Huang J.Y.Chao S.H.Lin C.M.CHUNG-CHE CHOU2026-03-122026-03-122024https://www.scopus.com/record/display.uri?eid=2-s2.0-105027916309&origin=resultslisthttps://scholars.lib.ntu.edu.tw/handle/123456789/736252Understanding the spatial distribution of velocity pulse (VP) is essential in engineering seismology. However, near-fault observations are lacking in worldwide application to constrain a regionalized empirical VP model, so additional scenario-based VP simulations are needed. This study presents a site-and fault-specific evaluation case in a sparse seismicity region, such as a target site located directly on a suspected fault trace of the Sanchiao Fault in northern Taiwan. Instead of using a time-consuming three-dimensional theoretical ground-motion simulation method (i.e., spectral element method) or a spatially limited empirical Green’s function (EGF) method, a comprehensive procedure derived based on a stochastic finite-fault ground motion simulation technique was conducted in this research. The restriction of the stochastic point source simulation technique cannot be directly used to predict the acceleration time history due to random phases. Several validation cases were performed by comparing observed VP and simulated VP for recent large earthquakes, such as occurred in the 2016 Kumamoto, Japan, 2018 Hualien, Taiwan, and 2022 Guanshan and Chihshang, Taiwan. If the localized source, path effects, and an empirical transfer function of Huang et al. (2017) were used, when the simulation procedure combined with a proposed moving random asperity model, reasonable evaluations for capturing pulse period (Tp) of near-fault VP for large earthquakes could be performed. Good agreements in velocity time history and extracted Tp of VP for the validated cases indicated that the size and location of asperities mainly control the long-period near-fault VPs on the fault from the finite-fault process as well as the directivity effect but are not sensitive to the random phase of each point-source simulations. Finally, the proposed procedure was used in a site-specific deterministic seismic hazard analysis case. The evaluation of the case study showed that the current design spectrum may be insufficient if the site-specific near-fault effect is not adequately considered.falseconference paper2-s2.0-105027916309