Sensitivity Analysis and Uncertainty Propagation Based on Surrogate Model for an Atmospheric Dispersion Code

The CEA has developed the CERES-MITHRA (C-M) application to model the radionuclide atmospheric dispersion in order to evaluate the consequences on human health of radionuclide releases in the environment. This application is used either for crisis management or to perform assessment calculations for regulatory safety documents relative to nuclear facilities. C-M code is a time consuming complex spatio-temporal model based on a Gaussian puff model and depends on many uncertain variables related to radionuclide, release or weather conditions. These variables are of different kind: scalar, functional and qualitative. Given the uncertain parameters, C-M code provides spatial maps of radionuclide concentration for various moments defined by the scenario. The objective is to assess how theses uncertainties can affect the C-M forecasts. In order to achieve this objective, a global sensitivity analysis of C-M code and an uncertainty propagation are performed. The sensitivity analysis is used to identify the most influential uncertain parameters, the noninfluential ones and the interaction between parameters. We consider here the global sensitivity analysis which often calls for the estimation of variance-based importance measures, called Sobol’ indices.