A global spatially and temporally complete reflectance anisotropy database to improve surface characterization for environmental modeling

This effort utilizes the operational MODIS (Moderate Resolution Imaging Spectroradiometer) reflectance anisotropy products to create spatially and temporally complete bases to support regional and global modeling and monitoring efforts. By considering the surface reflectance anisotropy or BRDF (Bidirectional Reflectance Distribution Function), it is possible to produce more accurate land surface albedos and nadir reflectances for better quantification of the terrestrial biosphere. Although a nine-year record of global BRDFs from NASA's Terra and Aqua satellites now exists, inclusion in regional and global models has been limited by the extensive data-gaps caused by persistent clouds and ephemeral snow cover. Therefore this research focuses on bridging the gaps in the MODIS BRDF products to provide consistent, long-term, climate-quality records.;Significant progress is achieved by applying rigorous temporal interpolation techniques based on vegetation development curves. Comparison of the resulting gap-free, snow-free, global MODIS products with the direct BRDF retrievals possible from the Polarization and Directionality of the Earth's Reflectance (POLDER-3) sensor shows a very good linear relationship between these two remotely sensed products.;Merging data from sensors on different satellites is effective in increasing the number of high-quality retrievals and reducing gaps in long-term datasets. The operational MODIS BRDF products are already derived from both Terra and Aqua observations, resulting in fewer gaps than if derived from only one sensor. This potential is explored further by combining observations from the polar-orbiting sensors with those aboard geostationary platforms to reduce gaps in regional datasets.;The importance of a priori information in increasing the number and quality of BRDF retrievals is also investigated in this work. Gap-filled products are used to inform subsequent BRDF retrievals; even when very few new observations are available. An evaluation verifies the efficacy of this approach in maximizing the influence of available observations while reducing the number of gaps occurring in the overall record.;This research establishes several effective techniques to both gap-fill the current MODIS BRDF, albedo, and nadir reflectance datasets and reprocess subsequent versions in order to provide the consistent, high-quality, long-term reflectance anisotropy databases required by the regional and global modeling and monitoring communities.