LABORATORY AND FIELD MEASUREMENTS OF LEAF SPECTRAL PROPERTIES AND CANOPY ARCHITECTURE AND THEIR EFFECTS ON CANOPY REFLECTANCE

The vegetation radiative transfer model, Cupid, was tested by comparing simulations with measured canopy reflectances viewed from nadir. The model was used to evaluate the effects of leaf spectral properties and canopy structure on canopy reflectance. Non-Lambertian leaf spectral properties, characterized from laboratory measurements of bidirectional reflectance and transmittance of individual corn and soybean leaves in the visible and near-infrared (NIR) wavelength bands, were incorporated into Cupid to evaluate the effects of specular reflectance. Forward scattering was a prominent feature with increasing source incidence angle. Hemispherical reflectance generally increased with increasing incidence angle while transmittance values decreased. Leaf spectral properties, LAI, leaf angle distributions, bare soil reflectance and diffuse beam fractions of incoming radiation were measured concurrently with nadir-viewed diurnal canopy reflectances and provided input information for model simulations. Normal incidence hemispherical reflectance and transmittance of in situ corn and soybean leaves were measured throughout a growing season. Spectral properties in corn remained fairly constant. Visible reflectance and transmittance decreased during soybean leaf expansion but increased at senescence. The opposite occurred with NIR reflectance. LAI and leaf angle distributions were inferred from light penetration measurements along a transect. The effects of leaf spectral properties, LAI and leaf angle distribution on canopy reflectance were separated by varying one input parameter while holding other input variables constant. Canopies composed of senescing soybean leaves had the highest reflectance in the visible and NIR wavebands. Canopies of low LAI had a relatively higher reflectance in the visible waveband and a relatively lower reflectance in the NIR than did canopies with higher leaf area. Canopies with more horizontally inclined leaves had the least variability in diurnal nadir viewed reflectance and the highest reflectance. The opposite was true for canopies with more erect leaves. Simulated non-Lambertian canopy reflectances were in good agreement with canopy specular reflectance measurements reported in the literature. Predictions from the Cupid model generally agree with experimental results