Broadband data aided velocity model building and imaging

Abstract

During the past decade, marine seismic acquisition evolved from narrow-azimuth (NAZ) to wide-azimuth (WAZ) geometries toward the latest offering of long-offset full-azimuth (FAZ) surveys. This advance in acquisition design is mainly driven by the demands of improved illumination in complex areas such as Gulf of Mexico (GOM) subsalt environments. Despite these acquisition and processing improvements, marine data still suffers from the ghost effect of free-surface reflections and resulting in constructive and destructive interference between the up- and down-going wavefields. If uncorrected, the ghost significantly reduces the bandwidth and distorts the phase of the recorded data and prevents us from producing a broadband image. Fortunately, the ghost effect is straightforward to simulate and, therefore, to correct, either prior to or during two-way wave equation inversion and imaging. Optimal use of this improved data, two-way wave equation algorithms such as reverse time migration (RTM) and full waveform inversion (FWI) are required. Their high-fidelity nature enables them to honor complex velocity fields with the accuracy required, particularly for migrating long-offset data, and as a result, for producing satisfactory model refinements and images when other methods fail. Finally, even with the best acquisition and imaging algorithms, RTM images can be further improved by post-imaging enhancements using vector image partitions (VIPs). We demonstrate an efficient workflow to maximize the power of broadband acquisition by applying velocity model building with FWI, imaging with RTM compensation for ghost effects in both steps, and outputting partial images for further optimization.