Mapping Subfield-Scale Evapotranspiration to Assess Agricultural Drought Sensitivity

Abstract:

Assessing crop response to drought on the subfield-scale is critical for efficient agricultural water management and yield forecasting. Evapotranspiration provides a direct physical link between the soil, crop canopy, and the atmosphere, and is hence highly sensitive to changes in water availability. Here, we introduce a new surface energy balance model (High Resolution Mapping of Evapotranspiration; HRMET) that can map ET at very high resolution (meter-scale) requiring only canopy surface temperature, canopy structure, and meteorology as inputs. HRMET can be used in both open and closed canopy conditions. We validate HRMET over two commercial cornfields in the Yahara River Watershed (south-central Wisconsin, USA) and investigate the spatially variable ET response to severe drought conditions during the 2012 growing season. Results show that the magnitude of within-field ET variability is much larger when the drought is more severe. We then introduce a new metric, Relative ET (ETR), which normalizes ET on a field scale and allows for direct comparison across measurement dates, despite differences in meteorological conditions and crop growth stage. Using a novel paired-image technique, we use persistent patterns of ETR identify portions of the field that are most susceptible to drought, and portions that are consistently productive across measurement dates. These results have implications for precision agriculture and irrigation efficiency in addition to water management and yield forecasting, as identification of persistent patterns in crop productivity during low-stress periods allows farmers to direct resources to the most sensitive areas early in droughts.