考虑参数不确定性的数字滤波法在梅江流域基流分割中的应用

Digital Filter Method with Parameter Uncertainty in the Base Flow Separation of Meijiang Basin

参数不确定分析方法,如蒙特卡罗法、重现期法和直接积分法等,作为处理系统不确定的一种有效方法,一直在水文模型方面有着广泛的应用。该文以广东省梅江流域3个流量站为例采用数字滤波法进行基流分割,分析了基流的空间分布特征,基流分割的结果表明,基流量和基流指数自上游到下游递增,梅江流域的基流指数在0.25~0.31之间;分别假设滤波参数服从均匀分布和正态分布,采用蒙特卡罗法求得基流量及相应的可靠度,计算结果表明,常规方法计算得到3个站基流量分别为3.461亿m^3/a、12.373亿m^3/a和24.834亿m^3/a,对应的可靠度在两种分布中均不足60%。

Parameter uncertainty analysis methods,such as the Monte Carlo,the return period and direct integral methods,etc.,as effective methods for system uncertainty,have been widely used in hydrological models. Three gauging stations of Meijiang watershed in Guangdong Province are chosen as case study. The spatial distribution characteristics of the base flow are analyzed and the results show,the base flow volume and base flow index both increase upstream to downstream,and base flow index ranges from0. 25 to 0. 31. The filter parameter is assumed to be uniform distribution and normal distribution,and the annual average baseflow and its accuracy are obtained. The calculation results indicates,the base flow volume of the three stations with the conventional method are calculate as 3. 461＊ 108~3/a,12. 373 ＊ 108 m3/a and 24. 834108 m3/a respectively,but their reliabilities are all less than 60% in both distributions.