一种数据驱动的VSP多次波压制方法

A Data-driven Method of VSP Multiple Suppression Waves

V SP资料用于提供地下结构的高分辨率信息,但其成像质量通常受到多次波的干扰,影响了地下真实反射界面的识别。本文介绍一种数据驱动的V SP多次波压制方法,该方法基于反馈迭代原理,综合利用地面和井中地震数据,资料中的多次波进行预测并基于最小能量准则实现自适应相减;同时,利用V SP资料中直达波,可以预测出下行波场,进而提取上行波记录。简单模型结果表明,该方法可以有效实现V SP资料多次波压制及上下行波场分离,且无需地下模型先验信息。

VSP data can provide high-resolution information about subsurface,but the imaging is often blurring due to multiples.This paper introduces a data-driven multiple attenuation method basing on feed-back model.It combines surface seismic data with VSP data to predict the interfering multiples and then uses minimum-energy criterion for adaptive subtraction.Besides,by using VSP direct wave,the down-going wavefield can be predicted and up-going wavefield can be separated.A simple synthetic model illustrates that this method is feasible for multiples prediction as well as wavefields separation.