The study of least-squares migration method based on split-step DSR

Abstract

In order to adapt for the exploration and development of complex structures and lithologic hydrocarbon reservoirs, least-squares migration method applying inversion is proposed. In this paper, based on the basic theory of least- squares migration (LSM), we deduced the forward propagator and its adjoint of back propagator, and then LSM based on split-step DSR was implemented. We tested this algorithm via four layers and SEG/EAGE Marmousi velocity model. The image results demonstrate that LSM has not only a high image resolution for the complex media, in addition, there will be a better image quality than conventional migration method after several iterations of conjugate gradient method for some local regions in the follow aspects: 1) the image resolution is highly improved; 2) A better imaging energy compensation for the deeper part of the model.