males and females likely to be caused by chronic low-LET radiation exposure and associated analysis of statistical and model-related errors, the uncertain factor W was taken to be approximately lognormally distributed with expectation 0.0004 cSv"! and SD, = 0.864. Based on the analysis of interindividual variability in expected dose, it was calculated that the expected value_of 30-y integral summarized in Figure 20 indicatejthat 99.5% of hypothetical 1995 Rongelap res#ttiers would never receive a 1-y dose greater than 1 mSvif imported foods were routinely co. , but that ~25% would receive maximum 1-yj doses greater than 1 mSv if only local foods wer¢ consumed at twice the caloric intake rate Jindicated in Table 20. population-average dose, (D(30)) is ~0.58 cSv, 1 and that the chance that (D(30)) > 2.0 cSvis ~1%, e.g., indicating that 2 cSv is the 30-y dose most likely to be incurred by the fifth highest exposed among 500 hypothetical Rongelap residents (Figure 17). The predicted relationship between cumulative exposure time t 2 0.38} S = o.6} $ ‘@ 0.4 e3 0.2 and interindividual variability in (D(t)) (Figures 18 and 19) indicates that the lower and upper 95% confidence limits on (D(t)) variability are ~2-fold below and ~2.5-fold above, respectively, the population-average expected- value function (D(t)). Oo 0 expected dose due to 137Cs ingestion than predicted by the LLNL imports-available diet. The distribution corresponding to the LNL 95th and 99th percentile values 2.0 cSv, respectively. The results 15 1.75 2 (cSv) 0.48, 1.2 and 2 1.2 RB o 4 Cos ® Z 0.6} o 33% of residents, respectively. The distribution corresponding to the local-foods-only diet with twice the local calorie intake indicated in Table 20 (light curve in Figure 20) has a mean of 0.83 mSv, and has 50th, 95th, 99th and 99.8th percentile values of 0.61, 2.2, 3.9 and 5.8 mSv, respectively, with maximal doses predicted to population of 500 (NRC, 1993). i Bt the 2nd and 3rd years of residence for 66% and respectively. Note that a 99.8th percentile dose indicated in Figure 20 corresponds to the most likely value of the greatest maximum 1-y dose predicted assuming a 1995 resettlement 1 dose corresponding to hypothetical residence on Rongelap Island starting in J1995. This distribution has a mean of 0.58 FSv and 50th, 95th, 99th and 99.8th percentile values of 0.21, 0.52, 0.87, and 1.3 mSv, respectively. The maxima of expected annual doses under this dietary scenario are estimated to occur during residence for 44.5%, 53%, and 2.5% of residents, 1 Figure 17. Estimated disfribution of interindividual variability in qxpected 30-y imports-available model diet (bold curve in Figure 20) has a mean of 0.25 mSv, and has 50th, occur during the 2nd, 3rd and 4th years of 0.25 05 0.75 Cumulative 30-y doge The calculated interindividual variability in expected maximum 1-y dose is shown in Figure 20, contrasted to variability in that dose estimated assuming a hypothetical LLNL-type localfoods-only diet with twice the local calorie intake shown in Table 20. Such a local-foodsonly diet implies a nearly 5-fold greater ey c 3B 0.45 5 0.2} 0 Time (y) Figure 18. Population-average Pxpected dose from hypothetical residence on Rbngelap Island starting in 1995 (middle curve) corresponding two-tail 95% confidencef limits on interindividual variability in exgected dose, as a function of residence time. 48