低信噪比地震资料的子波处理方法

WAVELET PROCESSING OF SEISMIC DATA WITH POOR S/N

叠后子波处理(求反射系数的估计)是波阻抗反演的第一步。传统的子波处理方法对噪声的适应性差,是造成传统波阻抗反演法(或伪速度测井)在低信噪比地区不能取得成功的关键之一。为了解决这一问题,本文采用了“广义线性反演子波提取”方法和“线性反演反褶积”方法。这两种方法联合使用构成一个完整的子波处理过程,其主要优点是稳定性好,适用于在低信噪比地区求取子波并最终求得反射系数。理论和模型试算证明方法是有效和实用的。

The post-stack wavelet processing (estimate of determining reflectivity) is considered the first step to perform the acoustic impedance inversion. The poor adaptivity of conventional way of wavelet processing to noise is one of the key problems, which causes the failure of using the conventional acoustic impedance inversion technique (or pseudo velocity well logging) in the poor S/N area. For resolving this problem, two methods are introduced, one is to extract wavelet using the Generalized Linear Inverse technique, the other is to perform deconvolution using the Linear Inverse. The combination of the 2 techniques makes up an integrated wavelet processing sequence. Its major advantage is that stability can be ensured and it's suitable for determining wavelet in the areas with poor signal-to-noise ratio and reflectivity can be estimated eventually, From theory and model trial-computational results one can see that the recommended apprach is proved efficient and of practical use.