Statistical Characterization of Radiation Doses from External Exposures and Relevant Contributors in Fukushima Prefecture

Abstract

1In areas contaminated by radioactive materials, well-designed dose assessment is necessary in order to protect people from radiation exposure and manage the exposure situation appropriately. Probabilistic dose assessment is a useful method for providing a more complete characterization of information on dose distributions in the population, and requires statistically characterized data on pathway-relevant contributors. The objective of this paper is to determine statistical features of contributors to external exposures, as well as to identify causes of variabilities of individual doses to the populations living in areas contaminated by the Fukushima Daiichi Nuclear Power Plant accident. To achieve these objectives, measurements of individual doses and ambient dose rates, as well as surveys of behavioral patterns, were performed between February and April, 2012. These were made with the cooperation of indoor workers, outdoor workers, and pensioners living in Fukushima prefecture. On the basis of these results, statistical analyzes were performed in order to identify variabilities of contributors. In addition, a multi-regression analysis was made to explore a significant relationship between individual doses and relevant contributors. As a result, the ambient dose equivalent rate also distributed with lognormal form and it had the variabilities attributable to the spatial distribution of deposited radionuclides. The distribution forms of time spent outdoors depend on the characteristics of occupation, and those of indoor workers and outdoor workers were lognormal and normal, respectively. Results of a multiple-regression analysis suggested that such valiabilities of contributors give rise to significant differences of individual doses among the populations.

A Probabilistic Approach to Assess External Doses to the Public Considering Spatial Variability of Radioactive Contamination and Interpopulation Differences in Behavior Pattern

“…The measured ambient dose rate was in the range of 0.07–0.55 μSv h −1 inside houses and 0.07–1.76 μSv h −1 outside houses. The GM and GSD of those measured outside houses were 0.35 μSv h −1 and 1.70, respectively . This result does not contradict the results measured by Fukushima prefecture described in Section .…”

“…By contrast, if the measured individual doses are classified in terms of population groups, the ratio can be reduced to approximately 3 in case of the participants from the Fukushima City Office, the Senior Citizens’ Club, and the Contractors’ Association. Our previous study, in which we statistically analyzed individual doses measured from February to April 2012, suggested that the homogeneity requirement could be satisfied for these three population groups. However, even after the data were classified with respect to participant group, clear nonhomogeneity was found in individual doses received by participants in the Agricultural Cooperative population.…”

“…Thus, the ratio between the maximum and the minimum values should be less than 3. Here, we used GMs for evaluating the ratio because the individual doses were distributed in lognormal form, as mentioned in our previous study …”

“…The adjusted R ‐squared statistic for this correlation is 0.45, which means that there was no clear correlation. However, based on the results of the F ‐test, the ambient dose rates measured inside the houses are explained well by the dose rates measured outside . We used this correlation to develop a dose assessment model as described in Section .…”

“…In this study, we focus on the variability resulting from the spatial distribution of contaminants and the interpopulation difference of behavioral patterns. These factors also give rise to a statistical spread in the distribution of doses among the people living in the areas contaminated by the Fukushima nuclear accident …”