2019 年 97 巻 1 号 p. 175-190
A set of atmospheric dispersion-deposition model (ADM) integrations were conducted with a hypothetical emission of radioactive materials consisting of 137Cs, 131I, and 134Cs from the Tomari Nuclear Power Plant (NPP) in Hokkaido, Japan, which is a snow climate site. Each integration was driven by Japan Meteorological Agency's (JMA's) mesoscale model (MSM) analysis data with 5-km horizontal resolution. The initial conditions were those on each day from January 2010 to December 2016, and the integration period was at most four days. The target was the area within 30 km of the plant. Extending a unit-mass emission concept, the measure of the relative risk is the probability of exceeding the threshold of the maximum effective dose rate based only on exposure from groundshine. Considering that the measure increased monotonically with the ratio of the total emission amount to the threshold, we evaluated the probabilistic risk with its median. The results suggested that the risk was higher in the eastern part of the target area owing to the prevailing westerly flow. The frequent snowfall in winter drags radioactive materials down in the target region, even under an active turbulent condition with strong vertical shear. The composite analysis of wind direction averaged over the target area revealed that the risk was high in the leeside but that mountains effectively blocked the inflow of the radioactive materials. The results were insensitive to wet deposition parameterization. The risk was reduced when we replaced the emission altitude with a higher one than the standard setting. The snow shielding effect was negligible on the short-term radioactivity just after the emission but was substantial on the seasonal change in radioactivity.
