基于流溪河模型的中小河流洪水预报方法

针对中小河流缺乏河道断面资料而无法实际应用分布式水文模型的问题,以流溪河模型为原型,假定河道断面为矩形,采用卫星遥感影像估算河道断面尺寸,通过实测洪水资料优选模型参数,提出了中小河流洪水预报的流溪河改进模型。应用于江西省典型中小河流——太平江流域的结果表明:河道分级越多,模型模拟的洪水的洪峰流量越大、次洪径流系数越大;洪峰出现时间越提前,模拟的洪水过程的效果越好,精度越高;改进的流溪河模型对50场洪水模拟的相关系数达0.92,洪峰误差均值为5.54%,最大不超过20%,平均峰现时间差为0 h;洪水过程的预报结果可评定为甲等。该模型可应用于中小河流实时洪水预报中。

Liuxihe Model-based method for flood forecast of medium and small-sized river

Aiming at the problem that the distributed hydrological model cannot be applied to medium and small-sized river due to lack of the relevant data of river channel cross-section and taking Liuxihe Model as the prototype， the cross-section size of river channel is estimated with the satellite remote sensing image concerned under the assumption of that the cross-section is rectangular， and then a revised Liuxihe Model for flood forecast of medium and small-sized river is proposed herein by optimizing the model parameters through the measured flood data. The result from the application of this model to a typical medium and small-sized river—Taipingjiang River Watershed in Jiangxi Province shows that the more the stages of the river channel are， the larger the peak flood flow simulated by the model is to be and the larger the runoff coefficient of individual flood is to be， while the earlier the flood peak occurs， the better the simulating effect of the flood process is to be and the higher the accuracy is to be as well. The correlation coefficient from the simulations made by the model on 50 flood events reaches 0.92 with the mean error of the peak flood of 5.54% and the maximum under 20%， while the flood peak appearance time is 0 h. The result of the flood process forecast can be judged to be grade A. This model can be applied to the real-time flood forecast of medium and small-sized river.