Estimation of 3-D Complex Deformation of Surface-Rupture Earthquake With Automated Fault Trace Identification
Accurate characterization of three-dimensional (3D) coseismic surface deformation is essential for analyzing earthquake deformation patterns and estimating fault slip distribution. Existing methods, including strain models like SISTEM and SM-VCE, offer improved solutions for coseismic 3D deformation over least-squares methods. However, these approaches encounter challenges in accurately capturing the 3D deformation of complex surface rupture earthquakes, primarily due to their limited capabilities in identify InSAR deformation heterogeneities on the sides at the cross-fault boundary. To address these limitations, this study introduces a novel method that automatically identifies fault traces from InSAR deformation to improve 3D deformation inversion for complex surface rupture earthquakes. Our method leverages InSAR deformation gradients to extract fault traces’ pixel information (FTPI), categorize FTPI to building fault traces model. By distinguishing heterogeneous points on either side of the fault through the encoded fault traces model, the method utilizes uniform deformation from homogeneous points to reconstruct the 3D deformation field in the vicinity of the fault. Simulation experiments reveal a significant enhancement in the esteemed 3D components by 45.6% (east), 47.8% (north), and 31.4% (vertical) when compared to SISTEM and SM-VCE. Applied to the 2021 Maduo earthquake, our method demonstrates considerable improvements in extracting detailed coseismic 3D deformation field and strain field. Furthermore, the study highlights that, in addition to errors arising from predominant deformations on opposing fault sides, steep gradients on the same side can also introduce inaccuracies. The method can restore discontinuous deformation characteristics along the trace of surface ruptures and achieve a more refined coseismic 3D surface deformation. Consequently, a more accurate coseismic strain field can be calculated, providing essential reference for seismic hazard analysis.
Authors
- M. Chen (ORCID: https://orcid.org/0009-0003-9829-363X)
- Lin Liu (ORCID: https://orcid.org/0000-0002-9581-1337)
- Guangyu Xu
Institutions
- Chinese University of Hong Kong (HK)
- East China University of Technology (CN)
Publication Details
- Journal
- IEEE Transactions on Geoscience and Remote Sensing
- Published
- 2025-01-01
- DOI
- https://doi.org/10.1109/tgrs.2025.3573809
- Citations
- 1
- Primary Topic
- earthquake and tectonic studies
- Type
- article
- Field-Weighted Citation Impact
- 1.87
Funders
- National Science Foundation
- National Natural Science Foundation of China