坡地崩塌是高破壞性的地質事件。台灣位處於歐亞大陸與菲律賓板塊聚合帶上，因此地震活動頻繁、造山速度快、地形陡峭且岩體破碎。這些先天存在的誘發坡地崩塌不良條件，加上台灣島位於西太平洋颱風路徑上，每年平均有三個半以上的颱風經過本島，伴隨颱風而來的快速強降雨致使島上山崩及土石流等事件十分頻繁。為了避免崩塌事件引發災情，事先的調查及防護工作是第一要務。本計畫將利用遙測科技於山區的坡地崩塌發生潛勢調查及發生規模評估。計畫規劃為三年期，先建置並測試崩塌潛感模式雛型，利用已發生之崩塌事件分析促崩因子對崩塌潛感模式之影響。建置準確性之崩塌潛感模式及崩塌潛勢調查圖之後，未來若有颱風降雨或大地震發生，將降雨資料或強地動水平量分布資料輸入模式中，即可以評估崩塌發生之規模及分布。主要工作內容包括：1) 遙測影像收及整理與建檔；2) 衛星影像分析，包括地物、地質與地貌之辨識與分類；3) 地形資訊萃取與分析；與5) 建立山崩塌潛感模式及崩塌潛勢圖。預期成果包括：1) 地形變遷資料庫可提供學術或工程研究應用，2) 增進了解台灣南部山區地形細部特徵及地形與構造地質的關係，3) 瞭解大型山崩的各種潛在因子與促崩因子，建立大型山崩之分類及各類山崩之因子統計分布及其參數特性，4) 建立地質災害潛勢分佈資料庫，提供防救災及公共基礎建設規劃的重要依據。Landslide is a high-destructive geological event. Taiwan island is located on the convergent point between the Eurasian plate and Philippine Sea plate. Phenomena associated with this special location include: high seismic activity, fast mountain building rate, steep terrains and broken rocks. Existence of these inherent adverse conditions for landslide, coupled with the very frequent typhoon events, an annual average of more than three and a half, caused that landslides and debris flows on this island are also very frequent. In order to avoid the potential landslide disaster, the survey and protection work before typhoon event is the most important task. This project aims at applying the remote sensing technology in landslide investigation and its potential assessment. The time to go through this project is taken into account as three years and this year is the first year. We will first build and test prototype of landslide susceptibility model by using of the occurred landslide events. Once the landslide susceptibility model and the landslide potential map have been established, we can assess the size and distribution of landslides that may triggered by large earthquake or typhoon event in advance. Main tasks of this project include: 1) remote sensing imagery collection and sorting; 2) satellite image analysis, including surface, geological and geomorphological identification and classification; 3) terrain information extraction and analysis; and 5) establishment of the landslide susceptibility model and the landslide potential map. Expected results include: 1) the terrain database, which can provide academic or engineering research and application; 2) better understanding of the detailed structures in southern Taiwan and the relationship between topography and geological structure; 3) understanding the factors of large landslides, establishing the statistical model for landslides; 4) the establishment of potential geological disasters database, which can provide important basis knowledge for public infrastructure planning. 研究期間：10008 ~ 10107
