would have the benefits of increasing the precision of inventory estimates over what would likely result if stratification were not done.

The first two safety-shot sites to be studied intensively by the NAEG
were Area 13 (Project 57) and Area 5 (GMX) in that order.
Based on the
results in Eberhardt and Gilbert (1972), the decision was made to use
stratified random sampling as the basic approach to estimating Pu inventory.
This was a conservative approach in the sense that we would not
be relying on the correlation between Am and Pu concentrations to estimate Pu inventory, i.e., Pu inventory would be estimated by obtaining
wet chemistry Pu determinations on all soil samples collected.
Ge(Li)
scans for Am on soil samples (same aliquots as used for Pu determinations) would be done on all samples primarily (i) to estimate Pu to Am
ratios,

(ii)

to relate Am concentrations in soil to those in other

ecosystem components like vegetation and small mammals, and (iii) to
evaluate more fully the feasibility of using Am to help estimate Pu
inventory,

An extensive number of FIDLER readings at 1-foot height were taken on

grid systems to estimate the spatial distribution of Am.
This information was used to divide the area into activity strata within which soil
samples were collected at random locations.
Before each surface soil
sample* was collected at 1-foot height, net FIDLER reading was taken
over the sampling location.
These FIDLER-Pu data were used to evaluate
the correlation between FIDLER count per minute readings and Pu wet
chemistry concentrations on 10-gram soil aliquots removed from the soil

samples. The details of the design of these studies are given in Gilbert
and Eberhardt (1974), Gilbert et al. (1975), and Gilbert et al. (1976a).

Once the studies at Project 57 and GMX were under way, FIDLER surveys
began at Clean Slates 1, 2, 3, and Double Tracks on TTR, and in Area ll

(Plutonium Valley) on NTS for the purpose of establishing strata.

The

same basic design used at the Area 13 (Project 57) and GMX sites (stratified random sampling relying on Pu wet chemistry analyses) was used at
these new study areas.
The first estimates of inventory to be published appeared in Gilbert and
Eberhardt (1974) for Area 13.
Estimates and their standard errors were
given for individual activity strata as well as for the total area (sum
of the six strata areas).
A year later, estimates of Pu inventory were

published (Gilbert et al., 1975) for all 10 safety-~shot sites except A

Site in Area 11 where data were not then available.
The inventory
estimates for the four TTR sites and Sites B, C, D, CD overlap, and ABCD
overlap in Area 11 were preliminary since only 40 percent of the samples
collected had been chemically analyzed and reported up to that time.
Estimates of 239?240py and total uranium inventory in vegetation are

given in Romney et al., 1977.

*5-inch-diameter ring to a depth of 5 cm.

453

Select target paragraph3