SWAT模型在淮河息县流域的参数敏感性及不确定性研究  

The Parameters Sensitivity and Uncertainty of SWAT Model and Its Application in Xixian River Basin,the Huaihe River

作　　者：杨小强[1] 石朋[1,2] 包为民[1,2] 侯爰冰[1] 周密[1] 

机构地区：[1]河海大学水文水资源学院,南京210098 [2]水文水资源与水利工程科学国家重点实验室,南京210098

出　　处：《中国农村水利水电》2013年第10期33-36,41,共5页China Rural Water and Hydropower

基　　金：国家自然科学基金重大项目资助项目(51190090/51190091);国家自然科学基金资助项目(51279057/40901015/41001011);中国国家重点实验室专项基金资助项目(2009585412/2009586412)

摘　　要：在淮河息县流域利用LH-OAT法进行SWAT模型的参数敏感性分析,得到了该地区模型参数敏感性排位,并从中选出6个敏感参数进行人工率定,大大提高了参数率定效率。模拟结果率定期(1980-1990年)月径流系列NS系数达到0.80,验证期(1991-1995年)为0.75,模型模拟效果较好。利用SUFI-2法进行参数的不确定性分析后,率定期和验证期的NS系数分别达到0.84、0.81,进一步提高了模型模拟精度和应用效果。In this paper, SWAT model is applied to the Xixian River Basin, located at the headwaters of the Huaihe River, for monthly runoff simulation. After the parameter sensitivity analysis（LH-OAT method）, the sensitive parameters in this region were got and the six most sensitive parameters for artificial calibration were selected. The Nash-Suttcliffe coefficient of calibration and validation period were 0. 80 and 0. 75, which shows that the SWAT performs well in the Xixian River Basin. Whafs more, after the un- certainty analysis（SUFb2 method）, the Nash-Suttcliffe coefficient of calibration and validation period came to 0. 84 and 0. 81, which shows that the model accuracy and application effect has been improved.

关 键 词：SWAT 淮河息县流域 参数率定 LH—OAT SUFI-2 

分 类 号：TV121[水利工程—水文学及水资源]