SWAT模型灵敏性分析模块在云州水库流域的应用

SWAT模型是基于水文过程的,具有很强物理机制的流域分布式水文模型.介绍了AVSWAT2005灵敏性分析模块,并在华北土石山区云州水库流域得到较好的应用与验证.通过灵敏性分析,辨析出影响该流域产流模拟结果精度的主要参数因子.SCS径流曲线系数(CN2)、土坡可利用水量(SOL-AWC)、土壤饱和导水率(SOL_K)的影响是显著的,灵敏性等级高;平均坡度(SLOPE),浅层地下水径流系数(GWQMN)有一定影响,灵敏性等级中;依据上述结论调整参数值,采用连续10 a的实测月流量数据对模型进行了校准和验证.结果表明,模型对月流量模拟的相对误差在模型校准期和验证期均小于15%,Nash-Suttcliffe系数Ens高于0.7,SWAT模型对云州水库流域产流的模拟结果良好.因此,灵敏性分析模块的应用极大地提高了模型在华北土石山区的可用性,具有一定的推广意义.

Application of Sensitivity Analysis Module of SWAT Model in Yunzhou Reservoir Basin

The soil and water assessment tool (SWA T) is a physically-based continuous event hydrologic model. This paper mainly introduces the sensitivity analysis model of the AVSWA T2005 and its verification in Yunzhou Reservoir basin. Based on the sensitivity analysis, the main parameters affecting the simulated values in the watershed are determined. Results show that the CN2, SOL_AWC, and SOL_ K parameters have the most significant influence on the simulated values of runoff from Yunzhou water shed, the sensitivity being in a high level ; and the SLOPE and GWQMN have the more significant influence on runoff, the sensitivity being in a general level . The parameters are adjusted based on the above analyzed results and the model is verified using over 10 years data of monthly runoff . Satisfied simulation results are obtained with the relative error under 15 % and Nash-suttcliffes coefficient (Ens) more t han 0. 7. The simulation result of SWA T model for Yunzhou Reservoir basin is good. Therefore, the application of sensitivity analysis module of SWAT model in t he soil-gravel area of North China may improve the accuracy of simulation dramatically.