Geophysical joint inversion based on mixed structural and rock-physics coupling constraints

Joint inversion algorithms continue to receive a lot of interest in the application of geophysical interpretation due to the reduc-tion of uncertainty of the inverse solutions. Over the past de-cade, several coupling constraints, structural or rock physics, have been introduced to the joint inversion of multiple geo-physical data sets. However, a single coupling constraint may not provide the optimum solution for all geologic scenarios. To improve the application of the multiple geophysical data sets, in which some rock-physics information is available, we have developed a new joint inversion algorithm based on a mixed structural and rock-physics coupling constraints. The presented algorithm incorporates the cosine dot product gradient function, a structural coupling constraint, with a guided fuzzy c-means clustering strategy, which is a rock-physics coupling constraint. The former imposes the structural similarity among different model parameters, whereas the latter improves the linear or non-linear correlation among model parameters. We use the exact structural resemblance algorithm to numerically solve the mixed coupling constraint joint inversion objective function. The developed algorithm is applied on 2D synthetic and field data examples. Our results find a significant improvement in the res-olution of the reconstructed models, in comparison with tradi-tional single coupling algorithms, even when the rock-physics information is not complete. This algorithm is more conducive to obtaining clear geologic unit boundaries and provides a favorable basis for delineating metallogenic target areas.