| 摘要: | 在歐亞大陸與菲律賓海板塊的碰撞與隱沒作用下，台灣地區的地體孕震構造呈現相當程度的複雜度。根據地震網所蒐集的資料，也因逐年的深入研究，許多台灣及其附近的地下地體構造特性也逐漸瞭解。1995 年進行的台灣深部構造先導性計畫(TAICRUST)提供了國內研究同仁們參與資料分析、處理、解釋與成果發表等的研究經驗。1999 年的921 集集地震亦引發了許多地震學上重要但複雜度增加與企待解決的問題，包括如嘉南地區詳細的孕震構造、台灣海峽與對岸福建廈門地區若發生大地震則對台灣本島的影響又如何等等。這些尚待瞭解的科學性問題皆仍待國內同仁們逐步的瞭解。因此，針對西部麓山帶、台灣海峽及福建廈門地區所處的地體架構、孕震構造及地震活動的詳細研究為延續前期整合計劃研究成果的必須步驟。本計劃將結合中央大學、中正大學及中央研究院地球所的研究力量進行跨兩岸的整合型計劃。其目標為增加國內的自主研究經驗，透過兩岸合作的機制擴大研究的地域。研究的重點區域包括，台灣西海岸平原、兩岸間的台灣海峽及福建地區的深部地體孕震構造與地震活動等的系列研究。本計劃分年的研究重點為：第一年：配合台灣地區大尺度深部地體孕震帶資料的整合與分析工作，主要工作集中在資料的統整與分析，除延續前期整合計劃的研究工作之外，還包括了波場的模擬與逆推方法的準備、發展、測試及應用為前期計畫的工作重點。同時，亦擬透過寬頻地震網進行深部莫荷面三維空間分佈的分析與研究工作。大尺度波場模擬，著重於地形的效應、探討二維聲性及彈性波場；及黏聲性與黏彈性波傳理論。強調透過理論與數值計算一併考慮位移、速度與加速度場等基本物理量，評估數值模擬理論基礎與數值法的適用性，必要時引進更適合的數值方法，如Spectral Element Method 或改進後的Pseudospectral Method 等本計劃第一、二年的目的將包括：理論與技術發展目標：（1）著重於對二、三維黏彈性地震震波傳遞的研究與發展（Chen, 1996）。及由二維的問題延伸為三維問題的探討。（2）將理論與計算地震學結合，發展平行計算技術使其成為應用於本計劃的主題－三維地震學研究相關課題的主要發展與運用的工具。實際應用目標：（3）深部地體構造與形貌及側向速度變化的逆推成像研究。探討地震震波受震源、路徑及場址放大效應的基本原因。（4）根據接收函數訂定Conrad 與 Moho 不連續面的三維形貌分佈。(5)對振幅衰減應受距離、深度、頻率等的影響加以定性及定量的分析，並評估逆推的可行性。第二、三年的計畫目標為：除利用被動式與主動式震源收集地震資料進行模擬與逆推成像外，同時計畫透過不同研究與學術單位間的分工合作，預計參加 2006 年兩岸炸測野外地震網的建置與資料記錄的工作。擬按資料的特性，以被動式地震資料為主軸進行深部孕震構造的探索與逆推。透過線型地震網陣列的配置，本計劃預計運用波場模擬與反演方法進行資料的分析與相關研究。 This is a three-year project mainly focus on real data analysis, simulation, and imaging. This project will work closely with other colleagues under joint projects and related institutes from NTU, CCU, NCCU, and ASIES. Base on the previous experiences from deep crust structure studies in TAICRUST experiment and impacts from Chi-Chi earthquake, it is necessary to extend our view in seismology studies. Under the collaboration between main land and Taiwan, the first target will focus on deep seismogenic structure studies along Fuchian-Taiwan strait-Taiwan western plan. The prestack modeling including 2D and 3D wavefield simulation will based on the improved finite-difference or pseudospectral methods. If necessary, spectral-element methods maybe included to accurately and reliably perform data simulation and processing. Prestack wavefield imaging technique will be developed, tested and implemented for both active and passive source data. To perform accurate seismic modeling, both time-domain computation of viscoacoustic and viscoelastic responses will be investigated. Current 3D spatial distribution of P-, S-velocity will be studied, modified, and evaluated. Effects of source mechanism, propagation, and site response due to local geology and topography changes will be studied. The proposed research is to study 2D and 3D anelastic seismic responses. Numerical model building is a pre-requisite for prestack forward modeling. The field data study will focus on strong motion records that is widely available through TSMIP and data management system (DMS) since 1990. Another source of data will potentially from the study of ocean-bottom seismograph and multi-channel marine seismic data gathered in 1995 and 2004. The objectives of the three year proposal are: Theoretical studies and development: (1) to extend the previous work of 2-D viscoacoustic and viscoelastic simulation developed by Chen, (1996), Carcione (1988a, b, c) and Tal-Ezer et al. (1990) to 3-D case. (2) to combine theory and strength of computational seismology for better understanding of realistic wave propagation phenomena, parallel computing will the main tool developed for practical application. Practical applications are: (3) to perform prestack imaging technique for both earthquake and OBS data (4) to simulate earthquake source mechanism and rupture process when large earthquake occurred under or in the immediate vicinity of Taiwan Island. (5) to investigate source, path and site responses including topography effects; amplification due to effects of shallow soil/sedimentary layers basin shape, lateral velocity variation, Q distribution and others. (6) to investigate the effects of attenuation as the function of distance, depth, frequency, scattering and the feasibility of reverse-time wave propagation or inversion. 研究期間：9308 ~ 9407 |
