多子波分解在地震反射系数研究中的应用

Application of Multiwavelet Decomposition in Seismic Reflection Coefficien

反射系数反演是高分辨率地震解释的常用技术之一,广泛使用的是基于奇偶分解的谱反演技术。本文采用基于匹配追踪(Matching Pursuit,MP)的多子波分解技术,将地震信号分解成一系列包含时间、频率、振幅的子波集,然后进行井震能量标定,选定研究层段内最强能量处,将分解的振幅比例到对应的反射系数。楔形正演模型表明,该思路可以得到准确的反射系数,实际资料表明对反射系数求取相对阻抗,能准确识别气层边界,拓展了多子波分解技术的应用领域。

Seismic reflection coefficient inversion is one of the commonly used techniques for high-resolution seismic interpretation. The widely used technique is based on spectral inversion. In this paper, a multi-wavelet decomposition technique based on Matching Pursuit (MP) is used to decompose the seismic trace into a series of wavelet sets including time, frequency and amplitude. In the decomposition process, adjacent trace are used as lateral constraints , then use seismic well tie, select the strongest reflection coefficient and seismic amplitude, divide the seismic amplitude by the reflection coefficient. The wedge model shows that it can obtain accurate reflection coefficients by this method. The field data shows that the boundary of the gas layer can be accurately identified by the relative impedance, so the application fields of the multi-wavelet decomposition technology are expanded.