一种改进表层多次波压制方法  

An improved method of surface-related multiple elimination

作　　者：董烈乾 李振春[1] 杨少春[1] 李凤磊 李志娜 

机构地区：[1]中国石油大学(华东)地球科学与技术学院,青岛266580 [2]中国石油集团新疆油田分公司,克拉玛依834000

出　　处：《地球物理学进展》2013年第6期3148-3152,共5页Progress in Geophysics

基　　金：国家科技重大专项(2011ZX05006-002);中国石油大学自主创新科研计划资助项目(11CX04010A)联合资助

摘　　要：由Berkhout等提出的自由表面多次波压制方法(SRME),由于在多次波预测过程中采用了数据的空间褶积,因此对输入的数据有严格的要求.本文设计了一种改进的表层多次波压制方法.该方法对SRME公式进行重新推导,将数据的空间褶积转化为数据相关,提高了对近偏移距缺失地震数据的适应性,并且能够将SRME方法中所需的地表因子隐含在迭代求取多次波模型过程中.在匹配相减阶段,本文引入了一种非平稳匹配滤波器方法,该方法在很大程度上克服了L2模匹配方法一次波和多次波正交的要求.模型数据试算说明了该方法在压制多次波的同时能够很好地保持一次波的运动学和动力学特征.Surface-related multiple elimination（SRME） proposed by Berkhout adopts data spatial convolution in the stage of multiple prediction. Therefore, the method needs rigorously regular input data. In this paper, an improved method by the derivation of SRME is proposed, which can transform the data spatial convolution to data correlation. It eliminates the need to synthesize near-offset traces, at the same time, also eliminates the need for an explicit surface operator, which are both required by conventional SRME. In the stage of subtraction, a non-stationary matching filter is designed, which overcomes the limit of orthometric requirement of primary and multiple of L2-norm matching. Applying this method to synthetical data sets indicate that it can suppress multiple effectively, meanwhile, also greatly maintain the dynamical and kinematical characteristics of primary.

关 键 词：多次波压制 空间褶积 数据相关 地表因子 非平稳匹配滤波器 

分 类 号：P631[天文地球—地质矿产勘探]