确定三维地层速度结构的地震层析成像技术

SEISMIC CT TECHNIQUE FOR DETERMINING THE 3-D SUBSURFACE VELOCITY STRUCTURE

观测数据与地层参数之间呈线积分关系的反演称作地球物理层析成像。本文介绍在连续曲界面均匀各向同性层状介质情形下,根据地震反射资料反演三维地层结构及地震速度的地震射线层析成像方法。正演模型射线追踪是根据Fermat原理实现的,即求解由时间最小原理列出的非线性方程组得到射线与界面的交点。反演的目标就是求解这样的地层模型,使其射线追踪走时与观测走时的残差为极小。具体做法是根据走时残差求取模型参数修正量的阻尼极小二乘解,循环迭代直至满足要求为止。 文中给出了两个合成资料反演的例子。一个是合成的P波走时数据(加随机噪声)的反演,其结果与真实模型能较好吻合。另一个例子则同时考虑P波及反射P—SV转换波走时,反演包括P波速度和SV波速度对三维速度结构。利用P—SV波进行P波和SV波速度结构的成像是本文的一个进步。

By Geophysical Tomography, it is meant inversion which presents a linear integral relationship between the data observed and formation parameters. The paper introduces that based on seismic reflection data, the inversion can be performed for 3-D subsurface structure and by the same token it can come true for the imaging of velocity by seismic ray tomography with the continuous curved surface in homogeneous, isotropic layered media. The forward modelling is performed by ray-tracing based on Fermat's principle, i.e. to determine the intersection of rays and interfaces by solving the non-linear coupled equation from the least time principle. The purpose of performing seismic C-T technique is to find an earth model where the travel time residuals of calculated data and observed data are minimized. In normal practice, the way we obtain the correction of the updated value of model parameters from travel time residuals is by solving a damped least square problem and then processing it iteratively until we are happy with the result turned out. Two numerical examples are given in the paper, one is the inversion of synthetic P-wave travel time data (with random noise) and its result turned out is very consistent with the real model, the other is, both P-wave and P-SV wave travel time are taken into consideration while performing the inversion of 3-D sabsurface velocity structure including 3-D structure of P-wave and SV-wave velocities. As compared with the descriptions in the REFERENCES, one can see from the improvement made over the imaging P-wave and SV-wave simultaneously on the converted P-SV wave travel time data. By the time you've finished reading the paper, you will feel progress is extended in imaging the P-wave and SV-wave velocities using P-SVwave.