|Loechel, Sara - UNIVERSITY OF MARYLAND|
Submitted to: International Conference Agricultural Mechanization for Precision Farming
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: December 1, 2000
Publication Date: July 16, 2000
Interpretive Summary: Nitrogen (N) is an essential element for plant growth and is frequently the major limiting nutrient in most agricultural soils. Farmers must balance the competing goals of supplying enough N to their crops, while minimizing the loss of N to the environment. This represents a threat to water quality and an economic loss. The economic penalties of reduced yields from not supplying adequate N are substantial. Our objectives were to simulate canopy reflectance and to evaluate a strategy for detecting N status of corn using remotely sensed images. Leaf chlorophyll concentration is an indicator of plant N status. Changes in leaf chlorophyll concentrations produce rather broad band differences in leaf reflectance and transmittance spectra. However, the transition from leaf spectra to canopy reflectance spectra is complicated. Variations in background reflectance and LAI confound detection of subtle differences in canopy reflectance due to changes in leaf chlorophyll concentration. Some spectral vegetation indices minimize background reflectance contributions, while other indices responded more to leaf chlorophyll concentrations. Combining these two groups of spectral vegetation indices in the same figure produced isolines of leaf chlorophyll concentrations. We demonstrated using simulated data and hyperspectral images. Pairs of spectral vegetation indices can estimate leaf chlorophyll concentrations with minimal confounding effects due to LAI and background reflectance. This approach holds promise as a management tool because leaf chlorophyll concentrations were determined with minimal confounding due to variations in background reflectance and LAI.
Technical Abstract: Nitrogen (N) is an essential element for plant growth and is frequently the major limiting nutrient. Farmers must balance the competing goals of supplying adequate N for their crops and minimizing N losses to the environment. Variations in soil reflectance and leaf area indexes often confound the assessment of leaf N by remote sensing techniques. Our objective was to examine a strategy for detecting leaf chlorophyll status of corn plants using aerial hyperspectral imagery. Field-grown corn was supplied with eight levels of N to establish a wide range of leaf chlorophyll levels. Reflectance and transmittance spectra of upper fully expanded leaves were acquired over the 400-1000 NM wavelength range. Corn canopy reflectance was simulated for a wide range of conditions. Variations in background reflectance and leaf area indexes (LAI) confounded detection of the subtle differences in canopy reflectance due to changes in leaf chlorophyll concentration. Some spectral vegetation indices minimized contributions of background reflectance, while others responded to both leaf chlorophyll concentrations and background reflectance. Pairs of the spectral vegetation indices plotted together produced isolines of leaf chlorophyll concentrations. Analysis of the hyperspectral imagery showed consistent patterns of leaf chlorophyll concentrations even where a leaf area index more than doubled. This approach holds promise as a management decision aid because leaf chlorophyll concentrations were determined with minimal confounding due to variations in background reflectance and leaf area indexes.