Joint inversion of seismic and gravity data using Gramian constraints and Gauss-Newton optimization

This study has developed and implemented a 3-D joint inversion algorithm for seismic and gravity data. The joint inversion algorithm is based on Gramian constraint, which is used to reduce the non-uniqueness of the inversion problem and improve the model resolution. The objective functional is optimized using the Gauss-Newton algorithm, and various forms of Gramian operators are incorporated into the objective functional to achieve the joint inversion of direct coupling and structural coupling. To validate the algorithm, we compared the slice images obtained from separate inversion and joint inversion of a classical synthetic model. Additionally, we analyzed the physical property parameters and the gradient cross-plots. The results indicate a significant improvement in the residual density model after joint inversion. Furthermore, the cross-plots of velocity and residual density show a strong correlation, thus validating the reliability and effectiveness of the joint inversion algorithm.