Combined weighted inversion of electrical resistivity data arising from different array types

In this work we examine the case of combined weighted inversion of resistivity data arising from different array types. In general, there is no single optimum array which can always give valid and useful results, independent of the target characteristics. Moreover, the geoelectrical models produced by the inversion of different arrays over the same structure can be different. In order to use all available information and produce a potentially more reliable geoelectrical model of the earth, an algorithm for the 2-D combined inversion of the most commonly used arrays (dipole-dipole, pole-dipole, Wenner-Schlumberger, Wenner) is examined. The use of this algorithm on several data sets showed that some arrays dominate over others. To overcome this problem an extra weighting factor is introduced to calibrate the participation of the data of each array into the inversion procedure. To determine the value of this factor we have used the Jacobian matrices which are, computed for the data set of each array. The 2-D combined weighted inversion algorithm is tested with various synthetic data sets. Furthermore, a real field data application is presented. The results indicate that the algorithm provides an inverted model of all available data which is not necessarily the optimum but more effective when compared to just jointly inverting all data sets without weighting. The combined weighted inversion algorithm is proved to be a useful tool for data interpretation, especially when the area of study has complicate structures for which there is no a priori information and each array type can not solely produce a reliable geoelectrical model of the subsurface. (c) 2006 Elsevier B.V. All rights reserved.