H15-108: Estimation of the cesium-137 source term from the Fukushima Daiichi nuclear power plant using air concentration and deposition data

A major difficulty when inverting the source term of an atmospheric tracer dispersion problem is the estimation of the prior errors: those of the atmospheric transport model, those ascribed to the representativeness of the measurements, the instrumental errors, and those attached to the prior knowledge on the variables one seeks to retrieve. In the case of an accidental release of pollutant, and especially in a situation of poor observability, the reconstructed source is very sensitive to these assumptions. This sensitivity makes the quality of the retrieval dependent on the methods used to model and estimate the prior errors of the inverse modeling scheme. In order to use all the available data, we propose to extend the methods developed in Winiarek et al. (2012), which were designed for the inversion of one type of data, to the use of several types of data in the same inversion, such as activity concentrations in the air and fallout measurements. The idea is to simultaneously estimate the prior errors related to each dataset, in order to fully exploit the information content of each one. Using the activity concentration measurements, but also daily fallout data from prefectures and cumulated deposition data over a region lying approximately 150 km around the nuclear power plant, we can use a few thousands of data in the inverse modeling algorithm to reconstruct the cesium-137 source term. As expected, the different methods yield closer results as the number of data increases. The updated cesium-137 releases are estimated to be in the range 12-19 PBq, with a std. of 15-25%, depending on the methods and the data sets used in the inversion.