Investigating the performance of combined resistivity model using different electrode arrays configuration

The use of a single geophysical method is prone to more ambiguity and loss of vital geological information at geophysical interpretation stage. To circumvent this problem, a statistical approach and image processing technique were used to combine three electrode arrays (dipole-dipole, pole-dipole, and Werner-Schlumberger). The combined models were tested using both synthetic and field models. For synthetic models, the mean absolute error (MAE), mean absolute percentage error (MAPE) and mean resistivity values were used as criteria to assess the combined models. For field data, the three protocols were combined and the models were compared with borehole data. Considering the overall performance, maximum value model has a low error and correlated well with the borehole data than the other combined models analyzed although the performance of these models depends on resistivity in addition to the nature of geologic structure under investigation. The correlation between resistivity and lithological changes with depth (borehole data) used in this research provided a more quantitative comparison approach than the ones available in the literature. The image processing technique served as an aid to make post inversion analysis such as combining data sets and quantitative comparison among/between the models.