Two-dimensional weighted stack determination using signal-to-noise ratios and probability statistics

Abstract

To improve the stacking effect in seismic signal processing, we use a weighted stack algorithm that considers the differences in signal-to-noise ratios (SNR) between seismic traces. The statistical properties of the weights in this weighted stack algorithm, such as variance and standard deviation, are the same as those of the SNR in each trace. We propose using a windowing technique to expand the weight in each seismic trace in “time,” resulting in a two-dimensional weight that varies with both “space” and “time.” A simulation test is conducted under the same conditions. The result shows that the improved two-dimensional weight is more effective. Then, we apply the two-dimensional weight to the multiple detection technique and the multiple coverage technique to determine the optimal condition for the use of such a weighted stack; we find that it is most effective when there are differences in the SNRs or regular noise interference in each seismic trace. It is thus clear that our weighted stack algorithm is more suitable for the multiple coverage techniques.