Stress Detection in Crops with Hyperspectral Remote Sensing and Physical Simulation Models

Abstract

Progress made on the detection of stress in heterogeneous crop canopies with hyperspectral remote sensing imagery is presented. High-spatial resolution multispectral remote sensing imagery was collected in 2002, 2003 and 2004 over vineyard and olive orchards in Spain. Imagery acquired with the Compact Airborne Spectrographic Imager (CASI) and the Reflective Optics System Imaging Spectrometer (ROSIS) in the visible and near infrared wavelength regions 400-950 nm at 1 m resolution, and with the Airborne Hyperspectral Scanner (AHS) in the reflective and thermal regions at 2 m resolution enabled the study of narrow-band vegetation indices and model simulation for estimation of chlorophyll content for chlorosis detection at the tree and vine level, as well as deriving thermal information function of the stress status. Ground data collection consisted of measurements of crown transmittance with a PCA LAI-2000 and geometrical measurements of crown projected area, height, crown cross-section, and biochemical constituents such as chlorophyll a+b and carotenoids, enabling the estimation of crown leaf area index, crown leaf density, biophysical variables related to the crown intercepted radiation, such as crop yield and canopy fractional cover, as well as crop functioning through chlorophyll content estimation. Leaf and canopy simulation models, such as PROSPECT, SAILH, FLIM, and rowMCRM were used and the scaling up methodology presented.