A radiosity model for heterogeneous canopies in remote sensing

A radiosity model has been developed to compute bidirectional reflectance f rom a heterogeneous canopy approximated by an arbitrary configuration of pl ants or dumps of vegetation, placed on the ground surface in a prescribed m anner. Plants are treated as porous cylinders formed by aggregations of lay ers of leaves. This model explicitly computes solar radiation leaving each individual surface, taking into account multiple scattering processes betwe en leaves and soil, and occlusion of neighboring plants. Canopy structural parameters adopted in this study have served to simplify the computation of the geometric factors of the radiosity equation, and thus this model has e nabled us to simulate multispectral images of vegetation scenes, Simulated images have shown to be valuable approximations of satellite data, and then a sensitivity analysis to the dominant parameters of discontinuous canopie s (plant density, leaf area index (LAI), leaf angle distribution (LAD), pla nt dimensions, soil optical properties, etc.) and scene (sun/ view angles a nd atmospheric conditions) has been undertaken. The radiosity model has let us gain a deep insight into the radiative regime inside the canopy, showin g it to be governed by occlusion of incoming irradiance, multiple scatterin g of radiation between canopy elements and interception of upward radiance by leaves. Results have indicated that unlike leaf distribution, other stru ctural parameters such as LAI, LAD, and plant dimensions have a strong infl uence on canopy reflectance. In addition, concepts have been developed that are useful to understand the reflectance behavior of the canopy, such as a n effective LAI related to leaf inclination.