近年台灣因氣候變遷影響導致豪雨不斷,土石流邊坡滑動等天然災害層出不窮,而台北文山區地區歷經過96年至97年卡玫基、辛克樂、哈格等颱風所挾帶大量雨水侵襲下,導致台北木柵猴山岳地區發生了土石崩塌;就防救災的觀點來看,若能充分了解崩塌區之地質構造,災前、後地形地貌之變化,便可助於災後整治規劃,設計施工上之參考,甚者可提供其他地區規劃研究之參考。 由於本區域之滑坡體屬小規模崩塌,一方面在本研究區內缺乏災前、後較詳實之相關地形、地質資料,另再以國內已有之DTM來分析滑動體之地形、地貌特徵及變化情形,有實際的困難性。依循上述動機,本研究猴山岳運用災前與災後影像資料,結合多種軟體,嘗試建置高精度數值地形模型,由既有高精度LIDAR座標資料搭配航空影像及動態即時全球定位系統(RTK GPS)來設置及量測地面控制點;配合高機動性、即時快速的無人載具影像資料,建置出災前時期、崩塌後與整治完工後之解析度2m DTM。利用建置出各時期之DSM分析此地形地貌變化,利用Arc Map查詢功能求得縱剖面與橫剖面之地表高程,比對94年至99年期間高度變化,並搭配正射影像初探樹木侵蝕範圍。並根據所建置之數值地形模型,利用精度管控指標先行評估DSM平均垂直誤差值、標準差、較差極大值與極小值,並呈現出誤差常態分布圖,藉以高峰圖形了解精確度。 本研究根據已有高精度LiDAR DEM及DSM資料為基礎,來評估本研究中所建置的數期數值地形模型之模型高程精度;另外,根據所建置之不同時期之DTM,來探求崩塌區附近之地質概況,及滑坡體之運動特性。
Recently, the natural calamities due to the climate change and torrential rain during typhoon season have been threatening the safety of local resorts and the residents’ lives. It is difficult to analyze the change to the area’s landform by using 5m DTM for small-scale shallow landslides which occur in this area. In addition, Houshanyue is devoid of related data on the change of landform before and after disasters. Based on above reasons, the research is aimed at structuring a 2m DTM consisting of the landform before and after landslides, and the rebuilding by exploiting aerial image and positioning system with precise LIDAR coordinates loaded to measure control point and mobile Remotely Piloted Vehicle image data. Furthermore, according to the Digital Terrain Model (DTM), the Precision Control Index is used to evaluate the DSM Average Vertical Error, Standard Deviation, Worse Maximum, and Worse Minimum, forming Error Normal Distribution whose Peak Graph help us ensure that the accuracy is in ideal condition. The range of evaluation includes the whole area, slide land and skirts of the slide land to find the difference among three sections. We compare the variations of altitude from 2005 to 2010 by constructing DSM analyzing topographic change in every period and using the search function of Arc Map obtaining Surface Elevation of longitudinal section and transverse section. Besides, orthoimage is also exploited to detect the range of wood erosion.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。