Reverse Time Migration in Angle Domain Based on Gradient Field of Traveltime

Reverse time migration (RTM) is a standard imaging tool for complicated geologic structures. The imaging gathers using RTM are important for migration velocity analysis (MVA), amplitude variation with angle (AVA) analysis and angle-related inversion, but the RTM in angle domain is still challenging for computational efficiency and stability. The calculation of Poynting vector is one of typical methods for angle gather generating of RTM, which uses spatial derivatives and time derivatives of the wavefield to calculate the propagating direction of subsurface imaging points. The traditional Poynting vector method has low computational amount and low storage memory requirement, but it suffers instability and inaccuracy when the source or receiver wavefield is complex. In this paper, we develop a method of generating angle gathers using RTM based on the gradient field of traveltime. Firstly, the gradient field of traveltime is calculated by dynamic programming method, which can be implemented efficiently before imaging. Then, we use the gradient field of traveltime to calculate the propagating direction of the source instead of using Poynting vectors, and this process can be merged into the imaging step conveniently. Finally, we can combine the source direction and the reflector normal direction to produce the common imaging gathers in angle domain. In order to avoid storing the huge wavefield on the disk and to reduce the I/O volume of data further, the RTM is implemented using source wavefield reconstruction scheme. Comparing to the Poynting vector method, our method using traveltime information instead of wavefield information to obtain the subsurface angles, which is more efficient and stable. The numerical examples of synthetic data demonstrate the effectiveness of the proposed method.