Incorporation of topography into two-dimensional dc resistivity and CSAMT joint inversion

Abstract

One approach to increasing the reliability of geoelectric data interpretation involves the application of multiple techniques and processing the data by using a joint inversion scheme. However, most of the current joint inversion techniques are based on one-dimensional models with a flat ground surface. Topographic effects are addressed with a correction factor. Therefore, good results can only be obtained for simple geologic structures. To avoid the influence of a possible incomplete topographic correction, the topography should be incorporated in the inversion model before being undertaken. Here, a 2-D joint inversion algorithm for direct current (dc) resistivity and controlled source audio-frequency magnetotellurics (CSAMT) data in which finite source effects are not considered is developed. This algorithm is able to invert sounding data collected on an irregular earth surface without presuming a flat earth surface. Dc resistivity and CSAMT modeling are based on the finite-element method; and, the iterative joint inversion scheme is derived from the second-order Marquardt damped least-squares method. The algorithm has been tested on both synthetic and field dc resistivity and CSAMT data with explicit incorporation of topography in the joint inversion. Both synthetic and field studies indicate that the technique is computationally efficient in providing an improved geologic interpretation for complex subsurface structures.