Séance Séminaire

NEMESIS Seminar

Efficient Simulation of Electrical Resistivity Tomography

This work focuses on the optimization of forward simulations for electrical resistivity tomography in fractured media. The main objective is to reduce the computational cost of these simulations in the context of inverse problem solving, in particular when using Monte Carlo methods, which require a large number of forward model evaluations. More specifically, the goal is to reduce the time spent solving the linear systems arising from a finite volume discretization of the direct problem. To this end, we investigate a static condensation strategy, in which some unknowns are eliminated in order to obtain a reduced system. This approach aims at taking advantage of the structure of the problem, especially the fact that different sources mainly affect the boundary unknowns in different ways.