Determination of moisture content in a deformable soil using time‐domain reflectometry (TDR)

Summary Time‐domain reflectometry (TDR) is being used increasingly for measuring the moisture content of porous media. However, successful application for measuring water in soil has been limited to non‐deformable soils, and it would be a valuable extension of the technique if it could be used for soils that shrink on drying. We have recently investigated its application to soils rich in clay and organic matter and peats. Here we propose a method for determining moisture content in deformable soils based on the relation between the dielectric constant, K , and the volumetric moisture content, Θ, measured by TDR. Parallel TDR probes with a length of 15 cm and a spacing of 2 cm were placed horizontally in soil cores with a diameter of 20 cm and height of 10 cm taken from a forest. The soil is very porous with large proportions of both silt and clay. The sample weight and travel time of the electromagnetic wave guided by parallel TDR probes were simultaneously measured as a function of time, from saturation to oven‐dryness during which the core samples shrank considerably. Vertical and horizontal components of shrinkage were also measured to take the air‐exposed region of TDR probe into account in the determination of K . The effect of deformation on volumetric moisture content was formulated for two different expressions, namely actual volumetric moisture content (AVMC) and fictitious (uncorrected) volumetric moisture content (FVMC). The effects of air‐exposure and expressions of volumetric moisture content on the relation between K and Θ were examined by fitting the observations with a third‐order polynomial. Neglecting the travel time in the air‐exposed part or use of the FVMC underestimated the Θ for a given K . The difference was more pronounced between AVMC and FVMC than between two different dielectric constants, i.e. accounting for air‐exposure, K ac , and not accounting for air‐exposure, K au . When the existing empirical models were compared with the fitted results, most underestimated the relation based on the AVMC. This indicates that published empirical models do not reflect the effect of deformation on the determination of Θ in our forest soil. Correct use of the Θ expression has more impact on determining moisture content of a deformable soil than the accommodation of travel time through the air‐exposed region of TDR probe.