Minimization of Shadow Effects in Forest Canopies for Chlorophyll Content estimation using Red Edge Optical Indices through Radiative Transfer: Implications for MERIS

Abstract

This paper reports on progress made to develop methods to accurately estimate pigment content in forest canopies from airborne hyperspectral data. Radiative transfer approaches and red edge optical indices were applied to twelve sites of Acer saccharum M. in the Algoma Region, Ontario (Canada), where field measurements and hyperspectral CASI reflectance data have been collected between 1997 and 2000 deployments. Turbid-medium models such as SAILH and Kuusk were successfully applied to closed forest canopies minimizing the effects of canopy shadows and structure by using red edge optical indices in the merit function. Pigment content was successfully estimated linking PROSPECT leaf model to SAILH canopy model, minimizing the merit function by iterative process in MERIS-derived spectra from hyperspectral CASI data. The selection of the merit function used for minimization has been demonstrated to be critical, showing that shadow effects are removed when red edge optical indices are used. Implications for MERIS sensor by ESA on ENVISAT are discussed in this paper.