Davidson, A, Wang, SS (2004). The effects of sampling resolution on the surface albedos of dominant land cover types in the North American boreal region. REMOTE SENSING OF ENVIRONMENT, 93(2-Jan), 211-224.

The central role that land surface albedo (7) plays in the physical climate system makes it a key component of climate and ecosystem models. However, this parameter remains one of the largest radiative uncertainties associated with modelling attempts. Uncertainty occurs because models commonly prescribe albedo using in situ observations, which are rarely sufficiently dense to accurately characterize albedo at a regional scale. This is especially problematic over seasonally snow-covered landscapes such as the boreal forest. The aims of this study are to (a) analyze and compare the local- and regional-scale albedo characteristics of the dominant land cover types found within the North American boreal region, (b) assess the effects of snow cover on these patterns, and (c) quantify the potential bias that can result from using local-scale observations to describe surface albedos across larger geographical extents. Our study is based on local-scale in situ observations and regional-scale satellite (Geostationary Operational Environmental Satellite-GOES) measurements that were collected as part of the Boreal EcosystemAtmosphere Study (BOREAS). Our results show (a) that the albedo patterns among land cover types are generally consistent at local and regional scales, (b) that snow cover not only increases the albedo of all cover types, but also their sensitivities to changes in solar zenith angle, and (c) that weekly averaged in situ observations provide a reasonable characterization of regional-scale albedo when under snow-free conditions, but a poor characterization when snow is present. Land cover albedo characteristics are caused by canopy properties that influence within-canopy shadowing. The disparity between in situ albedo observations and those collected over low-density needleleaf forest are particularly a concern because this cover type comprises a significant proportion of the boreal region, and its mis-specification in climate models could lead to large errors in energy balance. Further studies should focus on reducing the disparity between albedo data sets over snow-covered surfaces. They should also consider the effects of diffuse radiation, as well as finer time scales, on the above relationships. (C) 2004 Elsevier Inc. All rights reserved.