Long-term prediction and fluctuations of the concentration of Cs-137 aerosol in Chernobyl , Fukushima and Kawasaki

Statistical analysis was applied to the time series of the airborne concentration of Cs-137. In order to extract the fractal characteristics of the fluctuations, we employed the Hurst analysis. Interestingly, the Hurst index is around 1/3, which is common to the Chernobyl data, Fukushima data, and Kawasaki data. The Kawasaki data have been measured by the Tokyo City University, located at 30km south to Tokyo. We proposed a stochastic differential equation, based on the advection equation with winds fluctuating probabilistically. The averaged solution of the equation was compared to the measured data. We found that the index of the power of the time series is almost -4/3, which is common to the three cases, Chernobyl, Fukushima and Kawasaki. (author)