Scaling-up and model inversion methods with narrowband optical indices for chlorophyll content estimation in closed forest canopies with hyperspectral data

Study of the merit function in the numerical inversion showed that red edge optical indices used in the minimizing function such as R/sub 750//R/sub 710/ perform better than when all single spectral reflectance channels from hyper-spectral airborne CASI data are used, and in addition, the effect of shadows and LAI variation are minimized.

Abstract

Radiative transfer theory and modeling assumptions
\nwere applied at laboratory and field scales in order to study the
\nlink between leaf reflectance and transmittance and canopy hyperspectral
\ndata for chlorophyll content estimation. This study was
\nfocused on 12 sites of Acer saccharum M. (sugar maple) in the Algoma
\nRegion, Canada, where field measurements, laboratory-simulation
\nexperiments, and hyperspectral compact airborne spectrographic
\nimager (CASI) imagery of 72 channels in the visible and
\nnear-infrared region and up to 1-m spatial resolution data were
\nacquired in the 1997, 1998, and 1999 campaigns. A different set
\nof 14 sites of the same species were used in 2000 for validation of
\nmethodologies. Infinite reflectance and canopy reflectance models
\nwere used to link leaf to canopy levels through radiative transfer
\nsimulation. The closed and dense ( 4) forest canopies of
\nAcer saccharum M. used for this study, and the high spatial resolution
\nreflectance data targeting crowns, allowed the use of optically
\nthick simulation formulae and turbid-medium SAILH andMCRM
\ncanopy reflectance models for chlorophyll content estimation by
\nscaling-up and by numerical model inversion approaches through
\ncoupling to the PROSPECT leaf radiative transfer model. Study of
\nthe merit function in the numerical inversion showed that red edge
\noptical indices used in the minimizing function such as 750 710
\nperform better than when all single spectral reflectance channels
\nfrom hyperspectral airborne CASI data are used, and in addition,
\nthe effect of shadows and LAI variation are minimized. Estimates
\nof leaf pigment by hyperspectral remote sensing of closed forest
\ncanopies were shown to be feasible with root mean square errors
\n(RMSE’s) ranging from 3 to 5 5 g cm2. Pigment estimation by
\nmodel inversion as described in this paper using these red edge indices can in principle be readily transferred to the MERIS sensor
\nusing the 750 705 optical index.