Inversion of Critical Atmospheric 137Cs Emissions Following the Fukushima Accident by Resolving Temporal Formation from Total Deposition Data

Understanding the transport of 137Cs emitted during the Fukushima accident is challenging because the critical emissions that produced the high-deposition area are not adequately resolved in existing source terms. This paper presents an objective inverse reconstruction of these emissions by fusing atmospheric concentrations with a-priori emissions extracted from total depositions. This extraction, previously considered impossible for complex real-world accidents, is achieved by identifying the critical temporal formation process of depositions in the high-deposition area and estimating the corresponding emissions by using an atmospheric transport model. The reconstructed source term reveals two emission peaks from 10:00–11:00 and 14:00–15:00 on March 15, which agree with the in situ pressure measurements and accident analysis, suggesting that they came from pressure drops in the primary containment vessels of Units 3 and 2, respectively. This finding explains the environmental observations of spherical 137Cs particles. The source term also objectively and independently confirms the widely used reverse estimate. The corresponding 137Cs transport simulations better match the various observations than those produced by other source terms, proving that the two-peak emission creates a high-deposition area. The proposed method outperforms the direct fusion of deposition and atmospheric concentration observations, providing a robust tool for multiobservation fusion.