Imaging Steeply-Dipping Fault Zones Using Elastic Reverse-Time Migration with a Combined Wavefield-Separation and Poynting-Vector Imaging Condition

Abstract

Elastic reverse-time migration needs to properly handle the shear-wave polarization during imaging using compressional-to-shear or shear-to-compressional converted waves. Polarity distribution analysis of Poynting vectors is a computationally efficient method for polarization corrections. A limitation of the Poynting vector imaging condition is that it assumes only one dominant direction of wave propagation at each spatial point for any given time and thus is not very accurate in complex regions with complicated wavefields. We develop a computationally efficient imaging condition for elastic reverse-time migration to directly image steeply-dipping fault zones. We first separate the forward-propagating source wavefield and the backward-propagating receiver wavefield into left- and right-going, or down-going and up-going waves, and then apply the Poynting-vector imaging condition to the separated wavefields. After wavefield separation, the inaccuracy of the Poynting-vector imaging condition in complex regions is alleviated. We build a geophysical model using geologic features found at the Soda Lake geothermal field and generate synthetic multi-component seismic reflection data. The model contains several steeply-dipping fault zones. We validate the improved imaging capability of our new imaging condition for elastic reverse-time migration using the synthetic data. Our numerical results demonstrate that our new elastic reverse-time migration with the combined wavefield-separation and Poynting-vector imaging condition produces high-resolution images of steeply-dipping fault zones, while its computational cost is approximately one order of magnitude lower than that with the space-lag imaging condition for the 2D case.