Elastic Full Waveform Inversion Tests for Shallow Targets Reconstruction from Surface Waves Analysis Based Initial Models

Seismic data from land acquisitions are dominated by highly energetic surface waves (SW), showing complex behaviour when interacting with local structures. SW analysis is, therefore, an important tool for accurate nearsurface imaging, even if most of the conventional SW analysis techniques are limited by the lateral invariance assumption. This study attempts to enhance the resolution of shallow targets reconstruction by integrating SW dispersion curves (DC) analysis techniques with a spectral element based elastic full-waveform inversion (FWI) workflow. Multi-parameter elastic FWI tests have been conducted over a synthetic dataset related to a benchmark model that mirrors the characteristics of a real test site. The initial S-and P-wave velocity (Vs and Vp) models have been retrieved by DC analysis with increasing detail resolution. The FWI results showed a better model reconstruction when starting from a more detailed initial model, obtained using the full-DC analysis. Further, we improved the initial Vs model (reconstructed from SW information) by a preliminary mono-parameter FWI step. Starting from this improved initial configuration, we obtained more accurate results when performing the multiparameter FWI.