参数组合优化的随机边界条件及其在地震RTM中的应用

Parameterized random boundary condition and its application on seismic RTM

地震数据叠前逆时偏移(RTM)在高精度成像中具有重要应用,但其对存储量具有极高要求。使用随机边界条件会大大降低逆时偏移的存储成本,文中以自相关函数类型、相关长度、速度扰动标准差3个参数描述的随机介质表达方式来构建参数化的随机边界,对其三参数的意义及取值进行了分析,认为指数型自相关函数、相关长度与波长接近、速度扰动标准差为30%左右为相对优化的随机边界参数组合。模型试验表明,在RTM中采用参数组合优化的随机边界条件能有效破坏人工边界反射波场的相干性,在节省存储量的同时获得较高精度的成像结果。

Pre-stack seismic reverse time migration has important application in high accuracy subsurface imaging. However enormous memory is needed in RTM.Using random boundary in RTM can reduce memory need of RTM effectively.In this paper, random media expressed by correlation function, correlation length, root mean square velocity perturbation is used, the meanings and values of parameters are analyzed and the optimal parameter group（ exponential correlation function ,the correlation length being equal to wave length and root mean square velocity perturbation being 30% ） are obtained. Numerical simulation shows that using random boundary with optimal parameter group can destroy the coherence of artificial boundary reflection wave fields effectively and obtain high accuracy imaging resuh with less memory at the same time.