包气带土壤水分入渗在水文循环中起着至关重要的作用,深入了解降雨-积水-入渗过程的相互关系对评估荒漠环境下土壤水分补给和降水再分配具有重要意义。本研究以自然沙丘的丘间低地为研究对象,通过土壤剖面入渗原位观测试验。采用Kostiakov、Green-Ampt、Philip入渗模型和Hydrus-1D、Hydrus-2D/3D水文模型模拟了土壤水分垂直入渗过程,旨在寻求土壤积水入渗可接受的方法。模拟与实测结果比较表明：综合考虑平方和误差、均方根误差等验证指标,Philip模型能够预测砂土入渗率、累积入渗量和湿润锋推进过程,并且整个湿润区,Hydrus-3D模拟土壤含水量的效果明显优于Hydrus-2D,RMSE均值和R²系数分别达到了0.018 cm³·cm^-3和0.93。综上所示,Philip与Hydrus-3D模型结合可有效地估算砂土积水入渗参数和模拟水分传输过程,进而提高土壤水分入渗研究效率。

Water infiltration of the vadose zone plays a key role in the water cycle. A better understanding of the relationships among rainfall, ponded water, and soil infiltration processes is critical to estimate groundwater replenishment and redistribution in desert environments. However, research on the moisture distribution in the vadose zone of sandy soil is still limited. In particular, few studies have examined the water transport processes in the desert vadose zone. Therefore, we aimed to determine the soil moisture distribution and variation in homogenous sandy soil under ponded infiltration and to validate models using data from these observations. This study was conducted in a nature dune slack through dynamic in situ observations using profile infiltration experiment and continuous soil moisture measurements at 1 min intervals. A single-ring infiltrometer (diameter, 20 cm; insertion depth, 5 cm) was used to keep a constant water head. Kostiakov-Lewis, Green-Ampt, Philip, Hydrus-1D/2D/3D numerical software based on Richard’s equation were used to describe the vertical infiltration and soil water movement processes and to identify suitable models for ponded infiltration processes in sandy soil. The validation indices included the sum of squared error and root mean squ are error. A comparison between the simulated and measured results indicated that the Philip model could predict the infiltration rate, cumulative infiltration, and wetting front advancement correctly, with an RMSE value of 0.0003, 0.04, and 0.24 cm·min^-1 and R ² of 0.95, 0.99, and 0.94, respectively. These values were clearly less than those of the Kostiakov-Lewis, Green-Ampt, and Hydrus-1D models, although the other models also showed good overall agreement with the field measured cumulative infiltration. The Hydrus-3D model yielded a better fit to the simulated soil moisture than the Hydrus-2D in the wetted zone based on the coefficient of determination, average RMSE (0.018 cm³·cm^-3) and R ² (0.93). Altogether, it appears that the combination of the Philip and Hydrus-3D models is a highly effective approach to estimate ponded infiltration parameters and simulate the water transport process in variably saturated sandy soil. The three-dimensional soil water diffusivity must be considered to predict water infiltration in sandy soils. Our results also demonstrate that the integration of numerical simulation and field measurements can improve the research efficiency of soil water infiltration in dry environments. In practice, the accurate prediction of soil water infiltration can be done by the selection of the proper models based on the soil properties of the particular sites.