Robust Scheme For Ert Inverse Modeling

Improvements in the automated interpretation of cross-borehole electrical resistivity measurements are presented<br>comparing results from both simulations and recent field surveys. Improvements were achieved by means of a<br>statistically robust inversion scheme and include faster convergence, availability of a stopping criterion, and<br>decreased dependence on data quality. The theoretical basis for the new method is explained concisely to concentrate<br>on its main features and capabilities that arise from the analysis of the results. Particularly, the effect of standard<br>noise and systematic errors present in the data on the reconstructed resistivity images are discussed. The aim is to<br>provide a simple guide for the correct choice of the parameters that characterize an inversion, as for a correct<br>understanding of the images obtained.