Soil Texture From Tdr Waveform Analysis

Abstract

The dispersive electromagnetic (EM) behavior of soils is strongly related to the mineralogy, soil structure, and pore fluid characteristics. However, time domain reflectometry (TDR) measurements (TDR waveforms) are predominantly used for soil water content and dry density estimation. These two parameters are calculated based on empirical equations that relate them to the soil dielectric permittivity (Ka) and the bulk electric conductivity (ECb). Ka and ECb are obtained in the time domain from few data points of the TDR waveform, disregarding most of the acquired data (usually 2048 data points) that reflects the EM response of the material over a broad frequency range. The complexity of the soil-water interaction in the presence of a time-varying EM field, and the presence of a non-transverse propagation mode in the TDR system limit the characterization of soils by dielectric spectroscopy. This paper presents a semi-empirical method for soil texture identification based on an integrated numerical and experimental analysis of the effect of the EM soil dispersive behavior on TDR waveforms. Evaluation of TDR tests conducted at 20°C on sands, silts, and clays using tap water at different water contents and dry densities shows that a simple time-domain signal processing of the first reflection from the probe section captures the effects of the EM soil-water interaction. Considering that the coefficients of the TDR empirical equations for soil water content and dry density estimation are soil-type dependent, the developed method allows self-calibrating the TDR system. The result of this work provides the basis for making the TDR technique a tool not only for water content and dry density estimation, but also for soil characterization.