Joint Inversion of Direct Current and Electromagnetic Soundings

Abstract

Authors: Henry Ekene Ohaegbuchu, Boniface Ikechukwu Ijeh and  Marry Ihechiluru. Ojiaku

It was necessary to first build equivalent synthetic datasets to concurrently invert direct current (DC) and frequency-domain electromagnetic (FDEM) soundings. Declaring the initial number of subsurface layers, the spacing between the AB/2 and MN/2 electrodes, the range of acceptable frequencies, and the coil spacing employed to achieve this. Therefore, using these characteristics, we were able to create DC resistivity and FDEM data that resemble the actual field data that was previously obtained by traditional geophysical surveys. The DC and FDEM datasets were each given 3% and 1% of Gaussian white noise, with regularization strengths of 500 and 300, respectively. Across a homogeneous half-space with three layers and a thickness of 15 m, this was defined. We started a model transformation after creating the synthetic data, which turned the data into logarithms with upper and lower bounds. The DC and FDEM datasets were then independently inverted for comparison with their joint inversion. A combined forward operator was subsequently developed that takes into account the unique characteristics of the various geophysical datasets. By integrating the datasets, transformations, and related errors, we were able to invert the DC and FDEM jointly using the joint forward operator. When the combined inversion's results were shown side by side with the results from the separate inversions, it was found that the joint inversion offered a more accurate picture of the subsurface (model) as the computational errors were much less than that associated with the separate inversions of the individual geophysical datasets.