基于抗差估计方法的流域水文模型的不确定性分析

Uncertainty Analysis of Xinanjiang Hydrological Model Based on Robust Estimation

流域水文模型的不确定性主要来自水文气象资料的不确定性、水文模型结构的不确定性及模型参数的不确定性等。以危水水库流域中的乌溪沟以上流域为研究对象,采用抗差估计方法探讨了由于雨量资料误差引起新安江模型的不确定性。研究结果表明:误差的大小,发生的站数,对模型的不确定性均有影响。并且误差越大,发生站数越多,对模型不确定性影响越大。采用雨量观测误差三步修正方法先对雨量资料进行处理,然后采用新安江模型进行计算,采用该方法可以有效地抗御雨量资料误差,减少模型的不确定性,提高洪水预报精度。

The uncertainty of hydrological model mainly results from the uncertainty of hydrological and meteorological input data,uncertainty of the structure of the model and uncertainty of model parameters.The robust estimation method is applied to the uncertainty analysis of Xinanjiang hydrological model,with the Wuxigou basin,a control input gauged station of Weishui Reservoir basin,as the research object.The results show that the magnitude of error and the occurring numbers of the station will influence the uncertainty of the model.The error is larger and the occurring numbers of station are more,then the influence to the model uncertainty is bigger.Three-stepwise correction is proposed to calculate the precipitation data first and then the Xinanjiang model is applied to simulate the discharge of the outlet.The results demonstrate that the correction method can prevent the error of precipitation data and reduce the uncertainty of the model,therefore improve the accuracy of flood forecasting.