Angle‐domain common‐image gathers in reverse‐time migration by combining the Poynting vector with local‐wavefield decomposition

ABSTRACT Angle‐domain common‐image gathers are an important tool in the post‐processing of seismic images and for reservoir characterization. The generation of angle gathers is a very important issue when dealing with angle‐domain images. Efficiency and robustness are the main concerns in the generation of angle gathers. In this paper, we propose two methods for producing angle gathers based on the implementation of a reverse‐time migration. In the hybrid method, we adopt the local‐plane‐wave decomposition method to extract the local plane waves and obtain two possible opposite propagation directions in the time‐wavenumber domain. Then, Poynting vectors are used to determine the correct propagation direction. The hybrid method achieves a satisfactory balance between robustness and computational efficiency. Furthermore, in the improved hybrid method, additional computational acceleration is obtained by separating the overlapping and non‐overlapping wavefield areas. The hybrid method is only applied in these areas with overlapping wave fronts, and the Poynting‐vector‐based method is adopted in the other areas. The location of the overlapping events is determined using the eigenvalues of the structural tensor. Finally, the two‐dimensional synthetic and field examples demonstrate the effectiveness of both methods.