Abstract

With the goal of an effective method for calibrating dielectric soil moisture sensors that can be implemented rapidly, the applicability of a two-point α-mixing model was examined for a variety of sensor types. A large number of sensor outputs (X) were taken for four different types of ECH2O soil moisture sensors under varying volumetric water contents (θ) using four test sands with various grain sizes. The two-point α-mixing model was fitted to the X–θ data and a fitting parameter, α, was determined for each sensor type. Two other widely used sensors, the CS616 and the ThetaProbe ML2x, were also tested using the calibration functions and data provided by the manufacturers, including data for a soil with high electrical conductivity. For the sensors tested in this study, the effects of sensor characteristics were lumped into α. The value of α was 1.0 to 4.6, generally with high r2 values. Some of these were as accurate as polynomial models that require a large number of X–θ data. After the α value is fitted for a specific sensor type, the two-point α-mixing model requires only the measurement of porosity and two extreme X values, i.e., sensor outputs in dry and saturated soils, and does not require soil samples with intermediate water content. Therefore, the two-point α-mixing model can be considered as an alternative method for calibrating dielectric soil moisture sensors, especially when porosity is known with small uncertainty such as relatively homogeneous field or laboratory experiments.

You do not currently have access to this article.