论文部分内容阅读
本文首次将三维直角坐标系下导出的三维立体层析反演方法用于南海深水三维实际数据的偏移速度建模。三维立体层析反演的成功应用依赖于射线参数水平分量的正确提取以及则化项的正确实施,二者缺一不可。本文首先通过结构张量算法实现了对三维立体层析数据空间中最重要的数据分量:射线参数水平分量的高效率提取;其次,由于海洋三维数据具有炮线间距较大同时垂直于炮线方向的射线参数水平分量无法得到的特殊性,强化了规则化项的应用,最终够保证了算法的稳健收敛,反演得到的模型可以很好的服务于叠前深度偏移。基于三维理论数据与南海某实际三维数据的数值实验证实了算法以及实现策略的稳健性和可靠性。
This paper, for the first time, uses the three-dimensional tomography inversion method derived from a three-dimensional Cartesian coordinate system to model the migration velocities of three-dimensional real-world data in the South China Sea. The successful application of three-dimensional tomography inversion depends on the correct extraction of the horizontal components of the ray parameters and the correct implementation of the chemical terms, both of which are indispensable. In this paper, the most important data component of the three-dimensional tomography data space: the high-efficiency extraction of the horizontal components of the ray parameters is realized by the structure tensor algorithm. Secondly, because the oceanic three-dimensional data has the large spacing between the gun lines and the perpendicular to the gun line direction Which can not obtain the particularity of the horizontal component of the ray parameters, enhances the application of the regularized term and finally ensures the robust convergence of the algorithm. The inverted model can well serve the prestack depth migration. Based on the three-dimensional theoretical data and a practical three-dimensional data in the South China Sea, numerical experiments prove the robustness and reliability of the algorithm and the implementation strategy.