Paper submitted to SEG FWI Beyond Structure Interpretation Workshop, by James Sheng, Guanghui Huang, Terence Krishnasamy, Jaime Ramos-Martínez and Carlos Calderón-Macías (TGS)
Summary
Velocity model building (VMB) and seismic imaging in land environments present additional challenges compared to marine settings. We describe a multiparameter fullwaveform inversion (MP-FWI) approach that jointly estimates velocity and reflectivity, adapted for land scenarios. The scale-separation strategy applied in MP-FWI improves background velocity estimation from reflected energy beyond the penetration depth achievable with refracted and diving waves. We employ the dynamic matching (DM) objective function, which enhances velocity estimation at low frequencies under strong noise conditions. Additionally, iterative reflectivity inversion within FWI improves shallow subsurface illumination, where conventional imaging often performs poorly. The reflectivity kernel in MP-FWI boosts high-wavenumber content, improving coherence and resolution. We illustrate the method using a dataset acquired in the Midland Basin, onshore U.S.