評估坡地土壤沖蝕情形為集水區治理重要之基礎。現今評估土壤沖蝕量多以經驗模式(USLE)為主,雖可快速估算各集水區之沖蝕量,但由於經驗模式之參數為簡化沖蝕過程而得,尤其在評估農地時往往會因耕作制度、作物種類和水土保持處理等不同因素產生誤差。有鑑於此,美國農部為彌補USLE公式之不足而開發Water Erosion Prediction Project (WEPP)模式。本研究為評估WEPP模型於台灣南部熱帶氣候環境下土壤沖蝕量之預測能力,在沒有矯正WEPP參數值的狀況下利用GeoWEPP模擬老埤台地不同土地利用於2014~2015之年土壤沖蝕量。本研究同時於不同土地利用下,在現地埋設沖蝕針實際量測土壤沖蝕結果。結果顯示,比較WEPP模式推估值與現地針蝕針實測值,flowpath模式有較小的差異出現於鳳梨田、裸露地與檳榔園,其十年模擬平均值相對誤差介於19%~44%之間,而watershed模式於未修正參數狀態下不適用於推估老埤台地不同土地利用之年平均土壤沖蝕量。
Well understanding of soil erosion situation can help us to make a suitable management strategy in a watershed. Universal Soil Loss Equation (USLE) is a popular equation to predict soil loss contents, but its parameters should be still revised based on local climate conditions, topography and etc.. United State Agricultural Department (USDA) attempts to develop another model, Water Erosion Prediction Project (WEPP), to predict sediment production on slopelands to compensate USEL in precision of soil loss prediction. This study used WEPP model to predict sediment production in different land uses at Laopi terrace under a tropical climates. We also set up erosion piles in different land uses at Laopi terrace to validate the prediction values. The results indicated that WEPP model might effectively predict sediment production in pineapple, bare land, and areca farms due to small relative error ranged from 19% to 46% found between predicted and measured values, particularly in “flowpath” model. We concluded that some parameters of WEPP model should be further revised when we used it under tropical climates, particularly in “watershed” model.