Earthquake potential of active faults in Taiwan from GPS observations and block modeling

Wu Lung Chang, Kuo-En Ching, Chiou Hsien Lee, Yi Rui Lee, Chi Fang Lee

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Taiwan is located at the boundary between the Philippine Sea plate and the passive continental margin of the Eurasian plate and is one of the most seismically active regions in the world. In an attempt to evaluate the seismogenic potential of active faults in Taiwan, we separated the region into 34 blocks with 27 known active faults as their boundaries and employed a 3D elastic block modeling method to invert the Global-Positioning-System-measured surface deformation for block rotations and the fault coupling. Additional constraints from an upto-date dataset of geologic fault-slip rates were introduced to reconcile the discrepancy between the geodetically and geologically determined long-term slip rates. Our results show that the Hsinhua fault and the southern part of the Longitudinal Valley fault may be weakly coupled near the surface and therefore experience shallow creeping in the interseismic period. The slip-deficit rates, which relate to how fast the elastic strain is accumulated on faults, are relatively low (0:8-2:2 mm/yr) for faults in northern Taiwan compared with up to 4 mm/yr in the Western foothill of the central and southwestern Taiwan. Evaluations of earthquake potential based on our new modeling results indicate that the frontal thrust and the westernmost branch faults of central Taiwan and the northern Longitudinal Valley fault of eastern Taiwan are capable of generating Mw 6.0-7.3 earthquakes in the next few decades.

Original languageEnglish
Pages (from-to)1274-1286
Number of pages13
JournalSeismological Research Letters
Volume87
Issue number6
DOIs
Publication statusPublished - 2016 Nov 1

All Science Journal Classification (ASJC) codes

  • Geophysics

Fingerprint Dive into the research topics of 'Earthquake potential of active faults in Taiwan from GPS observations and block modeling'. Together they form a unique fingerprint.

Cite this