Weighted stacking of seismic AVO data using hybrid AB semblance and local similarity

Abstract

The common-midpoint (CMP) stacking technique plays an important role in enhancing the signal-to-noise ratio (SNR) in seismic data processing and imaging. Weighted stacking is often used to improve the performance of conventional equal-weight stacking in further attenuating random noise and handling the amplitude variations in real seismic data. In this study, we propose to use a hybrid framework of combining AB semblance and a local-similarity-weighted stacking scheme. The objective is to achieve an optimal stacking of the CMP gathers with class II amplitude-variation-with-offset (AVO) polarity-reversal anomaly. The selection of high-quality near-offset reference trace is another innovation of this work because of its better preservation of useful energy. Applications to synthetic and field seismic data demonstrate a great improvement using our method to capture the true locations of weak reflections, distinguish thin-bed tuning artifacts, and effectively attenuate random noise.