Comparison of unconfined and confined unsaturated hydraulic conductivity

Abstract

The field tension infiltrometer (TI) and the laboratory unit hydraulic gradient (UHG) methods are widely applied to determine the near-saturated soil hydraulic conductivity, K. Comparison between the two methods is relevant given that they differ in the explored soil volume (undetached or detached) and in the flow process (unconfined or confined). The objective of this investigation was to compare unconfined and confined measurements of unsaturated hydraulic conductivity. Twenty TI experiments were conducted in a relatively coarse-textured soil having an appreciable hysteretic behavior by using two different dry-to-wet-to-dry (DTWTD) sequences of pressure head, h0, values that differed by the highest h0 value imposed within the sequence (i.e. h0 = - 150, - 75, - 30, + 5, - 30, - 75, - 150 in site A or h0 = - 150, - 75, - 30, - 10, - 30, - 75, - 150 in site B). The same pressure head sequences were applied on twenty undisturbed soil cores, collected at the exact location of the TI measurements, to perform the laboratory UHG measurements. Regardless of the type of experiment (i.e. unconfined or confined) and the applied pressure head sequence (i.e. site A or B), higher K0 values were obtained with a drying sequence of h0 values (K0,d) than with a wetting one (K0,w)and the discrepancies between K0,w and K0,d decreased as the imposed h0 value increased, as it was expected due to hysteresis. A tendency of the UHG method to overestimate the K0 values was detected (ratios of mean K0,1D to mean K0,3D values ranging from 0.93 to 4.35), but the statistical significance of the observed differences varied with the considered sequence of pressure head values. It was concluded that both the TI and the UHG methods were effective in detecting hysteresis effects on K0, but the laboratory method resulted in K0 values that were higher and more variable probably as a consequence of a more substantial effect of macropore flow on the measured flow rates.