全过程联合校正的洪水预报修正方法

The entire-process correction approach for flood forecasting

洪水实时校正是洪水预报的重要组成内容,也是水文预报研究的热点与难点。为进一步提高洪水预报精度,提出了一种全过程联合校正的洪水预报修正方法。该方法的主旨是在雨量站数目较多的流域,构建雨量站网密度与洪水预报误差分配比例之间的定量关系;对雨量站数目较少的流域,借用该比例关系,将洪水预报总误差按比例划分为面雨量输入误差和模型误差两部分;再基于系统响应理论,对研究流域的这两部分误差进行修正,从而实现输入误差与模型误差的全过程联合校正。在淮河三个典型流域的应用结果表明,相较于目前单变量的系统响应曲线修正方法,全过程联合校正方法可进一步提升误差修正效果,提高洪水预报精度。

Real-time correction is an important component of flood forecasting,and also a hot and difficult topic in hydrological forecasting research.For this purpose,an entire-process correction approach was proposed to improve the accuracy of real-time flood forecasting.Firstly,we took a basin with plentiful rain gauges.Based on rainfall data of that basin,the relationship between the density of rain gauges and the ratio of flood forecasting error was established.Then,for the basin with a small number rain gauges,ratio of the real-time flood forecasting error into areal rainfall error and model error was used to calculate the error of the flood forecasting.Finally,based on system response theory,the two forecasting errors were simultaneously corrected as to realize the entire-process correction of model input and model structure.The application results in three typical basins of Huaihe River showed that the entire-process correction approach could improve the effect of error correction and the accuracy of flood forecasting compared to the current single-variable dynamic system response curve method.