Besides the radioactive contamination on Janet, there was also some chemical contamination by beryllium contained in rocket engine fuel. The engine was being tested in 1968 on a pad near Hardtack Station 1312 when it malfunctioned, damaging Station 1312 and contaminating the area with beryllium. The combination of decontamination efforts at the time of the incident and erosion since then should have removed most of the beryllium before the radiological cleanup began. 1972 Survey Results Because of its size and importance to the Enewetak people, Janet was sampled intensively during the 1972 survey. Out of a total of 140 soil sampling locations, ten were profile sampled to 185 em, two were 125 em profiles, one was a 65 em profile, one was a 35 em profile, and the remaining 126 were 0-15 em core samples. To help investigate relationships between radioactivity in the soil and in the food chain, a numberof plant and animal samples were taken. The 239,240py activity in profile samples generally declined steeply with depth, falling to less than 1 BSi/g by 30 em or shallower. Of three locations which were exceptions to the pattern, two had no 39,240py activity greater than 1 pCi/g at depth even though the activity was rising. The third location with an anomalous pattern was near the Easy and xoRay sites, where buried contamination wasalready suspected to exist. Table 7-3 summarizes the 239,240 py results for Janet. The depth distribution of 137Cs and 99Sr was similar to the pattern for 239,240py, although activity of these two isotopes did not decline_as steeply as 239,240py activity. The one profile which showed a significant increase in Cs and 9p activity below 30 cm was the same location near Easy and X-Ray which had the anomalous 239,240pu depth distribution. Summaries of the 37Cs and 90sr results are in Tables 7-1] and 7-2, respectively. In 1975, as part of the follow-up on the 1972 survey, Lawrence Livermore Laboratory established a garden plot on Janet to study radionuclide uptake in food plants. The results would aid in building dose-assessment models, and specifically to help determine when Janet might again be suitable for agriculture and habitation. Additional] soil samples were also collected in the garden area to provide better information on soil-to-plant transfer coefficients for radionuclides. A study of radionuclides in groundwater was also begun at this time; this involved drilling several wells and taking water samples. Both the garden and groundwater studies continued throughout and beyond the radiological cleanup, so care was taken during cleanup to try to avoid damageto the study areas. Surface Characterization Because Janet wasthefirst island measured with the in situ system, several preliminary experiments and sets of IMP measurements were done on Janet to develop procedures and evaluate the system. Details of these early efforts are in Section 6.4. After the initial break-in period, a 25 m grid, known as the Test Grid, was staked and measured in August 1977 to provide a test of the data collecting system and also data for preliminary statistical analyses. Although the absolute coordinates of the Test Grid were never established, its approximate boundaries are shown in Figure 7-65. As described in Section 6.4, the statistical analysis of the Test Grid data led to the conclusion that 50 m spacing for the Janet grid would give enough data for acceptable estimates. Meanwhile, part of the west tip of the island had already been staked at 25 m spacing and the IMP had nearly completed measurements in that area. (This 25 m grid was inadvertently shifted from its intended location. See Section 6.4). The IMP survey of the 25 m grid was therefore completed at that spacing in September 1977, and that block of data was handled separately in the statistical analyses. The remainder of Janet was initially staked and measured at 50 m spacing. The 50 m grid was located correctly, so it was extended far enough west to makecertain that estimates of TRU activity from the 25 m and 50 m grids would completely cover the island. It was further concluded that the kriging method (see Section 5.1) gave acceptable estimates, and the data satisfied the assumptions made in using this method. 283