Implemented predictions of drought tolerance using canopy temperature as proxy of water status

Surface temperature measurements have become a largely utilized high-throughput method to obtain empirical evidence while testing plant-environmental interactions. Unfortunately, the mechanistic relevance of temperature data at large spatial and temporal scales remains a challenge. Satellite remote sensing methods need vast integrations that often masking leaf-level physiological responses to the environmental changes. The heterogeneity of surfaces further lowers the predictive capability of current models of plant productivity due to the spatial and temporal variation of emissivity. Thus, ground-truth studies are essential to improve the mechanistic significance of canopy thermal signatures. We aimed to utilize high-throughput leaf temperature methods as proxies for stomatal conductance in sustained drought conditions. Thermal imagery and infrared radiometer techniques were used to monitor the progression of drought response for species with different susceptibility to soil water limitation. Ground-truthing was conducted on broadleaf (Populus tremuloides) and needleleaf (Pinus ponderosa) trees that are represented tree species in North America. Specifically, thermal measurements of the canopy were tested against leaf-level measurements of gas exchange. Hierarchical Bayesian Penman-Monteith energy balance models were developed to scale changes of leaf temperature and obtain estimates of transpiration and canopy resistance under drought. Energy balance models were validated across diverse vegetation including crops (Brassica rapa), subalpine forest, and desert shrubs. Within five days of drought, leaf temperature measured from infrared radiometer increased (5-17%), with leaf temperature explained 68% and 48% of the decline in stomatal resistance in broadleaf and needleleaf species, respectively. While leaf temperature from thermal images accounted for less (24-43%) of the canopy resistance. Overall, our results show a robust correlation between thermal observations and drought progression validating the use of the high-throughput methods for improved predictions of plant water status.