霧社水庫集水區土壤沖蝕預測模式研究

Study on Erosion Model to Predict Soil Loss for Wushoh Reservoir Watershed

楊凱翔(Kai-Hsiang Yang)

指導教授 ： 張德文

淡江大學/工學院/土木工程學系碩士班/碩士(2017年)

https://doi.org/10.6846/TKU.2017.00549

摘要

本研究參考林韋成(2014)及楊尚霖(2016)對石門以及曾文水庫集水區土壤沖蝕特性研究，探討霧社水庫集水區土壤沖蝕情形，以供後續應用參考。研究以WEPP(WaterEerosion Prediction Project)為分析工具，針對WEPP模式所需輸入參數逐一說明，並依輸入參數進行敏感度分析，瞭解何項土壤沖蝕參數對沖蝕量影響較鉅，同時建立WEPP二階段數值模擬分析，第一階段為特定年降雨量分析，依此模式可快速預測年沖蝕量，以預測長期沖蝕變遷趨勢。另以分析結果率定邊坡坡型、坡度分布範圍。第二階段為單場降雨事件分析，以單場降雨量及降雨延時為基礎，將監測時距各降雨事件進行雨場切割，建立降雨參數，逐一分析其沖蝕量，以探討各別雨場沖蝕狀況。主要成果如下:
1. 經敏感度分析可發現：降雨量、臨界剪切力、有效水力傳導度及坡度為影響沖蝕量的重要參數。
2. 以WEPP模擬霧社水庫集水區邊坡沖蝕行為，改變坡型、坡度及坡長之參數，依結果得知當坡長越長、坡度越陡，皆使沖蝕量增加，且坡型不同皆會影響沖蝕深度的大小，其沖蝕深度由大至小為凸坡>S坡≧平坡>凹坡。
3. 特定年雨量分析可模擬土壤沖蝕趨勢，但並不準確，因年降雨量包括所有降雨事件，而許多零星的降雨事件僅造成入滲而不造成逕流。單場降雨分析將零星降雨事件剔除，使分析結果較準確，惟缺點為步驟繁多，分析時間較冗長。
4. 以二階段分析模擬將各沖蝕點位分析結果累加與實際值比較。研究發現，坡度較陡之邊坡能準確預測土壤沖蝕深度，準確值達95%。惟需注意緩坡案例，故建議以WEPP分析時，坡度需大於25度以上，方可有效模擬沖蝕量。
本研究另嘗試規劃一套符合臺灣本土化土壤沖蝕渠槽試驗，以便於未來研究土壤沖蝕參數率定及沖蝕相關應用模擬。能有效地模擬試驗區域土壤沖蝕情形，節省大量時間，以探究土壤沖蝕力學特性、降雨逕流與土壤沖蝕互制行為及局部區域之土壤沖蝕力學機制等。

關鍵字

土壤沖蝕 ； 敏感度分析 ； WEPP ； 特定年雨量分析 ； 單場降雨事件分 ； 土壤沖蝕渠槽試驗

並列摘要

This study refers to the soil erosion characteristics of Shimen and Zengwen reservoir watershed studied by Lin W.C. (2014) and Yang S.L. (2016), and discusses the soil erosion situation in Wushoh reservoir watershed. This study adopts WEPP (Water Erosion Prediction Project) as analysis tool. By varying the WEPP input parameters, the sensitivity analysis was carried out accordingly to understand which the soil erosion parameters have great effects on the soil erosion. As the same time, establishing the WEPP two-stage numerical simulation analysis is conducted. In predicting the amount of soil erosion of the Wushoh reservoir watershed, the main results are concluded as follows:
1.The sensitivity analysis result suggested that rainfall, critical shear, effective hydraulic conductivity and slope are important parameters for soil erosion.
2.The WEPP is used to simulate slope erosion behaviors for Wushoh reservoir watershed. The results suggest that the steeper and the longer the slope is will increase the amount of erosion. The erosion depthfrom large to small is convex slope> S slope≧uniform slope>concave slope.
3.Two-stage numerical simulation analysis was made by continuous daily simulation analysis and single-storm analysis. Continuous daily simulation analysis can simulate soil erosion trends, but it is not accurate due to annual rainfall of many sporadic rainfall events where only infiltration occurred without causing runoff. Single-field rainfall analysis excludes sporadic rainfall events, so that the results are more accurate, but the shortcomings of having many steps let longer time of analysis.
4.The results are compared with the actual values by the two-stage analysis. It is found that the slopes with steep slope angle can accurately predict the depth of soil erosion, the accuracy is about 95%. But attention should pay to gentle slope case. In WEPP analysis, the slope angle should be least 25 degrees in order to simulate the amount of erosion effectively.
This study intends to plan a set of soil erosion flume test with Taiwan's reservoir watershed in order to facilitate the future study of soil erosion parameters and erosion-related applications simulation, which can effectively simulate the soil erosion by experiments and saving a lot of time to explore the soil erosion mechanics characteristics, rainfall runoff and soil erosion intercourse behavior.

並列關鍵字

Soil Erosion ； Sensitivity analysis ； Single Storm ； Rainfall Erosion ； Erosion Flume test

參考文獻

27. Tsai, K.J. (2009),” Application of Soil Erosion Control Measures on Slopeland Located at Mudstones in the Southwestern Taiwan.” CIMME Annual Convention

44. 江介倫(2013)，「氣後變遷對臺灣年降雨沖蝕指數潛在影響」，地理學報，第68期。

45. 何幸娟、林伯勳、冀樹勇、伊孝元、施美琴、羅文俊 (2012)，「神木集水區土壤沖蝕特性」，中華水土保持學報，第275-283頁。

46. 吳明峰，范正成，楊智翔，鐘宜全，楊文仁(2009)，「石塊敷蓋對紋溝間土壤沖蝕之影響」，中華水土保持學報，第40(2)期，第159-171頁

54. 林俐玲及張舒婷(2008)，「土壤沖蝕性指數估算公式之研究」，水土保持學報，民國97年第12期，第355-366頁。