千岛湖地区上梧溪流域基流磷输出定量研究

Estimation of baseflow phosphorus export of the Shangwu River watershed in the Qiandao Lake region

基于双参数递归数字滤波(ERDF)、LOADEST模型和遗传算法,建立一种递归滤波基流负荷分割算法(RFLSA),对千岛湖地区上梧溪流域的基流总磷(TP)负荷进行分割定量.结果表明:利用遗传算法对双参数滤波日尺度退水常数和最大基流指数(BFI_(max))进行同步优化,可以有效提高ERDF基流分割结果的准确性和可靠性(NSE=0.92,RSR=0.29,R^(2)=0.92);以此为基础建立的RFLSA能够实现上梧溪流域基流TP负荷的准确定量(NSE=0.79,RSR=0.46,R^(2)=0.95),可以作为流域尺度上基流非点源污染定量评价的一种有效方法;2020年11月—2021年10月,上梧溪流域以基流形式输出的TP负荷量为0.167 kg·hm^(-2),占总径流负荷量(0.302 kg·hm^(-2))的比例高达55.30%.基流已经成为上梧溪流域非点源TP的主要输出途径,是该地区地表河流水体一个不容忽视的重要污染源.

Based on the two-parameter recursive digital filter(ERDF),LOADEST model,and genetic algorithm,a recursive digital filter load separation algorithm(RFLSA)was established for estimating the baseflow total phosphorus(TP)export of the Shangwu River watershed that located in Qiandao Lake region.Results indicated that the accuracy and reliability of the ERDF-based baseflow separation can be improved efficiently(NSE=0.92,RSR=0.29,R^(2)=0.92)by optimizing the daily recession constants and the maximum baseflow index(BFI_(max))simultaneously.The RFLSA,based on the improved ERDF,can simulate the baseflow TP loads of the Shangwu River watershed accurately,and it can be an effective method for baseflow non-point source pollution estimation at the watershed scale.From Nov.2020 to Oct.2021,baseflow TP export(0.167 kg·hm^(-2))accounted for about 55.30%of the total TP loads in streamflow(0.302 kg·hm^(-2)).Baseflow has become the main pathway of non-point source TP export in the Shangwu River watershed,and it is an important pollution source for the riverine water in this area.