Temperature-corrected calibration of GS3 and TEROS-12 soil water content sensors 

Abstract

The continuous monitoring of soil water content is increasingly carried out using low-cost capacitance sensors that require a site-specific calibration to correlate sensor readings to apparent dielectric bulk permittivity (Kb) and soil water content (θ). Factory calibrations do not take into account the temperature effects on the bound water fraction associated with clay mineral surfaces, making it challenging to convert Kb to θ especially in fine-textured soils. Here, we present a new multi-point calibration approach that accounts for temperature effects in two soils with medium to high clay content. For this purpose, a laboratory calibration strategy for capacitance sensors was developed using repacked soil samples in which the Kb-θ relationship was determined for temperature steps from 10 to 40 °C. We tested this approach using the GS3 and TEROS-12 sensors (METER Group, Inc. USA; formerly Decagon Devices), which measure Kb and soil temperature (T). Direct measurements of θ were obtained using the thermo-gravimetric method. We found that Kb has an impact on T for both the loam and clay soils with contrasting T-Kb relationships. Despite the identical measuring principle, GS3 and TEROS-12 showed distinct T-Kb relationships, indicating that sensor- and site-specific calibrations are needed. The measured data were fitted using a linear function θ = a √Kb+b with temperature-dependent coefficients a and b. We found that the slope, a(T), and intercept, b(T), of the loam soil were different from those of the clay soil. By including a temperature correction in this equation resulted in a low RMSE, ranging from 0.0070 to 0.033 cm3 cm-3, which was considerably lower than the RMSE obtained from factory calibration (0.046 to 0.11 cm3 cm-3). Finally, we demonstrated how the proposed calibration method can be applied at two Mediterranean test sites.  The new calibration method, unlike the factory calibration, led to a significant improvement in the agreement between θ-values from a network of GS3 sensors, installed in these test sites, and independent θ-values from Cosmic Ray Neutron Sensors (CRNS).