Abstract
Groundwater recharge is strongly influenced by the infiltration process. In this research, the Philip, Horton, Kostiakov, and Green–Ampt infiltration models were tested for the ability to describe the infiltration process in the ephemeral stream beds located in Al Madinah Al Munawarah Province in Saudi Arabia. Infiltration data were obtained from double-ring infiltrometer tests in 14 locations distributed over the province. The method of least squares through an objective function optimization formalism is utilized to estimate the parameters of each model. The results show high variability in the parameters of each model over the tests. Individual tests showed that some models were better for representing specific tests than other models. On average, the Kostiakov empirical model was the best at describing the 14 infiltration tests with only 2 empirical parameters, since it had the minimum root mean square error (RMSE) for the cumulative infiltration depth F (1.13 cm), and it also had the same RMSE for the infiltration rates f (0.1 cm/min), similar to other models. Moreover, the Kostiakov model had an acceptable correlation coefficient R = 0.61 for f, and R = 0.99 for F. The results imply significant variability in the groundwater recharge rates from flash floods in the region.