Quantitative characterization of the vegetation red edge reflectance 1. An inverted-Gaussian reflectance model

Publisher Website
Google Scholar
Abstract
An inverted-Gaussian model for the vegetation red edge reflectance is evaluated with respect to its applicability as a simple four-parameter descriptor of vegetation reflectance in the 670 to 800 nm spectral region under a wide range of environmental/measurement conditions. The model has been fitted to laboratory spectral reflectance measurements of single leaves, leaf stacks and needle clump stacks for a number of species. For all of these data the model has been found to provide an effective quantitative representation of the shape and position of the vegetation red edge reflectance in terms of four parameters of physical significance: 1R shoulder reflectance R s, chlorophyll-well minimum reflectance R 0, red edge inflection point wavelength λp and reflectance minimum wavelength λ0. Provided that an appropriate strategy has been adopted to select the initial guess model parameters and the spectral range of reflectance data to be fitted, the values of derived model parameters can be used for a quantitative description of the temporal and species-dependent behaviour of the characteristics of the red edge. Sequential measurements of the same leaf stack in which the first leaf is changed and as well as remeasurement of the same sample indicate that appropriate confidence limits to the derived model parameters are: R 0 (±0·5 per cent), R s ( ± 4 per cent), λ0 and λp (±2·0nm). It is also suggested that in the analysis of high spectral resolution vegetation reflectance data, inverted-Gaussian model estimates of the spectral position of the red edge obtained with discrete filter-passband radiometers (or multispectral imagers) can be usefully compared directly to that obtained with more sophisticated field or laboratory spectrometers.