Canopy optical indices from infinite reflectance and canopy reflectance models for forest condition monitoring: application to hyperspectral CASI data

Abstract

This paper reports on progress made to link physiologically-based indicators to optical indices from hyperspectral remote sensing. This study is carried out on twelve sites of Acer saccharum M. in the Algoma Region, Ontario (Canada), where field measurements and hyperspectral CASI imagery have been collected in 1997 and 1998 deployments. Individual tree samples were collected at each site for biochemical analysis and measurement of leaf chlorophyll, chlorophyll fluorescence and carotenoid concentrations, as well as leaf reflectance and transmittance. Physiological indices and derivative analysis indices extracted from leaf spectral reflectance have been tested at canopy level using CASI data of 72 channels and 2 m spatial resolution at 3 simulation scales which progressively more closely represent the observed above-canopy reflectance spectra from the sites: single leaf reflectance data, infinite reflectance calculated from optically-thick leaf simulation formulae, and canopy reflectance models using nominal site canopy architecture data. This study shows that selected algorithms connecting leaf reflectance and transmittance data to corresponding bioindicators at the leaf level can be expressed at canopy level through canopy models yielding predictions of bioindicators in airborne imaging spectrometer with coefficients of determination as high as 0.91.