The Sichuan-Yunnan region,at the southeastern edge of the Tibetan Plateau,exhibits highly complex crustal structures and frequent seismicity driven by the ongoing collision between the Indian and Eurasian plates.Physi...The Sichuan-Yunnan region,at the southeastern edge of the Tibetan Plateau,exhibits highly complex crustal structures and frequent seismicity driven by the ongoing collision between the Indian and Eurasian plates.Physics-based simulations of strong ground motion are critical for reliable seismic hazard assessment in this tectonically active region.In recent years,multiple 3D velocity models have been developed for the Sichuan-Yunnan region using various datasets and methods.However,systematic studies evaluating how different velocity models affect strong ground motion simulation results are still scarce.In this study,we simulate strong ground motion for the M_(w)6.6 Luding earthquake that occurred on September 5,2022,using a finite-fault rupture model derived from observational data and a suite of different velocity models.We assess the reliability of each velocity model by comparing simulated and observed PGV at individual stations,and evaluate inter-model consistency by analyzing the spatial distribution of PGV across different models.The tested models include:SWChinaCVM-1.0/2.0,USTClitho1.0/2.0,CSES_VM1.0,ChinaM-S1.0,ShallowVs,and two 1D models with and without topography.Results show that the adopted fault and 3D velocity models can accurately reproduce PGV up to 0.3 Hz,with increasing errors observed at higher frequencies.Averaging simulation results from multiple models yields lower overall errors and reduces systematic biases introduced by a single model,thereby enhancing the reliability of ground motion prediction.Among the tested models,the SWChinaCVM series,USTClitho series,and CSES_VM1.0 showed similar spatial distributions of PGV and also produced results that closely match observed PGV at individual stations.In contrast,the ShallowVs model tends to overestimate PGV,while the ChinaM-S1.0 model tends to underestimate it.This study offers practical guidance on velocity model selection for ground motion simulations in the Sichuan-Yunnan region and offers insights for further refinement of regional velocity structures.展开更多
基金supported by the National Key R&D Program of China(Grant No.2023YFC3007304)the Guangdong Science and Technology Programme(Grant No.2023A1111120028)the Guangdong Provincial Pearl River Talents Program(Grant No.2019QN01G801)。
文摘The Sichuan-Yunnan region,at the southeastern edge of the Tibetan Plateau,exhibits highly complex crustal structures and frequent seismicity driven by the ongoing collision between the Indian and Eurasian plates.Physics-based simulations of strong ground motion are critical for reliable seismic hazard assessment in this tectonically active region.In recent years,multiple 3D velocity models have been developed for the Sichuan-Yunnan region using various datasets and methods.However,systematic studies evaluating how different velocity models affect strong ground motion simulation results are still scarce.In this study,we simulate strong ground motion for the M_(w)6.6 Luding earthquake that occurred on September 5,2022,using a finite-fault rupture model derived from observational data and a suite of different velocity models.We assess the reliability of each velocity model by comparing simulated and observed PGV at individual stations,and evaluate inter-model consistency by analyzing the spatial distribution of PGV across different models.The tested models include:SWChinaCVM-1.0/2.0,USTClitho1.0/2.0,CSES_VM1.0,ChinaM-S1.0,ShallowVs,and two 1D models with and without topography.Results show that the adopted fault and 3D velocity models can accurately reproduce PGV up to 0.3 Hz,with increasing errors observed at higher frequencies.Averaging simulation results from multiple models yields lower overall errors and reduces systematic biases introduced by a single model,thereby enhancing the reliability of ground motion prediction.Among the tested models,the SWChinaCVM series,USTClitho series,and CSES_VM1.0 showed similar spatial distributions of PGV and also produced results that closely match observed PGV at individual stations.In contrast,the ShallowVs model tends to overestimate PGV,while the ChinaM-S1.0 model tends to underestimate it.This study offers practical guidance on velocity model selection for ground motion simulations in the Sichuan-Yunnan region and offers insights for further refinement of regional velocity structures.
