Full-azimuth angle domain reverse time migration

With the advent of wide-azimuth and high-density acquisition technologies, seismic data has become richer and more informative, necessitating advanced seismic imaging techniques. This paper presents a full-azimuth reverse time migration imaging approach constrained by wavefield gradients, grounded in wave theory. The method initiates with an iterative determination of the propagation direction vector, leveraging gradient information from both amplitude and phase to bolster the algorithm’s stability and precision. An angle filter is subsequently formulated within the angle domain imaging condition to diminish the impact of large-angle energy interference. Following this, full-azimuth angle domain common imaging gathers are derived through integration across azimuth and reflection angles. The CPU-GPU collaborative parallel algorithm and the encoding-decoding-based data compression technology are also introduced to tackle the challenges of high computational load and limited storage capacity. Ultimately, numerical experiments validate the efficacy of the proposed algorithm and its suitability for imaging complex geological structures.