LAUSR

Light scattered from surfaces is partly polarized, which has implications for remote sensing of vegetation. To study the contribution of polarized light to multi-angular reflectance over vegetation covers, we produced a hyperspectral bidirectional polarized reflectance dataset for four types of vegetation cover acquired over a wide viewing range in the laboratory and field. The datasets were analyzed with respect to basic physical polarized reflectance mechanisms. Moreover, we compared the modeled polarization from five existing polarimetric models (1D) with our measured polarization of vegetation covers at selected wavelengths. The analyses indicated that our measured polarized reflectance factors of a special vegetation cover, in which the leaves of vegetation were horizontally oriented, were much greater than previous results. All of the models produced good fits to the measurements, except for the special vegetation cover, which leaded to markedly different distribution patterns for the five models. We also found that the maximum relative difference between modeled and measured polarization was in the backward-scattering direction near the principal plane. These results suggested that we can model the polarization accurately using the existing polarimetric models in visible and near infrared bands; more measurements should be performed in the laboratory and the field to illustrate the inherent polarization property of vegetation covers and to avoid underestimating the contribution of vegetation covers on the polarization of atmosphere in the polarized remote sensing study. Our study not only deepens our understanding of the scattering properties of vegetation covers, but also provides the polarization properties of vegetated surfaces for the study of atmosphere polarization.