本文研究坡面型土石流之潛勢分析方法,亦探討其土石流堆積土方量評估公式。採用莫拉克颱風高屏溪流域44處坡面型土石流與44處未發生土石流之坡面資料,萃取出14項特性因子,配合SHALSTAB模式預測模式研究淺層崩塌後,利用區別分析方法進行坡面型土石流發生潛勢分析。區別分析結果以有效集水指標、q/t值面積百分比、集水區面積、土石流發生區面積、集水區平均坡度、土石流發生區平均坡度為預測變數時其正判率最佳,整體正判率為81.8%。特性因子中以淺層崩塌預測模式中,q/t值面積百分比對正判率影響最大,代表若精準的預測坡面崩塌之潛勢更有助於判別坡面型土石流之發生。本研究進一步擬訂潛勢分級準則,將坡面型土石流分為低中高潛勢。 其次針對坡面性土石流堆積土方量,本研究擬定出兩項評估公式,一者為評估受極端氣候影響豪雨觸發之坡面型土石流堆積土方量公式為V=107812A0.75;其次為評估大地震後短期內坡面型土石流之堆積土方量公式為V=5409A0.44。另坡面型土石流堆積長度方面,於輕度變質岩地區,可參酌採用VD=40.61A0.166 ,進行堆積長度評估之用。
This study proposals a potential analysis model for the hillslope debris flow. 44 hillslope debris flows and 44 non-debris flow in Kao-ping River induced by Typhoon Morakot are employed in the analysis model. 14 factor were selected, included the index of landslide area ratio that predicted by SHLASTAB model. The result of the discriminant analysis shows the classification rate reached 81.8% in which the catchment indicators, watershed area, area percentage of prediction landslide, effective watershed area, watershed average slope, and effective watershed average slope are the most effective indexes to the analysis. The area percentage of prediction landslide is the maximum impact factor in all of factors meaning that a good prediction to the landslide potential does increase the prediction of hillslope debris flow. The potential model is proposal as well which also classified the hillslope debris flow into low, medium, and high potential. Last, this thesis proposal the evaluation model on deposition volume. Two evaluation formulas are composed from the deposition volume data form field in two circumstances, the extreme weather heavy rain and strong earthquake conditions. First, the formula V=107812A0.75 is proposal on the extreme weather heavy rain case. Nest formula V=5409A0.44 is used for the case of the deposition volume of hillslope debris flow after strong earthquake shortly. Furthermore, the deposition length for a hillslope debris flow is proposed as VL=40.61V0.166.