Comparison of methods for determining unsaturated hydraulic conductivity in the wet range to evaluate the sensitivity of wetting front detectors

Abstract

The design of passive lysimeters or wetting front detectors determines the tensions at which they collect a water sample from an unsaturated soil. When deployed in the field to help manage irrigation, it is necessary to know the minimum flux of water that can be sampled by a passive lysimeter and how this relates to the drainage flux at field capacity. This requires a good estimate of the unsaturated hydraulic conductivity characteristic, K(h), in the wet range (< 10 kPa). We compared various field, laboratory and theoretical approaches for obtaining the K(h) function and compared these to a reference K(h) function derived by applying inverse modelling approaches to field drainage experimental data. The Van Genuchten model and three of the pedotransfer models produced K(h) functions with a root mean square error of less than 5% compared to the reference, and appear to be simple methods of obtaining a reasonable estimate of unsaturated hydraulic conductivity. However, despite the goodness of fit, there can be a 10-fold difference in conductivity at a given tension < 10 kPa estimated from the different methods. Moreover, water content at field capacity depends entirely on whether field capacity is defined as time elapsed after saturation, a set tension or a minimum flux.