Spatial Bootstrapping of High Spatial and Temporal Resolution Thermal Infrared Imagery: A Canopy Wetness Case Study

Abstract

We measure and compare temperatures of dry and wet plant canopies with a high spatial resolution thermal imaging camera. We then evaluate measurement variability and how this changes during evapotranspiration. Our assessments examine mean temperatures and their variability under constant evapotranspiration conditions. We partition our study by time since wetting (Tn), plant replicate (Pn), moisture condition (Mn), measurement height above target (Hn), and observation day (Dn). Treatments involved wetting half of each plant’s canopy, imaging each replicate from two heights (1 and 2 m) nadir to the canopy, and then assessing temperatures within wet and dry zone masks. Images were acquired every 5 minutes for 1 hour and treatments were replicated over two days. Spatial bootstrapping (SBS) was performed independently within each zone for each treatment using 50 random placements of a 3×3 pixel window. Our methods show that natural heterogeneity in dry canopies presents less variable temperatures than wetted canopies, where there was an influence of variable wetting. The allotted treatment duration successfully permits full evaporation of water from the wetted canopy and therefore allowed us to identify the point at which wetted zones returned to a localized equilibrium with the dry zones.