Y(v,v) = 0.5214 w L. It may also be applied to spherical variograms as long as L is small compared with the range, since then the variogram is practically linear. Applying this formula to our fitted variogram models, we get: Y(v,v) = (1.287)2 x 0.5214 x 1.53 x 10 * x 100 + 0.5214 x (3/2) x (0.025/600) x 100 = 0.016 and the multiplicative factor is 1.04. RESULTS The figures that follow show the different steps of the estimation (all concentration results are in uCi/m): Figure 8 shows contours of Pu values derived from the 100-foot FIDLER grid data through straight regression (no kriging) using Equation 2 followed by taking antilogarithms. Blanks within the 100-foot grid indicate missing values. We note that the FIDLER stratum boundaries of 50,000, 25,000, 10,000, and 5,000 counts per minute (Figure 1) become Pu concentrations of 192, 79, 24, and 9.9 uCi/m“, respectively, using Equation 2. Figure 9 shows contours of the mean 100~ x 100-foot cell Pu concentrations estimated by kriging using Equation 9 on the basis of FIDLER data, i.e., [oe * s x fio wtofK/2 - Dupe i , where o2. F (15) is the kriging variance for the FIDLER data for a given ceft. In this estimation, all FIDLER data were used including those at random locations and the 400~-foot grid (useful in the edges). The bias correction factor, i.e., 2,2 2 arises from Equation 14 applied to the guess-field (10 oF(v) + B neglecting the 10107) term. The bias correction factor Note the general similarity ranges between 1.05 and 1.07. between Figures 8 and 9. The principal difference is that the kriging contours are "smoother" in appearance. 387 )>