用新安江模型和PBK模型对伊河上游东湾站洪水进行了模拟预报,然后分别用先集合预报后实时校正的方法和先实时校正后集合预报的方法对预报结果进行修正,以洪峰相对误差、洪量相对误差和确定性系数作为评价指标分析了集合预报和实时校正后的预报精度.结果表明:利用加权平均法集合预报和KNN法实时校正均能有效减小洪水预报误差,先集合预报后实时校正方法的洪峰预报精度优于先实时校正后集合预报方法,而洪量预报精度不如先实时校正后集合预报法.
The Xinr anjiang model and PBK model were applied to forecast the flood of Dongwan Station in the upper Yihe River. The methodsof ensemble forecasting-real-time correction and real-time correction-ensemble forecasting were separately used to correct the forecastingresults. The simulated results were analyzed by the evaluation indices of the flood peak relative error,flood volume relative error and the certaintycoefficient. The results show that the forecasting error can be efficiently reduced by using ensemble forecasting and real-time correctionmethod. The ensemble forecasting-real-time correction method is superior to real-time correction-ensemble forecasting method in flood peakforecast,and it is inferior to real-time correction-ensemble forecasting method in flood volume forecast.