Multi-parametric intensive stochastic simulations for hydrogeology on a computational grid

Numerical modelling is an important key for the management and remediation of groundwater resources. Numerical simulations must be performed on domains of a large size, at a fine resolution to take into account the scale of geological heterogeneities. Numerical models are based on probabilistic data and rely on Uncertainty Quantification methods (UQ). In this stochastic framework, non intrusive methods require to run multiple simulations. Also, each simulation is governed by multiple parameters and a complete study requires to carry out analysis for more than 50 sets of parameters. We have identified three levels of distributed and parallel computing: subdomain decomposition in one simulation, multiple simulations for UQ methods, multiparametric studies. Our objective is to use the computing and memory resources of computational grids to deploy these multiple large-scale simulations. We discuss our implementation of these three levels, using an object-oriented approach. We present some preliminary results, with a strategy to choose between the first and second level.