Elastic FWI for 3D and 4D subsalt imaging

Abstract Full-waveform inversion (FWI) has found great success in different geologic settings and has become a must-have tool for velocity model building (VMB), particularly in salt environments where geology and velocity are often highly complex. While still acoustic, FWI has already significantly improved salt models and marked a step change in subsalt imaging compared to conventional VMB workflows driven by manual salt interpretation. Furthermore, the introduction of an innovative imaging method — FWI imaging — has delivered another step change in subsalt imaging, providing subsalt images of greatly improved signal-to-noise ratio (S/N) and better-balanced amplitudes compared to conventional migration methods. Most recently, the evolution of FWI from acoustic (AFWI) to elastic (EFWI) has represented another key breakthrough, yielding velocity models and FWI images with better-defined details and higher S/N, especially in salt provinces where impedance contrasts are large and elastic effects are most pronounced. Our elastic time-lag FWI (E-TLFWI), which integrates an elastic modeling engine with a time-lag cost function, has demonstrated the ability to reduce salt halos, better delineate structures around salt bodies, and enhance subsalt S/N in both velocity models and FWI images. We illustrate these improvements using two ocean-bottom node (OBN) data sets and one streamer data set. Given the proven advantages of EFWI imaging in 3D, it is natural to extend its application from 3D to 4D, particularly for subsalt 4D imaging, which often suffers from weak and imbalanced illumination and poor S/N with conventional 4D processing. The first application of E-TLFWI imaging on 4D data sets over the Atlantis Field in the Gulf of Mexico reveals clearer subsalt 4D signals and more accurate 4D responses that were not previously observed and are confirmed by both well data and production history.