基于SWAT模型的缺资料流域径流模拟研究

Runoff simulation in data-scarce basins based on SWAT hydrological model

为实现流域水功能区水质目标管理,需要知道河道各控制断面的径流情况,然而中小河流常因水文站点缺乏而难以满足要求,为此,本文以赣江袁河流域为例,利用袁河上游芦溪、茅洲两站的实测资料构建了分布式SWAT水文模型,以月径流模拟的相对误差、相关系数和Nash-Suttcliffe效率系数3个指标为标准,对模型的敏感参数进行率定和验证.然后,利用已建立的袁河流域分布式SWAT模型模拟分析了下游无资料地区的径流数据.结果表明:以流域内有资料的区域为基础构建分布式SWAT分布式水文模型推求中下游无资料地区径流过程的方法具有物理成因、能够充分考虑流域的水循环特征,是一种缺资料流域推求径流比较实用的方法.

It is necessary to know the runoff conditions in controlling reaches of a river in order to carry out the water quality management in different water function zones of the river basin. However, hydrological gauge stations are usually scarcely distributed or ever lack the stations in a river basin. For the prediction of runoff in a data scarce watershed, a distributed hydrological model of SWAT is established in a case study of Yuanhe River, a tributary of Guanjiang River which belongs to the Yangtze River. During the case study, the SWAT model is constructed by using the data from the Luxi and Maozhou hydrological stations located at the upper to middle reaches of the Yuanhe River. The model calibration and verification are made according to criteria of relativity error, correlation coefficient and Nash-Suttcliffe effective coefficient. The accuracy of the calibration and verification meet the criteria requirements quite well. Runoffprediction in the downstream reaches without historical observation data is simulated. Results indicate that this method can fully considerate the characteristics of water cycle in the watershed because the distributed hydrological SWAT model has the physical mechanism of a river basin. Therefore, this is an applicable method for the prediction of runoff in data scarce watersheds.