大甲溪流域崩塌潛勢分析

Abstract

一、 計畫目標： 本研究主要探討崩塌潛勢之分析，以2009年莫拉克颱風在此集水區造成的崩塌為例。崩塌潛勢即用現知的數據資料，倚靠不同的模式分析，得出研究區域之崩塌潛在趨勢。集水區範圍為大甲溪東勢以上之範圍，因崩塌率及崩塌量皆較下游平地多，且德基水庫、梨山、合歡山等重要水利及觀光景點皆在此處。

二、 分析方法： 研究方法為應用地理資訊系統(GIS)進行分析，並將原始資料做彙整與數據分析，得出大甲溪東勢以上的各種現態及未來的崩塌潛勢，利用頻率比法、證據權重法得知該集水區之崩塌潛勢。

三、 結論與建議： 總結出兩項大標。選定因子方面，以相同因子做分析，證據權重法確實有更高的正確率，即代表因子之代表較契合分析方法之計算方式；其二為崩塌之不確定性，惟因子數據之採用皆以「已發生崩塌」之過往紀錄為依據，故對於崩塌是否發生以及發生之必要原因有其不確定性，未來可以增加更多考慮因子來讓崩塌潛勢及其分析之正確性更為接近事實之發生率，以達預防災害之目地。

Report Title: Potential collapse Dajia River Basin Analysis This study focuses on the potential collapse of the analysis to the collapse in 2009 of Typhoon Morakot caused in this catchment area for example. Potential collapse that is now known to use data, rely on different modes analysis, the collapse of the underlying trend study area. Catchment area in the range of the potential nature of the terrain above the east, because the collapse of the rate and amount of collapse are more than the lower plains and Deji Reservoir, Lishan, Acacia mountain water conservancy and other important tourist attractions are here. Methods for the application of Geographic Information Systems (GIS) to analyze raw data and make aggregated data analysis, the nature of the terrain east of the various potential than current state and future potential collapse, using the frequency ratio method, weights of evidence obtained The collapse of known potential catchment area. Summed up the two large scale. Selected aspects factor to do the same factor analysis, weights of evidence does have a higher accuracy rate, and representatives of the more fit factor calculation method of analysis methods; the other for the collapse of the uncertainty, but begin by factor of data, Past record of the "collapse has occurred," as the basis, it is necessary for reasons of occurrence of the collapse has occurred and there is uncertainty, the future can add more factors to consider to make the potential collapse of the correctness of its analysis closer to facts The incidence, in order to reach head to prevent disasters.