Seismic anisotropy tomography: New insight into subduction dynamics

Dapeng Zhao, Sheng Yu, Xin Liu

Research output: Contribution to journalReview articlepeer-review

90 Citations (Scopus)


Body-wave and surface-wave tomography, receiver-function imaging, and shear-wave splitting measurements have shown that seismic anisotropy and heterogeneity coexist in all parts of subduction zones, providing important constraints on the mantle flow and subduction dynamics. P-wave anisotropy tomography is a new but powerful tool for mapping three-dimensional variations of azimuthal and radial seismic anisotropy in the crust and mantle. P-wave azimuthal-anisotropy tomography has been applied widely to the Circum-Pacific subduction zones, Mainland China and North America, whereas P-wave radial-anisotropy tomography was applied to only a few areas including Northeast Japan, Southwest Japan and North China Craton. These studies have revealed complex anisotropy in the crust and mantle lithosphere associated with the surface geology and tectonics, anisotropy reflecting subduction-driven corner flow in the mantle wedge, frozen-in fossil anisotropy in the subducting slabs formed at the mid-ocean ridge, as well as olivine fabric transitions due to changes in water content, stress and temperature. Shear-wave splitting tomography methods have been also proposed, but their applications are still limited and preliminary. There is a discrepancy between the surface-wave and body-wave tomographic models in radial anisotropy of the mantle wedge beneath Japan, which is a puzzle but an intriguing topic for future studies.

Original languageEnglish
Pages (from-to)24-43
Number of pages20
JournalGondwana Research
Publication statusPublished - 2016 May 1


  • Azimuthal anisotropy
  • Earthquakes
  • Mantle dynamics
  • Mantle wedge
  • Radial anisotropy
  • Seismic tomography
  • Subducting slabs
  • Subduction zones


Dive into the research topics of 'Seismic anisotropy tomography: New insight into subduction dynamics'. Together they form a unique fingerprint.

Cite this