最小极化度滤波

The least polarization filtering

中国石油化工股份有限公司油田事业部 北京100029

SINOPKC Oilfield Department, Beijing 100029, China

地震信号和随机噪声具有不同的极化特性,可以利用极化分析的方法压制地震记录上的随机噪声。以Samson（1973）提出的谱矩阵极化分析方法为基础,结合复地震道技术给出最小极化度滤波方法。假设地震记录由有效信号和不相干随机噪声组成,由不相干随机噪声的复地震道构造协方差矩阵,将此协方差矩阵作为谱矩阵,根据不相干随机噪声的极化度最小这一准则,推导出最小极化度滤波方法的算子。该方法既可应用于三分量地震记录,也可用于单分量的地震记录。

Due to the fact that the polarization characteristics of seismic signals and random noises are different, polarization analysis can he used to attenuate random noises on seismic records. Based on polarization analysis via spectrum matrix proposed by Samson(1973) and complex trace analysis, this paper put forward a method called the least polarization filtering. Assumed that seismic records are composed of signals and incoherent noises, we construct a covariance matrix for complex seismic traces of incoherent noises. The resulting matrix was viewed as the spectrum matrix. Operator of the least polarization filtering is derived according to the rule that degree of polarization of incoherent noises is the least. The method can be applied to three component seismic records as well as single component records.