基于稀疏反演的OBS数据多次波压制方法

Estimation of primaries and multiples by sparse inversion for OBS data with integration of streamer data

自由表面多次波压制是海底地震仪(Ocean Bottom Seismometer,OBS)数据处理和成像中的难点,OBS数据多次波能量强,周期长,严重影响深层一次反射波的处理和成像.不同于常规拖缆观测系统,OBS数据站点一般相隔较远,仅仅利用检波点稀疏的波场信息难以压制OBS数据中的自由表面多次波.本文采用拖缆数据与OBS数据联合,利用稀疏反演估计(Estimation of Primaries and Multiples by Sparse Inversion,EPSI)方法,研究了OBS数据自由表面多次波压制理论,分析了OBS多次波产生的机理,详细推导了拖缆数据与OBS数据联合预测OBS多次波的EPSI方法基本原理.通过利用拖缆数据的信息,实现了OBS检波点稀疏数据多次波的压制问题.EPSI方法通过稀疏反演直接估计一次反射波,避免了SRME(Surface Related Multiple Elimination)方法中自适应相减对有效信号的损害,保真了一次反射有效信号,理论模拟OBS数据验证了方法的有效性.

Free surface related multiple elimination is very difficult for ocean bottom seismometer （OBS） data processing and imaging. The multiples of OBS data have the characteristics of strong energy and long period, which seriously impact the processing and imaging of primaries. The interval of OBS stations usually is large, which is different from conventional streamer acquisition. Therefore, the conventional wave-equation-based method of free surface related multiple elimination cannot deal with the sparse OBS data. In this paper, we introduce estimation of primaries and multiples by sparse inversion （EPSI） to OBS multiples elimination with integration of streamer data. We first analyse the mechanism of multiples, and then give the methodology of OBS multiples elimination using EPSI with integration of streamer data. By using the information of streamer data, we achieve the multiples elimination of sparse OBS data. EPSI directly estimates primaries using sparse inversion, and avoids adaptive subtraction in surfacerelated multiple elimination （SRME） method. Therefore, EPSI can preserve the primaries. Synthetic data example demonstrates that the proposed method can effectively primaries of OBS data. effective estimate