On the validation of an Eulerian model with the Fukushima Nuclear Power Plant (FNPP) accident. Global and local results for Europe and Japan

A large debate about the exact emissions after the Fukushima NPP accident is still ongoing more than 3 years after the original disaster. Terada et al. (2012) reported the total release of 137Cs to be 13 PBq (x1015 Bq), based on an inverse modelling using Japanese data only, whereas the IRSN reported releases of 137Cs to be 20.6 PBq (IRSN, 2011). In the present study, we used the emission inventories for 137Cs and 133Xe reported by Stohl et al. (2012) estimated by inverse modelling using the CTBTO (Comprehensive Nuclear Test Ban Treaty Organisation) and Japanese networks (36.7 PBq of 137Cs and 15.3 EBq of 133Xe). For the simulations of the accident, three different versions of the LMDZORINCA model were used; A regular one with a grid resolution of 2.50 deg. x 1.27 deg. for the global comparison with the CTBTO network (19 and 39 vertical layers), and a zoom version over Europe and Asia (0.45 deg. x 0.51 deg. for 19 levels) resulting after 'stretching' the grid using the same number of grid points to assess what happened in Greece, and Japan. Cesium isotopes were treated as sub-micronic aerosols, whereas 133Xe as a passive tracer within the model, whereas several other radionuclides were estimated from reported isotopic ratios. Our results for the global assessment fit well to the observations. They differ about 0.04% from the measurements for 137Cs, and around 40% for xenon. The most significant deviations were observed for the northernmost stations due to both scavenging processes and transport over the Arctic. Scattered measurements of several radionuclides from Japan were adopted from literature (Shininaga et al., 2014). The comparison showed a significant quality of our model, although some isotopes are miscalculated. This shows that the reported isotopic ratios might be biased somehow. Finally, in Greece, few measurements of 131I, 134Cs and 137Cs were adopted from Potiriadis et al. (2013), as well as from a northernmost region (Bulgaria, NCRRP, 2013). The correlation improves to the north probably due to the higher abundance of the radionuclides in the atmospheric aerosol. A general comment is that the source must be re-assessed by the official authorities. Our model, as well as many others employed by several modeling groups, concludes that the initial emission of 137Cs and 133Xe must be doubled. (authors)