基于褶积模型的动校正及其实现方法

Convolution model-based NMO and implementation approach

为了消除基于传统地震动校正过程中出现的动校正拉伸效应,本文重新审视了动校正的物理意义,即通过非零炮检距地震记录求取零炮检距处的自激自收旅行时,以此为出发点从数学上给出了基于褶积模型的动校正(即所谓的真动校正)概念。在此基础上,提出了利用空间—时间(滤波)叠加法实现向真动校正叠加的逼近。理论合成记录的试算和实际资料的处理结果均显示了该方法的可行性和有效性。

In order to remove the NMO stretching effect appeared in traditional seismic NMO processing, the paper reviewed the physical meaning of NMO: computing auto-exiting and auto-receiving seismic records at zero-offset position on non-zero offsets seismic records, and starting from that point, gave convolution model-based NMO conception mathematically (so-called true NMO). On that bases, the paper presented using space-time (filtering) stack approach to realize approaching to true NMO stack. The tests of synthetic records and processing results of practical data showed the feasibility and effectiveness of the approach.