Determination of soil bulk density with thermo-time domain reflectometry sensors

Surface soil bulk density varies with time and location. In most cases, destructive soil sampling has been used to monitor soil bulk density, rho(b). Recently the thermo-time domain reflectometry (TDR) technique was shown in principle to be able to estimate rho(b). Thermo-TDR sensors can measure soil volumetric heat capacity (rho c) and soil water content (theta). Since rho c depends on theta and rho(b), measurements of rho c and theta can be used to estimate rho(b). Previous studies indicated, however, that there were large deviations between thermo-TDR estimates and gravimetric measures of rho(b). Deviations were attributed mainly to the change of thermo-TDR needle-to-needle spacing during sensor insertion into the soil. The objective of this study was to improve the thermo-TDR sensor design to improve estimation of rho(b). The ability of three new prototype thermo-TDR sensors, along with an existing thermo-TDR sensor, to estimate rho(b) was investigated. Evaluation results indicated that a three-needle sensor design with 2-mm needle diameter, 40-mm needle length, and 8-mm needle-to-needle spacing provided the most accurate estimation of soil rho(b). For this sensor, the RMSEs of rho(b) estimates compared with gravimetric measures of rho(b) in laboratory evaluations were 0.055, 0.051, and 0.046 Mg m(-3) for a silt loam, a clay loam, and a sand soil, respectively, and was 0.095 Mg m(-3) in a field evaluation. In terms of relative error, this specific design was generally within 5% under laboratory conditions and within 10% under field conditions. The improved thermo-TDR sensor provides an opportunity for obtaining accurate, nondestructive, repeated estimates of soil rho(b).