Assessing the performance of various infiltration models to improve water management practices

Abstract Infiltration plays a key role in stormwater management and irrigation scheduling. A review of previous studies reveals that the effectiveness of infiltration models varies significantly depending on soil characteristics and field conditions. Accurate predictions depend on selecting appropriate models for specific sites because of soil spatial variability. This requires extensive testing and recording of infiltration rates at each location. This study assesses various infiltration rate measurement models to enhance water management efficiency. Infiltration rate measurements were conducted at three sites in Dehradun using a double-ring infiltrometer. Well-established models, such as Philips JR, Green, Ampt, Horton, Kostiakov, modified Kostiakov, and the Soil Conservation Service (SCS) model, were evaluated. Data from infiltration tests were used to calibrate these models, facilitating better irrigation system design and stormwater management. In assessing their effectiveness and efficiency, key evaluation criteria such as Nash–Sutcliffe efficiency (NSE), R -squared ( R 2 ), root mean square error (RMSE), mean absolute error (MAE), and mean bias error (MBE) were employed. Our findings highlight the superiority of the Philips JR model, offering the highest overall accuracy with the highest average value R 2 = 0.9557 and NSE = 0.9553, lowest MAE = 0.6717 cm/h, MBE = − 0.0160 cm/h and RMSE = 1.0077 cm/h. These results underscore the model’s ability to synthesize infiltration data effectively, even in the absence of direct measurements. This insight positions the Philips JR model as a valuable tool for estimating infiltration rates in similar conditions.