论文详情
提出了一种联合结构张量与运动学反偏移的立体层析数据空间提取与反演策略。首先在数据域搜索得到初始数据空间,可获得初始反演结果与初始成像数据体,然后基于初始成像数据体内的共偏移距成像剖面人工拾取出比较可靠的数据点位置,在这些位置搜索构造倾角与剩余曲率,最后利用运动学反偏移将上述成像域运动学信息反偏移到数据空间并实施校正,获得更为可靠与均匀的立体层析数据空间。无论是第一步的数据域搜索还是第二步的成像域搜索都是基于高效率、高精度的结构张量算法的实现,可认为是一种联合结构张量与运动学反偏移的两步法立体层析数据空间提取与反演策略。同时还探讨了当射线参数水平分量信息引入层析反演的数据空间后,实施反演算法的各种可能性。理论数据证实了上述策略的可靠性,为后续的实践奠定了理论基础。
Using a structure tensor and kinematic demigration,we propose a two-step scheme for data space construction and inversion in stereo-tomography.First,the scheme searches a prestack data volume to obtain the initial data space and obtains the initial inversion model and the initial prestack depth migrated (PSDM) image.Next,the scheme selects reliable data point locations in the initial PSDM image manually to search for the corresponding residual moveout (RMO) and structural dip.Then,it inverts the kinematic information from the imaging domain to the data space by kinematic demigration followed by correction,to obtain a reliable and uniform data space for stereo-tomography.Note that both the first step search in the data domain and the second step search in the imaging domain are performed using an efficient structure tensor algorithm with high precision.In addition,the paper discusses the types of possible approaches and strategies to accomplish the tomography when the horizontal component information of the ray parameter is introduced into the data space of the tomographic inversion.Finally,the reliability of the proposed scheme was verified through a synthetic data test,thus laying the foundation for its application in field data.
国家自然科学基金面上项目(41574098,41630964)、国家科技重大专项子课题(2016ZX05026-001-03)共同资助。