Common-image gathers in the angle domain from reverse time migration in 2D isotropic acoustic, elastic and VTI media

Wavefield polarization is used to decompose wavefields and to extract common-image gathers in the incident angle domain (ADCIGs). For the first time, a theory of vector wavefield decomposition in isotropic and VTI media is derived in the wavenumber domain from both mathematical and physical properties; and implemented with examples.;Angle-domain common-image gathers are obtained by reordering the data in common-image gathers by local incident angle. Incident angles can be calculated as the difference between the propagation direction of the incident wavefield and the normal to the reflector. The reflectors are defined by stacking over common-source reverse-time migration (RTM) images produced by prestack depth migration. The normals to the reflectors are estimated by instantaneous wavenumber. For calculating propagation angles, there are different methods to be used in isotropic and anisotropic, 2D and 3D media.;In isotropic media, the propagation angles are calculated as the elastic P-wave polarization direction. In elastic media, the P- and S-wave vector decomposition is required, but is too expensive to practical. We use a vector separation method to get the P-propagation angle as well as separating the P- and S-wave in isotropic elastic media. In isotropic acoustic media, the propagation angles are calculated by the wavefield gradient directions.;In transversely isotropic media with a vertical symmetry axis (VTI), the local incident angles are local phase propagation angles, which are the difference between the phase angle and the normal to the reflectors. The phase angle can be expressed as the local polarization direction, so in VTI media, qP-wave vector decomposition is applied. We use a modified image-condition to reduce the cost of wavefield vector decomposition in the wavenumber domain. So, for the first time, we implement the RTM in elastic anisotropic media and obtain ADCIGs from VTI RTM images. The potential application is to anisotropic velocity estimation.