Comparison of alternative methods for deriving hydraulic properties and scaling factors from single‐disc tension infiltrometer measurements

Abstract

Analysis of single‐disc tension infiltrometer data is commonly based on the interpretation of the steady state infiltration rate derived from large time measurements. In this study, three analytical models describing the cumulative infiltration curve were used to estimate the sorptivity and the steady state infiltration rates. Hydraulic conductivity was estimated from (1) two steady state infiltration rates, (2) a sorptivity value and a steady state infiltration rate, and (3) two sorptivity estimates. In total, seven different methods were used to calculate the hydraulic conductivity. We compared the spatial variability of sorptivity, steady state infiltration rates, unsaturated hydraulic conductivity, and scaling factors of these seven methods. Cumulative infiltration curves were measured at 50 locations and three pressure heads along a 40 m long transect on a silty loam Eutric Regosol. In a first step the amount of water and the time needed to wet the contact sand under the disc was successfully filtered from the raw data using nonlinear regression techniques. The analytical models described the infiltration curves very well but gave different estimates of the sorptivities and steady state infiltration rates. Method 2 using both sorptivity and steady state infiltration rate resulted in the largest estimates of the mean unsaturated hydraulic conductivity, whereas the other two methods (1 and 3) resulted in similar mean values. Linear scaling analysis described well the observed variability in the hydraulic conductivity. Overall, different methods produced quite different unsaturated hydraulic conductivity estimates at specific locations, while their field mean or scaled properties were better comparable. Results of this study may prove to be important for deciding appropriate disc infiltrometer data analysis procedure while addressing water flow and chemical transport behavior at different spatial scales.