| 摘要: | 深層重力變形(DSGSD)通常發生在地形起伏大造山環境有關，其崩塌體積大、平時運動緩慢，往 往在短時間內造成重大破壞。然而，當它們轉變成快速移動和災難性的大規模滑移時，其移動距離 很遠會造成造成顯著的危險。受到氣候變遷之影響，坡地災害類型及規模遠超過過去歷史事件，尤 其莫拉克風災期間高雄市甲仙鄉小林村獻肚山發生之大規模崩塌，造成全村450 人失蹤與死亡，對 於大規模崩塌災害之相關研究與災害防治對策研擬更趨重要。台灣造山帶存在著超過1000 餘處之大 規模崩塌潛勢區位。若從後續災害防治與經濟效益之觀點考量，如何評估其活動性、抑或轉換成劇 變式山崩的可能性，成為政府單位進行山崩災害防治的重要議題。在過去幾年中，我們的研究團隊 通過使用空載光達LiDAR 產置的高精度數值地形模型（DEM）判釋了超過1600 個DSGSD。在這個 計劃中，我們提出了綜合性的科學研究，包括地形、構造、遙感和大地測量，進行台灣造山帶之深 層重力變形活動與監測。使用多時序時域相關點雷達干涉技術(TCP-InSAR)，配合日本衛星雷達影像 (ALOS/PALSAR, ALOS2)L 波段SAR 數據，可以穿透高植被覆蓋區域。TCP-InSAR 技術改善傳統合 成孔徑雷達影像分析方法，在植被茂密的山區有效觀測點數量過少的問題，可得到大面積的地表資 料，並將地表GPS 連續觀測資料納入TCP-InSAR 之解算，將其分析成果提高至公釐(mm)級精度。
Deep-seated gravitational slope deformations (DSGSD) are generally linked to high relief mountain environments and are voluminous, short travelling and slow moving failures (Agliardi et al., 2001). However, they sometimes transform into fast moving and catastrophic long-runout rockslides, which might pose a significant hazard for areas, situated in substantial distances from source zones (Brideau et al., 2005, 2009; Ganerød et al., 2008). Due to its geodynamical context and climatic setting dominated by active mountain building and high precipitations, DSGSD are commonly observed in mountainous region of Taiwan. The 2009 Typhoon Morakot induced the catastrophic Hsiaolin landslide that caused 450 casualties is one of the famous examples occurred in the island. Therefore, how to detect sites of DSGSD and monitor their activity and susceptibility becomes an urgent task for natural hazard mitigation of the island (Lin et al., 2011; Kuo et al., 2011; Tsou et al., 2011). In Taiwan, over thousand landslides that including some DSGSD transforming to fast moving occurred annually. In the past few years, our research team has identified over 1600 DSGSD sites by using airborne LiDAR derived Digital Elevation Model (DEM). In this project we propose a multi-disciplinary approach including Geomorphology, Tectonics, Remote Sensing, and Geodesy to monitor their activity and susceptibility in mountainous region of Taiwan. Our main target focused on the multi-temporal InSAR analysis ( TCP-InSAR) on the potential deep-seated landslide area. Using available L-band SAR data of ALOS-2 satellite, which could penetrate dense vegetation and used to estimate more ground deformation information to evaluate the activity and susceptibility of deep-seated gravitational deformation (DSGSD) which has potential transform to catastrophic landslide. In this project, our study areas include whole mountainous region of Taiwan Island. In order to overcome the steepness and heavy vegetation of mountains, L-band radar are selected due to its longer wavelength which can penetrate vegetation and can apply the TCP-InSAR techniques to determine the displacement of deformed slope by using images acquired by ALOS and ALOS-2 satellite. With the periodic satellite observation and continuous ground observation, a better knowledge of the activity and susceptibility of deep-seated gravitational slope deformation in an active orogenic and subtropical environment can be obtained. |
