Diffraction Extraction and Least-Squares Reverse Time Migration Imaging for the Fault-Karst Structure With Adaptive Sampling Strategy

Abstract

The ultra-deep fault-karst structure discovered in the Tarim Basin in Western China is a fractured-vuggy carbonate reservoir with great potential for development. The diffraction generated by fractures, small-scale caves and vugs is often used for reservoir identification and seismic interpretation. Since the diffraction is much weaker than the reflection, it is difficult to separate the diffraction from the full wavefield. We use a plane-wave destruction (PWD) filter to extract the diffraction from the full data. In order to obtain a high-accuracy and amplitude-preserving imaging profile, we use the least-squares reverse time migration (LSRTM) method to image the separated diffraction. The large amount of calculation is the most challenging problem of the LSRTM algorithm. In view of this, we develop an adaptive sampling strategy to improve computing efficiency and reduce memory requirement. We use a fault model, a vugs-fractures model, and a fault-karst model to demonstrate the effectiveness and practicability of the proposed method. The numerical examples show that the proposed method can enhance the imaging resolution of the fault-karst structure and save computing cost without losing accuracy. In addition, a test on field data processing demonstrates the advantages of our algorithm.