[S-SS04] New trends in data acquisition, analysis and interpretation of seismicity

[SSS04-P02] Probabilistic Seismic Hazard Assessment in Complex Fault Systems: Exploring the Longitudinal Valley of Taiwan

*Ting-Ying Lu¹、Chung-Han Chan¹ (1.Department of Earth Sciences, National Central University, Taiwan)

キーワード：probabilistic seismic hazard assessment, Longitudinal Valley, Taiwan, Gutenberg-Richter law, pure characteristic earthquake model, Seismic Hazard and Earthquake Rates In Fault Systems (SHERIFS) model

This study validated the credibility of several seismic models frequently implemented in probabilistic seismic hazard assessments and evaluated their impacts on hazard levels. For model validation, we applied these approaches to the Longitudinal Valley, Taiwan, with high seismic activity and observation quality. We first introduced the Gutenberg-Richter law (G-R law) based on regression of observed seismic activity in the study region. It shows a good fit with observations for small to moderate magnitudes. However, when applying this model to the part of large magnitudes, the uncertainty may increase due to the limited number of events. Then we implemented the pure characteristic earthquake model to evaluate the recurrence rate of the maximum earthquake for each seismogenic structure and active fault based on structure parameters. Since the recurrence interval of most structures is longer than the observation period, validating the credibility of the pure characteristic earthquake model is rather difficult. To implement the Seismic Hazard and Earthquake Rates In Fault Systems (SHERIFS) model, we incorporated the b-value of the G-R law and seismogenic structure parameters to discuss rupture behavior of each individual fault. In the results, we took into account the observations at various distances from the fault. The model well forecasted the seismicity rate along each seismogenic structure as well as regional activity in the Longitudinal Valley and fit historical records by considering multiple-structure rupture. Subsequently, we integrated all seismic models into a probabilistic seismic hazard assessment and examined their similarities and differences. The results of this research will be vital for grasping the physics of earthquakes and offering more accurate parameters for seismic hazard evaluations, particularly in the context of complex fault systems.