Modeling water flow and contaminant transport in the unsaturated zone is a difficult task that relies heavily on good hydrodynamic soil characterization. This article presents a complementarity between experimentation, direct modeling and inverse modeling in order to provide a better estimate of the hydrodynamic parameters of stratified alluvial soil in the El Haouareb region of the Kairouane plain in Tunisia. A field sampling campaign was carried out. The samples collected underwent particle size analysis, bulk density measurements and infiltration tests using a mini-Muntz. In parallel, simple evaporation tests were applied to separate strata. In addition, a 2 m soil column was reconstituted and fitted with sensors to monitor water content, tension, temperature and electrical conductivity. An internal drainage test was performed on this monolith. Three methods were applied using experimental data to estimate soil hydrodynamic parameters. In the first method, pedotransfer functions were used (Rosetta platform) based on granulometric results and bulk density. In the second, water tension and water content monitored during the simple evaporation test were used to plot the soil–water retention curve (SWRC) using SWRC-Fit. In the third method, inverse modeling was applied to the internal drainage test. A comparison of the results showed that the inverse method had the lowest RMSE. Uncertainty analysis has been implemented for both the experimental and numerical set up.