Pre-stack gather optimization technology based on an improved bidimensional empirical mode decomposition method

Abstract

This study proposes a pre-stack gather optimization technique based on the improved BEMD (bidimensional empirical mode decomposition) method with high stability and adaptability. An original gather signal was decomposed into different components with varying characteristics by the BEMD method, which combines morphological extremum point extraction and the bicubic spline interpolation method to obtain different-scale component signals. Then, wavelet threshold orthogonal filter was used to process the components, which correspond to noise, mostly in small-scale domain. Finally, correlation weighting reconstruction was performed on each component. The weight coefficient is determined according to the relationship between each component of the amplitude-preserving superposition gather and the original gather-preserving superposition gather. The testes on the data from Northeast Sichuan confirm that the method can optimize the pre-stack gather steadily and efficiently. The comparison of the AVO analysis results before and after optimization shows that the method is effective and feasible, which lays a good foundation for a prediction method based on pre-stack data.