An improved reflectometric method for soil moisture measurement exploiting an innovative triple-short calibration

Time-domain reflectometry (TDR) techniques have become increasingly attractive for soil moisture evaluation thanks to their adaptability, low cost, and measurement accuracy. Many different TDR-based approaches are currently available for deriving moisture content: empirical calibration curve, empirical dielectric models, frequency-dependent dielectric mixing models, etc. Generally, TDR soil moisture measurements resort to multiple-rod probes, which provide good adhesion to the soil and ease of insertion. However, the typical configuration of such probes does not allow performing a short-open-load (SOL) calibration procedure, which is definitely necessary for retrieving accurate dielectric characterization from TDR waveforms. To overcome the impracticability of the traditional SOL calibration, in this paper, an innovative triple-short calibration (TSC) procedure for commercially available three-rod probes is proposed. First, the robustness of the TSC procedure is validated on well-referenced liquids, demonstrating that its application to TDR measurements leads to a substantial enhancement of the final accuracy in the evaluation of the frequency-dependent reflection coefficient. Successively, the TSC procedure is applied to TDR measurements performed on moistened sand samples. In particular, it is demonstrated that the proposed TSC method, applied to traditional TDR measurements in combination with a dielectric mixing model, leads to an accurate and effective moisture evaluation procedure. This goal is reached through an efficient optimization algorithm that minimizes the difference between experimental and theoretical reflection coefficients. Results show that the proposed strategy is a suitable candidate for low-cost, highly accurate, and easy-to-perform moisture content estimation of soils. © 2010 IEEE.