Canopy Radiative Transfer Models for Spherical and Known Leaf Inclination Angle Distributions: A Test in an Oak-Hickory Forest

Publisher Website
Google Scholar
Abstract
Two canopy radiative transfer models were tested against solar radiation measurements made in an oak-hickory forest. One model assumes that the distribution of leaf inclination angles in spherical (Norman 1979). The other model uses measurements of the distribution of leaf inclination angles to stimulate the radiation regime within the canopy (de Wit 1965). Vertical profiles of downward directed insolation components computed with Norman''s model are in better agreement with measured values than are those computed with de Wit''s model. Insolation flux densities computed with both models, however, underestimate total incoming shortwave radiation and its components direct and diffuse shortwave radiation, total photosynthetically active (PAR) and near IR radiation (NIR). This underestimation results primarily from clumping of the foliage, which enhances the penetration of direct and diffuse radiation into the canopy and minimizes the conversion of beam to scattered diffuse radiation. Although observed leaf inclination angles are different from a spherical leaf inclination distribution these differences do not account for differences between simulated and measured fluxes of solar radiation. This is because the leaf orientation function (G) computed with observed leaf angles is greater than the G-function for a spherical canopy, which should lead to an overestimation in beam penetration. Possible errors in the optical parameters are also not a factor since most of these parameters were measured directly and because the agreement between measured and simulated diffuse shortwave radiation, which is affected by optical parameters, is relatively good. Canopy reflectivities of PAR are overestimated by a factor of three. This is because deep penetration of insolation reduces the amount of light that is intercepted and scattered upward in the upper canopy. De Wit''s model estimates shortwave reflectivity reasonably well. However, this is an artifact of compensating errors.