Hence, the two-stage kriging results place the 9.9 uCi/m? line further from GZ than do the straight regression estimates (no kriging) given in Figure 8. Study of Figures 8 and 9 suggests the Ist stage kriging estimates (Figure 9) are in close agreement in this regard with the regression results (Figure 8). This is not surprising since the lst stage results are based solely on FIDLER data. In regions near GZ where more data is available, the inner two Pu contour lines of 192 and 79 uCi/m? (corresponding to 50,000 and 25,000 cpm using Equation 2) are in about the same locations for both the lst and 2nd stage kriging results. * Figure 13 shows contours of estimates 10° (v) i.e., of * 10 * oF (v) +R8+ T (v) (17) ’ P*(y) P(v) where 10 (v estimates the geometric mean 10 Vv’. These contours are most similar to those obtained for the two-stage results. The two differ by factors less than 1.2 (see Figure 14). Figure 14 shows contours of the overall bias correction factor, i.e., 2152/2 - [rol [opy/ & 272 LE 4] * [rol ote & 2 Hef 1] , (18) as it results from Equation 14, neglecting 107/10") Hence Figures 11 and 12 are the product of Formulae 17 and 18. Formula 18 is the product of the bias correction factors for FIDLER regression (stage 1) and T (v) (stage 2). Figure 15 shows 107° for use in Formulae 13. To get a 95 percent confidence interval for the geometric mean 10° ) | it suffices to divide and multiply Formula 17 by 10°°x, Up to a multiplication by 1.04, this interval is also valid for the mean Pu concentrations Pu(v) in Figure 12. Our final (two-stage) estimates (Figures 11 and 12) seem to be in good agreement with the observed Pu data at most locations (see Figures 3 and 4). However, reservations must be made in the region of very high levels of activity in stratum 6 near GZ where estimates appear to be too low. This may result from the log-transformation which has the effect of downweighting large Pu concentrations. The estimates for the two cells adjacent to GZ are 826 and 832 uCi/m?, while soil sample values of 1,170, 1,300, and 14,300 uCi/m? were observed in that general area (Figure 4). Very high concentrations probably do exist in the immediate vicinity of GZ, but it is not clear to what extent local highs may be "extended" to larger areas. Note, for example, that such highs are 394