Uncertainty and variability in updated estimates of potential dose and risk at a U.S. nuclear test site--Bikini Atoll.

Uncertainty and interindividual variability were assessed in estimated doses for a rehabilitation scenario for Bikini Island at Bikini Atoll, in which the top 40 cm of soil would be removed in the housing and village area, and the rest of the island would be treated with potassium fertilizer, prior to an assumed resettlement date of 1999. Doses were estimated for ingested 137Cs and 90Sr, external gamma-exposure, and inhalation+ingestion of 241Am + 239+240Pu. Two dietary scenarios were considered: imported foods are available (IA); imported foods are unavailable with only local foods consumed (IUA). After approximately 5 y of Bikini residence under either IA or IUA assumptions, upper and lower 95% confidence limits on interindividual variability in calculated dose were estimated to lie within a approximately threefold factor of its in population-average value; upper and lower 95% confidence limits on uncertainty in calculated dose were estimated to lie within a approximately twofold factor of its expected value. For reference, the expected values of population-average dose at age 70 y were estimated to be 16 and 52 mSv under IA and IUA dietary assumptions, respectively. Assuming that 200 Bikini resettlers would be exposed to local foods (under both IA and IUA assumptions), the maximum 1-y dose received by any Bikini resident is most likely to be approximately 2 and 8 mSv under the IA and IUA assumptions, respectively. Under the most likely dietary scenario, involving access to imported foods, this analysis indicates that it is most likely that no additional cancer fatalities (above those normally expected) would arise from the increased radiation exposures considered.