Two-Grid Genetic Algorithm Full Waveform Inversion of Surface Waves: Two Actual Data Examples

We present the application to two field data cases of our two-grid full waveform inversion (FWI) of surface waves for the prediction of 2-D shear-wave velocity models. The inversion employs a genetic algorithm as the optimization tool and a finite difference forward modeling. We aim at estimating models reproducing the long-wavelength structures of the velocity field. Frequency marching up to 30 Hz is included in the inversion. The use of a global optimization method allows us to relax the requirement of local optimization methods of having a “good” starting model from which to launch the inversion. In fact, we show that fair results can be retrieved even in the case of null a-priori information. In the first example, we find a satisfactory matching between the predicted model and the long-wavelength velocity structure exhibited by velocity logs from nearby boreholes. The predicted seismograms match well the observed data and when mismatches occur, the time shifts are within half-periods. In the second example, the inversion well predicts the fundamental mode and partially predicts the 1st higher mode. The fair inversion results indicate that the models derived from our stochastic approach could be proper initial models for a local optimization FWI.