An estimate of the average intake of 239,240py per day by a cow grazing the
inner compound is computed from average 239,240by concentrations for shrubs
Eurotia lanata and Atriplex canescens. This result (620 + 130 nCi/day) agrees
quite well with the 560 nCi/day estimated by Smith et al. (1976a) using average
239,240p, activities in the vegetation and liquid components of rumen contents
of fistulated steers grazing the inner compound of Area 13.
An estimate of
770 nCi/day was obtained using Martin and Bloom's model assuming 500 grams of
soil is ingested per day by grazing cattle.
The second part of this paper discusses some aspects of displaying and analyzing
skewed data sets. Some relationships between the arithmetic mean, geometric
It is shown that estimates of inventory
mean, median, and mode are discussed.
can be greatly influenced by the particular measure of central tendency (average)
The point is made that the choice of
used to characterize a skewed data set.
a measure of central tendency depends on the objective of the study and the
Stem-and-leaf displays for displaying the
use to be made of the estimate.
information content of a data set are discussed and illustrated using soil and
rodent plutonium concentrations.
The third and final portion of this paper reviews some of the bias problems we
have encountered in attempting to estimate plutonium concentration contours in
Area 13 (Project 57) and Area 5 (GMX site).
The possibility of estimating
inventory by first estimating the spatial pattern of plutonium concentrations
in surface soil is suggested.
An evaluation of Universal Kriging for this
purpose is encouraged.
Finally, some estimated contours for Area 5 (GMX) are
displayed to illustrate what appears to be improved estimates using log-trans-
formed data.
Contours obtained using polynomial fits are shown to be susceptible
to large biases in areas of the study site where few data were collected.
FUTURE PLANS
We expect to:
1.
Continue to provide statistical help in the synthesis of safety-shot data
for 239Pu and for other radionuclides (238py, 24 lam, 238y),
2.
Continue our assistance to program investigations in the analysis of
3.
Help design environmental transuranic studies at nuclear sites on NTS.
4.
Identify or develop efficient field sampling designs for estimating
spatial patterns (geographical distribution).
5.
Continue to experiment with estimating Pu concentration contours using
their data.
computer algorithms.
270