30-m Land Surface Albedo by Integrating Landsat directional reflectance and MODIS anisotropic information

Land surface albedo as a key physical variable determining the solar energy absorbed by the land surface, and can affect climate through ecosystem feedback processes. Some studies have highlighted that positive radiative forcing (warming) induced by increased forest cover and decreased albedo in temperate and boreal forest regions could offset the negative forcing expected from carbon sequestration (Betts 2000). However, these studies have not used data at the spatial resolution of human land dynamics (e.g. 30m Landsat resolution). Therefore, there is a need for improved estimates of land surface albedo at high resolution to fully understand the role of land cover change in climate forcing and carbon cycle. Following our initial "concurrent" approach applied to Landsat data acquired during the post-2000 MODIS era (Shuai et al.2011), we have developed a "pre-MODIS era" approach to generate 30-meter albedos using Landsat surface directional reflectance (1970s-2000) and Look-Up-Tables (LUT) of anisotropy information extracted from MODIS BRDF data. We use a NLCD (National Land Cover Dataset)-class-based LUT for non-disturbed land cover. Disturbed forest patches are identified from the Monitoring Trends in Burn Severity (MTBS) and North American Forest Dynamics (NAFD) datasets. For each category, high quality MODIS BRDF parameters (MCD43A1 product) are retrieved and used to populate the LUT. Each entry in the LUT reflects a unique combination of land cover type, disturbance age and type, season/month, and sensor bands. The initial BRDF LUT generated for the Pacific Northwest of the United States exhibits various BRDF evolution trajectories for disturbed classes, including different recovery trajectories for fire and non-fire disturbance. The albedo-to-nadir-ratio method (Shuai et al., 2011) is applied to the BRDF LUT to calculate spectral albedos, followed by a narrow-to-broadband conversion (Liang 2000) to generate broad-band shortwave albedo. Our preliminary validation over three Ameriflux sites shows that the LUT-based land surface albedo, has a good agreement with both ground measurements and those from "concurrent approach", with values slightly higher than the current 500m 8-days MODIS operational retrievals. Future work will expand the validation to various land cover types across the full range of North American vegetation regimes. We then intend to evaluate the net change in both winter and summer albedo across the continent due to land cover change processes occurring since the 1970's. This information will also feed into a global analysis of albedo changes occurring since the 1700's parameterized using the IPCC Global Land Use harmonization data set (Hurtt et al., 2009).