We demonstrated the effectiveness of weighted stacking for seismic AVO data containing class II polarity-reversal anomalies using a hybrid framework of AB semblance based NMO velocity analysis and local-similarity-weighted stacking with the near-offset trace as the reference trace. In the synthetic data examples, the SNR in the improved CMP stacked trace is greatly enhanced compared with the conventional velocity analysis and stacking methods. At the same time, the locations of reflectors for the polarity-reversal data in the improved CMP stacked traces are clearly visible that appear as thin-bed tuning in the conventional stacking methods. In the field CMP gather examples, deep reflectors of the good pre-processed CMP gather are captured in the proposed approach that has the highest SNR. For low-SNR prestack marine data, a demultiple technique is required before stacking. The ultimate advantages of the proposed method in event continuity and artifact attenuation are demonstrated in the comparison of field stacked sections. Both the synthetic and field examples illustrate the feasibility and advantages of my proposed approach with both high- and low-SNR data. A beneficial probe has been done for structural imaging by weighted stacking using AVO data. Future research would improve the AVO information for better lithology and fluid prediction, e.g., fluid-factor estimation and gas detection through weighted stacking.
Weighted stacking of seismic AVO data
using hybrid AB semblance and local similarity