山洪预报模型历史数据影响及解决方法探析

Analysis on the Impact of Historical Data of Flash Flood Forecasting Model and its Solutions

随着山洪灾害防治工作的深入开展,山洪预报也面临着从经验性预报向实时数值精准预报转变的要求。针对山洪预报模型数据普遍存在缺失和质量不佳的现状,提出了参数移植法和参数回归法两种使用区域性代表参数搭建山洪预报模型的方法,并以海南白沙河山洪小流域为试点,采用牛尼头测站历史数据和NAM(Nedbor-Afstromnings-Model)模型进行参数的率定和验证,然后通过参数移植法搭建了整个流域水文产汇流模型,并针对模型的结构提出了改进模型精度的方法和建议。

With the development of the prevention and control of flash flood disaster,flash flood forecasting is also facing the requirement of changing from experiential forecast to real-time accurate forecast.In view of the current situation of missing and poor quality of mountain flood prediction model data,this paper proposes two methods of establishing mountain flood prediction model using regional representative parameters,namely,parameter transplantation method and parameter regression method.Taking Baishahe mountain flood watershed in Hainan as a pilot,the historical data of Niunitou station and NAM(Nedbor-Afstromnings-Model) Model were used to calibrate and verify the parameters.Then,the hydrological runoff Model of the whole basin was built by means of parameter transplantation method,and suggestions for improving the Model accuracy were put forward based on the structure of the Model.