Image-Domain Least-Squares Reverse Time Migration Based on an Adjoint-Consistent Inverse-Scattering Imaging Condition

Abstract

Image-domain least-squares reverse time migration (ID-LSRTM) improves the resolution, illumination balance, and amplitude fidelity of reverse time migration (RTM) images by approximating the Hessian matrix with point-spread functions (PSFs) in the image domain. However, the conventional zero-lag cross correlation imaging condition (CCIC) often generates severe low-wavenumber artifacts in the presence of strong velocity contrasts. Such undesirable low-wavenumber artifacts can greatly degrade the inversion of high-wavenumber reflectivity via a linear migration-deconvolution inverse problem. To address this problem, we propose an ID-LSRTM method based on the inverse-scattering imaging condition (ISIC). By reformulating the ID-LSRTM framework with an adjoint-consistent ISIC operator pair, we ensure that both migrated images and PSFs are constructed in a numerically adjoint-consistent manner. Numerical examples demonstrate that the proposed method suppresses low-wavenumber artifacts, improves the conditioning of the Hessian matrix, and yields cleaner and better-resolved reflectivity images compared to conventional CCIC-based ID-LSRTM. Finally, a field data application shows the feasibility of the proposed method for seismic field application.