4.4 Article

Segmentally Iterative Ray Tracing in Complex 2D and 3D Heterogeneous Block Models

期刊

出版社

SEISMOLOGICAL SOC AMER
DOI: 10.1785/0120090155

关键词

-

资金

  1. National Natural Science Foundation of China [40721003, 40830315, 40874041, 40404009]
  2. National Probing Project [SinoProbe-02-02]
  3. Chinese Academy of Sciences [KZCX2-YW-132]
  4. Special Project for the Fundamental R & D of Institute of Geophysics, China Earthquake Administration [DQJB09B11]
  5. Knowledge Innovation Program of the Chinese Academy of Sciences

向作者/读者索取更多资源

We describe a complex geologic model as an aggregate of arbitrarily shaped blocks separated by cubic splines in 2D and triangulated interfaces in 3D. Recently we have introduced a segmentally iterative ray-tracing (SIRT) method based on Fermat's principle of stationary travel time, which has been documented to be robust and fast for a complex block model with a constant velocity defined in each block. In this work, we extend the constant velocity to a generally continuous distribution with an analytical expression of travel time, and develop SIRT in the redefined velocity distribution. As a three-point perturbation scheme, SIRT requires an explicit analytical travel time between two intersection points expressed as a function of coordinates of the two points. In these situations, we derive a general midpoint perturbation formula, and further a detailed perturbation formula for familiar media with a constant velocity gradient. SIRT is a scheme in which we perturb the intersection points of an initial-guess ray path in sequence by the first-order explicit formulas instead of using traditional iterative methods. A key consideration is the fact that the number of intersection points may be variable during the iteration process. Numerical tests demonstrate that SIRT is effective in implementing kinematic two-point ray tracing in complex 2D and 3D heterogeneous media.

作者

我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。

评论

主要评分

4.4
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据