Comparing Ground and Uav Remote Sensing for NUE and Variety Selection in Wheat.

Twelve wheat (Triticum aestivum) varieties were examined at two locations in Virginia: the Tidewater AREC in Suffolk and the Eastern Virginia AREC in Warsaw.  These twelve varieties were subjected to two fertility rates: a low fertility treatment with a total of 60 lbs/ac of spring Nitrogen and a normal fertility treatment with a total of 120 lbs/ac of spring Nitrogen. Measurements were taken at several times throughout the growing season with both ground based sensors and aerial based sensors with a UAV. Ground taken measurements included NDVI, leaf area index, and canopy temperature depression. The UAV taken measurements were taken with three different sensors that were individually mounted on the UAV: a RGB digital camera, a near-infrared multispectral camera, and a infrared camera. Images taken with the RGB camera were processed in Image J software to compute color space characteristics such as hue angle, intensity, saturation, as well as RGB-derived vegetation indices Green Area (GA) and Greener Area (GGA). The images from the near-infrared multispectral camera were used to derive NDVI, while the images from the infrared camera were used to obtain canopy temperature depression.

UAV-taken NDVI was successful in estimating distinguishing between the two N rates, and there was up to a 0.85 correlation between ground-measured NDVI and UAV-taken NDVI. At each growth stage when the measurements were taken, UAV-taken NDVI was better correlated with yield than ground-taken NDVI. The infrared camera was successful in discriminating varieties with high and low nitrogen use efficiency by showing which varieties exhibit greater sensitivity to nitrogen stress as temperature increase. Therefore, UAV collected data has the potential to replace ground collected data for variety selection.