Unmanned Aerial Vehicles for Nitrogen Management of Corn (Zea mays L.): A Framework for Predicting Spatial and Temporal Variability of N Requirement.

Traditional canopy spectroscopy methods can provide estimates of nitrogen (N) rates in corn production, but are constrained by labour requirements, cost, and poor early-season detection of N differences. Vegetative indices (VIs) generated by ground-based sensors have had some success in predicting nitrogen requirements in corn, but issues persist in differentiating nitrogen status prior to the V8 development stage. Unmanned aerial vehicles (UAVs) are an emerging crop monitoring tool able to rapidly measure spectral reflectance of crop canopies. UAV-derived VI maps may improve N rate determination through estimation of spatial and temporal variation in corn leaf-N concentration. However, the sensitivity of aerial sensors depends on accurate correction of radiometric and flight conditions, and is not well known for low-altitude UAV surveys. Additionally, their ability to differentiate nitrogen sufficiency has not been tested in corn. During the 2016 season, a fixed-wing UAV carrying consumer digital cameras modified to intercept near infrared and visual wavebands will capture images from the V4 to V12 development stages for seven pre-plant N treatments. The 16-bit images in TIF format will be calibrated for noise, lens distortion, and radiometric differences for use in time series analysis. Data will be extracted from orthorectified maps of VIs including normalized difference vegetative index (NDVI), soil adjusted vegetative index (SAVI), and chlorophyll index (CI) using a novel machine vision software, and plot values will be used to test the relationship between VIs and leaf-N concentration, as well as the most economically optimal rate of N. Ground reference data including soil- and tissue-N concentrations, spectroscopy, and yield response will be used to validate VI maps of the trial. Accurate estimation of N status from UAV imagery, combined with weather and soil-N supply data, may improve N requirement calculations.

Key words: Unmanned aerial vehicles, nitrogen, corn, multispectral