泰平溪集水區臨溪崩塌空間分布及土砂量之研究

Abstract

台灣河川流域集水區之保育治理已成為水資源運用管理重要的一環，溪流臨岸崩塌之土砂直接匯入河道，為河道主要的土砂來源，故為保育治理之重點，為此如何快速且準確評估溪流臨岸之土砂生產，即成為重要且關鍵的課題。本研究利用空載光達（Airborne LiDAR）技術，以95年至98年間建置之高精度數值地形及同步施測之航照，界定野溪深水線及水路流動範圍，再由坡度明顯陡降變化點作為臨溪坡面與河道分界點，以劃定臨溪崩塌區位；最後，以LiDAR產製之數值地形與判釋之崩塌地區域，建立臨溪崩塌土砂運移經驗式，作為未來快速預估臨溪崩塌的土砂運移量之依據。研究成果並與以往常用分析方法比較，以瞭解土砂運移經驗式之適用性及應用範圍。本研究之歸納結論及重要觀察摘要如下：(1) 利用坡度等值區及航照劃定臨溪區域，其各期臨溪裸露地佔集水區總裸露地面積皆超過48%；(2) 本研究以「裸露地面積」及「事發前坡度」建立臨溪崩塌土砂產量經驗式，其可靠性較以往土砂產量推估經驗式高；(3) 航照之陰影區內裸露地乃可能產生大量土砂，故需詳加考量陰影區域之影響，否則將可能低估土砂產量；(4) 含有事件累積雨量之經驗式，因無法確實釐清何雨場為主控崩塌產生之外在誘因，導致經驗式估算結果不佳。

Watershed management has become the important task for water resource management in Taiwan. The landslides adjacent riverbank always induce debris materials which will transport to the river channel directly and the mitigation of the landslides become the focal work in watershed management. Consequently, it is the key issue that how to estimate the volume of debris materials of the landslides adjacent riverbank quickly and accurately. In this study, airborne LiDAR is adopted to generate DTM and is combined with aerial photographs to determine flow line of the river and the boundary of riverbank. Furthermore, the area of landslides adjacent riverbank can be obtained by identifying the point where the slope of hillside near the river drops suddenly. The empirical equation of the volume of debris materials can be established between the difference of DTM and the landslide area. The results of this study can be concluded as follows: 1.The landslide area adjacent riverbank is over 48% to the total landslide area in the watershed. 2.The empirical equation of the volume of debris materials in this study will induce more accuracy of volume estimation compare to other empirical equations. 3.Because the landslide in the shadow area on aerial photograph image is always ignored and that leads to underestimate the volume of debris materials. 4.The empirical equation of the volume of debris materials including the accumulative rainfall factor will induce less accuracy of volume estimation because of the uncertainty of rainfall event and rainfall duration.