Inversion géostatistique par itérations de la tomographie de résistivité électrique

Abstract Electrical resistivity tomography (ERT) is a geophysical method used to create an image of the subsurface due its sensitivity porosity, water saturation, and pore fluid salinity. This has been widely applied in investigation mineral groundwater resources, as well archaeological, environmental, engineering studies. The prediction properties, such electrical conductivity, from measured ERT data requires solving challenging inversion problem. work proposes iterative geostatistical using stochastic sequential simulation co-simulation model perturbation update techniques. models are generated conditioned target histogram, often retrieved available borehole data, assuming spatial continuity pattern described by variogram model. From models, finite-volume approximation Poisson’s equation compute synthetic data. misfit between predicted observed drives convergence procedure conditions new subsequent iterations. methodology two-dimensional case, set profiles obtained survey carried out southern Portugal. In both application examples, final predict that match ones while reproducing imposed models. results sets compared against commercial deterministic methodology, showing ability proposed small-scale variability assess uncertainty.