Inversion of a physically based bidirectional reflectance model of vegetation

The challenge in using remote sensing data for the characterization of vegetation is to exploit reflectance measurements to infer biophysical and structural properties of the medium under study. In the present paper, the resolution of this problem was achieved, in the case of an homogeneous canopy, through a physically based bidirectional reflectance factor model. In a first validation phase, some inversions were performed against actual data obtained during the FIFE'89 experiment, and the retrieved parameters were found to be in good agreement with the corresponding in situ measurements. A sensitivity study was conducted afterward with synthetic data sets in order to identify and address some of the difficulties encountered while inverting such models, with a view to improving the accuracy of the retrievals and evaluating the inversion costs. The tackled points concerned not only the influence of the initial estimate values or the repercussions of errors made on the estimation of one of the variables, but also the angular sampling and the effect of simplifying assumptions made at the modeling step.