Water content is an important parameter of soil impacting the physical, chemical and biological properties and processes within. Presently, most of the soil water content sensors on the market are based on electromagnetic (EM) field and dielectric permittivity theory. Water content is essentially determined by soil dielectric properties yielding accurate determinations in most cases because of the high permittivity of water (80) relative to other soil constituents. When it comes to soil texture, many EM sensors show good performance in 'non-relaxing' sand, but not so satisfying in 'relaxing' clay nor in high electrical conductivity (EC) conditions. The objectives of this study were to develop a standardized methodology for evaluating EM sensors under difficult soil conditions associated with relaxation and electrical conductivity. Considering the advantages of liquid dielectrics, which are homogeneous and stable, easy to acquire, standardized and repeatable, we employed glycerol, water, silica flour and potassium chloride mixtures to simulate high clay content soil with varied electrical conductivity. One of the goals in this case was to create dielectric properties similar to clay soils in the frequency domain. Eleven different water content sensors were tested, including time domain reflectometry (TOR), time domain transmissometry (TOT), capacitance-based, impedance-based and fixed frequency sensors. The results reveal that the response of sensors operating at low frequency (less than 100 MHz) were more dramatically altered by increasing electrical conductivity than sensors operating at relatively high frequencies (i.e., TOR and TOT), which showed greater tolerance for variation in electrical conductivity. . A method for evaluating the measurement error of both permittivity-output sensors and analog-voltage-output sensors was applied to quantify each sensor's performance.
Available at: http://works.bepress.com/scott_jones/97/