Prediction of Soil Solution Electrical Conductivity by the Permittivity Corrected Linear Model Using a Dielectric Sensor

Abstract

In the present study, the electrical conductivity of the soil solution (σp) was predicted using a linear model in which the bulk soil electrical conductivity (σb) effect on the apparent dielectric permittivity (ϵs) was considered. The performance of the proposed model was evaluated by measurements with a dielectric sensor (the WET sensor) in four porous media at four different levels of electrical conductivity of the moistening KCl solution (σw). It was found that the relationship between the square root of the permittivity (ϵs) and soil volumetric water content (θ) was dependent on soil type, which is consistent with the low operating frequency of the sensor. Establishing a soil specific θm–ϵs relationship substantially increased the θ measurement accuracy compared to the factory calibration. It was shown that the new approach for the σp prediction gave reasonably accurate results in sands irrespective of the σp values. For the finer porous media, it improved the prediction of σp only for the higher salinity levels, but the σp values appear to be underestimated. The relationship between the corrected dielectric permittivity ϵR and σb is strongly linear for σw and σb values up to 6 and 1.7 dS·m−1, respectively.