中文摘要 地震災害防治是急迫在眉睫的全球性議題，然而，有效的地震災害評估卻是 地震災害防治最重要的一環。本論文由五個單元組成：第一章概括介紹臺灣地體 構造、活斷層、地質分區、地震週期和地震週期模型。第二章則描繪如何利用現 代地震儀記錄的地震資料來制訂一個本地規模(Local magnitude)與震央震度 (Epicentral intensity)的經驗公程式：ML = 0.08I0 2 – 0.04I0 + 3.41，ML 係本地規模， 而I0 則是震央震度。通過此關係式，我們可以系統地依相同標準規納比較臺灣從 17 到20 世紀發生的歷史地震。歷史地震研究對於地震災害評估是很重要的，這 是因為現代地震儀所記錄的地震資料時間跨距很短，地震紀錄十分有限；藉由研 究歷史地震資料我們可較有效地評估大地震的發生週期。第三章廷續第二章歷史 地震的初步研究結果，比對區域的現今地表變形與地震分佈的關係。我們利用永 久散射體差分干涉法（Persistent Scatterers InSAR method）研究桃園-新竹-苗栗 地區近期構造活動，且從歷史地震事件的角度探討地殼變形與區域構造。第四章 分析台灣地震震度規模的問題，因為規模與震度估算的方法差異常導致不正確的 地震災害評估。最後，第五章以誤差百分比來驗証第二章所制訂的經驗公程，此 外平均和累加的PGA 和PGV 災害圖也在這章中呈現與討論。Abstract Earthquake hazard mitigation is a world-wide matter of great urgency and an effective seismic hazard assessment is vital to it. In this dissertation, five chapters are organized: Chapter I provides background information about tectonic setting of Taiwan, active faults, geologic domains, earthquake cycle and recurrence models. Chapter II describes how the instrumental seismic data can be utilized to formulate an average empirical relationship between ML, the local magnitude, and I0, the epicentral intensity: ML = 0.08I0 2 – 0.04I0 + 3.41, which, in turn, summarizes historical data for three hundred years in a homogeneous way in Taiwan. The study of historical seismicity is imperative in seismic hazard assessment since the instrumental seismic recorded history is too short to record a sufficient amount of large earthquakes for study. Historical data provide a window for us to qualitatively analyze large earthquakes in the first approximation. Following the first approximation in Chapter II, the seismogenic zone of Taoyuan-Hsinchu area (zone II) becomes the test region for investigating recent tectonic activity using Persistent Scatterers InSAR (PSI) method. We shift our angle of view in assessing seismic hazard from historical seismic events to crustal deformation from the space in Chapter III. In Chapter IV, this PhD dissertation addresses the inaccurate seismic hazard assessment arisen from earthquake intensity scale problem in Taiwan. Characteristics of damaging seismic events in Taiwan are also depicted. At last, in Chapter V, percentage error is especially performed to test the validation and reliability of the empirical relationship done in Chapter I: ML = 0.08I0 2 – 0.04I0 + 3.41. Average and cumulative PGA- and PGV-depicted seismic maps are compared to our results in historical seismicity study in Chapter II and radar interferometry analysis in Chapter III.
