[S-CG49] Integrative seismic and secondary hazard/risk assessment

[SCG49-01] Probabilistic seismic hazard assessment for Taiwan: TEM PSHA2025

★Invited Papers

*Jia-Cian Gao^1,2,3,4 (1.Earthquake-Disaster & Risk Evaluation and Management (E-DREaM) Center, National Central University, Taiwan、2.Graduate College of Sustainability and Green Energy, National Central University, Taiwan、3.Institute of Applied Geology, National Central University, Taiwan、4.Cian Yu Earthquake Disaster and Risk Evaluation and Management Consultants Ltd.)

キーワード：probabilistic seismic hazard assessment, ground motion prediction equation, Taiwan earthquake model

To develop the next generation Taiwan probabilistic seismic hazard assessment, we refined earthquake and seismogenic structure databases and updated seismic source parameters to reflect the latest information. The areal sources surrounding Taiwan was divided into 30 zones to estimate seismic hazards, and seven earthquake catalogs from 1900 to 2023 were combined to calculate background seismicity, excluding foreshocks and aftershocks. Onshore and offshore structures have been updated in the model, as well as newly identified multiple structure ruptures based on Coulomb stress and historical evidence. The slip rate determination for onshore structures now incorporates geodetic rates alongside geological evidence-based calculations for a more comprehensive assessment. Seismicity in the Ryukyu and Manila subduction zones was reassessed with updated parameters, and intraslab seismic activity was reanalyzed using a 3D plate model.
Choosing a ground motion prediction equation is a key step. Based on new strong ground motion databases from 1991 to 2020, this study uses weights to ensure that the amount of data is consistent across a variety of magnitudes and distances. The model was objectively evaluated using log-likelihood function values (Log-Likelihood, LLH, Scherbaum et al., 2009) and modified Euclidean distance-based ranking (EDR, Akkar and Kale, 2014). The performance of each model is comprehensively sorted in order to determine the weight of the logic tree based on the matching degree of 16 domestic and foreign crustal and subduction zone earthquake models under various natural vibration periods.
Finally, by utilizing the aforementioned parameters combined with a new database of the average shear-wave velocity in the top 30 meters at each site and considering uncertainties caused by different GMPEs, we have developed a seismic hazard map. The results of this study will contribute to subsequent research on seismic risk assessments and the promotion of disaster prevention strategies in the public sector (such as earthquake scenarios, disaster evacuation and rescue drills, etc.).