基流分割方法在疏勒河上游流域的应用对比分析

基于疏勒河上游昌马堡水文站2007—2016年日径流资料， 运用单参数数字滤波法、 平滑最小值法、 递归数字滤波法和HYSEP法（固定步长法、 滑动步长法、 局部最小值法）对其进行基流分割计算。结果表明： 4种方法均可实现基流的自动分割， 且得到的基流过程线平滑缓慢， 所获得的基流序列稳定， 变异性小。4种方法得到基流指数具有差异性， 最大为0.72， 最小为0.49， 其中递归数字滤波法估算的结果稳定性最高。单参数数字滤波法中， 当滤波参数β取值0.95时为最优参数。平滑最小值法、 HYSEP法中的SI法、 LM法和单参数数字滤波法（β=0.95）具有较高的Nash-Sutcliffe效率系数， 且平均相对误差小于10%， 说明这些方法基流估算结果稳定可靠且精度高， 均适用于疏勒河上游流域。

Applied comparison analysis of base-flow separation methods in upper Shule River basin

Base-flow is main supply source of runoff during the dry season. The reliable method of base-flow separation plays an important role in base-flow research and the management and utilization of water resources. Based on the daily runoff data from Changmabao Hydrological Station in the upper reaches of the Shule River from 2007 through 2016， in this study， using single parameter digital filtering method， minimum smoothing method， recursive digital filter technique， fixed interval method， sliding interval method and local minimum method， the base-flow is separated. The results show that the four methods can accomplish the automatic separation of base-flow. The above separated base-flow hydrograph is smooth. There are four base-flow separation methods， which have different base-flow index， ranging from 0.49 to 0.72. The results of recursive digital filtering method have the highest stability. When the filtering parameter β takes a value of 0.95， it is an optimal parameter of the single parameter digital filtering method. The minimum smoothing method， sliding interval method， local minimum method and single parameter digital filtering method （β=0.95） have high Nash-Sutcliffe efficiency coefficients with the average relative error less than 10%. The base-flow simulated results show that these methods were stable， reliable， accurately and could be used as the optimal separation methods to estimate base-flow in the study area.