Angle domain common image gathers from reverse time migration by combining the Poynting vector with directional decomposition

Abstract

ABSTRACTAngle domain common image gathers are crucial for seismic applications such as the migration velocity analysis and amplitude variation with angle analysis. Angle domain common image gathers are the gather of reflection coefficients when seismic waves incident at different angles on the same image point. The curvature of gathers can be used to update the migration velocity model. The amplitude variation of gathers characterizes different rock properties. Estimating the wavefield propagation direction is the most important step in the process of generating angle domain common image gathers. The wavefront propagation direction is perpendicular to the wavefront. Accurately estimating the wavefield propagation direction is crucial for calculating the reflection angle. Existing methods cannot provide the propagation direction with the high angular resolution when the subsurface structure is complex, meaning overlapped wave events exist. We propose a new method to overcome this problem. The proposed method separates overlapped waves with a directional filter bank after the up‐going and down‐going wavefield decomposition. Then, we apply the Poynting vector method to decomposed wavefield slices to determine the propagation directions of corresponding wavefield components. The key feature of the proposed method is that it does not require spatial local windows, and it is computationally efficient. Numerical tests demonstrate that our proposed method can accurately obtain the propagation direction of seismic waves in a wavefield including overlapped waves. The angle domain common image gathers generated using our proposed method have improved image quality with more accurate amplitudes and less noise compared with the angle domain common image gathers produced by the conventional Poynting vector method. It means our proposed method provides high‐quality prestack angle gathers for seismic attribute analysis.