水文模型在山洪模拟中的比较应用

Comparison and Application of Hydrological Models in Mountain Flood Simulation

摘要: 山洪灾害频繁发生，给社会造成了重大经济损失，因此水文模拟对山洪灾害防治具有重要现实意义。为研究不同水文模型在山洪预报中的模拟效果，以官山河流域为例，使用新安江模型、TOPMODEL模型和改进的SCS模型等三种模型对多场场次洪水进行模拟研究。使用1973~1987年的逐日逐时降雨径流资料对模型参数率定，取其中15场典型洪水进行模型验证。从三种模型模拟效果来看，模拟效果一般。TOPMODEL模型对洪峰流量拟合得较好，平均相对误差为23.14%，合格率为40%；改进的SCS模型与TOPMODEL模型对峰现时间模拟较好，合格率均为93%；改进的SCS模型对洪量模拟效果最佳，相对误差为20.73%，合格率为67%。TOPMODEL半分布式水文模型能从物理层面更为真实地描述自然界的洪峰流量响应过程，改进的SCS模型提高了地下径流的贡献而使得洪量模拟效果相对较好。

Abstract: Mountain flood disasters occur frequently, causing major economic losses to society. Therefore, hydrological simulation has important practical significance for the prevention and control of mountain flood disasters. In order to study the simulation consequence of different hydrological models in mountain flood forecasting, taking the Guanshan river basin as an example, Xin’anjiang model, TOPMODEL model and improved SCS model were used to simulate multi-field floods. The model parameters were determined using the time by day rainfall runoff data from 1973 to 1987, and 15 typical floods were used for model verification. From the simulation results of the three models, the simulation effects are general. The TOPMODEL model fitted the peak flow better. The average relative error is 23.14%, and the pass rate is 40%. The improved SCS model and the TOPMODEL model have better peak-to-peak time simulation, and the pass rate is 93%; the improved SCS model’s simulation effect on the flood amount is the best, the relative error is 20.73%, and the pass rate is 67%. The TOPMODEL semi-distributed hydrological model can describe the flood peak flow response process in the natural world from the physical level more realistically. The improved SCS model improves the contribution of underground runoff and makes the flood simulation effect relatively better.