Using Thermal Infrared Imagery To Estimate Soil Hydraulic Parameters: A Novel Approach

In this study, skin temperature measured with a thermal infrared (TIR) camera was used to estimate soil hydraulic parameters. These physical properties that control how soils transport and retain water are notoriously difficult to measure in the field due to spatial variability. Laboratory experiments were set up to record surface skin temperature response in a clean soil column using a TIR camera after an artificial wetting event. An array of thermocouples, a net radiometer, heat flux sensor and weather station were used to constrain the TIR data and the energy budget during the experiment. The soil column surface was then wetted with a known amount of water over a controlled time period and the thermal response recorded at five minute intervals over the course of 18 hours. Soil hydraulic parameters were then estimated by fitting a water-energy conservation model (ECH2O) to the observed data using a Marqhart-Levenberg least squares minimization method. The estimated parameters obtained were then compared to several sets of known values based on soil textural classification. This inversion of ECH2O was able to estimate the Brooks-Corey λ for sand with a relatively high degree of precision; however, the inversion was unable to provide reasonable estimates of air entry pressure for sand, air entry pressure for soil, or the Brooks-Corey λ for soil. These results indicate that soil hydraulic parameter estimation based on TIR skin temperature data could potentially be a fast and useful new tool to characterize the distribution and spatial heterogeneity in some soil hydraulic parameters. However, future studies should test the method with dedicated groundwater flow models and accurately account for surface emissivity before conducting field tests