Submesoscale Mixing on Initial Dilution of Radionuclides Released From the Fukushima Daiichi Nuclear Power Plant

Abstract

AbstractThis study developed a submesoscale eddy‐resolving oceanic dispersal modeling system comprising a double‐nested oceanic downscaling model and an offline oceanic radionuclide dispersion model. This was used to investigate the influences of submesoscale coherent structures (SCSs) and associated ageostrophic secondary circulations (ASCs) on the three‐dimensional (3‐D) dispersal of dissolved cesium‐137 (137Cs) released from the Fukushima Daiichi Nuclear Power Plant (FNPP1). Extensive model‐data comparison demonstrated that the innermost high‐resolution model, with a lateral grid resolution of 1 km, could successfully reproduce transient mesoscale oceanic structures, the Kuroshio path and stratification, and spatiotemporal variations of 137Cs concentrations. Using an accompanying mesoscale eddy‐resolving model (grid resolution: 10 km) as a guide, we showed that submesoscale dynamics are important for improved representation of both the eddy field and the resultant 3‐D dispersal of 137Cs, with the temporal variability of surface 137Cs near the FNPP1 being equivalent to that in the coarse‐resolution model. According to energy conversion and spectral analyses, SCSs and ASCs occur most intensively on the submesoscale, primarily because of shear instability. However, baroclinic instability serves as a secondary mechanism. SCSs have prominent seasonality, reflected by intensification in the colder months, which is when the FNPP1 accident occurred. Analysis of the vertical flux of 137Cs was performed by decomposition of the variables into eddy, mesoscale, and submesoscale components using frequency and wave number filters. It revealed that 42.7% of the FNPP1‐derived 137Cs was transported downward below the mixed layer by eddies with the major contribution being from ASCs induced by submesoscale eddies.