Canopy Level Detection of Nitrogen Stress In Corn Using Remote Sensing Techniques.

How to more efficiently utilize Nitrogen (N) fertilizer is an important issue facing the agricultural community. The environmental impact of excessive N application and considerations of economically optimizing input to yield ratios warrant research in topics such as the optimal time to apply N fertilizer, effectiveness of variable rate applications to improve Nitrogen Use Efficiency (NUE), and remote sensing techniques to detect the onset of N stress and the spatial variability of that stress. A multi-year assessment of the efficacy of spectral remote sensing derived vegetation indices (VIs) to discern N stress in corn has been conducted in Maryland from 2004 to 2011. Aerial hyperspectral imagery, ground based passive sensor reflectance data, and ground based active sensor reflectance data were collected. Additionally, biophysical parameters including plant height, leaf area index (LAI), leaf chlorophyll content, and development stage were acquired weekly from growth stages V6 to R1. Vegetation indices derived from spectral remote sensing can be grouped into two broad classes, biomass responsive indices and plant pigment responsive indices. As N stress in corn is expressed in reduced leaf chlorophyll levels, plant pigment responsive indices seem more appropriate for its detection. It was found that VIs effectiveness for discerning N stress in corn plants varies the most at growth stages prior to canopy closure, when the effects of background components and variations in plant biomass confound the detection of the leaf chlorophyll levels. The performance of various plant pigment responsive VIs converge as LAI increases and the plant canopy closes. Normalizing a VI appropriate for detecting leaf chlorophyll levels with a VI well correlated to plant biomass was shown to improve the ability to extract information about leaf chlorophyll variation from canopy level remote sensing spectra collected at growth stages prior to canopy closure.