A Novel Nonlinear Solver for Geophysical Inversions

We present the application of a preconditioned nonlinear conjugate gradients (PC-NLCG) optimizer to a common-image-point (CIP) seismic tomography. The new solver operates in our simultaneous joint inversion (SJI) framework, automatically being available for any possible single-domain inversion or SJI configuration with other seismic or non-seismic datasets. In all the tests we carried out the new PC-NLCG has always outperformed the standard NLCG. We show an example of performance improvement using data from a CIP tomography case.