zy
zyxwvutsrqponmlkjihgfe
zyxwvutsrqponm
zyxwvutsrqponm
zy
+
Heahh Phy$ics Vol. 59, No. 5 (November), pp. 659-668, 1990
Printed in the U S A .
0017-9078/90 $3.00 .00
0 1990 Health Physics Society
Pergamon Press pic
Analyses and Modeling for Internal Dose Estimates
MODELS OF RADIOIODINE TRANSPORT TO POPULATIONS
WITHIN THE CONTINENTAL U.S.
Andre Bouville *
Radiation Effects Branch, National Cancer Institute, Bethesda, MD 20892
and
Mona Dreicer
Lawrence Livermore National Laboratory, Livermore, CA 94550
and
Harold L. Beck
Department of Energy, Environmental Measurements Laboratory, New York, NY 100 14
and
Walter H. Hoecker
Air Resources Laboratory, National Oceanic and Atmospheric Administration, Silver Spring, MD 209 10
and
Bruce W. Wachholz
Radiation Effects Branch, National Cancer Institute, Bethesda, MD 20892
Abstract-A methodology is being developed to estimate the exposure of Americans to I3'I originating from atmospheric nuclear weapons tests carried out at the Nevada Test Site (NTS) during the 1950s and early 1960s.
Since very few direct environmental measurements of I3'I were made at that time, the assessment must rely on
estimates of 13'1 deposition based on meteorological modeling and on measurements of total B activity from the
radioactive fallout deposited on gummed-film collectors that were located across the country. The most important
source of human exposure from fallout I3'I was due to the ingestion of cows' milk. The overall methodology used
to assess the "'I concentration in milk and the I3'I intake by people on a county basis for the most significant
atmospheric tests is presented and discussed. Certain aspects of the methodology are discussed in a more detailed
manner in companion papers also presented in this issue. This work is carried out within the framework of a task
group established by the National Cancer Institute.
INTRODUCTION
portant tests, the I3'I exposures from fallout for representative individuals and for the populations of each county
of the contiguous U.S. during the time of the tests.
The most significant atmospheric weapons tests with
respect to fallout occurred in the 1950s, during which
time most of the monitoring of environmental radioactivity consisted of gross /3 or y measurements. Because
the radioactive half-life of I3'I is about 8 d, the I3'I activity
present in the samples collected more than 25 y ago has
completely decayed and cannot be measured retrospectively. Therefore, the estimation of I3'I exposures dating
back to the 1950s must be derived from the original measurements of gross ,k? or y activity, from current or past
measurements of radionuclides other than I3'I, or from
mathematical models.
zyx
ONE PART of Section 7 (a) of Public law 97-4 I4 directs
the Secretary of Health and Human Services to "conduct
scientific research and prepare analyses necessary to develop valid and credible assessments of the exposure to
I3'I that the American people received from the Nevada
atmospheric bomb tests."
The National Cancer Institute (NCI) was requested
to respond to this mandate. In doing so, a task group,
established to assist the NCI in this effort, suggested that
it might be possible to estimate, for each of the most im-
* Author to whom correspondence should be addressed.
659
zyxw
zyxwvutsrq
zyx
660
Health Physics
November 1990, Volume 59, Number 5
In addition to the present study, two other efforts
are concerned with the exposure of more specific populations to I3lI from fallout: The Off-Site Radiation Exposure Review Project of the Department of Energy
(Church et al., this issue) is estimating exposures of
downwind residents to fallout, and the University of Utah
is conducting an epidemiological study of thyroid disease
among populations of Utah (Wachholz, this issue).
Transport models used in the three studies to estimate the extent to which individuals or populations were
exposed to I3II are similar. There are some differences
that distinguish this study from the other two, however,
because of the larger geographic scope of this study. The
models used here are of a more generic nature since the
level of detail required in the other two studies is not
practical for a continent-wide assessment. Moreover, because most of the fallout in the eastern part of the country
was associated with precipitation (i.e., “wet” fallout),
whereas “dry” deposition was predominant in the western
part of the country (Beck et al., this issue), precipitation
receives a greater emphasis in this study than is required
for the other two.
Once l 3 l I from fallout has been deposited on vegetation, the main pathway to man is, for most individuals,
via the grass-cow-milk chain (Eisenbud and Wrenn 1963;
Garner and Russell 1966). In the assessment of 1311 exposures on a continental scale, certain assumptions and
methodologies enable estimates of the following parameters to be made for each of the nearly 3,100 counties in
the contiguous United States:
testing at Nevada Test Site (NTS), show that 1 3 1 1 from
weapons tests is partitioned among three physicochemical
forms: gaseous organic, gaseous inorganic, and particulate
(Perkins 1963; Perkins et al. 1965; Voilleque 1979). From
measurements taken after a Chinese nuclear weapons test,
the partitioning between these three forms was shown to
vary with the time elapsed following the detonation
(Voilleque 1979). Voilleque ( 1986) speculated that more
than half the l3II would be associated with particular diameters of less than about 20 pm, with the remainder
presumably being in organic and inorganic gaseous forms.
In the absence of better data, it has been assumed that all
I 3 l 1 was in particulate form.
the activities of I3II deposited on soil and vegeta-
Dispersion of the radioactive cloud
The amount of I3’I produced in each explosion was
derived from Hicks ( 1981). The I3lI activity per unit yield
was found to be about 5 PBq kt-I of fission for the shots
considered in the assessment. The apportionment of the
I3II activity between the cloud top, the cloud stem, and
the local deposition in the immediate vicinity of the test
site was estimated as follows, according to the type of test:
0
tion,
METEOROLOGICAL MODELING
The radioactive cloud that is formed after an atmospheric detonation near the ground surface usually is
in the shape of a mushroom, extending from the ground
surface to the highest layers of the troposphere, and occasionally reaching into the stratosphere. It contains
hundreds of different radionuclides, including I3’I.
The meteorological prediction of I3II deposition,
presented in detail in another paper (Hoecker and Machta,
this issue) and discussed here briefly, involves two steps:
zyxw
zyxwvut
zyxwvu
0 the amount of l3’I consumed by dairy cows and
the resulting I3II concentrations in cow’s milk, and
0 the I3II ingested by people.
The purpose of this paper is to present the overall
methodology currently used in the assessment of the I 3 l I
exposures from fallout resulting from the atmospheric
nuclear weapons tests camed out at the Nevada Test Site
during the 1950s and early 1960s. Although all aspects of
the methodology are subject to revision before completion
ofthe study, it is likely that changes will be minimal. Parts
of the methodology are described in more detail in companion papers presented in this issue (Beck et al.; Dreicer
et al.; Hoecker and Machta).
ESTIMATION OF ACTIVITIES DEPOSITED
ON THE GROUND
Meteorological modeling and reanalysis of historical
monitoring data are two complementary methods used
to estimate I3II deposited on the ground following each
test.
For both approaches, the assumption is made that
the I3’I was in particulate form, as were the majority of
radionuclides produced in the atmospheric nuclear weapons tests. Limited measurements, unrelated to weapons
a) dispersion of the radioactive cloud across the US., and
b ) estimation of the amount of I3’I deposited on the
ground.
Type of test
Surface or
tower
Balloon or
airdrop
Apportionment of 13’1 activity
Cloud top
Cloud stem
Local deposition
0.8
0.1
0.1
0.9
0.1
0
The 1311 activities in the cloud are assumed to be homogeneously distributed within the cloud top and cloud stem.
The dispersion of the radioactive cloud has been analyzed for each important atmospheric test using routine
weather maps that depict airflow at constant pressure levels. These maps, which were provided twice a day by
weather services, were used to construct, at several altitudes ranging between 3 and 13 km, 6-h trajectories of
air parcels originating at the Nevada Test Site and moving
across the U.S. In general, trajectories at those various
zyxw
zyxw
zyxw
Models of radioiodine transport to populations within the continental U.S. 0 A. BOUVILLEet al.
altitudes diverged in both direction and speed after leaving
the detonation site.
The radioactive cloud was often stretched by vertical
wind shear to many hundreds of kilometers before it left
the U.S. This large shear resulted in a great dilution of
I3'I. Additional distribution was caused by lateral spreading of the cloud by eddy or turbulent diffusion, which
Hoecker and Machta (this issue) assumed to occur at a
rate of about 7 km h-I. The meteorological model predicts
the spatial coverage of the radioactive cloud at each 6-h
interval and the I3'I activities per unit area contained in
the radioactive cloud at each county centroid of the continental U.S. Hoecker and Machta (this issue) assumed
in their model that at any given time, the 13'1distribution
is uniform within the boundaries of the cloud segments
created by lateral spreading and vertical shearing between
the altitudes at which the trajectories are determined.
66 1
Environmental Measurements Laboratory (EML) in
cooperation with the U S . Weather Bureau (Beck 1984;
Beck et al., this issue; Harley et al. 1960).
The EML network effectively fulfilled its original
purpose of indicating quickly where and when fallout occurred. The network was not designed to derive the
amounts of specific radionuclides deposited on the
ground; however, since it represents the only radiological
monitoring data available on a daily basis over the entire
U.S. during most of the atmospheric testing period, it was
considered that, by careful reevaluation of the data, it
might be possible to derive estimates of 13'1deposition at
the location of the gummed-film stations.
The EML deposition network across the US. evolved
gradually from the use of trays of water at 10 locations
in 1951, to the use of gummed-paper collectors at 93 locations in 1952, and finally to the use of gummed-film
collectors at about 100 locations until the end of the decade. A "gummed-film collector" consisted of a 0.3 m
X 0.3 m exposed area of gummed film that was positioned
horizontally on a stand 0.9 m above the ground. Usually
two films were exposed during a 24-h period beginning
at 1230 GMT. The samples collected were ashed and
counted for total p activity.
Beck ( 1984) reviewed and reanalyzed the available
gummed-film data that could be found in the HASL/
EML archives, together with other less extensive fallout
data, in order to derive depositions of 137Cs,I3'I, and 1331.
One of the difficulties in the reanalyses of monitoring data
was that original data may have been mislabeled or not
assigned to the appropriate nuclear weapons test. In an
effort to alleviate this problem, the areas of observed deposition systematicallywere compared with the areas predicted by the meteorological model to be covered by the
radioactive cloud. In case of a significant discrepancy,
a more detailed trajectory analysis was carried out to
determine whether the suspect gummed-film results
could be explained from meteorological considerations
(Hoecker and Machta, this issue). If not, the suspect results were discarded.
The resulting data set includes daily depositions of
13'1at up to approximately 100 locations in the U.S. during
most of the atmospheric testing period. Those 13'1depositions are associated with information on the precipitation amounts occurring during the same 24-h periods.
Some of the gummed-film results are presented in another
paper in this issue (Beck et al., this issue).
zyxwvu
zyxwvutsr
zyxw
Deposition on the ground
Deposition of "'I on the ground results from two
processes: impaction of aerosols on the ground surface
(dry) and precipitation (wet). In the western part of the
country, most I3'I deposition was from dry processes, since
weapons testing generally was not allowed under atmospheric conditions such that wet deposition was likely to
occur within a few hundred kilometers from the NTS.
That operational precaution, however, did not extend
to the eastern part of the country, where most 13'1deposition occurred as a result of wet processes (Beck et al.,
this issue). Therefore, the amount of rainfall is an important parameter in attempting to estimate the extent to
which wet deposition occurred. In order to approximate
the amount of rain that occurred across the country during
the time of interest, recorded daily rainfall amounts reported by the National Oceanic and Atmospheric Administration (NOAA) were averaged on a county basis.
The endpoint of the meteorological model is the estimation of the amounts of 13'1deposited by precipitation.
This involves not only the knowledge of daily rainfall
amounts but also that of many other uncertain factors,
among which are the efficiency of the rain-out process,
the exact location of the radioactive cloud segment, the
location and dimensions of the precipitating cloud, and
the physicochemical form of I3'I. Because of the complexity of the problem, an empirical method was established based on the use of relationships between the column content of I3'I in the overhead cloud and the 13'1
deposition estimated from the monitoring data (gummedfilm) , which are discussed in the following section.
zyxwvutsrq
REVIEW AND REANALYSIS OF HISTORICAL
MONITORING DATA
For counties near the NTS, monitoring consisted
mainly of exposure-rate measurements using portable
survey instruments (Beck and Anspaugh 1990; Thompson, this issue). This close-in monitoring network will
not be discussed here.
Over the remainder of the U.S., monitoring of fallout
deposition in the 1950s was carried out primarily by the
ESTIMATION OF 1311 DEPOSITION
IN ANY GIVEN COUNTY
The daily deposition densities of I3'I need to be estimated for each of the nearly 3,100 counties in the contiguous U.S. For this purpose, both meteorological modeling and reanalysis of historical data have limitations.
EML was at that time called the Health and Safety Laboratory
( HASL) .
662
zyxwvutsrqponm
zyxwvuts
Health Physics
The position of the radioactive cloud is not always in
agreement with the areas of deposition derived from
monitoring data, usually because of the simplifying assumptions used in the model. In particular, measured depositions often occurred over a longer time period than
that predicted by the meteorological model. In addition,
although the meteorological model has the potential of
predicting 13'1 deposition by wet processes, it can only do
so in a crude way for those areas where precipitation occurred during the predicted passage of the radioactive
cloud. On the other hand, the reanalysis of historical
monitoring data provides the best available estimates of
I3'I deposition per unit area but, under the best conditions,
only for up to about 100 locations.
In order to estimate the daily 1311 deposition in any
given county, the following procedure, in which preference
is systematically given to the monitoring data, has been
applied:
November 1990, Volume 59, Number 5
zyxw
zyx
zyxwvutsr
0 For the tests for which gummed-film data are
available (from October 1951 to November 1958), the
deposition densities were obtained in most cases by interpolating between the counties with measured data using
a kriging procedure.
0 For tests in which gummed-film data are not available (before October 1951 or after November 1958), meteorological modeling was used to estimate deposition
densities in the counties where precipitation occurred
during the passage of the radioactive cloud. Counties
where precipitation did not occur during the passage of
the radioactive cloud were assigned a zero deposition.
ESTIMATION OF 1311 CONCENTRATIONS
IN FRESH COWS' MILK
The transfer of I3'I from deposition on the ground
to fresh cows' milk is relatively well documented (e.g.,
Bergstrom 1967; Black and Barth 1976; Garner 1967;
Kirchner et al. 1983; Wicker and Kirchner 1987). Figure
1 illustrates the parameters involved in that transfer. The
time-integrated concentration of I3'I in milk (IC) corresponding to an estimated deposition density on the ground
(DG) on a given day and in a given county was calculated
as:
IC
=
* interception
* retention
by vegetation
on vegetation
v
Time-in tegrated concentration
in veaetation
* pasture intake
intake-to-milk
by dairy cows
transfer coefficient
Fig. I . Iodine-I31 transfer from deposition on the ground to
fresh cows' milk.
ical conditions and the type and density of vegetation.
Values of interception coefficients obtained in laboratory
or field experiments conducted under dry or light spray
conditions with artificial radionuclides show a large range
of variation between 0.02 and 0.82 (Miller 1980). A much
narrower range of 1 to 4 m2 kg-' (dry) is obtained for
the mass interception coefficient ( F * ), defined as the interception coefficient ( F ) divided by the standing crop
biomass ( Y ) .
Chamberlain ( 1970) proposed that the interception
coefficient could be estimated as:
F = 1 - e-ay
(2)
where the numerical value of a , the foliar interception
constant, is 2.8 m kg-' (dry weight) for elemental iodine
and small-size aerosols under dry or light spray conditions.
The value of F* is then obtained as:
1 - e-aY
F* =
(3)
Y
-
zyxwvutsrq
zyxwvut
Ten X P I X F,,
DG X F* X In 2
(1)
in which F* is the mass interception coefficient [ m 2 kg-'
(dry weight)], Tenis the effective half-time of retention
by the vegetation (d), PI is the pasture intake [ kg (dry
weight) d-'1 , and F,,, is the intake-to-milk transfer coefficient (d L-'). Each of these parameters will be discussed.
MASS INTERCEPTION COEFFICIENT
Iodine- 1 3 1 concentration in milk is directly proportional to that fraction of activity deposited that is intercepted by vegetation ( F ) ,or the interception coefficient,
which depends, among other factors, on the meteorolog-
If the product N X Y is much smaller than 1, which
is often the case, eqn ( 2 ) can be approximated as:
F=aXY,
(4)
and the numerical value of the mass interception coefficient ( F * ) is equal to that of the foliar interception constant.
There is evidence that the value of a decreases as the
particle size increases (Romney et al. 1963; Anspaugh et
al. 1986; Whicker and Kirchner 1987) and, therefore, that
the interception coefficient decreases as the particle size
increases. In the case of atmospheric nuclear weapons
tests, large-size particles fall out near the detonation site
and smaller particles are deposited as the radioactive cloud
moves farther away. Simon (this issue) estimates that the
zyxw
zyxw
zyxwvutsrqp
zyxwvutsr
663
Models of radioiodine transport to populations within the continental U.S. 0 A. BOUVILLEet al.
'
zyxwvutsr
variation of a (m kg-', dry weight) as a function of distance, D ( km ) ,can be expressed as:
a(D) = 7.01
X D'.I3.
(5)
Using this expression, the value of a increases with distance from the NTS and is approximately equal to 2.8
m2 kg-' (dry) for D = 1600 km. Beyond that distance,
the value of a is taken to remain constant at 2.8 m 2 kg-'
in order to remain consistent with the value proposed by
Chamberlain ( 1970) for elemental iodine and small-size
aerosols.
All of the laboratory and field experiments were conducted under dry or light spray conditions (Miller 1980)
and do not, therefore, provide any information on the
values to be expected in the case of moderate or heavy
rainfall. In a limited number of cases, however, I3'I fallout
was measured in rain and vegetation after atmospheric
nuclear weapons tests. The interception coefficient values
derived from those measurements show a large range of
variation, from less than 0.09 to about 0.9, with a high
scatter for any given rainfall levels but with a tendency
to decrease as the rainfall amount increases (Voilleque
1986). Adapting an expression originally developed by
Horton (1919) for the initial retention of rainwater by
vegetation, Voilleque ( 1986) proposed that the variation
of the mass interception coefficient as a function of the
rainfall amount P (mm) can be estimated by:
tigate the dependence of the mass interception coefficient
on the nature and physicochemical form of radionuclides,
the rainfall amount and intensity, and the type and height
of vegetation (Hoffman et al. 1989). The results of these
experiments are in general agreement with those derived
from the model. For 13'1 in soluble form, the experimental
values of the mass interception coefficient are about 10
times lower than those predicted by the model. However,
for the case in which I3'I is attached to particulates, which
is the form likely to have been predominant in fallout,
there is good agreement between experimental and predicted values of the mass interception coefficient, especially for amounts of rainfall in excess of 10 mm.
EFFECTIVE HALF-TIME OF RETENTION
After 13'1 is deposited on vegetation, environmental
removal processes combine with radioactive decay to reduce the initial amount on the vegetation surface (Miller
and Hoffman 1979). The time necessary for one-half of
the activity to be removed by environmental processes is
referred to as the environmental half-time ( T,). This time
value, together with the radioactive half-life ( T,) , determines the effective half-time ( T g ) :
zyxwvuts
S
F*=E+-
P'
where E = 1.3 m kg (dry weight) is the in-storm evaporation fraction per unit areal density of vegetation, and
S = 16 mm kg-' (dry weight) m-2 is the rainfall storage
capacity per unit areal density of vegetation. According
to this expression, the mass interception coefficient is inversely related to the rainfall amount.
The application of eqn (6), however, yields values
of the interception coefficient that are greater than one
for low rainfall amounts associated with high-standing
crop biomasses, which is physically impossible. In order
to resolve this inconsistency, the inverse relationship between the mass interception coefficient and the rainfall
amount proposed in eqn (6) was assumed only to apply
to daily rainfall amounts in excess of 5 mm.
The linkage between eqn ( 3 ) , used for dry deposition,
and eqn (6), used for wet deposition, is done assuming
that, for light rain (less than PI = 5 mm), the value of
F* for a distance D and a daily rainfall amount P is obtained by linear interpolation:
P
F * ( D , P ) = F * ( D , O ) + [F*(Pt)-F*(D,O)] X - .
PI
(7)
Given the importance of the interception coefficient
in the assessment of I3'I exposures, and the limited information on its value under conditions of moderate or
heavy rainfall, a research program was designed to inves-
Measured values of T, for particulates on vegetation
range from 9 to 71 d, with most values between 10 and
20 d (Miller and Hoffman 1979).
Values of T, may be expected to vary markedly as
a function of the growth of vegetation and of meteorological conditions. Given the short radioactive half-life of
13'1,however, the effective half-time Tefis not particularly
sensitive to large variations of the environmental halftime T,. The average value of T, was assumed to be 14
d, yielding an effective half-time, Tef, of slightly more
than 5 d.
PASTURE INTAKE BY DAIRY COWS
Significant concentrations of I3lI in milk can only
arise from the occurrence of fresh fallout deposition followed by the consumption of fresh pasture by the cow
reasonably soon thereafter. Because of the climatological,
geographical, and agricultural diversity of the U.S., a large
data base is necessary to estimate the amount of pasture
eaten by cows in counties across the U.S. Such a large
data base is provided by the Dairy Herd Improvement
Association ( DHIA ) , which maintains extensive records
on dairy cattle and milk production.
The method by which the pasture intake by dairy
cows is estimated is described in detail in another paper
in this issue (Dreicer et al., this issue). The daily pasture
intake PI,, in week wand state s is obtained as the product
of the maximum daily intake of dry matter DM, by cows
in state s and of the fraction of the maximum dry matter
intake that is due to pasture FPw,s
in that week and that
state:
PI,,, = DM, X FP,,.
(9)
zy
664
zyxwvu
zyxwvut
zyxwvuts
zyxwvutsr
zy
zyxwvut
Health Physics
The maximum daily intake of dry matter, DM,, can
be estimated from the cow's body weight and its milk and
fat production (NRC 1978) . These data are available in
the DHIA records for a large number of herds across the
U.S. and for each year of the atmospheric testing period.
For the purpose of this study, state averages for the entire
testing period have been used in determining the daily
intake of dry matter.
The fraction of daily dry matter intake by cows that
is obtained from pasture FP,,s in each state has been estimated on a weekly basis using the expert opinions of
U.S. Department of Agriculture Extension Specialists and
of other knowledgeable persons asked to help reconstruct
pasture feeding practices during the 1950s. Although subjective, these estimates are the best obtainable information
on the seasonal variation of pasture practices at that time.
Table 1 presents the estimated pasture intakes
obtained from eqn ( 9 ) , for states of the northeastern part
of the country. More detailed results are given in Dreicer
et al. (this issue).
November 1990, Volume 59, Number 5
The type of vegetation ingested by the cow has been shown
to have an influence on the value of F,,, but there is not
enough information to take this variation into account.
d L-' used in this study is assumed
The value of 4 X
to be independent of any influencing parameter.
ESTIMATION OF
1311
INTAKES BY HUMANS
Assuming that the time-integrated 13'1 concentrations
in fresh cow's milk produced in any county of the U.S.
have been estimated, it is necessary to determine how
much milk was produced and where it was consumed.
Accordingly, information is needed on the milk production in each county, on the milk distribution pattern
within each county and each state, on the delay between
production and consumption of milk, and on the consumption of milk as a function of factors, such as race,
age, and sex. The time-integrated I3II concentrations in
the milk consumed by man and the corresponding I3'I
intakes are assessed for each atmospheric nuclear test of
significance.
INTAKE-TO-MILK TRANSFER COEFFICIENT
The intake-to-milk transfer coefficient, F, (d L-'),
is defined as the time-integrated concentration of l3II in
milk per unit of I3II activity consumed by the cow. This
transfer coefficient has been determined experimentally
in a large number of studies, including tracer experiments
with stable or radioactive iodine and field studies in which
pasture was contaminated by I3'I resulting from releases
from nuclear facilitiesor from fallout from nuclear weapons tests. Reported values range from 2 X 10-3 to 4 X 10-2
d L-I (Hoffman 1979; Ng et al. 1977), but it seems that
fallout studies yielded values in the lower part of the range.
In this study, it is assumed that the average value of F,,,
for 13'1and for cows is 4 X lop3d L-I.
There is conflicting evidence regarding the influence
of milk yield on the value of the transfer coefficient, F,.
Time-integrated 311concentrations in consumed milk
The time-integrated 13'1 concentrations in consumed
milk are assessed in each county as follows:
0 The amount of milk consumed annually in each
county is estimated as the product of the population of
the county and of the per capita consumption rate in the
state.
The amount of milk consumed within the county
is compared to the fluid milk available from within the
county for consumption purposes. The difference between
the quantity of available fluid milk and the amount of
milk consumed represents the amount of fluid milk for
consumption purposes that is imported into or exported
out of the county. The available fluid milk for consumption purposes has two components:
Table 1. Estimated pasture intake by dairy cows in the 1950s in Northeastern US.
Estimated pasture intake (kg(dry weight) d-l)
zyxwvutsrq
~
State
averaged over the
pasture season
for the first week of June
Connecticut
5.6
Delaware
4.9
Maine
7.9
Maryland
6.5
7.6
Massachusetts
6.8
11.4
New Hampshire
7.6
9.8
New Jersey
6.1
7.6
New York
5.1
6.8
Pennsylvania
4.4
7.7
Rhode Island
8.3
9.7
Vermont
7.3
9.9
8.8
5.5
9.6
zyxwvut
zyxw
zyxw
z
zyxwvu
Models of radioiodine transport to populations within the continental U.S. 0 A. BOUVILLE
et al.
( 1 ) The amount of milk consumed on farms,
which is reported by year and by state and is apportioned by county within the state using the number
of farms in each county as a guide. A delay of 1 d
between production and consumption is assumed
for this category of milk.
( 2 ) The amount of milk sold for fluid use (i.e.,
that which is either retailed directly from the farm
or sold to a distributor). This amount is obtained in
the manner estimated by Dreicer et al. (this issue),
in which the total milk production in the county and
the amount used for manufactured products (cheese,
butter, yogurts, etc.) are also considered.
In order to model the local distribution of the milk
sold for fluid use in the 195Os, each state is divided into
regions determined by the state agricultural service reports,
milk marketing order areas, major population areas, or
state topography. The surpluses and deficits of milk in
the counties of the established regions are balanced, to
the extent possible, within each of the regions by pooling
the milk in the surplus counties of the region and distributing it to the deficit counties of that region.
If a milk surplus or deficit remains after intra-regional
pooling, the milk can be shipped to, or from, another
region. Milk flow is estimated on the basis of available
marketing statistics and upon the advice of experts. Milk
is assumed to have been consumed within 2 d of its production if it was produced in the same county, within 3
d if it was produced in the same region, and within 4 d if
it was produced from outside the region.
In summary, four categories of fluid milk for human
consumption with different delay times are considered:
Milk consumed on the farm (delay: 1 d);
0 Milk sold for fluid use:
(a) produced in the same county (delay: 2 d);
(b) produced in the same region (delay: 3 d);
and
(c) produced in another region (delay: 4 d).
0
Iodine-I 3 I activity intake from milk by humans
Iodine-13 1 intake from milk by humans is the product of the time-integrated concentration of I3'I in the milk
ingested and of the milk consumption rate. Individual
I3'I intakes from milk vary widely from person to person
because of variability in such factors as environmental
parameters, patterns of milk production and distribution,
and dietary habits. Therefore, realistic estimates of individual intakes can be made only if specific information
is available on the individual considered (age, sex, place
of residence, source of milk, delay between production
and consumption, milk consumption rate).
In the absence of personal data, only average intakes
over large or homogeneous groups of people can be estimated with reasonable accuracy. For this reason, the I3'I
intakes of milk by humans estimated in the NCI study
for each county and for each nuclear test are averages
over specified population groups deemed to be represen-
665
tative of a large spectrum of individuals. Information on
average milk consumption rates according to age, sex,
and region of the country can be found in Dreicer et al.
(this issue).
Although ingestion of cows' milk is generally the
predominant contributor to I3'I intake, other exposure
routes need to be considered for individuals who consume
little or no cows' milk. These exposure routes, which include inhalation and ingestion of goats' milk, cottage
cheese, leafy vegetables, and eggs, are considered in the
NCI study but are not discussed in this paper.
ILLUSTRATIVE RESULTS
The methodology previously discussed is to be applied to each county of the contiguous U.S. following each
significant atmospheric nuclear test. It is too early at this
phase of the study to provide final estimates, but a hypothetical example can demonstrate how the various
models and procedures are used.
Many factors influence 13'1intake through ingestion
of cows' milk in a given county: the magnitude and type
of deposition (dry or wet), the distribution of deposition
in the total area that supplies milk in the county considered, and the pasture intake by cows in that area, which
depends on the time of year in which the weapons test
took place.
The importance of deposition distribution is illustrated in the following examples, in which it is assumed
that following a nuclear test, a total dry deposition of 10
TBq occurred in either one of two different regions of
New York state (Fig. 2):
1) Deposition in the New York City region only
and nowhere else, resulting in an I3'I deposition density
of about 2,000 Bq m-' in that region ( 1954 population:
10 million people; area: 5,100 km2). Milk production for
fluid use in the New York City region was approximately
2 X lo7 L in 1954, while the expected milk consumption
was 2 X lo9 L. Most milk consumed in the region was
therefore imported.
2) Deposition in the New York North region only,
yielding an I3'I deposition density of about 700 Bq m-'
in that region ( 1954 population: 0.2 million people; area:
14,000 km ') . In that region, the milk production for fluid
use was about 3 X lo8 L in 1954, greatly exceeding the
estimated 3 X lo7 L needed by the population at that
time. Most milk produced for fluid use in the region was
therefore exported, about 90% of the surplus being shipped
to the New York City region.
Although the deposition patterns are hypothetical
and chosen to be very simple for the purposes of these
examples, the deposition levels that are used are within
the range of values observed in a northeastern state after
an atmospheric nuclear test of moderate yield. The calculation of average individual I 3 'I intakes corresponding
to the two examples considered has been carried out using
the following parameter values:
666
zyxw
zyxwvutsrqpon
Health Physics
November 1990, Volume 59, Number 5
New York City
zyxw
zyxwvutsrqponm
zyxwvutsr
zyxwvut
Fig. 2. Hypothetical I3'I deposition in New York state.
0
m -2.
Standing crop biomass Y
=
0.3 kg (dry weight)
Foliar interception constant LY = 2.8 m 2 kg-' (dry
weight).
0 Effective half-time of retention by vegetation, T e ,
= 5.1 d.
0 Pasture intake: 6.8 kg (dry weight) d-I.
0 Milk distribution: as described in Dreicer et al.
(this issue).
Table 2 presents the time-integrated I3'I concentrations obtained for the four categories of milk considered.
For the two distributions of deposition, there are large
differences in the average concentrations in the various
categories of milk. As expected, the highest concentrations
are obtained in the milk produced and consumed locally
when deposition occurs in the region. A maximum of 680
Bq d L-' is found in the milk hypothetically consumed
on a farm in the New York City region for the assumed
deposition of 10 TBq in that region.
However, the volume-weighted average I3'I concentration in milk consumed in the New York City region
is about four to five times lower when deposition occurs
in that region (7 Bq d L-') compared to when it occurs
in the New York North region (30 Bq d L-' ). This reflects
milk transfer from the New York North region to the
New York City region and shows the importance of assessing the movement and dilution of milk from its site
of production to its site of consumption.
Iodine- 131 intakes by representative individuals
through ingestion of cows' milk can be calculated using
the time-integrated concentrations in milk presented in
Table 2 and the milk consumption rates given in Dreicer
et al. (this issue). For example, in the case of deposition
in the New York City region, individual intakes of people
living in the New York City region are found to range
from 7 Bq d L-' X 0.19 L d-' = 1.3 Bq for a 70-y-old
female drinking milk from a supermarket, to 680 Bq d
L-' X 0.75 L d-'= 510 Bq for a male teenager drinking
milk from a local farm.
Similar calculations are being carried out for several
population groups in each of the contiguous counties of
the U.S. and for each atmospheric nuclear test of significance. It is to be noted that, as indicated by Wachholz
(this issue), assigning thyroid doses per unit activity intake
of 13'1 and assessing the risk of developing thyroid cancer
per unit of thyroid dose from I3'I are topics considered
by other task groups at NCI.
SUMMARY AND CONCLUSIONS
A methodology is being developed by an NCI task
group to estimate I3'I exposures that Americans received
from the NTS atmospheric weapons tests that were detonated in the 1950s and early 1960s. The most important
contributing factor to those exposures is due to the intake
of cows' milk. Various steps in assessing milk concentration and intake by humans on a county basis for the most
significant atmospheric tests are discussed. The most important factors that influence 13'1intake from ingestion
of cows' milk in a given county appear to be the atmospheric process by which 13'1is returned to the earth (i.e.,
dry or wet deposition), the magnitude and distribution
zyxw
zyx
zyxwv
Models of radioiodine transport to populations within the continental U.S. 0 A. BOUVILLEet al.
667
Table 2. Time-integrated concentrations of 13’1in the four categories of milk (Bq d L-’) consumed in each
region for the hypothetical cases of 10 TBq deposition in either the New York City region or the New York
North region.
1311 in milk (Bq d L-l)
Deposition in
New York City region
zyxw
Deposition in
New York North
Category
NYC
NY North
NYC
Milk consumed on farm
680
0
0
Milk sold which originated
from the same county
110
0
0
230
Milk from the region pool
-a
-a
-a
-a
0
-a
30
-a
Milk from other regions
250
zyxwvut
zy
zyxwvu
7
Volume-weighted
NY North
0
30
240
a There is no volume of milk corresponding to this category.
of deposition in the total area that supplies milk to the
county considered, the amount of fresh-pasture ingested
by dairy cows, and the consumption rate of milk by man.
Acknowledgments-The authors would like to thank the following people
for their assistance: L. R. Anspaugh, Lawrence Livermore National Laboratory; R. 0. Gilbert, Battelle Pacific Northwest Laboratories; C. V.
Gogolak and A. Hutter, Environmental Measurements Laboratorv:
F. 0. Hoffman,Oak Ridge National Laboratory; L. Machta and hi:
Smith, National Oceanic and Atmospheric Administration; T. B. Kirchner and F. W. Whicker, Colorado State University: S. L. Simon. Office
of Resident Scientist, Republic of the Marshall Islands;J. Till, Radiation
Assessments Corporation; P. Voilleque, Science Applications International Corporation; D. Wheeler, Nevada Operations Office of the US.
Department of Energy.
REFERENCES
Anspaugh, L. R.; Koranda, J. J.; Ng, Y. C. Internal dose from
ingestion. In: Anspaugh, L. R.; Koranda, J. J., eds. Assessment
of radiation dose to sheep wintering in the vicinity of the
Nevada Test Site in 1953. Las Vegas, NV: U.S. Department
of Energy; DOE-239; 1986.
Beck, H. L. Estimates of fallout from Nevada weapons testing
in the western United States based on gummed-film monitoring data. New York, NY: U S . Department of Energy,
Environmental Measurements Laboratory; EML-433; 1984.
Beck. H. L.; Anspaugh, L. R. The county data base: Estimates
of exposure rates and times of arrival in the off-site radiation
exposure review project (ORERP) Phase I1 Area. Washington, D.C.: U.S. Department of Energy; U.S. DOE report
NVO-320; 1990.
Beck, H. L.; Helfer, I. K.; Bouville, A.; Dreicer, M. Estimates
of fallout in the continental U.S. from Nevada weapons testing based on gummed-film monitoring data. Health Phys.
59:565-576; 1990.
Bergstrom, S. 0. W. Transport of fallout 1-131 into milk. In:
Environmental contamination by radioactive materials. New
York: Pergamon Press; 1967: 159- 174.
Black, S. C.; Barth, D. S. Radioiodine prediction model for nuclear tests. Washington, D.C.: U.S. Environmental Agency;
Report EPA-60014-76-027; 1976.
Chamberlain, A. C. Interception and retention of radioactive
aerosols by vegetation. Atmos. Environ. 457-78; 1970.
Church, B. W.; Wheeler, D. L.; Campbell, C. M.; Nutley, R. V.
Overview of the Department of Energy’s Off-Site Radiation
Exposure Project (ORERP). Health Phys. 59503-5 10; 1990.
Dreicer, M.; Bouville, A. C.; Wachholz, B. W. Pasture practices,
milk distribution, and consumption in the continental U.S.
in the 1950s. Health Phys. 59:627-636; 1990.
Eisenbud, M.; Wrenn, M. E. Biological deposition of radioiodine-A review. Health Phys. 9:1133-1139; 1963.
Garner, R. J. A mathematical analysis of the transfer of fission
products to cows’ milk. Health Phys. 13:205-2 12; 1967.
Garner, R. J.; Russell, R. S. Isotopes of iodine. In: Russell,
R. S., ed. Radioactivity and human diet. New York: Pergamon Press; 1966: 297-3 15.
Harley, J. H.; Hallden, N. A.; Ong, L. D. Y. Summary of
gummed-film results through December 1959. New York,
NY: U.S. Atomic Energy Commission, Health and Safety
Laboratory; HASL-93; 1960.
Hicks, H. G. Results of calculations of external y radiation exposure rates from fallout and the related radionuclide compositions. Livermore, CA: Lawrence Livermore National
Laboratory; UCRL-53152; 1981: parts 1-8.
Hoecker, W. H.; Machta, L. Meteorological modeling radioiodine transport and deposition within the continental United
States. Health Phys. 59:603-6 17; 1990.
Hoffman, F. 0. The coefficient for the transfer of radionuclides
from animal intake to milk, F(m).In: A statistical analysis of
selected parameters for predicting food chain transport and
668
zyxwvutsrqponm
zyxwvutsrq
Health Physics
internal dose of radionuclides. Oak Ridge, TN: Oak Ridge
National Lab; ORNL/NUREG/TM-282; 1979: 64-79.
Hoffman, F. 0.; Frank, M. L.; Blaylock, B. G.; von Bernuth,
R. D.; Deming, E. J.; Graham, R. V.; Mohrbacher, D. A.;
Waters, A. E. Pasture grass interception and retention of I3'I,
'Be, and insoluble microspheres deposited in rain. Oak Ridge,
TN: Oak Ridge National Laboratory, Environmental Sciences
Division; Report ORNL-6542; May 1989.
Horton, R. E. Rainfall interception. Mon. Weather Rev. 49:
603; 1919.
Kirchner, T. B.; Whicker, F. W.; Otis, M. D. PATHWAY: A
simulation model of radionuclide transport through agricultural food chains. In: Analysis of ecological systems: Stateof-the-art in ecological modelling 5. New York Elsevier Applied Science Publishers; 1983: 959-968.
Lloyd, R. D.; Gren, D. C.; Simon, S. L.; Wrenn, M. E.; Hawthorne, H. A.; Lotz, T. M.; Stevens, W.; Till, J. E. Individual
external exposures from Nevada Test Site fallout for Utah
leukemia cases and controls. Health Phys. 59:723-737; 1990.
Miller, C. W. An analysis of measured values for the fraction of
a radioactive aerosol intercepted by vegetation. Health Phys.
38:705-712; 1980.
Miller, C. W.; Hoffman, F. 0. The environmental loss constant
for radionuclides deposited on the surfaces of vegetation, A,.
In: A statistical analysis of selected parameters for predicting
food chain transport and internal dose of radionuclides. Oak
Ridge, TN: Oak Ridge National Lab; EG/TM-282; 1979:
43-50.
National Research Council. Nutrient requirements of dairy cattle. 5th revised ed. Washington, D.C.: National Academy
Press; 1978.
Ng, Y . C.; Colsher, C. S.; Quinn, D. J.; Thompson, S. E. Transfer
coefficients for the prediction of the dose to man via the
October 1990, Volume 59, Number 4
forage-cow-milk pathway from radionuclides released to the
biosphere. Livermore, CA: Lawrence Livermore National
Laboratory; UCRL-5 1939; 1977.
Perkins, R. W. Physical and chemical form of "'I in fallout.
Health Phys. 9: 11 13; 1963.
Perkins, R. W.; Thomas, C. W.; Nielsen, J. M. Measurements
of airborne radionuclides and determination of their physical
characteristics. In: Radioactive fallout symposium series 5.
Springfield, VA: National Technical Information Service;
CONF-765; 1965.
Romney, E. M.; Lindberg, R. G.; Hawthorne, H. A.; Bystrom,
B. G.; Larson, K. H. Contamination of plant foliage with
radioactive fallout. Ecology 44:343-349; 1963.
Simon, S. L. An analysis of vegetation interception data pertaining to close-in weapons' test fallout. Health Phys. 59:
619-626; 1990.
Thompson, C. B. Estimates of exposure rates and fallout arrival
times near the Nevada Test Site. Health Phys. 59:555-563;
1990.
Voilleque, P. G . Iodine species in reactor effluents and in the
environment. Palo Alto, CA: Electric Power Research Institute; EPRI Report NP- 1269; 1979.
Voilleque, P. G. Initial retention by vegetation of "'I in wet
depositions of fallout. Report prepared for the National Cancer Institute; 1986. (Available from Paul G. Voilleque, SAIC,
101 South Park Avenue, P.O. Box 50697, Idaho Falls, ID
83405-0697.)
Wachholz, B. W. Overview of the National Cancer Institute's
activities related to exposure of the public to fallout from the
Nevada Test Site. Health Phys. 59:5 1 1-5 14; 1990.
Whicker, F. W.; Kirchner, T. B. PATHWAY A dynamic foodchain model to predict radionuclide ingestion after fallout
deposition. Health Phys. 52: 717-737; 1987.
zyxwvu
zyxwvutsrqpo