LTPP Computed Parameter: Moisture Content

Chapter 1. Introduction

Time domain reflectometry (TDR) information has been collected as a key component of the Long Term Pavement Performance (LTPP) program's seasonal monitoring program (SMP) to monitor subsurface moisture conditions in pavement structures. The TDR waveform data do not provide in situ moisture contents directly. Rather, the data must be analyzed to determine parameters-such as volumetric moisture content (VMC)-that are of use in pavement design and performance prediction.

Interpretation that included the development of empirical-based methodologies to convert waveform characteristics and in situ soil properties to moisture parameters was performed on a portion of these TDR data under previous analysis.^([1]) The computed parameter data from this effort are currently available in the LTPP Pavement Performance database. Since then, approximately 175,000 more automated TDR measurements have been added to the database but have not been interpreted. The current study was performed to not only compute the moisture parameters for these additional TDR measurements but also to assess other feasible computational procedures.

The objectives of the current study were to:

• Investigate the differences between the automated and manual TDR trace interpretation methods.
• Investigate the adequacy of the models used to estimate VMC from the TDR interpreted dielectric constant.
• Investigate the adequacy of data used to compute gravimetric moisture content from VMC.
• Determine adequacy of data to support computation of other moisture-related indices such as degree of saturation.

A significant portion of the analysis was focused on alternative computational processes, wherein previously uninterpretable traces could be utilized. Based on this investigation, a new approach was recommended that computed the soil dielectric constant using a solution of the transmission line equation (TLE) for each TDR trace and the computation of the dry density and the moisture content using a micromechanics model. The new approach eliminated many of the issues related to the trace interpretations and provided a relatively accurate assessment of the in situ moisture content.

In phase 1 of this project, the basic procedures of the new approach were developed and evaluated using measured moisture contents from the SMP Installation Reports and other sources.^([2] The new approach was shown to work and to produce reasonable estimates of the in situ moisture contents compared with ground truth measurements in both field and laboratory settings. Phase 2 of the project entailed the development of a new computer program to automate the computation process, the computation of the moisture content for 274,000 TDR traces, and the uploading of this data into the LTPP database. Quality control (QC) checks were developed and performed on all of the computed data.

This report presents the development of the new computational procedures, evaluation of results from the new approach, development of the computer program automating the process, and the QC initiatives implemented to ensure data reported in the LTPP database are of research quality.