标签：nbsp eth cti may mis oge tran research nsf
The DSurfTomo is short for ‘Direct surface wave tomography method‘, intiailly developed in
Fang, H., Yao, H., Zhang, H., Huang, Y. C., & van der Hilst, R. D. (2015). Direct inversion of surface wave dispersion for three-dimensional shallow crustal structure based on ray tracing: methodology an applications. GJI, 201(3), 1251-1263
and can be found on github (https://github.com/HongjianFang/DSurfTomo).
The code uses the fast marching method (FMM, implemented by Rawlinson et al., 2004) to perform the 2-D ray tracing at each period, and uses the eigenvalue solver of Herrmann (2004) for 1-D forward simulation. The basic idea in Fang et al. (2015) is to construct a pseudo-3D kernel by combing the horizontally 2-D kernel and 1-D kernel in vertical direction.
I personally find the code quite easy to use. In addition, the method has already been extended in the 3-D joint inversion framework (e.g., Fang et al., 2017 JGR).
However, some new problems evolves as the methd to be used in high frequency dispersion inversion within a relatvely small area (such as in the Hefei and Jinan city). One of the problem is the topography effect on the dispersion, because the current code does not take into consideration of how the surface wave propagate along the surface with topography. At first sight, the problem can be approximately solved by forcing the ray raypath within the surface, which can be done easily and has been done previously (a G3 paper and Xu XL2019). However, the problem is that, surface wave in different period seems to sense different wavelength spatially. Two approaches may be promising to solve the problem. The first one is suggested by Marteen De hoop in summar 2018, who visited USTC for several days. He suggested to transform the heterogeneous model to be flat, at the sacriface of introducing some anisotropy (He has a nice manuscript to describe the theory). Another approach may be to extend the method of ChenXF1999, which can calculate the dispersion of a heterogeneous media, and then static correction can be applied to the observed dispersion data. The 2-D Love case has already been considered by QiuHR in his bachelor degree thesis.
Of course, the topography effects have been investigated by many forward simulations (such as WangLM in JAG), and can be avoided in dispersion waveform inversion (LiJ 2017, 2018). Approaches to handle the topography effect in a more conventional framework, however, may also be valuable, because the waveform inversion is still not feasible for most researchers, and traditional inversion method requires less computer memory and time currently.
标签：nbsp eth cti may mis oge tran research nsf
原文地址：https://www.cnblogs.com/shaoqianHu/p/11310890.html
