快速叠后偏移反褶积算法及应用

Fast post-stack migration deconvolution algorithm and its application

偏移反褶积方法是一种提高偏移剖面分辨率的有效方法,求取偏移反褶积所需的格林函数基于采集观测系统以及速度模型只随深度变化的假设,但求取偏移反褶积滤波因子的计算量巨大,对于三维数据更是如此。本文提出基于均匀介质假设的叠后偏移格林函数的快速算法,其思路是:对于叠后数据来说,震源点与检波点是重合的,同时,可以认为其观测系统是规则的,再根据偏移格林函数平移不变性的假设,在每个深度上只需计算一个偏移格林函数即可。通过快速偏移反褶积方法,采用常速扫描来设计优化的偏移反褶积算子以处理横向变速情况。针对不同数据的初步实验表明,快速偏移反褶积算子能有效地提高计算效率和偏移剖面的成像质量。

Migration deconvolution method is effective for raising resolution of the migrated sections,the Green Function which was needed by migration deconvolution method is based on the assumption in which the acquisition geometry and velocity model only vary with the depth,estimation of migration deconvolution filtering factor takes a large amount of time,especially for 3d datasets.In order to solve the problem a fast algorithm of post-stack migration Green Function which is based on homogenous medium assumption was proposed in this paper,its main idea is as below:For post-stack datasets,the source and the receiver are in a same location,meanwhile by regarding the geometry as regular and based on the assumption of migration Green Function's shift invariance we only need to calculate a migration Green Function at each depth level.Through fast migration deconvolution method,the constant velocity scanning was used to design optimized deconvolution operator which fits the circumstance of lateral velocity variation.The preliminary experiments for different datasets show that the fast migration deconvolution operator could effectively raise the calculation efficiency and the imaging quality of the migrated sections.