Joint migration inversion continuous equations and discretized solution via multiparameter Gauss-Newton method

摘要

Joint Migration Inversion (JMI) is an independent approach to solve the seismic inverse problem, based on decoupled imaging and inversion operators. Here, we review the background equations in their continuous form. We then proceed to test the JMI methodology using the multiparameter Gauss-Newton method to estimate simultaneously image and slowness updates, and we compare the results to those of the conventionally used steepest-descent method. Our numerical results show that the Gauss-Newton method can provide velocity models with improved resolution, albeit at a higher cost. These results demonstrate that the JMI implementation under the assumptions discussed here can provide a good depth migrated image and a satisfying initial velocity model for a subsequent Full Waveform Inversion.