LTPP Computed Parameter: Moisture Content

Chapter 8. Recommendations for Future Research

This section provides recommendations on future research needed to improve and/or further the TLE/micromechanics approach to estimating soil parameters. These recommendations would provide complete validation of the process and improve the overall accuracy while reducing variability in the estimates.

Investigate the effects of soil suction on the composite dielectric constant

As discussed in chapter 5 of this report, a shift was observed in the relationship between composite dielectric constant and VMC. This shift typically occurred at dielectric constant values below five. Upon investigation, the research team believed that this phenomenon was caused by soil suction influences. However, the data needed to fully examine this issue is not currently available.

It is recommended that additional research be conducted to more fully investigate the effects of soil suction on the composite dielectric constant as it is used to compute moisture content using the micromechanics models used in this project. This would require laboratory evaluation of TDR traces in soils at different stages of the soil water characteristic curve.

Investigate the use of conductivity and reflectivity

As a result of this project, conductivity, reflectivity, and dielectric constant were computed and reported. Only the dielectric constant was used for estimating the VMC and dry density on this project. In some cases this seemed to lead to unrealistic values for the dielectric constant relative to the moisture content. While the resulting moisture content values were reasonable as compared to the ground truth calibration moisture content, the composite dielectric constant was very low relative to the moisture content estimate. It is likely that the soil reflectivity and conductivity information obtained from TDR traces would account for these anomalous data. It is recommended that further research in this area be conducted to fully document the relationship of all components of the TDR trace on soil parameters.

Investigate Repeatability of TDR Traces

This project did not address the relative accuracy, variability, or repeatability of the computation process. Because the project dealt with existing data, there was no option available to obtain better information on the relative accuracy of the procedure other than to provide basic estimates from observations.

It is recommended that future research be conducted to investigate the accuracy and repeatability of the micromechanics-based procedure used on this project. The relative accuracy of the procedure was tested against the existing data from the site installations, as well as the data from Klemune's thesis data. ⁽⁵⁾

It is also recommended that more in situ information be collected and utilized to quantify the repeatability of the models. One such source of information would be from additional soil samples collected at the end of the TDR data collection sequence. This information could be used to compare estimates from the last TDR trace to ground truth data at a point in time other than during installation. In addition, it is recommended that repeated TDR traces be taken to determine the resultant variation in soil parameter estimates.

Sensitivity of MicroMoist program output to variation in input values

The data needed to fully evaluate and document the sensitivity of the MicroMoist program were not available for this study. This would require a series of laboratory tests where moisture content and/or dry density is changed using a constant soil sample. The TDR traces from multiple conditions over a range of soil types and conditions could be used to fully document sensitivity. This information could also be compared to ground truth values derived from moisture content and dry density laboratory testing.