波动方程可变参数偏移方法

Parameter-variable wave equation migration

在波动方程偏移方法中，采用何种计算方法是提高地震资料处理效果的关键，而在计算方法中使用何种差分方程取代偏微分方程则是问题的核心。笔者在分析现有波动方程偏移方法"WEMIG"程序过程中发现，用差分方程近似取代偏微分方程时，不考虑地质构造的特点，一律采用固定不变的权重比例系数，会导致地震反射的连续性变差、反射层断点不清晰、画弧现象严重等缺陷。本文阐述了现有方法中各系数的来龙去脉及其物理含义，并根据地震波的传播规律，在建立差分方程的过程中，通过采用地震道间可变权重系数W和邻近深度、时间点之间的可变权重系数的方法，实现对不同特点的地震资料选用不同的权重系数，以达到改善偏移剖面质量的目的。处理中在该连续而连续性不好的部位可取较小的W（1，2或3）；在断点不清晰的部位，可取较大的W（＞4）；若需减少画派现象时，则应取较大的W值和θ值（θ＞2）。本文所提出的方法，可作为常规处理方法使用，但更适合于地震资料的精细处理。

In wave equation migration, computation method is the key to good data pro-cessing effect l and in a given computation method,the very essential point is whatdifference equation we introduce to replace the partial differential equation.By analysing 'WEMIG'program of existing wave equatinn migrations,we canfind that if we take constant weighting coefficients without considering variable geo-logical structures in replacing the partial differential equation by approximate differ-ence equation,we will have an unsatisfactory seismic section that shows some de-fects such as poor1y continuous seismic horizons,blurred fault points,severe arcs,etc.The article describes both derivations and physical meanings of different coeffi-cients- According to seismic wave propagation law,the intertrace variable weightingcoefficient W and the variable weighting coefficient Obetween adjacent depth andtime are used in difference equation construction,so that different weighting coeffi-cients can be nicely applied to different seismic data,thus improving seismic sec-tion. In data processing,small W(l, 2,or 3)peay be applicable-for the poorly contin-uous seismic horizons which should be nicely continuous, big W(above 4)for theseismic horizons which shoW some blurred fault points,and big W and 6(0>2)forthe part where arcs must be removed desirably.This method can be used desirably in conventional seismic data processing,par-ticularly in detailed processing of seismic data.