Superior resolution through multiparameter FWI imaging: A new philosophy in seismic processing and imaging

Abstract

Seismic processing and imaging workflows have been refined over many decades to attenuate aspects of the recorded wavefield which would be improperly mapped into the image domain by legacy migration algorithms such as Kirchhoff prestack depth migration. These workflows, which include techniques such as deghosting, designature, demultiple, and regularization, have become increasingly complex and time-consuming due to the sequential fashion in which they must be tested and applied. The single-scattering (primary-only) preprocessed data are then migrated and used in extensive model building workflows, including reflection residual moveout tomography, to refine low-frequency subsurface models. Obtaining optimal results at each stage requires subjective assessment of a wide range of parameter tests. Results can be highly variable, with different decisions resulting in very different outcomes. Such workflows mean that projects may take many months or even years. Full-waveform inversion (FWI) imaging offers an alternative philosophy to this conventional approach. FWI imaging is a least-squares multiscattering algorithm that uses the raw field data (transmitted and reflected arrivals as well as their multiples and ghosts) to determine many different subsurface parameters, including reflectivity. Because this approach uses the full wavefield, the subsurface is sampled more completely during the inversion. Here, we demonstrate the application of a novel multiparameter FWI imaging technique to generate high-resolution amplitude variation with angle reflectivity simultaneously with other model parameters, such as velocity and anisotropy, directly from the raw field data. Given that these results are obtained faster than the conventional workflow with a higher resolution, improved illumination, and reduced noise, we highlight the potential of multiparameter FWI imaging to supersede the conventional workflow.