Seismic traveltime inversion based on tomographic equation without integral terms

Guangnan Huang, Bing Zhou, Hongxing Li, David C. Nobes

Research output: Contribution to journalArticlepeer-review

13 Scopus citations


The Jacobian matrix in the seismic traveltime tomographic equations usually contains several integral terms. These integral expressions not only greatly increase the computational complexity of seismic traveltime tomography, but also increase difficulty for programming these expressions. Therefore, if these integral expressions of the Jacobian matrix can be eliminated, the program of seismic traveltime tomography can be greatly simplified. In order to solve the computational complexity of the traditional seismic traveltime tomography, we found an anisotropic seismic traveltime tomographic equation which does not contain integral expressions. Then, it is degenerated into an isotropic seismic traveltime tomographic equation. In order to verify the effectiveness of this seismic traveltime tomographic equation based on the node network, a program has been coded to execute seismic traveltime inversion. For a crosswell checkerboard velocity model, the same results are obtained by this proposed tomographic method and the traditional method (with integral terms). Besides, two undulating topography velocity models are used as testing models. Numerical simulation results show that this proposed tomographic method can achieve good tomograms. Finally, this proposed tomographic method is used to investigate near surface velocity distribution near a power plant. Tomogram indicates that contaminated liquid diffuses and aggregates along strata at a certain depth. And velocity is lower near pollutant source than that away from it.

Original languageBritish English
Pages (from-to)29-34
Number of pages6
JournalComputers and Geosciences
StatePublished - 1 Jul 2017


  • Checkerboard model
  • Model parameterization
  • Seismic traveltime tomography
  • Velocity inversion


Dive into the research topics of 'Seismic traveltime inversion based on tomographic equation without integral terms'. Together they form a unique fingerprint.

Cite this