A new elastic least-squares reverse-time migration method based on the new gradient equations

Compared with single-component seismic data, multicomponent seismic data contain more P- and S-wave information. Making full use of multicomponent seismic data can improve the accuracy of seismic exploration. Elastic reverse-time migration (ERTM) is the most advanced migration technology for imaging multicomponent seismic data from complex subsurface structures. However, most conventional ERTM methods often use the adjoint operator of forward operator for approximation to the inverse operator. When the multicomponent seismic data suffer from a finite recording aperture, limited bandwidth, and imperfect illumination, the image quality of conventional ERTM is greatly reduced. In this study, we propose an elastic least-squares reverse-time migration (ELSRTM) scheme to improve the image quality of ERTM through multiple iterations. We first review the ERTM method; then, we derive the Born modeling equations, adjoint wave equations, and gradient equations of P- and S-wave images of ELSRTM. The new gradient equations, which use the time derivative of stress to replace the spatial derivative of particle velocity for improving the accuracy of gradients near the boundary, are also proposed. We compare the performance of ERTM with ELSRTM via synthetic experiments in numerical examples. Synthetic examples reveal that ELSRTM can generate high-quality images with higher resolution, fewer artifacts, and more balanced amplitude than ERTM.