Estimation of Crop Cover and Chlorophyll from Hyperspectral Remote Sensing

Over the last two decades, there has been extensive development in hyperspectral remote sensing. Interest is rapidly growing in the application of hyperspectral data to precision farming. This paper investigates the potential of hyperspectral remote sensing data for providing crop information for use in precision farming. Ground measurements and airborne hyperspectral Probe-1 data were simultaneously acquired in July 1999 near Clinton, Ontario, Canada. Specifically, percent ground cover and chlorophyll estimations derived from the Probe-1 data are being validated. Constrained linear unmixing was conducted on the airborne hyperspectral surface reflectance and at-sensor radiance data to determine crop endmember fractions. Chlorophyll maps were generated from Probe-1 reflectance data using three different methods. Correlations between ground data and Probe-1 derived image products were significant and produced encouraging results. Although based on a limited range of chlorophyll values available in this study, Probe-1 derived chlorophyll index values were sensitive to differences in SPAD-502 measurements taken in the field. Crop ground cover was significantly correlated with spectral fractions derived from the radiance or the reflectance data.