本研究於利用國立屏東科技大學水土保持系戶外教室內有一面積0.32 ha示範坡地芒果園,2013年6月配合設置山邊溝及滯洪沉砂池後,隨即於同年8月起進行暴雨期間逕流量觀測至2014年10月止,藉以探討初設山邊溝後之坡地果園進行降雨量及逕流量觀測,於暴雨下之流量歷線變化,並利用SCS法來比較有效降雨量及實測逕流量間之關聯,並嘗試建立坡地果園於暴雨下之逕流係數及逕流量。結果得知坡地果園於2013年初設山邊溝期間及2014年山邊溝恢復地表植被期間,所測得之最大洪峰流量分別為0.0431及0.0119 cms,其值對照於合理化公式與修正三角形單位歷線法所估算之2年重現期距之洪峰流量分別為0.059、0.118及0.054 cms,此表示坡地果園於施作山邊溝後,洪峰流量未有高於估算值情形。觀測期間測得之最大洪峰逕流係數為0.21,整場降雨逕流係數在0.08~0.10之間,而美國農業部SCS逕流曲線法之有效降雨量估算逕流量有高於實測逕流量趨勢,經修正集水區最大蓄水量及入滲量估測逕流歷線及實測逕流歷線相近。
The purpose of this study was focused on the runoff variation by the hillside ditches constructed in the demonstrated mango orchard on the slopeland. The orchard with an area of 0.32 hectare was located at the outdoor classroom of soil and water conservation, National Pingtung University of Science and Technology. The hillside ditches system and sedimentation pond had set on the slopeland orchard in August, 2013, and then measured the rainfall and runoff until October 2014.All the measurements of rainfall and runoff after each storm events had analyzed the hydrograph and hyetograph to comprehend the hydrological parameters of hillside ditches system on the slopeland orchard. The SCS method of Soil Conservation Service, U.S.A. was used to realize runoff coefficient and runoff discharge by the efficient rainfall and practical runoff. The results were shown as the following, initial stage of hillside ditches construction in 2013 and afterward there was ground canopy recovered in 2014, the peak flows were 0.0431 and 0.0119 cms. The predicted peak flow by rational formula, triangular unit hydrograph and modified triangular unit hydrograph in 2-year return period were 0.059, 0.113 and 0.054 cms, and the comparison with the practical peak flow after hillside ditches constructed on slopeland orchard could result in the decreasing tendency of runoff. The runoff coefficient of maximum peak flow and rainfall period were 0.21 and 0.08-0.10. The runoff predicted by efficient rainfall of SCS method was more than practical value, but modified retention and infiltration of watershed could obtain a moderate prediction.