|Abstract: || 本研究以QUAL2K水質模式檢討人工溼地改善水質濃度之效果，以大漢溪中下游河段及其8座人工溼地為研究區域，研究目的為：(1) 評估人工溼地改善大漢溪水質（BOD、SS及NH3-N）之效果、(2) 情境模擬流量（Q90、Q75、Q50）對水質改善之影響、及(3) 探討各人工溼地對水質改善之貢獻度。|
首先，QUAL2K水質模式參數（溫度修正係數、祛氧係數及再曝氣係數）等分別以相關係數（r）、決定係數（R2）及模式有效性係數（EF）及平均絕對值誤差率（MAPE）作率定及驗證，結果顯示模擬值除NH3-N外，BOD及SS均於可接受範圍內（r > 0.5、R2 > 0.7、EF > 0及MAPE < 50），顯示本研究模式模擬參數值合理。
大漢溪沿岸人工溼地設置後可改善水體水質，以流量Q75為例，水質濃度改善程度BOD、SS及NH3-N分別為16%~41%、9%~16%及7%~23，依序為BOD > NH3-N > SS。情境模擬流量對水質濃度改善影響之結果，顯示流量越大則削減BOD及SS比例越小，BOD及SS水質濃度改善程度受情境流量影響，但NH3-N則不受影響。改善後之大漢溪中下游水質僅有SS符合丙類及丁類水體水質標準，BOD及NH3-N仍無法達到丙類標準，故需從源頭減量著手，減少大漢溪沿岸之生活污水後應能使BOD及SS濃度降低。
此外，各別人工溼地改善水體水質之貢獻度及占整體削減比例依序為，BOD：浮洲人工溼地（38%）> 華江人工溼地（22%）> 新海一期人工溼地（13%）。SS：城林人工溼地（47%）> 鹿角溪人工溼地（15%）> 浮洲人工溼地（13%）。NH3-N：浮洲人工溼地（33%）> 新海一期人工溼地（27%）> 華江人工溼地（21%）。顯示人工溼地因其設計處理水量及對BOD、SS及NH3-N處理濃度不同而使改善程度有差異。於不同設計情境流量下之模擬結果亦顯示源頭流量越大，人工溼地所能削減污染物之比例降低，削減污染物能力：Q90 > Q75 > Q50。
The water quality model QUAL2K is applied in this study to simulate the improvement of water quality by constructed wetlands. The study area of this study is midstream and downstream of Da-Han Creek with its eight constructed wetlands. the purpose of this study is to: (1) evaluate the improvement and effects of water quality (BOD, SS and NH3-N) of Da-Han Creek by constructed wetland, (2) evaluate the effect of flow scenarios (Q90, Q75, Q50) on water quality improvement, and (3) explore the contribution of each constructed wetlands for water quality improvement.
First, the fitness of the parameters (temperature correction factor, deoxygenation rate coefficient and reaeration coefficient) in QUAL2K were verified through calibration and validation based on four performance indexes, which are correlation coefficient (r), coefficient of determination (R2), model efficiency (EF) and mean absolute percentage error rate (MAPE), the results of calibration and validation indicate that the simulated values of BOD, SS are all located within acceptable range and have a good performance(r> 0.5, R2> 0.7, EF> 0 and MAPE<50)except NH3-N, thus it can be seen, the calibrated parameters were reasonable and applicable in this study.
The setting of constructed wetland on Da-Han Creek can improve the water quality. Take Q75 flow rate as an example, the improvement of BOD, SS and NH3-N were form 16% to 41%, 9% to 16% and 7% to 23, respectively, and the improvement of BOD is better than NH3-N and SS. The other major purpose of this study is to measure the impacts of different flow scenarios on water quality improvement. The results display that the reduction portion of BOD and SS concentration will decrease with increasing the flow rate. On the contrary, NH3-N are not affected. The improvement of water quality on middle and lower reaches of Da-Han Creek shows that only SS can meet the Surface Water Classification Standards of class C and D, however, BOD and NH3-N are still unable to meet the Surface Water Classification Standards of class C. The results point out that if we want to improve the water quality significantly, it proceed from the reduction sewage from the sources or upstream.
Furthermore, this study explores the efficiency and contribution of water quality improvement of each constructed wetland. For BOD: Fuchou constructed wetland (38%)> Hwajiang constructed wetland (22%)> Sinhai-I constructed wetland (13%). SS: Chenglin constructed wetland (47%)> Lujiao Creek constructed wetland (15%)> Fuchou constructed wetland (13%). For NH3-N: Fuchou constructed wetland (33%)> Sinhai-I constructed wetland (27%)> Hwajiang constructed wetland (21%). The results indicates that the improvement efficiency of Constructed wetlands is effected by the capacity of wastewater treatment and the concentration of BOD, SS and NH3-N. Through Different flow scenario simulation, the results also show that the higher flow from upstream, the lower pollutants reduction proportion of constructed wetland. In this study, the reduction proportions under different flow rate are Q90> Q75> Q50.