Le 0021620 used in Shots Diablo and Shaste .24 The Coulomb C samples were surface soil samples which, by definition, had no sampling bias. The sample bias factors, q/q, are used in the next section, where applicable, to increase the values of a (fissions/sq ft) for calculation of M,(1) from I(1) and a/m values. For the stations at which I(1) was not observed, the average value of the ratio, q/q, was used to increase the FOD (fraction of device) per sq ft values and, by use of Eq.. 25, to give estimates of I(1). 4.6 COMPUTATION OF K(x,W) FROM MASS CONTOUR RATIO DATA The values of K(x,W) can be determined by means of Eq. 24 and the observed or estimated values of M.(1). In order to increase the mmber of data points in determining the dependence of K(x,W) on x, the I(1) values were estimated for the stations at which observed values were not available (i.e. mainly the floating stations for Operations CASTLE and REDWING) by the method described in Section 4.5. In addition, . correlations were made of the variation of the specific activity of the fallout with distance from the data of Table 1. Fig. 4. These are shown in The data are quite scattered with respect to variation with distances; in the calculations, the empirical equations for Zuni and Flathead were used but for Tewa and Navajo, the geometric means were used. Activity~-particle size data and specific activity data from PLUMBBOB Shot Shasta are given in Figs. 5 through 7. The mean values of the sizes and specific activities are summarized in Table 13 for each station; these values and an «xtrapolated value were used in Table 1 for the calculated values of M.(1). ‘The very small amounts of activity in small sizes (Figs. 5 and 6), if neglected, would result in distribu- tion curves with a fairly small particle size range for each station. In Fig. 7, it may be noted that, for the range of distances given, the specific activity is nearly proportional to xi/2. DOE/NV) The values of K(X,W)/q are summarized in Table 14. For the JANGLE and CASTLE data (except for Stations How and Victor), no correction was applied for sampling bias since no estimate was available to apply to the collectors used. No bias was assumed for the PLUMBBOB data. The K(X,W)/q values are plotted against the 1 MT Scaled Distance in Figs. 8, 9, and 10; the 1 MP Scaled Distance was used to adjust the numerical values of the distances for each shot to a convenient common range for plotting.