奇异值分解在垂直地震剖面中的应用

APPLICATION OF SINGULAR VALUE DECOMPOSITION (SVD) TO VSP

1. 地矿部北京计算中心,北京100083;2. 中国地质大学,北京100083

1. Beijing Computer Centre, MGMR, Beijing 100083;2. China University of Geoseiences, Beijing 100083

垂直地震剖面(VSP)资料处理中的关键一环是上、下行波的分离。本文介绍了奇异值分解(SVD)法应用于地震图象重建的方法原理,说明了用SVD法分离VSP资料中的上、下行波是一种较新的可行方法。对理论模型数据用低通、带通和高通SVD滤波得到了下行波、上行波和噪声剖面,证明了本方法的有效性;对不等间隔采样和下行初至精度不高的VSP资料做了模型试验,结果表明并不影响用本方法分离上、下行波的效果。文中还给出了某油田实际零偏和非零偏VSP资料的应用实例。

The key to process vertical seismic profile (VSP) data is to separate upgoing and dovvngiong waves. This paper describes the singular value decomposition (SVD) technique for seismic image reconstruction, a new and feasible way to separate upgoing and downgoing waves in VSP data. Separated upgoing waves?downgoing waves, and noise sections have resulted from a set of theoretical model data respectively via the low-pass, band-pass,and high-pass SVD filtering, which proved the validity of the SVD method. Moreover, the fact that the separation of upgoing and downgoing waves is independent from irregularly sampled VSP data and inaccurate downgoing first arrivals was confirmed via a series of model experiments. Some applied examples related to zero and non-zero offset VSP data from a certain oilfield are also presented in the paper.