SWAT模型水文过程参数区域差异研究

探讨SWAT模型的敏感参数和参数值的区域差异,有助于减少水文过程模拟的不确定性,提高模型在不同区域的应用水平。论文在收集SWAT模型的国内应用文献基础上,依据干燥指数将流域划分为湿润区、半湿润区、半干旱区和干旱区,主要采用定性分析方法,分析模型的敏感参数和参数值的空间差异及其影响因素。结果表明,SWAT模型的敏感参数和参数值具有一定的区域差异。对于地表水文过程,径流曲线系数(CN2)和土壤可利用水量(SOL_AWC)在各个区域都比较敏感,土壤蒸发补偿系数(ESCO)在湿润区和半湿润区较为敏感;由于均受土壤质地的影响,湿润区CN2 值高于半湿润区,SOL_AWC则正好相反;受气温和植被覆盖的影响,湿润区和干旱区的ESCO取值高于半湿润区和半干旱区。地下水参数在湿润区和干旱区较为敏感。降雪融雪参数仅在干旱区融雪对径流贡献大的流域才比较敏感。由于气候、土壤、植被条件的区域异质性,导致了流域水文过程参数的区域差异。结果还表明,由于水文过程还受到局部气象条件、下垫面等多种因子共同作用,以及模型尚未揭示的其它因子的影响,模型参数在相同区域的不同流域也存在着一定的差异。

A Study on Regional Difference of Hydrological Parameters of SWAT Model

Discussing the parameter sensitivity in different regions and their regional difference for SWAT model would be helpful for reducing the uncertainty of hydrological process simulation and improving the application level in different regions. Based on the domestic articles related to application of SWAT model, firstly, different regions including humid, semi-humid, semi-arid, and arid regions were divided according to aridity index. Then the variation of sensitivity and values of hydrological parameters in different regions were investigated and its influencing factors were analyzed qualitatively for SWAT model. The results indicate that the parameters sensitivity and parameter value of SWAT model have some regional difference. During surface water process, runoff curve coefficient (CN2) and soil water availability (SOL_AWC) are sensitive in all regions, soil evaporation compensation coefficient (ESCO) is more sensitive in humid and semi-humid regions. Because influenced by soil texture, the value of CN2 is much bigger in humid region than in semi-humid region, and that of SOL_AWC is opposite. Because influenced by temperature and vegetation cover, the value of ESCO is much bigger in humid and arid regions. The groundwater parameters are more sensitive in humid and arid regions. Snow parameters only show sensitivity in arid region where melt water accounts for very great proportion of runoff. Moreover, the main reason for the difference of parameters is due to regional heterogeneities of climate, soil and vegetation. That spatial heterogeneities of several parameters sensitivity and values in the same region also exist, because the hydrological process is also affected by local meteorological conditions and underlying surface, and other factors of model which have not been revealed.