中文摘要：

为实现饱和、非饱和带水分运动耦合模型参数的有效率定,以内蒙古河套灌区永联试验区的地下水位、土壤含水率实测数据为基础,选用MODFLOW-HYDRUS耦合模型,采用试错法与PEST优化算法相结合的方法,对水文地质参数和土壤水力参数进行优化。优化结果表明：通过PEST算法进行参数优化后,耦合模型模拟精度有所提高,决定系数提高了2%~35%,纳什系数提高了6%~11.3倍,均方根误差减小了5%~20%,标准化均方根误差减小了5%~20%。其中验证期土壤含水率的各项评价指标提高或减小的幅度均为最大,决定系数、纳什系数、均方根误差和标准化均方根误差变化幅度分别为35%、11.3倍、20%和20%。

英文摘要：

To optimize parameters for saturated-unsaturated coupling models efficiently, this research selected Yonglian Irrigation District, Inner Mongolia, China as study area, and combined with trial-and-error method and PEST, an automatic optimization algorithm, to optimize hydrogeological parameters and soil hydraulic parameters with measured data of water table and soil moisture for MODFLOW-HYDRUS in which the soil water movement was descried as one-dimensional vertical movement and groundwater flow as three--dimensional movement. The results show that simulation accuracy of MODFLOW-HYDRUS after parameter optimization was improved which was reflected in 2% - 35% increases of determination coefficient, 6%- 11.3 times increases of Nash coefficient, 5%- 20% decreases of root mean square error and 5%c 20% decreases of normalized root mean square error. The determination coefficient and the Nash coefficient of soil moisture in validation period had the largest increases which were 35% and 11.3 times respectively; besides, the root mean square error and normalized root mean square error of soil moisture in validation period had the largest decreases which were 20% and 20% respectively.