流溪河模型是一个主要用于流域洪水预报的分布式物理水文模型,包括流域划分、蒸散发计算、产流计算、汇流计算、参数推求5个模块。针对流溪河流域上游的流溪河水库流域建立了空间分辨率为100m的流溪河模型,对流域进行了单元划分,估算了河道断面尺寸,依据1场实测洪水推求出了模型参数,对实测的12场洪水,采用推求的模型参数发进行了洪水模拟,取得了较好的模拟效果。将研究流域划分成了52853个单元流域,进行一个时段的洪水模拟计算,在普通桌面PC机上平均只需要12s。研究结果表明,饱和含水率、土壤层厚度和河道糙率为高度敏感的模型参数,田间持水率、土壤特性参数b、饱和水力传导率、边坡糙率为模型的敏感参数,凋萎含水量、蒸发系数、潜在蒸发率和地下径流消退系数是模型的不敏感参数。
The Liuxihe Model is a physically based distributed hydrological model for catchment flood forecast which consists of 5 components, including catchment dividing, evaporation, runoff production, runoff routing and parameter deriving. The Liuxihe Model was set up in the Liuxihe Reservoir basin of upstream Liuxihe catchment at a spatial resolution of 100 m. The grid type was categorized and the river channel cross-section size was estimated, and also the model parameters were derived referring to the sim- ulation results from an observed flood event. Twelve observed flood events were simulated with the de- rived model parameters and reasonable results were acquired. The studied basin was divided into 52853 grids. It only took 12 seconds to simulate one stage flood discharge on a general desktop PC. The results show that water content at saturation, soil thickness and channel Manning's coefficient are highly sensitive model parameters, water content at field condition, soil porosity parameter b, hydraulic conductivity and hillslope Manning's coefficient are sensitive model parameters, while water content at wilting condition, potential evaportranspiration, evaportranspiration coefficient and underground water recession coefficient are insensitive model parameters.