基于SGA和SRFR的畦灌入渗参数与糙率系数优化反演模型Ⅱ——模型应用

Optimized inverse model used to estimate Kostiakov infiltration parameters and Manning's roughness coefficient based on SGA and SRFR model:Ⅱ Application

利用作者开发的基于SGA和SRFR模型构建的畦灌土壤入渗参数和田面糙率系数优化反演模型,在不同田间灌溉实例条件下进行参数优化估值,分析评价依据地表水流推进(消退)数据估算得到的土壤入渗参数和田面糙率系数进行畦灌地表水流运动过程模拟的效果。结果表明,利用基于地表水流推进数据估算的参数可很好地模拟非稳定状态下的水流推进过程,但其模拟地表水流消退过程的效果却普遍较差,而采用基于地表水流消退数据估算的参数虽可较好地模拟稳定状态下的水流消退过程,但用其模拟水流推进过程的精度却有所下降,只有采用基于地表水流推进与消退组合数据估算的参数,才能获得较佳的综合模拟效果。与人工试算法和离散变量法估算的参数相比,基于SGA优化估值参数模拟的畦灌地表水流运动过程效果更好,SGA是改善参数估值精度的有效手段。

The optimized inverse model based on SGA and SRFR model for border irrigation developed by the authors is applied to determine the infiltration parameters and roughness coefficient according to the field observation data under different irrigation practices. The estimated parameters are used to simulate the water advance and water recession phases, and the differences between the observation and simulation for each given irrigation events are analyzed. It shows that the estimated parameters obtained only from water advance data can result in a good simulation for water advance phase under unsteady regime, while the simulation of the water recession phase is poor. On the other hand, if estimated parameters are obtained only from water recession data the simulation of water recession phase under steady regime may remarkably improved, but the precision of simulation of water advance phase is relatively decreased. For this reason, the best simulation for both water advance phase and water recession phase can be achieved if the parameters are optimized by using both water advance data and water recession data. The comparison of the results obtained from the proposed model with those based on composite algorithm of discrete variables as well as the trial and error method verifies the effectiveness of the model.