Correction of Temperature and Electrical Conductivity Effects on Dielectric Permittivity Measurements with ECHO Sensors

The measurement accuracy of low-cost electromagnetic soil water content sensors is often deteriorated by temperature and soil bulk electrical conductivity effects. This study aimed to quantify these effects for the ECH2O EC-5 and 5TE sensors and to derive and test correction functions. In a first experiment, the temperature of eight reference liquids with permittivity ranging from 7 to 42 was varied from 5 to 40 degrees C. Both sensor types showed an underestimation of permittivity for low temperature (5-25 degrees C) and an overestimation for high temperature (25-40 degrees C). Next, NaCl was added to increase the conductivity of the reference liquids (up to similar to 2.5 dS m(-1) for a permittivity of 26-42, up to similar to 1.5 dS m(-1) for a permittivity of 22-26). The permittivity measured with both sensors showed a strong and complicated dependence on electrical conductivity, with both under-and overestimation of permittivity. Using these experimental data, we derived empirical correction functions. The performance of the correction functions for the 5TE sensor was evaluated using coarse sand and silty clay loam soil samples. After correcting for temperature effects, the measured permittivity corresponded well with theoretical predictions from a dielectric mixing model for soil with low electrical conductivity. The conductivity correction function also improved the accuracy of the soil moisture measurements, but only within the validity range of this function. Finally, both temperature and electrical conductivity of the silty clay loam were varied and a sequential application of both correction functions also resulted in permittivity measurements that corresponded well with model predictions.