i
Figure 10 shows contours of the mean block correction term computed as
lao * ) ; [rotlode!2 “Eu R 1]
,
(16)
where OF ae is the kriging variance for a given block for the
*
residuals T (x).
Formula 16 was calculated by kriging using
Equation 8 on the basis of the 145 residuals T(x,) (see Equation 11).
1.04.
The bias correction factor is between 1.01 and
We note that the mean block correction term is greater
than a factor of two only to the south and east of GZ where it
ranges between 2 and 4.
The increase south of GZ is caused by
soil Pu concentrations that were higher than would be expected
on the basis of FIDLER reagings on the 100- x 100-foot grid in
the general vicinity**.
T (v) should be greater than 1 when
,
measured Pu is higher than expected from FIDLER, and less than
1 when Pu is less.
Figure 11 shows contours of the product of Formulas 15 and 16.
These
estimates are those obtained using the entire two-stage estimation procedure.
Note from Formula 15 that we are ignoring the
drift m(x) within 100- x 100-foot cells as discussed in the
previous section,
Figure 12 gives the numerical estimates shown in contour form in Figure 1}.
The FIDLER strata boundaries in Figure 1 are superimposed on
the two~stage kriging results in Figure 12 for comparison
purposes.
The reader may wish to draw on Figure 12 a contour
line connecting cell averages of 9.9 uCi/m2,
the Pu concentra-
tion corresponding to the 5,000 cpm FIDLER line using Equation 2
(see discussion under Figure 8 above).
You will find that the
total land area enclosed by this line is larger than the area
The
increase occurs mostly to the south and east of GZ, and this
results from the rather large positive block correction terms
in those regions (see discussion above concerning Figure 10).
**kIn 1976 six new soil samples were collected south of GZ,
2 north of GZ,
and 2 east of GZ. Their locations and Pu concentrations (Ci/m?)
are given in Figures 3 and 4. Four of the samples south of GZ (Pu
concentrations of 11.2, 173, 326, and 97.8 uCi/m2) were used in the
kriging computations.
FIDLER values (taken from the 100-foot-grid
readings) used for these Pu concentrations are 2,300, 4,300, 5,300, and
2,400 cpm, respectively.
The data collected at the remaining 2 locations
within the 100- x 100-foot-grid area (south of GZ) were inadvertently
left out of the analysis.
If these two concentrations (328 and 573
uCi/m2) had been included, the kriging estimates would have been larger
in that area.
The four new observations north and east of GZ were not
used since they were collected outside the 100- x 100-foot-grid area.
390
ae eMee ee ee
enclosed by the 5,000 cpm FIDLER line drawn in Figure 12.