Determining Moisture Content of Soil Layers with Time Domain Reflectometry and Micromechanics

Abstract

Although the moisture condition of pavement sublayers can significantly affect pavement performance, accurate interpretation of in situ soil moisture measurements has been difficult to achieve because of the limitations of existing methods. Time domain reflectometry (TDR), originally developed to detect breaks or shorts in electrical conductors, has been used for measuring parameters related to the in situ soil moisture content. However, the apparent length method currently used to determine dielectric constant ignores other electrical properties of the conducting medium that may affect the interpretation of TDR trace to determine soil moisture. Furthermore, the existing methods for computing volumetric water content ignore the variations of dry density and determine the model parameters with assumption or regression analysis. These deficiencies can, in many cases, create a significant systematic error in the final determination of volumetric water content. To minimize these errors and improve the accuracy of moisture content estimate, a new three-step approach was proposed. The approach uses the transmission line equation to calculate the dielectric constant, conductivity, and reflectivity of a soil mixture. A micromechanics and self-consistent scheme was used to determine the volumetric moisture content and dry density on the basis of calibrated values of the solid and water dielectric constants. The system identification method was used iteratively to solve for dielectric parameters, soil moisture content, and dry density values. The validation of the new approach with ground-truth data indicated that the calculated errors were significantly less than those of existing method.