Abstract
Full-waveform inversion (FWI) is a high-resolution earth model-building technique based on recorded seismic data. Conventional FWI usually relies on diving and refracted waves to update the low-wavenumber/background components of the model; however, the update based on transmitted energy is often depth limited due to the limited offset range of the acquired data. To extend the FWI updating depth beyond the transmitted energy limits, we must use reflection data. Recently, industry interest has resumed in the potential for automated subsurface model-building, especially in complex geologic settings (e.g., salts), through data-driven minimization such as FWI. The business impact of such an automatic model-building technique would be significant in that it is proposed to improve the efficiency of any model-building exercise involving structural complexity and high uncertainty in seismic image interpretation. An ultimate expectation for the FWI technique is to build or update the salt geometry because these complex bodies have a first-order impact on image quality. We evaluate several examples using FWI for building a subsurface model, including salt boundary and salt velocity delineation, in geologically complex areas in the western Gulf of Mexico. The geology there comprises rugose and deformed shallow salt bodies with intracanopy high dip and close-proximity structures, resulting from regional basinward gliding and associated compressional mechanisms. Given the challenges for model building in such a complex setting, a data-fitting approach such as FWI with access to the full reflectivity record is proposed to provide practical solutions for an effective salt model update. Improving confidence in the seismic image and subsequent geologic understanding remains the core objective.
