台灣地動衰減式準確性分析;Study of the Predictability of Taiwan’s Ground Motion Prediction Equation

NCU Institutional Repository > 工學院 > 土木工程學系碩士在職專班 > 博碩士論文 > Item 987654321/97232

jsp.display-item.identifier=請使用永久網址來引用或連結此文件: https://ir.lib.ncu.edu.tw/handle/987654321/97232

题名:	台灣地動衰減式準確性分析;Study of the Predictability of Taiwan’s Ground Motion Prediction Equation
作者:	吳江瑩;Wu, Jiang-Ying
贡献者:	土木工程學系在職專班
关键词:	台灣地動衰減式;概似函數;地震動觀測資料;地盤效應;台灣地震;Loh et al. (1992) 模型;Ground Motion Prediction Equation;Likelihood Function;Ground Motion Data;Site Effects;Taiwan Earthquakes;Loh et al. (1992) model.
日期:	2025-07-24
上传时间:	2025-10-17 11:00:46 (UTC+8)
出版者:	國立中央大學
摘要:	台灣位處於歐亞板塊與菲律賓海板塊交界，地震活動頻繁，因此找尋適合本島特性的地震動預測模型對於地震防災與工程設計具有重要意義。本研究旨在探討台灣地區台灣地動衰減式（Ground Motion Prediction Equations, GMPE）之準確性，在已知的眾多GMPE中尋找出最適合台灣地區的GMPE，並評估其在不同地震條件下的適用性。本研究蒐集過去二十年內九場有感地震之地震動觀測資料，選取十組以台灣地區為時空背景的GMPE模型，運用概似函數（Likelihood Function）計算各模型之預測準確性，並以對數近似值（Log-likelihood）進行排序分析。同時探討震源距離、地震規模與場址條件等因素對模型表現之影響。研究結果顯示，不同GMPE模型在台灣地震條件下之適用性存在顯著差異。其中，Loh et al. (1992)之模型整體表現最佳，能較準確預測地表最大加速度（Peak Ground Acceleration, PGA）。本研究建議未來可擴大資料樣本，深入考量地盤效應，並發展台灣本地化之地震動預測模式，以提升地震災害評估之準確性與可靠性。研究成果可供後續地震工程設計、都市防災規劃與學術研究參考。 ;Taiwan is located at the boundary between the Eurasian Plate and the Philippine Sea Plate, making it a region of high seismic activity. Therefore, identifying ground motion prediction models suitable for Taiwan’s geological characteristics is crucial for seismic hazard mitigation and engineering design. This study aims to evaluate the predictive accuracy of Ground Motion Prediction Equations (GMPEs) in the context of Taiwan and identify those most applicable to local seismic conditions. The study collected ground motion data from nine significant felt earthquakes that occurred in Taiwan over the past two decades. Fourteen GMPE models, developed based on Taiwan-specific data, were selected for analysis. The Likelihood Function was employed to evaluate the predictive accuracy of each model, with performance ranked using Log-likelihood values. In addition, the effects of source-to-site distance, focal depth, and site conditions on model performance were examined. The results indicate that the predictive performance of GMPEs varies significantly under Taiwan’s seismic conditions. Among the models evaluated, the one proposed by Loh et al. (1992) demonstrated the best overall performance in estimating Peak Ground Acceleration (PGA). This study suggests that future research could incorporate a larger dataset, consider site effects in greater detail, and focus on the development of localized GMPEs tailored to Taiwan. The findings provide valuable insights for seismic engineering design, urban disaster prevention planning, and academic research related to seismic hazard assessment.
显示于类别:	[土木工程學系碩士在職專班] 博碩士論文

文件中的档案:

档案	描述	大小	格式	浏览次数
index.html		0Kb	HTML	12	检视/开启

在NCUIR中所有的数据项都受到原著作权保护.

社群 sharing

数据加载中.....