramagnetic centers as a function of wavelength 

and intensity of UV light and decay of these centers should be 

investigated in order to obtain a clear view of this effect. 

Study of spectrometric properties (e.g., saturation of the signals) may 

yield an approach to discrimination of UV- and radiation-induced 

signals by means of EPR technique. Investigation of depth profiles of 

UV-generated signals in teeth for different energies of UV photons and 

the UV component of daylight should help explain attenuation of UV 

light in enamel and could be used for target etching of exposed 

fractions of tooth enamel. Recommendations concerning accounting 

for and mitigating this effect should be issued as a final point of this 

research. 

Nonlinearity of Dose Response Curves 

Nonlinearity of dose response curves in the dose range below 1 Gy was 

observed in some teeth in the course of the dose reconstruction exercise 

in this project. Before, saturation of the dose-response curve was 

observed only at doses above 10 Gy; below this range, the curve was 

considered to be linear, and this property is widely used for extrapolation 

of calibration curves in the low-dose regions. Moreover, the techniques 

based on the utility of a single calibration factor (without 

additive dose) critically depend on linearity of the dose-response 

function. 

Nonlinearity of dose response curves may have a significant influence 

on the results of dose reconstruction. Not accounting for nonlinearity 

of calibration curves leads to substantial under- or overestimation of 

individual doses (as illustrated in figure 5). Advanced study of this 

effect, investigation of factors having impact on the dose-response 

curve, and the development of methods for extrapolating additive-dose 

curves are necessary for accurate and reliable retrospective dosimetry 

31


using teeth as a natural dosimeter. Since this effect takes place in only 

about 5% of cases, the scope of dose reconstruction should be large 

enough to provide consistent and statistically significant conclusions. 

32

Summary 

During the 14-month period covered by this contract, extensive research 

and technological developments were performed at SCRM AMS 

Ukraine in close collaboration with the University of Utah, USA. As a 

result of this effort, EPR dosimetry with teeth was brought to the level 

of a semiroutine technique for evaluation of doses received by individuals 

heavily exposed after the Chernobyl accident. 

Special attention was paid to quality assurance for this high-technology 

method in order to provide accurate and reliable individual dose 

assessments. The quality assurance program included several international 

cross-calibrations using a variety of specimens, from pulverized 

tooth enamel in the beginning to whole teeth from liquidators exposed 

in vivo during the final phase of intercomparison. 

Since the limited availability of samples from the individuals of interest 

is one of the important bottlenecks of EPR dosimetry now, a 

complete system for the reconstruction of doses to liquidators must 

include a means for acquiring the samples. Therefore, an organization 

pattern for acquisition of teeth extracted by medical prescription from 

the Chernobyl liquidators was presented. This infrastructure is being 

implemented in Ukraine now. 

The semiroutine technique developed and adopted over the period of 

consideration was used for retrospective dosimetry of a considerable 

group of liquidators. In total, doses were reconstructed for 135 individuals 

who took part in the Chernobyl cleanup in 1986-87. The cohort of 

liquidators studied was assembled randomly in the course of dental 

surgery in the Kiev central liquidators' polyclinic. 

Analysis of the data obtained revealed that the mean dose of this group 

is 287 mGy, ranging to the highest value of 2220 mGy. This is signifi- 

33


cantly higher than the officially reported mean dose of 110 mGy. 

Therefore, the widely accepted opinion that the official records are of 

low quality and underestimate the actual doses was supported by this 

first retrospective dosimetry effort involving an appreciable number 

of subjects. The fact that doses reconstructed instrumentally are much 

higher than those officially recorded gives additional justification for 

the investment in development and performance of retrospective 

dosimetry, particularly EPR. 

However, recent research and developments in the field of EPR dosimetry 

make obvious a need for further investigations. From the pragmatic 

point of view, these investigations should be conducted along 

the following lines: 

"* Investigation and development of approaches to account for EPR 

signals induced by lifetime medical x-ray exposure 

"* Comprehensive study of the effects in tooth enamel caused by UV 

light 

"* Investigation of the factors causing nonlinearity of the dose-response 

function in the dose range below 1 Gy and development of 

approaches to account for this effect in dose determination 

" Cross-validation of EPR dosimetry with independent methods of 

retrospective dosimetry; this may be achieved by parallel application 

of different methods (e.g., EPR, FISH, and analytical) to the 

same objects 

"* Methodological research aimed at improving the technological 

capabilities of EPR dosimetry and enhancing the productivity of 

the technique. 

Completion and success of the outlined efforts will bring EPR dosimetry 

from a quite exotic methodology to an ordinary dosimetric routine 

like gamma-spectroscopy and alpha counting. 

34

References 

1. Romanyukha AA, Ignatiev EA, Degteva MO, Kozheurov VP, Wieser 

A, Jacob P (1996) Radiation doses from Ural Region. Scientific 

Correspondence. Nature 381:199-200 

2. Chumak V, Sholom S, Likhtarev 1 (1995) Semi-routine ESR-dosimetry 

technique currently used in Ukraine. Presented at the 4th 

International Symposium on ESR Dosimetry and Application, Munich, 

Germany, May 15-19, 1995 

3. Chumak V, Sholom S, Pasalskaya L, Pavlenko Yu (1995) Ukrainian 

version of the EPR-dosimetric technique: An approach to the routine 

dose reconstruction. Second Workshop on Dose Reconstruction, 

Bad-Honnef, Germany, November 20-22, 1995 

4. Sholom S, ChumakV, Pavlenko Yu (1995) An account of diagnostic 

x-ray exposure in the problem of retrospective ESR dosimetry. 

Presented at the 4th International Symposium on ESR Dosimetry 

and Application, Munich, Germany, May 15-19, 1995. 

5. Chumak V, Baran N, Bugai A, Dubovsky S, Fedosov I, Finin V, 

Haskell E, Hayes R, Ivannikov A, Kenner G, Kirilov V, Khamidova 

L, Kolesnik S, Liidja G, Lippmmaa E, Maksimenko V, Meijer E, 

Pasalskaya L, Past J, Puskar J, Sholom S, Skvortzov V, Vaher U, 

Wieser A (1995) The first international intercomparison of EPR-dosimetry 

with teeth: First results. Presented at the 4th International 

Symposium on ESR Dosimetry and Application, Munich, Germany, 

May 15-19, 1995 

6. Haskell EH, Kenner GH, Hayes RB, Sholom S, ChumakV (1995) An 

EPR intercomparison using teeth irradiated prior to crushing. Sec- 

35


ond Workshop on Dose Reconstruction, Bad-Honnef, Germany, 

November 20-22, 1995 

7. Ivannikov A, Skvortzov V, Khamidova L, Eichhoff U (1995) Development 

of tooth enamel EPR spectroscopy method for individual 

dosimetry. Presented at the 4th International Symposium on ESR 

Dosimetry and Application, Munich, Germany, May 15-19, 1995 

36

Appendix 

Identification Form for Tooth Sampling 

1. Complete affiliation of the hospital which 

performed extraction 

2. ID number 3. Date of extraction ___ 

N General information Fragment 1 

1 Family name 

2 First name 

3 Second name 

4 Sex( male - 1, female - 2) 

5 Date of birth 

6 Liquidators pass (series and number) 

7 Year of work in Chernobyl 

8 Dose value, officially recorded (if available) 

9 Date of evacuation from the 30-km zone 

10 From what settlement 

N Postal address at present time Fragment 2 

1 ZIP code 

2 Region 

3 District 

4 Town 

5 Street 

6 House 

7 Building 

8 Appartment 

37


4. Places of stay since the accident (region, district, settlement) 

(1986 in all details, afterwards - reflect locations with period of stay more than 3 months). 

Year Settlement Period of stay 

Arrival Departure 

5. Professional contact with radiation (including military service) 

6. Information about x-ray examinations of skull, jaws, teeth (dates, type, approximate number during 

life span): 

7. General deseases affecting solid tissues of tooth 

8. Location of the tooth and reason of extraction: 

8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 

38

Appendix 

9. Affiliation during the Chernobyl recovery activities 

10. Notes 

11. Name of physician who extracted the tooth 

39



U.S. ARMY RESEARCH INSTITUTE OF ENVIRONMENTAL 

MEDICINE 

ARMED FORCES RADIOBIOLOGY RESEARCH INSTITUTE ATTN: DIRECTOR OF RESEARCH 

ATTN: PUBLICATIONS BRANCH 

ATTN: LIBRARY U.S. ARMY RESEARCH LABORATORY 


ARMY/AIR FORCE JOINT MEDICAL LIBRARY 

ATTN: DASG-AAFJML WALTER REED ARMY INSTITUTE OF RESEARCH 

ATTN: DIVISION OF EXPERIMENTAL THERAPEUTICS 


ATTN: AE 

ATTN: HA(IA) DEPARTMENT OF THE NAVY 

DEFENSE SPECIAL WEAPONS AGENCY 

BUREAU OF MEDICINE & SURGERY 

ATTN: TITL ATTN: CHIEF 

ATTN: DDIR 

ATTN: RAEM NAVAL AEROSPACE MEDICAL RESEARCH LABORATORY 

ATTN: MID ATTN: COMMANDING OFFICER 


NAVAL MEDICAL RESEARCH AND DEVELOPMENT COMMAND 

ATTN: ACQUISITION ATTN: CODE 42 

ATTN: ADMINISTRATOR 

NAVAL MEDICAL RESEARCH INSTITUTE 

FIELD COMMAND DEFENSE SPECIAL WEAPONS AGENCY ATTN: LIBRARY 

ATTN: DASIAC 

ATTN: FCIEO NAVAL RESEARCH LABORATORY 

ATTN: LIBRARY 


ATTN: DIRECTOR OFFICE OF NAVAL RESEARCH 

ATTN: BIOLOGICAL & BIOMEDICAL S&T 

LAWRENCE LIVERMORE NATIONAL LABORATORY 

ATTN: LIBRARY 


UNDER SECRETARY OF DEFENSE (ACQUISITION) 

ATTN: OUSD(A)/R&E BROOKS AIR FORCE BASE 

ATTN: AL/OEBZ 

UNIFORMED SERVICES UNIVERSITY OF THE HEALTH SCIENCES ATTN: OEHL/RZ 

ATTN: LIBRARY ATTN: USAFSAM/RZB 

OFFICE OF AEROSPACE STUDIES 

DEPARTMENT OF THE ARMY ATTN: OAS/XRS 


OFFICE OF THE SURGEON GENERAL 

ATTN: SLCSM-SE ATTN: HQ AFMOA/SGPT 

A-TN: HQ USAF/SGES 

OFFICE OF THE SURGEON GENERAL 

ATTN: MEDDH-N U.S. AIR FORCE ACADEMY 

ATTN: HQ USAFA/DFBL 

U.S. ARMY AEROMEDICAL RESEARCH LABORATORY 

ATTN: SCIENCE SUPPORT CENTER U.S. AIR FORCE OFFICE OF SCIENTIFIC RESEARCH 

ATTN: DIRECTOR OF CHEMISTRY & LIFE SCIENCES 

U.S. ARMY CHEMICAL RESEARCH, DEVELOPMENT, & 

ENGINEERING CENTER 

ATTN: SMCCR-RST OTHER FEDERAL GOVERNMENT 

U.S. ARMY INSTITUTE OF SURGICAL RESEARCH 

ARGONNE NATIONAL LABORATORY 

ATTN: COMMANDER ATTN: ACQUISITIONS 

U.S. ARMY MEDICAL DEPARTMENT CENTER AND SCHOOL 

BROOKHAVEN NATIONAL LABORATORY 

ATTN: MCCS-FCM ATTN: RESEARCH LIBRARY, REPORTS SECTION 

U.S. ARMY MEDICAL RESEARCH AND MATERIEL COMMAND 

CENTER FOR DEVICES AND RADIOLOGICAL HEALTH 

ATTN: COMMANDER ATTN: DIRECTOR 

U.S. ARMY MEDICAL RESEARCH INSTITUTE OF CHEMICAL 

GOVERNMENT PRINTING OFFICE 

DEFENSE ATTN: DEPOSITORY ADMINISTRATION BRANCH 

ATTN: MCMR-UV-R ATTN: CONSIGNED BRANCH 

U.S. ARMY NUCLEAR AND CHEMICAL AGENCY 

LIBRARY OF CONGRESS 

ATTN: MONA-NU ATTN: UNIT X


BRITISH LIBRARY 

ATTN: REPORT LIBRARY ATTN: ACQUISITIONS UNIT 

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION 

CENTRE DE RECHERCHES DU SERVICE DE SANTE DES ARMEES 

ATTN: RADLAB ATTN: DIRECTOR 

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION 

FEDERAL ARMED FORCES DEFENSE SCIENCE AGENCY FOR NBC 

GODDARD SPACE FLIGHT CENTER 

PROTECTION 

ATTN: LIBRARY ATTN: LIBRARY 

NATIONAL CANCER INSTITUTE 

FOA NBC DEFENCE 

ATTN: RADIATION RESEARCH PROGRAM ATTN: LIBRARY 


INHALATION TOXICOLOGY RESEARCH INSTITUTE 

ATTN: LIBRARY TECHNICAL SERVICES ATTN: LIBRARY 

U.S. DEPARTMENT OF ENERGY 

INSTITUTE OF NUCLEAR MEDICINE AND ALLIED SCIENCES 

ATTN: LIBRARY ATTN: DIRECTOR 

U.S. EMBASSY, MOSCOW 

INSTITUTE OF RADIOBIOLOGY, ARMED FORCES 

ATTN: DEPARTMENT OF ENVIRONMENT, HEALTH, MEDICAL ACADEMY 

OCEANS, AND FISHERIES ATTN: DIRECTOR 

U.S. FOOD AND DRUG ADMINISTRATION 

OAK RIDGE ASSOCIATED UNIVERSITIES 

ATTN: WINCHESTER ENGINEERING AND ATTN: MEDICAL LIBRARY 

ANALYTICAL CENTER 

RESEARCH CENTER OF SPACECRAFT RADIATION SAFETY 

U.S. NUCLEAR REGULATORY COMMISSION ATTN: DIRECTOR 

ATTN: LIBRARY 

RUTGERS UNIVERSITY 

ATTN: LIBRARY OF SCIENCE AND MEDICINE 

RESEARCH AND OTHER ORGANIZATIONS 

UNIVERSITY OF CALIFORNIA 

ACADEMY OF MEDICAL SCIENCES, SCIENTIFIC CENTER OF ATTN: DIRECTOR, INSTITUTE OF TOXICOLOGY & 

RADIATION MEDICINE 

ENVIRONMENTAL HEALTH 

ATTN: V. CHUMAK ATTN: LIBRARY, LAWRENCE BERKELEY LABORATORY 

AUSTRALIAN DEFENCE FORCE 

UNIVERSITY OF CINCINNATI 

ATTN: SURGEON GENERAL ATTN: UNIVERSITY HOSPITAL, RADIOISOTOPE 

LABORATORY 

AUTRE, INC. 

ATTN: PRESIDENT XAVIER UNIVERSITY OF LOUISIANA 

ATTN: COLLEGE OF PHARMACY


I 

FormApproved 

OMB No. 0704-OIBB 

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. 

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 

collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for information Operations and Reports. 1215 Jefferson 

Davis Highway, Suite 1204, Arlington, VA 22202-4302. and to the Office of Management and Budget, Paperwork Reduction Project 10704-0188). Washington. DC 20503 


I December 1997 Contract Report 


Retrospective Reconstruction of Radiation Doses of 

Chernobyl Liquidators by Electron Paramagnetic Resonance NWED QAXM 

6. AUTHOR(S) 

Chumak, V.V., Likhtarev, I.A., Sholom, S.S., 

Pasalskaya, L.F., Pavlenko, Y.V. 



REPORT NUMBER 

8901 Wisconsin Avenue CR97-2 

Bethesda, MD 20889-5603 





Approved for public release; distribution unlimited. 


Accurate, reliable dose reconstruction is a critical component in the epidemiological followup of 

liquidators. Dosimetry of teeth by electron paramagnetic resonance (EPR) is a state-of-the-art laboratory 

technique that is key to this effort. The Scientific Center of Radiation Medicine (SCRM) has developed 

and refined this technique in order to meet the practical demands of large-scale epidemiologic followup of 

the Chernobyl liquidators. Independent analysis using similar technology was performed by investigators 

at the University of Utah and showed good correlation with the SCRM results. The lower limit of detection 

for reliable dose reconstruction was 100 mGy. Techniques were applied to samples from approximately 

135 liquidators involved in cleanup activities within the first 2 years after the Chernobyl accident in 1986. 

Mean dose was 287 mGy, geometric mean was 205 mGy, and median dose value was 200 mGy. The 

reconstructed dose values range from 30 to 2220 mGy. Correlation of results between the two institutions 

was generally within 17%. This report also addresses some confounding factors (previous medical x-ray 

exposures, ultraviolet light effects on anterior teeth, nonlinearity of dose response curves below 100 mGy) 

and how to deal with them. 


56 

16. PRICE CODE 

17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF 

OF REPORT OF THIS PAGE OF ABSTRACT ABSTRACT 

UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED UL 

NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89) 

Prescribed by ANSI Sts. 239-18 

298-102

DTRA-TR -00-14 

,,: 

t I . 

~I 

I Neurocognitive & Physical Abilities Assessments 

Twelve Years After the Chemobyl Nuclear Accident 

. Approved for public release; distribution is unlimited. 

March 2001 

Prepared for: 

Defense Threat Reduction Agency 

8725 JohnJ. Kingman Road, Mail Stop 6201 

Fort Belvoir, VA 22060-6201 

20010504 158 

DSWA 01-97-C-0047 

Gerald L. Gamache, et. aI. 

Prepared by: KGA International 

8 Althea Street 

St. Augustine, FL 32084-2181

DESTRUCTION NOTICE: 

Destroy this report when it is no longer needed. 


Do not 

PLEASE NOTIFY THE DEFENSE THREAT REDUC 

TION AGENCY, ATTN: ADM, 8725 JOHN 1. KING 

MAN ROAD, MS6201, FORT BELVOIR, VA 22060- 

6201 ,IF YOUR ADDRESS IS INCORRECT, IF YOU 

WISH IT DELETED FROM THE DISTRlBUTION 

LIST, OR IF THE ADDRESSEE IS NO LONGER 

EMPLOYED BY YOUR ORGANIZATION.

I: 

,: 

I 

I: 

I: 


I: 

I; Ibis mailer is provided to enable DTBA to maintain current distribution lists for reports, (We would 

I; appreciate you providing the reguested information,) 

I: 

I: 

Ct Add the individual listed to your cflStnbution list 

Cl 

Cl 

Delete the cited organizationlindividual. 

Change of address, 

Nota: 

Please retum the maJDng label from the 

document so that any additions, changes, 

corrections or deletions can be made easily, 

For distribution cancellation or mora 

Information can DTRAlADM (703) 325-1036. 

NAME: __________________________________________________ __ 

·ORGANIZATION: ______________________________________________ __ 

z 

a: 

~ 

a: 

o 

z 

< 

w 

a: 

w 

::I: 

l 

=> 

U 

OLD ADDRESS 

NEW ADDRESS 

TELEPHONE NUM8ER: ~(~~) ____________________________________ __ 

DTRA PUBLICATION NUMBERITITLE 

CHANGESIDELETlONSlADOmONS, etc.) 

(Allach ShHt If mot8 ~,. ~ 

DTRAoro~erGOVERNMENTCONTRACTNUM8ER: ________________________ __ 

CERTIFICATION of NEED-TO-KNOW BY GOVERNMENT SPONSOR [If other than DTRA): 


------------------------------------------ 

CONTRACTING OFFICER or REPRESENTATIVE: 

SIGNATURE:

DEFENSE THREAT REDUCTION AGENCY 

ATTN: ADM 

45045 AVIATION DRIVE 

DULLES, VA 20156-7517 


ATTN: ADM 




Form Approved 

OMB No. 0704-0811 

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 

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, 

including suggestions for reducing this burden, to Washington Headquarters Se!vices, Directorate for Information Operations and Reports. 1215 Jefferson Davis Highway, Suite 1204, Arlington, 

VA 22202-4302, and to the OffICe of Manaaement and Budaet. Paperwork Reduction Pro· ect (0704-0188), Washinirton DC 20503. 

1. AGENCY USE ONLY (Leave Blank) 2. REPORT DATE 13. REPORT TYPE AND DATES COVERED 

990930 Technical 970519 - 990930 


Neurocognitive & Physical Abilities Assessments Twelve Years After C -DSWA01 - 97 - C - 0047 

the Chemobyl Nuclear Accident: Final Report 

PE -RTD&E 

PR -AF 

5. AUTHOR(S) 

Gerald L. Gamache, Dennis L. Reeves, Daniel M. Levinson, and 

TA-AD 

Peter I. Bidiouk 

WU-DH54909 


REPORT NUMBER 

KGA International 

8 Althea Street KGAI Report 97-1 

St. Augustine, FL 32095-2181 

9. SPONSORINGIMONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORINGIMONITORING 


Defense Threat Reduction Agency 

8725 John J. Kingman Road, MS-6201 DTRA-TR-00-14 

Fort Belvoir, VA 22060-6201 

TDACClKehlet 


This work was sponsored by the Defense Threat Reduction Agency under RDT &E RMC code B 4662 D 

AF AD 549094WlO A AF 25904D. 


Approved for public release; distribution is unlimited 

13. ABSTRACT(Maximum 200 words) 

In an effort to assess the effects of exposure to ionizing radiation on neuropsychological and physical 

abilities, a longitudinal study in and near Chemobyl, Ukraine was conducted. In this report are findings 

from 1995 to 1998. Participants were volunteers who resided in Ukraine during and since the Chemobyl 

Nuclear Power Plant accident. A translated subset of the Automated Neuropsychological Assessment 

Metrics battery and the Gamache Physical Abilities Battery were administered to a control and three 

experimental groups. Controls were healthy volunteers who resided well outside of the exposed area. 

Eliminators were decontamination and reconstruction workers with known levels of exposure. Forestry 

and Agricultural workers resided and worked in contaminated areas. Analyses of 1995 - 1998 4-year 

averaged results indicated the Eliminators were significantly impaired on all measures of neurocognitive 

14. SUBJECT TERMS 

Chemobyl GPAB radiation effects 

ANAM 

radionuclides 

17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 


UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED 

15. NUMBER OF PAGES 

117 

16. PRICE CODE 

20. LIMITATION OF 

ABSTRACT 

SAR 

NSN 754M1-280-5500 

Standard Form 298 (Rev.2-89) 

Prescribed by ANSI Sta. 239-18 

298-102

~ ___ UNCLASS~D ____ _ 

SECURITY CLASSIFICATION OF TInS PAGE 


NI A since Unclassified 


NI A since Unclassified 

13. ABSTRACT (Continued) 

and physical performance as compared to Controls. Forestry and Agricultural workers were impaired 

on subsets of the neurocognitive and physical batteries. Significant correlations between levels of 

radiation dosage and 4-year averaged physical and cognitive performance were obseIVed on 21 of24 

tasks for the combined exposure groups. The results appear to reflect the existence of clinically 

meaningful neurotoxic effects of both acute and chronic exposure to radionuclides. 

ii 

SECURITY CLASSIFICATION OF THIS PAGE 

UNCLASSIFIED

PREFACE 

This document represents the Final Report of a four year effort sponsored by the Defense Threat 

Reduction Agency (DTRA). This report provides the Tables and Figures to substantiate the effects of 

low-dosage radiation (less than 70 rads) on a Ukrainian population exposed as a result of the Chernobyl 

disaster. 

The success of this study was dependent on the active participation of a group of subject matter experts 

from military and civilian services on an international basis. Our first line acknowledgment and sincere 

appreciation is extended to our sponsor, the United States Defense Threat Reduction Agency (DTRA). 

Further, special thanks goes to Robert A. Kehlet, our Program Manager at DTRA. He has been our 

champion in this project, and successes to date are due to his dedicated scientific and program guidance. 

Development of the TWB/ANAM, which has served as the neuropsychological test instrument in this 

project, was sponsored by the U.S. Army Office of Military Performance Assessment Technology, 

Walter Reed Army Institute of Research, Washington, DC, Dr. F.W. Hegge, Program Director. The 

TWB/ANAM system was constructed at the Naval Computer & Telecommunications Station, NAS, 

Pensacola, FL, K. P. Winter, Principal Investigator. Authors of the ANAMUKR Battery were D. Reeves, 

G. Gamache, A. Chervinsky, & P. Bidiouk. Finally, the authors would like to express a special note of 

appreciation for the exceptional volunteer technical assistance provided by Dr. J. Wood, Krug Life 

Sciences during the English to RussianlUkrainian portion of this project. Her knowledge and experience 

with the NASA-MIR projects proved invaluable in helping us to not "re-invent" the wheel and launch 

our project "on-schedule." 

On-site Supervisors in the Ukraine were: Damian V. Kolisnyk, who supervised all aspects of testing the 

control group, Nikolay N. Kaletnik, who supervised all aspects of testing the forestry workers, and 

Victor G. Bondarenko, who supervised all aspects of testing the eliminators. Dr. Peter I. Bidiouk was 

the on-site project scientist and over-site supervisor for all aspects of the selection of test sites and the 

selection of participants, as well as being the on-site project administrator in Ukraine (see Appendix A). 

Finally, the authors would like to express sincere appreciation to Dr. A. J. Glasner, National Cognitive 

Recovery Foundation Editorial Board Member, who has been instrumental in production of this report 

and helping us remain "on-track" and in adherence with NCRF lAP A guidelines. 

111


Conversion factors for U.S. Customary to metric (SI) units of measurement. 

MULTIPLY • BY ~ TOGET 

TO GET ~ BY ~ DIVIDE 

angstrom 1.000 000 X E -10 meters 

atmosphere (normal) 1.013 25 X E +2 kilo pascal (kPa) 

bar 1.000 000 X E +2 kilo pascal (kPa) 

bam 1.000 000 X E -28 meter (mz) 

British thermal unit (thermochemical) 1.054 350 X E +3 joule (1) 

calorie (thermochemical) 4.184 000 joule (1) 

cal (thermochemical/em?) 4.184 000 X E -2 mega joule/mz (MJ/m 2 ) 

curie 3.700 000 X E +l *giga becquerel (GBq) 

degree (angle) 1.745 329 X E -2 radian (rad) 

degree Fahrenheit !I< = (tOf + 459.67)/1.8 degree kelvin (K) 

electron volt 1.602 19 X E -19 joule (1) 

erg 1.000 000 X E -7 joule (1) 

erg/second 1.000 000 X E -7 watt{W) 

foot 3.048 000 X E -1 meter(m) 

foot-pound-force 1.355 818 joule (1) 

gallon (u.s. liquid) 3.785 412 X E -3 mete~ (m 3 ) 

inch 2.540 000 X E -2 meter (m)' 

jerk 1.000 000 X E +9 joule (1) 

joulelkilogram (j/kg) radiation dose 

absorbed 1.000 000 Gray(Gy) 

kilotons 4.183 terajoules 

kip (1000 lbf) 4.448 222 X E +3 newton(N) 

kiplinch 2 (ksi) 6.894 757 X E +3 kilo pascal (kPa) 

ktap 1.000 000 X E +2 newton-second! m Z (N-slm 2 ) 

micron 1.000 000 X E -6 meter(m) 

mil 2.540 000 X E -5 meter(m) 

mile (international) 1.609 344 X E +3 meter(m) 

ounce 2.834 952 X E -2 kilogram (kg) 

pound-force (lbs avoirdupois) 4.448 222 newton(N) 

pound-force inch 1.129 848 X E -1 newton-meter (N"m) 

pound-force/inch 1.751 268 X E +2 newton-meter (N/m) 

pound-force/foof 4.788 026 X E -2 kilo pascal (kPa) 

pound-force/inch 2 (PSi) 6.894 757 kilo pascal (kPa) 

pound-mass (Ibm avoirdupois) 4.535 924 X E -1 kilogram (kg) 

pound-mass-foof (moment of inertia) 4.214 011 X E -2 kilogram-meter (kg. m 2 ) 

pound-massIfoof 1.601 846 X E +l kilogram-mete~ (kg/m 3 ) 

rad (radiation dose absorbed) 1.000 000 X E -2 **Gray(Gy) 

roentgen 2.579 760 X E -4 ooulomblkilogram (CIkg) 

shake 1.000 000 X E -8 seconds (s) 

slug 1.459 390 X E +1 kilogram (kg) 

torr ~mm Hg, (Y C) 1.333 22 X E -1 kilo pascal (kPa) 

• The becquerel (Bq) is the SI unit of radioactivity; 1 Bq = 1 event/so 

**The Gray (Gy) is the SI unit of absorbed radiation. 

IV


Section 

Page 

PREFACE ................................................................................................................................. iii 

CONVERSION TABLE ........................................................................................................... iv 

FIGURES ................................................................................................................................. vii 

TABLES ..................................................................................................................................... x 

1 INTRODUCTION ..................................................................................................................... 1 

2 METHOD ................................................................................................................................ 20 

2.1 PARTICIPANTS ......................................................................................................... 20 

2.2 INSTRUMENTS .......................................................................................................... 21 

2.2.1 Gamache Physical Abilities Battery (GP AB) ............................................... 21 

2.2.2 Automated Neuropsychological Assessment Battery-Ukraine .................... 21 

2.3 ASSESSMENT SITES AND ENVIRONMENT ........................................................ 22 

2.4 PROCEDURE .............................................................................................................. 22 

3 RESULTS ................................................................................................................................. 24 

3.1 OVERVIEW OF 4-YEAR RESULTS: 1995-1998 ..................................................... 24 

3.1.1 GPAB ............................................................................................................ 22 

3.1.2 ANAMUKR: Accuracy ................................................................................. 27 

3.1.3 ANAMUKR: Efficiency ............................................................................... 27 

3.1.4 ANAMUKR: Additional Measures .............................................................. 28 

3.1.5 A Clinical Neuropsychological Interpretation ofChemobyl-ANAM 

Data ............................................................................................................... 30 

3.2 CORRELATIONS BETWEEN DOSAGE OF RADIATION AND 4-YEAR 

AVERAGED PERFORMANCE LEVELS ................................................................ .31 

3.3 RESULTS OF 1995 INITIAL TEST SESSION ......................................................... .43 

3.3.1 GPAB ............................................................................................................ 44 

3.3.2 ANAMUKR: Accuracy of Performance .......................................................47 

3.3.3 ANAMUKR: Efficiency of Performance (Throughput) ............................... 50 


3.3.5 GPAB-ANAMUKR: Correlational Analyses ...............................................53 

3.4 RESULTS OF 1996 RETEST SESSION .................................................................... 56 

v

TABLE OF CONTENTS (Continued) 

Section 

Page 

3.5 GLOBAL ASSESSMENTS OF DECLINES BY EXPOSURE GROUPS ................. 56 

3.6 SPECIFIC ASSESSMENTS OF DECLINES IN EXPOSURE GROUPS ................. 60 

3.6.1 GPAB ............................................................................................................ 60 


3.6.3 ANAMUKR: Efficiency .............................................................................. ~61 



3.7.1 Global Assessments of Declines in Exposure Groups .................................. 64 

3.7.2 Specific Assessments of Declines in Exposure Groups ................................ 67 

3.7.3 GPAB ............................................................................................................ 69 



3.7.6 ANAMUKR: Additional Measures ............................................................. 70 


3.8.1 Assessments of Declines by Exposure Groups ............................................. 71 

3.8.2 GPAB ............................................................................................................ 71 




4 CONCLUSIONS ...................................................................................................................... 86 

5 REFERENCES ........................................................................................................................ 88 

Appendix 

A UKRANIAN PROJECT DEPARTMENTS AND PERSONNEL CONTACTED .............. A-I 

B PHOTOGRAPHS OF CHERNOBYL AND TESTING SITES ............................................ B-1 

C CONSENT FORM ................................................................................................................. C-1 

D GLOSSARY ......................................................................................................................... D-1 

DISTRIBUTION LIST ....................................................................................................... DL-l 

VI

FIGURES 

Figure 

Page 

3-1 4-year averaged performances on GPAB: BROADJUMP ........................................................ 25 

3-2 4-year averaged performances on GPAB: CARRYING WEIGHT ...........................................26 

3-3 4-year averaged performances on GPAB: SQUAT THRUSTS ...............................................26 

3-4 4-year averaged performances on GPAB: BALANCE BEAM ................................................27 

3-5 4-year averaged performances on ANAMUKR: 2CH-ACC ....................................................33 

3-6 4-year averaged performances on ANAMUKR: CDS-ACC .....................................................33 

3-7 4-year averaged performances on ANAMUKR: CDI-ACC ..................................................... 34 

3-8 4-year averaged performances on ANAMUKR: CDD-ACC ...................................................34 

3-9 4-year averaged performances on ANAMUKR: CPT -ACC ....................................................35 

3-10 4-year averaged performances on ANAMUKR: DGS-ACC ....................................................35 

3-11 4-year averaged performances on ANAMUKR: MSP-ACC ....................................................36 

3-12 4-year averaged performances on ANAMUKR: SPD-ACC ....................................................36 

3-13 4-year averaged performances on ANAMUKR: SRT-EFF .......................................................37 

3-14 4-year averaged performances on ANAMUKR: 2CH-EFF ......................................................37 

3-15 4-year averaged performances on ANAMUKR: CDS-EFF ......................................................38 

3-16 4-year averaged performances on ANAMUKR: CDI-EFF .......................................................38 

3-17 4-year averaged performances on ANAMUKR: CDD-EFF ...................................................... 39 

3-18 4-year averaged performances on ANAMUKR: CPT-EFF .......................................................39 

3-19 4-year averaged performances on ANAMUKR: DGS-EFF ......................................................40 

3-20 4-year averaged performances on ANAMUKR: MSP-EFF ......................................................40 

3-21 4-year averaged performances on ANAMUKR: SPD-EFF .......................................................41 

VB

FIGURES (Continued) 

Figure 

Page 

3-22 4-year averaged performances on ANAMUKR: T APPING-RlGHT .......................................41 

3-23 4-year averaged performance on ANAMUKR: TAPPING LEFT ...........................................42 

3-24 4-year averaged scores on ANAMUKR: SLEEP SCALE .......................................................42 

3-25 Mean % Decrement for Exposure Groups Relative to Controls-1995 .....................................43 

3-26 Mean % Decrement for Exposure Groups-1996 Relative to Controls-l 995 ............................ 57 

3-27 Mean % Decline for Exposure Groups: 1995-1996 ................................................................. 59 

3-28 Mean % Decrement for Exposure Groups-1997 Relative to Controls-1995 ............................ 64 

3-29 Mean % Decline for Exposure Groups: 1996-1997 ................................................................. 66 

3-30 Mean Performance on GP AB: BROAD JUMP ......................................................................... 74 

3-31 Mean Performance on GPAB: CARRYING WEIGHT ........................................................... 74 

3-32 Mean Performance on GPAB: SQUAT THRUSTS ................................................................. 75 

3-33 Mean Performance on GPAB: BALANCE BEAM .................................................................. 75 

3-34 Mean Performance on ANAMUKR: 2CH-ACC ...................................................................... 76 

3-35 Mean Performance on ANAMUKR.: CDS-ACC ...................................................................... 76 

3-36 Mean Performance on ANAMUKR: CDI-ACC ....................................................................... 77 

3-37 Mean Performance on ANAMUKR: CDD-ACC ..................................................................... 77 

3-38 Mean Performance on ANAMUKR: CPT-ACC ...................................................................... 78 

3-39 Mean Performance on ANAMUKR: DGS-ACC ..................................................................... 78 

3-40 Mean Performance on ANAMUKR: MSP-ACC ..................................................................... 79 

3-41 Mean Performance on ANAMUKR: SPD-ACC ...................................................................... 79 

Vlll

FIGURES (Continued) 

Figure 

Page 

3-42 Mean Performance onANAMUKR: SRT-EFF ....................................................................... 80 

3-43 Mean Performance on ANAMUKR: 2CH-EFF ....................................................................... 80 

3-44 Mean Performance on ANAMUKR: CDS-EFF ....................................................................... 81 

3-45 Mean Performance on ANAMUKR: CDI-EFF ........................................................................ 81 

3-46 Mean Performance on ANAMUKR: CDD-EFF ...................................................................... 82 

3-47 Mean Performance onANAMUKR: CPT-EFF ....................................................................... 82 

3-48 Mean Performance on ANAMUKR: DGS-EFF ....................................................................... 83 

3-49 Mean Performance on ANAMUKR: MSP-EFF ....................................................................... 83 

3-50 Mean Performance on ANAMUKR: SPD-EFF ....................................................................... 84 

3-51 Mean Performance on ANAMUKR: TAPPING-RIGHT ......................................................... 84 

3-52 Mean performance on ANAMUKR: TAPPING-LEFT ........................................................... 85 

3-53 Mean Ratings on ANAMUKR: SLEEP SCALE ...................................................................... 85 

lX

TABLES 

Table 

Page 

1-1 Comparison of radionuclide content released into the environment as a nuclear 

weapons tests with levels resulting from the Chemobyl accident ................................................ 3 

1-2 Volume ofCs J37 in the result of Dnieper reservoirs, measured in Curies .................................... .4 

1-3 Distribution of radio nuclides in Kiev and suburbs ....................................................................... 5 

1-4 Contamination by Cs 137 ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••• 6 

1-5 Contamination by Sr 90 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••• 7 

1-6 Radionuclide activity in suburbs of Kiev: May, 1986 .................................................................. 7 

1-7 Contamination of water by tritium ............................................................................................... 8 

1-8 Measurement of gamma radiation for 30-km zone and adjacent areas, May 16, 1986 ................ 9 

1-9 Gamma radiation in forest south of Chemobyl at various distances: May, 1 Device: 

DP-58 (milliradslhr) .................................................................................................................... 1 0 

1-10 Result of forest damage by radiation within the 30-kilometer zone ........................................... 10 

1-11 Radioactive particles from the Brown F orest ............................................................................. 11 

1-12 Contamination of forests by CS 137 ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••• 12 

1-13 Concentration ofSro and Pu in 1991 Kiev foliage ..................................................................... 13 

1-14 Contamination ofCs!37 in berries, mushrooms, and medical herbs ............................................ 13 

1-15 Contamination of Cs 137 in wood .................................................................................................. 14 

1-16 Disease rates for relocated individuals ........................................................................................ 16 

1-17 Disease rates among individuals relocated to Kiev .................................................................... 16 

1-18 Percentages of individuals who are considered healthy .............................................................. 17 

2-1 Demographic information and mean dose of radiation for the 4 groups .................................... 20 

2-2 Description of Gamache Physical Abilities Battery (GPAB) ..................................................... 21 

x

TABLES (Continued) 

Table 

Page 

3-1 4-Year averaged performance on GP AB: physical tasks ............................................................ 25 

3-2 4-Year averaged performances on ANAMUKR: Accuracy ....................................................... 28 

3-3 4-Year averaged performances on ANAMUKR: Efficiency ...................................................... 29 

3-4 4-Year averaged performances on ANAMUKR: Additional Measures ..................................... 29 

3-5 Correlations between dosage of radiation and performance levels ............................................ .32 

3-6 Mean % performance decrement for exposure groups relative to controls-1995 ...................... .43 

3-7 GP AB: 1995 Means (and Standard Deviations) for the four groups ......................................... .44 

3-8 GP AB: Results ofMANOV A and Univariate ANOV AS ......................................................... .45 

3-9 GPAB: Groups significantly lower on physical abilities tasks thanAC .................................... .45 

3-10 GPAB: Results of discriminant function analysis for groupsAC andAE ................................ .45 

3-11 GPAB: Means (and Standard Deviations) for females and males, either not exposed 

(Controls) or exposed to radiation ............................................................................................. .46 

3-12 GPAB: MANOVA and univariate tests for females and males, either exposed or not 

exposed (Controls) to radiation .................................................................................................. .47 

3-13 Accuracy (% Correct): 1995 Means (and Standard Deviations) ................................................ .48 

3-14 Accuracy: Results ofMANOVA and univariate ANOVAs ...................................................... .48 

3-15 Accuracy: Groups significantly less accurate thanAC .............................................................. .49 

3-16 Accuracy: Results of discriminant function analysis for groupsAC andAE ............................ .49 

3-17 Efficiency (Correct responses/min): 1995 Means (and Standard Deviations) ............................ 51 

3-18 Efficiency: Results ofMANOVA and univariate ANOVAs (and MANCOVA and 

univariate ANCOVAs with SRI as a covariate) ........................................................................ 52 

3-19 Efficiency: Groups significantly less efficient thanAC. ............................................................. 52 

xi


Table 

Page 

3-20 Efficiency: Results of discriminant function analysis for groupsAC andAE ........................... 53 

3-21 ANAMUKR Additional Measures: Means (and Standard Deviations) ..................................... .53 

3-22 Pearson correlations between GPAB and ANAMUKR: Accuracy (N=127) ............................. .54 

3-23 Pearson correlations between GPAB and ANAMUKR: Efficiency (N=127) ............................ 55 

3-24 Pearson correlations between GP AB and ANAMUKR: Additional Measures 

(N=127) ....................................................................................................................................... 55 

3-25 Mean % Performance decrement for Exp. Groups-1996 relative to Controls-1995 ................... 56 

3-26 Groups not significantly lower thanAC in 1995, but significantly lower in 1996 ..................... 58 

3-27 Mean % performance decline for exposure groups: 1995-1996 ................................................. 58 

3-28 Significant multivariate declines by exposure groups ................................................................ 59 

3-29 GPAB: 1996 Means (and Standard Deviations) for the exposure groups ................................. 60 

3-30 ANAMUKR Accuracy: 1996 Means (and Standard Deviations) for the 

exposure groups ........................................................................................................................... 61 

3-31 ANAMUKR Efficiency: 1996 Means (and Standard Deviations) for the 


3-32 ANAMUKR Additional Measures: 1996 Means (and Standard Deviations) ............................. 62 

3-33 Significant declines in performance by the exposure groups: 1995 to 1996 .............................. 63 

3-34 Mean % Performance decrement for expo groups-1997 relative to Controls-1995 .................... 64 

3-35 Groups not significantly lower thanAC in 1995 or 1996, but significantly lower 

in 1997 ......................................................................................................................................... 65 

3-36 Mean % performance decline for exposure groups: 1996-1997 ................................................. 66 

3-37 Significant multivariate declines by exposure groups from 1996 to 1997 ................................. 67 

3-38 GPAB: 1997 Means (and Standard Deviations) for the exposure groups .................................. 67 

xii


Table 

Page 


exposure groups .......................................................................................................................... 68 



3-41 ANAMUKRAdditional Measures: 1997 Means (and Standard Deviations) ............................. 69 

3-42 Significant declines in performance by the exposure groups: 1996 to 1997 .............................. 70 

3-43 GPAB: 1998 Means (and Standard Deviations) for the exposure groups .................................. 71 


exposure groups .......................................................................................................... , ................ 72 



3-46 ANAMUKR Additional Measures: 1998 Means (and Standard Deviations) ............................. 73 

xiii/xiv

SECTION 1 

INTRODUCTION 

The nuclear reactor accident at Chemobyl in 1986 resulted in geoecophysical damage and polluted 

farmlands and forests with radioactive contamination in im 

mediate and outlying areas. It has been estimated that 72% of the land mass of Ukraine is contaminated, 

and will be so for thousands of years (Yakovlev, 1992). By 1988, the Ukrainian registry contained 

names of347,619 civilians who had experienced medical symptoms that are frequently associated with 

exposure to the ionizing radiation fallout. In addition, 36,000 military personnel (Yakovlev, 1992) were 

listed as adversely affected. By 1992, over 1.5 million individuals were on government registries as 

having suffered medical problems associated with radiation exposures. This group included 350,225 

children and 180,144 persons assigned to "clean-up" duties at or very near the accident site 

(Awramenko, 1992). By 1993, about 7,000 people had died from apparent radiation-related illnesses, 

including heart, vascular, respiratory, and digestive diseases. One area of investigation that has been 

neglected during the aftermath of the Chemobyl accident has been the possible long term and subtle 

effects on neuropsychological functions. 

According to Chemousenko (1991) and Gittus et al. (1988), construction started on the Chemobyl 

Nuclear Power Plant (CNPP) in March, 1970. Plans called for construction of six High Power Channel 

Reactors, each having the capability to generate 1,000 megawatts of electricity. These are single-loop 

reactors, which means that steam travels directly to the turbine for electric generation. On September 

28, 1977, the fIrst of four reactors for CNPP became operational. Engineers working at the site warned 

of substantial problems because the reactor was unstable and emitting small doses of radioactivity. 

Despite these problems, more reactors were built, all operational until Saturday, April 26, 1986, when at 

1 :23 AM the fourth reactor exploded. The blast ripped off the roof and radioactive waste in the form of 

plutonium, cesium, and uranium dioxide were released (see Figure B-1 in Appendix B). 

At the reactor site, a 30-kilometer exclusion zone was established by the government to keep out nonscientifIc 

personnel. Cleanup and containment crews were dispatched to the scene where approximately 

660,000 volunteers and soldiers were employed. Only the Russian radiation monitors were outfItted for 

the task at hand (Figure B-2). The rest were "eliminators" composed of military units (infantry, 

helicopter crews, and engineer units), Ukrainian police, tractor and truck drivers from all Ukrainian 

provinces, medical doctors and nurses, scientists and engineers from Ukraine, Belarus, and Russia, and 

farm laborers (men and women) from Ukraine and Russia. The Ukrainian volunteers were 

approximately 396,515, while those living in Belarus and Russia comprised approximately 264,343. 

These cleanup crews were dressed only in surgical masks and lead aprons (Chemousenko, 1991). In 

neighboring Pripyat the population of 55,000 was not evacuated until 36 hours post-accident. 

Meanwhile in Kiev, government officials detected large doses of radiation; however, they chose not to 

cancel a planned May Day parade. 

Several years later, the accident continued to bring suffering to millions of people in the Ukraine, 

Belarus, and Russia. According to Baranov and Guskova (1988), and Laupa and Anno (1989), thirtyone 

workers at Chemobyl Nuclear Power Plant died a little over three months post-accident as a result of 

acute radiation sickness. 

1

Baryahtar (1991) reported that as many as 7,000 have died since then of radiation-related illnesses, and 

The International Chernobyl Project claimed that the scientific community lacked research and 

information on the medical consequences of radioactive pollution. Read (1993) stated that the number 

of fatalities ranged from thirty-one to "a projection that Chernobyl will ultimately claim more victims 

than did World War II." 

According to the newspaper Vestnik Chernobylia, Ukraine produces 54.4 million kilowatts-year of 

energy annually. Over 62% is generated from coal or oil burning plants, 8.3% from hydroelectric plants, 

23.5% from other nuclear power stations, and 5.5% from the two operating reactors at the Chernobyl 

Nuclear Power Plant. In the winter of 1992-1993, nuclear power plants accounted for 40% of all electric 

power generated. In 1990,5.8% of this power was generated by Chernobyl; in 1991 the rate droppecl to 

5.2% while in 1992 the rate was only 2.2%. 

Two reactors are still operational at Chernobyl (Figure B-8). The third reactor was destroyed in a fire in 

October, 1991. Each reactor is programmed for 30 years of service; however, the two remaining 

reactors were scheduled to be deactivated in 1993. Decommissioning will cause substantial problems 

because the four reactors have generated over 244 million kilowatts-year of energy, including 94 million 

kilowatts-year since the accident. The two remaining reactors have the capability of generating 10.3 

million kilowatts-year of electrical energy at a cost of $515 million dollars. Utilizing them would save 

2.6 million tons of oil or 6.1 million tons of coal, purchases the Ukrainians cannot afford. In October, 

1993, the law which would have closed Chernobyl by the end of 1993 was repealed. 

For the year 1989, the absorbed dose for people at Chernobyl was 1.3 rads compared with a lifetime 

dose of35 rads. In 1992, the absorbed dose was only .97 rads. The amounts released into the 

atmosphere by the cracking sarcophagus comprise only 20% of the rate of the fully functional reactor. 

Decommissioning will involve four steps. First, the reactors are shutdown with a one-year cooling 

period. In step two, the nuclear fuel is removed and kept in water for one to two years before placing it 

in special storage. In stage three, Chernobyl will be temporary closed for a period of 20 to 30 years to 

allow any radionuclides present in the facility to decay. Finally, in stage four, Chernobyl will be 

dismantled. 

When Chernobyl was fully operational, a total of 50,000 persons worked at four reactors. By 1987- 

1988, 90% of these workers had been replaced. Of the replaced workers, more than 90% were under the 

age of 45 and 60% graduated from a university or technical colleges. Today, the 25,000 present staff 

lived about 50 kilometers from Chernobyl in the town of Slavutych, which was constructed in 1986 after 

the town of Prypiat was evacuated. 

Debate over the future of Chernobyl is still going on in the Ukrainian Supreme Soviet. On one side of 

the debate are those who believe the economy of Ukraine will not improve by decommissioning. They 

point out the replacement cost of alternative methods of generating 

electricity. They also point out the Dnieper basin has seven other reactors with the same design as 

Chernobyl, and another similar reactor will soon be operational near Kursk. They argue that Chernobyl 

has a highly trained staff which will go elsewhere if the reactors are decommissioned, and each year 

savings from nuclear power can be turned to clean-up efforts and treating those still suffering from the 

accident. 

2

Finally, they argue that other third-world governments are not decommissioning Russian-made RBMK 

reactors; in fact, Russia is planning to double energy production by nuclear power by the year 2010, 

although they will not rely on the RBMK reactor. 

According to Yakovlev (1991), the Chemobyl incident resulted in waste emission of approximately 50- 

80 million Curies (Ci), including over a million Ci of Cs137, 200,000 Ci of Sr 90 and between 3,500 and 

5,500 Ci ofPu239124o into the environment. 

His research states that as of January 1, 1992, forty three thousand square kilometers, or 15% of the 

Ukraine, was contaminated by Cs137, with doses greater than one curie per square kilometer. Over 72% 

of Ukraine is contaminated above background radiation. This area has 3,200 towns and villages and a 

population, excluding Kiev, of over 4,000,000 people. Migratory birds and animals carry radiationrelated 

diseases along their routes, increasing the contaminated area. Table 1-1 (Sobodovych et al, 

1992) shows contamination from Chemobyl as compared with that resulting from nuclear weapons tests. 

While Chemobyl nuclear contamination is low compared with Russian nuclear weapons tests, the 

scientific community cannot afford to forget the fact that Chemobyl is adjacent to a heavily populated 

area. 

Table 1-1. Comparison of radionuclide content released into the environment as a 

result of nuclear weapons tests with levels resulting from the Chernobyl 

accident. 

Type of Half-life in Nuclear Total in Chernobyl Chernobyl %of 

radio- years weapons test reactor outburst % outburst NWT 

nuclide (million Ci) (million Ci) (million Ci) 

Sr 90 28.60 57.5 6.00 6.0 0.30 0.50 

Cs 137 30.17 87.0 7.02 30.0 2.10 2.40 

PU 238 87.74 0.0055 0.0254 9.0 0.00076 13.80 

PU 239 24118.00 0.96 0.0256 9.0 0.00077 0.20 

PU 240 6570.00 0.50 0.040 9.0 0.0012 0.20 

PU 24 ! 1435.00 23.00 4.97 9.0 0.15 0.70 

PU 242 .763x10 5 0.00045 0.000056 9.0 0.000002 0.40 

It should be noted that 50-80 million curies of radiation were released into the environment from the 

Chemobyl accident, as compared to only 14-20 from the Three Mile Island incident. 

During the accident, reactor cooling water was flushed into the Prypiat River, a tributary of the Dnieper, 

rather than allowing more radioactive steam to escape into the atmosphere (Chemousenko, 1991). The 

result of this action, and the break in the sewage system that serviced the Chemobyl cleanup effort, are 

the principle causes of radionuclide pollution in the Dnieper and Prypiat rivers, as well as their water 

reservoirs which serve to irrigate the region. Thus the contamination is spread. At present, the vertical 

migration of radionuclides approaches a depth of one meter from the surface. 

3

According to Professor V. Kopeikin (1993), water near the burial place ofChernobyl debris is about 4 

meters underground. Ten wells with filters installed were constructed on-site to monitor ground water 

contamination. These wells are monitored daily and were placed every four meters, with depths ranging 

from 8-9 meters. 

The damaged reactor is covered by a sarcophagus composed of220,000 m 3 of concrete and 15,000 m 3 of 

steel, around which there are 5 to 6 meters of "clean topsoil" (Yakovlev, 1991). 

However, the sarcophagus has approximately 700 square meters of heat cracks caused by the damaged 

reactor core underground. The cracks cause venting of radioactive materials from fragments of the 

reactor core and the graphite bed underneath it. About 75% of the residual nuclear fuel is composed of 

clinker (135 tons) and nuclear dust (10 tons). The maximum temperature at the surface is 60 degrees 

centigrade, while the temperature underground approaches 200 degrees centigrade. 

On the surface radioactive contamination is about 3,000 roentgen per hour. In 1990-1991 the outburst 

from radioactive dust was 1,000 times less than that of a working reactor. However, with time and 

clinker decomposition, the amount of nuclear dust will increase. On May 30-31, 1990, Chernobyl was 

struck by an earthquake measuring 4.0 on the Richter scale. No damage to the sarcophagus was 

reported. In 1992 the Ukrainian government opened bidding on a replacement sarcophagus. In June, 

1993, the lowest bidder, a French company, began studying the problem. During 1987-1990,270,000 

rubles were spent to increase safety of all reactors in the Ukraine. 

According to Woytsehowich (1991), as shown in Table 1-2, between 1986-1990 the amount ofCs l37 in 

the Kiev water reservoir increased more than 17 times. In the Kremenchug reservoir, approximately 170 

miles south of Kiev, the pollution has doubled; this signifies a spreading of radio nuclides along the 

Dnieper basin. 

Table 1-2. Volume orcs!37 in the Dnieper reservoirs, measured in curies. 

Reservoir 1986 1987 1988 1989 1990 

Kiev 413 850 -- 1000 7200 

Kanev 60 -- -- 570 2200 

Kremenchug 150-200 -- 218 294 294 

The maximum values of specific activity of radionuclides as well as heavy metals are in riverbeds and 

shallow bays, while the minimum values are in the floodlands and irrigated fields. From CNPP, along 

the Prypiat and Dnieper rivers, down to Kiev Reservoir, three contamination zones have been 

established (Woytsehowich, 1991) for those areas affected by Cs 137 • The Migration Zone lies along the 

Prypiat river between CNPP and its entrance into the Dnieper. The Accumulation Zone is from the 

mouth to 15 kilometers upstream from the reservoir dam. The Wash Away Zone is from the point 15 

kilometers from the dam to the dam itself. 

4

Table 1-3 (Sobodovych et al., 1992) shows soil contamination in Kiev and its suburbs to be .015 to 5.31 

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 

is the contamination by Sro, which is 50-750 Ci/Km 2 . The concentration of plutonium was reported by 

two sources. Source 1 (Ukrainian) shows the level of contamination to be 5 to 10 Ci/km 2 , while Source 

2 (Russian) shows the level of contamination to be much lower--0.1 to 2.7 Ci/km 2 . The authors were 

given no explanation for these disparities. 

Table 1 -3. Distribution of radionuclides in Kiev and suburbs. 

Note: 

Radionuclide Activity (CiIkg) Density of contamination 

Cs 137 2.6 x 10- 10 17.94 x 10- 8 

Ce l44 3.06 x 10- 10 11.12 x 10- 8 

RU lO6 4.31 x 10- 1 °/2.19 x 10- 8 

(1) Maximum Contamination Levels in Kiev 

1880 mCilkm 2 (Kudry, Pechersk) 

4524 mCilkm 2 (Montazbnikov, Sovley) 

1182 mCi/km2 (Davydova, Rusanovka) 

1367 mCi/km2 (Kybalchiucha, Voskresenka) 

1105 mCi/km2 (Kosmomantov, Otradny) 

(2) Pu Contamination: 

Source 1: 5 - 10 Ci/km 2 

Source 2: 0_1 - 2.7 Ci/km 2 (Ci/km 2 ) 

0.015-5.31 

0.02-0.80 

0_03-1.57 

The city of Kiev comprises 340 km 2 ; however, only 7% is contaminated, mostly with Cs 137 • As shown in 

Tables 1-4 and 1-5 (Sobodovych et aI., 1992), in the largest area in the Kiev region which includes the 

city plus surrounding countryside, fifty percent of the total landmass has received greater than 40 Ci/km 2 

of contaminated radiation. This exceeds the recognized standard for lifetime exposure rate of 40 Ci/km 2 ; 

however, the city itself is quite habitable. 

Dr. Kaletnik, Head of the Scientific and Technical Office for the Ministry of Forest, was questioned why 

the Kiev region, rather than the city itself, was so highly contaminated. He explained that this 

phenomenon was due to excessive forest fires in the area, and that rising smoke and soot which blankets 

an area after rainfall would increase already high levels of contamination. 

5

Table 1-4. Contamination by CS137. 

Concentration 

Intervals 

(CiJKm 2 ) 

0.5-1.0 1.0-5.0 5.0-15.0 15.0-40.0

Table 1-5. Contamination by Sr>°. 

Concentration 

Intervals 

CilKm2) 

0.05-.50 0.50-1.0 1.0-3.0

Finally, tritium, a radioactive isotope of hydrogen, was present at Chemobyl (Woytsenhowich, 1991). 

Tritium emits negative beta particles of 19,000 electron volts of energy and has a half-life of 12.5 years. 

Table 1-7 (Sobodovych et aI., 1992) shows the testing of water contaminated by tritium. Since tritium 

naturally occurs in water, probably the action of cosmic rays on atmospheric hydrogen, one would need 

prior testing to determine if these measurements were significant. If they were significant, the dates of 

these measurements might indicate serious problems. 

Table 1-7. Contamination of water by tritium. 

Type of Date of Number of 

Concentration (BklLiter) 

Water test test Range Median 

1. Atmospheric Jan 92-Feb 92 51 2.7 to 6.4 3.5 

(Snow, Water) 

2. Surface Water Nov 91-Jan 92 124 2.7 to 11.8 4.8 

3. Ground Water Nov 91-Jan 92 101 2.7 to 12.7 5.1 

4. Underground Sep 91-Jan 92 68 2.2 to 5.2 2.2 

Water 

Within weeks following the Chemobyl disaster, measurements to assess damage to forests surrounding 

the area were taken. Ultimately, about 500 hectars of forests were destroyed as a result of the accident. 

The government established a 30-kilometer fenced zone around the reactor with military patrols, whose 

purpose was to prevent unauthorized access. Even outside this zone, the density of contamination was 

10 to 80 Ci/km 2 • At least 28,000 km 2 of forest were contaminated. 

The first reported measurement of gamma radiation occurred on May 16, 1986,20 days post-accident 

(Sobodovych, 1992). Two devices were used for measurements. The first device, the Russian Army 

DP-5B, was available to units working in nuclear contaminated battlefields where it was used for rough 

estimates of nuclear contamination. The second device, SRP-6801, yielded more precise measurements 

and was carried by engineers who surveyed the site. Table 1-8 shows the results for measurements inside 

the 30-kilometer zone and adjacent forest areas. Approximately 100 hectare grids were laid out on maps 

by the Ministry of Forest. A sampling technique was then utilized. The number in parentheses 

following the forest name in the first column denotes the grid area. 

8

Table 1-8. 

Measurement of gamma radiation for 30-km zone and adjacent areas, 

May 16, 1986. 

DP-58 

SRP-6801 

Location of the measurement (milliradslhr) ( milliradslhr) 

Army device (more precise) 

1. Polissia Forest (18) 

Meadows 0.14 0.20 

Free in air 0.10 0.25 

Grass 0.12 0.25 

Carpet 0.20 Not Measured 

2. Radynskoye forest (12) 

Forest edge 0.13 0.25 

Meadows 0.20 0.41 

Trees 0.26 0.44 

Carpet 0.22 0.42 

Village of Cheremoshe 0.20 0.38 

3. Radynskoye forest 

Edge of pine trees 0.75 1.25 

Young mixed trees 0.75 1.50 

Free in air 0.75 1.40 

Crown of trees 0.75 1.45 

Grass 0.80 1.50 

Moss 1.50 1.50 

Oat field 0.45 1.25 

Ground 0.70 0.70 

Note: numbers in parentheses show grid number denoting location of sampling squares. 

Table 1-9 (Sobodovych, 1992) shows gamma radiation in forests south of the reactor. Measurements 

were also made in May, 1986. These forests contain coniferous trees, which are especially vulnerable to 

radiation. This table is useful because the distance from the reactor is measured, as well as the growth of 

pine trees in a particular forest. 

9

Table 1-9. Gamma radiation in forest south of Chernobyl at various distances: 

May,1986; Device: DP-58 (milliradslhr). 

Location Free in air Carpet 

1. Dymer forest (100 km*) 

Pine trees-50 years old 0.6 1.0 

2. Ivankov forest (80 km*) 



Pine trees-l 8 years old 1.3 3.2 

3. Chemobyl forest (20 km*) 


4. Novoshepelychi forest (l0 km*) 


Note: * denotes kilometers from reactor to center of forest. 

Table 1-10 is extremely interesting, since it shows the result of forest damage within the 30- kilometer 

zone. It was constructed based on conversations with Mr. Kaletnik. This table was cited in the Pacific 

Sierra Research Corporation's analysis of Landsat imagery (McClellan, G.E. et al., 1994). The distance 

from the reactor site is approximately one kilometer, and as Table 10 reflects, 100% of the trees located 

350 meters from the edge of the forest were recovered and sent to mills to be used as lumber. The rest of 

the trees were bulldozed into large pits and covered with topsoil. This action has increased ground 

contamination, and the Ukrainians are presently reviewing options to deal with this situation. 

Table 1-10. Result of forest damage by radiation within the 30-kilometer zone. 

Distance from Calculated absorbed % of tree Degree of 0/0 of 

the edge of the Dose-Rad x 10 3 crown harm recovered 

forest damage trees 

Edge of forest 10 100 Completely dry 0 

Wood 

35 meters 6.5 50 Very strong 25 

90 meters 4.9 20-30 Medium 50 

350 meters 0.5 Up to 10 Small 100 

10

In 1986, the crowns of trees contained about 50% of the radionuclides (Yakovlev, 1992). By 1988,95% 

of the radionuclides were in humus or carpet. Today, most radionuclide content of the foliage has 

migrated through the root system. 

Beginning in the summer of 1988 measurements were made of radioactive particles within the 30- 

kilometer zone in an area called the "Brown Forest", so named because the leaves and vegetation were 

discolored by radiation. Table 1-11 (Sobodovych, 1992) shows the results of these measurements, with 

the percentages of each isotope listed. The table also displays the particle properties as either irregular 

shaped flakes or round balls with varying composition of beryllium, cuprum, lead, silicon, tantalum, and 

iron. High concentrations of cerium, cesium, and ruthenium are shown. 

Table 1-11. Radioactive particles from the brown forest. 

Date of Form, size (mcm), and Basis of Element (percentages) 

Test properties of particles particle Ce l44 Cs 134 CS l37 Ru lo6 Co 6O 

Jun 1988 Black, hard, magnetic Oxides of 2.0 - 2.0 94.0 2.0 

balls, 46.7 mcm 

Fe 

Aug 1988 Irregular, 1.2-6.5 mcm and Be, Pb, Cu 54.2 5.4 15.5 24.9 - 

Balls, 4.1-5 mcm, dark br 

Jan 1989 Irregular, 1.6-88.0 mcm, Fe, Si, Pb 50.0 4.5 22.3 29.0 - 

dark br, nonmagnetic 

Jan 1989 Irregular, fragile, black, Fe, Si 4.0 8.9 92.6 24.5 - 

Nnnmagnetic, 2.0-20 mcm, 

Balls, 1.0-4.0 mcm 

Jan 1989 Irregular, 1.2-6.4 mcm, Si, Ta 88.2 0.8 3.8 7.2 - 

dense, black, with balls, 

0.6-2.4mcm 

Table 1-12 (Sobodovych, 1992) displays a list of forests contaminated by Cs 137 and the levels of 

contamination. These measurements were made in 1990 and 1991. Figures for 1992 were not available. 

Table 1-13 (Woytsehowich, 1991) shows measurements taken from Kiev in 1991 in which foliage 

samples were collected, burned, and analyzed for Sr 90 and Pu. These measurements were taken in three 

parks in Kiev proper. Leningradskaya Square is in the center of downtown Kiev, and the hydropark lies 

along the Dneiper River to the east of downtown. 

11

Finally, Tables 1-14 and 1-15 (Baryahtar & Bobyleva, 1991) reflect the Ukrainians' concern with 

individuals growing, harvesting, and consuming food from contaminated areas. 

Blackberries, mushrooms, and medical herbs were chosen for analysis because they are both plentiful 

and susceptible to the effects of radioactive contamination. The transition coefficient presented in these 

tables was designed to give a "density rating" by converting measurements taken in square meters to 

more useful kilograms. The specific activity for CS 137 is listed in becquerels per kilogram, with a level of 

confidence as shown. 

Table 1-12. Contamination of forests by CS137. 

Regio Year Total Levels of Contamination CiIkm~) 

n area Studied

Table 1-13. Concentration ofSro and Pu in 1991 Kiev foliage. 

Location Weight of Weight of Sro Pu 

dry sample ash (Bk/kg) (BkIkg) 

Leningradskaya square 987 gr 120 gr 2.5-22.9 0.03-0.10 

Lesnoy district 638 gr 67 gr 4.6-49.4 0.05-0.22 

Hydropark 836 gr 90 gr - 0.05-0.39 

Note: Free-in-Air: Pu = 10-70 Bklm 3 or 10- 13 to 10- 12 Ci/liter 

Table 1-14. Contamination OfCS 137 in berries, mushrooms, and medical herbs. 

Contamination in Ci/km 2 

Table 1-15. Contamination of CS137 in wood. 

Contamination in Ci/km l 

Prior to 1991 there were no cases of internal organ cancers among children, yet today it ranks as the 

second largest cause of infant mortality. In 1991, 37.1 % of newborns evidenced some kind of pathology 

and in Kiev there were 2,500 premature deliveries. Anemia increased four times among pregnant 

women post-Chernobyl. 

According to Ukrainian law, individuals who were employed at Chernobyl and those involved in the 

clean-up efforts were divided into five categories depending on the amount of radiation absorbed. 

Category I individuals were assigned to Chernobyl when the accident occurred and received at least 25 

Gy of radiation. Category II were men who were assigned clean-up duties who also received 25 Gy or 

more of radiation. Category III were individuals who received between 10 and 24.9 Gy of radiation. 

Category IV were individuals who received between 5.0 and 9.9 Gy, and Category V were those 

individuals who received between 0.1 and 4.9 Gy. One Gy is equivalent to 100 rads. 

According to the Office of Chernobyl Affairs in Kiev, the new disease rates for individuals who received 

at least 25 Gy has doubled over those receiving less dosages. The relationship between absorbed dose 

and effect has only been investigated since 1990. In 1990, it was noted that increased oncological rates 

were attributed to men who received 25 Gy or more and to women who received lOGy or more. 

Generally, endocrine disease rates among male clean-up crew workers increased almost 4 times the 1988 

rate. The endocrine disease rate among women who received lOGy or more has doubled each year 

between 1989 to 1990. Category II personnel were 80 - 100% more likely to suffer from digestive and 

nervous system diseases than individuals in other categories. In the Kiev, Zhytomir, and Chernigov 

regions the rates for newly acquired diseases for Category V are 26.2% hematic system, 18.2% 

respiratory system, and 12.6% nervous and digestive system (Baryahtar & Bobyleva, 1991). The 

maximum disease rates are in Ivankov, Polissia, Narodichi, and Ovruch. In the Rovno region, the lowest 

newly acquired disease rates were reported among children born in 1984 and 1985. Disease rates for 

children born after the Chernobyl accident were 1.5 to 3.0 times higher than those born prior. In the 

same region, respiratory illness among children accounts for 25 - 40% of all new illnesses, especially 

among children 1 - 3 years of age where they present with respiratory problems 8 - 10 times per year. 

The other sixty percent of children present with thyroid gland problems. Anemia has increased among 

children 2.5 to 3.2 times between 1985 and 1988. . 

Ukrainian law also divides people who suffered from Chernobyl into another five-group categorization. 

Group A includes 5,237 disabled individuals, 187 people diagnosed with acute radiation symptoms, and 

15,000 people who suffered diseases directly attributed to the Chernobyl accident. Group B includes 

180,000 personnel who took part in the clean-up efforts, 130,000 people who received doses in excess of 

250 mSv and were relocated from areas inside the 30 kilometer zone, and 12,000 children born to 

parents involved in the clean-up effort. Group C includes children who have thyroid gland radiation in 

excess of allowable standards, 60,000 people who took part in the clean-up effort from 1988 until 1990 

who received less dosages than those who were first on the scene, and approximately one million people 

who currently live in contaminated areas but who await relocation. Among them are 350,000 children, 

65,000 of whom were born after the accident. Group D includes 1.5 million persons who work or live 

permanently in areas still receiving radio-ecological monitoring, in addition to their 400,000 children. 

Some government figures include in Group D all the inhabitants of the Kiev, Chernigov, and Zhytomir 

regions, or approximately another 4.5 million people. 

15

In 1990, the fifth group was added, comprised of women, when it was realized they showed the highest 

rates for new respiratory and digestive systems diseases for ages of30 - 39. Today, the Ukrainian 

registry contains the names of347,619 civilians who suffered direct medical problems as a result of 

Chemobyl, plus 36,000 military personnel who were also affected. 

Table 1-16 (Awramenko, 1992) shows the newly acquired disease rates for relocated persons. These 

individuals lived in contaminated areas but were forced to relocate because of excessive levels of 

radiation. Most of these individuals belong to Group D and the majority of them have been relocated. 

Table 1-16. Disease rates for relocated individuals. 

Disease 1986 1990 Percent increase 

Heart and blood 0.74/1000 6.9411000 938% 

Endocrine, digestion, 12.67/1000 171.11/1000 1350% 

Immune 

Respiratory 23.6711000 136.6811000 577% 

Nervous 21.25/1000 106.28/1000 500% 

Table 1-17 (A wramenko, 1992) shows the disease rates for those people relocated to Kiev. 

Table 1-17. Disease rates among individuals relocated to Kiev. 

Disease 1986 1990 Percentage 

Increase 

Endocrine system 11.7911000 119.7311000 1016 % 

Respiratory system 26,8911000 163.1611000 607% 

16

Finally, Table 1-18 (Awramenko, 1992) shows the relative health of these four groups from 1988 to 

1991. Year-to-year percentages are decreasing due to survival rate. 

Table 1-18. Percentages of individuals who are considered healthy. 

Groups 1988 1989 1990 1991 

Group A 74.0 66.4 52.8 33.8 

Group B 

Adults 61.5 44.1 35.3 28.8 

Children - 43.9 35.2 29.1 

Group C 

Adults 35.4 35.4 26.0 31.7 

Children - 52.9 40.7 39.8 

GroupD 

(Only 400,000 children - 77.7 62.9 48.5 

were tested) 

As of January 1, 1992, 1,536,270 persons were registered by the Ukrainian government as having 

suffered medical problems as a result ofChernobyl. Among those were 350,225 children and 180,144 

personnel assigned clean-up duties after the Chernobyl accident (A wramenko, 1992). 

Seventy percent of workers in the Narodichi forest in the Zhytomir region received 0.44 rads per year of 

radiation, and 10% received more than 2.3 rads per year (Yakovlev, 1992). The highest content of CS137 

was absorbed by woodcutters and forestry workers in the Rovno region which contains two forests, the 

Vladimiretsky and Dubrovitsky. These workers received between 31.8 and 74.2 Bk/kg. 

Ukrainian law divides all forests into four categories, dependent on CS137 dose. Category I includes the 

Pollessky and Narodichi forests, where contamination exceeds 40 Cilkm? Category II contains the 

Dymer, Ivankov, Ovruch, Luginsk, and Slovechansky forests, where active monitoring shows 

contamination between 15 and 40 Ci/km 2 • Category III includes forests with contamination between 5 

and 15 Ci/km 2 • These forests require continuous monitoring. Category IV consists of forests containing 

no more than 5 Ci/km 2 • The Ministry of Forests in Kiev is concerned about forest workers who have 

monitoring and woodcutting duties in contaminated forests. 

In the 10 years following the Chernobyl nuclear accident (as of 1996), the number of healthy individuals 

living in contaminated areas decreased from 67.1 % to 33.1 %. Chronic pathologies increased from 

31.5% to 66.0%. 

17

Most cases reflect pathologies of the endocrine system, blood and blood-generating systems, nervous 

system, and gastroenterological system. The most substantial growth of this dangerous statistics is related 

to young people aged 15-17 years--6.6 times normal (8.1 times for boys and 5.6 times among girls). 

Death rates and disability rates have increased substantially (about 2.5-3.5 times), as compared to those 

of people who lived in normal conditions. Pathology of the thyroid gland constitutes 72.7% of all 

endocrine cases for women, and 62.5% for men. 

As of 1998, the official death rate among the "Eliminators" (who are still alive) is 1.8 (80% higher than 

normal). This is primarily due to higher incidences of cancer, diseases of blood and blood-generating 

organs, and pneumonia. 

The present study, an ongoing longitudinal project which commenced in 1995, entails assessments of 

neuropsychological and physical capabilities of four independent volunteer participant groups. The 

control group (Controls) consists of healthy volunteers that reside outside the immediate radiation 

exposure area. The second group consists of "Eliminators," who are individuals who were involved in 

the tasks of removing nuclear debris and assisting in construction of the containment chamber for the 

defective reactor facility. The third group of volunteers consists of Forestry workers who perform 

monitoring, woodcutting, and other related activities in the Narodichi forest, which is in close proximity 

of Chernobyl. It is known that 70% of workers in the Narodichi forest (in the Zhytomir region) received 

approximately 0.44 rads per year of radiation, and 10% received more than 2.3 rads per year (Yakovlev, 

1992). Finally, the fourth group is comprised of Agricultural workers from Rozumnytsia, which is 

approximately 150 km south of Kiev, and for whom knowledge of the level of radio nuclide is known. 

The instrument that was chosen to be the primary measure of cognitive performance is the Automated 

Neuropsychological Assessment Metrics (ANAM) Battery, which is a subset of the Office of Military 

Performance Assessment Technology (OMPAT) Tester's Workbench (TWB). The TWB is a library of 

automated tests that has been constructed to meet the need for precise measurements of cognitive 

processing efficiency. The ANAM batteries are unique combinations ofTWB tests that have been 

configured for neurocognitive assessment and evaluation of functioning in a variety of 

neuropsychological domains. Many of the component tests in ANAM were derived from the Unified 

Tri-Service Cognitive Performance Assessment Battery (UTCPAB; Reeves et al. 1991) and the Walter 

Reed Performance Assessment Battery (Thorne et al. 1985). The Ukrainian subset of ANAM 

(ANAMUKR) was designed by Reeves and Gamache (1994), and constitutes a specialized subset of the 

TWB-ANAM batteries. It consists of tests that have been configured for repeated measures testing for 

neurocognitive impairment due to exposure to radionuclides. It has been designed to assess levels of 

neurocognitive function ranging from superior to moderately impaired. ANAMUKR subtests also 

include a stand-alone module for assessing sustained attention (Running Memory Continuous 

Performance Test). 

In the present study, the ANAMUKR battery was combined with the Gamache Physical Abilities 

Battery (GP AB; Gamache, 1993), for testing the physical capabilities of individuals exposed to 

radionuclides. This battery, composed of tests derived from Fleishman and Quaintance (1984), is 

especially sensitive to the physical decrements in performance resulting from exposure to radionuclide 

contamination. The GP AB consists of tests designed to measure explosive strength (broadjump), static 

strength (carrying weight), dynamic strength (squat thrusts), and gross body equilibrium (balance beam). 

18

In this report, we describe data on the GP AB and ANAMUKR obtained from four independent groups 

from Ukraine in 1995 (initial test), and in 1996,1997, and 1998 (repeated tests). These data provide a 

reference point from which to gauge physical and cognitive performance of individuals who mayor may 

not have been exposed to varying levels of ionizing radiation resulting from the nuclear accident at 

Chemobyl in 1986. 

19

SECTION 2 

METHOD 

2.1 PARTICIPANTS. 

The participants in the initial phase of the study consisted of 127 volunteers (24 females, 103 males) 

who lived in Ukraine prior to 1986. Ages ranged from 11-61 years, averaging 40.21 years. The four 

groups into which they were divided included a non-exposed Control group (AC); and three exposed 

(exposure) groups: Eliminators (AE), Forestry Workers (AF), and Agricultural Workers (AG). 

Demographic information is presented in Table 2-1. Mean dose levels of exposure to radiation (in rads) 

for each group are included; these are based on the medical records of the individuals. 

Table 2-1. Demographic information and mean dose of radiation for the 4 groups - 

above background radiation. 

GROUP~ AC(n=31) AE(n=36) AF(n=29) AG (n=31) 

Age 

Mean (S.D.) 33.23 ( 7.86) 40.47 ( 6.81) 50.83 ( 7.83) 36.32 (14.26) 

Gender 

Male 24 33 29 17 

Female 7 3 0 14 

Mean dose 

In rads 0 62.95 12.61 8.81 

(FIA) 

These individuals were randomly assigned to their groups, which were counterbalanced by occupation. 

Further, participants in the control group were assigned to match by occupation, as closely as possible, 

the exposure participants. For example, if there was a truck driver in any of the three exposure groups, a 

truck driver was sought as a control. In addition, participants in the control group were matched for age 

and gender to those in the exposure groups. 

20

2.2. INSTRUMENTS. 

2.2.1. Gamache Physical Abilities Battery (GP AB). 

Physical testing involved two stations, each equipped with a stop watch and tape measure. Three tests 

were timed with a stopwatch: balance beam, squat thrusts, and carrying weights. A description of each 

test in the GPAB is presented in Table 2-2. 

Table 2-2. Description of Gamache Physical Abilities Battery (GP AB). 

Test 

Broad jump 

(BROADJMP) 

Carrying weight 

(CARRYWGT) 

Squat thrusts 

(SQUATTHR) 

Balance beam 

(BALBEAM) 

Measurement 

Distance covered in one broad 

jump (meter) 

Distance covered in 30 sec. 

(meter) 

Number of squat thrusts in 2 

mm. 

Distance covered in 20 sec. 

(meter) 

Apparatus 

Designated starting point. 

Tape measure. 

10-meter course. participants run 

back and forth in straight lines 

carrying 15 kilograms of sand 

(men) or 10 kilograms (women, 

children). Tape measure, stop 

watch, buckets of sand 

Any area where squat thrusts can 

be don. stop watch 

4-meter board, 12 centimeters 

wide, 15 centimeters from 

ground. Tape measure, stop 

watch. 

All physical test scores were recorded in laboratory notebooks. Two observers participated as test 

administrators for the physical abilities battery. One administrator read instructions to participants and 

ensured their understanding. The other timed and/or took measurements as appropriate for each test. 

Independent observer confirmation was required prior to recording scores. All subjects were tested 

according to standard procedures, with one exception. The hospital where the eliminators resided did 

not allow buckets of sand on the premises. Therefore, hand-held weights equivalent to the weight 

carried by other groups were substituted. 

2.2.2 Automated Neuropsychological Assessment Battery-Ukraine (ANAMUKR). 

The tests in the ANAMUKR battery includes the Stanford Sleepiness Scale (SLP), Code Substitution 

(visual search, immediate recall, and delayed recall: CDS, CDI, CDD), Running Memory Continuous 

Performance Task (CPT), Digit Symbol (DGS), Matching to Sample (MSP), Spatial Processing (SPD), 

Simple Reaction Time (SRT), Tapping-Right and Left Index Fingers (TAPR & TAPL), and Twochoice 

Reaction Time (2CH). These subtests of ANAM have been described previously (Reeves & 

Winter,1992; Levinson & Reeves, 1994). Each session required approximately 60 minutes. 

21

2.3 ASSESSMENT SITES AND ENVIRONMENTS. 

Participants in the Control group resided in Temopil (pop. 250,000). This city is located approximately 

450 km west of Kiev, the capital of Ukraine. All testing was conducted in High School Number 22 (Fig. 

B-7). The Eliminators were tested in the Ukrainian Center for the Radiation Protection of the Population, 

which is essentially a hospital environment (Fig. B-3). This special hospital is in the suburbs of Kiev, 

and was established to attend to the medical needs of these individuals. The Forestry Workers were 

tested in the Ovruch forest, approximately 250km northwest of Kiev. All testing was conducted in their 

barracks (see Fig. B-8). The Agricultural workers were tested in the village ofRozumnytsia, in the Kiev 

region, approximately 150 km south of Kiev. The testing site was a farmhouse (Figs B-4, B-5, B-6). 

2.4 PROCEDURE. 

On days assigned to specific groups, researchers prepared the testing site by installing laptop computers 

for administration of the ANAMUKR battery (Fig. B-4), and by setting up the apparatus for 

administration of the physical abilities test battery. A table and two chairs were situated at each of three 

ANAMUKR testing stations. Each participant sat in one chair and the test administrator sat in the other. 

This enabled the administrator to ensure that the participant understood instructions and was prepared 

for testing. Participants were instructed to ask as many questions as necessary to ensure full 

understanding prior to testing. One table was placed at the entrance for participant registration and 

orientation, which included having each participant read and sign an informed Consent Form in Russian 

and English (see Appendix C -only the English version is shown). 

During periods when all computer test stations were occupied, participants first completed the GP AB. 

The rotation of physical testing (i.e., order of balance beam, squat thrusts, etc.) was randomized to the 

extent that space was available. A rest area was established for participants while awaiting further 

testing. During test sessions the administrator ensured that there was no discussion among participants 

about up-coming tests. At the conclusion of the testing sessions, all participants were thanked and 

given the equivalent of two USA dollars. 

The ANAM battery test scores were stored on the hard disk drive of each computer. At the end of the 

day, all scores were copied to backup 3 Yz" floppy disks and marked for that group and year of testing. 

The backup disks were then re-copied to a second diskette. 

Participants in the four groups were tested on the GPAB and ANAMUKR in 1995, 1996, 1997, and 

again in 1998. All 1995 measures were deemed valid, as were the 1996 GPAB data obtained on the 

Controls. 

Due to extraneous factors, a major portion of the 1996 ANAMUKR data obtained from the Controls was 

not valid. As a result, these data were not included in analyses of the 1996 ANAMUKR results. 

Instead, all analyses of the 1996 ANAMUKR performances of the exposure groups relative to the 

Controls were based on the 1995 Control data. Further, the 1996 GPAB data from the Controls were 

virtually identical to those obtained in 1995. 

22

Therefore, the same procedure was used in analyses of the 1996 OP AB performances of the exposure 

groups relative to those of the Controls; i.e., 1995 Control data were also used for these comparisons. 

For the same reasons, similar analyses were performed on the 1997 and 1998 data. However, for 

purposes of data analyses based on 4-year averages, all data from all groups were used. 

23

SECTION 3 

RESULTS 

3.1 OVERVIEW OF 4-YEAR RESULTS: 1995-1998. 

The ANAM data files were first consolidated in a computerized spreadsheet and then inspected for 

completeness and invalid data. Invalid data were defined as "premature responses" which occurred in 

less than 100 ms, and/or an inordinate number of lapses which indicated that the participant did not 

understand the instructions for a test prior to administration. This initial screening resulted in unequal 

numbers of participants associated with each test, but it ensured that the preliminary data presented 

herein were derived from complete and valid test administrations. 

An ANOV A revealed a significant age difference among the groups. Subsequent Scheffe tests 

indicated that this was due to the higher ages of Group AF, as this group differed significantly from each 

of the others. No other significant differences in age were observed. Gender composition of the four 

groups also significantly differed, as revealed by results from a Chi-square test. This is most likely a 

result of the higher number of females in GroupAG, as none of the other groups differed from each 

other: all were predominantly male. Possible differences between the ACs and AEs on the demographic 

variables were of particular concern, but none were revealed. 

Analyses of 4-year averages for all groups on all measures were performed so as to obtain an overview 

of how the exposure groups performed relative to the controls on the physical and cognitive measures. 

Multivariate analyses of variance (MANOVAs) revealed significant differences among the 4 groups on 

all measures; further, pairwise comparisons indicated that with only a few exceptions, the average levels 

of performance of the exposure groups were significantly lower than those of the controls. 

Graphic illustrations of actual performance levels of the exposure groups on each task across the 4 years 

are presented in Figures 3-30 through 3-53, in the section describing the 1998 retest. In each figure, the 

4-year averaged performance level of the controls is used as a referent; it is denoted by a dotted line 

(typically across the top) on each figure. 

3.1.1 GPAB. 

The 4-year averages for the 4 groups on the physical tasks are presented in Table 3-1. The difference 

among the groups on BRODJMP was significant at the .01 level, while the other differences were 

significant at the .001 level. Pairwise comparisons of the 4-year averages revealed that with only one 

exception, all exposure groups performed at significantly lower levels than the controls. The 4-year 

averages for all groups on all GP AB tasks are graphically illustrated in Figures 3-1 through 3-4. 

24

Table 3-1. Four-year averaged performance on GPAB: physical tasks. 

GROUP~ AC AE AF AG 

TASK 

BRODJMP (meters) 1.59 1.33 1.43 1.50* 

CARRYWGT (meters) 51.61 30.96 40.45 42.51 

SQUATTHR (number) 66.66 21.24 40.35 48.41 

BALBEAM (meters) 22.19 15.96 20.26 19.13 

Note: *not significantly lower than controls. 

1.6 

en 1.5 

a:: 

w 

I- w .AC 

:E 1.4 

z 

.AE 

~ 

w 

:E 1.3 

.AF 

DAG 

1.2 

AC AE AF AG 

GROUP 

Figure 3-1. 4-year averaged performances on GPAB: BROADJUMP. 

25

en 

a:: 

w 

I W 

:E 

z 

« 

w 

:E 

AC AE AF AG 

GROUP 

Figure 3-2. 4-year averaged performances on GPAB: CARRYING WEIGHT. 

a: 

w 

m 

70 

60 

== ::;) 40 

Z 

Z 30 

en 

a: 

60 

50 

w 40 

I- 

w 

.AC 

:aE 30 

z 

.AE 

« 

w 20 .AF 

:aE 

DAG 

10 

0 

AC AE AF AG 

GROUP 

Figure 3-4. 4-year averaged performances on GPAB: BALANCE BEAM. 

3.1.2 ANAMUKR: Accuracy. 

The 4-year averages for the 4 groups on ANAMUKR-accuracy are presented in Table 3-2. Table 3-2 

represents the percentage of correct responses. All differences among groups were significant at the .001 

level. Pairwise comparisons of the 4-year averages revealed that, with only 3 exceptions, the exposure 

groups performed at significantly (most at .001) lower levels than the controls. Figures 3-5 through 3- 

12 graphically illustrate these findings. 

~.1.3 ANAMUKR: Efficiency. 

The 4-year averages for the 4 groups on ANAMUKR-efficiency are presented in Table 3-3. Table 3-3 

represents correct responses per minute. All differences among groups were significant at the .001 level. 

Pairwise comparisons of the 4-year averages revealed that, with only 1 exception, the exposure groups 

performed at significantly (most at .001) lower levels than the controls. Figures 3-13 through 3-21 

graphically illustrate these findings. 

27

3.1.4 ANAMUKR: Additional Measures. 

Performances averaged over 4 years on the tapping task (right, left), and on the sleepiness scale are 

presented in Table 3-4. Although all group differences were significant at the .001 level, the levels of 

the agricultural workers did not differ from those of the controls on any of these measures. The tapping 

rates for the other exposure groups were significantly (.001) lower that those of the controls, and levels 

of sleepiness were significantly higher for the Eliminators. These findings are graphically illustrated in 

Figures 3-22 through 3-24. 

Table 3-2. 4-Year averaged performances on ANAMUKR: Accuracy (Percentage of correct 

responses). 

GROUP -7 AC AE AF AG 

TASK 

2CH 97.36 92.76 92.91 93.98 

CDS 96.17 91.14 90.66 95.18* 

eDI 91.42 73.50 80.49 89.01 * 

-- -- 

eDD 89.24 72.35 78.99 85.32 

-- -- 

CPT 93.66 77.06 86.08 89.72 

-- -- 

DGS 87.96 73.00 81.32 80.40 

-- 

MSP 92.88 75.33 84.22 87.68 

-- 

SPD 88.21 81.18 83.96 91.94* 

Note: *not significantly lower than controls. Italicized, bolded, and underlined numbers are 

considered to suggest clinically meaningful impairment. 

28

Table 3-3. 4-Year averaged performances on ANAMUKR: Efficiency (Correct responses 

per minute). 


TASK 

SRT 163.51 111.58 111.67 147.24 

2CH 110.62 76.42 80.95 101.27 

-- -- 

CDS 39.58 24.22 26.12 34.33 

-- -- 

CDI 36.42 18.76 24.07 29.46 

-- -- 

CDD 38.19 20.77 26.46 31.57 

-- -- 

CPT 85.22 58.22 71.69 79.00 

-- 

DGS 36.19 24.28 30.54 29.30 

-- 

MSP 38.63 20.23 24.56 30.31 

-- -- 

SPD 27.67 19.11 22.81 26.38* 

-- 

Note: *not significantly lower than controls. Italicized, bolded, and underlined numbers are 


Table 3-4. 4-Year averaged performances on ANAMUKR: Additional measures. 


TASK 

TAP-R (Mean number of 57.57 

-- 

47.00 49.20 55.23* 

-- 

responses in 10 seconds) 

TAP-L (Mean number of 51.50 41.23 42.63 48.70* 

-- -- 

responses in 10 seconds) 

SLP** (Scores from 1-7) 1.63 2.46 1.77* 1.69* 

-- 

Note: *not significantly different than controls. 

**higher score = more sleepy. Italicized, bolded, and underlined numbers are 


29

3.1.5 A Clinical Neuropsychological Interpretation of Chernobyl-ANAM data. 

Overall results from the 4-year averages of ANAM test results, presented in Tables 3-2 through 3-4, 

provide evidence of clinically meaningful neurocognitive impairment associated with the Eliminator and 

Forestry groups. The Agricultural group was generally comparable to the Control group. Their test 

performance did not reveal any meaningful evidence of neuropsychological impairment, and scores were 

generally within normal limits. 

With respect to the Eliminator and Forestry groups, it appears that they have clinically significant 

neuropsychological deficits in several areas. These include deficits in the ability to sustain high levels 

of attention/concentration, to encode new information (i.e., learning ability), working memory (i.e., the 

ability to hold new information in memory for short periods of time), mental flexibility (i.e., the ability 

to shift mental sets rapidly), and psychomotor speed. The Eliminators are suffering the most severe 

impairment of neurocognitive and psychomotor abilities. Their test scores revealed impairment of 

mental power (the ability to produce correct responses to test items) and neurocognitive efficiency (the 

ability to do both quickly and accurately). In this study mental power was indexed by using percent 

accuracy scores, which are presented in Table 3-2. Neurocognitive efficiency was indexed by the 

ANAM "thruput" score, which literally translates to "number of correct responses per minute." These 

data are presented in Table 3-3. Psychomotor speed data are presented in Table 3-4. These data 

represent the average number of "Finger Taps" a subject is able to make on a mouse key during 5 

consecutive 10 second response trials. Finally, each subject's fatigue level was measured by a 

Sleepscale score that ranges from 1 =very alert and energetic to 7=sleep onset soon. A summary of 

clinically meaningful deficits for the eliminators is presented below. 

3.1.5.1 Learning and Memory. The Code Substitution, Learning, Immediate, and Delayed recall 

subtests were the primary instruments for assessing a subject's ability to learn and retain new 

information. This test entails having the subjects learn 8 pairs of associated digits and symbols. 

Following the learning trial, subjects are required to demonstrate the ability to remember the correct 

pairings immediately and then after a delayed time interval of approximately 20 minutes. Memory 

scores for immediate and delayed recall trials for Eliminators were 73 & 72% respectively as compared 

to 91 & 89% for the controls. These results suggest that the Eliminators have an impaired ability to 

encode, i.e., learn and store new information in short-term memory (e.g., CDI=73%). Although their 

learning ability is impaired, they are still able to learn and store new information to a certain degree. 

Further, they are able to retain and retrieve (i.e., remember) the newly learned information over 

meaningful time intervals. This is indicated by a delayed recall score (i.e., CDD=72%) that nearly 

matches the immediate recall score (i.e., CDI=73%). These results indicate that they do NOT have a 

rapid rate of forgetting over retention intervals, as would be the case in Alzheimer's and Alcohol 

Korsakoff s Dementias. The implication is that observed impairments in neurocognitive and memory 

processes in this sample are NOT a result of chronic alcohol abuse or an Alzheimer's-like eNS 

disorders. 

3.1.5.2 Attention/Working Memory. The Digit Set Comparison Test (DGS) was the primary 

ANAM subtest used to assess attention and working memory. It is comparable to the traditional 

Wechsler Adult Intelligence Scale Digit Span subtest. The DGS requires the subject to remember a 

series of numbers for a few seconds and then determine if a comparison sample is the same or different. 

30

Results from this test indicated that the Eliminators were meaningfully impaired with respect to both 

percent accuracy (power) and efficiency measures. For example, the Eliminators had an averaged 

accuracy score of73% as compared to a Control group score of 87%. Further, the Eliminators had an 

averaged efficiency score of24% vs. a score of36% for the Controls. These results suggest that the 

Eliminators are experiencing significant difficulties with the ability to attend to and retain information 

that is not personally meaningful; even for brief periods of time. 

3.1.5.3 Mental FlexibilitylExecutive Function. The Continuous Performance Task was the primary 

ANAM subtest used for assessing the ability to sustain high levels of concentration and rapidly shift 

mental sets. These are important "executive" functions that relate to frontal lobe functions. The test 

requires the subject to rapidly determine if a letter that is displayed on the computer screen is the "same 

or different" from the letter presented immediately before. Results from this test indicate that the 

Eliminators have serious deficits regarding the ability to sustain concentration and exercise mental 

flexibility. For example, the Eliminator's accuracy scores were 77% vs. Control's score of93%. 

Further, the Eliminators' efficiency scores were 52 vs. the Control's scores of85. 

3.1.5.4 Level of Subjective Energy. The ANAM Sleep Scale was used to assess the subjects' level of 

fatigue ... i.e., how energized or tired did they feel on a 1-7 scale. The results indicate that the 

Eliminators felt slightly more tired than the other groups ... however, the difference was barely 1 point 

higher (i.e., the Eliminator average was 2.46). This means that they did not really feel tired. This 

suggests that observed attention and memory deficits were not due to fatigue. 

3.1.5.5 Psychomotor Ability. The ANAM Finger Tapping Test was implemented as the primary test of 

psychomotor speed. The test requires the subject to "tap" on a mouse key as fast as possible during ten 

second test trials. The outcome measure is the average number of taps for a ten-second interval. The 

ANAM test results on this subtest revealed that the Eliminators were much slower that the other groups. 

For example, their averaged tapping scores for RT and LT index finger tapping were 47 & 41 vs. the 

Control group scores of 57 & 51. Since their subjective level of fatigue was minimal, results suggest 

that this psychomotor slowing was not due to being tired. 

3.1.5.6 Final Conclusion. The overall results that include loss of mental power and cognitive efficiency 

and psychomotor slowing strongly suggest impaired brain function. The pattern of impairment is similar 

to that commonly associated with white matter disease (white matter disease effects the myelin sheaths 

as a part of neurological functioning). 

3.2 CORRELATIONS BETWEEN DOSAGE OF RADIATION AND 4-YEAR AVERAGED 

PERFORMANCE LEVELS. 

For each participant in the study, the level of radiation exposure was obtained from medical records. The 

original dosage presented on Table 2-1, on page 21, included all subjects. However, our statement of 

work specifies that the government is interested in low dosage defined as below 70 rads. Therefore we 

eliminated from Table 2-1 all dosages higher than 70 rads. The following reflects only those dosages 

less than 70 rads. Mean dosage (and standard deviation) in rads for the four groups were as follows: AC: 

0.00 (.OO);AE: 43.41 (19.82);AF: 12.61 (2.10); andAG: 8.81 (5.63). 

31

Since the Controls received virtually no radiation, correlations were based only on participants in the 3 

combined exposed groups for whom measures on all tasks were obtained for all 4 years of testing, 

excluding Eliminators receiving over 70 rads. The 4-year average for each exposed individual on each 

measure was calculated, and these were used to obtain Pearson correlations between each measure and 

dosage of radiation. In addition, standard multiple regressions were used to calculate the correlations 

between combined groups of tasks and dosage. The results of these analyses are presented in Table 3-5. 

Table 3-5. Correlations between dosage of radiation and performance levels. 

TASK CORRELATION SIGNIFICANCE 

GPAB .70 .001 

BRODJMP -.18 -- 

CARRYWGT -.34 .01 

SQUATTHR -.55 .001 

BALBEAM -.56 .001 

ANAMUKR:ACCURACY .71 .001 

2CH -.07 -- 

CDS -.38 .01 

CDI -.62 .001 

CDD -.56 .001 

CPT -.47 .001 

DGS -.35 .01 

MSP -.54 .001 

SPD -.52 .001 

ANAMUKR: EFFICIENCY .68 .001 

SRT -.25 -- 

2CH -.44 .001 

CDS -.37 .01 

CDI -.50 .001 

CDD -.50 .001 

CPT -.48 .001 

DGS -.49 .001 

MSP -.45 .001 

SPD -.51 .001 

ANAMUKR:OTHERTASKS 

TPR -.34 .01 

TPL -.37 .01 

SLP .60 .001 

Significant correlations were revealed for 21 of the 24 measures. The only tasks not significantly 

correlated with dosage were broad jump, simple reaction time, and 2-choice accuracy. 

32

Unlike the other GP AB tasks, broadjump does not require sustained energy; therefore, it is not surprising 

that it is not related to dose level. The two ANAMUKR measures not related to dose are the simplest of 

all the tasks, requiring little effort to complete. Thus one would not expect them to be related to levels 

of radiation. Although tapping is also relatively simplistic, it requires fine motor coordination. Such 

coordination is reflective not only of integrity of the cerebral precentral gyri, but also of the cerebellum. 

Involvement of either of these areas would compromise the ability to perform this task. 

33

I 

U 

w 

a: 

a: 

o 

u 

eft. 

z 

« 

w 

~ 

lilAC 

.AE 

EAF 

DAG 

AC AE AF AG 

GROUP 

Figure 3-5. 4-year averaged performances on ANAMUKR: 2CH-ACC. 

I 

U 

w 

a: 

a: 

o 

u 

eft. 

z 

« 

w 

~ 

iliAC 

.AE 

rmAF 

DAG 

AC AE AF AG 

GROUP 

Figure 3-6. 4-year averaged performances on ANAMUKR: CDS-ACC. 

34

95 

I- 0 90 

w 

a: 

a: 85 

0 

.AC 

0 ~ 

80 .AE 

Z 

.AF 

« 

w 75 

DAG 

:l5 

70 

AC AE AF AG 

GROUP 

Figure 3-7. 4-year averaged performances on ANAMUKR: CDI-ACC. 

90 

I- 0 

W 85 

a: 

a: 

0 

iliAC 

80 

0 ~ 

.AE 

Z 

.AF 

« 

w 75 

DAG 

:l5 

70 

AC AE AF AG 

GROUP 

Figure 3-8. 4-year averaged performances on ANAMUKR: CDD-ACe. 

35

.... 

o 

w 

c:: 

c:: 

o 

o 

'ifz 

« 

w 

::2E 

I1AC 

IIAE 

I!!I AF 

DAG 

AC AE AF 

GROUP 

AG 

Figure 3-11. 4-year averaged performances on ANAMUKR: MSP-ACC. 

95 

90 

85 

80 

75 

&lAC 

I!IAE 

IIAF 

DAG 

70 

AC AE AF 

GROUP 

AG 

Figure 3-12. 4-year averaged performances on ANAMUKR: SPD-ACC. 

36

170 

c 160 . Ē 

....... 150 a: 

a: 

0 140 .AC 

() 

=1:1: 130 .AE 

Z 

« 120 

IIAF 

w 

DAG 

~ 110 

100 

AC AE AF AG 

GROUP 

Figure 3-13. 4-year averaged performances on ANAMUKR: SRT -EFF. 

c 

E 

a: 

a: 

o 

() 

=1:1: 

Z 

« 

W 

~ 

AC AE AF AG 

GROUP 

Figure 3-14. 4-year averaged performances on ANAMUKR: 2CH-EFF. 

37

40 

t: 35 

. Ē 30 

...... 

a:: 

a:: 25 

0 

20 

iliAC 

:#: .AE 

Z 

15 

Pfl AF 

AC AE AF AG 

GROUP 

Figure 3-17. 4-year averaged performances on ANAMUKR: CDD-EFF. 

c .- 

.E 

CC 

CC 

o 

:t:I:: 

Z 

40 

c 35 

. Ē '- 30 

a: 

a: 25 

0 

(.) 20 

:t:I: 

Z 

15 

« 10 

w 

~ 

5 

0 

AC AE AF AG 

GROUP 

lilAC 

I!!!IAE 

fmAF 

DAG 

Figure 3-19. 4-year averaged performances on ANAMUKR: DGS-EFF. 

c 

E 

a: 

a: 

o 

(.) 

:t:I: 

Z 

« 

W 

~ 

iliAC 

.AE 

I1AF 

DAG 

AC AE AF 

GROUP 

AG 

Figure 3-20. 4-year averaged performances on ANAMUKR: MSP-EFF. 

40

30 

c 25 

E 

-a:: 20 

a: 

0 .AC 

0 

=I*: 

.AE 

Z 10 

« 

.AF 

w 

5 

DAG 

== 

0 

AC AE AF AG 

GROUP 

Figure 3-21. 4-year averaged performances on ANAMUKR: SPD-EFF. 

iliAC 

II AE 

.AF 

DAG 

AC AE AF 

GROUP 

AG 

Figure 3-22. 4-year averaged performances on ANAMUKR: TAPPING-RIGHT. 

41

en 

Q. 

« 

l- 

#: 

Z 

« 

w 

:E 

liAC 

.AE 

IIAF 

DAG 

AC AE AF AG 

GROUP 

Figure 3-23. 4-year averaged performance on ANAMUKR: TAPPING LEFT. 

en 

w 

a: 

o 

en 

z 

« 

w 

:E 

iliAC 

.AE 

IlAF 

DAG 

AC AE AF 

GROUP 

AG 

Figure 3-24. 4-year averaged scores on ANAMUKR: SLEEP SCALE. 

42

3.3 RESULTS OF 1995 INITIAL TEST SESSION. 

Because perfonnance on the tasks within and among the test batteries is indexed in a variety of different 

ways, a measure was required which would describe comparisons between the exposure groups and the 

control group in similar tenns across the tasks and test batteries. This measure consisted of determining 

the proportion of the control group's mean on a given task equivalent to an exposure group's mean on 

that task (i.e. mean-E / mean-C). The complement of this proportion (I-prop.) reflects the proportional 

difference between the means of the two groups on the task, and when multiplied by 100 it reflects the 

percent difference between the means of the two groups. The mean % difference for a given test battery 

was then calculated by obtaining the mean of the % differences for all the tasks in that battery. In all four 

cases the mean of a given exposure group on any given task was either equal to, or (most typically) less, 

than that of the control group. Therefore, the mean differences of all exposure groups on all test 

batteries reflected decrements in perfonnance, as compared to the control group. Thus it seemed 

appropriate to describe the composite results of the 1995 test session in tenns of mean % decrements as 

shown by the exposure groups relative to the control group, on all test batteries. These are presented in 

Table 3-6 and graphically in Figure 3-25. 

Table 3-6. Mean % performance decrement for exposure groups relative to 

controls-1995. 

GROUP 

GPAB 

ANAMUKR-ACC 

ANAMUKR-EFF 

AE 

AF 

AG 

27.00 

16.88 

12.49 

11.00 

7.35 

3.09 

40.52 

23.91 

22.51 

60 

50 

40 

30 

20 

10 

o +==~_--lio 

GPAB 

ANAM-ACC 

ANAM-EFF 

DAG 

IIAF 

.AE 

GROUP 

Figure 3-25. Mean % Decrement for Exposure Groups Relative to Controls-1995. 

43

All of these decrements were significant, as evidenced by the results of multivariate analyses of variance 

(MANOVAs) described later. As dramatically illustrated in Figure 3- 25, the Eliminators were most 

drastically affected by their exposure to the radiation in the power station, on all test batteries. The other 

groups were also affected (although to a lesser degree) by the radiation in the Chemobyl region and 

showed sizeable decrements in performance on the test batteries as well. 

3.3.1 GPAB. 

Means and standard deviations for the groups on the GPAB are presented in Table 3-7. 

Table 3-7. GPAB: 1995 Means (and standard deviations) for the four groups. 


TASK 

BROADJMP 1.57 (.18) 1.40 (.17) 1.39 (.16) 1.43 (.45) 

(meter) 

CARRYWGT 43.42 (3.85) 36.61 (7.24) 37.17 (6.09) 40.58 (5.57) 

(meter) 

SQUATTHR 62.65 (10.01) 22.42 (8.78) 35.52 (12.72) 44.10 (17.39) 

(number) 

BALBEAM 19.61 (1.19) 16.22 (2.30) 19.93 (1.51) 18.65 (2.22) 

(meter) 

The results of a MANOV A (Table 3-28) revealed a significant difference on the composite 

MEASURES. Univariate ANOVAs and post-hoc Dunnett tests, the results of which are presented in 

Tables 3-8 and 3-9, indicated that theAEs were significantly impaired on all 4 tasks as compared to the 

ACs. TheAFs were significantly lower than theACs on BROADJMP, CARRYWGT and 

SQUATTHR, while theAGs were significantly lower on SQUATTHR only. Interestingly, theAFs 

performed somewhat better than the A Cs on BALBEAM; this may well have resulted from their 

training in forestry work, including balancing on logs. 

44

Table 3-8. GP AB: Results of MANOV A and univariate ANOV AS. 

TASK F p< 

COMP* 21.10 .001 

BROADJMP 2.92 .05 

CARRYWGT 9.33 .001 

SQUATTHR 59.65 .001 

BALBEAM 26.53 .001 

Note: *Wilks' Lambda = .21 

Table 3-9. GP AB: Groups significantly lower on physical abilities tasks than AC. 

TASK GROUP p< 

BRODJMP AE,AF .05, .05 

CARRYWGT AE,AF .001, .001 

SQUATTHR AE,AF,AG .001, .001, .001 

BALBEAM AE .001 

A discriminant function analysis, the results of which are presented in Table 3-10, indicated that the 

GP AB is extremely sensitive to the effects of exposure to ionizing radiation: 98.51 % of ACs and AEs 

were correctly classified. Table 3-10 represents numbers of subjects. 

Correlational analyses were performed for dosage of radiation and measures of physical performance for 

theAEs. Significant negative correlations ofrads with performance on BRODJMP and SQUATTHR 

(both rs>-.35, ps

Subsequent analyses included comparisons of performance scores of females and males as a function of 

non-exposure or exposure to radiation. The control group included only 7 females; thus 7 males from 

this group were selected for comparison purposes. The 3 exposed groups included a total of 17 females, 

so 17 males (matching the numbers offemales from each of these groups) were selected. Selection of 

males was not altogether random, since an attempt was made to ensure equivalence of ages between 

females and males. Means and standard deviations for the groups resulting from the combination of 

these variables are presented in Table 3-11, for each of the 4 physical tasks. The results of a MANOVA 

and univariate tests are presented in Table 3-12. MANOVA creates a composite measures based upon 

the corrolation relationship between individual measures. 

Significant differences on the composite measures were revealed for the main effects of gender and 

exposure; however, the multivariate interaction was not significant. The univariate tests indicated that 

overall, males performed significantly better than females on BRODJMP, CARRYWGT and 

SQUATTHR, but not on BALBEAM. Further, the controls were significantly better than the exposed 

people on SQUATTHR and BALBEAM. None of the univariate interactions were significant, 

however, indicating that the magnitudes of the female-male differences did not differ as a function of 

exposure to radiation. 

Table 3-11. 

GPAB: Means (and standard deviations) for females and males, 

either not exposed (controls) or exposed to radiation. 

TASK FEMALES- FEMALES- MALES- MALES- 

CONTROL EXPOSED CONTROL EXPOSED 

BROADJMP 1.36 (.11) 1.24 (.31) 1.72 (.13) 1.58 .37) 

(meter) 

CARRYWGT 38.14 (2.04) 33.88 (6.38) 44.71 (3.25) 39.88 (7.27) 

(meter) 

SQUATTHR 46.86 (5.49) 36.41 (15.12) 66.29 (4.19) 44.82 (22.55) 

(number) 

BALBEAM 19.21 (.91) 17.19 (1.91) 20.21 (1.43) 18.44 (2.79) 

(meter) 

46

Table 3-12. 

GP AB: MANOVA and univariate tests for females and males, 

either exposed or not exposed (controls) to radiation. 

SOURCE 

TASK 

F p< 

FEMALE- 

MALE 

CONTROL- 

EXPOSED 

INTERACTION 

COMP* 

BRODJMP 

CARRYWGT 

SQUATTHR 

BALBEAM 

COMP** 

BRODJMP 

CARRYWGT 

SQUATTHR 

BALBEAM 

COMP*** 

BRODJMP 

CARRYWGT 

SQUATTHR 

BALBEAM 

4.28 .01 

13.42 .001 

5.05 .05 

7.00 .01 

2.76 

3.34 .05 

1.59 

1.78 

9.19 .01 

7.85 .01 

1.25 

.002 

1.44 

1.10 

.03 

Notes: * Wilks' Lambda = .71 

** Wilks' Lambda = .75 

*** Wilks' Lambda = .89 


Means and standard deviations of accuracy scores for the 9 ANAM tasks are presented in Table 3-13. 

Since all scores on SRT were 100%, this task was not included in data analyses. The results of a 

MANDV A revealed a significant difference on the composite measures. Univariate ANOV As and posthoc 

Dunnett tests indicated that these results were primarily due to the AEs being significantly less 

accurate than the ACs on 6 of the tasks: CDS, CDI, CDD, CPT, DGS, and MSP. These data indicate 

that the AEs are experiencing remarkable decrements in short-term (working) memory, as 5 of these 

tasks assess integrity of this cognitive skill. TheAFs performed less accurately than theACs on 5 tasks: 

2CH, CDI, CDD, DGS, and MSP. Apparently, they are also experiencing some difficulties in shortterm 

memory. 

The A Gs were significantly less accurate on 2 tasks: DGS and MSP. Thus all the exposed groups are 

being affected by the ionizing radiation in varying degrees, as compared to the control group. The 

results of these analyses are presented in Tables 3-14 and 3-15. 

47

Table 3-13. Accuracy (% Correct): 1995 Means (and Standard Deviations). 


TASK 

SRT 100.00 (0.00) 100.00 (0.00) 100.00 (0.00) 100.00 (0.00) 

2CH 97.52 (3.51) 96.94 (3.19) 94.10 (7.05) 96.97 (2.83) 

CDS 97.74 (1.84) 95.39 (3.77) 95.97 (3.55) 96.35 (3.05) 

CDI 97.84 (4.10) 83.69 (15.48) 87.90 (8.33) 93.32 (6.67) 

CDD 97.10 (3.42) 84.00 (16.84) 88.14 (9.23) 93.00 (7.75) 

CPT 96.19 (4.29) 76.28 (15.72) 90.55 (8.28) 91.74 (5.56) 

DGS 89.84 (8.52) 73.69 (13.46) 82.03 (11.28) 83.13 (7.96) 

MSP 98.42 (2.98) 84.83 (10.64) 82.03 (11.28) 89.71 (15.18) 

SPD 90.16 (5.24) 87.08 (8.97) 87.76 (8.41) 96.97 (2.83) 

Table 3-14. Accuracy: Results of MANOV A and Univariate ANOVAs. 

TASK F p< 

COMP* 7.81 .001 

2CH 3.67 .01 

CDS 3.25 .05 

CDI 12.74 .001 

CDD 9.14 .001 

CPT 26.03 .001 

DGS 12.98 .001 

MSP 13.35 .001 

SPD 13.55 .001 

Note: *Wilks' Lambda = .28 

48

Table 3-15. Accuracy: Groups significantly less accurate thanAC. 

TASK GROUP p< 

2CH AF .01 

CDS AE .01 

CDI AE,AF .001, .001 

CDD AE,AF .001, .01 

CPT AE .001 

DGS AE,AF,AG .001, .05, .05 

MSP AE,AF,AG .001, .001, .01 

SPD 

NONE 

The results of a discriminant function analysis for theACs andAEs are presented in Table 3-16. The 

high percentage of correct classification (93%) is mainly a result of all the ACs being correctly 

classified, as only 14% oftheAEs were misclassified. Nonetheless, accuracy of performance would 

appear to be a good indicator of the effects of the hazardous environment surrounding Chemobyl, and it 

is as sensitive to these effects as to those resulting from traumatic brain injury (Levinson & Reeves, 

1997) and stroke (Goldstone et al. 1995). 

Levels of exposure to varying dosages of radiation were obtained for each of the AEs. Mean dose was 

62.95 (SD=32.64); these ranged from 18 rads to 144 rads. No significant correlations between rads dose 

and accuracy on any of the tasks were found. 

Table 3-16. Accuracy: Results of discriminant function analysis for groups AC and AE. 

ACTUAL PREDICTED PREDICTED 

GROUP GROUP--AC GROUP--AE 

AC (numbers) 31 0 

AE (numbers) 5 31 

49

3.3.3 ANAMUKR: Efficiency. 

Means and standard deviations of efficiency scores for the 9 ANAM tasks are presented in Table 3-17. 

Scores on SRT were included in these analyses. Once again, the results of a MANOV A (Table 38) 

revealed a significant difference on the composite measures. Univariate ANOV As and post -hoc 

Dunnett tests, the results of which are presented in Tables 3-18 and 3-19, indicated that theAEs were 

significantly impaired on all 9 tasks as compared to the ACs. These effects appear to be extremely 

profound: the mean response time on SRT was greater than that of a group of 70-year olds (Goldstone, 

et al. 1995); in fact, a discriminant function analysis, the results of which are presented in Table 3-20, 

resulted in over 91 % correct classification oftheACs andAEs, with 34 of the 36AEs being correctly 

classified. The global impairment seen here is reminiscent of that observed in individuals who have 

suffered moderate traumatic brain injury (Levinson & Reeves, 1997). As with accuracy, the effects were 

not limited to theAEs. TheAFs were significantly less efficient than theACs on 7 tasks: SRT, CDS, 

CDI, CDD, CPT, DGS, and MSP. The efficiency oftheAGs was similarly affected; they were 

significantly less efficient than theACs on 6 tasks: CDS, CDI, CDD, CPT, DGS, and MSP. 

Since efficiency is partially based upon response speed, and since it was significantly lower for the AEs 

on all tasks (and for theAFs andAGs on most of them), a MANCOVA was performed in which SRT 

was entered as a covariate. It was believed that adjusting for differences on this primarily motor 

measure would yield a clearer assessment of CNS-processing of information. Although a multiple 

regression analysis indicated that SRT was significantly correlated with the other measures combined 

(Multiple R=.72,p

Table 3-17. Efficiency (Correct responses/min): 1995 means (and standard deviations). 


TASK 

SRT 167.26 (46.34) 117.78 (43.14) 131.76 (47.05) 147.71 (48.24) 

2CH 110.87 (25.49) 86.94 (27.06) 98.62 (32.97) 116.42 (19.63) 

CDS 49.42 (14.44) 28.64 (8.74) 31.52 (18.69) 38.58 (13.95) 

CDI 49.77 (15.39) 21.72 (8.44) 30.55 (17.67) 32.64 (12.13) 

CDD 52.61 (13.04) 27.89 (10.94) 35.69 (19.85) 38.29 (10.81) 

CPT 96.74 (19.68) 58.86 (16.88) 82.90 (22.63) 81.35 (18.53) 

DGS 44.65 (10.45) 26.17 (8.07) 35.14 (10.11) 29.94 (8.06) 

MSP 51.55 (16.77) 26.39 (10.49) 32.59 (19.14) 29.29 (11.20) 

SPD 32.61 (9.85) 20.33 (6.05) 32.14 (11.88) 26.03 (7.27) 

51

Table 3-18. 

Efficiency: Results ofMANOVA and univariate ANOVAs (and 

MANCOV A and univariate ANCOVAs with SRT as a covariate). 

TASK F p< 

COMP* 6.79 (6.51) .001 (.001) 

SRT 7.00 .001 

2CH 8.15 (4.58) .001 (.01) 

CDS 13.77 (6.55) .001 (.001) 

CDI 24.27 (16.14) .001 (.001) 

CDD 17.98 (11.29) .001 (.001) 

CPT 22.14 (14.77) .001 (.001) 

DGS 24.75 (18.46) .001 (.001) 

MSP 19.20 (13.55) .001 (.001) 

SPD 14.11 (11.42) .001 (.001) 

Notes: *Wilks' Lambda = .28 (.33) 

Table 3-19. Efficiency: Groups significantly less efficient thanAC. 

TASK GROUP p< 

SRT AE,AF .001, .01 

2CH AE .001 

CDS AE,AF,AG .001, .001, .01 

CDI AE,AF,AG .001, .001, .001 

CnD AE,AF,AG .001, .001, .001 

CPT AE,AF,AG .001, .01, .01 

DGS AE,AF,AG .001, .001, .001 

MSP AE,AF,AG .001, .001, .001 

SPD AE,AG .001, .01 

52

Table 3-20. Efficiency: Results of discriminate function analysis for groups AC and AE. 

ACTUAL PREDICTED PREDICTED 

GROUP GROUP--AC GROUP--AE 

AC (numbers) 27 4 

AE (numbers) 2 34 

Correlational analyses were also perfonned for dose of radiation and efficiency of perfonnance for the 

AEs. Unlike accuracy, significant negative correlations ofrads dose with efficiency on CDI, CDS, 

CPT, DGS, and MSP (all of which assess short-tenn memory), as well as with SRT (all 7s>-.36, all 

ps 4.36,ps < .01). Compared 

to the ACs, tapping rates for both hands were significantly lower for theAEs (.001) and for theAFs 

(.05); while levels of sleepiness were higher for theAEs (.001). 

3.3.5 GP AB-ANAMUKR: Correlational Analyses. 

Significant correlations were found between many of the measures on the GPAB; these are presented in 

Table 3-22, along with intercorrelations between perfonnance on the tasks comprising the GPAB and 

accuracy ofperfonnance on the tasks in ANAMUKR for the combined groups (N=127). 

53

Table 3-22. 

Pearson correlations between GP AB and ANAMUKR: Accuracy 

(N=127). 

TASK 

BRODJMP 

CARRYWGT 

SQUATTHR 

BALBEAM 

CDD 

CDI 

CDS 

CPT 

DGS 

MSP 

SPD 

SRT 

2CH 

BRODJMP 

.59** 

.62** 

-.04 

.01 

.14 

.02 

.18 

.15 

.29** 

-.06 

.03 

-.07 

CARRY SQUATTHR BALBEAM 

WGT 

.65** 

.14 .30** 

.11 .27** .19* 

.32** .37** .17 

.06 .11 .21* 

.21 * .48** .45** 

.14 .37** .26** 

.31 ** .41 ** .05 

.11 .08 .14 

-.08 .06 .09 

-.02 -.05 -.01 

Notes: 

* p

Table 3-23. Pearson correlations between GP AB and ANAMUKR: Efficiency 

(N=127). 

TASK 

BRODJMP 

CARRYWGT SQUATTHR BALBEAM 

CDD 

CDI 

CDS 

CPT 

DGS 

MSP 

SPD 

SRT 

2CH 

.04 

.21* 

.23* 

.28** 

.24** 

.27** 

.16 

.35** 

.10 

.07 .42** .26** 

.20* .54** .24** 

.19* .45** .16 

.19* .49** .40** 

.23** .48** .27** 

.29** .49** .22* 

.15 .31 ** .36** 

.21* .48** .13 

.11 .24** .23* 

Notes: 

* p

3.3.5.1 Age. Since ages of all 127 participants were obtained, correlations between it and all other 

variables were calculated. Significant negative correlations were observed between age and three of the 

GPAB tasks (BRODJMP, CARRYWGT, and SQUATTHR: all rs>.36, allps

GPAB ANAM-ACC ANAM-EFF 

GROUP 

Figure 3-26. Mean % Decrement for Exposure Groups-1996 Relative to Controls-1995. 

The data of Figure 3-26 are similar to those illustrated in Figure 3-25; however,except for Groups AF 

andAG on the GPAB, the magnitude of the 1996 differences between the exposure groups and the 

Controls is noticeably greater than those of 1995. In addition to declines in performance in groups 

significantly lower than the Controls in 1995, the performance of several groups not significantly lower 

than the Controls on certain tasks in 1995 declined to significant levels in 1996; these are listed in 

Table 3-46. 

Using a procedure similar to that for calculating mean percent decrement of the exposure groups relative 

to the Controls, the mean percent declines for each of these groups on each battery of tests were 

calculated by using 1995 levels of performance as a baseline by which to gauge 1996 levels (i.e., mean - 

1996/ mean-1995). These declines are presented in Table 3-27 and illustrated graphically in Figure 3- 

27. With the exception of Groups AF and AG on the GP AB, all groups showed significant declines as 

revealed by MANOV As, the results of which are presented in Table 3-28, on all test batteries from 1995 

to 1996. These findings indicate that both the physical (in the case of Group AE) and cognitive 

performance levels in these groups of individuals are worsening over time. 

57

Table 3-26. Groups not Significantly Lower than AC in 1995, but significantly lower in 1996. 

BATTERY: GROUP p< 

TASK 

GPAB 

BRODJUMP AE .001 

ANAM-ACCURACY 

2CH AE,AG .01, .01 

CDS AF .01 

CDI AG .001 

CDD AG .001 

SPD AE,AF .001, .01 

ANAM-EFFICIENCY 

SPD AF .001 

Table 3-27. Mean % Performance Decline for Exposure Groups: 1995-1996. 

GROUP GPAB ANAMUKR-ACC ANAMUKR-EFF 

AE 9.78 1.95 11.89 

AF 0.00 2.84 15.22 

AG 0.00 3.18 9.03 

58

GPAB ANAM-ACC ANAM-EFF 

GROUP 

Figure 3-27. Mean % decline for exposure groups: 1995-1996. 

Table 3-28. Significant multivariate declines by exposure groups. 

TEST GROUP WILKS' F p< 

BATTERY 

LAMBDA 

GPAB AE .39 11.95 .001 

ANAM-ACC AE .51 3.07 .01 

AF .42 3.49 .01 

AG .38 4.60 .01 

ANAM-EFF AE .39 4.39 .001 

AF .28 5.01 .001 

AG .32 5.32 .001 

59

3.6 SPECIFIC ASSESSMENTS OF DECLINES IN EXPOSURE GROUPS. 

Means and standard deviations for the exposure groups on the 1996 GP AB retest are presented in Table 

3-29. Similar data for accuracy, efficiency, and the additional measures performance on the 1996 

ANAMUKR retest are presented in Tables 3-30, through 3-32, respectively. Significant declines, as 

revealed by the results of univariate analyses comparing the 1996 levels of performance of the exposure 

groups on the various tasks in the test batteries to their 1995 levels, are presented in Table 3-33, along 

with the figure number from Figures 3-30 through 3-53 illustrating these declines. 

3.6.1 GPAB. 

Already significantly weaker than the ACs (1995), the AEs continued to decline on all measures of 

strength, including explosive, static and dynamic. These findings indicate that they are physically 

deteriorating, even 10 years following their exposure to the radiation in the power station. In contrast, 

the AFs and A Gs maintained their 1995 levels--perhaps because they are working in occupations 

requiring physical exertion and were therefore able to "stay in shape". 

TABLE 3-29. GPAB: 1996 means (and standard deviations) for the exposure groups. 

TASK AE AF AG 

·BRODJMP 1.30 (.28) 1.37 (.14) 1.41 (.42) 

CARRYWGT 30.56 (11.68) 38.89 (5.93) 41.48 (6.03) 

SQUATTHR 19.47 (7.44) 35.39 (11.97) 43.74 (17.07) 

BALBEAM 15.85 (3.16) 20.02 (1.76) 19.11 (1.20) 


All three exposure groups showed significant declines in accuracy of performance on a variety of the 

tasks. Most of these declines occurred in Groups AF and A G, on tasks assessing both attention and 

memory skills. Nonetheless, theAEs declined as well, on tasks requiring sustained attention. 

Surprisingly, all three groups showed significant declines on SPD, indicating a developing difficulty in 

skills related to assessment of spatial relations. 

60

Table 3-30. ANAMUKR Accuracy (% correct): 1996 means (and standard deviations) 

for the exposure groups. 

TASK AE AF AG 

SRT 100.00 (.00) 100.00 (.00) 100.00 (.00) 

2CH 94.26 (5.30) 94.79 (4.53) 93.90 (7.19) 

CDS 93.59 ( 3.66) 94.07 (3.21) 95.39 (3.32) 

CDI 79.85 (8.07) 82.96 (8.54) 88.71 (10.55) 

CDD 79.53 (9.96) 81.11 (6.79) 84.00 (10.75) 

CPT 83.24 (16.15) 89.11 (6.96) 91.61 (6.80) 

DGS 74.82 (10.26) 83.43 (7.03) 85.16 (8.56) 

MSP 80.94 (13.30) 81.18 (16.88) 89.58 (9.14) 

SPD 80.88 (7.83) 81.79 (6.56) 88.55 (6.73) 


As with accuracy, all exposure groups exhibited significant declines in efficiency of performance on the 

majority of the tasks. Most of these were the same tasks upon which their accuracy declined. Thus the 

declines in efficiency are not surprising, since efficiency is based in part on accuracy of performance. 

Nonetheless, it should be noted that none of the groups showed significant declines in SRT, and 

therefore the declines in efficiency cannot be explained by slowing of response speed (the other factor 

upon which efficiency of performance is based). The fmding that declines in efficiency for the most part 

accompanied corresponding declines in accuracy would indicate that these individuals are experiencing 

difficulty in processing cognitive information on tasks entailing attention, memory and spatial abilities. 

61

Table 3-31. 

ANAMUKR Efficiency (correct responses/min): 1996 means (and standard 

deviations) for the exposure groups. 

TASK AE AF AG 

SRT 117.74 (34.16) 136.93 (52.23) 139.00 (52.65) 

2ca 76.47 (29.60) 90.39 (31.89) 98.13 (33.56) 

CDS 23.35 (8.03) 31.79 (21.41) 30.45 (11.55) 

CDI 17.35 (5.71) 27.57 (16.19) 26.29 (11.48) 

CDD 20.74 (8.47) 25.50 (13.26) 29.13 (12.77) 

CPT 61.26 (1S.0S) 74.14 (24.49) 79.74 (27.67) 

DGS 25.26 (7.57) 29.39 (7.4S) 30.S7 (10.S7) 

MSP 19.52 (S.52) 21.82 (11.30) 31.81 (15.30) 

SPD 18.41 (7.22) 21.18 (9.29) 24.97 (9.53) 


Means and standard deviations for the exposure groups are shown in Table 3-32. 

Table 3-32. ANAMUKR Additional measures: 1996 means (and standard deviations). 

TASK AE AF AG 

TAP-R (mean n 42.82 (9.74) 48.351 (11.38) 55.58 (13.37) 

of responses in 10 

sec) 

TAP-L (mean n 37.97 (8.11) 41.00 (10.04) 49.30 (12.07) 

of responses in 10 

sec) 

SLP (scores from 2.59 (.71) 1.27 (.46) 1.46 (.59) 

1-7) 

No significant changes were observed from 1995 to 1996 in rates of tapping for either hand by any 

group; however, all 3 groups showed significant decreases in levels of sleepiness (all Fs > 4.29, allps 

< .05). 

62

Table 3-33 shows the significant declines in performance from 1995 to 1996 for the GPAB tasks, and for 

accuracy and efficiency on ANAMUKR; the figure number refers to the figure which specifically 

illustrates the decline. 

Table 3-33. Significant declines in performance by the exposure groups: 1995 to 1996. 

BATTERY: GROUP F p< FIG. 

TASK # 

GPAB 

BRODJUMP AE 8.29 .01 30 

CARRYING WGT AE 31.58 .001 31 

SQUATTHRUSTS AE 16.32 .001 32 

ANAM-ACCURACY 

2CH AE,AG 7.90, 4.37 .01, .05 34 

CDS AE,AF 5.71, 7.90 .05, .01 35 

CDI AF,AG 5.52, 5.52 .05, .05 36 

CDD AF,AG 10.30, 12.53 .01, .001 37 

SPD AE,AF, 10.05, 12.04, 35.40 .01, .01,001 41 

AG 


2CH AG 7.45 .01 43 

CDS AE,AG 8.01, 10.96 .01, .01 44 

CDI AE,AG 6.15, 7.90 .01, .01 45 

CDD AE,AF, 11.02, 4.62, 10.89 .01, .05, .01 46 

AG 

DGS AF 7.51 .01 48 

MSP AE,AF 10.30, 8.01 .01, .01 49 

SPD AF 14.67 .001 50 

63

3.7 RESULTS OF 1997 RETEST SESSION. 

Because of unavailability for testing resulting from relocation, illness, or death, 1997 data were obtained 

on 34 Eliminators, 20 Forestry workers, and 28 Agricultural workers. However, some of these had not 

been available for testing on the GPAB and/or ANAMUKR in 1996, so the 1996-97 comparisons were 

based on Ns as follows: GPAB-Eliminators-32, Forestry workers-l 8, Agricultural workers-28; 

ANAMUKR-Eliminators-30, Forestry workers-15, Agricultural workers-28. Because many of the 

original forestry workers were unavailable for testing in 1997, a new group of 13 foresters was tested. 

However, since this was their first year of testing, their data could not be included in the analyses. 

3.7.1 Global Assessments of Declines By Exposure Groups. 

Mean percent decrements in performance for the exposure groups on the 1997 retest relative to the 1995 

control data are presented in Table 3-34 and graphically illustrated in Figure 3-28. The data of Figure 3- 

28 are similar to those illustrated in Figures 3-25 and 3-26, except that decrements in ANAMUKR 

accuracy for Groups AE and AF are more pronounced. 

Table 3-34. Mean % performance decrement for expo groups-1997 relative to 

controls-1995. 

GROUP 

GPAB 

ANAMUKR-ACC 

ANAMUKR-EFF 

AE 35.97 

AF 9.93 

AG 11.60 

23.24 

14.96 

8.46 

50.54 

43.68 

34.57 

DAG 

ImAF 

.AE 

GPAB 

ANAM-ACC 

GROUP 

ANAM-EFF 

Figure 3-28. Mean % decrement for exposure groups-1997 relative to controls-199S. 

64

In addition to continuing declines exhibited by all the groups on at least some of the measures, the 

performance of Groups AF and A G not significantly lower than that of the Controls in 1995 or 1996 

declined to significant levels on several tasks in 1997. These are listed in Table 3-35. 

Table 3-35. Groups not significantly lower thanAC in 1995 or 1996, but significantly 

lower in 1997. 

BATTERY: GROUP p< 

TASK 

GPAB 

BALANCE BEAM AF .05 

ANAM-ACCURACY 

CDS AG .05 

CPT AF .001 


SRT 

AG 

.01 

2CH AF,AG .001, .05 

U sing a procedure similar to that for calculating mean percent decrement of the exposure groups relative 

to the Controls, the mean percent declines for each of these groups on each battery of tests were 

calculated by using 1995 levels of performance as a baseline by which to gauge 1997 levels (i.e., mean - 

1997/ mean-1995). These declines are presented in Table 3-36 and illustrated graphically in Figure 3- 

29. With the exception of Groups AF and A G on the GP AB, all groups showed significant declines as 

revealed by MANOVAs, the results of which are presented in Table 3-37, on all test batteries from 1995 

to 1997. These findings indicate that both the physical (in the case of Group AE) and cognitive 

performance levels in these groups of individuals are worsening over time. 

Mean percent declines in performance by the 3 exposure groups from 1996 to 1997 are presented in 

Table 3-36 and graphically illustrated in Figure 3- 29. As revealed by MANOV As, the decline of Group 

AE on the GP AB was significant, as were those of all 3 groups on ANAM accuracy. The declines in 

ANAM efficiency, although noteworthy, were not significant. 

65

Table 3-36. Mean % performance decline for exposure groups: 1996-1997. 

GROUP 

GPAB 

ANAMUKR-ACC 

ANAMUKR-EFF 

AE 3.00 

AF 0.00 

AG 0.00 

12.43 

8.44 

2.43 

5.20 

14.52 

6.41 

DAG 

IIAF 

.AE 

GPAB 

ANAM-ACC 

GROUP 

ANAM-EFF 

Figure 3-29. Mean % Decline for Exposure Groups: 1996-1997. 

These findings indicate that the physical performance of Group AE and the accuracy of performance of 

all 3 exposed groups are continuing to significantly decline. Although a MANOV A indicated that the 

decline of Group AE on efficiency was also significant, a MANCOV A using SRT as a covariate 

revealed that the significance was a result of slowing of reaction time only (see Table 3-41). The results 

of the MANOV As assessing the significance of the changes in performance for all 3 groups from 1996 

to 1997 are presented in Table 3-37. 

66

Table 3-37. Significant multivariate declines by exposure groups from 1996 to 1997. 

TEST BATTERY GROUP WILK'S F p< 

LAMBDA 

GPAB AE .52 6.48 .001 

ANAM-ACC AE .14 17.31 .001 

AF .09 8.62 .01 

AG .47 2.84 .05 

ANAM-EFF AE .39 3.63 .01* 

Note: *n.s. when analyzed via MANCOVA using SRT as a covariate. 

3.7.2 Specific Assessments of Declines in Exposure Groups. 

Means and standard deviations for the exposure groups on the 1997 GP AB retest are presented in Table 

3-38. Similar data for accuracy and efficiency of performance on the 1997 ANAMUKR retest, and for 

the additional ANAMUKR measures, are presented in Tables 3-39, through 3-41 respectively. Figures 

3-30 through 3-53 illustrate changes in performance for the three exposure groups from 1996 to 1997 (as 

well as from 1995-1996). On each figure, the 1995 control data are again included as a point of 

reference. Significant declines, as revealed by the results of univariate analyses comparing the 1997 

levels of the exposure groups on the various tasks in the test batteries to their 1996 levels, are presented 

in Table 3-42 along with the figure number from Figures 3-30 through 3-53 illustrating these declines. 

Table 3-38. GPAB: 1997 means (and standard deviations) for the exposure groups. 

TASK AE AF AG 

BRODJMP 1.24 (.28) 1.42 (.20) 1.40 (.44) 

(meter) 

CARRYWGT 28.32 (12.13) 42.05 (7.99) 41.00 (6.77) 

(meter) 

SQUATTHR 19.71 (8.39) 41.65 (13.81) 45.04 (19.30) 

(number) 

BALBEAM 15.76 (2.55) 20.95 (1.94) 19.25 (1.85) 

(meter) 

67

Table 3-39. 

ANAMUKR: Accuracy (% correct): 1997 means (and standard deviations) 

for the exposure groups. 

TASK AE AF AG 

SRT 100.00 (.00) 100.00 (.00) 100.00 (.00) 

2CH 89.86 (6.91) 91.45 (12.00) 94.07 (16.82) 

CDS 87.06 (8.50) 86.92 (9.18) 93.80 (3.91) 

CDI 65.63 (14.29) 73.36 (11.19) 88.13 (8.74) 

CDD 61.33 (10.82) 68.44 (11.91) 84.17 (9.39) 

CPT 73.03 (14.13) 77.06 (10.84) 86.88 (12.91) 

DGS 68.70 (11.67) 72.02 (10.66) 76.39 (12.35) 

MSP 64.79 (17.04) 80.33 (11.34) 85.78 (13.75) 

SPD 75.63 (10.22) 81.00 (9.12) 90.67 (8.48) 

Table 3-40. ANAMUKR: Efficiency (correct responses/min): 1997 means (and 

standard deviations) for the exposure groups. 

TASK AE AF AG 

SRT 94.69 (30.61) 103.23 (42.00) 133.87 (37.75) 

2CH 67.28 (32.01) 75.70 (30.03) 91.45 (31.32) 

CDS 23.89 (11.47) 20.03 (9.29) 30.44 (15.20) 

CDI 19.28 (12.97) 21.26 (10.27) 27.19 (12.11) 

CDD 18.55 (10.87) 25.00 (9.79) 27.13 (13.06) 

CPT 54.01 (20.43) 58.27 (13.87) 71.16 (19.45) 

DGS 23.00 (9.13) 26.12 (8.81) 25.80 (6.72) 

MSP 18.43 (12.29) 24.09 (11.13) 27.04 (12.65) 

SPD 20.66 (12.68) 20.16 (6.67) 24.63 (8.42) 

68

Table 3-41. ANAMUKR: Additional measures: 1997 means (and standard deviations). 

TASK AE AF AG 

TAP-R (mean n 42.97 (11.99) 51.20 (11.89) 57.47 (11.01) 

of responses in 

10 sec) 

TAP-L (mean n 37.81 (9.30) 46.16 (12.40) 51.43 (10.79) 


10 sec) 


1-7) 

3.7.3 GPAB. 

From 1996 to 1997, Group AE declined on measures of explosive and dynamic strength, reflecting their 

continuing physical deterioration (see Table 3-42). Conversely, GroupsAF andAG showed some 

improvement in physical abilities. 


As seen in Table 3-42, GroupAE showed significant declines in accuracy from 1996 to 1997 on all 

tasks. This global decline in accuracy is most likely reflecting a general deterioration of their 

neurocognitive abilities; in fact, their levels of performance are similar to those observed in individuals 

with moderate-severe traumatic brain injuries (TBIs) (Levinson & Reeves, 1997; Levinson, et al, 1998). 

Similar declines were observed in GroupAF, on 6 of8 tasks. Accuracy of performance of Group AG 

also declined on 3 tasks, although not as sharply as that of the other exposed groups. 


GroupAE showed a significant decline in efficiency on SRT only, and when this was entered into a 

MANCOV A as a covariate, the multivariate difference in efficiency revealed by the MANOV A (see 

Table 3-37) was no longer significant. Group AF also showed a significant slowing of reaction time, 

and this most likely explains their corresponding decline in efficiency on CPT. Group A G did not 

exhibit any apparent slowing of reaction time, however, and therefore the significant declines on CPT 

and DGS are more likely reflecting difficulty in sustaining attention. At this point in time, the 

performance of all exposed groups on all tasks is significantly lower than that of the Controls. 

69

Table 3-42. Significant declines in performance by the exposure groups: 1996 to 1997. 

BATTERY: GROUP F p< FIG. # 

TASK 

GPAB 

BROADJUMP AE 7.87 .01 30 

SQUATTHRUSTS AE 3.49 .05* 32 

ANAM-ACCURACY 

2CH AE 14.59 .001 34 

CDS AE,AF 18.77, 11.88 .001, .001 35 

CDI AE,AF 23.02, 14.43 .001, .001 36 

CDD AE,AF 69.38,14.21 .001, .001 37 

CPT AE,AF, 11.76, 18.07, .001, .001, .05 38 

AG 4.07 

DGS AE,AF, 6.32, 16.09, .01, .001, .001 39 

AG 15.49 

MSP AE,AF, 12.26, 9.29, .001, .01, .05 40 

AG 4.04 

SPD AE 5.98 .05 41 


SRT AE,AF 11.93, 7.24 .001, .01 42 

CPT AF,AG 5.56, 8.79 .05, .01 47 

DGS AG 13.89 .001 48 

Note: * I-tailed 


No significant declines in tapping rates for either hand were observed. GroupAE showed a significant 

(.001) decrease in levels of sleepiness, while Group AF showed a significant increase. 

3.8 RESULTS OF 1998 RETEST SESSION. 

For similar reasons described for the 1997 retest, 1998 data were obtained on 22 Eliminators, 21 

Forestry workers, and 29 Agricultural workers. Since some of these had not been available for testing 

on the GPAB and/or ANAMUKR in 1996, but were in 1997; therefore, the 1997-98 comparisons were 

based on Ns as follows: GPAB-Eliminators-22, Forestry workers-19, Agricultural workers-29; 

ANAMUKR-Eliminators-22, Forestry workers-20, Agricultural workers-29. Of the new group of 13 

foresters from 1997,8 were retested. Since this was only their second year of testing, their data were not 

included in the analyses. 

70

3.8.1 Assessments of Declines by Exposure Groups. 

Multivariate and univariate tests revealed no significant declines from 1997 levels on any measure in 

any of the test batteries, for any group. In fact, some of the groups showed significant increases in levels 

of performance compared to the 1995 control groups; these will be described in the relevant sections of 

this report. 

3.8.2 GPAB. 

Means and standard deviations on the 1998 retest are presented in Table 3-43. TheAEs showed 

significant increases in performance on CARRYWGT, SQUATTHR and BALBEAM (ps < .01), while 

theAGs improved on CARRYWGT (.05). 

Table 3-43. GP AB: 1998 means (and standard deviations) for the exposure groups. 

TASK AE AF AG 

BRODJMP 1.29 (.16) 1.42 (.24) 1.51 (.40) 

(meter) 

CARRYWGT 31.82 (12.40) 41.62 (8.56) 44.58 (6.98) 

(meter) 

SQUATTHR 23.91 (7.51) 40.52 (12.02) 51.35 (19.01) 

(number) 

BALBEAM 17.36 (1.59) 19.93 (1.88) 19.96 (2.81) 

(meter) 


Means and standard deviations for the 1998 accuracy retest are presented in Table 3-44. Although some 

improvements over the 1997 levels were observed in some groups, none of these was significant. 

71

Table 3-44. 

ANAMUKR: Accuracy (% correct): 1998 means (and standard 

deviations for the Exposure Groups. 

TASK AE AF AG 

SRT 100.00 (.00) 100.00 (.00) 100.00 

2CH 89.90 (5.56) 88.89 (10.25) 92.86 

CDS 89.70 (6.62) 88.20 (2.75) 94.99 

CDI 63.83 (8.99) 74.06 (11.52) 85.29 

CDD 64.20 (11.92) 72.19 (10.23) 79.53 

CPT 73.07 (16.74) 82.29 (9.35) 87.96 

DGS 68.75 (10.11) 72.29 (9.59) 76.01 

MSP 67.57 (16.75) 79.63 (8.39) 85.63 

SPD 82.05 (10.98) 83.25 (5.45) 91.72 

(.00) 

(14.32) 

(3.72) 

(10.92) 

(13.56) 

(11.58) 

(13.94) 

(17.57) 

(5.39) 


Means and standard deviations in efficiency are presented in Table 3-45 compared to the 1995 control 

group. Significant (.05) increases in efficiency of performance were made by the A Gs on SRT and 

SPD. Although other increases in levels of performance were observed (see, e.g., Figures 3-43,3-47, 

and 3-49), none were significant. 

72

Table 3-45. ANAMUKR: Efficiency (correct responses/min) : 1998 means 

(and standard deviations) for the exposure groups. 

TASK AE AF AG 

SRT 96.72 (38.46) 86.39 (20.87) 150.20 

2CH 60.31 (30.64) 63.69 (22.84) 100.70 

CDS 18.62 (7.61) 21.56 (13.28) 33.82 

CDI 15.97 (9.63) 18.32 (13.11) 28.06 

CDD 14.25 (6.10) 16.58 (8.19) 29.02 

CPT 53.60 (16.77) 63.26 (13.80) 76.97 

DGS 19.64 (5.66) 22.73 (8.38) 27.36 

MSP 15.38 (9.89) 19.32 (13.10) 30.10 

SPD 16.32 (9.07) 19.45 (9.27) 28.15 

(31.82) 

(26.94) 

(11.59) 

(12.38) 

(10.22) 

(15.54) 

(7.04) 

(13.31) 

(6.85) 

3.8.5 ANAMUKR: Additional measures. 

Means and standard deviations for these are presented in Table 3-46. GroupAG showed a significant 

increase (.05) in TAP-L. 

Table 3-46. ANAMUKR: Additional Measures: 1998 Means (and Standard Deviations). 

TASK AE AF AG 

TAP-R (mean n 47.64 (6.48) 49.65 (6.09) 59.48 (10.29) 


10 sec) 

TAP-L (mean n 40.80 (5.36) 43.53 (6.62) 53.84 (10.14) 


10 sec) 


1-7) 

Figures 3-30 through 3-53 graphically illustrate yearly changes in perfonnance from 1995-1998 on all 

measures. The 4-year averaged levels of the Controls are represented as a straight dotted line (typically 

across the top) on each figure. Each graph represents the mean and the standard deviation are identified 

in their appropriate tables. 

73

BROADJUMP 

.......................................... 

1.6.,...---------------------. 

C/) 

a::: 

w 

I- W 

~ 

c: 

CO 

Q) 

~ 

1.5 

1.4 

1.3 

..--.-....... ......-.. --.....- 

. 

~-~.------.------ 

GROUP 

• AE 

..........------r-- 

• AF 

1.2 

... AG 

~-----_r------~-------~ 

1995 1996 1997 1998 

YEARS 

Figure 3-30. Mean performance on GPAB: BROADJUMP. 

fQ 

W 

tu 

~ 

c: 

CO 

Q) 

~ 

CARRYING WEIGHT 

55~-------------------------~ 

50 

45 

40 

35 

30 

.......................................... 

-- 

..----.......--.. ........ -.-......~..,...-.--- 

.... ---.--- 

------.. -- 

• AE 

25 

• AF 

20 

... AG 

1995 1996 1997 1998 

YEARS 

Figure 3-31. Mean performance on GP AB: CARRYING WEIGHT. 

74

SQUATTHRUSTS 

70~------------------------------------. 

60 

.. - -- .. -.-- 

~ 50 __ - 

W IP------..------- --- 

III 

2 40 ..------..--- -- ... ~ 

c: 

m 

~ 30 

GROUP 

~ -_ .... ---- .. 

10~ __________ ~ __________ ~~ ________ ~ .... AG 

1995 1996 1997 1998 

YEARS 

Figure 3-32. Mean performance on GPAB: SQUAT THRUSTS. 

BALANCE BEAM 

23~------------------------------------, 

~ 

UJ 

tu 

:2 

c: 

m 

Q) 

:2 

22 •••••••••••••••••••••••••••••••••••••••••• 

21 

20 

19 

18 

17 

-_ .......... - 

_..._- 

............. ....- .... 

-- 

....... _._._.JJr--_.-- 

.---- 

• AE 

16 

• AF 

15 

.... AG 

~----------~-----------,------------4 

1995 1996 1997 1998 

YEARS 

Figure 3-33. Mean performance on GP AB: BALANCE BEAM. 

75

2CH-ACC 

.......................................... 

98~------------------------------------~ 

96 ~ 

• 

0::: 94 . 

0::: 

"" 

t; 

" • 

W 

---~- 

0 

() 

COI-ACC 

100~------------------------------------------------------~ 

.. -: :-:-. .. ....,.............................. --- - 

. 

90 .... -~- 

.... 

.. 

........ .......... 80 

" .-.- 

~ 

" 

" " '-... GROUP 

----- 

70 e AE 

--........................................ -l. AF 

60~ _______________ ~ __________________ ~ _______________ ~ ... AG 

1995 1996 1997 1998 

YEARS 

Figure 3-36. Mean performance on ANAMUKR: CDI-ACC. 

COO-ACC 

100~------------------------------------------------------~ 

b 

w 

c:: 

c:: 

8 

::R 0 

~ 

90 

80 

70 

60 

..... ...,"="-,.,-....... 

•••••••••...••••.•••••.••.•••••••• 

, . .........." ------ ....... 

. ....... 

1~--------~1L .-. ......... 

' .... , .... 

-- 

' ... --- 

50 

1995 1996 1997 

-.._---,eAE 

1998 

• AF 

... AG 

YEARS 

Figure 3-37. Mean performance on ANAMUKR: CDD-ACC. 

77

CPT-ACC 

100~----------------------------------~ 

90 

80 

......................................... 

...... 

~.- 

.... - .....,' .. 

" 

' .., 

, , 

.......-._.- 

" , , 

, --' .... - 

lI'-- 

• AE 

• AF 

70~ __________ ~ __________-,__________--4 

.... AG 

1995 1996 1997 

1998 

YEARS 

Figure 3-38. Mean performance on ANAMUKR: CPT -ACC. 

OGS-ACC 

t; 

W 80 

0::: 

0::: 

8 

'$. 

~ 70 

............................................. 

---- ._111 -.~'" ~ 

90~----------------------------------~ 

~~ 

" ".~ 

, . 

.. 

,~.-.-. 

----- 

• AE 

• AF 

60~ __________ ~ __________ ~~ __________ ~ 

.... AG 

1995 1996 1997 1998 

YEARS 

Figure 3-39. Mean performance on ANAMUKR: DGS-ACC. 

78

MSP-ACC 

t; 

w 

0::: 

0::: 

o 

() 

"#. 

z 

U) 

~ 

100~---------------------------------------------------~ 

90 

80 

70 

......................................... 

-"'~~ 

~ ......... 

" ..............-.-.- 

" " , " ...... - 

1I ...... - 

GROUP 

• AE 

'-___---------1· AF 

60~---------------~---------------~---------------~ 

1995 1996 1997 1998 

... AG 

YEARS 

Figure 3-40. Mean performance on ANAMUKR: MSP-ACC. 

SPD-ACC 

100~---------------------------------------------------~ 

90 

.'-, . 

., . - .--_ 

.A.----- 

............ ~ .......................... 

80 

ROUP 

• AE 

• AF 

70~ _______________ ~ _______________ ~ _______________ --4 ... AG 

1995 1996 1997 1998 

YEARS 

Figure 3- 41. Mean performance on ANAMUKR: SPD-ACC. 

79

Z 

~ -.. 

t; 

LU 

c:: 

c:: 

0 

(,) 

=*I: 

~ 

~ 

SRT-EFF 

......................................... 

---.--. ,- 

~ ,,'" 

170~--------------------------------------------------------~ 

160 

150 

140 

130 

120 

110 

100 

90 

80 

70 

1995 

1996 

--.. - 

~.~. 

--"- " 

GROUP 

...----...·AE 

11-__ • 

--- AF 

... AG 

1997 1998 

YEARS 

Figure 3-42. Mean performance on ANAMUKR: SRT -EFF. 

110 

Z 

~ 100 

~ 

90 

~ 

0 

(,) 

80 

'*I: 

~ 70 

~ 

2CH-EFF 

120~---------------------------------------------------~ 

.. ~ ............................................ . 

.. """ ........ 

"-.. 

'. ,. 

, ."",-- 

.. , .... ,. 

~ - .... ......... GROUP 

....... - 

----- • AE 

~ • AF 

50 .... AG 

1995 1996 1997 1998 

YEARS 

Figure 3-43. Mean performance on ANAMUKR: 2CH-EFF. 

80

CDS-EFF 

50~------------------------------------~ 

z 

~ 40 .., § 

........ 

O~ 

30 

() 

~ 

~ 

-- ....... '..,e--._._....- 

.. ............. 

• • 

" 

'.... 

GROUP 

::2: 20 • AE 

• AF 

10~ __________ ~ ____________ ~ __________ -4 A AG 

1995 1996 1997 1998 

YEARS 

Figure 3-44. Mean performance on ANAMUKR: CDS-EFF. 

CDI-EFF 

40~------------------------------------~ 

"' .. 

.... ., 

....... , 

'....... . ..-- 

-... ~.-.-.~-.-- 

'-.... ........ ........ 

..... 

........ 

.......... 

............. • AE 

GROUP 

• AF 

10~ __________ ~ __________ ~~ __________ 4 A AG 

1995 1996 1997 1998 

YEARS 

Figure 3-45. Mean performance on ANAMUKR: CDI-EFF. 

81

CDD-EFF 

50~--------------------------------------------------------, 

Z 40 

~ 

W 

0::: 

0::: 30 

o 

=*I: 

~ 

~ 20 

......................................... 

" 

, .. 

". 

,, '. 

, 

~--. 

---...------ 

, , --_ ...... 

.--- -- 

• AE 

~-.. .... AF 

10~ _______________-,__________________ ~ _______________ ~ ... AG 

1995 1996 1997 1998 

YEARS 

Figure 3-46. Mean performance on ANAMUKR: CDD-EFF. 

CPT-EFF 

100~------------------------------------------------------~ 

90 

80 

70 

. --..,._ ....... 

• ~.

DGS-EFF 

50~------------------------------------------------------~ 

Z 40 

~ . ~."': ~ ...................................... 

-----... ~ "'-'~- _ 

0::: .... 'It----L. 

0::: ..-.- ~ ........ 

o 30 - - ::::-..... 

_ .. .. 

.. () 

'*I: ---... -.:. _ GROUP 

~ --- 

~ 20 ~----_.J.. • 

AE 

• AF 

10~ _______________ ~ __________________ ~ _______________ -4 • AG 

1995 1996 1997 1998 

YEARS 

Figure 3-48. Mean performance on ANAMUKR: DGS-EFF. 

z 

~ 

hl 30 

~ 

() 

'*I: 

~ 20 

:2 

MSP-EFF 

.......................................... 

40~-------------------------------------------------------, 

, -_.-4 ...... 

, -.. .~ 

, -- ,~ 

....-....-- . 

, ...... ......... ...... 

~ 

, , , -, .. , 

'---- ~ 

,.- ' .... GROUP 

' .... 

~-..-..-..-..-..-e~ __________ I·AE 

• AF 

10~ _______________ ~ __________________ ~ _______________-4 

& AG 

1995 1996 1997 1998 

YEARS 

Figure 3-49. Mean performance on ANAMUKR: MSP-EFF. 

83

SPD-EFF 

40~------------------------------------, 

z 

~ , 

j:: 30 , 

() , 

w •••• ~ ••••••••••••••••••••••••••••••• .......... 

~ ~- 

, 

~ .-~-. .-...-. ",- 

O 

() 

' 

, 

.......-.,.-. 

'*I: , 

~ 20"' ___ ~",,-=-:-:::::-:~ 

:2 

• AF 

10 ... AG 

~----------~----------~------------4 

1995 1996 1997 1998 

YEARS 

Figure 3-50. Mean performance on ANAMUKR: SPD-EFF. 

TAPPING-RIG HT 

S2~------------------------------------~ 

so 

,.,. .... 

58 ~ .. 

••• • • •• •• • ••• ••• • ••• •• •• ••• • •• -:..;..r-~••••••• 

~ 56 

-M-" 

/_.-.-- 

~ 54 

_ - 

# 

'*I: #/ 

~ 52 ~#/ 

itJ, _ .... ---__ GROUP 

:2 50)./. .._ .... - _ 

~ --1~. 

48 .". .... - ~ ..__ 

AE 

....-..........-----·-------T.. 

~~-- 

--- 

4S I '" • AF 

44~----------__ ------------__ ----------_4 A AG 

1995 1996 1997 1998 

YEARS 

Figure 3-51. Mean performance on ANAMUKR: TAPPING-RIGHT. 

84

TAPPING-LEFT 

54~-----------------------------------. 

52 

................................... ..•. ,;,;. .. 

50 

------ 

~ 48 .-.Jit--- 

I- 

=II: 46 

. 

_ .. ..ttl!"./ '" 

11-- 

~ 

~ / ---... 

:2: 44 ~ 

GROUP 

, 

42 

40 

~ 

~ 

, • AE 

~~,:-................ """" ........ ~----------~. AF 

38~ __________ ~ __________ -r __________ -4 .6. AG 

1995 1996 1997 1998 

YEARS 

Figure 3-52. Mean performance on ANAMUKR: TAPPING-LEFT. 

SLEEPINESS 

UJ 

~ 

OC/)° 

~ 

:2: 

2.0 

1.5 

, • 

, , • 

,, ., 

,, ., ------ 

~-- 

, . ~ 

, . -- 

~ 

, .... ~.-.~--. 

........, , .. ......... ".. .................. 

... ~ ........ 

~~ .......... 

~ -- 

1.0 ....------------.------------r-------------j, 

1995 1996 1997 1998 

• AE 

• AF 

.6. AG 

YEARS 

Figure 3-53. Mean ratings on ANAMUKR: SLEEP SCALE. 

85

SECTION 4 

CONCLUSIONS 

Taken collectively, the results of the data analyses are rather frightening. Initial dosages were from 1 rad 

to 183 rads. Our research suggests neurocognitive and physical decrements in performance 12 years 

AFTER a nuclear accident. They indicate that not only have cognitive and physical functioning of the 

AEs been severely compromised by exposure to the environmental effects of the ionizing radiation being 

emitted by the Chemobyl power station, but also that cognitive and physical performance of the other 

groups (AFs, AGs) in the vicinity of Chemobyl have been affected as well, although to a lesser degree. 

At this point it is hard to specify whether the observed impairments are permanent or temporary 

(although they would appear to be permanent in the AEs). It is also difficult to determine whether they 

are a result of direct effects of radiation on the CNS itself, as opposed to being an indirect result of 

bodily illness resulting from chronic radiation perhaps paralleling that experienced by people diagnosed 

as suffering from reporting heat exhaustion (Gestaldo, et. al., 1997). It is also possible that the 

individuals in and around the Chemobyl area are experiencing symptoms somehow similar to those of 

the "neurasthenic syndrome," which was reported by Soviet workers exposed to non-ionizing microwave 

radiation in the 1980's. As described by Akoyev and Justesen (personal communication), this syndrome 

included fatigue, malaise, and achiness. Such factors would most certainly result in compromised 

performance on neurocognitive tasks requiring attention and working memory, and on physical tasks 

requiring explosive and sustained energy. Since this study was not a "medical" study, we were not 

equipped to identify aplastic anemia (pancytopenia). 

The results of the 1996 retest indicated that both the physical and cognitive abilities of the individuals 

initially exposed (Eliminators) to the ionizing radiation resulting from the Chemobyl nuclear accident 

were seriously declining. Although the 1997 retest indicated that the forestry and agricultural workers 

were actually improving on the physical tasks, the cognitive performance of these groups was becoming 

globally impaired; i.e., they showed significant impairments on the majority of the ANAMUKR tasks. 

This global impairment is reminiscent of that observed in survivors of moderate-to-severe traumatic 

brain injuries (Levinson & Reeves, 1997; Levinson, et al, 1998). Unlike those people, however, whose 

cognitive performance improved over time, these individuals continued to experience increasing 

difficulty in neurocognitive function. As of eleven years after the accident, they continued to decline 

and had not yet plateaued. The results of the 1998 retest indicate that the declines of the exposure 

groups appeared to be leveling off, and improvements in performance were observed in some cases. 

Nevertheless, the results of analyses of the 4-year averaged scores indicate that the effects of exposure to 

radionuclides in and around the area of Chemobyl have resulted in clinically meaningful impairments in 

both physical and cognitive performance. Further, the fmding that significant correlations between 

dosage and 4-year averaged performance occurred on 21 of24 tasks for the combined exposure groups is 

extremely disconcerting. Retests performed during the next several years will be extremely valuable in 

determining whether the physical and neurocognitive performance of these individuals resumes to 

decline, continues to plateau, or begins to improve. 

There are in the literature, several views on low-dosage radiation. One view holds that low-dosage does 

not cause an increase in cancer. However, our end was "performance" and physical and 

neuropsychological performance, according to our study, has been severely compromised. We believe it 

is time to take into account performance as well as health consequences. 

86

Ukrainians have suffered serious medical problems as a result of the Chernobyl disaster, and their 

country has suffered economically since the accident. The burden of medical treatment is enormous and 

containment of nuclear pollution is almost impossible. Ukrainian scholars and scientists are generally in 

agreement that Chernobyl was one of the causes for Ukrainian independence from the former Soviet 

Union. For Russia the economic expenditures for environmental clean-up and treatment of the 

population were prohibitive. 

In 1991, the Ukrainian Supreme Soviet enacted a law, which would decommission Chernobyl by the end 

of 1993; however, in October 1993, this law was repealed. The cost of decommissioning Chernobyl, 

especially in light of obtaining alternate energy sources, remains too great. The Chairman of the 

Ukrainian Supreme Soviet Committee on Chernobyl, Volodymyr Javorivsky said, "We will be 

extracting problems from the well ofChernobyl for a very long time." In light of this report, that is an 

understatement. 

To put this report into some perspective it needs to be stated that while contamination of the ecostructure 

is considerable, much of the contamination affecting the Ukraine is below international 

standards for life-time dosages. Poor diet, the stress of living in a country where the minimum wage is 

$1.50 per month, and where food is a daily preoccupation, are not adequately addressed in medical 

reports which blame Chernobyl for all increases in disease. Yet research has shown that exposure to 

ionizing radiation is harmful to humans and the environment. The effects of living in contaminated 

areas, as well as growing and consuming contaminated food, have received little research support. The 

psychological concomitants of Post-Traumatic Stress Disorder are being considered in terms of recent 

child development studies in the Ukraine. 

The Ukrainians are desirous for research support. The Ministry of Forests is particularly interested in 

research support to develop models for cognitive and physical decrements in performance associated 

with forestry workers exposed to contaminated forests. The Director of the Ukrainian Psychological 

Research Institute is also desirous of research support dealing with psycho-sociological problems 

associated with adults and children living in contaminated areas, or who have relocated from these areas. 

The need for assistance to help plan and guide longitudinal research has been expressed by those who 

have been working with the children affected by the Chernobyl disaster. 

The benefits of these research proposals would be enormous for the United States. Nuclear energy is a 

fact worldwide. Data gathered by research efforts involving Chernobyl could impact Federal 

Emergency Management Administration procedures for relocating victims of similar accidents, and their 

medical and psychological treatment. Unfortunately, in the conduct of this research, a wealth of data 

was obtained that is available, however, funds were not allocated for analyses. In addition, little is 

known about cognitive and physical decrements in performance or stress associated with living and 

working in contaminated areas. Preparedness should be the bottom line in research. 

Read (1993) mentions that because radiation technology was so new, there was no way to be prepared 

for all possible problems resulting from its use. Nonetheless, it is important that people learn from such 

accidents about the effects of radiation, especially considering the aftermath of nuclear war or terrorist 

attack. This remains one of the objectives of the present, ongoing study of the physical and 

neurocognitive effects of the Chernobyl accident. 

87

SECTION 5 

REFERENCES 

Awramenko, O. (1992). Health is deteriorating: Problems remain. Your Health, 9, 23. 

(UNCLASSIFIED) 

Baranov, A.E. & Guskova, A. K. (1988). Acute Radiation Disease in Chernobyl Accidents Victims. 

Paper presented at the Institute of Biophysics, Ministry of Health of the USSR, Moscow, USSR. 


Baryahtar, V. & Bobyleva, O. (1991). Generalized Data of the Academy of Sciences of the 

Ukraine and the Ministry of Health Characterizing the Consequences of the Chernobyl Disaster with 

Respect to the Health ofthe Population. Ukrainian Ecological Bulletin, 4. (UNCLASSIFIED) 

Chernousenko, V. M. (1991). Chernobyl: Insightfrom the inside. Berlin: Springer-Verlag. 


Fleishman, E.A. & Quaintance, M.K. (1984). Taxonomies of human performance: The 

description of human tasks. Orlando, FL: Academic Press. (UNCLASSIFIED) 

Gamache, G. (1993). Gamache physical abilities battery. (Report No. KGAI-93-97-TR). 

St. Augustine, FL: KGA International. (UNCLASSIFIED) 

Gastaldo, E. Reeves, D. Levinson, D., & Winger, B. (1996). ANAM USMC normative data, Series I: The 

effects of hyponatremia on neurocognitivefunction. (Report No. NCRF-TR-96- 01). San, Diego, CA: 

National Cognitive Recovery Foundation. (UNCLASSIFIED) 

Gittus, J., Hicks, D., Bonnell, P., Clough, P., Dunbar, I., Egan, M., Hall, A., Hayns, M., Nixon, W., 

Bullock, R., Luckhurst, D., Maccabee, A., and Edens, D. (1988). The Chernobyl Accident and its 

Consequences. London: Atomic Energy Authority. (UNCLASSIFIED) 

Goldstone, A., Reeves, D., Levinson, D.M., & Pelham, M. (1995). Effects of stroke on cognitive 

performance as assessed by Automated Neuropsychological Assessment Metrics (ANAM V3.11 a). 

Report No. NCRF-95-01). San Diego, CA: National Cognitive Recovery Foundation. 


Guskova, A. et al. (1988). Medical Aspects of the Chernobyl Accident. Proceedings of All-Union 

Conference organized by the USSR Ministry of Health and the All-Union Scientific Centre of Radiation 

Medicine, USSR Academy of Medical Sciences, held in Kiev, Ukraine, 11-13 May, 1988. 


Kay, G. (1995). Cogscreen: Aeromedical edition; professional manual. Odessa, FL: Psychological 

Assessment Resources, Inc. (UNCLASSIFIED) 

88

Kopeikin, V. (January, 1993). Answersfrom Professor Kopeikinfrom Kazan. Vestnik Chernobylia 

9 (442). (UNCLASSIFIED) 

Laupa, A. & Anno, G. (1989). Chernobyl Accident Fatalities and Causes. Pacific-Sierra Research 

Corporation Technical Report 1865 prepared for the Defense Nuclear Agency, Washington, D.C. 


Levinson, D.M. (1990, August). Normative data on a subset of the Unified Tri-Service Cognitive 

Performance Assessment Rattery (UTC-PAR). Presented at the 98th American Psychological 

Association Annual Convention, Boston, MA. (UNCLASSIFIED) 

Levinson, D. M., & Reeves, D. L.(1994). Automated Neuropsychological Assessment Metrics 

(ANAM): ANAMVi.ONormative Data. (Report No. NCRF-TR-94-01). San, Diego, CA: National 

Cognitive Recovery Foundation. (UNCLASSIFIED) 

Levinson, D.M., & Reeves, D.(1997). Monitoring recovery from traumatic brain injury using 

Automated Neuropsychological Assessment Metrics (ANAM Vl.O). Archives of Clinical 

Neuropsychology, 12, 155-r66. (UNCLASSIFIED) 

Levinson, D.M, Reeves, D.L., Wild, M.R, & Lewandowski, A.G. (1998). Classifying level of 

neurocognitive impairment in individuals with acquired brain injury. Archives of Clinical 

Neuropsychology, 13, 73. (UNCLASSIFIED) 

Lewandowski, A.G., Reeves, D.L., & Dietz, A.J. (1994, April). Assessing the pharmacodynamics of 

CNS compounds through repeated neuropsychological measures under clinical trial conditions. 

Presented at the 14th Frontiers Symposium of the American College of Clinical Pharmacology, 

Rockville, MD. (UNCLASSIFIED) 

McClellin, G.E., Anno, G.H., Whicker, F.W. (1994). Chernobyl Doses Volume i-Analysis of Forest 

Canopy Radiation Response from Multispectral Imagery and the Relationship to Doses. (Technical 

Report DNA 001-87-C-0104). Washington DC Defense Nuclear Agency. (UNCLASSIFIED) 

Read, P.P. (1993). Ablaze: The story of the heroes and victims ofChernobyl. New York: Random 

House. (UNCLASSIFIED) 

Reeves, D. & Gamache, G. (1994). Ukrainian subset of ANAM battery: ANAMUKR. (Report No. 

KGAI-94-106-TR). St. Augustine, FL: KGA International. (UNCLASSIFIED) 

Reeves, D., Kane, R., Winter, K., Raynsford, K., & Pancella, T. (1993). Automated 

NeuropsychologicalAssessment Metrics (ANAM): Test administrator's guide Version i.O. St. Louis: 

Missouri Institute of Mental Health. (UNCLASSIFIED) 

Reeves, D., Kane, R., & Winter, K. (1995). Automated Neuropsychological Assessment Metrics 

(ANAM): Test administrator's guide Version 3.11. (Report No. NCRF-SR-95-01). San Diego, CA: 

National Cognitive Recovery Foundation. (UNCLASSIFIED) 

89

Reeves, D., Schlegel, R, Gilliland, K., & Crabtree, M. (1991). UTC-PAB and the NATO/AGARD 

STRES Battery: Results from standardization studies. Proceedings of the 1991 Medical Defense 

Bioscience Review. Aberdeen Proving Grounds, MD. (UNCLASSIFIED) 

Reeves, D.L., Thome, D.R, Winter, S.L., & Hegge, F.W. (1989). The Unified Tri-Services 

CognitivePerformance Battery (UTCPAB): II Hardware and software design and specifications. (Report 

No. SR 89-1). Pensacola, FL: Naval Aerospace Medical Research Laboratory. (UNCLASSIFIED) 

Reeves, D.L. & Winter, K.P. (1992). ANAM documentation volume I: Test administrator's 

guide. Washington, DC: Office of Military Performance Technology, Scienceboard. (UNCLASSIFIED) 

Sorodovych, E., Dudkin, V., Kaletnik, N., Los, I., Guclzenko, V., Bobyleva, D., & Tabachny, L. 

(1992). Chernobyl Catastrophe: Reasons and Consequence. Kiev, Ukraine: Academy of Science. 


Thome, D.R (1990). Throughput: a simple performance index with desirable characteristics. 

Washington, DC: Office of Military Performance Technology, Scienceboard. (UNCLASSIFIED) 

Thome, D., Genser, S., Sing, H., & Hegge, F. (1985). The Walter Reed Performance Assessment 

Battery. Neurobehavioral Toxicology and Teratology, 7,415-418. (UNCLASSIFIED) 

Vestnik Chemobylia. (May, 1993). Energy Generated by the Ukraine. 37 (472). (UNCLASSIFIED) 

Woytsehowich, O. (1991). The Problems of Radioactive Pollution of the Dnipro River System After 

Five Years of the Chemobyl Catastrophe. Ukrainian Ecological Bulletin, 5. (UNCLASSIFIED) 

Yakovlev, E. A. (1991). Geological and Economical Aspects of Decommissioning the Chemobyl 

Nuclear Power Plant. Ukrainian Ecological Bulletin, 2. (UNCLASSIFIED) 

Yakovlev, E.A. (1992). We are hewing the bough. Rabochaya Gaseta (August 12). (UNCLASSIFIED) 

90

APPENDIX A 

UKRAINIAN PROJECf DEPARTMENTS AND PERSONNEL 

CONTACTED 

Ukrainian Psychological Institute 

2 Pankovskaya Street 

Kiev, Ukraine 252033 

Director, Professor A. K. Wasilewich 

Director of the Children's Center for Chernobyl, Dr. S. Yakovenko 

Project Manager, Adults of Chernobyl, Dr. Y. Sewalb 

Child Psychologist, Natalia A Bastun 

Director, Ukrainian Guidance Center, Dr. V. Panok 

Institute for Psychiatry 

103 Frunze Street 


Director, A. P. Chuprinov, M.D. 

Ministry of Forests 

5 Kreshechatik Street 


Head of the Scientific and Technical Office, N. Kaletnik 

Minister of Chernobyl Affairs 

1 Lesi Ukrainki Square 


Chiet: Protection of the Population Department, L. Tabachny 

First, Vice-Minister, Boris S. Prister 

Director, International Cooperation, Y. Pavlov 

Department Head, Statistics, N. Repina 

Shevchenko University 

64 VJadimirskaya Street 


Chairman, Psychodiagnostics and Medical Psychology, Prof L. Burlatchuk 

First, Vice-Rector, Prof O. Tretyak 

Vice-Rector of Education, Prof L. Gubersky 

A-I

Ukrainian Academy of Sciences 

Dr. V. V. Gudzenko 

Dr. E. V. Sobodovych 

A-2

APPENDIXB 

PHOTOGRAPHSOFCHERNOBYL 

AND 

TESTING SITES 

Figure B-1. ChernobyI site after explosion, 26 April 1986. 

B-1

Figure B-2. Russian monitors, 26 April 1986. 

B-2

Figure B-3. Setting up to test eliminators. 

B-3

Figure B-4. Agricultural workers ready for testing on At~AMUKR 

B-4

Figure B-S. Agricultural workers walking on balance beam. 

B-5

Figure B-6. Agricultural workers performing broad jump. 

B-6

Figure B-7. Control person carrying weights. 

B-7

Figure B-8. Forestry workers performing ANAM. 

B-8

Figure B-9. Dr. Gamache at Chernobyl Nuclear Power Plant, 1997. 

B-9/B-I0

APPENDIXC 

CONSENT FORM 

Dear 

---------------------------------------------------------------- 

You are invited to participate as a VOLUNTEER in a 4-year MINIMAL RISK research project 

designed to test your physical and cognitive perfonnance. This research is conducted by KGA 

International, Kiev Polytechnic Institute, and the Ukraine Center for Radiation Medicine. The research 

is planned for 1995-1998, and the testing will be carried out in the summer months and will require you 

to be tested, as scheduled. Complete testing will take one-half day. 

Prior to testing you will be provided with instructions how to perfonn the tests. The physical 

part of testing is based on simple exercises that are easy to perfonn for any individual. The cognitive 

test will be perfonned on a computer with appropriate instructions in Russian. 

If, during the testing, you feel ill or want to ask a question, notify your instructor immediately. 

If you are ill, the instructor will refer you to the medical staff. Inquries regarding instructions given 

MAY necessitate starting testing over again. Make sure to ask ALL questions prior to commencing the 

test procedure. 

I, certify that I am a volunteer and all procedures and risks have been thoroughly explained to 

me. I also have the choice NOT to participate at any time. 

Signature __________________________________ _ 

Dare _________________ _ 

C-lIC-2

APPENDIXD 

GLOSSARY 

2CH 

AC 

ACC 

AE 

AF 

AG 

ANAM 

ANAM-ACC 

ANAM-EFF 

ANAMUKR 

BALBEAM 

BROADJMP 

CARRYWGT 

CDD 

CDI 

CDS 

COMP 

CPT 

DECL 

DECR 

DGS 

EFF 

GPAB 

MSP 

SLP 

SPD 

SQUAI"I'HR 

SRT 

TAP-L 

TAP-R 

Twq-choice Reaction Time 

Control group 

Accuracy 

Eliminator group 

Forester group 

Agricultural group 

Automated Neuropsychological Assessment Matrices 

ANAMUKR - accuracy scores 

ANAMUKR - efficiency sores 

Special subset of ANAM created for this study 

Balance Beam 

Broad jump 

Carrying weights 

Code Substitution - delayed recall 

Code Substitution - immediate recall 

Code Substitution - visual search 

Composite measure 

Running Memory Continuous Perfonnance Task 

Decline 

Decrement 

Digit Symbol 

Efficiency 

Gamache Physical Abilities Battery 

Matching to Sample 

Stanford Sleepiness Scale 

Spatial Processing 

Squat thrusts 

Simple Reaction Time 

Tapping -left index finger 

Tapping - right index finger 

D-lill-2



DIRECTOR 


INFORMATION SERVICES DEPARTMENT 

8901 WISCONSIN AVENUE 

BETHESDA, MD 20889-5603 

ATTN: BHS 

ATTN: EXH 

ATTN: MRA 

ATTN: RSD 



8725 JOHN J. KINGMAN ROAD, SUITE 0944 

FORT BELVOIR, VA 22060-6218 

ATTN: DTIC/OCP 




ATTN: TDA,C.MCFARLAND 

ATTN: TDAC 

ATTN: TDACC 

ATTN: TDANP, TRC 


ALBUQUERQUE OPERATIONS 

1680 TEXAS STREET, SE 

KIRTLAND AFB, NM 87117-5669 

ATTN: TDTO 

DIRECTOR 

DISA 

TECHNICAL RESOURCES CENTER 

7010 DEFENSE PENTAGON 

WASHINGTON, DC 20301-7010 

ATTN: JNGO 

PRES~DENT 


FORT LESLEY J. MCNAIR 


ATTN: ICAF, TECH LIB 

ATTN: LIBRARY, STOP 315 

DIRECTOR 


OFFICE OF THE SEC OF DEFENSE 

ROOM 3A930, THE PENTAGON 

WASHINGTON, DC 20301 


HEADQUARTERS 

US EUROPEAN COMMAND 

UNIT 30400 BOX 1000 

APOAE 09128 

ATTN: ECJ5-N 

USSTRATCOMlJ531T 

901 SAC BOULEVARD, SUITE-BB11 

OFFUTT AFB, NE 68113-5160 

ATTN: J-521 


INSTITUTE FOR DEFENSE ANALYSES 

CONTROL & DISTRIBUTION 

1801 N. BEAUREGARD STREET 

ALEXANDRIA, VA 22311 

ATTN: D. SCHULTZ 

ITT INDUSTRIES 

ITT SYSTEMS CORPORATION 

ATTN: AODTRAlDTRIAC 

1680 TEXAS STREET, SE 

KIRTLAND AFB, NM 87117-5669 

ATTN: DTRIAC 

ATTN: DTRIAC/DARE 

JAYCOR 

1410 SPRING HILL ROAD, SUITE 300 

MCLEAN, VA 22102 

ATTN: DR C. P. KNOWLES 

PACIFIC-SIERRA RESEARCH 

OPERATING COMPANY OF VERID 

29801 28TH STREET, 2ND FLOOR 

SANTA MONICA, CA 90405 

ATTN: G. ANNO 

ATTN: H. BRODE 

ATTN: L. MATHESON 

PACIFIC-SIERRA RESEARCH CORPORATION 

WASHINGTON OPERATIONS 

1400 KEY BOULEVARD, SUITE 700 

ARLINGTON, VA 22209 

ATTN: G. MCCLELLAN 

PSYCHOLOGICAL RESEARCH CENTER 

8 ALTHEA STREET 

ST AUGUSTINE, FL 32095 

ATTN: D. L. REEVES 

ATTN: D. M. LEVINSON 

ATTN: L. GAMACHE 

ATTN: P. I. BIDIOUK 

DL-1

SCIENCE APPLICATIONS INT'L CORPORATION 

10260 CAMPUS POINT DRIVE 

SAN DIEGO, CA 92121-1578 

ATTN: D. KAUL, MS-33 

ATTN: E. SWICK, MS-33 


MAIL LOCATION 577 

CINCINNATI, OH 45267 

ATTN: E. SILBERSTEIN 


UNIVERSITY OF CALIFORNIA 


P. O. BOX 808 

LIVERMORE, CA 94551-9900 

ATTN: Z. DIVISION LIBRARY 


ATTN: MAIL SERVICES 

P. o. BOX 5800 

ALBUQUERQUE, NM 87185-0459 

ATTN: TECH LIB, MS-0899 



600 CHENNAULT CIRCLE 

BUILDING 1405 - ROOM 160 

MAXWELL AFB, AL 36112-6424 

ATTN: AUL-LSE 

ATTN: LIBRARY 



STUDIES AND ANALYSES AGENCY 

1570 AIR FORCE PENTAGON 


ATTN: AFSAAISAMI 

DIRECTORATE OF NUCLEAR 

AND COUNTERPROLIFERA TION 



ATTN: XOOSS 

US AIR FORCE ACADEMY 

2354 FAIRCHILD DRIVE, SUITE 3A22 

USAF ACADEMY, CO 80840-6214 

ATTN: LIBRARY 



DEPUTY CHIEF OF STAFF FOR 

OPERATIONS AND PLANS 

THE PENTAGON 




US ARMY MEDICAL RESEARCH & 

DEVELOPMENT COMMAND 

FORT DETRICH, BUILDING 722 

FREDRICK, MD 21702-5012 

ATTN: SGRD-PLE 

US ARMY MODEL IMPROVEMENT STUDY 

MANAGEMENT AGENCY 

CRYSTAL SQUARE 2 

1725 JEFFERSON DAVIS HIGHWAY SUITE 808, 

ARLINGTON, VA 22202 

ATTN: SFUS-MIS 

COMMANDER 

US ARMY NUCLEAR & CHEMICAL AGENCY 

7150 HELLER LOOP, SUITE 101 

SPRINGFIELD, VA 22150-3198 

ATTN: MONA-NU 

US ARMY RESEARCH LAB 

5232 FLEMING ROAD 

ABERDEEN PROVING GROUND, MD 21001-5067 


ATTN: SLCBR-D 

ATTN: SLCBR-DD-T, TECH LIB 


US ARMY RESEARCH LABORATORIES 

2800 POWDER MILL ROAD 

ADELPHI, MD 20783-1197 


COMMANDER 

US ARMY TEST & EVALUATION COMMAND 

BUILDING 860 

ABERDEEN PROVING GROUND, MD 21005 

ATTN: STECS-NE,70 

DIRECTOR 

US ARMY VULNERABILITYILETHALITY 

ASSESSMENT MANAGEMENT OFFICE 

ABERDEEN PROVING GROUND, MD 21005-5001 

ATTN: AMSLC-VL-NE, DR J. FEENEY 

DL-2

SUPERINTENDENT 

US MILITARY ACADEMY 

WEST POINT, NY 10996 

ATTN: DEPT OF BEHAVORIAL 

SCI & LEADERSHIP 

DIRECTOR 

WALTER REED ARMY INSTITUTE OF RESEARCH 

6825 16TH STREET, NW 


ATTN: SLRD-UWI 


PRESIDENT 


686 CUSHING ROAD 

NEWPORT, RI 02841-1207 

ATTN: LIBRARY, CODE 1E22 


NATIONAL ARCHIVES & RECORDS 

ADM I NISTRA TION 

8601 ADELPHI ROAD, ROOM 3360 

COLLEGE PARK, MD 20740-6001 

ATTN: USER SERVICE BRANCH 

DL-3

SIMULATING WET DEPOSITION 

OF RADIOCESIUM 

FROM THE CHERNOBYL ACCIDENT 

THESIS 

Aaron M. Kinser, Captain, USAF 

AFIT/GM/ENP/OIM-05 



AIR FORCE INSTITUTE OF TECHNOLOGY 

Wright-Patterson Air Force Base, Ohio 

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED. 

I'J 

C=» 

C=» 

~ 

C=» 

-........ 

«....PrI 

C=» 

C=» 

«....PrI 

'-C) 

/

The views expressed in this thesis are those of the author and do not reflect the official 

policy or position of the Department of Defense or the United States Government.

AFIT /GM/ENP /01M-5 

SIMULATING WET DEPOSITION OF RADIOCESIUM 


THESIS 

Presented to the Faculty of the Department of Engineering Physics 

Graduate School of Engineering and Management 

of the Air Force Institute of Technology 

Air University 

In Partial Fulfillment of the 

Requirements for the Degree of 

Master of Science 

Aaron M. Kinser, B.S. 

Captain, USAF 

March 2001 

Approved for public release; distribution unlimited

AFIT /GM/ENP /OlM-5 



Aaron M. Kinser, B.S. 

Captain, USAF 

Approved: 

~1f{~ 

Lt. Col. Michael K. Walters (Chairman) 

~ 

i 2fo..tR 0 I 

Date 

/2.. JI1~ {J/ 

Date 

Dr. Dennis W. Quinn (Member) 

Date

· ----------------------------------------------------------------------------------------~ 

Acknowledgements 

Craig Sloan from AFTAC offered the initial inspiration for this research, 

and his continued coordination has been crucial to the work. He also personally 

introduced me to Roland Draxler, paving the way for our productive 

relationship. The greatest and most generous contributions to this 

work came from the lead author of HySPLIT software, Roland Draxler of 

NOAA's Air Resources Laboratory. He has convincingly earned the title 

of Chief Scientific and Technical Advisor for this thesis. The amount 

and variety of work, knowledge, and collaboration he contributed (almost 

totally via e-mail) merits serious consideration as co-author. Thank you, 

Roland. 

Thesis Committee Chairman and inspiring mentor, Lt Col Michael Walters, 

laid critical groundwork of scientific principles and theory during my 

graduate level studies at AFIT. Major Vince Jodoin helped me catch 

the excitement of modeling nuclear events, and quickly got me back on 

the track whenever I derailed my "nuclear" train of thought. Dr. Dennis 

Quinn introduced sound, fundamental approaches to the data at hand, 

and I have fed on his contagious optimism and enthusiasm. 

My appreciation goes also to Peter J. Rahe, AFIT Meteorology Laboratory 

Technician, for equipment set-up and software installation. I 

am grateful to Capt Lisa C. Shoemaker, Air Force weather officer and 

AFTAC Meteorologist, for technical advice, data templates, and "been 

there" encouragement. Ginger Caldwell from UCAR granted and coordinated 

access to ECMWF data at NCAR. Joey Comeaux from NCAR 

supplied download and conversion instructions for their mass storage system. 

He also assembled a program for reading and processing GRIB format 

data. Thanks to the European Center for Medium-range Weather 

Forecasting (ECMWF) for reanalyzing and sharing ECMWF model data 

with the global weather research community. 

Noted editor and scientist, Giovanni Graziani, from the European Commission's 

Joint Research Centre (JRC) in Ispra, Italy contributed deposition 

data files, key information on handling their contents, the full 

Radioactivity Environmental Monitoring (REM) Chernobyl database report, 

and a handsome atlas of European caesium deposition. 

I dedicate this thesis to a hard-working heroine, my beautiful wife, 

Aaron M. Kinser 

iii

Table of Contents 

Page 

Acknowledgements 

III 

List of Figures 

vii 

List of Tables 

xii 

Abstract ....... . 

Xlll 

I. 

Introduction .......... . 

1.1 Problem and Objective. 

1.2 Thesis Organization 

1 

2 

2 

II. 

Background . . . . . . . . . . 

2.1 Background Overview 

2.2 The Chernobyl Accident as a Wet Deposition Case Study 

3 

3 

3 

2.3 Weather Patterns During the Chernobyl Accident . . 5 

2.4 Cesium-137 Transport from the Chernobyl Accident . 5 

2.5 Other Long-Range Transport Modeling Exercises 8 

2.6 HySPLIT Model Description. . . . . . . . . . . . 11 

III. 

Methodology . . . . . . . . . . . . . . 

3.1 Methodology Chapter Overview . 

3.2 Incorporation of Meteorological Input Fields 

3.3 Comparison to Chernobyl Simulation by ARL 

3.4 In-Cloud Wet Scavenging Rate Sensitivity Runs 

3.5 Modeled Cloud Base Modification. . . . . . . 

3.5.1 Daily Phases of Chernobyl Emissions 

14 

14 

14 

16 

16 

19 

19 

IV

Page 

3.5.2 Modified-Cloud-Base Motivation and Procedure 22 

3.5.3 Modified-Cloud-Base Procedures, Simulation of 

Daily Deposition . . . . . . . . . . . . . . . . 25 

3.5.4 Modified-Cloud-Base Procedures, Simulation of 

April Deposition in Germany and Austria .. 28 

IV. 

Results 

37 

4.1 In-Cloud Scavenging Sensitivity Test Results. 

37 

4.1.1 ICS Sensitivity Over Germany, 86.04.26.06Z 37 

4.1.2 ICS Sensitivity Over Germany, 86.04.26.12Z 44 

4.2 Modified-Cloud-Base Height Simulation Results . . . 44 

4.2.1 Modified-Cloud-Base Performance Over Time 44 

4.2.2 Modified-Cloud-Base Performance in April Over 

Germany / Austria . . . . . . . . . . . . . . . . 62 

V. 

Conclusions . . . . . . . . 

5.1 Sensitivity Runs 

5.2 Cloud Base Modification Runs 

5.3 Future Research Opportunities 

69 

69 

69 

73 

Appendix A. Glossary of Acronyms 75 

Appendix B. 

Radioactivity Primer. 

B.1 Ionizing Radiation 

B.2 Cesium-137 .... 

77 

77 

78 

Appendix C. 

Reanalyzed Precipitation Fields from the ECMWF Model 

80 

Appendix D. HySPLIT Settings . . . . . . . . 

D.1 Release Height Sensitivity Runs. 

D.2 Comparison to ARL Chernobyl Simulation. 

87 

87 

91 

v

----------------------------------------------------------------------------------- 

Page 

D.3 In-Cloud Scavenging Sensitivity Control Run 91 

D.4 Daily Deposition Control Run. . . . . . . . . 91 

D.5 Chernobyl Control Run - Cumulative Deposition on Germany 

and Austria to OOZ, 1986MayOl ......... 92 

D.6 Greece Diagnostic Run HySPLIT Settings - Emission 

10m to 1750m ...................... 92 

Appendix E. Political Map of Europe 93 

Appendix F. HySPLIT Source Code Modification 94 

Appendix G. 

Investigation of Greece Exclusion from Modeled April 

Cs-137 Deposition ........... . 

96 

G.1 Evaluation of Source Term Height - Wet 

G.2 Evaluation of Source Term Height - Dry 

97 

97 

Bibliography 

105 

Vita ..... 

108 

vi

Figure 

1. 

List of Figures 

Simplified 12Z Surface Weather Maps 25 - 28 Apr, 1986 

Page 

6 

2. Cumulative Cs-137 Deposition Measurements 27Apr1986 to lOMay1986, 

from ATMES Report .................. 10 

3. 

4. 

5. 

6. 

7. 

8. 

9. 

10. 

11. 

12. 

13. 

14. 

15. 

16. 

17. 

18. 

19. 

20. 

Illustration of HySPLIT Cloud Base Parameterization 

Domain of ECMWF Input Meteorological Data . . . . 

ARL's Chernobyl Cs-137 Deposition Using HySPLIT . 

Attempted Duplicate of ARL's Chernobyl Cs-137 Deposition 

Using HySPLIT ........................ . 

Chernobyl Cs-137 Emissions in Twelve Phases, Summary Bargraph 

.......................... . 

Emission Height Sensitivity Run, 1500-meter Release. 




Daily Deposition Cities with Measurement Start Dates . 

Surface-Based Cesium-137 Measurement Sites in Europe. 

Modeled April Chernobyl Trajectories from 2100Z, 25Apr1986 

Modeled April Chernobyl Trajectories from 0400Z, 26Apr1986 

Modeled April Chernobyl Trajectories from OOOOZ, 27 Apr1986 

Modeled April Chernobyl Trajectories from OOOOZ, 28Apr1986 



Deposition of Phase I Emissions Over 04.25.21Z - 05.01.00Z . 

12 

15 

17 

18 

20 

23 

23 

24 

24 

27 

29 

31 

32 

33 

34 

35 

36 

36 

21. ECMWF 6-hr Model Precipitation During 86.04.26.06Z Sensitivity 

Run ............................ 38 

Vll

Figure 

Page 

22. In-Cloud Scavenging Sensitivity Test, Control Run Deposition, 

86.04.26.06Z . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 

23. In-Cloud Scavenging Sensitivity, 86.04.24.06Z, Deposition Difference 

with 1% ICS Efficiency Boost ............. 40 


with 1% ICS Efficiency Reduction . . . . . . . . . . . 41 






with 10% ICS Efficiency Boost 43 


with 10% ICS Efficiency Reduction .......... 43 

29. ECMWF 6-hr Model Precipitation During 86.04.26.12Z Sensitivity 

Run ............................ 45 

30. In-Cloud Scavenging Sensitivity Test, Control Run Deposition, 

86.04.26.12Z .......................... . 46 


with 1 % ICS Efficiency Boost ............. 46 








with 10% ICS Efficiency Boost 48 


with 10% ICS Efficiency Reduction .......... 49 

viii

Figure 

37. 

38. 

Page 

Summary of Helsinki Daily Cs-137 Deposition 1986Apr27 - 1986May14, 

Measured, Control Run, and Modified Run . . . . . . . . . . 51 

Summary of Bratislava Daily Cs-137 Deposition 1986Apr27 - 

1986May14, Measured, Control Run, and Modified Run . . . 51 

39. Summary of Moravsky Krumlov Daily Cs-137 Deposition 1986Apr27 

- 1986May14, Measured, Control Run, and Modified Run .. 52 

40. Summary of Hof Daily Cs-137 Deposition 1986Apr27 -1986May14, 


41. Summary of Pass au Krumlov Daily Cs-137 Deposition 1986Apr27 


42. Summary of Schwandorf Krumlov Daily Cs-137 Deposition 1986Apr27 


43. Summary of Hradec Kralov Daily Cs-137 Deposition 1986Apr27 


44. Summary of Kosice Daily Cs-137 Deposition 1986Apr27 - 1986May14, 


45. Summary of Budapest Daily Cs-137 Deposition 1986Apr27 - 


46. Summary of Mol Daily Cs-137 Deposition 1986Apr27 -1986May14, 


47. Summary of Harwell Daily Cs-137 Deposition 1986Apr27 - 1986May14, 


48. Summary of Aachen Daily Cs-137 Deposition 1986Apr27 -1986May14, 


49. Summary of Emden Daily Cs-137 Deposition 1986Apr27 - 1986May14, 


50. Summary of Koblenz Daily Cs-137 Deposition 1986Apr27 -1986May14, 


51. Summary of Schleswig Daily Cs-137 Deposition 1986Apr27 - 


ix

Figure 

Page 

52. Summary of Bilthoven Daily Cs-137 Deposition 1986Apr27 - 


53. Summary of Offenbach Daily Cs-137 Deposition 1986Apr27 - 


54. Summary of Glasow Daily Cs-137 Deposition 1986Apr27 - 1986May14, 


55. Summary of Berlin Daily Cs-137 Deposition 1986Apr27 - 1986May14, 


56. Summary of Giessen Daily Cs-137 Deposition 1986Apr27 - 1986May14, 


57. Summary of Muenchen Daily Cs-137 Deposition 1986Apr27 - 


58. Summary of Berkeley Daily Cs-137 Deposition 1986Apr27 - 


59. Summary of Risoe Daily Cs-137 Deposition 1986Apr27 -1986May14, 

Measured, Control Run, and Modified Run 

62 

60. 

Cloud Base Modification Test, Control Run 

63 

61. Cloud Base Modification Test, 75%-RH Continental Cloud Base 

Run. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 

62. April Chernobyl Deposition Difference: (April Deposition Modified- 

Cloud-Base Run - April Deposition Control Run) . . . . . . . 65 

63. Normal Probability Plot of April Cumulative Cs-137 Deposition 

Measurements in Germany and Austria . . . . . . . . . . . . 67 

64. 

65. 

66. 

67. 

Set 1. Pearson Correlation Coefficients of Daily Deposition by 

City ............................... . 

Set 2. Pearson Correlation Coefficients of Daily Deposition by 

City ............................... . 

Typical Shielding Requirements for Different Ionizing Radiation 

Types from VIC, 00 . . . . . . . 

Current Political Map of Europe 

71 

71 

78 

93 

x

Figure 

68. 

69. 

70. 

Greece Diagnostic Run, Cumulative April Deposition [Bq/m2], 

Release Height Profile from 10m to 1750m . ......... . 

Greece Diagnostic Run, Cumulative April Deposition [Bq/m2] , 

Release Height Profile from 10m to 2990m. . . . . . . . . . . 

Greece Diagnostic Run, Cumulative April Deposition [Bq/m2], 

Release Height Profile from 10m to 9000m . ......... . 

Page 

98 

99 

100 

71. Five-Day Deposition from Exaggerated Phase I Emissions, Precipitation 

Turned Off . . . . . . . . . . . . . . . . . . . . . . 102 

72. Five-Day 1000-m Average from Exaggerated Phase I Emissions, 

Precipitation Turned Off . . . . . . . . . . . . . . . . . . . . 102 


Precipitation 'furned Off .. . . . . . . . . . . . . . . . . . . 103 


Precipitation 'furned Off . . . . . . . . . . . . . . . . . . . . 103 


Precipitation 'furned Off .... . . . . . . . . . . . . . . . . 104 


Precipitation 'furned Off . . . . . . . . . . . . . . . . . 104 

xi

List of Tables 

Table 

Page 

l. Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr25. 81 

2. Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr26. 82 





Xll

AFIT /GM/ENP /OlM-5 

Abstract 

In response to the Chernobyl nuclear power plant accident of 1986, cesium-137 

deposition was measured in Europe at sites equipped to do so. The resulting deposition 

dataset is uniquely applicable to atmospheric transport model validation. 

Most of the airborne Chernobyl cesium was wet deposited, i.e., either via interception 

by falling raindrops (below-cloud scavenging) or via absorption into cloud 

droplets destined to become raindrops (in-cloud scavenging). The model used in 

this work is the Hybrid Single-Particle Lagrangian Integrated Transport (HySPLIT) 

model developed at Air Resources Laboratory. A cloud base modification is tested 

and appears to slightly improve the accuracy of one HySPLIT simulation of daily 

Chernobyl cesium-137 deposition over the course of the accident at isolated European 

sites, and degrades the accuracy of another HySPLIT simulation of deposition in Germany 

and Austria accumulated in the month of April, 1986. Large uncertainties in 

the emission specifications, model precipitation fields, and deposition measurements 

prevent designating the results as conclusive, but most evidence points to improved 

performance within 500km of the emission source. Trial and error lessons learned 

from hundreds of preliminary model runs are documented, and the exact HySPLIT 

settings of successful and meaningful simulations are appended. 

xiii



1. Introduction 

The United States Air Force Technical Applications Center (AFTAC) is charged 

with observing global environmental conditions to detect and identify activities peculiar 

to nuclear weapons testing. AFTAC's global array of seismic, atmospheric, 

and other environmental sensors makes up the U.S. Atomic Energy Detection System 

(USAEDS). By means of USAEDS and a full complement of world-class analytical 

laboratories, AFTAC monitors signatory nations' compliance with international 

nuclear test ban treaties (Hagans 00). In a role supporting this mission, AFTAC 

meteorologists generate routine and special atmospheric pollutant transport simulations. 

The simulations can, from a given source, gauge how much pollutant will 

arrive where and when. Long-range meteorological simulations require accounting 

for precipitation scavenging, both in-cloud and below-cloud. In-cloud scavenging 

(or rain-out), hereafter referred to as ICS, is the process of cloud droplets or ice 

crystals assimilating pollutant within clouds, aggregating, and falling to the ground. 

Below-cloud scavenging (or wash-out), hereafter referred to as BCS, is the process 

of pre-formed precipitation cleansing pollutant from the air below clouds on its way 

to the surface. The combined processes of ICS and BCS result in wet deposition 

at the earth's surface, and playa significant role in removing long-term pollutants 

from the atmosphere. So, improvements to wet deposition modeling are important 

to improving the accuracy of long-range transport simulations. Because wet deposition 

dominated the other long-range Chernobyl fallout deposition mechanisms, 

the Chernobyl case, though severely limited by uncertainty in initial conditions and 

deficiencies in measurement data, is uniquely applicable to wet deposition scheme 

1

validation. AFTAC meteorologists have proposed tests of wet deposition schemes 

in Chernobyl deposition simulations using the Hybrid Single-Particle Lagrangian Integrated 

Trajectory (HySPLIT) model (Draxler 98b). The rest of this document 

expands on the motivations, procedures, and results of requested HySPLIT simulations 

of the Chernobyl accident fallout deposition. 

1.1 Problem and Objective 

Reasonable cloud base parameterization is crucial to realistic model assignments 

of ICS and BCS. To that end, the sensitivity of wet deposition modeling to 

ICS is tested, then used to interpret results of a test of a HySPLIT modified-cloudbase 

scheme, namely, reducing cloud bases to 75% relative humidity over land masses 

from 80%, in pursuit of improved HySPLIT wet deposition parameterization. 

1.2 Thesis Organization 

The chapters of this thesis are structured so as to clarify specific challenges 

to modeling wet deposition of Chernobyl Cs-137. Chapter II provides background 

on aspects of the Chernobyl accident relevant to wet deposition modeling, including 

emission characteristics and prevalent weather patterns. Chapter II also puts 

major atmospheric transport studies in perspective with respect to wet deposition 

modeling. Chapter III describes the methods used to introduce weather variables, 

validates the basic simulation parameters by comparison to previous work, and details 

the methods used for further simulation runs of ICS sensitivity and cloud base 

modification. Chapter IV presents separately the results of said sensitivity runs 

and cloud base modification runs. Chapter V ties the other chapters together and 

gives the reader direction for further wet deposition investigation. The appendix is 

designed to aide the reader in reconstructing and customizing the simulations herein. 

The reader may find the Glossary of Acronyms in Appendix A frequently useful. 

2

---------------, 

II. Background 

2.1 Background Overview 

This chapter includes a review of the Chernobyl nuclear power plant accident of 

April 1986 with emphasis on the role of Cs-137 deposition, a brief description of prevailing 

weather conditions, a short discussion of major long-range transport studies, 

and an overview of HySPLIT, the transport model used for all Chernobyl simulations 

in this thesis. These topics provide a foundation both for an ICS sensitivity 

study described in Section 3.4, and for a cloud base modification study described in 

Section 3.5. 

2.2 The Chemobyl Accident as a Wet Deposition Case Study 

At 2123 UTC (0123L) on 25 April 1986, during a sequence of tests, reactor 

unit number four at the Chernobyl nuclear power plant experienced an unmanageable 

increase in power due to a number of "design deficiencies and operator errors" 

(DeCort 98: 11). Emergency actions by the attendant technicians were fruitless as 

increasing temperatures in the cooling system brought on two violent steam explosions, 

ejecting reactor components into the reactor room, blowing apart portions of 

the building including the roof, and setting dozens of fires at the site. Heat from 

the initial steam explosion and subsequent graphite fire lifted a cloud of radioactive 

particulates at least a kilometer up into the atmosphere. The emission rate gradually 

tapered off until 2 May when heroic efforts to contain the fire caused, instead, 

increasing emissions over the next four days. The ruptured unit was finally sealed 

in a concrete sarcophagus on 6 May (Klug 92:2). In total, the event released 6000- 

8000kg of radioactive material and probably much more inactive material into the 

atmosphere. About one-third of the particles were transported more than 20km 

from the power plant (Pollanen 97). Over the course of the 10-day emission an 

estimated 85 ± 26P Bq of Cs-137 was released (Metivier 95). Later, in Section 3.5.1, 

3

complete Chernobyl emission specifications for simulations used in this thesis are 

described. 

Wet deposition played a dominant role in long-range cesium deposition during 

the Chernobyl accident. Based on estimates of time-integrated concentrations 

and measured dry deposition velocities of cesium (or "caesium") after the accident, 

Lauritzen and Mikkelsen suggest "that only approx. 10% of the total deposition of 

caesium is due to dry deposition, while the remaining 90% stems from wet deposition" 

(Lauritzen 99). Wet deposition modeling, then, must be included in any 

realistic simulation of Chernobyl cesium deposition. Long-range (dry) transport is 

itself an inexact science. Superimposing another process as complicated as rain or 

snow modeling onto the dry transport process takes transport modeling to a new 

level of uncertainty. Lauritzen and Mikkelsen call the distribution of atmospheric 

transport deposition "multi-fractal" and believe that this randomness on all scales 

"implies that standard atmospheric dispersion models (i.e., deterministic models) 

cannot explain details of the deposition pattern, but only its gross, average structure" 

(Lauritzen 99:3271). Case studies of wet deposition are difficult and rare because of 

this compounded uncertainty. Severe patchiness in deposition measurements from 

the localizing effects of precipitation scavenging oblige transport experiment designers 

to carefully schedule experiments so as to avoid the complications of precipitation 

effects. Likewise, nuclear weapons tests are performed on clear days, avoiding dangerous 

radioactive hot spots associated with precipitation (Glasstone 77:418). The 

Chernobyl accident is a unique case study in that it involves a massive quantity of 

radioactive tracer wet-deposited and measured hundreds and even thousands of miles 

away. Because the Chernobyl case is a unique case of measured wet deposition, it 

presents a unique opportunity to validate wet deposition in transport and dispersion 

models. 

4

2.3 Weather Patterns During the Chernobyl Accident 

Although Chernobyl pollutants were eventually detected throughout the northern 

hemisphere, much of Chernobyl's emissions were deposited in Europe because 

of low level circulations and widespread precipitation typical for the season. The 

weather patterns during the accident provided a range of changing conditions throughout 

the continent. Figure 1 presents simplified surface weather charts for the first 

four days of the accident from Knap, 1988 (Knap 88:151). Synoptic weather analysis 

reveals a prevailing cold continental high pressure system to the northeast of Chernobyl. 

Meanwhile, a North Atlantic semi-permanent low pressure system off the 

west coast of Great Britain spawned a series of precipitating troughs across western 

and central Europe during all phases of the Chernobyl accident emissions. While 

the effects of large-scale features west and east of Central Europe on the Chernobyl 

plume were apparent during the time of the accident, the plume was often directly 

steered by smaller, weaker weather features such as the shallow fronts associated 

with these short-wave troughs. 

2.4 Cesium-137 Transport from the Chernobyl Accident 

For particles to travel far enough (100's to 1000's of km) to be considered 

long-range emissions, they must be aerodynamically small enough for turbulence to 

keep them suspended and carried on the wind for days. The Chernobyl fire generated 

massive quantities of sub-micron particles carrying Cs-137 (cesium-137). Small 

particles, 111m in aerodynamic diameter and smaller, have a fall speed of less than 

about l.Omml s. Accordingly, the dry processes that have the greatest influence on 

their transport and deposition are turbulent eddies and Brownian diffusion (small 

particles spread out by random collisions with air molecules) (P611anen 97). Particles 

that tiny remain suspended long enough for precipitation, when present, to 

play a major role in their deposition. Large particles, 20l1m in aerodynamic diameter 

and larger, have a fall speed of greater than about 1O.0mml s in the lower 

5

1030_-.::!..,-__ 

~ 

Figure 1 Simplified 12Z surface weather maps for 25 - 28 Apr, 1986 (Knap 88) 

6

atmosphere. As a result, the processes that have the greatest influence on their 

deposition are gravitational settling (large particles pulled earthward) and turbulent 

dispersion (particle-bearing air mass grows by entrainment). The relatively short 

transport life of large particles prevents rain from playing a significant role in their 

deposition. Particles with a size between small and large present a special challenge 

to the transport and deposition modeler. A single fall speed parameterization for 

medium-sized particles is elusive. Which dry transport mechanisms determine the 

effective fall speed of medium-size particles depends on whether the particles are 

large-medium or small-medium and on several specific weather conditions at each 

location (Seinfeld 86). Fortunately for Chernobyl plume modelers, particles from 

the Chernobyl accident containing Cs-137 are confined mainly to the small category 

because of how they were formed in the fire and smoke at Chernobyl. 

Cs-137 and Sr-90 (strontium-90) are signature long-range fallout isotopes both 

of nuclear power production and of nuclear weapons testing. The human body is 

less susceptible to harm from ingesting Cs-137 than from ingesting the same amount 

of Sr-90. The biological half-life of Cs-137 is 50 to 200days. Sr-90, on the other 

hand, is chemically similar to calcium, so the body tends to concentrate the isotope 

in bone tissue where it remains in the body much longer. So, even though Cs-137 is 

just as easy to measure, has a slightly longer radioactive half-life, and is slightly more 

abundant than Sr-90 in nuclear weapons fallout, nuclear scientists normally characterize 

long-range weapons fallout by patterns of Sr-90 deposition (Glasstone 77:604). 

The accident at Chernobyl, due to the nature of the explosion and fire, produced 

relatively little Sr-90 outside the 30-km evacuation zone (DeCort 98:13). Therefore, 

Cs-137 is the best species for characterizing the long-range radioactive deposition 

pattern from the Chernobyl accident as a whole (Klug 92). 

Because Cs-137 is radioactive, it is detectable in very small concentrations, 

an ideal property for a long-range plume tracer. The radionuclide itself has a' 

radioactive decay half-life of more than thirty years (Serway 92). Such a long half- 

7

life affords meaningful cumulative measurements over periods of months and even 

years. More details are available in Appendix B, a primer on radioactivity and 

Cs-137. An experiment releasing sub-micron particles bearing Cs-137 would, in 

theory, be ideal for validating and improving operational wet deposition modeling. 

Atmospheric nuclear weapon tests in the 1950's and 1960's injected radiocesium 

into the stratosphere which, to this day, continues to trickle radioactive particles 

back into the troposphere, especially at mid-latitudes near the jetstream, although 

one could argue that the amount is negligible (Glasstone 77:448). At any rate, a 

large and hazardous emission would be required to discern long-range experimental 

concentrations above measurement background noise, and the political repercussions 

of such an experiment would be prohibitive. An experimental case study using Cs- 

137 as a tracer is not feasible. So, the Chernobyl case is likely to remain a unique 

opportunity to model and compare Cs-137 deposition on a large scale. 

2.5 Other Long-Range Transport Modeling Exercises 

In November 1986, an international effort emerged to coordinate a transport 

modeling study within the context of the Chernobyl accident. In response to the accident 

and its environmental repercussions, the IAEA (International Atomic Energy 

Agency) collaborated with the WMO (World Meteorological Organization) to develop 

the Atmospheric Transport Model Evaluation Study (ATMES). The study was 

designed both to test the emergency response capability of current transport modeling 

agencies, and to "intercalibrate" their various models (Klug 92:v). Twenty-two 

agencies from fourteen countries volunteered their transport models for the study. 

Each model was developed independently, some for purposes other than nuclear accident 

response, so results varied widely. Some models were designed for meso-scale 

application and were specially adapted for the ATMES exercise. Some simulated 

puff emissions, others tracked individual particles. Some were based on integrations 

in an Eulerian frame, others were based on Lagrangian integrations. Each 

8

model's configuration and results were compiled in the ATMES Report along with 

comparisons to each other and to available measured deposition data. The writers 

of the ATMES Report constructed a contour plot of accumulated Cs-137 deposition 

(Figure 2) from available surface-based measurement data. Though based on all 

available deposition measurements at the time, the figure is not representative of 

the whole pattern of Chernobyl Cs-137 deposition. For instance, even though the 

highest concentrations of Cs-137 were found near Chernobyl at 51.38° latitude, 30.1° 

longitude, Figure 2 suggests a minimum there. So, due to large data sparse regions, 

it appears Figure 2's content may be less representative of area-averaged Cs-137 deposition 

values than of the geographical density of observation sites in the ATMES 

Cs-137 deposition dataset (map in Section 3.5.4). 

A key result of the ATMES project was the realization of the strong need for 

an experimental case, i.e. a transport experiment with known emission specifications 

and synchronized, homogeneous measurements, to confidently evaluate even 

the relative performance of long-range transport models (Klug 92). Controlled experiments 

have several advantages over accidental cases. To date, major long-range 

atmospheric transport modeling experiments have all released non-depositing tracers 

to maintain detectable pollutant concentrations over distance and remove uncertainties 

involved in deposition. Controlled experiment observations are planned 

at regular time and space intervals to generate output grids that are homogeneous 

(Rodriguez 95:800). Perhaps most importantly, the source rate and height are 

known precisely in a controlled experiment, in stark contrast to the typically vague 

specifications of accidental emissions. 

Following the guidance from ATMES conclusions, and the lessons learned from 

the Across North America Tracer EXperiment (ANATEX), the same agencies that 

organized the ATMES Report designed and executed the European Tracer Experiment, 

or ETEX (Rodriguez 95). This time, NOAA's Air Resources Laboratory 

(ARL) was a participant in the study, contributing deposition simulations created 

9

70.oo,..------------::4'i~-~"""'--_, 

61.25 

(J) 

~ 52.50 

~ 

43.75 

I 

............- 

(,/ 

I 

I 

I 

I 

I 

I 

//) 

, , 

I 

I 

I 

I 

I 

I 

I 

I 

I 

I 

I 

I 

.... ~~ ...... ' 

~.oo~~~~~~~~.-~~~~~~~~ 

-10.0 2.5 15.0 27.5 40.0 

Longltlx:le 

MEAS ------. 2 ----. 5 --- 10 -- 50 

Figure 2 Geographic plot of contours of accumulated Cs-137 deposition [kBqjm 2 ] 

from 27 Apr1986 to 10May1986. ATMES Report Figure 10 (Klug 92). 

with HySPLIT Version 4.0 (ARL OOb). 

The organizers of ETEX, like the ANA 

TEX designers, chose a non-soluble perfiuorcarbon chemical as a tracer species. 

The chemical resists deposition by both wet and dry mechanisms, optimizing the 

homogeneity of the tracer's transport pattern, and so making for better simulation 

comparisons. 

Both ANATEX and ETEX were initiated in part to provide a 

dataset for future model evaluations. The careful completeness of each experiment's 

design makes them ideal for long-range (dry) transport model evaluations and improvements. 

However, since the tracer could not be rained out, their datasets are 

not suited for a wet deposition study (Graziani 97). Again, the Chernobyl accident 

10

stands alone as a case study for validating long-range wet deposition schemes against 

in situ measurements. 

2.6 HySPLIT Model Description 

HySPLIT, the Hybrid Single-Particle Lagrangian Integrated Transport model 

calculates either the trajectories of air parcels, or the transport, dispersion and deposition 

of pollutant particles or puffs. User-supplied inputs for HySPLIT calculations 

are pollutant species characteristics, emission parameters, gridded meteorological 

fields, and output deposition grid specifications. Input meteorological fields can be 

on Polar Stereographic, Lambert Conformal, or Mercator map projections. The 

horizontal deformation of the wind field, the wind shear, and the vertical diffusivity 

profile are used to compute dispersion rates. The model can be configured to treat 

the pollutant as particles, or as Gaussian puffs, or as top-hat puffs. The term hybrid 

refers to the additional capability of HySPLIT to treat the pollutant as a Gaussian 

or top-hat puff in the horizontal, while treating the pollutant as a particle for the 

purposes of calculating vertical dispersion. An advantage to the hybrid approach 

is that the higher dispersion accuracy of the vertical particle treatment is combined 

with the spatial resolution benefits of horizontal puff-splitting. All model runs for 

this work were made in the default hybrid particle/top-hat mode. 

HySPLIT calculates wet deposition by scavenging pollutant from portions of 

the plume in (ICS) and below (BCS) precipitating model clouds. All of the scavenged 

pollutant is assumed to deposit on the ground directly below the clouds. To identify 

precipitating model clouds, HySPLIT's wet deposition algorithm checks the input 

meteorological data at each surface gridpoint for precipitation. Where precipitation 

is non-zero, it searches upward from the top of the surface layer (i.e., no fog modeling) 

for the lowest model level with an RH (relative humidity) greater than or equal to 

80%. This 80% threshold establishes the modeled cloud base. A 75% threshold is 

tested later in Chapter III. The cloud top is determined by the lowest level above 

11

Figure 3 

* * 

Illustration of cloud representation in HYSPLIT. Stars represent pollutant 

plume. 80% RH is the default cloud base, 75% RH is the tested 

modification. 

the base where the RH is below 60%. Above the first (lowest) cloud layer, HySPLIT 

diagnoses no more clouds. Figure 3 illustrates a precipitating cloud, both modeled 

cloud bases, and a pollutant plume. 

Below the bases of precipitating clouds HySPLIT scavenges the pollutant 

plume, reducing its concentration by an amount equal to the product of the pollutant 

concentration, the user-specified BGS rate [8- 1 ] (below-cloud scavenging rate), 

and the time increment [8]. The amount of below-cloud pollutant reduction (concentration 

reduction times plume volume) is then added to the surface deposition 

output grid. Deposition from IGS (in-cloud scavenging) is calculated (and in-cloud 

pollutant concentration is reduced accordingly) using a user-specified IGS efficiency 

[L/ L] defined as the ratio of pollutant concentration in air (grams of plume pollutant 

per liter of air) to pollutant concentration in rain (grams of deposited pollutant per 

liter of precipitated water). The amount of deposition from IGS is found by multiplying 

the in-plume pollutant concentration [L- 1 ] by the rain accumulation [mm] 

and dividing by the IGS efficiency as derived in the equations below (unit conversion: 

1m 2 x 1mm = 1Liter = 0.001m 3). 

This approach to IGS requires rain rate [mm] 

12

from the input meteorological grid. HySPLIT is not able to calculate wet deposition 

without input precipitation fields (Draxler 98a:16). 

In-Cloud Plume Concentration ICS Effi . 

. . = C18ncy 

Ram ConcentratIOn 

R · P 11 [ IL] _ Air Pollutant[gl L] 

am 0 utant 9 - ICS Effi 

C18ncy 

. 

R · P 11 [I 2] I R . [ ] _ 0.001 x Air Pollutant[glm3 ] 

am 0 utant 9 m am mm - ICS Effi . 

clency 

[I 2] _ O.OOl(Air Pollutant[glm3 ])(Rain[mm]) 

R am · P 11 

0 utant 9 m - ICS Effi 

clency 

. 

13

III. Methodology 

3.1 Methodology Chapter Overview 

HySPLIT has seen frequent algorithm updates to assimilate current findings in 

the field of atmospheric transport modeling, and HySPLIT's user interface has been 

continually enhanced to improve its usability as an operational tool. This chapter 

supplies the details on exactly how to use HySPLIT to perform selected atmospheric 

transport simulations. Section 3.2 describes how meteorological fields including 

wind, temperature, pressure, humidity, and precipitation data were incorporated in 

simulations for this thesis. To ensure that correct meteorological fields and other 

inputs are being used, a duplicate Cs-137 deposition simulation is attempted and 

compared to results produced by ARL. Section 3.3 explains how the attempted 

duplicate simulation is performed. Section 3.4 describes the method used to evaluate 

the sensitivity of a Chernobyl simulation to various scavenging rates. Finally, Section 

3.5 describes the procedures used to validate a proposed cloud base modification in 

the model against Chernobyl deposition measurements. Results of sensitivity runs 

and of cloud base modification runs are presented later, in Chapter IV. 

3.2 Incorporation of Meteorological Input Fields 

The meteorological input fields for all simulations are reanalyzed ECMWF 

data from NCAR. Using HySPLIT for Chernobyl plume transport and deposition 

calculations requires conversion of ECMWF GRIB format meteorological fields to 

ARL packed format (Draxler 99). The conversion utility program provided with 

HySPLIT requires platform-dependent GRIB decoder libraries typically available 

from the source of raw GRIB data, in this instance NCAR. As it converts a file to 

ARL-packed format, the utility interpolates the data linearly to a polar stereographic 

lat/lon grid in the horizontal, and to internal terrain-following sigma levels in the 

vertical. HySPLIT uses the smallest domain of input meteorological fields as the 

14

Figure 4 

Model domain and resolution of ECMWF meteorological input data grid, 

also used for concentration calculations in HySPLIT. Crosses at every 

fourth gridpoint for clarity. . 

total domain for a given simulation. The domain and resolution of the ECMWF 

input files for all simulations in this work is depicted in Figure 4. Upper air data 

fields for all simulations include temperature in [0C], u and v wind components in 

[m/s], w wind component in [hPa/hr], and specific humidity in [g/kg]. The surface 

data fields provided are 2-m temperature in [0C], 10-m u and v wind components in 

[m/ s], surface pressure in [hPa], and 6-hr prior accumulated precipitation in [mm]. 

For comparison to simulations and for informal diagnosis of wet deposition effects, a 

full set of six-hourly ECMWF re-analyzed precipitation fields over the model domain 

for the first five days of the accident are provided as shaded plots in Appendix C. 

15

3.3 Comparison to Chernobyl Simulation by ARL 

Before addressing the methods for sensitivity runs and cloud base modification 

runs (Sections 3.4 and 3.5), evidence is presented here to show that the 

working installation of HySPLIT is performing as designed, and that appropriate 

meteorological fields and user-specified parameterization settings are being utilized 

properly. The evidence takes the form of results from an attempted duplicate 

of web-published Chernobyl deposition contours generated at ARL (Air Resources 

Laboratory) (ARL OOa). The attempted duplicate simulation uses an abbreviated 

Chernobyl source term, releasing pollutant at a constant rate for 24hrs only. Gross 

features of the 84-hr deposition patterns from the ARL simulation (Figure 5) and 

the attempted duplicate (Figure 6) are in agreement, suggesting that HySPLIT is 

functioning properly, the proper time period of meteorological data has been applied, 

wet scavenging is actually being modeled, etc. Differences (e.g., deposition southeast 

of Chernobyl) between the ARL simulation and the attempted duplicate are 

attributable to ARL's undocumented inclusion of some emissions beyond the first 24 

hours (Draxler OOa). A copy of the control file settings used to create the HySPLIT 

duplicate simulation is furnished in Section D.2 of Appendix D. The setup for this 

simulation serves as a baseline for simulation setups for the remainder of this work. 

3.4 In-Cloud Wet Scavenging Rate Sensitivity Runs 

To gauge the relative importance of ICS in wet deposition modeling, a simplified 

scenario is required mainly because the actual emissions from Chernobyl were 

continuous for days, making it difficult to attribute given deposition to a particular 

release time. So, an abbreviated emission is used in the sensitivity run, and the 

country of Germany is chosen as the deposition domain because the weather conditions 

modeled there also apply to cloud base modification studies in Section 3.5. 

The sensitivity run emission rate, 6.65 x 10 14 Bq / hr, and the emission's uniform vertical 

profile from 1250m to 1750m, mirror the first phase of the Chernobyl emission 

16

DEPOSITION FROM 002 27 APR TO 12Z 30 APR (UTe) 

12:2 215 APR CHNB FOREeAS!' IN1TIALIZATION 

GROUND-LEVEL DEPOSlTlON t ;1112) 

·1.0E+04 • LOE+02 1.0E+OO >0.0 

7.0E+D4 llAXIMUM A.T SQUARE 

Figure 5 

Chernobyl Cs-137 deposition as modeled by ARL. Deposition velocity 

set at 0.lcm/8. In-cloud scavenging ratio set at 3.2 x 10 5 1/1. Belowcloud 

scavenging rate set at 5.0 x 10- 5 8- 1 . Deposition contoured on a 

logarithmic scale 

simply because it is useful to verify the model's ability to accommodate values with 

these magnitudes. Apr 26 is the chosen time period because precipitation is present 

that day. Six-hourly accumulations of pollutant deposition are recorded to coincide 

with the time resolution of the precipitation fields. The coordinate 48.0 0 latitude, 

11.0 0 longitude is the chosen release location because the spot is immediately upstream 

from Germany during the chosen period. If the modeled release were chosen 

at Chernobyl, it would not be possible to observe the immediate influence of ICS. 

ARL suggests a value of 3.2 x 10 5 for the user-specified ICS efficiency after 

Hicks (Hicks 86) and an empirical mean value of 5.0 x 10- 5 8- 1 for the BCS rate. 

To evaluate the sensitivity of HySPLIT's wet deposition scheme to ICS parameters, 

17

Deposition from OOz 27 Apr to 12z 30 Apr (UTe) 

12Z 25 Apr 86 ECMF FORECAST INITIALIZATION 

w 

o 

..... 

a 

CO? 

Z 

.. I".~j""." __ '/_" __ 

o 

...... 

1;; 

* c 

o 

.~ 

u 

.3 

Q) 

~ 

::J 

o 

if) 

.1.0E+04 

3.2E+Q4 MAXIMUM AT SQUARE 

1.0E-02 

Figure 6 Attempted duplicate of ARL's Chernobyl Cs-137 deposition. Source 

modeled as uniform vertical line source from 750m to 1500m at a rate of 

10 15 Bq/hr for 24 hours. Deposition velocity set at 0.1cm/8. In-cloud 

scavenging ratio set at 3.2 x 10 5 [/1. Below-cloud scavenging rate set at 

5.0 x 10- 5 8- 1 . Deposition contoured on a logarithmic scale. 

these ARL-recommended values are employed as the baseline values for the sensitivity 

control run. In addition to the sensitivity control run, a simulation is performed 

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 , 

3.04 X 10 5 , 2.88 X 10 5 . These values reflect boosts and reductions of the ICS efficiency 

by 1%, 5%, and 10%. HySPLIT control file settings for the sensitivity control run 

are recorded in Section D.3 of Appendix D. HySPLIT output concentration grid 

files are converted using HySPLIT utility program, 'con2bin.exe' to GRADS format 

for field differencing. Results of the sensitivity simulations accompanied by plots of 

simultaneous accumulated model precipitation are presented in Section 4.1. 

18

3.5 Modeled Cloud Base Modification 

In this Section, a method is presented for assessing performance of a modifiedcloud-base 

height parameterization in simulations of Chernobyl Cs-137 deposition. 

Subsection 3.5.1 develops a reasonable source term (emission specification). Subsection 

3.5.2 gives the motivation and procedure for the specific cloud base modification 

tested later. The available dataset lends itself best to two main model comparisons: 

a comparison to simulations of daily deposition 1986Apr28 - 1986May15 at 21 measurement 

sites spread across Europe described in Subsection 3.5.3, and a separate 

comparison to simulations of April-cumulative deposition from the onset of Chernobyl 

emissions at 2123Z, 1986Apr25, up to OOOOZ, 1986May01 at a cluster of 395 

measurement sites in Germany and Austria, described in Subsection 3.5.4. 

3.5.1 Daily Phases of Chernobyl Emissions. The best-guess Chernobyl 

source term (i.e., the emission specifications) is segmented into daily phases, except 

that the plume is treated separately during the first seven hours because it is believed 

to have risen significantly higher than subsequent emissions. The twelve phases, I 

through XII, used in this research are adapted from Table 1 of the ATMES Report 

(Klug 92:358). The ATMES Report provided a daily Cs-137 emission rate, specified 

a center of mass height for each phase of emission, and required that a step-function 

be used for the project's simulations. However, the ATMES organizers "still gave 

a certain degree of freedom to the participants, e.g. on the mass distribution with 

height" (Klug 92:2). According to the ATMES report, revised Russian release height 

estimates presented to the ATMES Steering Committee in January, 1989 are the only 

authoritative estimates (Klug 92:1,2). Today, though some documented evidence 

supports a change to the official release height, the source term estimate has not yet 

been updated by consensus (Graziani 00). Generally, it is accepted that the initial 

plume escaped the boundary layer, and that, after the first two days, the initial 

plume did not exceed an altitude of 400m (Persson 87). For this research, Phases I 

and II were recalculated (based on equal total emission amounts to OOOOZ, 27 Apr) 

19

Chernobyl Emission Rate in Twelve Phases 

+/- 50% Error Bars 

IOCS-137 [TBq/hr] I 

1400 

1200 +-----~~~------------------------------------------ 

1000 ~~-------------------------------------------- 

800 

600 

400 

200 

o ~~,-LJ-,L-~~~~~,L-L~~~LJ"L-~~L,~~,-~ 

Figure 7 

Complete Chernobyl Cs-137 modeled source term in twelve phases. 

Plume modeled as uniform vertical line source at an hourly rate in becquerels. 

Phase I initial plume 1250-1750m, Phases II and III initial 

plumes 350-850m, Phases IV through XII initial plumes 200-400m. 

as a compromise between ATMES' 26.0000Z 6-hr initial plume and the known time 

of Chernobyl's initial explosion, 25.2123Z. Each phase was modeled in HySPLIT 

as a uniform vertical line source. Figure 7 displays the resulting twelve Chernobyl 

emission phases. 

The first few hours of Chernobyl emissions are the release time period with the 

greatest vertical location uncertainty. It is agreed that the initial steam explosion 

and ensuing fire at Chernobyllaunched radioactive particles well above the accidentaveraged 

boundary layer top at roughly 500m. Pollanen et al. present evidence for 

a higher release height based on large particle trajectory calculations. 

"In northeastern Poland, 500-700 km from Chernobyl, particles up to 

rv 60 microns in aerodynamic diameter were found. Their sedimentation 

velocity is so large (up to rv O.lms- I ) that turbulent dispersion, rapid 

transport in a prefrontal low-level jet, warm frontal conveyor belt ... or 

20

even effective release height of 3000 m cannot explain these findings" 

(Pollanen 97:3581). 

A series of four preliminary 84-hr Chernobyl simulations is accomplished here 

to demonstrate how much the pattern of modeled pollutant deposition in Europe 

changes for a range of distinct, reasonable point source heights, namely, 1500m, 

3000m, 4000m, and 5000m. These preliminary simulations are not referenced outside 

of this section of the thesis. 

Each simulation's emission mirrors Chernobyl's 

initial emission rate of 6.65 x 10 14 Bq/hr (see Subsection 3.5.1 for a complete time 

profile of best-guess Chernobyl emissions). 

Other particle parameterizations are 

empirical estimates of those properties typical of Cs-137-bearing particles from a 

nuclear reaction. The four preliminary simulations are identical except for release 

height. Exact HySPLIT settings used for the 1500-m run are presented in Section 

D.1 of Appendix D along with detailed descriptions of each setting. These descriptions 

serve to familiarize the reader with HySPLIT concentration model setup. 

Should the reader consider downloading and using HySPLIT, further instructions 

are available on the internet from ARL (and more details are in Appendix D). 

surface deposition concentration grid is computed in each simulation over the lat/lon 

grid centered at 48°/13° and spanning 26° of latitude, 36° of longitude. The main 

difference between a 1500-m release (Figure 8) and a 3000-m release (Figure 9) is 

an overall decrease in deposition, presumably because the particles take longer to 

settle from a higher release point, and because greater wind speeds at 3000m carry 

more particles beyond deposition grid boundaries. 

A 

Changing the release height to 

4000m (Figure 10) produces a distinct southward shift in the 84-hr deposition pattern. 

Less pollutant deposits in the Nordic countries (e.g., none in Finland), while 

higher and more widespread pollutant concentrations are modeled from Ukraine and 

Romania to Italy, France and even Algeria. A release height of 5000m (Figure 11) 

produces a further southward shift in the deposition pattern: less deposition from 

Lithuania and Belarus to Sweden, more deposition from Ukraine and Romania to 

21

Italy and Algeria. For the reader's reference, Appendix E offers a map of Europe 

with current political boundaries. 

The pollutant is transported completely across the Mediterranean Sea into 

Algeria when the initial plume exceeds 3000m. The concentration of modeled 

deposition is on the order of only IBq/m2, well below background levels of 2000 

- 3000Bq/m2. However, modeled deposition underestimates hot spots due to precipitation 

field smoothing. So, if the model is broadly accurate in the region, detection 

would not be impossible, especially if daily measurements are available within 

deposition hot spots. Though beyond the scope of this work, this clue would be 

of special interest to those interested in refining the Chernobyl source term. There 

is a distinct shift in the deposition pattern to include deposition south and east of 

Chernobyl for release heights above 3500m. This change in general plume direction 

supports the view that release heights above 3000m would seriously alter the 

deposition pattern of Chernobyl's day one emissions. 

3.5.2 Modified-Cloud-Base Motivation and Procedure. Since empirical values 

for Cs-137 ICS efficiency and BCS rate have been documented, a logical place 

to look for wet deposition improvement is the cloud base parameterization since it 

directly determines vertically where modeled BCS stops and modeled ICS begins. 

The complexities of cloud formation are immense and many. Sophisticated prognostic 

cloud models are available, but are computationally expensive and gain little 

accuracy over diagnostic parameterizations since large uncertainties remain in the 

accounting of "advective transports of cloud variables, sub-grid scale processes, cloud 

microphysics, and cloud optical properties" (Tiedke 93:3040). Most current transport 

models and even some global meteorological models still use simple diagnostic 

schemes to model clouds. Future generations of transport models may just accommodate 

liquid and ice cloud fields from the input meteorological model rather than 

calculating their own cloud limits. For now, a parameterization is still needed and 

HySPLIT's simple scheme of cloud diagnosis from relative humidity is considered 

22

• 1.0E+04 • 1.0E+02 " 1.0E+OO 1.0E-02 

9,4E104 MAXIMUM AT SQUARE 

Figure 8 

Chernobyl deposition [Bq/m2] from OOZ 1986Apr26 to OOZ 1986MayOl. 

Emission from 2123Z Apr25 for 24hrs at Chernobyl (star in the graphic) 

at 6.65 x 10 14 Bq / hr from 1500-m point source height. 

w 

~ ',,1,; 

c:i 

C'l 

1ll 

* c: 

o 

] 

(I) 

[J 

~ 

"40 

/---- 

C"> 

III 

:0 

(tI 

ifi 

III 

III 

(tI 

(J) 

w 

~ 

(tI 

0.. 

~ 

~ 

N 

~ 

» 

"Q 

~ 

g 

·1.0E+04 

4,8E; 04 MAXIMUM AT SQUARE 

1.0E-02 

Figure 9 



at 6.65 x 10 14 Bq/hr from 3000-m point source height. 

23

LlJ 

0 

r- 

c:i 

C'l 

Z 

co 

iD 

·f5 

III 

to'> 

~ 

~ 

Iii ;:+ 

iii 

* 

e: ~ 

~ 

] }:- 

Q) "Q 

u 

0 

'.p 

N 

:s 

~ 

(") 

{IJ 

:0 

(tI 

(ll 

ID 

0- 

W 

M- 

C 

~ 

9 

1 ,2E t 04 MAXIMUM AT SQUARE 

1.0E-02 

Figure 10 



at 6.65 x 10 14 Bq/hr from 4000-m point source height, 

UJ 

0 

..- 

c:i 

:0 

(tI 

C'l " 

iD 

4:5 III 

Z 

{IJ 

41 

co 

rJ) 

~ 

Iii 

iii 

* 

e: ~ 

0 

'.p 

N 

~ 

...J }:- 

al "Q 

~ 

:J 

~ it 

(") 

{/) 

iif 

g 

0- 

~ 

~ 

~ 

;;,OEto.1 MAX.lMUM AT SQUARE 

1.0E-02 

Figure 11 



at 6.65 x 10 14 Bq/hr from 5000-m point source height. 

24

an over-generalization (Draxler OOb). To attempt an improvement to HySPLIT's 

cloud base parameterization simple enough to test in a short time, and as a starting 

point for exploring cloud base parameterization, a simple HySPLIT modification 

is proposed following work at NCEP (National Center for Environmental Prediction). 

As part of a comprehensive cloud model algorithm in NOAA's Meso ETA 

model, meteorologists at NCEP split the cloud base scheme into a marine part and 

a terrestrial part. It is believed that an 80% RH cloud base over continents represents 

too little cloud condensation. Cloud bases over land are modeled at 75% 

RH while cloud bases over water are modeled at 80% RH (Staudenmaier 96). HyS 

PLIT's hard-wired 80% RH cloud base is not unreasonable. If limited to a single 

value, long-range transport models should weight a unified scheme in favor of the 

marine environment since the earth's surface is mostly water. So, if NCEP's scheme 

approximates reality, 80% RH-modeled global cloud bases should outperform 75% 

RH-modeled cloud bases on a global scale. However, since HySPLIT has the ability 

to distinguish land use types, there is no need to compromise. The requisite cloud 

base modification in HySPLIT requires a change to the model source code as given in 

Appendix F. HySPLIT source code was provided for this thesis courtesy of Roland 

Draxler at ARL. Once the source code is edited and recompiled, the comparison 

runs can be accomplished identically to the April deposition control runs. Test results 

of April deposition control runs and April deposition modification runs against 

the April deposition data are presented in Section 4.2. 

3.5.3 Modified-Cloud-Base Procedures, Simulation of Daily Deposition. 

The available dataset used for comparing surface-based Cs-137 measurements to simulation 

data is from the REM (Radioactivity Environmental Monitoring) Chernobyl 

archives at the JRC /Ispra (Joint Research Centre - Ispra, Italy) of the EC (European 

Commission) (DeCort 90). Deposition from above-ground nuclear weapon tests in 

the mid 1950's and early 1960's has blanketed the surface of the entire globe with a 

thin layer of Cs-137. Just before the Chernobyl accident, typical Cs-137 concentra- 

25

tions on the ground in Europe were between 2000 and 3000Bq/m 2 (DeCort 98:15). 

Post-accident deposition measurements on the ground near that range of values cannot 

be attributed to Chernobyl with confidence unless measurements were recorded 

at the same site before the accident. Since measurement records were mainly near 

large cities and near other nuclear power plants, pre-accident measurement data is 

sparse and geographically irregular. These limitations prevent construction of a 

more complete surface-based measurement dataset. Aerial gamma spectrometry 

measurements were taken over Eastern Europe and Western Russia weeks after the 

accident and deposition maps of these measurements exist (DeCort 98). However, 

these data were not available for this research, so this work is based entirely on 

the available surface-based Cs-137 measurements (DeCort 90). The surface-based 

deposition data for the Chernobyl case, available from the JRC, are segregated into 

two distinct datasets: a daily deposition dataset described in this section, and a 

cumulative deposition dataset described in Section 3.5.4. Figure 12 identifies, for 

the period during and just after Chernobyl's emission, the 23 sites where daily deposition 

readings are available and the first date at each site that daily measurements 

were recorded. Section 4.2.1 presents a bar graph for each city with daily deposition 

measurements. 

Daily deposition data in the REM dataset is recorded at most sites in [Bq/m2]. 

Passau, Koblenz, Glasgow, and Berkeley reported deposition in [Bq/ LJ, i.e., becquerels 

per liter of rain. For calculation of daily deposition totals at these sites, the 

REM dataset supplies daily precipitation amounts in [mm] for some sites. For those 

sites with measurements in [Bq/ L] and no precipitation, precipitation at the nearest 

weather station is used to convert to [Bq/m2]. No weather station precipitation 

was recorded for Koblenz, so model precipitation amounts are used to convert measurements 

from [Bq/ L] to [Bq/m2] for the Koblenz data only. European weather 

station precipitation records are courtesy of the Air Force Combat Climatology Center 

(AFCCC). Precipitation is missing from the Koblenz record in the REM dataset 

26

; 

'" 

> 

,',-' 

..... '-~- _." -: ....... -" 

Figure 12 

European cities where daily Cs-137 deposition measurements were taken 

with measurement start dates. Dates 29-30 are April 1986, 01-08 are 

May 1986 

and from precipitation records supplied by AFCCC. Absent human-recorded precipitation, 

ECMWF model precipitation is substituted for the conversion of Koblenz 

deposition from [Bqj L] to [Bqjm 2 ]. HySPLIT limits pollutant release specifications 

to a constant emission rate, so twelve separate runs are required to model emissions 

from the entire accident, one for each phase of emission. Time series of surface 

deposition at measurement sites are extracted from the daily deposition control run 

output grids of each run and summed. To produce corresponding deposition time 

series modeled with the proposed cloud base modification, the process is repeated 

exactly as the daily deposition control run, but executed with a recompiled HyS 

PLIT model. The resulting daily deposition control and modification time series 

are presented and examined in Section 4.2.1. To allow the reader to reproduce the 

27

simulations, HySPLIT settings used to produce Phase I model output from the daily 

deposition control run are given in Section D.4 of Appendix D. 

3.5.4 Modified-Cloud-Base Procedures, Simulation of April Deposition in Germany 

and Austria. Measurement sites from the REM cumulative deposition 

dataset are depicted geographically in Figure 13, from ATMES Report Figure 8. It 

is evident from Figure 13 that the dataset's data points in Germany, Austria, and 

Greece are uniquely dense and homogeneous. Further examination reveals that the 

German, Austrian, and Grecian data in the cumulative dataset is also simultaneous. 

Because isolated measurement data points are not generally representative of a region 

due to unpredictable hot spots and holes in the long-range deposition pattern, the 

portion of the REM cumulative Cs-137 deposition dataset in these three countries is 

extracted for cloud modification run comparisons in this thesis. Unfortunately, the 

HySPLIT April deposition control run with best-guess Chernobyl source term and 

reanalyzed meteorological input data (Section 4.2.2) does not yield any deposition in 

Greece up to 86.05.01.00Z, while several separate measurements taken on that day 

in Greece indicate cumulative Cs-137 concentrations abqve 10 5 Bq/m 2 • Appendix G 

addresses possible reasons for the exclusion of Greece from modeled April deposition 

patterns in this thesis. May 1 German and Austrian deposition data remains the 

most homogeneous cumulative Chernobyl deposition data and is used exclusively for 

the April-cumulative model comparison (results in Section 4.2.2). 

Because HySPLIT only accommodates a constant emission rate, a separate 

model run must be accomplished for each phase of Chernobyl emissions to account 

for the total Cs-137 emission, then the deposition from each phase can be added 

together. To minimize the number of deposition simulations required to model Aprilcumulative 

Chernobyl deposition on Germany and Austria, air parcel trajectories 

are calculated for each of the first six phases of Chernobyl emissions. Figures 14 

through 19 are the modeled atmospheric trajectories of air parcels from Chernobyl 

during April (86.04.25.21Z - 86.04.30.24Z). The six figures correspond to the first six 

28

·

15 trajectories from 350m and 500m and Figure 16 trajectory from 850m imply that 

the only modeled Chernobyl accident air parcel trajectories that cross Germany or 

Austria in April originate from Phase II emissions (1986.04.26.04Z - 27.00Z) and 

Phase III emissions (27.00Z - 28.00Z). These three trajectories trace the plume's 

path northwestward toward the Baltic Sea under the influence of northeast high pressure 

before getting wrapped southwestward through Germany and Austria around 

the back side of a shallow, transient trough. The remainder of modeled April Chernobyl 

trajectories once again exhibit anticyclonic curvature, characteristic of a high 

pressure system, and curve away from Germany and Austria (Figures 17, 18, and 

19). April 28 and 29, the northeast high retreats eastward leaving a loose arrangement 

of very weak frontal boundaries. Without a strong pressure gradient to boost 

winds, the plume trajectories slow and meander more. By April 30, high pressure 

to the west begins to build and move eastward. 

Since only Phases II and III produced plumes over the area of interest (Germany 

and Austria), only two runs are required for each April cumulative deposition 

simulation. April deposition control run cumulative output concentration grids from 

HySPLIT are converted to GRADS format and summed in the GRADS program, 

available from the Institute of Global Environment and Society (IGES 01). The 

process is repeated using the recompiled HySPLIT model, and results for April deposition 

control run and April deposition modification runs are presented in Section 

4.2.2. Exact HySPLIT settings for April cumulative Phase II model output from 

the April deposition control run over Germany and Austria are given in Section D.5 

of Appendix D. To confirm that Phase I emissions did not diffuse from their mean 

path (i.e., the trajectories in Figure 14) all the way to Germany, the results of a 

full deposition simulation for Phase I appears in Figure 20. The results show that 

modeled Phase I emissions do not contribute in April to the initial 5-day deposition 

on Germany and Austria, but deposit instead largely in Belarus and Lithuania. 

30

w 

0 

ci 

C') 

z 

~ 

10 

-ru 

• t: 

0 

~ 

0 

0 

-I 

Ql 

!:? 

::I 

0 

C/) 

.. , .: 

.0 

, 

o 

s: 

m 

G) 

» 

! L ii . d • ...--;: .. 

Figure 14 

Modeled trajectories of Chernobyl air parcels in 6-hr increments. Originating 

at 2100Z on 25Apr19S6 from 500m (triangles to Sweden), 1000m 

(circles), 1500m (squares), and 2000m(triangles to Russia) 

31

--------------------------------, 

w 

0 

{) 

.... < 

ci 

l'TI 

t") 

(:> i:!; 

Z ~ 

co s: 

~ 0 

,... c!'. 

Lf) 0 

:l 

iii s: 

it ~ 

0 

r:: a. 

0 I 

~ 

0 

...J 

m 

G) 

Q) 

~ :t> 

:::I 

0 

en 

-.... 

& 1 100 

.r:: 900 

a 

?: 

Figure 15 


at 0400Z on 26Apr1986 from 350m (triangles), 500m (squares), 

and 850m (circles) 

32

LU 

o 

ci 

C') 

c 

o 

.,0:; 

ttl 

o 

....J 

QJ 

1\ 

, 

o 

s:: 

m 

G) 

» 

0.. ttl ~700 800 

..c: 900 

000 * ,. .Ic" 

Figure 16 


at OOOOZ on 27 Apr 1986 from 350m (triangles), 500m (squares), 


33

w 

- -56 

0 

< 

ci (1) 

en ?a 

40 42 ::l- 

34 36 ;38 n' 

z !!. 

C(I 

s;: 

~ 

!l 

..- O· 

It) 

:::I 

1U 

ir :T 

0 

c 

c.. 

0 

".j:; 

t'II 

0 

0 s:: 

0 

.J 

m 

(;) 

GI 

~ 

» 

:::I 

0 

(/J 

.. ~'. 

s;: 

(1) 

. ~ ~900 #I a I • • • • • I ' 

L: 000 

L a • 

• 

Figure 17 


at OOOOZ on 28Apr1986 from 200m (triangles), 300m (squares), 

and 400m (circles). 

34

w 

0 

ci 

(') 

Z 

~ 

10 

26 

'-- 

36 40 

"Iii 

ir 

t: 

0 

~ 

0 

0 

--l 

CII 

~ 

::J 

0 

f/) 

o 

So: 

m 

G) 

» 

..c ~~~ 000 

.. J • £! 3 

Figure 18 




35

ill 

o 

d 

M 

31 32 

33 

-+c 

c 

o 

'';:::; 

t1I 

o 

...J 

~ 

:;:) 

o 

C/J 

.................... ··········50·· 

o 

s:: 

m 

G) 

» 

c I. 

• 

IS 

•• 

Figure 19 



and 400m (circles). 

UJ 0 

w 

0 

0 

;u 

(') 

!l!. 

if) 

::z 1}: 

co 

III 

I"l [g 

lti 

~ 

'iii a. 

* 

~ 

~ 

13 -f 

(f) 

.D 

~ 

N 

C 

1.SE-(I5 \1A)(n~tJ\1 ATf.QI)ARF 

Figure 20 

Modeled Deposition of Phase I Chernobyl Emissions Accumulated Over 

04.25.21Z - 05.01.00Z 

36

IV. Results 

4.1 In-Cloud Scavenging Sensitivity Test Results 

The results of seven diagnostic HySPLIT sensitivity test runs are explored in 

this section. Design of the runs follows the method for diagnosing ICS sensitivity 

presented in Section 3.4. The seven runs include the sensitivity control run, 

three runs with ICS efficiencies boosted by different amounts, and three runs with 

ICS efficiencies reduced by different amounts. The first time period, 86.04.26.00Z - 

86.04.26.06Z, is examined in Subsection 4.1.1, then the second time period, 26.06Z- 

26.12Z is examined in Subsection 4.1.2. In the sensitivity test scenario, a plume of 

particles like those carrying Cs-137 from a nuclear event is initiated in southernmost 

Germany at 48.0 o N, 1l.ooE and travels north-northeast as evidenced in deposition 

plots to follow. In HySPLIT, the units of emission per hour translate to the units of 

surface deposition per square meter. So, emissions in [Bq/hr] translate to deposition 

in [Bq/m2]. 

4.1.1 ICS Sensitivity Over Germany, 86. 04.26. 06Z. Figure 21 provides 

6-hr-accumulated precipitation from the model to aide interpretation of deposition 

plots in Figures 22, 23, 24, 25, 26, 27, and 28. For example, Figure 21 identifies a 

dry area (no rain) in southeast Germany that corresponds to a deposition minimum 

between two maximums in the sensitivity control run deposition plot in Figure 22. 

The effects of boosting and reducing default ICS efficiency by various percentages 

in deposition model runs can be seen clearly by subtracting sensitivity control run 

deposition from each test run's results. The amount of deposition difference from 

the sensitivity control run for each boosted-ICS or reduced-ICS test run appears in 

Figures 23 through 28. 

In Figure 23 subtracting the sensitivity control run deposition from the (1% 

ICS efficiency boost) test run deposition yields a change in deposition on the order 

37

TPP6 ( mm) AT HEIGHT: 1.000 

• 1.00E+OO • 2.00E+OO • 3.00E+OO 

II 4.00E+OO 5.00E+OO r 6.00E+OO 

Figure 21 

Reanalyzed ECMWF model precipitation in [mm] accumulated from 

86.04.26.00Z to 86.04.26.06Z for comparison to in-cloud scavenging sensitivity 

run. Contours in 1-mm increments. 

of -O.5Bq/m2 near the source at 48.0DN, n.ODE. Logic dictates that deposition 

should instead be initially heavier when ICS is boosted, and initially lighter when 

ICS is reduced. The deposition change in Figure 23 implies that the immediate result 

of boosted ICS is decreased deposition. About lOOkm further north (downstream) 

increased deposition is observed. Two questions arise from these observations. 

The first question is, "How could deposition change amounts be opposite in sign 

if scavenging efficiency is increased only?" The second question is, "Why does it 

appear that increased scavenging immediately causes a decrease in deposition?" 

38

6-hr In-Cloud Default Accum. [Bq/m"2] 04.26.06z 

56N.-------~------~--~~~--~----. 

55N 

54N 

53N 

52N 

51N 

50N 

49N 

48N 

~ ... 

JQQ 

47N 

46N 

45NSE 

6E 7E 8E 9E 10E11E12E13E14E1SE16E17E18E19E 

Figure 22 

Six-hour-accumulated Cs-137 deposition [Bq/m2] from in-cloud scavenging 

sensitivity control run, 19S6.04.26.06Z 

In answer to the first question, the deposition differences from altered model 

scavenging can be opposite in sign upstream versus downstream from a point that 

could be termed the " scavenging error crossover point," or "SECP." At this unique 

point along the plume path, if total scavenging errors remain somewhat constant, 

the impact of the scavenging errors on deposition reverses. For example, if one 

assumes that modeled net scavenging is always and everywhere over-estimated, there 

must be a point in time and space where excess scavenging upstream has depleted 

the model plume so much that over-estimated scavenging downstream cannot make 

up for the concentration deficit in the plume. The resulting pattern of modeled 

deposition concentration would be too heavy upstream from the SECP and too light 

downstream from the SECP. Conversely, if net scavenging is always and everywhere 

under-estimated, the resulting deposition pattern would be too light near the source, 

and too heavy far from the source. The SECP principle applies as well to ICS 

errors alone if BCS and dry deposition are held constant as in these sensitivity tests. 

39

----------------------------------------------------, 

6- hr (In - Cloud + 1 %) - Default 04.26.06z 

56N.-----~~--------~--------------_. 

55N "- ---- 

54N 

53N 

52N 

, 

51N 

50~~ 

~ 

49N 

4BN 

«.~ 

-0.5 

47N 

46N 

45N+-~~--r-,-_._,--~~~~r_~_r_._1 

5E 6E 7E 8E 9E 10E 11 E 12E 13E 14E 15E 16E l7E 18E 19E 

Figure 23 

Difference in 6-hr-accumulated Cs-137 deposition [Bqjm 2 ]; Deposition 

from a test run with a 1% boost of in-cloud scavenging efficiency less 

deposition from control run, 86.04.26.06Z 

However, in the tests, ICS efficiency changes do not necessarily apply everywhere 

and always (e.g., a 1% boost of the default constant ICS efficiency may be an overestimate 

in some places and an under-estimate in others), so the SEep principle 

cannot be applied to model results blindly. In fact, the SEep principle cannot 

explain the answer to the second question. The SEep principle is noted, later, in 

Section 4.2.2 in an interpretation of deposition pattern changes from modifying the 

modeled cloud base. 

In answer to the second question, increased model scavenging appears to immediately 

cause decreased model deposition due to grid resolution and interpolation 

issues within the model. The feature of interest at 48.5°N, 11.0oE in Figures 23 

- 28, just north of the emission source, is not a physical phenomenon, but a computational 

one. HySPLIT only uses IeS efficiency (to calculate wet deposition) 

at gridpoints where precipitation is present. Since no precipitation is modeled at 

40

6-hr (In-Cloud - 1%) - Default 04.26.06z 

56N.-----~~--------~----------------_, 

55N 

54N 

53N 

52N 

51N 

50N 

49N 

48N 

....::=, 

( "':',>' 

\f,'-~?) 

('i/ 

0.5 

47N 

46N 

45NSE 

6E 7E BE 9E 10E l1E 12E 13E l4E l5E 16E 17E lEE 19E 

Figure 24 

Difference in 6-hr-accumulated Cs-137 deposition [Bq/m 2 ]; Deposition 

from a test run with a 1% reduction of in-cloud scavenging efficiency 

less deposition from control run, 86.04.26.06Z 

48.5°N, 1l.ooE, and the only difference between the sensitivity control run and each 

test run is the rcs efficiency parameter, there must be a reason for the feature that 

is unrelated to rcs efficiency. The scale of the feature, the symmetry of the feature 

with the feature just north of it, as well as the persistence of both features in the 

nearly identical 12Z runs, leads one to believe that the cause of the phenomenon is 

initial deposition grid interpolations within HySPLrT. The impact of the feature on 

cloud base modification test results in Section 4.2 is negligible because of its relatively 

small magnitude, and is irrelevant because no deposition observations in the 

dataset selected for this work are available near Chernobyl to diagnose the cause of 

the feature. Further investigation of the phenomenon is beyond the scope of this 

work. Finally, to answer the second question explicitly, the immediate decrease in 

deposition is not caused by increased scavenging, but instead is likely an artifact of 

HySPLIT's interpolation of continuous variables using a discrete 60-km grid. 

41

56N 

55N 

54N 

53N 

6-hr (In-Cloud + 5%) - 

_.' 

Default 04.26,06z 

52N 

If) 

51N 0 

50N 

49N 

48N -2 

'. 

47N 

46N 

45N 5E 6E 7E 8E 9E 10E l1E 12E 13E 14E l5E 16E 17E l8E 19E 

Figure 25 

Six-hour-accumulated Cs-137 deposition [Bq/m2] difference from control 

run with a 5% boost of in-cloud scavenging efficiency, 86.04.26.06Z 


56N.-----~~----c---~--------------_, 

55N 

54N 

53N 

52N 

51N 

J 

50N 

49N 

48N 

47N 

46N 

45N+-~-.--.-~~~--.-~~~--~~~~ 

5E 6E 7E BE 9E 10El1E12E13E14E15E16E17E18E19E 

Figure 26 


run with a 5% reduction of in-cloud scavenging efficiency, 

86.04.26.06Z 

42

6-hr (I n-Cloud + 10%) - 

Defa ult 04.26.06z 

56N,-------~--~----~---------------~ 

55N 

54N 

53N 

52N 

51N 

SON 

49N 

48N 

47N 

46N 

q;. 

-5 

45N+-~~--r_,-_r_,--~~~~r_~_r~~ 

SE 6E 7E BE 9E IOE11E12E13E14E1SEl6E17E18E19E 

Figure 27 




56N.---~--~--~----~---------------~ 

55N 

54N 

53N 

52N 

51N 

SON 

49N 

48N 

47N 

46N 

- , 

, 

( \. 

~,'0 

45NSE 6E 7E BE 9E IOE11E12E13E14E15E16E17E18E19E 

Figure 28 



86.04.26.06Z 

43

4.1.2 JCS Sensitivity Over Germany, 86. 04.26. 12Z. Because of the initial 

interpolation errors described in Subsection 4.1.1, a sensitivity test beyond the first 

six hours is required to estimate the impact of leS on total deposition. A 12Z sensitivity 

test, covering the period 86.04.26.06Z - 86.04.26.12Z, is presented here. Figure 

29 provides 6-hr-accumulated precipitation from the model to aide interpretation of 

the 12Z sensitivity test deposition plots in Figures 30 through 36. As in Subsection 

4.1.1, sensitivity control run deposition is presented first, in Figure 30, then six figures 

displaying difference plots where sensitivity control run deposition is subtracted 

from the deposition from each test run. The 12Z sensitivity control run deposition is 

an order of magnitude greater than the 06Z sensitivity control run deposition. Each 

difference plot in Figures 31, 32, 33, 34, 35, and 36 exhibits the initial interpolation 

error like those in Subsection 4.1.1. However, after the plume travels about 100km, 

the deposition pattern is straightforward. As expected, a boost of leS efficiency 

results in increased deposition as generally indicated in Figures 31, 33, and 35. A 

reduction in leS efficiency results in decreased deposition as generally indicated in 

Figures 32, 34, and 36. The response in the deposition pattern is approximately 

proportional to changes in the leS efficiency, e.g., the change in deposition caused 

by a 5% boost in leS efficiency (Figure 33) is approximately 5 times as much as 

the change in deposition from a 1% boost in leS efficiency (Figure 31). It is also 

noted that, assuming the sensitivity control run is truth, a plume with a 10% error 

in leS efficiency, traveling 600km in precipitation falling at 1mm/hr, does not reach 

its SEep. Otherwise, there would be a change in sign of downstream portions of 

the deposition difference patterns, at least in Figures 35 and 36. 

4.2 Modified-Cloud-Base Height Simulation Results 

4.2.1 Modified-Cloud-Base Performance Over Time. Daily deposition output 

from HySPLIT runs, produced as specified in Subsection 3.5.3, is presented in 

Figures 37 - 59 in order by earliest deposition measurement at each city. 

Raw 

44

Valid Time (UTe): 86/04/26/12 

~---n-..., 


II 1.00E+00 • 2.00E+OO • 3.00E+OO 

• 4.00E+OO 5.00E+OO; 6.00E+00 

Figure 29 

Reanalyzed ECMWF model precipitation in [mm] accumulated from 

86.04.26.06Z to 86.04.26.12Z for comparison to in-cloud scavenging sensitivity 

run. Contours in I-mm increments. 

45

6-hr In-Cloud Default Accum. [8q/m"2] 04.26.12z 

56N.-------~------~--~~~--~----, 

55N 

.--:- 

54N 

53N 

52N 

51N 

50N 

49N 

48N 

47N 

46N 

45N5E 6E 7E BE 9E 10El1E12E13E14E15E16E17E1BE19E 

Figure 30 

Six-hour-accumulated Cs-137 deposition [Bq/m2] from in-cloud scavenging 

sensitivity control run, 1986.04.26.12Z 

56N 

6-hr (In-Cloud + 1%) - 

Default 04.26.12z 

55N - 

54N 

53N 

52N 

51N 

50N 

49N 

48N 

47N 

46N 

45N 

5E 

6E 7E BE 9E 10El1E12E13E14E15E16E17E18E19E 

Figure 31 



46

---------------------------------- 

6-hr (In-Cloud - 1 %) - Default 04.26.12z 

56N~---~---~---------, 

55N 

54N 

53N 

52N 

51N 

50N 

49N 

48N 

47N 

46N 

45N5E 

6E 7E 8E 9E 10El1E12E13E14E15E16E17E1BE19E 

Figure 32 

Six-hour-accumulated Cs-137 deposition [Bqjm 2 ] difference from control 

run with a 1 % reduction of in-cloud scavenging efficiency, 

86.04.26.12Z 

6-hr (In-Cloud + 5%) - 


56N.--~-~---~---------. 

55N .

------------------------------------------ 


56N.--~-~-----~-------_. 

55N 

54N 

53N 

52N 

51N 

50N 

49N 

48N 

47N 

46N 

45NSE 

6E 7E BE 9E 10E l1E 12E 13E 14E lSE 16E 17E l8E 19E 

Figure 34 



86.04.26.12Z 

6-hr (In-Cloud + 10%) - 

56N , 


55N 

54N 

53N 

52N 

51N 

50N 

49N 

48N 

47N 

46N 

45N SE 

6E 7E BE 9E 10El1E12E13E14E15E16E17E18E19E 

Figure 35 



48

----------------------------------------------- 


56N '. 

55N 

54N 

53N 

52N 

J 

51N 

50N 

49N 

48N 

47N 

46N 

45NSE 

6E 7E 8E 9E 10El1E12E13E14E1SE16E17E18E19E 

Figure 36 



86.04.26.12Z 

measurement data (bars with shade gradient) for all 23 figures are from the REM 

databank at JRC - Ispra, Italy, and are based on total daily Cs-137 deposition 

measurements (DeCort 90). At many cities, gross features of modeled deposition 

distribution are in good agreement with those of measured deposition distribution. 

For instance, bimodal distributions are often indicated in both the measurement data 

and the modeled data with peaks synchronized to within about one day. The Pearson 

correlation coefficient between the entire daily deposition control run dataset and 

the daily deposition measurements is 0.5037. The Pearson correlation coefficient between 

the daily deposition modified-cloud-base run dataset and the daily deposition 

measurements is 0.5050. The formula for correlation follows, where n is the number 

of datapoints, and the variables X and Yare either daily measurement data and 

daily control run data, or daily measurement data and daily modified-cloud-base run 

49

data. 

Often, the daily deposition control run and daily deposition modified-cloudbase 

runs produce identical deposition, even days after emissions begin. Three 

different scenarios could produce identical daily deposition control and modifiedcloud-base 

amounts of deposition for a given day and site. If deposition for the 

whole day took place without the benefit of precipitation, no cloud base is calculated 

in HySPLIT, eliminating any cloud base modification effect. The second scenario 

finds precipitation on site, but no plume present between 75%-RH and 80%-RH levels 

above the deposition site, i.e., control run and modified-cloud-base run both find the 

plume either entirely beneath the cloud base, or entirely above the cloud base. The 

third scenario is a vertical resolution issue arising when the 75%-RH and 80%-RH 

are at effectively the same level. Such a discontinuity is possible in the model since 

the wet deposition algorithm uses discrete layers of humidity data. Even in some 

meteorological situations, such as a warm, moist air mass over-running a cold, dry 

air mass, a discontinuous relative humidity vertical profile is not an unreasonable 

approximation. 

50

-----------------~- 

Helsinki (Konala) (60.13,25.0) Daily Cs-137 Deposition [Bq/m A2] 

1 986 A P r2 8 - 1 986 May 1 4 

10000 

1000 

100 

10 

1 

91 

1901 

8 

f--- 

30 32 

1 19 20 19 18 20 

14 

- 

e-- - - I- e-- 1- - - 

1 

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 

Dmeasured 18 911 1901 87 30 32 19 20 19 18 14 55 301 20 111 27 8.4 

Oconlrol 

_modified 

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 

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 

301 

111 

27 

8.4 

~ 

17 

Figure 37 

Summary of Helsinki (Lat,Lon in decimal degrees) daily total Cs-137 deposition 

1986Apr27 - 1986May14; measured, control run, and modified 

cloud base run. 

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 

-1986May14 

10000 

1000 

100 

10 

1 

om e a su re d 

Dcontrol 

.m odified 

34 

15 

11 

9 

I-- - 

2 

117 

79 

~5 

41 

3 

f-- - . 

1 1 1 1 1 1 

rL 

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 

1 117 349 157 26 91 65 32 79 46 117 5 1 1 1 1 1 

1 1371 1012 693 133 213 27_9 65.2 30.2 1067 149 1.21 1.01 1 1 1 1 

1 1371 1022 687 135 208 29.4 64 29.8 1067 149 11.21 1.01 1 1 1 1 

---------------------- 

Figure 38 

Summary of Bratislava (Lat,Lon in decimal degrees) daily total Cs- 

137 deposition 1986Apr27 - 1986May14; measured, control run, and 

modified cloud base run. 

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) 

1 986 A P r2 8 • 1 986 May 1 4 

10000 

1000 

100 

80 

17 

76 

~ .~ J 

2 24 

391 

10 

1 

om easured 

Dcontrol 

.m odlfled 

- - 

1 1 1 1 1 1 

1 2 3 4 5 6 7 8 9 10 11 12 

- 

13 14 15 16 17 

1 24 801 171 53 76 59 53 24 52 391 58.8 1 1 1 1 1 

1 50122231 2601 119 53.1 46.6 41.3 21 673 29.9 1.2 1 1 1 1 1 

1 50122268 2578 107 52.1 47 40.3 20.8 673 29.9 1.2 1 1 1 1 1 

Figure 39 

Summary of Moravsky Krumlov (Lat,Lon in decimal degrees) daily total 

Cs-137 deposition 19S6Apr27 - 19S6May14; measured, control run, and 


H 0 f (50.31,11.93) Daily C 5·137 De p 0 5 itio n [8 q 1m A 2) 

1986 A P r2 8 • 1986 May 14 

10000 

1000 

100 

751 

14 

40< 

t------ I- 

9 

- 

3 

238 

10 

r--- - j--- -- r- 

1 

om e a su re d 

Dcontrol 

.m 0 d ifie d 

1 1 1 1 1111 1 1 

1 2 3 4 5 6 7 8 9 10 11 12 13 

1 751 146 30 92 462 238 1 1 1 1 1 1 

1 504 6634 389 156 50.5 218 29 18.7 25.5 2.35 1 1 

1 504 6634 406 145 59.4 205 28.4 18.9 25.1 2.34 1 1 

1 1 1 1 

14 15 16 17 

1 1 1 1 

1 1 1 1 

1 1 1 1 

~----------------------------------------------------------------------~ 

Figure 40 

Summary of Hof (Lat,Lon in decimal degrees) daily total Cs-137 deposition 

19S6Apr27 - 19S6May14; measured, control run, and modified 


52

~ 

r--- 

~ 

r--- 

Passau (48.58,13.47) Daily Cs·137 Deposition [Bq/mA2] 

1 986 A P r2 8 • 1 986 May 1 4 

10000 

1000 

84 

43 

100 

r--- 

75 

10 

1 

om easured 

Dcontrol 

.m odified 

r--- 

- - - - 

9.8 

6.52 

1 1 1 1 1 1 1 1 1 

~.~ ~ 

1 2 3 4 5 6 7 8 9 10 11 12 13 14 

1 844 439 1 1 1 1 1 1 75 9.8 6.52 1 1 

1 1989 62912317 832 19.3 103 20.7 16.1 35.3 4.07 1 1 1 

1 1989 62982364 778 23.2 99.4 20.1 16.2 34.9 4.07 1 1 1 

1 1 1 

15 16 17 

1 1 1 

1 1 1 

1 1 1 

Figure 41 

Summary of Passau (Lat,Lon in decimal degrees) daily total Cs-137 deposition 



Schwandorf (49.33,12.11) Daily Cs·137 Deposition [Bq/mA2] 

1986Apr28 ·1986M ay14 

10000 

1 000 

417 

814 

1 00 

- - 

12 

22 

78 

31 33 

1 

10 - 

3.2 

2.3 2.3 

1 

rt.1 

1 1 

1 

n n 

1 2 3 4 5 6 7 8 9 1 0 11 12 1 3 14 15 16 17 

om easured 1 417 479 123 814 22 47 78 31 16 3.2 1 2.3 33 2.3 1 1 

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 

.m odified 1 129568641761 863 32.9 1 06 1 2.1 1 3.5 27.52.08 1 1 1 1 1 1 

Figure 42 

Summary of Schwandorf (Lat,Lon in decimal degrees) daily total Cs- 



53

Hradec Kralov (SO.21,1S.83)Daily Cs-137 Deposition [Bq/mI\2] 

1 986 A P r2 8 - 1 986 May 1 4 

10000 

3101 

1000 

I-- 

241 251 

100 

10 

1 

Dm e a su re d 

oco"l.ol 

.modified 

8 

- - - _IU - 

17.6 

1 

- - I-- - 

1 1 1 1 1 1 1 1 1 1 

J ...... 

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 

1 1 3101 87 241 251 61 1 17.6 13 1 1 1 1 1 1 1 

1 39983011195413.7 3.88 67 .9 24.4 12.5 1711 154 1.21 1 1 1 1 1 

1 39983038193010.6 6.65 65 .8 23.6 12.5 1711 154 1.21 1 1 1 1 1 

Figure 43 

Summary of Hradec Kralov (Lat,Lon in decimal degrees) daily total 

Os-137 deposition 19S6Apr27 - 19S6May14; measured, control run, and 


Kosice (48.73,21.2S)Daily Cs-137 Deposition [Bq/mI\2] 

1 986 A P r2 8 - 1 986 May 1 4 

100000 

10000 

1000 

0 

100 

- 

.il 

32 

87 9 

10 

1 

omeasured 

Dcontrol 

.modifled 

rr 

nI - 

I-- 

1 1 1·~1 lr. 1 

... 

1 2 3 4 5 6 7 8 9 10 11 12 13 14 

lrJ 

1 1 405 55 32 4.2 1 41 41 87 91 1001 1 1 

1 926 954 482 7.64 15.7 1.61 1.83 9.27 56.5 1243178761.4 3.66 

1 926 954 482 7.55 15.3 1.55 1.82 9.27 56.5 1243178761.4 3.66 

1 

.-- 

1 1 

15 16 17 

1 1 1 

1.33 1 1 

1.33 1 1 

Figure 44 

Summary of Kosice (Lat,Lon in decimal degrees) daily total Os-137 deposition 

19S6Apr27 - 19S6May14; measured, control run, and modified 


54

----------------------------------------- 

Budapest (47.5,19.1) Daily Cs-137 Deposition [Bq/mA2] 

1986Apr28 - 1986M ay14 

10000 

1000 

100 

10 

1 

301 

239 

- f---s-: 

863 

1 32 124 

92 

--- 

2 27.4 

22 

3 

f-- - 

1 1 

1'1 -. 

1 2 3 4 5 6 7 8 9 10 11 12 13 14 

1 1 1 

15 16 17 

omeasured 1 1 301 547 239 61 92 132 28 13 863 124 27 22 1 1 1 

Dcontrol 1 331 185 262 72 168 9.3 24 59 275 915 1.9 2.4 1.2 1 1 1 

.m odified 1 331 185 263 78 162 10 23 58 275 915 1.9 2.4 1.2 1 1 1 

Figure 45 

Summary of Budapest (Lat,Lon in decimal degrees) daily total Cs-137 

deposition 1986Apr27 - 1986May14; measured, control run, and modified 


Mol (51.18,5.12) Daily Cs-137 Deposition [Bq/mA2] 

1986Apr28 - 1986M ay14 

10000 

1000 

1191 

371 341 

100 

.0 

f-- 

108 

.0 

10 

- f-- f-- 

16 

10.8 

1 

1 

1 

1 1 

1 f1 ~ 11 

1 

1 1 1 1 

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 

Dmeasured 1 1 1 58 119 371 108 341 16 1 58 11 1 1 1 1 1 

Dcontrol 1 1 1 3.5 39 57 1.8 4.4 2.3 1.1 1 1 1 1 1 1 1 

.m odlfied 1 1 1 3.5 40 57 2.1 I 4 2.3 1.1 1 1 1 1 1 1 1 

Figure 46 

Summary of Mol (Lat,Lon in decimal degrees) daily total Cs-137 deposition 



55

1 

Harwell (Chilton) (51.61,-1.3) Daily Cs-137 Deposition [Bq/mA2) 

1986Apr28 - 1986M ay14 

1 000 

1 00 

9 

10 - 

0 

4 

1 

2. 

2 

1.34 1.4 

1 1 1 1 1 1 1 

1 

1 1 

r n n D 

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 

omeasured 1 1 1 1 1 2.1 6 1 1 4 2 1.3 1 1.4 1 1 9 

Dcontrol 1 1 1 1 361 47 1 1 1 1 1 1 1 1 1 1 1 

.m odified 1 1 1 1 361 47 1 1 1 1 1 1 1 1 1 1 1 

~ 

Figure 47 

Summary of Harwell (Lat,Lon in decimal degrees) daily total Cs-137 deposition 



Aachen (50.76,6.1) Daily Cs-137 Deposition [Bq/mA2) 

1986Apr28 - 1986M ay14 

1 000 

1 7 9 1 99 

1 00 

- 

10 

1 

r--- 

20 

8 

2 

1 1 1 If) 1 1 1 

1 1 1 

'--- 

~ -- n 

1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14 1 5 16 1 7 

om easured 1 1 1 1 1 1 1 79 1 99 7 20 8 7 1 2 1 1 1 

Dcontrol 

.m odified 

I 

1 1 1 3.3 82 87 1 2 11 3.8 1 .2 1 

--- 

1 1 1 1 1 1 

1 1 1 3.3 84 86 1 2 1 0 3.8 1 .2 1 1 1 1 1 1 

Figure 48 

Summary of Aachen (Lat,Lon in decimal degrees) daily total Cs-137 deposition 



56

,------------------------------------------------, 

Emden (53.35,7.21) Daily Cs-137 Deposition [8q/mA2)1986Apr28· 

1986M ay14 

1000 ,------- 714 - 

------ 

250 

1 1 5 

1 00 78 79 

47 

32 

1 2 

1 0 - - - ------ f-- f--- 

1 

1 1 1 1 If 

1 1 1 1 

1 

1 2 3 4 5 6 7 8 9 1 0 11 12 1 3 14 1 5 16 17 

Dm easured 1 1 1 1 1 1 250 714 1 1 1 5 78 47 79 1 1 12 32 

Dcontrol 1 1 1 1 1 .7 1 62 1 66 53 11 4.6 1 1 1 1 1 1 1 

.m od ifie d 1 1 1 1 14 1 68 1 52 49 11 4.5 1 1 1 1 1 1 1 

Figure 49 

Summary of Emden (Lat,Lon in decimal degrees) daily total Cs-137 deposition 



Koblenz (50.35,7.6) Daily Cs-137 Deposition [8q/mA2) 

1986Apr29 - 1986M ay15 

1 000 

481 

253 

1 00 

- I-- 

1 22.9 

1 0 

1 

Om easured 

Dcontrol 

.m odified 

1 

27.4 

,~ 

- 

1 1 1 1 1 1 

-- 

~~ 

1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14 

1 1 1 1 1 1 48 1 2 53 1 123 27 1 1 1 

1 1 22 7 2 5 5 1 25 28 1 0 7.1 2 .5 1 1 1 1 

1 1 22 8 .1 2 62 1 1 8 26 9 .7 7.1 2 .5 1 1 1 1 

1 1 1 

15 1 6 17 

1 1 1 

1 1 1 

1 1 1 

Figure 50 

Summary of Koblenz (Lat,Lon in decimal degrees) daily total Cs-137 deposition 



57

Schleswig (54.45.9.53) Daily Cs-137 Deposition [Bq/mA2] 

1986Apr28 - 1986M ay14 

1 000 

8 54 

o • , 

-- 

1 00 

1 0 

1 

om easured 

Dcontrol 

.m odlfled 

40 

23 

26 

t 

17 

- - I--- 

1 1 1 1 1 1 1 1 

1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14 

1 1 1 1 1 1 8 54 1 583 40 23 26 1 7 1 

1 1 1 1 1 47 2 1 1 79 1 9 14 1 .6 1 1 1 

1 1 1 1 9 68 1 86 75 1 9 14 1 .6 1 1 1 

8 

- 

1 

15 1 6 17 

8 1 69 

1 1 1 

1 1 1 

Figure 51 

Summary of Schleswig (Lat,Lon in decimal degrees) daily total Cs-137 


cloud base run_ 

Bilthoven (52.11.5.18) Daily Cs-137 Deposition [Bq/mA2] 

1986Apr28 - 1986M ay14 

1 000 

22 1 

1 00 

S--7 

95 

46 

22 

1 0 

- 

5 

3 

1 1 1 1 1~ 1 

1 

1 -.1 1 1 1 

~ 0 

.or 

1 2 3 4 5 6 7 8 9 1 2 13 14 1 5 1 6 17 

Dm easured 1 1 1 1 1 1 67 9 5 221 22 1 1 1 1 46 5 3 

Dcontrol 1 1 1 1 2.7 94 2 1 1 5 2.8 1 .4 1 1 1 1 1 1 1 

_modified 1 1 1 1 4.1 94 2 1 14 2.8 1 .3 1 1 1 1 1 1 1 

Figure 52 

Summary of Bilthoven (Lat,Lon in decimal degrees) daily total Cs-137 


cloud base run_ 

58

,---------------------------------------------------------------------------~ 

Offenbach (50.2,8.65) Daily Cs-137 Deposition [Bq/m A 2) 

1986Apr28 -1986M ay14 

1 000 

1 00 

f-- ----6-6 

62-- 

33 

35 

1 0 

-- f-- f-- f-- 

13 

11 

r-- 

1 

Dm easured 

Dcontrol 

.m odified 

1 1 1 1 1 1 1 ~1 1 1 

1 2 3 4 5 6 7 8 9 1 0 11 12 1 3 14 

1 1 1 1 1 1 1 66 13 11 1 33 1 1 

1 1 1 58 46 260 1 57 76 16 11 4 1 1 1 1 

1 1 1 58 49 272 1 4 9 70 1 5 11 3.9 1 1 1 1 

1 

1 5 16 1 7 

62 35 1 

1 1 1 

1 1 1 

Figure 53 

Summary of Offenbach (Lat,Lon in decimal degrees) daily total Cs- 


modified cloud base run_ 

Glasgow (56.0,-4.49) Daily Cs-137 Deposition [Bq/mA2) 

1986Apr28 - 1986M ay14 

1 0000 

1 71 1 

1 000 

1 00 

1 3 1 

r-- r--- 

20 1 20 1 

1 6 

1 0 

-- 1- 1-- 1- -- 

1 

1 1 1 1 1 1 1 

1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14 

r:.. 

1 1 1 1 1 

1 5 1 6 17 

om easured 1 1 1 1 1 1 1 1 31 171 201 1 6 201 1 1 1 1 1 

Dcontrol 1 1 1 1 1 31 56 1 3 1.2 1 1.1 1 1 1 1 1 1 1 

.m ad ifie d 1 1 1 1 1 31 56 13 1.2 1 1.1 1 1 

I 

1 1 1 1 1 

Figure 54 

Summary of Glasgow (Lat,Lon in decimal degrees) daily total Cs-137 



59

Berlin (52.53,13.42) Daily C s-137 De positio n [B q/m A 2) 

1986Apr28 - 1986M ay14 

1 00 00 

1 000 

1 057 

285 

1 00 

--- -- 

45 

1 0 

1 

om easured 

Dcontrol 

.m odlfled 

- 

1 1 1 1 

1 .1r 

1 1 1 1 1 

1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 1 4 

1 1 1 1 1 1 1 1 105 285 45 1 1 1 

1 1 1 1 1 4.1 1 29 44 19 1 1 3 29 1 1 1 

1 1 1 1 2.5 24 1 09 42 19 1 1 3 29 1 1 1 

1 1 1 

15 16 1 7 

1 1 1 

1 1 1 

1 1 1 

Figure 55 

Summary of Berlin (Lat,Lon in decimal degrees) daily total Cs-137 deposition 



Giessen (50.58,8.67) Daily Cs-137 Deposition [Bq/mA2) 

1986Apr28 - 1986M ay14 

1 000 

743 

1 00 

- 

1 0 

5.3 

- - 

f-- 

1 1 1r. 1 • 1 1 1 1 1 1 1 1 1 1 1 

1 

r 

1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14 1 5 16 17 

Dm easured 1 1 1 1 1 1 1 1 743 5.3 1 1 1 1 1 1 1 

Dcontrol 1 1 1 .3 1 66 230 1 73 39 1 8 5.5 1 1 1 1 1 1 1 

.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 

Figure 56 

Summary of Giessen (Lat,Lon in decimal degrees) daily total Cs-137 deposition 



60

----------------------------------------------------------------------- 

M uenchen (48.13,11.5) Daily Cs-137 Deposition [Bq/mA2] 

1986Apr28 -1986M ay14 

1 0000 

1 397 

1 000 

- - 

1 00 

- - - 

72 

10 

- - - - I-- I-- 

r-- 

1 

Dm easured 

Dcontrol 

.m odified 

1 1 1 1 1 1 1 1 1 1[1 1 1 1 

1 2 3 4 5 6 7 8 9 1 0 11 1 2 1 3 14 

1 1 1 1 1 1 1 1 1 39 1 1 1 1 1 

1 199 1 85 1 52 1 1 9 34 32 1 2 14 3 1 1.7 1 1 1 

1 199 1 85 1 57 1 13 33 33 1 2 14 30 1.7 1 1 1 

1 1 

15 16 1 7 

1 1 72 

1 1 1 

1 1 1 

Figure 57 

Summary of Muenchen (Lat,Lon in decimal degrees) daily total Cs- 



Berkeley (51.69,-2.42) Daily Cs-137 Deposition [Bq/mA2] 

1986Apr28 - 1986M ay14 

1 000 

1 00 

1 0 

~ 

1 1 1 1 1 1 1 

'" n 1 ~ 1 1 1 

---- ~-- 

1 2 3 4 5 6 7 8 9 1 0 11 12 13 14 

1 1 1 

1 

1 5 1 6 17 

Dm easured 1 1 1 1 1 1 1 1 1.4 1 6.1 1 1 1 1 1 1 

Dcontrol 1 1 1 1 305 27 1 1 1 1 1 1 1 1 1 1 1 

.m odified 1 1 1 1 305 27 1 1 1 1 1 1 1 1 1 1 1 

Figure 58 

Summary of Berkeley (Lat,Lon in decimal degrees) daily total Cs-137 



61

------------------------------------------------- 

Risoe (55.7,12.07) Daily Cs-137 Deposition [8q/mA2] 

1986Apr28 - 1986M ay14 

1 000 

487 

_.- 

1 00 

1 09 

74 

2 3 

1 0 

1 

1 

1 

Dm easured 1 

Dcontrol 1 

.m odified 1 

I-- - - --- 

5 

2 3 4 5 6 7 8 9 1 0 11 12 1 3 14 

1 1 1 1 1 11 1 1 

~ 

1 1 1 1 1 1 1 1 487 1 09 74 23 5 

1 1 1 1 1 23 37 12 1 1 0 75 1 1 1 

1 1 1 1 5.3 19 37 12 1 1 0 75 1 1 1 

1 1 1 

1 5 16 1 7 

1 1 1 

1 1 1 

1 1 1 

Figure 59 

Summary of Risoe (Lat,Lon in decimal degrees) daily total Cs-137 deposition 



4.2.2 Modified-Claud-Base Performance in April Over Germany/Austria. 

Herein are described the results of two 5-day Chernobyl plume simulations and their 

deposition in Germany/Austria. The first simulation is an April deposition control 

run using default Chernobyl settings as described in Section 3.5.4. The second 

simulation is an April deposition cloud base modification run identical to the April 

deposition control run, except with 75%-RH cloud bases over land, instead of the 

default 80%. Deposition contours from the April deposition control run over the 

selected region are depicted in Figure 60. Total deposition from the April deposition 

cloud base modification run is shown in Figure 61. The modified-cloud-base deposition 

pattern exhibits generally greater deposition than the control run. Also, the 

contours are less smooth, indicating higher variability. The difference between the 

two fields, displayed in Figure 62, confirms the general increase in deposition with 

a lowered cloud base. A quantitative analysis is presented next, then qualitative 

analysis of major features of Figure 62. 

62

Cu Cs-137 to 1 May Phases 2+3 

Control 

56N~------~--~--~------~----~~~ 

55N 

53N 

52N 

50N 

49N 

48N 

47N 

45N+-~~--~~~--~-r~~~~~j~·.-.-~--~ 

5E 6E 7E BE 9E WE i 1 E 12E 13E 14E 15E i 6E 17E 18E 

Figure 60 April deposition control run. Modeled Chernobyl Cs-137 deposition 

concentration [Bq/m2] accumulated over 86.04.26.00Z - 86.05.01.00Z. 

63

Cu Cs-137 to 1 May Phases 2+3 

75% RH cloud base 

56N?-------~----~~-------------~~~ 

55N 

54N 

5.3N 

52N 

51N 

50N 

49N 

48N 

47N 

46N 

Figure 61 April deposition modified cloud base run. Modeled Chernobyl Cs- 

137 deposition concentration [Bq/m2] accumulated over 86.04.26.00Z 

- 86.05.01.00Z. 

64

Difference contours 

175% Runl - lControl Run~ 

56N~------~--~~~~--~~~~~~ 

55N 

54N 

5.3N 

52N 

51N 

50N 

49N 

48N 

47N 

46N 

Figure 62 

Difference in modeled Chernobyl Cs-137 deposition concentration 

[Bqjm 2 ] between cloud-base-modified run April deposition and control 

run April deposition (modified run deposition minus control run deposition), 

accumulated over 86.04.26.00Z - 86.05.01.00Z. 

65

Concentrations from each run are extracted at points corresponding to the 

395 April-cumulative measurement sites (Subsection 3.5.4) and are compared to the 

measurements. As may be suspected from earlier explanations, the cloud base modification 

is a relatively subtle adjustment and its impact in a Chernobyl simulation 

is masked by large errors in other parameters. For display and statistical purposes, 

a value of 1 was added to avoid taking the logarithm of zero, then the base-lO logarithm 

was taken at each data point, measured and modeled. Adding a constant to 

a distribution does not at all affect its correlation with another variable. A normal 

probability plot of log-transformed measured deposition values, Figure 63, strongly 

confirms the assertion by Rodriguez et al. that the distribution of surface concentration 

is log-normal (Rodriguez 95:811). Measurements were of total deposition, 

including any Cs-137 deposits prior to the Chernobyl accident. To adjust modeled 

quantities for pre-Chernobyl deposition the lesser of the corresponding measurement 

value or 2500Bq was added to each modeled data point. No null measurements are 

found in the REM database, so the data is log-transformed without adding a value 

of 1 to each data point. Correlation is 0.6059 for a point by point comparison of 

the log transformed 5-day Cs-137 measurements in Germany/Austria to log transformed 

deposition from the April deposition control run. Correlation is 0.5843 for 

the same comparison to the modified-cloud-base run. So, a slightly lower correlation 

to measurements is observed using modeled cloud bases lowered to the 75% humidity 

threshold. While cloud base parameterization improvement has not been shown 

for 5-day Chernobyl deposition in Germany and Austria, qualitative analysis of the 

results does unearth some clues to the possible role of wet deposition mechanisms at 

work in the Chernobyl case. 

66

Normal Probability Plot 

7 

- 

6 -c- 

• 

t/) 

m 5 +- 

-CI) 

E 

4 -c- 

0 3 ,,- 

"l""" 

C) 2 +- 

0 

1 -1-- 

0 I I I I 

0 20 40 60 80 100 

Sam pie Percentile 

Figure 63 

Normal Probability Plot of Log Transformed April Cumulative Cs-137 

Deposition Measurements in Germany and Austria 

The change from 80% to a 75% RH continental cloud base threshold in effect 

lowers the cloud base over land without changing the cloud top, the plume height, 

or the horizontal pattern of precipitation. In the sensitivity studies in Subsection 

4.1.1, the BCS rate was left undisturbed. With the cloud base lower, as long as 

the plume is near the cloud base, more of the pollutant will be within the cloud 

and less will be below it. So, for modified-cloud-base simulation runs, BCS applies 

to less of the plume, and ICS applies to more of the plume. BCS is dependent 

on rain duration which, in the model, is always 6hrs. ICS is dependent on rain 

amount which varies with each gridpoint, therefore, more variability appears in the 

modified-cloud-base run deposition pattern because more ICS is occurring relative 

to April deposition control run ICS. Since ICS always counts more in the cloud base 

modification run relative to the control run, the sign of the change in deposition 

67

depends on which scavenging provided more deposition, lCS or BCS. The fact that 

an overall increase in deposition is observed for a lowered cloud base indicates that 

modeled lCS is removing more Cs-137 overall than modeled BCS over Germany in 

April, 1986. While lCS is proportional to total amount of rain, BCS is proportional 

to the amount of time it rains. lCS tends to dominate the wet deposition from 

a ten-minute downpour, and BCS is the predominate deposition mechanism for a 

3-day drizzle. 

The general precipitation pattern over the sampled domain for late April is 

heaviest over Austria and gradually diminishing northward (Appendix C). The best 

opportunity for BCS to dominate lCS is where the model precipitation is lightest, 

north in this case. The patch of negative deposition difference values over the north 

half of the Germany/Poland border in Figure 62 is consistent with this thinking. 

Since this area is the most likely place for lCS to apply to more of the plume at 

the expense of the more dominant BCS. The other negative anomaly, over the 

heart of Czechoslovakia, is more difficult to assess. Since, between April deposition 

control and April deposition modified-cloud-base runs, no changes in scavenging 

occur above the 80%-RH level or below the 75 %-RH level, one only needs to assess 

what happens in the layer between those levels. Perhaps, during one or more 

precipitation events in the April deposition control run, BCS only slightly reduces 

pollutant concentration in the 75%-80% layer, leaving plenty of pollutant in the 

layer for deposition in a downstream location. Then, in a parallel April deposition 

modified-cloud-base run, during the same precipitation events, lCS depletes the 75%- 

80%-RH layer completely. So, the downstream location will have less pollutant 

available for scavenging, therefore, deposition will amount to less than that in the 

April deposition control run, as in the Czechoslovakia negative feature in Figure 62. 

If the April deposition modified-cloud-base run were truth, and the April deposition 

control run included the exact scavenging error, the SECP would be somewhere 

upstream from the negative feature. 

68

V. Conclusions 

5.1 Sensitivity Runs 

It has been observed that for a 6-hr time step and a polar stereographic computational 

grid as coarse as 60km in resolution, a HySPLIT transport and deposition 

simulation requires over 6hr s (one time step) to produce a realistic deposition pattern. 

Anomalous modeled deposition occurs within 100km of the source and is 

probably an artifact of interpolations used to initialize the pollutant plume. It has 

been shown that relative changes in deposition due to altering the ICS efficiency 

parameter in HySPLIT are nearly proportional to the ICS efficiency alteration. In 

other words, when the efficiency of ICS is doubled, a doubling of rain-out deposition 

is the result as long as pollutant concentrations are not significantly reduced. The 

acronym, SECP (Scavenging Error Crossover Point), has been coined describing the 

location at which the effect of scavenging errors on plume concentration bottoms 

out and begins to have the opposite effect. For example, over-scavenging initially 

produces excess deposition, but at the SECP no excess deposition occurs because 

the plume concentration has dropped enough. Downstream from the SEep, deposition 

is underestimated as the plume concentration continues to drop too fast. 

ICS efficiency sensitivity test simulations were run out to 12hrs and 600km in light 

rain (approximately Imm per 6hrs). At these limits, no SECP was apparent in the 

deposition pattern. Future studies may help refine the understanding of the SECP 

in general and of its possible range of influence on Chernobyl simulations. 

5.2 Cloud Base Modification Runs 

The predictive ability of HySPLIT appears to improve very slightly when modified 

to model continental cloud bases at 75% RH instead of its default 80%. The 

slight improvement is based on a higher correlation with measured data (0.5050) 

of the total bulk results of a modified-cloud-base daily deposition run than that 

69

(0.5037) of the total bulk results of the control daily deposition run. Given the 

combined severe uncertainties of Chernobyl emissions, of model precipitation and 

wind fields, and of the measurements themselves, and given that a 0.50 Pearson 

correlation is equivalent to raw guessing, the results in this format are inconclusive. 

However, an interesting clue arises from a city-by-city analysis. When correlation 

coefficients are calculated for each city, a trend of degraded predictive ability with 

time emerges. Figures 64 and 65 summarize the correlation of both control run and 

modified-cloud-base runs with measured data. The cities are in order by plume arrival, 

i.e., by earliest non-zero measurement recorded at each city. The cities where 

the modified cloud base has improved HySPLIT's predictive ability were Helsinki, 

Hof, Schwandorf, Hradec Kralov, Budapest, and Mol. There is virtually no change 

in predictive ability at Bratislava, Moravsky Krumlov, Passau, or Kosice (map of 

daily deposition cities in Section 3.5.3, Figure 12). So, the modified cloud base has 

performed slightly better than the control at the cities nearest Chernobyl (Figure 

64, except for Harwell) and the same as, or worse than, the control run at the cities 

furthest from Chernobyl (Figure 65, and Harwell). This trend could be an indication 

that the modified cloud base has induced an actual predictive ability that degrades, 

as expected, with distance from the source. 

Comparisons of HySPLIT control and cloud-base-modified runs of deposition 

in Germany and Austria in April, 1986 indicate a decrease in the predicitive ability 

of a HySPLIT Chernobyl simulation. Evidence of a SECP just upstream from 

Germany was examined in Section 4.2.2. These results are not inconsistent with the 

daily deposition run results since the modified-cloud-base run decreased accuracy at 

three of five German cities. 

By means of a HySPLIT cloud base parameterization revision (75%-RH continental 

cloud bases), very small improvement has been demonstrated in the accuracy 

of a HySPLIT Chernobyl Cs-137 daily deposition simulation, while a decrease in accuracy 

has resulted from a 5-day-cumulative Chernobyl deposition simulation over 

70

Correlation of Control Run & Cloud-Base-Modified Run Daily Deposition to 

Measured Daily Deposition 

j 

1.000 

0.800 r-- 

- 

- 

0.600 r- - r- -rr- 

f---- 

- 

I!! D.400 - r-- r- - - r-- r- r--= 

~ 

0.200 - - 

f--- f--- - f--- I- f--- 

0.000 L.- L- '- L- r-. L- L- 

- 

L- 

- 

Bratislav Moravsk Schwan Hradec Budapes 

Helsinki Hot Passau Kosice Mol Harwell 

a y dorf Kralov t 

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 

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 

Figure 64 

Set 1. Pearson correlation coefficients of modeled Chernobyl Cs-137 

daily deposition (control run and modified-cloud-base run) against measured 

Chernobyl Cs-137 daily deposition, by city. In order by earliest 

deposition measurement at each city. 

Correlation of Control Run & Cloud-Base-Modified Run Daily Deposition to 

Measured Daily Deposition 

1.000 

0.800 

c 0.600 

~ r- r- 

..!!! 0.400 r-- t-- t-- 

- 

~ 

u 

0 0.200 '--- 

t-- I-- - 

0.000 -- 

~ '---- 

-0.200 

r- 

KI 

~ LII 

Aache Emde Schle Koble Biltho Offen Glasg Giess Muen Berkel 

Berlin 

Risoe 

n n swig nz yen bach ow en chen ey 

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 

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 

Figure 65 Set 2. Pearson correlation coefficients of modeled Chernobyl Cs-137 

daily deposition (control run and modified-cloud-base run) against measured 

Chernobyl Cs-137 daily deposition, by city. In order by earliest 

deposition measurement at each city. 

71

Germany and Austria. The slightly improved HySPLIT performance in the daily 

deposition case does not prove indisputably that a 75%-RH continental cloud base 

is more true to the Chernobyl environmental conditions. A credible source term 

construction could be artificially devised such that 80%-RH continental cloud bases 

would produce the more "accurate" daily deposition. The 50% margin of error 

in Chernobyl's estimated daily emissions provides wide latitude to do so. Simulations 

of atmospheric plume transport and deposition are vulnerable to limitations 

in the representation of several aspects of the problem. These aspects include 

the particle size distribution and its space and time variations, turbulence, atmospheric 

instability, and other dry transport processes, the solubility of the particles, 

cloud formation and dissipation processes, fog deposition, and particle resuspension 

(Knap 88:48). Cumulative measurements representing all or several days of the 

Chernobyl deposition are likely to have a large positive bias since Cs-137 deposition 

measurement locations would tend to be where the highest radioactivity levels had 

been detected. The slightly positive results from the daily deposition simulation 

should in no way, then, be taken as proof of a general improvement in modeled 

cloud base. One can expect much larger gains in the accuracy of wet deposition 

modeling by improving the location, timing, and amount of precipitation inputs. It 

appears, in fact, that verifiable improvements to wet deposition parameterizations 

must wait both for increased resolution and accuracy of precipitation modeling, and 

for an experimental wet deposition case with more precise emission specifications 

and more homogeneous, higher resolution deposition measurements. Although the 

specific wet deposition parameterization test yielded no conclusive evidence of either 

better or worse performance, the exercise constitutes a meaningful starting point for 

a researcher interested in either refining the Chernobyl source term, or using Chernobyl 

data for validating transport or deposition mechanisms in a model where wet 

deposition is a factor, or learning how to use the HySPLIT model and becoming 

familiar with its capabilities and limitations. 

72

5.3 Future Research Opportunities 

HySPLIT lacks the modeling of fog deposition. No clouds (fog) are diagnosed 

in the surface layer in HySPLIT, so only dry deposition occurs within the surface 

layer in the model. Modification of HySPLIT to include accurate fog modeling 

and the increased surface layer deposition that results, especially in up-slope wind 

instances, and investigation into its impact on Chernobyl could bring model results 

more in line with Chernobyl deposition measurements. Fog parameterization is an 

even larger challenge than cloud parameterization, so unless approached carefully, 

adding fog deposition to a model could easily hurt the accuracy of modeled deposition 

more than it helps. One could probably make the same argument about cloud 

parameterization. Until liquid and ice cloud water content variables are available 

routinely from meteorological models, transport model cloud parameterization in 

general would still benefit from more accurate cloud diagnosis. The MRF model 

run by the National Weather Service (NWS) in addition to treating clouds differently 

over land and sea, makes finer cloudiness distinctions by relative humidity in 

predefined latitude regions and in four predefined vertical layers based on Real Time 

Nephanalysis (RTNEPH) data from USAF (NWS 01). Slingo (Slingo 87) poses and 

validates a more complex diagnostic cloud parameterization scheme accounting for 

relative humidity, vertical velocity and static stability (specifically, potential temperature 

change in the vertical). His approach holds promise for improving regional 

deposition distinctions between cumuliform and stratiform precipitation events. 

Further model comparisons to Chernobyl deposition should include as much 

measurement data as possible, increasing the span and resolution of observations in 

space and in time to strengthen confidence in results. Since surface-based observations 

near the source are scarce, aerial gamma spectrometry measurements taken 

over Russia could serve that purpose (DeCort 98). Even though there was about 

2000 - 3000Bq/m 2 of Cs-137 from weapons fallout on the ground before the accident, 

these readings could help improve model representation of the Chernobyl plume early 

73

in the accident because millions of Bqjm 2 of Cs-137 were deposited on Ukraine and 

Belarus (DeCort 98). The difficulties of source term uncertainty are not unique to 

Chernobyl. Transport analysts responding to any urgent, short-notice call for an 

emission simulation normally have only a crude estimation of the source term. Since 

the effective release height is highly dependent on static stability, one could investigate 

the modeled vertical profiles at Chernobyl and the nearest observed atmospheric 

soundings. Principles developed in this project could be applicable and valuable to 

operational simulations. The future of transport modeling, like the future of general 

meteorological forecasting, may look like ensemble forecasting. Motivated by a 

statistical view on stochastic processes like weather, an ensemble forecast is a set of 

simulations made up of a best-guess control run and a set of perturbation runs, each 

with a slightly different reasonable departure from the control run. An ensemble of 

forecasted patterns should provide helpful information about the spectrum of possible 

outcomes and about the confidence of any particular run (Draxler OOb). This 

method also provides an ongoing opportunity to generate further clues about which 

variables are important under specific synoptic regimes. 

74

--------------------------------------------- ._-- 

Appendix A. Glossary of Acronyms 

• AFTAC - Air Force Technical Applications Center 

• AGL - Above Ground Level 

• ARL - Air Resources Laboratory 

• ATMES - Atmospheric Thansport Model Evaluation Study 

• BCS - Below-Cloud Scavenging 

• CEC - Commission of the European Communities (or just EC, European Commission) 

• EC - European Commission (see CEC) 

• ECMWF - European Centre for Medium-range Weather Forecasts 

• ETA - Not an acronym; weather model named after coordinate system with 

vertical coordinate, eta (the Greek letter, rt) 

• GRADS - GRidded Analysis Display System 

• GRlB - GRldded Binary format 

• HTML - HyperText Markup Language 

• HySPLIT - Hybrid Single-Particle Lagrangian Integrated Thajectories 

• ICS - In-Cloud Scavenging 

• JRC - Joint Research Centre 

• NEA - Nuclear Energy Agency 

• NOAA - National Oceanographic and Atmospheric Administration 

• NCAR - National Center for Atmospheric Research 

• REM - Radioactivity Environmental Monitoring 

• RH - Relative Humidity 

75

• RTNEPH - Real Time Nephanalysis 

• SECP - Scavenging Error Crossover Point 

• UCAR - University Corporation for Atmospheric Research 

• USAEDS - United States Atomic Energy Detection System 

• USAF - United States Air Force 

76

Appendix B. Radioactivity Primer 

B.l Ionizing Radiation 

A radioactive atom is an unstable isotope characterized by the high-energy 

radiation its nucleus emits upon spontaneous decay (or disintegration) to a more 

stable state. This "ionizing radiation" packs enough energy to strip electrons from 

materials it hits. Ionizing radiation from a radioactive atom can take the form of: 

1. an alpha particle (2 protons with 2 neutrons, i.e., an electron-stripped 

helium nucleus) 

- can be shielded by a few inches of air 

2. a beta particle (stripped electron) 

- can be shielded by several inches of plastic 

3. gamma ray or x-ray (high-frequency electromagnetic wave) 

- can penetrate lead 

4. a neutron (stripped) 

- can penetrate thick lead shields 

Figure 66 (UIC 00) illustrates typical shielding requirements for each of the 

four types of ionizing radiation. 

77

Figure 66 

Typical Shielding Requirements for Different Ionizing Radiation Types 

from UIC, 00 

There are (at least) three ways to measure ionizing radiation. 

• Radiation Activity, a measure of the number of atomic disintegrations per unit 

time [e.g., in Bq = S-l] 

• Radiation Exposure, a measure of the amount of gamma or x-rays present [e.g., 

in coulombs/kg] 

• Radiation Dose, a measure of the amount of radiation absorbed by a subject 

[e.g., in sieverts] 

- See http://physics.nist.gov/cuu/Units/SIdiagram.html for an extensive summary 

of SI (International System) units. 

B.2 Cesium-13? 

The becquerel is a common unit of Cs-137 deposition radioactivity. One becquerel 

of Cs-137 is the amount of Cs-137 substance in which 1 unstable cesium atom 

per second undergoes atomic disintegration (emitting a beta particle and gamma 

radiation) (MSE 00). The average radiation dose in 1998 from 1kBq/m 2 of Cs-137 

deposited in 1986 is about 1 to 2f-LSv. Where soils are more conducive to human 

exposure, the average dose is closer to 20f-LSv (DeCort 98:22). Radioactive xenon 

78

gas, an abundant product of nuclear fission, decays to Cs-137 which then readily 

condenses onto particles present (Glasstone 77:389). Cs-137 decays to barium-137 

(Ba-137) with an ionizing radiation (beta particles and gamma rays) of O.662M e V 

per decay (Serway 92). The effective dose is highly dependent on the pathway 

(respiratory system, skin, digestive system). Although the health effects of exposure 

to Chernobyl's fallout should not be trivialized, to date the only clear evidence 

for a confirmed correlation between Chernobyl fallout dose and illness is thyroid 

cancer in children induced by exposure to the iodine isotope, 1-131. Because the 

detrimental health effects of Cs-137 are not sudden, and because of deficient human 

health records before the accident, it is difficult to isolate the effects of exposure to 

Chernobyl accident radiation from the existing widespread decline in the Russian 

population's general health. 

For emergency planning purposes (one application of atmospheric transport 

modeling), the uncertain concentration effects of local land use, runoff, and population 

habits (DeCort 98:22), combined with the uncertain health effects of radionuclide 

exposure/ingestion introduce enough uncertainty to cloud the relative importance 

of the magnitude of operational concentration estimates and the location and 

timing of radionuclide deposition. To illustrate, consider the evacuation planner 

who may be much more interested in which side of a mountain (and when) a radioactive 

plume may settle than in exactly how much fallout will land in a certain 

neighborhood. 

79

Appendix C. Reanalyzed Precipitation Fields from the ECMWF 

Model 

To facilitate informal diagnosis of wet deposition in simulations within this thesis, 

6-hrly model precipitation contours are furnished on the following pages. The valid 

time for each plot is the end of the six-hour accumulation. Graphics are produced 

with display.exe utility included with HySPLIT software. 86/04/31/00 UTC implies 

86/05/01/00 UTC. 

80

Table 1 

Reanalyzed ECMWF 6-Hour Precipitation Fields 1986Apr25. 

1986042506 missing, 

but not needed for comparison 

to Chernobyl deposition 


• 1.00E+OO .2.00E+OO • 3.00E+OO • 4.00E+OO 

• 5.00E+OO t. 6.00E+OO O.OOE+OO O.ODE+OD 

TPP6 ( mm) AT HEIGHT: 1,000 

• 1,OOE+OO • 2,OOE+OO • 3,OOE+OO III 4,OOE+OO 

• 5,OOE+OO r:;'" 6,OOE+OO O,OOE+OO O,OOE+OO 


• 1.00E+OO • 2.00E+DO • 3.DOE+OO III 4.00E+OO 

• 5.00E+OO '6.00E+OO O,ODE+OO O.OOE+OO 

81

Table 2 



III 1.00E+OO III 2.00E+OO III 3.00E+OO III 4.00E+OO 

m 5.00E+OO r" 6.00E+OO O.OOE+OO O.OOE+OO 

III 1.00E+OO III 2.00E+OO 

m 5.00E+OO ~X! 6.00E+OO 

III 4.00E+OO 

O.OOE+OO 


III 1.00E+OO III 2.00E+OO III 3.00E+OO III 4.00E+OO 

ij 5.00E+OO JZ.:. 6.00E+OO O.OOE+OO O.OOE+OO 

III 1.00E+OO III 2.00E+OO 

r'! 5.00E+OO v:' 6.00E+OO 

III 4.00E+OO 

O.OOE+OO 

82

Table 3 


mm) 

.1.00E+OO 

• 5.00E+OO r;;;£ 6.00E+OO 

.4.00E+OO 

O.OOE+OO 

• 1.00E+OO • 2.00E+OO 

• 5.00E+OO ~Y' 6.00E+OO 

.4.00E+OO 

O.OOE+OO 


• 1.00E+OO • 2.00E+OO • 3.00E+OO III 4.00E+OO 

• 5.00E+OO !hl 2 [ 6.00E+OO O.OOE+OO O.OOE+OO 

TPPS ( mm) AT HEIGHT: 1.000 

• 1.00E+OO • 2.00E+OO • 3.00E+OO III 4.00E+OO 

• 5.00E+OO ~;;2 6.00E+OO O.OOE+OO O.OOE+OO 

83

Table 4 


mm) 

• 1.ooE+OO • 2.00E+OO 

• 5.00E+OO 5tJ~ 6.00E+OO 

.4.00E+OO 

O.OOE+OO 

• 1.00E+oO • 2.ooE+oo 

• 5.00E+OO ~:i! 6.00E+OO 

.4.00E+OO 

O.OOE+OO 

NOAA AIR RESOURCES LABORATORY 


• 1.00E+OO • 2.00E+OO • 3.00E+OO .4.00E+OO 

• 5.00E+OO ~l;;~ 6.00E+OO O.OOE+OO O.OOE+OO 

• 1.00E+OO • 2.00E+OO 

iL~ 5.00E+OO ~::;; 6.00E+OO 

.4.00E+OO 

O.OOE+OO 

84

Table 5 



• 1.00E+OO • 2.00E+OO • 3.00E+OO • 4.00E+OO 

• 5.00E+OO ~F 6.00E+OO O.OOE+OO O.OOE+OO 


• 1.00E+OO • 2.00E+OO .3.00E+OO • 4.00E+OO 

• 5.00E+OO fi;i1! 6.00E+OO O.OOE+OO O.OOE+OO 

( mm) 

• 1.00E+OO • 2.00E+OO 

• 5.00E+OO I:~: 6.00E+OO 

iii! 4.00E+OO 

O.OOE+OO 

( mm) AT ,"~",o.rr 

• 1.00E+OO • 2.00E+OO 

• 5.00E+OO ~"I( 6.00E+OO 

iii! 4.00E+OO 

O.OOE+OO 

85

Table 6 



• 1.00E+OO • 2.00E+OO • 3.00E+OO • 4.00E+OO 

• 5.00E+OO r+:~: 6.00E+OO O.OOE+OO O.OOE+OO 

mm) 

• 1.00E+OO • 2.00E+OO 

• 5.00E+OO fjj';j 6.00E+OO 

.4.00E+OO 

O.OOE+OO 


TPPG ( mm) AT HEIGHT: 1.000 

• 1.00E+OO • 2.00E+OO • 3.00E+OO • 4.00E+OO 

• 5.00E+OO I~~ 6.00E+OO O.OOE+OO O.OOE+OO 


• 1.00E+OO • 2.00E+OO • 3.00E+OO • 4.00E+OO 

r;;; 5.00E+OO f:.' 6.00E+OO O.OOE+OO O.OOE+OO 

86

Appendix D. HySPLIT Settings 

This appendix provides exact HySPLIT model settings for all representative simulations 

run for this thesis. An attempt has been made to provide enough supplementary 

comments to enable the reader to reproduce the simulations in this thesis with 

a functioning version of HySPLIT. Australian Meteorological Magazine carried an 

article covering the general capabilities of HySPLIT Version 4, the transport modeling 

software used for this thesis (Draxler 98b). A complete description of the model 

including dispersion equations is available in NOAA Technical Memorandum ERL 

ARL-224 (Draxler 98a). The HySPLIT executable program and documentation is 

available for download at the following website: 

http://www.arl.noaa.gov/ss/models/gethysplit.html 

Initial settings for HySPLIT runs are adapted from settings used for Chernobyl 

simulation by Air Resources Laboratory. Some details are available online at: 

http://www.arl.noaa.gov/ss/transport/chernobyl.html 

D.l Release Height Sensitivity Runs 

Following this paragraph is a line-by-line breakdown of the 'Control' file settings 

for the HySPLIT diagnostic run from Section 2.5 with a 1500-m point source. 

HySPLIT 'Control' file (an ordinary text file) format requires that each numbered 

item appears (without the number) on a new line in the 'Control' file. Each numbered 

item below is followed by its description. Zeroes in line 27 would trigger 

calculations of gaseous emissions. The nominal 1.0 values in line 27 signal to the 

model that the pollutant is in particle form. A specified non-zero deposition velocity 

in line 28 eliminates the need for the model to calculate fall speed from particle 

attributes in line 27. See the HySPLIT User's Guide (Draxler 99) for more details 

87

on the 'Control' file settings. The only changes for the 3000-m, 4000-m, and 5000- 

m release height sensitivity simulations are the release heights in line 3 and unique 

output grid names in line 20. 

1. 86 4 25 21 

Simulation Starting Time [yy mm dd hh] 

2. 1 

Number of Emission Starting Locations (double it for uniform vertical 

line sources) 

3. 51.38 30.1 1500.0 

Emission Latitude [decimal degrees] Longitude and Emission Height 

[mAGL] 

4. 123 

Total Simulation Run Time [hours] 

5. 0 

Vertical Coordinate Type for Simulation Run (0 defaults to met. model's) 

6. 10000.0 

Ceiling, or Top of Model [m AGL] 

7. 1 

Number of Setup Meteorology Files 

8. D:/HySPLIT /hysplit4/metdata/ChernMet/ 

Path to Meteorology File 

9. analysisp. bin 

Filename of Meteorology File 

10. 1 

88

Number of Pollutants 

11. Cs 

Pollutant Identification 

12. 6.65E+ 14 

Pollutant Emission Rate [hr-l] (Concentration [m- 3 ] or Deposition [m- 2 ] 

output will be in these units) 

13. 24 

Hours of Emission 

14. 86 4 25 21 23 

Release Start Time [yy ma dd hh mn] 

15. 1 

Number of Output Grids to Generate 

16. 48.0 13.0 

Center of Output Grid Latitude [decimal degrees] Longitude 

17. 0.5 0.5 

Spacing (Resolution) of Output Grid Latitude [decimal degrees] Longitude 

18. 26.0 36.0 

Span (Length and Width) of Output Grid Latitude [decimal degrees] 

Longitude 

19 .• j 

Path Specification for Output Grid File 

20. 1500 m 

Output Grid Filename 

89

21. 1 

Number of Vertical Levels in Output Grid 

22. 0 

Height of Level [m AGL] (0 Triggers Deposition Calculation) 

23. 864 26 0 0 

Output Grid Sampling Start [yy ma dd hh mn] 

24. 86 5 100 

Output Grid Sampling End [yy ma dd hh mn] 

25. 0 120 0 

Output Grid Concentration Type (O=Average or l=Snapshot) and Interval 

[hh mn] 

26. 1 

Number of Deposition Setups (Must Match Number of Pollutants) 

27. 1.0 1.0 1.0 

Particle Diameter [11m], Density [g/ ee] and Shape Factor 

28. 0.0001 0.0 0.0 0.0 0.0 

Deposition Velocity [m/ s], Molecular Weight [g], A-Ratio, D-Ratio, 

and Effective Henry's Constant 

29. 0.0 3.2E+05 5.0E-05 

Actual Henry's Constant [M/atm] , In-cloud Scavenging Efficiency Ratio 

[L/ L], and Below-cloud Scavenging Rate [S-l] 

30. 10976.0 

Pollutant Radioactive Decay Half-life [days] (Airborne and Deposited) 

31. 0.0 

90

ReSllspension Factor [m -1] 

D.2 Comparison to ARL Chemobyl Simulation 

The following HySPLIT control file contents correspond to those used by ARL 

(ARL OOa) and discussed in Section 3.3. 

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, 

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, 

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, 

0.0 3.2E+05, 5.0E-05, 10976, ° 

D.3 In-Cloud Scavenging Sensitivity Control Run 

The following HySPLIT control file contents are used for the ICS sensitivity 

control run described in Section 3.4. 

86 4 25 21, 2, 48.0 11.0 1250.0, 48.0 11.0 1750.0, 147, 0, 10000.0, 2, E:j, 

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, 

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, 

.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 

0.0, 0.0 3.2E+05 5.0E-05, 10976.0, 0.0 

D.4 Daily Deposition Control Run 

The following HySPLIT control file contents are used for the cloud base modification 

control run described in Section 3.5.3. 

86 4 25 21, 2, 51.38 30.1 1250.0, 51.38 30.1 1750.0, 483, 0, 10000.0, 2, E:j, 

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, 

14.032.0, .j, dlyO 1. hyc , 1, 0, 86427 ° 0, 99 1231 00, ° 240, 1, 1.0 1.0 1.0, 0.0001 

0.0 0.0 0.0 0.0, 0.0 3.2E+05 5.0E-05, 10976.0, 0.0 

91

D.5 Chernobyl Control Run - Cumulative Deposition on Germany and Austria to 

OOZ, 1986MayOl 

The following HySPLIT control file contents are used for the cloud base modification 

control run described in Section 3.5.4. Control run resolution increased to 

0.05 deg lat & Ion for compatibility with display software, GRADS. 

Phase II Germany I Austria Control Run HySPLIT Settings 

864264,2,51.3830.1350.0,51.3830.1850.0,116,0,10000.0, 1, E:/Chernmet/, 

analysisp.bin, 1, Cs, 8.8E+14, 20.0, 8642640,1,45.518.0,0.050.05,21.026.0, .1, 

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 

3.2E+05 5.0E-05, 10976.0, 0.0 

Phase III Germany I Austria Control Run HySPLIT Settings 

864270,2,51.3830.1350.0,51.3830.1850.0,96,0,10000.0, 1, E:/Chernmet/, 

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, 

.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, 

0.0 3.2E+05 5.0E-05, 10976.0, 0.0 

D.6 Greece Diagnostic Run HySPLIT Settings - Emission 10m to 1750m 

The following HySPLIT control file contents, and minor variations on it, are 

used for the April Greece omission diagnostic runs in Appendix G. 

8642521,2,51.3830.110.0,51.3830.11750.0,123,0,10000, 1, E:/Chernmet/, 

analysisp.bin, 1, Cs, 7.66E+13, 123,8642521 0, 1,45.5 18.0, 0.50.5,21.026.0, .1, 

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, 

0.0 3.2E+05 5.0E-05, 10976, 0 

92

Appendix E. Political Map of Europe 

Faroe 

Isl~ 

" 

Figure 67 

Current Political Map of Europe 

93

Appendix F. HySPLIT Source Code Modification 

HySPLIT is configured to handle up to eleven land use types, i.e. one water type 

(type number 7), and ten terrestrial types. 

HySPLIT source code was provided 

for this thesis courtesy of Roland Draxler at ARL. Modifying the HySPLIT source 

code for the cloud base split parameterization requires a simple code change. The 

following is an excerpt from the HySPLIT subroutine, 'depelm.f' with all necessary 

modification. 

Additions to the original code include all full lines preceded with 

'ccc' and one 'ccc' at the start of the line following '0 L D COD E.' 

C test for wet removal processes 

IF(DIRT(KT)%DOWET.AND.RAIN.GT.O.O)THEN 

C determine bottom and top of the precip layer (80% to 60%) 

CCC ******************************************************* 

CCC with modification: (BASE AT 75% OVER LAND, 80% OVER SEA) 

CCC ******************************************************* 

KBOT=O 

KTOP=NLVL 

DO K=1,NLVL 

KRH=QQ(K)*100.0+0.5 

CCC ******************************************************* 

CCC CLOUD BASE MODIFICATION PROPOSED BY aaron@gimail.af.mil 

CCC SEE: http://nimbo.wrh.noaa.gov/wrhq/96TAs/TA9629/ta96-29.html 

CCC WESTERN REGION TECHNICAL ATTACHMENT 

CCC NO. 96-29 

CCC NOVEMBER 19, 1996 

94

CCC THE EXPLICIT CLOUD PREDICTION SCHEME IN 

CCC THE MESO ETA MODEL 

CCC Mike Staudenmaier, Jr. - WRH-SSD/NWSFO SLC 

CCC * * * * * * 0 L D COD E * * * * * * 

CCC IF(KBOT.EQ.O.AND.KRH.GE.80)KBOT=K 

CCC * * * * * * NEW COD E * * * * * * 

IF(LAND.EQ. 7.AND.KBOT .EQ.O.AND.KRH.GE.80)KBOT=K 

IF(LAND.NE. 7.AND.KBOT.EQ.O.AND.KRH.GE. 75)KBOT=K 

CCC ******************************************************* 

IF(KBOT.NE.O.AND.KTOP.EQ.NLVL.AND.KRH.LE.60)KTOP=K 

END DO 

95

Appendix G. Investigation of Greece Exclusion from Modeled April 

Cs-131 Deposition 

Since several cumulative Cs-137 concentration measurements exceeding 10 5 Bq/m2 

were recorded in Greece on 19S6MayOl, the conspicuous omission of Greece from 

April deposition patterns in this thesis bears investigation. No other simultaneous 

sources of atmospheric Cs-137 are documented, and Cs-137 arrives in Greece in 

the model early in May, so it is safe to assume that the measured deposition came 

from Chernobyl. This investigation considers three other possible explanations for 

the Greece exclusion from the cumulative control run. The first possibility is an 

over-estimated deposition velocity, i.e. the modeled dry fall speed is too fast. The 

distribution of particle sizes is not known nor is the change of the distribution in 

space and time, yet the modeled constant deposition velocity implies a uniform modeled 

size distribution. It's conceivable that actual particles of smaller aerodynamic 

diameter could have been carried further than the modeled plume before depositing. 

Another possible reason the model failed to show transport to Greece is trajectory 

looping of the actual plume, i.e. pollutant could leave the model domain and return 

(HySPLIT ignores particles that leave the domain of the meteorological model). 

Figure 17 suggests this is a strong possibility since the air parcels 'disappear' from 

the simulation (i.e, HySPLIT omits particles from any further calculations) when 

they cross the meteorological grid boundary at about 40.5 degrees longitude. The 

third possible reason for a modeled Cs-137-free Greece is a combination of modeled 

wind direction error and modeled wind speed error. Direction errors (especially 

near the source) or speed errors (especially in the vertical) could prevent the model 

from transporting pollutant to Greece by OZ, May 1. Since no compelling evidence 

for control run adjustments has come to light, settings for the simulation should 

not be tuned solely to provide for April deposition in Greece, and the data is ignored. 

Modeled emission release heights of up to 2990m and 9000m in diagnostic 

96

simulations yielded deposition nearer to Greece, but not in Greece. The first (wet) 

part checks different source term profiles in the vertical to isolate control run release 

heights that could result in Cs-137 deposited in Greece; none do. The second (dry) 

part employs the ultimate source term profile in the vertical to isolate control run 

plume layers that could result in Cs-137 deposited in Greece; none do. Reference 

Appendix E for a map of Europe with political boundaries and designations. 

G.l Evaluation of Source Term Height - Wei 

Since the source term height is uncertain, it is prudent to see if an error in the 

modeled source term height could be the cause of the Grecian deposition omission. 

Three diagnostic simulations are presented. Figure 68 is the result of a run set up 

the same as the 5-day cumulative control run, but with the emission in a uniform 

vertical line from 10m to 1750m. If the reason for Greece's omission was a control 

run plume base estimate that was too high, then a plume base low enough would 

lead to model deposition in Greece. Figure 69 checks the effects of centering, for the 

run's duration, the uniform vertical line source at 1500m, the recommended center 

of mass for the first six hours of emission (Klug 92:358). Figure 70 checks the effects 

of a source term extending to 9000m. None of these diagnostic simulations deposit 

pollutant in Greece. 

The contents of the HySPLIT Control file for the simulation in Figure 68 are 

recorded in Section D.6. Control file settings for the other two diagnostic simulations 

simply reflect the revised top of emission in line 4, and output grid filename in line 

21. For a summary of all primary settings in this thesis see Appendix D. For full 

details on settings see the HySPLIT User's Guide (Draxler 99). 

G.2 Evaluation of Source Term Height - Dry 

The exclusion of the Greece region from the model's control run output deposition 

fields is not a result of the model over-scavenging the plume by precipitation 

97


Deposition from 21 z 25 Apr to OOz 01 May (UTe) 

OOZ 01 May 86 ECMF INITIAL DATA 

.1.0E+05 

1.6E+0Cl MAXIMUM AT SQUARE 

1.0E-01 

Figure 68 

Greece Diagnostic Run, Cumulative April Deposition [Bq/m2], Release 

Height Profile from 10m to 1750m 

98


Deposition from 21 z 25 Apr to ODz 01 May (UTe) 

aoz 01 May 86 ECMF INITIAL DATA 

• 1.0E+05 • 1.0E+03 1.0E+01 1.0E-01 

1.5Et05 MAXIMUM AT SQUARE 

Figure 69 



99


Deposition from 21 z 25 Apr to OOz 01 May (UTe) 

ooZ 01 May 86 ECMF INITIAL DATA 

\' ~. 

( 1M2) 

• 1.0E+03 • 1 .OE+01 1.0E·01 1.0E-03 

6.0E ... 03 MAXIMUM AT SOUARE 

Figure 70 



100

washout upstream. Dry model runs (i.e., with precipitation fields absent from the 

meteorological input file) indicate that the modeled pollutant plume (even with a 

vertical line source from 10 to 14000m) did not reach Greece. Figures 71 through 

76 display slices of the total 5-day-averaged dry plume at several vertical levels. 

The output grids of this simulation imply that no particle emitted from Chernobyl 

at any height between 10m and 14000m at any time between 21Z1986Apr25 and 

24Z1986Apr30 floated over Greece. This result is consistent with the null Greece 

deposition of the 'wet' runs, because the meteorological model contains significant 

precipitation Greece-wide, especially late in the simulation (see precipitation fields 

in Appendix C). 

The dry runs were identical to the moist with these exceptions: 

• The input meteorological model lacked precipitation fields (identical otherwise). 

• All runs used the same source term layer (10m to 14000m). 

• The transport model top (ceiling) was set at 20000m instead of at 10000m (the 

default). 

• Each run recorded an average concentration field at a different level (one at 

Om, i.e., deposition, as in the wet runs). 

• 2000 particles were tracked in each simulation instead of the default 500 (set 

in file 'setup.cfg'). 

101

NOAA AIR RESOURCES LA130RATOIW 

Depos itio n from 21 z 25 Apr to 21 z 30 Apr (UTe) 

"1 Z 30 Ai>' 86 ECMF 1~ITIAl DATA 

40 (") 

'" 

:0 

!l!- 

iD 

m 

iD 

fn 

iii 

4 

'iD 

Q. 

W. 

~ 

N 

~ 

» 

~ 

S 

g 

~ 5E-Qt \IAXtMUM AT 5QtJARE 

Figure 71 

Five-day accumulated deposition due to Phase I emissions modeled as 

vertical line source from 10m to 14000m. Precipitation turned off. 


Average from 21z 25 Apr to 21z 30 Apr (UTe) 

21Z 30 Apt" 86 ECMF INITIAL DATA 

Figure 72 

Five-day average concentration at 1000m due to Phase I emissions modeled 

as vertical line source from 10m to 14000m. Precipitation turned 

off. 

102

--------------------------------------------------- 

NOM AIR RESOURCES LABORATORY 

Average from 21z 25 Apr to 21z 30 Apr (UTC) 

21 Z 30 Ap' 86 ECMF 1~ITIAl DATA 

u..J 

o 

g 

(') 

'" 

• 1.0E+OO ,. 1_0E-02 c- 1.0E-04 1.0E-06 

1.SE-ao \lAXII.r.J~' f


Average from 21z 25 Apr to 21z 30 Apr (UTC) 

21 Z 30 AV 65 EC MF IN ITIAl DATA 

'. 7E-ao \IAl

Bibliography 

ARL OOa. ARL. The Chernobyl Reactor Accident. 

http) /www.arl.noaa.gov/ss/transport/chernobyl.html, October 

2000. Air Resources Laboratory, Transport Modeling and Assessment 

- Silver Spring, MD. 

ARL OOb. ARL. ETEX Simulations Using HYSPLIT4. 

http)/www.arl.noaa.gov/ss/transport/etex.html, October 2000. 

Air Resources Laboratory. 

DeCort 90. 

DeCort 98. 

M. DeCort, G. Graziani, F. Raes, D. Stanners, G. Grippa & I. Ricapito. 

Radioactivity Measurements in Europe after the Chernobyl Accident, 

Part II, Fallout and Deposition. Rapport technique EUR 12S00 EN, 

Commission of the European Communities Joint Research Centre - 

Ispra Site, 1990. 

M. DeCort, G. Dubois, Sh D. Fridman & M. G. Germenchuk et al. 

Atlas of Caesium Deposition on Europe After the Chernobyl Accident. 

Luxembourg, Office for the Official Publications of the European Communities, 

1998. 176 pp. 

Draxler 9Sa. Roland R. Draxler & G. D. Hess. Description of the HYSPLIT4 modeling 

system. Rapport technique NOAA Tech. Memo. ERL ARL-224, 

National Oceanographic and Atmospheric Administration, December 

1998. 

Draxler 98b. Roland R. Draxler & G.D. Hess. An overview of the HYSPLIT4 modelling 

system for trajectories, dispersion and deposition. Aust. Met. 

Mag., vol. 47, pages 295-308, 1998. 

Draxler 99. 

Roland R. Draxler. Hybrid single-particle Lagrangian integrated trajectories 

(HY-SPLIT): Version 3.0 - User's guide and model description. 

Rapport technique NOAA Tech. Memo. ERL ARL-195, National 

Oceanographic and Atmospheric Administration, 5285 Port Royal Rd, 

Springfield, VA 22161, 1999. Available from National Technical Information 

Service, 5285 Port Royal Road, Springfield, VA 2216l. 

Draxler OOa. Roland Draxler, 2000. Electronic Message, 19Sep2000, ARL, Silver 

Spring, Maryland. 

Draxler OOb. Roland Draxler, 2000. Personal Communication, 23March2000, Air 

Force Technical Applications Center, Patrick AFB, Florida. 

Glasstone 77. Samuel Glasstone & Philip J. Dolan. The Effects of Nuclear Weapons, 

3rd edition. United States Department of Defense and United States 

Department of Energy, 1977. 653 pp. 

105

Graziani 97. Giovanni Graziani. European Tracer Experiment. 

http://www.ei.jrc.it/etex/. August 1997. 

Graziani 00. Giovanni Graziani, 2000. Electronic Message, 18sep2000, JRC, Ispra, 

Italy. 

Hagans 00. 

Hicks 86. 

IGES 01. 

Klug 92. 

Knap 88. 

Fred Hagans. http://www.aftac.gov/mission.htm. 2000. 

Bruce B. Hicks. Differences in wet and dry particle deposition parameters 

between North America and Europe. Aerosols: Research, 

Risk Assessment, and Control Strategies, pages 973-982, 1986. Atmospheric 

Turbulence and Diffusion Division, Air Resources Laboratory, 

ERL/NOAA, Oak Ridge, TN - Lewis Publishers, Chelsea, MI. 

IGES. GRid Analysis and 

http://grads.iges.org/grads/, January 2001. 

Environment and Society. 

Display System. 

Institute of Global 

W. Klug, G. Graziani, G. Grippa, G. D. Pierce & C. Tassone. ATMES 

Report, Evaluation of long range atmospheric transport models using 

environmental radioactivity data from the Chernobyl accident. Elsevier 

Science Publishers, England, 1992. 366 pp. 

Anthony H. Knap. The Long-Range Atmospheric Transport of Natural 

and Contaminant Substances. Kluwer Academic Publishers, 1988. 

NATO Advanced Science Institutes Series C: Mathematical and Physical 

Sciences - Vol. 297, 321 pp. 

Lauritzen 99. B. Lauritzen & T. Mikkelson. A probabilistic dispersion model applied 

to the long range transport of radionucleides from the Chernobyl accident. 

Atm. Environ., vol. 33, pages 3271-3279, 1999. Riso National 

Laboratory, 4000 Roskilde, Denmark. 

Metivier 95. Henri Metivier. Chernobyl Ten Years On, Radiological and Health 

Impact. http://www.nea.fr/htmljrp/chernobyljallchernobyl.html, 

November 1995. OECD Nuclear Energy Agency. 

MSE 00. 

NWS 01. 

Persson 87. 

Curie (measurement) Microsoft Encarta Online Encyclopedia 2000. 

http) /encarta.msn.com 1997-2000 Microsoft Corporation. All Rights 

Reserved, 2000. 

The Medium-Range Forecast Model status as of January 09, 2001. 

http://sgi62.wwb.noaa.gov:8080/research/mrf.html, 2001. National 

Weather Service. 

C. Persson, H. Rodhe & L. E. De Geer. The Chernobyl accident - a 

meteorological analysis of how radionuclides reached and were deposited 

in Sweden. Ambio, vol. 16, pages 20-31, 1987. 

106

P5llanen 97. Roy P5llanen, Ilkka Valkama & Harri Toivonen. Transport of radioactive 

particles from the Chernobyl accident. Atm. Environ., vol. 31, pages 

3575-3590, 1997. Finnish Centre for Radiation and Nuclear Safety 

(STUK), P.O. Box 14, 00881 Helsinki, Finland; and Finnish Meteorological 

Institute (FMI), P.O. Box 503, 00101 Helsinki, Finland. 

Rodriguez 95. D. Rodriguez, H. Walker, N. Klepikova, A. Kostrikov & Y. Zhuk. 

Evaluation of two pollutant disperson models over continental scales. 

Atm. Environ., vol. 29, no. 7, pages 799-812, 1995. University of California, 

Livermore, CA 94550, U.S.A.; and Institute of Experimental 

Meteorology, Obninsk, Kaluga Region 249020, Russia. 

Seinfeld 86. J. H. Seinfeld. Atmospheric Chemistry and Physics of Air Pollution. 

John Wiley and Sons, New York, 1986. 324 pp. 

Serway 92. 

Slingo 87. 

Raymond A. Serway. Physics for Scientists and Engineers with Modern 

Physics, 3rd edition. Saunders College Publishing, 1992. 1444 pp. 

J. M. Slingo. The development and verification of a cloud prediction 

scheme for the ECMWF model. Quart. J. Roy. Meteor. Soc., vol. 113, 

pages 899-927, 1987. European Centre for Medium Range Weather 

Forecasts, Reading, England. 

Staudenmaier 96. Mike Staudenmaier Jr. The explicit cloud prediction scheme in 

the Meso ETA model. Rapport technique Western Region Technical 

Attachment 96-29, National Weather Service Western Region Headquarters, 

Salt Lake City, 1996. 

Tiedke 93. 

UIC 00. 

M. Tiedke. Representation of Clouds in Large-Scale Models. Mon. Wea. 

Rev., vol. 121, pages 3040-3061, 1993. European Centre for Medium 

Range Weather Forecasts, Reading, Berkshire, England. 

UIC. Radiation and Life. http://www.uic.com.au/ral.htm. 2000. URA 

NIUM INFORMATION CENTRE, Ltd., A.C.N. 005 503 828. 

107

Vita 

Captain Aaron M. Kinser was born in Kettering, Ohio in 1966, and was raised 

in nearby Centerville. Centerville High School granted him a diploma in May, 

1984. In 1989 he enlisted in the US Air Force as a computer operator and served 

one tour assigned to the 354th Communications Squadron, Eielson AFB, Alaska. 

While assigned there he was selected for the Airmen Education and Commissioning 

Program (AECP) administered by the Air Force Institute of Technology (AFIT), 

Wright Patterson AFB, Ohio. Per AECP, Captain Kinser earned a Bachelors Degree 

in Atmospheric Sciences in 1996 from the University of Arizona in Tucson, Arizona 

and graduated from Officer Training School at Maxwell AFB, Alabama on December 

6, 1996. He then served as wing weather officer for the 57th Operational Support 

Squadron at Nellis AFB, Nevada until he was selected to earn a Masters Degree in 

Meteorology at AFIT starting in the Fall of 1999. Following graduation, Captain 

Kinser will be assigned to the Air Force Technical Applications Center (AFTAC), 

Patrick AFB, Florida. Captain Kinser can be contacted for the foreseeable future 

by text-only email ataaron@gimail.af.mil. 

108


Form Approved 

OMB No. 0704-0188 

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 

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, 

including suggestions for reducing this burden to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway. Suite 1204, Arlington, 

VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project 10704-01 BB1, Washington, DC 20503. 

PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS. 

1. REPORT DATE (OO-MM-YYYY) 12. REPORT TYPE 3. DATES COVERED 

03-2001 Master's Thesis Aug 2000 - Mar 2001 

4. TITLE AND SUBTITLE Sa. CONTRACT NUMBER 

SIMULATING WET DEPOSITION OF RADIOCESIUM FROM THE 

CHERNOBYLACCIDENT 

5b. GRANT NUMBER 

5c. PROGRAM ELEMENT NUMBER 

6. AUTHOR(S) 5d. PROJECT NUMBER 

Kinser, Aaron, M., Captain, USAF 

5e. TASK NUMBER 

Sf. WORK UNIT NUMBER 

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) S. PERFORMING ORGANIZATION 

REPORT NUMBER 

Air Force Institute of Technology (AFIT/EN) 

Bldg 640 Rm 100 

2950 PSt 

Wright Patterson AFB, OH 45433-7765 

AFIT/GMIENP/01M-05 

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR'S ACRONYM(S) 

Air Force Technical Applications Center 

AFTAC/TMAR 

Attn: Mr. Craig Sloan 

11. SPONSOR/MONITOR'S REPORT 

1030 S. Highway AlA 

NUMBER(S) 

Patrick AFB, FL 32925-3002 DSN: 854-7781 

12. DISTRIBUTION/AILABILITY STATEMENT 

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED. 


14. ABSTRACT 

In response to the Chernobyl nuclearle0wer plant accident of 1986, a cesium-137 de!i0sition dataset was assembled. Most of the 

airborne Chernobyl cesium was wet eEosited, either via interception blr falling rain rops or via abs0r.ption into cloud dn~tets 

destined to become raindrops. The Hy rid Single-Particle Lagrangian ntCfated Transport (HYSPLI ) model, develope at Air 

Resources Laborat02" is used to simulate the transport and deposition of ernobrl cesium-137. A cloud base barameterization 

modification is teste and ap~ears to slightly imfarove the accuracy of one HYSPL T simulation of dail Cherno (I cesium-137 

deposition over the course 0 the accident at iso ated European sites, and degrades the accuracy of anot h er HYSP IT simulation of 

deposition in Germany and Austria accumulated in the month of April 1986. Large uncertaintIes in the emission specifications, 

model precipitation fields, and deJ'0sition measurements prevent designating the results as conclusive, but most eVidence points to 

improved performance within 50 kilometers of the emission source. Trial and error lessons learned from hundreds of preliminary 

model runs are documented, and the exact HYSPLIT settings of successful and meaningful simulations are appended. 


Hybrid Single-Particle Lagrangian Integrated Transport (HYSPLIT) model, atmospheric transport modeling, wet deposition, 

cesium-I37 deposition, cloud base parameterization, Chernobyl 

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 

a. REPORT b. ABSTRACT c. THIS PAGE ABSTRACT 

U U U UU 

18. NUMBEF 

OF 

PAGES 

123 

19a. NAME OF RESPONSIBLE PERSON 

Lt Col Michael K. Walters, AFITIENP 

19b. TELEPHONE NUMBER (/nc/ude area code) 

(937)237-7166 

Standard Form 298 (Rev. 8/98) 

Prescribed by ANSI Std. Z39.18

Radiation Injuries after the Chernobyl 

Accident: Management, Outcome, and 

Lessons Learned 

Scientific Medical Effects of Ionizing Radiation 

(MEIR) Course 

July 30, 2008 

Alla Shapiro, M.D., PhD 

Medical Officer, 

US Food and Drug Administration 

Center for Drug Evaluation and Research 

Office of Counter-Terrorism and 

Emergency Coordination

Presentation Outline 

I. What could have been done at Chernobyl to lessen 

the effects of radiation damage? 

II. Health consequences 

• Thyroid gland and radiation exposure 

• Acute Radiation Syndrome (ARS) and its outcome 

• Neuro-psychological impact 

• Chernobyl experience of Cutaneous Radiation Syndrome 

(CRS) 

• Other medical problems caused by the accident 

III. Conclusions

What if….

Basic Information on the Radionuclide Releases 

and the Types of Exposure at Chernobyl 

• 100% of gaseous fraction of the noble gases and nuclides 

may have escaped from the plant 

• Cesium, Iodine and Tellurium isotopes accounted for up to 

10-20% of the nuclides inventory 

• Transuranic elements (Plutonium, Curium and Americium) 

were found only in the lungs 

• Neutron irradiation was not significant 

• ARS was caused by - and gamma-irradiation of the whole 

body and by beta-irradiation of the skin surface 

Ref: International Atomic Agency. Summary Report on the Post-Accident Review 

Meeting on the Chernobyl Accident, Vienna, 1986.

Sheltering 

• Sheltering is an effective preventive action in the 

area within a radius of 3-10 km from the point of 

the accident even in the case of absence of 

confirming radiation measurements 

• At Pripyat information about the need for 

sheltering was delayed 

• For other populations including Kyiv 

recommendations on sheltering were distributed 

on May 10, 1986 after the spread of radioiodine 

• The efficiency of sheltering could not be 

assessed

Relocation from the site 

The preliminary decision to evacuate the town of Pripyat, 

which is located less than 3 km from the ChNPP, was 

taken on the afternoon of 26 April 1986, when the dose 

rate in some parts of the town reached several mSv/hour 

By 9 pm on 26 April 1986, 1,350 buses, 2 railway trains 

and 3 motor ships were brought into the Chernobyl 

district (12 km from the town of Pripyat) 

At 10 pm the USSR Ministry of Public Health decided that 

the emergency evacuation of the town was necessary

Relocation from the site 

(continued) 

The organized evacuation of the town of Pripyat 

(49,360 including about 17,000 children and 80 bedbound 

patients), was carried out on 27 April 1986, 

between 2 pm and 5 pm

Iodine Prophylaxis 

Official information from 1986 

Total: It was administered to about 5 million people, including 

1.6 million children 

Pripyat town: It was administered to about 70% of the total 

population, including 60% on April 26 

Kiev Oblast: Department of the Ministry of Health made a 

decision on iodine prophylaxis on May 6, 10 days after the 

accident 

The Russian Federation: It was administered to 71,930 people, 

including 25,060 children, from June to the middle of August 

1986

Number of thyroid cancer cases in children and adolescents of Ukraine 

(aged 0-180 

yrs) at the time of the Chernobyl accident 

400 

350 

300 

250 

200 

150 

100 

50 

0 

19 

11 

8 

14 

11 

15-18 

0-14 

25 22 

118 

35 46 

62 69 20 23 

22 37 83 83 

42 46 

13 

11 

11 

24 

147 

118 129 43 

104 

251 

67 

192 

183 

81 

197 66 

117 

86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 

54 

138 

61 

136 

170 

284 

217 

249 

69 

180 

359 

85 

274 

350 

331 

88 

243 

105 

245 

370 

92 

278 

1

Distribution of thyroid cancer cases depending on 

patients’ age at the time of the accident 

50% 

45% 

40% 

35% 

30% 

1986 - 1989 

1990 - 1995 

1996 - 2001 

2002 - 2004 

25% 

20% 

15% 

10% 

5% 

0% 

0 - 4 5 - 9 10 - 14 15 - 18 

4

Papillary Carcinoma 

11

Chronic Thyroiditis 

350 

Controlled territories residents 

300 

Evacuees 

Recovery operation workers 

274.1 

306.1 

250 

225 

200 

150 

100 

50 

0 

182.1 

169 

147 

138.4 

126.2 

108.3 

115 

96 

86.4 82.8 84.7 

75 

57.7 

34.2 

9.7 13.5 16.9 18 

19 

21 

25 

6.5 

187.4 

27.3 

1992 1993 1994 1995 1996 1997 1998 1999 2000 

Data of Ministry of Health of Ukraine

Leukemia 

• Consensus exists on the absence of leukemia 

excess in inhabitants of the contaminated 

territories (French-German Initiative study) 

• There is a controversy in data on the leukemia 

incidence in children exposed in utero 

• Preliminary data from 2003 - 2005 demonstrate 

dose-effect relationship in operation recovery 

workers irradiated over 100 mGy (US Natl. Cancer 

Institute-RCRM joint study of leukemia among clean-up 

workers of Chernobyl in Ukraine)

Mutation AML1 in ARS Survivors with 

Myelodisplastic Syndrome 

During L1 gene sequenation in ARS 

survivor, who had suffered MDS, it was 

revealed punctuated mutation as 

repeating of 6 nucleotides 

Patient No 24 / AML1 wt 

Patient No 263 / appel with ins 

1502 (CGGCAT) 

T S G I G I G I G M S A M 

consecution with mutation

Conclusions 

Studies in Ukraine have shown: 

• an excess of thyroid cancer and non-cancer 

thyroid disease in children & other exposed 

groups (recovery operation workers,evacuees, 

adult population) 

• a controversy in data on the leukemia 

incidence in children exposed in utero 

• a dose-effect effect relationship on the leukemia 

incidence in recovery operation workers 

exposed to over 100 mSv

Conclusions (continued) 

• an increase in the breast cancer 

incidence rate in females participating 

in recovery operation works in 1986/87 

and female subpopulation still living in 

the most contaminated areas 

• an increase in all forms of cancer 

incidence rate only among recovery 

operation workers of 1986-1987 

1987 in 

comparison with national level

Liquidators in Action

Chernobyl, April 26, 1986: Sequence of the 

Initial Intervention 

Time Intervention Treatment/Disposition 

30 min – 

3-4 hours 

Initial treatment on the site 

• evacuation from the site, 

antiemetics, sedative, 

cardiotonic 

4 hours - 

12 hours 

12 hours - 

36 hours 

Evaluation and treatment at 

the nuclear plant medical 

facility 

Specialized team arrived 

• discharged if condition is OK 

• remained hospitalized 

• Assessment, blood tests, 

administration of KI, priority for 

hospitalization established 

Guskova et al, 1986. Acute radiation effects in exposed persons at the Chernobyl 

Atomic Power Station Accident. Medical Radiology, (477), pp.3-18

Sequence of the Initial Intervention (continued) 

• Within the first three days, 299 persons were sent to the 

specialized treatment center in Moscow and to hospitals in 

Kiev 

• Over the subsequent days hundreds of additional persons 

were admitted for examination 

• Criteria for hospitalization included for patients with the 

suspected ARS 

» Presence, time of onset and intensity of nausea and vomiting 

» Primary erythema of the skin 

» Decrease of the lymphocyte count in the peripheral blood 

Primary Diagnostic Criteria of ARS: 

Diagnostic Coefficient (DC) 

Assessment of irreversible myelosuppression according to DC in cases of ARS 

Time to the onset of vomiting Hours 

0-0.4 

0.41-0.8 

0.81- 1.2 

1.21 – 1.6 

>2.0 

Diagnostic Score 

+8 

+4 

+2 

-2 

-10 

Lymphocyte count 10 9 x1-1 Diagnostic Score 

Lymphocyte count on Day 2 0-0.2 

+6 

Lymphocyte count on Day 2 0.61- 0.8 

-15 

Lymphocyte count Days 4 – 7 0.01 

+5 

Lymphocyte count Days 4 – 7 >0.15 

-15 

A sum of +10 is the basis for the prognosis of irreversible myelosuppression; a 

sum of -10 is a prognosis for NO irreversible myelosuppression.

The Severity and Outcome of ARS in 

Chernobyl Victims 

ARS 

Dose 

Number of Patients 

Grade (Gy) Total Alive Died (days to death) 

I 0.8 -2.1 31 31 0 

II 2.0 - 4.0 43 42 1 96 

III 4.2 - 6.3 21 14 7 16 - 48 

IV 6.0 –16.0 20 1 19 14 - 91 

TOTAL 115 88 27 

Baranov et al, Antibiotics and Chemotherapy, 1989, 34, 7, 555-558; Guskova et al, 

“Acute Radiation Effects in Exposed Persons at the Chernobyl Atomic Power 

Station Accident” Medical Radiology, 1986, pp. 3-18.

The Bone Marrow Syndrome and its Treatment 

in Chernobyl Victims (1) 

• Antiseptic regimen 

• Supportive therapy 

» Isolation 

» Air sterilization 

» Changes of underclothing for patients at least once/day 

» Maintaining the micro-organism population at less than 500/mm3 

in the room air 

» Antimicrobial decontamination of the intestine 

» Administration of systemic antibiotics 

» Acyclovir 

» Transfusions of blood cells (e.g. fresh donor platelets and RBC) 

Ref: UNSCEAR Report, 1986; Robert Gale et al, The Lancet, April 23,1988. Guskova A. et al, 1986.

Supportive Therapy for Neutropenia 

• Oral quinolones, fluconazole, acyclovir 

prophylactically 

• Standard care for hemopoietic failure 

• All blood products irradiated at 25 Gy 

• Other supportive measures ad libitum

The New Concept of the ARS 

6-8 Gy 

( MODS) 

( MOF) 

1 Gy 

4 Gy 

SUBCLINICAL 

BONE MARROW 

(SOF) 

Reversible if 

heterogenous irradiation 

30 Gy 

50 Gy 

INCREASING DOSE

Bone Marrow Syndrome and its Treatment in 

Chernobyl Victims (2) 

• HLA-matched unrelated bone marrow donors from 

large HLA-typed volunteer donor pools – 13 patients 

• Fetal liver cells – 6 patients 

• Bone marrow syndrome combined with other Injuries 

• Skin 

• GI 

• Oropharyngeal 

• Radiation pneumonitis

Number of Deaths from Direct Radiation 

Effects in first 3 months 

Number of patients 

died (TOTAL = 27) 

Days of death after 

the exposure 

22 14 - 34 

Comments 

In 20/22 patients -burns 

were the main cause of 

death 

5 48 – 99* 

Died after the bone marrow 

recovery stage 

* Patient on Day #96 died 

from ischemic stroke 

Baranov et al, Antibiotics and Chemotherapy, 1989, 34, 7, 555-558; Guskova et al, 

“Acute Radiation Effects in Exposed Persons at the Chernobyl Atomic Power 

Station Accident” Medical Radiology, 1986, pp. 3-18.

Indications for an Allogenic BMT or an 

Embryonic Live Cell Transplantation 

• Whole body 

-irradiation dose 6.0 Gy -16.0 Gy 

• Irreversible degree of myelosuppression using a 

Diagnostic Coefficient (DC) 

plus additional criteria 

• Vomiting during the first 30 minutes 

• Diarrhea during 1-2 hours after the exposure 

• Swelling of the parotid glands during the first 24-36 hours 

Ref: UNSCEAR 1988 Report

Outcome (Survival or Cause of Death) 

in Patients Receiving BMT 

Dose 

range (Gy) 

Bone marrow transplant patients 

Number of 

patients 

Deaths* Deaths** Number of 

survivors 

< 6.5 4 0 3 1 

6.5 – 9.0 3 2 1 1 

> 9.0 6 5 0 0 

TOTAL 13 7 

*skin and GI 

injuries 

4 

**GVHD + 

infection 

2 

Gale et al, 1988; UNSCEAR report, 1988; Baranov et al, 1989; Guskova et al, 1989;

Hemopoetic Stem Cell Transplants ??? 

• Never an emergency! 

• Not if MODS! 

• Always consider heterogeneity of irradiation and 

possibility of autologous hemopoietic recovery 

• HLA typing immediate 

• Transplant never before day 14-21 

• Low immunosuppression: fludarabine ± ATG 

• High cell dose 2x10 6 CD34/kg (peripheral blood), 

2x10 8 nucleated cells/kg (bone marrow) and 3x10 7 

nucleated cells (cord blood)

Problems that Complicated the Use of 

BMT for Chernobyl Victims 

• Determination of the radiation dose 

• Several kinds of irradiation (external -and , and 

inhaled and ingested isotopes) 

• Partial shielding of body parts by physical structures 

• Rapid onset of lymphocytopenia made HLA typing 

difficult. Donor-recipient histocompatibility was not 

accurately determined 

• Most individuals who received a sufficiently high dose 

of irradiation had thermal burns as well as injuries to 

the GI tract and other tissues

Causes of Death among ARS Survivors 

(1986 through 2006) 

Cause of death 

Oncological and 

oncohematological pathology 

Grade I 

ARS 

1 

Grade II Grade IIII Total 

ARS ARS 

2 2 5 

Sudden cardiac death 2 2 2 6 

Internal organ systems and 

neurological diseases 

1 

3 1 5 

Traumas and accidents 2 - - 2 

TOTAL 6 7 5 18

Oncological Diseases in ARS Survivors and 

non-ARS Patients 

No Group Diagnosis First revealed 

Outcome 

1 non-ARS Sarcoma of hip soft tissues 1992 Died in 1993 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

non-ARS 

non-ARS 

non-ARS 

non-ARS 

non-ARS 

non-ARS 

non-ARS 

non-ARS 

ARS 1 d. 

ARS 2 d. 

ARS 2 d. 

ARS 2 d. 

Leiomyosarcoma of shin 

Cancer of colon 


Cancer of kidney 

Cancer of stomach 

Cancer of stomach 

Cancer of lung 

Cancer of prostate 

Cancer of throat 


Cancer of thyroid gland 

Cancer of thyroid gland 

Neurinoma of lower jaw 

1998 

1999 

2001 

2000 

2004 

2004 

2001 

2001 

2000 

1997 

2000 

2000 

2003 

Operated in 1998 


Died in 2005 


Died in 2004 

Died in 2005 


Died in 2003 

Died in 2001 




Died in 2004

Non-bone Marrow Syndromes 

Caused by Radiation Exposure 

Acute 

Radiation 

Syndrome 

Skin burns 

(%) 

Oropharyngeal 

Syndrome 

(%) 

Gastrointestinal 

(%) 

Radiation 

Pneumonitis 

(%) 

115 56 

(48.6) 

80 

(69.5) 

17 

(14.7) 

7 

(6.1) 

Barabanova A., Vojnosanit Pregl. 2006 May;63(5):477-80 

Ministry of Health, Clinical Department of the Institute of Biophysics, Moscow, 

Russia. abarabanova@rambler.ru

Stages of CRS 

Stage Onset Symptoms 

Prodromal 24-72 

hours 

Manifestation Days – 4 

weeks 

Transient erythema, pruritis 

Intense erythema, edema, pruritis, pain, 

blisters, erosions, ulcerative necrosis 

Subacute 4-6 weeks Erythema, edema, ulcers 

Chronic 

3 months- 

2 years 

Keratosis, fibrosis, ulcer, atrophy, 

pigment alteration, subcutaneous 

vasculitis, ulceration 

Late Decades Ulcers, angioma, fibrosis, keratosis, 

basal cell carcinoma 

Stages of the CRS according to Second Consensus Development Conference on 

the Management of Radiation Injuries, Bethesda, MD, 1993

Early and Late Skin Lesions in Radiation-exposed 

Patients after the Chernobyl Accident 

ARS 

(Grade) 

Number 

of 

patients 

Body area 

Early skin 

lesions 

(1986) 

Late skin 

lesions 

Basal Cell 

Carcinoma 

(BCC) 

I 5 

feet, LE, 

trunk, hands 

Erythema, 

edema 

Atrophy, pigment 

alteration, xerosis 

II 6 

LE, UE, 

trunk + LE 

Erythema, 

edema 


alteration, xerosis, 

keratosis, ulcers 

III 9 

Combinations 

of the above 

Erythema, 

edema, 

blisters, ulcers 


alteration, fibrosis 

keratosis, ulcers 

IV 1 

trunk + 

extremities 

Blisters, ulcers 

Same as Grade III 

plus carcinomas 

2 BCC 

lesions 

Nonconfirmed 

group 

1 

TOTAL 22

Chronic Cutaneous Radiation Syndrome (CRS) 

Patient N. In 1986 suffered from severe ARS (3rd degree) and moderatesevere 

acute skin damage (2nd - 3rd degree) of right foot. Essential 

keratosis and fibrosis. Nail bone of 1 finger was amputated in 1986, the 

focus with transplanted skin are well visible. 

Courtesy of Drs Belyi and Bebeshko, Kiev, Ukraine


Patient K. In 1986 suffered from severe ARS (3rd degree) and severe acute 

radiation skin damage of both shins (3rd degree). On the frontal surface foci 

of hyperpigmentation and telangiectasis are visible (15 years had passed) 



Patient N. In 1986 suffered from severe ARS (3rd degree) and moderate 

acute skin damage (2nd degree) of right leg. After 15 years following skin 

changes dominate: telangiectasis, hyperpigmentation, keratosis, fibrosis. 


Treatment Experience of Skin Injuries in 

Chernobyl victims 

• Systemic treatment 

Hemoperfusion, plasmapheresis, continuous 

heparinization and administration of freshly frozen 

plasma 

• Local treatment 

Use of Combutec-2 for local treatment of skin injuries 

Aerosol Lioxanol 

Solution Balis-2 

• Pain management 

was challenging and not effective due to an 

absence of the local anesthetics in the treatment arsenal 

• Necessity of surgical operations at an early stage 

Guskova et al, 1988, Baranov et al, 1991, Selezneva, 1990, Barabanova, 2006

Non-radiological Health Effects 

• Psychological effects 

– Can overwhelm 

radiological physical effects 

– Symptoms are similar for 

different radiation 

emergencies (different 

scales) 

– Need for comprehensive 

strategy directed to 

different population 

groups before/during/after 

an emergency

“The largest public health problem 

unleashed by the accident is the mental health 

impact” 

(WHO report of the UN Chernobyl expert group, August 2005) 

• Stress-related symptoms 

• Chronic Fatigue Syndrome 

• Effects on the 

developing brain 

• Organic brain disorders 

in highly exposed 

clean-up workers 

• Suicide

Brain Damage in Clean-up Workers 

“Today it is recognized that the Central 

Nervous System (CNS) is a radiosensitive 

organ whose degree of dysfunction can be 

quantified by electrophysiological, 

biochemical and/or behavior parameters. 

Abnormalities in CNS function defined by 

these parameters may occur at a low dose of 

whole body radiation”

Impact of Low Level Radiation on 

Brain Development 

1. Children irradiated in 

utero in the first 4-5 

months of gestation 

have: 

• reduced verbal IQ at age 

11 

• ECG changes in the left 

hemisphere 

2. Treatment with low dose 

radiation in infancy leads 

neuropsychological 

disorders later in life 

120 

118 

116 

114 

112 

110 

108 

106 

104 

102 

100 

p

Schizophrenia Incidence 

Excess in Chernobyl 

Exclusion Zone Personnel 

6 

5 

4 

3 

2 

1 

Chernobyl 

exclusion zone 

Ukraine 

Linear 

approximation 

(Exclusion zone) 

0 

At Issue: Schizophrenia Spectrum Disorders in Persons Exposed to Ionizing Radiation as a Result of the 

Chernobyl Accident by Konstantin N. Loganovsky and Tatiana K. Loganovskaja Schizophrenia Bulletin, 

26(4):751–773, 2000.

Neuro-Psychological Consequences: 

Summary 

• Genetic predisposition to schizophrenia can be provoked by 

environmental stressors including effects of exposure to ionizing 

radiation 

• Left hemisphere is 

the most radiovulnerable 

• Neuroimaging abnormalities 

are revealed following exposure 

to >0.3 Sv 

• The CNS effects that could be attributed to exposure to ionizing 

radiation are as follows: schizophrenia spectrum disorders; 

hronic Fatigue Syndrome; accelerated aging processes and 

neurodegeneration; and suicide

‰ 

40 

35 

30 

25 

20 

15 

10 

5 

0 

The risk of development of cerebrovascular 

diseases is higher in recovery operation workers 

with doses of 0.25 Gy and higher as compared to 

those with an exposure of less than 0.1 Gy 

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 

Years 

- < 0,05 Gy; - < 0,25 Gy; - 0,25-0,7 Gy 

trend: < 0.25 Gy trend: < 0.05 Gy trend: 0.25 - 0.7 Gy

Lessons Learned: Twenty years of Follow-up after 

the Chernobyl Accident (1) 

• Cutaneous component of the ARS had significantly 

complicated the clinical prognosis and contributed to or 

caused death in patients 

• Severe beta-burns of the skin remain an unsolved problem 

as a result of their spreading 

• The severity of the skin damage could have been avoided 

by removing the contaminated clothing 

• The prevention of late skin effects depends upon the 

effective management of acute lesions

Lessons Learned: Twenty years of Follow-up 

after the Chernobyl Accident (2) 

• Communications: who/how to contact, how to verify and 

confirm information 

• Confidentiality: different understanding of what was 

classified and what was not, limited information available 

for International professional community 

• Public health implications of the radiological accident: 

International significance was not as well understood as for 

communicable diseases incidents



• The outcomes and late effects of the skin lesions depended 

on the depth-dose distribution and on the size of the area 

affected 

• Radiation-induced fibrosis is a predominant clinical problem 

• Appearance of secondary ulcerations presents treatment 

challenges 

• No malignant melanoma or squamous cell carcinoma have 

been detected so far

"An accident has occurred at Chernobyl 

nuclear power station. One of the atomic 

reactors has been damaged. Measures are 

being taken to eliminate the 

consequences of the accident. Aid is 

being given to the victims. A government 

commission has been set up."

Be clear of what you are trying to say!



• Effective medical care is generally not possible for accident 

victims with high-dose TBI 

• Most individuals will not receive a sufficiently high dose to make 

a bone marrow transplant necessary for hematological recovery 

• Only a small number of patients will have bone marrow 

syndrome without other life-threatening non-bone marrow 

related complication 

• Transplants should probably be considered for victims receiving 

more than 7 to 8 Gy of external radiation



• Maximize the education of physicians 

• Provide medical community with practical tools how to 

identify and assess radiation victims 

• Explain situation in plain language and avoid conflicting 

information 

• Stay in touch with collaborating centers in 

European and other countries experienced 

in managing radiation emergencies

Invisible danger still exists

What was the most unexpected for us? 

• Diversity of clinical manifestations of skin lesions 

• Unaccustomed course of clinical phases of a 

radiation injury to skin 

• Significant severity of injuries 

• Serious influence of skin burns on the general state 

of a patient 

• Need for surgical interventions at an early stage

LUCK FAVORS THE PREPARED! 

Thank You!

Acknowledgements: 

• M.Tronko, T.Bogdanova et al, G. Thomas et al, 

Institute of Endocrinology and Metabolism, Acad. Med. Sc., Ukraine; Swansea University, 

UK; University, Japan; Institute of Oncology and Radiology, Acad. Med. Sc., Ukraine’’. 

• D.A.Bazyka, V.G.Bebeshko, D. Belyi, A.E.Romanenko, V.A.Buzunov, A.E.Prysyazhniuk, 

K.M.Loganovsky, M.I.Omelyanets 

Research Centre for Radiation Medicine (Kyiv) 

• Ihor J. Masnyk, Ph.D., NCI, Epidemiology BranchU.S. Director Chornobyl Research 

Projects 

• M.N. 

• Savkin, L.A. Ilyin, A.K. Guskova 

• State Research Center – Institute of Biophysics, Moscow, Russia 

• Albert L. Wiley, MD, PhD, Director, Radiation Emergency Assistance Center Training Site 

(REAC/TS) 

• Patrick Gourmelon, T.M.Fliedner and V. Meineke, Institute for Radiation and Nuclear 

Safety (France), University of ULM (Germany) 

• Pierre Flor-Henry, Konstantin Loganovsky, Alberta Hospital Edmonton, Canada, Research 

Centre for Radiation Medicine, AMS of Ukraine, Kyiv

Radioprotectants Currently Approved in 

Russian Federation 

ANTI-RADIATION FIRST- AID KIT CREATED 

Includes: 

• INDRALIN – neutralizes Cesium and Strontium 

• LIOXAZOL (spray) – for early treatment of 

radiation sickness and spray for skin burns 

• ZASHITA (PROTECTION) – deactivation and 

protection of skin

Radioprotectants approved in 

Russian Federation (continued) 

• DEZOXYNATUM stimulates the proliferation of hemopoetic 

cells 

• Chemical structure: Sodium salt of DNA extracted from the 

milt of sturgeon species 

• Mechanism of Action: stimulates 

proliferation of hemopoetic cells 

• Indication: ARS, hypo-and aplastic 

anemia secondary to chemo 

or radiation therapy 

• Contraindications: none 

• Side effects: low grade fever (infrequent)

USING ATMOSPHERIC 137 CS MEASUREMENTS AND HYSPLIT TO CONFIRM 

CHERNOBYL AS A SOURCE OF 137 CS IN EUROPE 

Erik L. Swanberg 1 and Steven G. Hoffert 2 

Veridian Systems 1 , Autometric 2 

Sponsored by Defense Threat Reduction Agency 

Contract No. DTRA01-99-C-0031 

ABSTRACT 

The Chernobyl nuclear reactor accident released considerable amounts of radioactive material into the 

environment, including a large amount of 137 Cs. A large fraction of the 137 Cs was deposited on the ground in the 

surrounding areas. Two atmospheric monitoring stations that contribute data to the Prototype International Data 

Centre (PIDC), one in Stockholm, Sweden, and the other in Helsinki, Finland, routinely measure 137 Cs. It is 

believed that the source of this 137 Cs is the ground contaminated by the Chernobyl accident. The PIDC 

routinely uses HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory) atmospheric modeling 

software to determine probable source locations of radionuclides detected during normal operations. In this 

paper, HYSPLIT was used in conjunction with the data from the PIDC to more firmly establish the link 

between Chernobyl and 137 Cs measurements. The results indicate that an air mass containing 137 Cs has a higher 

likelihood of having recently been in the Chernobyl area than an air mass that does not contain 137 Cs. The 

inverse seems true also: an air mass that does not contain 137 Cs is far less likely to have been in the vicinity of 

Chernobyl in the recent past. These results, while not definitive, are very encouraging. The results also 

improve the confidence in HYSPLIT. 

KEY WORDS: Chernobyl, Cesium, PIDC 

OBJECTIVE 

In 1986, the Chernobyl accident released large amounts of many different radionuclides into the atmosphere. 

Starting with its first sample in 1996, the Prototype International Data Centre (PIDC) has measured 137 Cs at 

radionuclide monitoring stations in Europe. This paper attempts to link 137 Cs re-suspension in the vicinity of 

Chernobyl to the 137 Cs regularly measured by monitoring stations. To this end, HSYPLIT (HYbrid Single- 

Particle Lagrangian Integrated Trajectory) atmospheric modeling software was used in conjunction with 

radionuclide concentration measurements to show that the Chernobyl region is likely the source of 137 Cs. 

There are about 150,000 km 2 of land contaminated with 137 Cs at greater than 37 kBq/m 2 (UNSCEAR, 2000). It 

is not surprising that conditions exist that re-suspend this 137 Cs. One example is a forest fire in the Chernobyl 

area that re-suspended measurable amounts of 137 Cs, that was detected by the PIDC in May, 2000. Once 

airborne, the 137 Cs is transported considerable distances in the atmosphere. In particular, the 137 Cs is transported 

to European atmospheric monitoring stations before the 137 Cs concentration falls below the minimum detectable 

concentration. 

There are four PIDC monitoring stations in Europe, all of which regularly measure 137 Cs. Two of these, one in 

England and the other in Germany, sample air for seven days. For the purposes of this study, these 

measurements are too long to make reasonable predictions about the origin of the 137 Cs re-suspension. Two 

other stations, one in Stockholm, Sweden, and the other in Helsinki, Finland, sample air for 24 hours. This 

shorter sampling time allows better estimation of an air mass's previous position. This study, therefore, only 

uses data from the Swedish and Finnish stations. 

To model atmospheric transport, the PIDC runs HYSPLIT daily and generates a Field of Regard (FOR). The 

FORs show areas where it is most likely a parcel of air originated during a specified time period, and hence help 

864

Report Documentation Page 

Form Approved 

OMB No. 0704-0188 

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 

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, 

including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington 

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 

does not display a currently valid OMB control number. 

1. REPORT DATE 

OCT 2001 

2. REPORT TYPE 

3. DATES COVERED 

00-00-2001 to 00-00-2001 

4. TITLE AND SUBTITLE 

Using Atmospheric 137CS Measurements And Hysplit To Confirm 

Chernobyl As A Source of 137CS In Europe 

5a. CONTRACT NUMBER 

5b. GRANT NUMBER 

5c. PROGRAM ELEMENT NUMBER 

6. AUTHOR(S) 5d. PROJECT NUMBER 

5e. TASK NUMBER 

5f. WORK UNIT NUMBER 

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 

Veridian Systems,6 Pecan St.,California,MD,20619 

8. PERFORMING ORGANIZATION 

REPORT NUMBER 

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S) 

12. DISTRIBUTION/AVAILABILITY STATEMENT 

Approved for public release; distribution unlimited 

11. SPONSOR/MONITOR’S REPORT 

NUMBER(S) 


Proceedings of the 23rd Seismic Research Review: Worldwide Monitoring of Nuclear Explosions held in 

Jackson Hole, WY on 2-5 of October, 2001. U.S. Government or Federal Rights. 

14. ABSTRACT 

See Report 


16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 

ABSTRACT 

a. REPORT 

unclassified 

b. ABSTRACT 

unclassified 

c. THIS PAGE 

unclassified 

Same as 

Report (SAR) 

18. NUMBER 

OF PAGES 

7 

19a. NAME OF 

RESPONSIBLE PERSON 

Standard Form 298 (Rev. 8-98) 

Prescribed by ANSI Std Z39-18

to indicate possible sources of radionuclides detected. These FORs were used to show a link between 137 Cs 

measurements and Chernobyl. About 18 months of FOR data was used for this study. 

RESEARCH ACCOMPLISHED 

This section begins by describing the data that was used from the PIDC. It then covers conditions that are 

favorable for transporting 137 Cs from Chernobyl to the monitoring stations. It also describes how HYSPLIT 

was used to generate FOR data. Then the link between 137 Cs measured at the monitoring stations and 

Chernobyl is established. 

PIDC Data 

The PIDC receives data from a global system of atmospheric monitoring stations. The Swedish and Finish 

systems are two of these stations. The stations send High Purity Germanium (HPGe) gamma-ray spectra to the 

PIDC on a regular basis, usually daily. The spectra are analyzed to determine which radionuclides are present 

and what their concentrations are. This includes concentrations for 137 Cs. The Swedish and Finish stations have 

been providing the PIDC with this data since 1996. 

The PIDC uses atmospheric data and atmospheric modeling software to estimate previous locations of air that 

was sampled by a monitoring station. The model is run daily. This allows the PIDC to estimate the source of 

any unusual radionuclides that are observed at a station. The modeling software results were available from 

May 1999 to September 2000, which are the dates included in this study. 

137 Cs Transport 

In order for 137 Cs to travel from Chernobyl to a European monitoring station located over 1000 kilometers away 

it must first be re-suspended in the air. Then atmospheric conditions must be such that the 137 Cs is transported 

to the monitoring station. Some possible mechanisms for re-suspension include high winds stirring up 

contaminated soil (Gillette and Porch, 1978), burning of contaminated vegetation (Garger et al, 1998), or 

perhaps a disturbance in the sarcophagus surrounding the reactor. Atmospheric conditions consist mostly of 

prevailing winds blowing in the correct direction. This study did not attempt to look at processes that could resuspend 

the 137 Cs. Hence discrepancies could result where HYSPLIT indicates air transport in the correct 

direction but no 137 Cs was detected. On the other hand, if 137 Cs from the Chernobyl region was measured, then 

both of the above conditions must have been met. 

HYSPLIT 

The PIDC uses HYSPLIT atmospheric modeling software to simulate particle trajectories in the atmosphere. 

These trajectories are then used to create FORs as follows. Simulated particles were released every hour from a 

2 X 2 grid formed from points located at the intersection of even lines of latitude with even lines of longitude. 

HYSPLIT used measured atmospheric data to simulate the trajectories real particles would follow if released 

under the same conditions. If a particle passed near a station in the simulation, its point of origin was noted as 

having a particle reach the detector. The simulations were run for 24, 48, and 72 hours. After a run was 

complete, the number of particles that reached the detector was totaled, and the grid point tallies were 

normalized so that the sum of all grid point values in an FOR is one. 

A few remarks are in order. First, the 2 X 2 grid square is somewhat coarse, resulting in coarse resolution. 

Second, the model was only run for 72 hours. The distance between Chernobyl and the monitoring stations is 

large enough that it could easily take more than 72 hours for an air mass to travel between the two. Therefore, 

some 137 Cs measurements were possibly not associated with Chernobyl because the air mass took longer than 

72 hours to reach the station. Also remember that the atmosphere is a complex system and no model is perfect. 

Combining HYSPLIT and 137 Cs Concentrations with Atmospheric Transport 

Because FORs are normalized, combining them is accomplished by simply adding them together. The figures 

that follow were created with three different objectives in mind. Figure 1 is an average of all FORs that were 

965

available for the Swedish station. It serves as a baseline and allows any unusual features to be seen. Figure 2 

only includes FORs for days when the concentration of 137 Cs was above 4 Bq/m 3 , which indicates likely 

origins of high 137 Cs concentrations. Figure 3 only includes FORs for days when 137 Cs was not measured, 

indicating where 137 Cs does not originate. The FORs are drawn as contour plots. Figures 4 through 6 are 

corresponding FORs for the Finnish station. 

The figures show that when 137 Cs was measured, air was much more likely to have come from Chernobyl. In 

figures 2 and 5, high likelihood regions are near the stations and Chernobyl. There are other regions where the 

air might have originated, but they have smaller associated probabilities. Since the stations sample air for 24 

hours, it is possible for more than one air mass to pass over a station and be included in a sample. Also, air in 

transit for more than 72 hours is not included in the FORs. Data from both stations indicate Chernobyl as being 

the most likely source for 137 Cs. 

It is also interesting to note where the air originates when no 137 Cs is detected (Figures 3 and 6). Air appears to 

come from similar places as the average (Figures 1 and 4) but with lower probabilities of originating near the 

Chernobyl region. This is what would be expected if Chernobyl were the source of 137 Cs. Since none of the 

other places where air normally reaches the stations from are a source of high 137 Cs concentrations, Chernobyl 

must be the source of the elevated measurements. 

Instead of selecting FORs based on whether or not 137 Cs was measured, they can be selected based on having a 

high probability of air reaching a radionuclide monitoring station from the Chernobyl area. This was done 

using four grid points near Chernobyl. For Sweden, 137 Cs was seen 11 out of 13 times when one of these four 

points had a probability of greater than .05. For Finland the same analysis yields 6 of 9 times. The days with 

higher likelihood also include the highest measurements of 137 Cs made at both stations. Table 1 contains a list 

of probabilities and 137 Cs concentrations. 

For several particularly high concentrations of 137 Cs, HYSPLIT trajectory models were run. Figure 7 shows the 

result for one of these runs. It can clearly be seen this day (which was the highest concentration measured to 

date at the Swedish station) that the air mass goes almost directly from the Chernobyl region to Stockholm. 

Table 1. Probabilities of air originating at a grid point near Chernobyl and 137 Cs concentrations actually 

measured. 

Sweden 

Finland 

Probability Concentration (Bq/m 3 ) Probability Concentration (Bq/m 3 ) 

0.053 0.59 0.065 Not Detected 

0.053 0.58 0.071 0.86 

0.055 Not Detected 0.071 0.86 

0.065 1.9 0.14 5.7 

0.065 1.9 0.17 3.3 

0.067 1.8 0.17 3.3 

0.069 6.4 0.19 Not Detected 

0.069 6.4 0.21 5.7 

0.11 45 0.22 Not Detected 

0.17 16 

0.19 Not Detected 

0.32 45 

0.39 16 

10 66

Swedish 

Station 

>.01 

>.1 

>.5 

>1 

Figure 1. Average of all available FORs for the Swedish station. 


>.02 

>.1 

>.2 

>.3 


Swedish 

Station 

Figure 2. FORs for the Swedish station when the concentration of 137 Cs was greater than 4 Bq/m 3 . 

>.01 

>.1 

>.5 

>1 

Swedish 

Station 

Figure 3. FORs for the Swedish station when no 137 Cs was measured. 


11 67

Figure 4. Average of all available FORs for the Finish station. 


Finish 

Station 

>.01 

>.1 

>.5 

>1 

>.02 

>.1 

>.2 

>.3 


Finish 

Station 

Figure 5. FORs for the Finish station when the concentration of 137 Cs was greater than 4 Bq/m 3 . 

>.01 

>.1 

>.5 

>1 

Finish 

Station 


Figure 6. FORs for the Finish station when no 137 Cs was measured. 

12 68

Swedish 

Station 


Figure 7. Trajectory for the day the highest concentration of 137 Cs was measured in Stockholm. 

CONCLUSIONS AND RECOMMENDATIONS 

A link was established between 137 Cs re-suspended in the Chernobyl region and 137 Cs measured by two 

European monitoring stations in several ways. First, it was shown that when 137 Cs was measured, it was more 

likely for the air that was sampled to have come from the Chernobyl region. Secondly, when no 137 Cs was 

measured the sampled air was much less likely to have been in the Chernobyl region. Likewise, if air did not 

come from the Chernobyl region, 137 Cs wasn't seen. It was also shown that if air did move from the Chernobyl 

region to a monitoring station, it was likely to contain detectable concentrations of 137 Cs. While these results 

are qualitative in nature, the evidence strongly suggests that a correlation exists. 

Another benefit of this study is an increased confidence in the tools used and the manner in which they are used. 

Data analysis using two different stations yielded equivalent results. These results agree with what common 

sense dictates, namely that re-suspension of 137 Cs in the Chernobyl region is the source of 137 Cs in Sweden and 

Finland. One can thus conclude that using radionuclide concentrations obtained with currently fielded systems 

in conjunction with HYSPLIT produces good results. 

If more conclusive evidence of the link between 137 Cs re-suspension near Chernobyl and high 137 Cs 

concentrations is desired, there are a couple of investigations that could be performed. First, 134 Cs was also 

released during the accident. In fact, several of the earlier samples measured by the PIDC contain both 137 Cs 

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 

detection methodology with reduced minimum detectable concentrations, to measure the ratio of 137 Cs to 134 Cs 

in both soil in the Chernobyl region and in the samples from the monitoring stations. If the ratios match, this 

would be strong evidence for one causing the other. Also, determining conditions good for the re-suspension of 

137 Cs into the atmosphere could help to improve the data used in this study. 

13 69

REFERENCES 

Gillette, Dale A. and William M. Porch (1978), The Role of Fluctuations of Vertical and Horizontal Wind and 

Particle Concentration in the Deposition of Dust Suspended by Wind, J. of Geophysical Research, 

January 20, 1978, 409-414 

Garger, E. K., V. Kashpur, H. G. Paretzke, and J. Tschiersch (1998) Measurement of Resuspended Aerosol in 

the Chernobyl Area Part II. Size Distribution of Radioactive Particles, Radiation and Environmental 

Biophysics Vol 36 Number 4, 275-283 

HYSPLIT4 (HYbrid Single-Particle Lagrangian Integrated Trajectory) Model (1997). Web address: 

http://www.arl.noaa.gov/ready/hysplit4.html, NOAA Air Resources Laboratory, Silver Spring, MD. 

United Nations Scientific Committee on the Effects of Atomic Radiation, UNSCEAR 2000, Sources and Effects 

of Ionizing Radiation, Volume II: Effects, Report to the General Assembly, with Scientific Annexes. 

14 70

Chernobyl Nuclear Accident Defense Department Reports

Create successful ePaper yourself

Delete template?

Save as template?