本研究以二重疏洪道人工溼地為模擬對象,使用AQUASIM建構的地下流式人工溼地水質模式(許,2009)討論影響淨水效果的重要機制。首先對水質觀測資料進行單因子變異數分析(One-way ANOVA),決定敏感度分析項目,水質模式參數則藉由現地實驗觀測(郭,2009)資料經回歸計算求得適合值,最後對生化需氧量(BOD)、溶氧(DO)、硝酸態氮(NO3--N)與懸浮固體(SS)四項水質進行參數敏感度分析,計算反應速率係數與出流物質濃度的絕對-相對敏感度值(SensAR),比較礫間處理槽內各種去除機制的影響力與差異性。 敏感度分析結果顯示,生物膜吸附作用在大部分水質循環中都具有一定程度的影響力,唯有懸浮固體主要是透過物理性的過濾作用去除,故具有較多孔隙空間提供生物膜生長的牡蠣殼介質溼地在水質淨化上具有較好的效果。
This study focuses on the sensitivity analysis of parameters in a water quality model for subsurface-flow constructed wetlands (Hsu, 2009). An one-way ANOVA of experimental data which is implemented to identify the difference in removal quality, and to decide the objects of sensitivity analysis. The model is built by AQUASIM 2.1 and the parameters are regressed by experimental data (Kuo, 2009). Absolute-relative sensitivity (SensAR) of BOD, DO, NO3--N and SS submodels can estimate the importance of each parameter or reaction. Therefore, the significant mechanisms in the constructed wetlands can be determined. By the result of sensitivity analysis, biofilms adsorption have significant influence in most of removal of substance except SS which is mainly removed by filtration. The treatment efficiency in oyster-shell wetlands is better than the gravel-bed wetlands, because the oyster-shell can provide more habitat for microorganism growing.