Frequency‐domain elastic reverse‐time migration using wave‐field separation

Abstract

Many researches based on wavefield separating migration techniques have been suggested for processing multi‐component data properly. Those works were carried out in the time domain, however. In this paper, we propose the frequency‐domain migration technique with wavefield separation. In the frequency domain, reverse‐time migration can be implemented efficiently by the zero‐lag convolution between virtual source and back‐propagated wavefield. Our new migration can also be defined as the zero‐lag convolution between the decomposed virtual source and the decomposed back‐propagated wavefield derived by exploiting Helmholtz decomposition. P‐and S‐wave migration images can be obtained separately by defining the virtual source separately. We verify our new migration algorithm on synthetic seismic data generated from the Marmousi‐2 model. We demonstrate that the results given by the new algorithm are more accurate than those of the conventional method without wavefield separation.