عنوان مقاله [English]
Soil infiltration models play an important role in the design and evaluation of surface irrigation systems. The purpose of this study was to evaluate, compare and calibrate water infiltration models in some areas of western Iran including Davood Rashid and Honam in Lorestan province and Kalat in Ilam province. Influence models including Kostiakov model, Modified Kostiakov model, Novel, Horton, SCS and Philip’s models were investigated. To evaluate the accuracy of the models, the calibration indices including RMSE, MAE, MR, PE and R2 were used. Model calibration was performed using Solver Excel software. Evaluation indices showed that Novel infiltration model with mean values of RMSE, MAE, MR, PE and R2 had 0.0221, 0.0178, 0.999, 0.13 and 0.9785 respectively. It performs better than the other five models in the whole study area. On the opposite side of the SCS model with mean values of RMSE, MAE, MR, PE and R2 were 0.07, 0.0575, 1.07, 7.86 and 0.8021, respectively. It was the most inappropriate performance. Whereas the analysis of variance of the models showed no significant difference between the calculated and measured values. In the regional study of these models, the Novell model was also chosen as the most suitable model for all three regions of Davoodershad, Honam and Kalat. Evaluation of model performance in different tissues indicates less sensitivity of modified Horton, Novel, Kostiakov and modified Kostiakov models to soil texture changes and better SCS model performance in heavy tissues and better model performance Philip was in lighter-textured textures. Therefore, Novell model can be used to estimate infiltration under different soil texture conditions as well as under similar conditions in the study area.