Please use this identifier to cite or link to this item:
標題: Modelling for Delineation of the Large-scale Landslide Potential in Laishe Creek Watershed
作者: Chia-Hung Lin
Shu-Fen Tsai
Chao-yuan Lin
關鍵字: Large-scale landslide;Risk analysis;Environmental index;大規模崩塌;風險度分析;環境指標
Project: 水土保持學報, Volume 48, Issue 3, Page(s) 1749-1762.
Recently, it has been widely reported that the duration and intensity of the most extreme rainfall events is increasing and consequently large-scale landslide, debris flow and river sediment occurred more severe in Taiwan. Such events caused catastrophe disaster which damage infrastructures, properties and human living. Laishe Creek watershed devastated by Typhoon Morakot was selected to assess the risk and scale of landslide potential. The landslide risk estimated from the relevant normalized factors i.e. inverted Normalized Difference Vegetation Index (iNDVI), distance from road, and distance from river. The landslide scale was derived from the combination of modified Normalized Difference Vegetation Index (mNDVI), Topographic Wetness Index (TWI), river curvature, and headwater erosion. Results show that riparian with risk value over 0.7 is the main landslide area of the watershed, and the landslides with high index of scale (> 0.8) were mostly occurred at headwater. In addition, the non-landslides with high risk (>0.7) and scale (>0.7) were located at the watershed outlet which was vulnerable to large-scale landslide occurrence.

近年來因極端氣候影響,強降雨事件的規模漸增,颱風暴雨所帶來的高強度及長延時降 雨,常造成集水區山坡地崩塌、河道淤積、土石流等災情,嚴重破壞森林原有保土蓄水功能及 國人生命財產損失。本研究以莫拉克颱風事件災損嚴重之來社溪集水區為樣區,採集水區事件 前之植生綠劣指數、距道路距離、距河道距離等指標,評估集水區崩塌風險值;另以植生覆蓋 指數、地形濕度指數、河道凹岸、向源侵蝕等指標推估莫拉克之崩塌規模。結果顯示,近岸崩 塌為來社溪集水區莫拉克風災之主要崩塌區位,其崩塌風險多大於 0.7;而崩塌規模大於 0.8 之崩塌區位多屬源頭崩塌;另外,崩塌風險大於 0.7 且規模大於 0.7 而未崩塌之區位,主要位 於集水區出口,為未來大規模崩塌亦發生之區位。
Appears in Collections:第48卷 第03期

Files in This Item:
File Description SizeFormat
98588-1.pdf1.65 MBAdobe PDFView/Open
Show full item record

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.