TTI介质弹性波随机边界逆时偏移的实现

从TTI介质弹性波波动方程出发，推导出逆时偏移高阶有限差分算子，应用完全匹配层(PML)吸收边界条件压制边界反射，通过高阶有限差分与通量传输校正(flux-corrected transport,FCT)方法压制网格频散，建立起常规的TTI介质弹性波有限差分逆时偏移流程。在此基础上针对逆时偏移存储量大的问题，采用计算换存储的思路，引入随机边界条件有效减少存储空间，给出了便于实际应用的TTI介质弹性波随机边界逆时偏移方法实现流程。最后采用传统的拉普拉斯滤波方法压制低频成像噪声。多层模型和逆掩断层模型试算结果表明，随机边界逆时偏移流程与常规有限差分逆时偏移流程成像界面位置一致，只是存在少量可以忽略的随机噪声，证明了该方法的正确性与有效性。

Starting from the TTI elastic wave equation,the high-order finite difference operator of elastic reverse time migration is obtained,the boundary reflection is suppressed by using perfectly matched layer (PML) absorbing boundary condition and the grid dispersion is suppressed by using high-order finite difference and flux-corrected transport (FCT) method,then the traditional reverse time migration process is established.Aiming at the massy memory demand,we adopt the random boundary condition for reducing the memory demand which increasing the computation cost.Then the TTI media reverse time migration process facilitated to practical application based on random boundary is established.Finally Laplace filtering was adopted to eliminate the low-frequency noise.The random boundary RTM and conventional finite difference RTM were tested on a multi-layer model and an overthrust-fault model.The result shows that the position of imaging interface is coincided and only little random noise exists,which proves the correctness and effectiveness of random boundary RTM.