rameters a, a2, and a; were determined by the
X(2k—-1)
M,=a,P,S8,+Pe™ 3 De ?
k=1
t+o\,,(1—P,)
[8]
where
M, =milk level in the n‘* month expressed in
micromicrocurles per liter
P,=proportion of cows on pasture in the
a‘ month
Sn ==cumulative level of strontium 90 in the
soil in the m** month expressed in millicurles per square mile
A ==pasture depletion constant, the reciprocal
of the half-residence time
D,;, = deposition during the &* month expressed in millicuries per square mile per
month
Ms average milk level over the closest pre-
ceding harvest season expressed in micromucrocuries per liter
This equation has the form of a multiple
regression
M,=a0;X\, +arNon +03Xan
[9
with independent and dependent variables as
indicated. The basic time period will be taken
as l month with 7 = 1, 2, 3, .. ., 638 months as
the total time covered extending from July 1957
to September 1962. The values of the variables
Xin, Vo, and 3, were computed and the paTable 2.
Station
least squares analysis using a computer. Computation times involved are on the order of 14
hour per station. Values for deposition and soil
levels, when available, were extracted from the
Health and Safety Laboratory report made by
the Atomic Energy Commission.
Unfortunately, there were a few missing
monthly deposition values in each of the
stations. However, if the particular monthly
station values are compared with the overall
average deposition rates by month, a constant
relation is observed. These results are plotted
in figure 2 for the five stations. Asan example,
suppose that the value for New York is missing
for June, but that the corresponding overall
average is 4 millicuries per square mile, Then,
the estimated station value is 6 millicuries per
square mile. A conservative estimate for the
spring of 1962 is that the deposition rates will
be the same as during the spring of 1959. The
maximum deposition rates which would occur
under the worst possible assumptions are
assumed to be about twice the expected rates.
The average soil levels are based on those
estimated by Knapp.
The estimated milk levels do not differ greatly
with different values for 71/2, although they get
progressively worse with the longer half-times.
The optimum half-time for each station was
Average proportion of cows in pasture during each month
| Jan.
Feb. | Mar.
Apr.
May
June
July
Aug. | Sept.
Oct.
Nov.
0.20;
. 00
0620)
. 00
0.22)
. 00
0.36,
. 00
1.00;
. 62
1.00
1.00
1.00);
1.00
1.00
1.00
1.00
1.00
0.98
. 87
0.78
. OL
0. 37
. 01
St. Louis.____________-- 1,01
. 02
13
94
OF
.o7
97
.938
. 45
. 09
0:
Cincinnati__.._..2-_ 2 New York City. _____.-
Sacramento _____ lee eee
Salt Lake City. ..__.___- |
. 08
. 00
Table 3.
. 10
. 00
. AT
. 00
. 81
. 00
.84)
. 61
. 84
. 84
. 71
5
. 84
.95 ,
. 75 |
. 83
. 78 |
.7
. 45
. 02
J 1
. Ol
Model parameters yielding best fit for each station
oo
Station
Ti
Cincinnati_.._.-...--._.- ee. woe ene
New York City... ---- © 7+ - oe aoe
Sacramento__-.--.-- 2 eee ee
0. 75
.75
2. 00
0. 10
. 10
. 02
Salt Lake City__...___-----__--- + --_2-- Lie ee
2. 00
. 04
St. Louis.._.--......0-- == ee ee eee eee
1060
. 32
Dee.
25
ay
21
Qs
3.7
2.7
3. 2
13.7
1.1
as
Percent
variation
accounted
for
1.1
.9
1.0
6
4
~
.8
4
1.0 |
é
Public Health Repor