Earthquake Science (Feb 2023)

A comparative study of seismic tomography models of Southwest ChinaKey points

  • Xuezhen Zhang,
  • Xiaodong Song,
  • Feiyi Wang

Journal volume & issue
Vol. 36, no. 1
pp. 15 – 39

Abstract

Read online

The margin of the Tibetan Plateau of Southwest China is one of the most seismically active regions of China and is the location of the China Seismic Experimental Site (CSES). Many studies have developed seismic velocity models of Southwest China, but few have compared and evaluated these models which is important for further model improvement. Thus, we compared six published seismic shear-wave velocity models of Southwest China on absolute velocity and velocity perturbation patterns. The models are derived from different types of data (e.g., surface waves from ambient noise and earthquakes, body-wave travel times, receiver functions) and inversion methods. We interpolated the models into a uniform horizontal grid (0.5° × 0.5°) and vertically sampled them at 5, 10, 20, 30, 40, and 60 km depths. We found significant differences between the six models. Then, we selected three of them that showed greater consistency for further comparison. Our further comparisons revealed systematic biases between models in absolute velocity that may be related to different data types. The perturbation pattern of the model is especially divergent in the shallow part, but more consistent in the deep part. We conducted synthetic and inversion tests to explore possible causes and our results imply that systematic differences between the data, differences in methods, and other factors may directly affect the model. Therefore, the Southwest China velocity model still has considerable room for improvement, and the impact of inconsistency between different data types on the model needs further research. Finally, we proposed a new reference shear-wave velocity model of Southwest China (SwCM-S1.0) based on the three selected models with high consistency. We believe that this model is a better representation of more robust features of the models that are based on different data sets.

Keywords