In order to put in the stakes for the 50 m grid on Janet, most of the island required extensive devegetation efforts. The primary method was to bulldoze the vegetation into long, east-west windrows. One effeet of this method was to reduce the apparent effects of the wind on the distribution of TRU activity (Section 6.4), and another was to decrease the measured surface activity. Because the raw variogram (Section 5.1.1) was also affected, the statistical results on the Test Grid data could not be used. Therefore, the analysis was repeated, the two candidate models on the 50 m data were tested, and the better one chosen to estimate 0.25 hectare average TRU activity. There were two areas of Janet where neither model estimated well due to higher variability in the physical distribution of contamination. These areas were staked and measured on a 25 m grid to provide more data. The in situ sampling of the west area began 23 August 1977, and this area plus the 50 m grid were completed 16 November 1977. The two additional 25 m areas were sampled by the IMP from 6 January to 8 February 1978. Figure 7-66 shows the areas estimated to have TRU activity above 40 pCi/g on the 0.25 hectare averages, using all the 25 m data as well as the 50 m data. (Note that these estimates were based on original data. See Tech Note 23 for discussion of original versus final data). The total area shown in Figure 7-66 as having 0.5 s upper bounds on the TRU activity estimates above 40 pCi/g is 20.75 ha, where s is the standard deviation of the kriging error; without the additional 25 m data, the estimate was 21.25 ha. In orderto arrive at estimates of TRU from IMP 24l4m data, soil samples were taken to determine the ratio of TRU to *4lam. Two composites were taken at each of 29 locations, using the method described in Section 4.2.1, for a total of 58 samples.* The locations sampled are shown in Figure 7-65. The estimated ratios of TRU to 24!Am fell into two distinct groups corresponding to location on Janet. All the samples from the Easy/X-Ray area on the west tip had higher ratios than the samples from elsewhere_on Janet. The change from one ratio to the other was abrupt, matching an abrupt change in the 24lam data from the IMP, as well as a distinct change in soil characteristics. The change in the soil, visible on the 1972 aerial photographs, also matched an abrupt drop in gamma activity measured in the 1977 aerial survey (see Section 3.1). The boundary between populations of ratio of TRU to 241Am was therefore drawn on the basis of the 1972 aerial photographs, and is shown in Figure 7-65. The ratios of TRU to 241 am used for the initial characterization and cleanup were 5.34 + 0.69 for the west area and 3.32 + 0,42 for the rest of the island. Surface Cleanup The surface cleanup of Janet was accomplished in stages, with the first lifts coming from the areas with the highest activity. All areas with average TRU activity exceeding 60 pCi/g had already been measured by the IMP at 25 m spacing as part of the additional work on the two small areas, About half the area with TRU activity between 50 and 60 pCi/g had also been measured by the IMP on a 25 m grid. No further fine grid surveys were made until all the areas with average TRU activity exceeding 50 pCi/g had been lifted. It was recognized at that point that the total amountof soil to be removed could be minimized by taking more data to refine the excision boundaries, The remaining areas with TRU activity greater than 40 pCi/g were therefore measured with the IMP at 25 m spacing before being lifted, The fine grid survey was also extended 25 m beyond the above 40 pCi/g areas to allow better revised estimates. After each soil lift, the lifted area plus a boundary of points beyond the lift were measured with the IMP. New estimates were computed by averaging the IMP data values, since kriging is not the best method to use for data from a 25 m grid; the detector field of view includes most of the surface at 25 m spacing (see Section 5.1.1). If the new TRU estimatestill exceeded 40 pCi/g, the sequence of lifting and remeasuring was repeated, although very few areas actually required additional lifts. To save time and maintain a smooth operation, fine grid IMP surveys, lifts in areas already measured, and post-lift IMP surveys were done concurrently in different parts of the island. *Results from only 50 of the samples were actually used in the ratio computation. See Tech Note 2.6. 284