Biophysical and Biochemical Sources of Variability in Canopy Reflectance

to nadir-viewed vegetation reflectance data. Variability INTRODUCTION in tissue optical properties was wavelength-dependent. During the past several decades, the tools for vegetation For green foliage, the lowest variation was in the visible remote sensing have evolved significantly. Optical remote (VIS) spectral region and the highest in the near-infrared sensing has expanded from the use of panchromatic and (NIR). For standing litter material, minimum variation multispectral sensors to off-nadir looking instruments and occurred in the VIS/NIR, while the largest differences imaging spectrometers. Ecological remote sensing now enwere observed in the shortwave-IR (SWIR). Woody stem compasses a wide range of applications including vegetamaterial showed opposite trends, with lowest variation in tion mapping, land-cover change detection, disturbance the SWIR and highest in the NIR. Leaf area index (LAI) monitoring, and the estimation of biophysical and bioand leaf angle distribution (LAD) were the dominant chemical attributes of ecosystems (Asner et al., 1998a). As controls on canopy reflectance data with the exception of increased value has been placed on remote sensing for soil reflectance and vegetation cover in sparse canopies. ecological research, management and modeling, a conLeaf optical properties (and thus foliar chemistry) were comitant increase in understanding the factors that influexpressed most directly at the canopy level in the NIR, ence vegetation radiometric signals has been only parbut LAI and LAD strongly controlled the relationship betially realized. tween leaf and canopy spectral characteristics. Stem maBased on experimental and modeling evidence, vegterial played a small but significant role in determining etation reflectance is known to be primarily a function of canopy reflectance in woody plant canopies, especially tissue (leaf, woody stem, and standing litter) optical propthose with LAI,5.0. Standing litter significantly affected erties, canopy biophysical attributes (e.g., leaf and stem the reflectance characteristics of grassland canopies; area, leaf and stem orientation, and foliage clumping), soil small increases in the percentage of standing litter had a reflectance, illumination conditions, and viewing geometry disproportionately large affect on canopy reflectance. The (e.g., Ross, 1981; Goel, 1988; Myneni et al., 1989; Jacquestructural attributes of ecosystems determine the relative moud et al., 1992). Foliage and nonphotosynthetic vegetacontribution of tissue, canopy, and landscape factors that tion (NPV5woody stems, standing litter, etc.) affect the