NOAA/AVHRR Bidirectional Reflectance: Modeling and Application for the Monitoring of a Temperate Forest

In this article, bidirectional effects on NOAA/AVHRR short wavelength data were analyzed for the regional monitoring of the temperate pine (Pinus pinaster) Landes forest. Because of differences in soil hydrologic characteristics, the forest is described by five ecotypes, which determine the understorey type and the potential for pine growth. Bidirectional visible and near-infrared reflectances were analyzed from a daily NOAA-11 archive through years 1990–1994 using the bidirectional reflectance function model of Rahman. Rahman’s model fits 72% of AVHRR visible reflectance, and 93% of AVHRR near-infrared reflectance. A typical example shows that 1) in the principal plane, the ratio of hot spot to forward scatter reflectance is 2.9 and 2.1 for visible and near-infrared, respectively, 2) bidirectional effects become negligible across the principal plane, and 3) NDVI shows a shape opposite to that of reflectances. After normalization, the seasonal dynamics of AVHRR reflectances were related to the phenological cycle of pine trees and understorey vegetation. Visible reflectance drops earlier on coastal than on inland sites, due to an advance in the onset of pine growth. Near-infrared reflectance increases in summer for annual understorey, and remains steady for evergreen understorey. Differences in the seasonal dynamics of NDVI consequently characterize each Landes ecotype. From the modeled reflectances, we estimated the seasonal dynamics of visible and near-infrared albedo. We calculated the solar albedo by a combination of the last two. The visible albedo displays a seasonal dynamic similar to that of normalized visible reflectance. The solar albedo was found constant and around 12.7% on each ecotype. In both cases, albedo strongly increases with solar zenith angle, consistent with ground measurements. These results clearly show the usefulness of coarse spatial resolution satellite data for the regional monitoring and modeling of temperate coniferous ecosystems.