Land Seismic Data Multi-parameter Waveform Inversion

Acoustic full waveform inversion of low-frequency, long-offset, and wide-azimuth data allows updating the long-to-intermediate wavelength components of the velocity model. For migration of reflection data, anisotropy needs to be accounted for in the inversion, since full waveform inversion mainly retrieves the horizontal velocity. The simultaneous inversion for the anelliptic parameter and the normalized NMO velocity is attractive because, under the vertical transverse isotropy assumption, this parameterization is sufficient to model the kinematics of the acoustic waves and is computationally more efficient than a sequential inversion. We present the result of a simultaneous multi-parameter inversion of a real data set from North Oman which shows complexity due to a major fault zone. The velocity model obtained after the inversion properly retrieves the main geological features. The quality of the gathers is better to the one obtained from the initial model and a PSTM model. Because FWI requires a simpler data pre-processing than standard pre-stack depth-domain reflection-based tomography, the approach presented could potentially shorten the cycle time of an imaging project when low-frequency, long-offset, and wide-azimuth land data are available. Indeed, acoustic anisotropic FWI may provide a good model for subsequent velocity update (notably deeper update) using conventional reflection-based velocity inversion.