Estimation of soil hydraulic parameters using double-ring infiltrometer data via inverse method

In this study, HYDRUS2D/3D software was used to estimate the hydraulic parameters of van Genuchten-Mualem model via inverse modeling using double-ring infiltrometers data in 3 different soil textures. Nine scenarios of inverse modeling (divided in three groups) were considered with different number (5, 4 and 3) of fitted hydraulic parameters for optimization. In the first group, simulation was carried out solely using cumulative infiltration data. In the second group, cumulative infiltration data plus water content at h = −330 cm (i.e. field capacity, FC) were used as inputs. In the third group, cumulative infiltration data plus water contents at h = −330 cm (FC) and h = −15000 cm (i.e. permanent wilting point, PWP) were used simultaneously as predictors. The results indicated that by reducing the number of hydraulic parameters involved in the optimization process, simulation error is reduced and the prediction accuracy of other soil hydraulic parameters would be increased. Including FC as an additional data was important to better optimize/define soil hydraulic functions. So using of (Saturated hydraulic conductivity) Ks, (Shape parameter of soil water characteristic curve) n and (the parameter that inversely related to the air entry value) a as predictor parameters an FC as an additional data was the best scenario. RMSE(cm3)¡ NRMSE¡ AIC¡ and R2 were respectively 1259, 528.2, 0.0081 and 0.9999 in Sandy Loam soil, 242, 79.0, 0.0059 and 0.9988 in Loamy soil and 298, 153.6, 0.0174 and 0.9983 in Silty Clay soil. Using PWP as additional data, increased the simulation error in all 3 soil textures.