Performance of some infiltration models based on obtained data from double-ring and HYDRUS-1D software

Water infiltration is one of the most important properties of soil. The importance of infiltration process has led to development of several theoretical and empirical infiltration models. However, the applicability of these models is strongly subjected to the method of infiltration measurement. In this study, double ring infiltration data were collected from different regions of Iran with different soil textures. On the other hand, HYDRUS-1D model was used to simulate vertical infiltration through forward solution of the Richards’ equation. Van Genuchten-Mualem model was used to quantify the soil hydraulic properties. For this purpose, the hydraulic parameters of van Genuchten-Mualem model were optimized using inverse modeling in the HYDRUS, for each region's soil and were used for simulation. In order to evaluate the accuracy of the Horton, Kostiakov, Kostiakov-Louis and Philip models, based on measured and simulated infiltration data, mean error (ME), root mean square error (RMSE), mean absolute mean error (MAME), Pearson correlation coefficient (r) and modeling efficiency (EF) statistics were used. The results indicated that the Kostiakov-Lewis model has had the best performance in different soil textures based on measured double ring infiltration data. Horton model has had the best performance based on HYDRUS simulated infiltration data in different soil textures. The Philip model had the least efficiency in estimating cumulative infiltration based on both measured and simulated infiltration data.