Hyperspectral mapping of crop and soils for precision agriculture (Proceedings Paper)

Abstract

Precision agriculture requires high spectral and spatial resolution imagery for advanced analyses of crop and soil conditions to increase environmental protection and producers' sustainability. GIS models that anticipate crop responses to nutrients, water, and pesticides require high spatial detail to generate application prescription maps. While the added precision of geo-spatial interpolation to field scouting generates improved zone maps and are an improvement over field-wide applications, it is limited in detail due to expense, and lacks the high precision required for pixel level applications. Multi-spectral imagery gives the spatial detail required, but broad band indexes are not sensitive to many variables in the crop and soil environment. Hyperspectral imagery provides both the spatial detail of airborne imagery and spectral resolution for spectroscopic and narrow band analysis techniques developed over recent decades in the laboratory that will advance precise determination of water and bio-physical properties of crops and soils. For several years, we have conducted remote sensing investigations to improve cotton production through field spectrometer measurements, and plant and soil samples in commercial fields and crop trials. We have developed spectral analyses techniques for plant and soil conditions through determination of crop water status, effectiveness of pre-harvest defoliant applications, and soil characterizations. We present the most promising of these spectroscopic absorption and narrow band index techniques, and their application to airborne hyperspectral imagery in mapping the variability in crops and soils.