背景噪声互相关去噪的滑动窗加权SVD滤波方法

Denoising noise cross-correlations via moving time-window weighted-SVD method

假设噪声互相关函数中的面波为相干信号和椭圆极化的,而噪声信号一般是非相干的,本文采用滑动窗加权SVD方法对噪声互相关函数去噪,通过对一些列时间窗内的波形进行SVD分解,将前两个特征图像叠加重构波形,叠加过程中引入与极化强度有关的权重因子达到滤波目的.该方法显著的增加了噪声互相关函数的信噪比,面波前驱信号也得到有效压制,有效增加了远距离台站对射线数目,射线覆盖和交叉程度的增加一定程度上提高了成像的分辨率.

In this paper, we present Moving Time-Window Weighted-SVD(MTWS) method based algorithm for polarization denoising of three component noise cross-correlation functions based on the assumption that surface wave green function is elliptically polarized and coherent. The signal is reconstructed by the sum of the first two eigenimages derived from data SVD, Weighting factors, dependent on the intensity of the polarization, are used to filter the data. A major advantage of this method is that it can greatly improve the signal to noise ratio, in addition, it can suppress the precursor wave and noise. From the results of actual data processing, we can recover more travel time measurements from longer distances and the path coverage. This improvement in path coverage should increase the resolution of surface wave tomography.