Uncertainty in Aerosol Rainout Processes through the Case of the Radioactive Materials Emitted by the Fukushima Dai-ichi Nuclear Power Plant in March 2011.

The process of an aerosol rainout in wet deposition induces large uncertainties among atmospheric aerosol simulations, especially for particles in the fine mode. In this study, we performed an intercomparison study of four different rainout schemes on the model (the nonhydrostatic icosahedral atmospheric model or NICAM) to simulate particulate Cs-137 in the emission scenario of the March 2011 accident at the Fukushima Dai-ichi Nuclear Power Plant. The schemes include global climate models (GCMs) approach with a simple tuning parameter to determine the scavenging coefficient, and another optimized for cloud resolving models (CRMs) to account for prognostic precipitation and realistic vertical transport. The third approach was the conventional method under the assumption of a pseudo-first-order approximation based on the surface precipitation flux. The fourth approach was involved in offline chemical transport models (CTMs) with a simplified parametric analysis approach to clouds and precipitation flux. In most experiments, statistical metrics of the Cs-137 concentrations using in-situ measurements were calculated to be within ±30 % (bias), 0.6 - 0.9 (correlation), 67 - 112 Bq m-3 (uncertainty), and < 40 % (precision within a factor of 10). The CRM-type method yielded the best results but required a lower limit of tuning parameters to compensate for the results. Both the GCM-type and the conventional methods were also useful by setting proper tuning parameters. The CTM-type yielded better correlation and lower uncertainty but larger negative bias. These analyses suggest the overestimation of the conversion rate from cloud droplets into raindrops by the NICAM. However, this cannot be resolved by simply interchanging cloud microphysics schemes. It was found that the sensitivity of the rainout scheme has a stronger influence on the Cs-137 concentration than the different treatments of cloud microphysics. Thus, to replicate the observed Cs-137 distribution, it is essential to have a better meteorological field as well as a proper rainout scheme. [ABSTRACT FROM AUTHOR]