利用多次波和二维匹配滤波的近偏移距数据重构方法

在地震资料采集时,受地表条件、震源等因素的影响,几乎无法采集到近偏移距数据,不利于后续数据的处理。针对这一问题,研究了一种利用多次波和二维匹配滤波的近偏移距数据重构方法。首先给出了由表层多次波构建准一次波的互相关方法;然后在时间-空间滑动窗口内利用准一次波计算二维匹配滤波器,并用该滤波器重构近偏移距数据;最后使用均方根振幅校正方法校正重构数据的振幅。整个数据重构的过程在频率域进行,避免了基于维纳滤波方法求取滤波器的大量计算;同时使用快速傅里叶变换算法,提高了计算效率。与一维匹配滤波的近偏移距重构方法相比,采用二维匹配滤波方法的重构数据在空间上连续性更好。通过对Sigsbee 2B数据的处理及实际资料的测试,验证了该方法的正确性和有效性。

Reconstruction Method of near Offset Data Using Multiples and 2D Matched Filters

In seismic data acquisition,it is always difficult to collect near offset data due to the influences of surface conditions,source and so on,affecting the follow-up data processing.We proposed a reconstruction method of near offset using multiples and two dimensional matched filters to solve the problem.First,we constructed pseudo primaries from surface-related multiples with cross-correlation method.Then we computed matched filters from the pseudo primaries and reconstruct near offset using the filters in time and space window.Finally,root mean square amplitude correction method was used to correct the amplitude of reconstruction data.The reconstruction process was designed in frequency domain to avoid large amount of calculation based on Wiener filtering method to get the matched filters,the computational efficiency was improved by using fast Fourier transform algorithm.The spatial continuity of the data reconstructed by two dimensional matched filters is better than that of one dimensional matched filters.We proved the correctness and effectiveness of the method by testing on Sigsbee 2B data and real seismic data.