定量降雨與崩塌機率在土石流潛勢分析之應用

Abstract

至今土石流潛勢溪流分析方式仍常以村里為單位，因此常發生數條土石流潛勢溪流共用同一參考雨量資訊(雨量站)與警戒值之情況，所以單點雨量站觀測到的降雨，並無法代表鄰近溪流集水區亦發生降雨。本研究以集水分區的概念出發，直接以溪流為單位進行土石流潛勢溪流之分析。接續利用物理崩塌模式模擬崩塌規模，而為考慮降雨空間不均勻分佈之影響，崩塌模式輸入之降雨資料採用定量降雨資料(QPESUMS)，最後藉由坡地崩塌機率與分佈情況，作為土砂來源的分析依據，藉此研判土石流潛勢分析之基礎。分別以石門水庫集水區於艾利颱風及南投仁愛鄉於辛樂克颱風和海棠颱風為例，結果顯示本研究提出之分析方法可提升評估之正確性。

Up to date, the method of analyzing potential debris flow is still based on the area of village. So, it often happened that the same rainfall information (same rainfall gauge) and alert value were used to analyze the potential debris flow creeks which may belong to neighboring different sub-watershed. Therefore, the rainfall observed by a single rainfall gauge may not indicate that the rainfall also occurred in neighboring watershed around the gauge. This study is performed based on concept of watershed and taking the stream as the basic unit to analyze potential debris flow creek. The physical landslide model is used to quantify the landslide scale occurred in the watershed. In addition, to consider the effect on non-uniformity of spatial distribution of rainfall, the rainfall data of landslide model input is given from the quantitative precipitation data (QPESUMS). In the end, by taking into account landslide distribution in the watershed and its related probability, the source of sediment is analyzed as the base to examine the potential occurrence of debris flow. Taking the typhoon Aere in the Shihmen Reservoir watershed, typhoons Sinlaku, and Haitang in Renai township (Nantou) as examples, the proposed analysis method raises the evaluation accuracy for the potential debris flow up to a certain extent.