Estimating Arctic sea-ice shortwave albedo from MODIS data

Shortwave surface albedo of the Arctic sea-ice zone is one of the most important parameters in the surface energy budget over the northern hemispheric cryosphere. Although many global and regional broadband surface albedo products have been generated from satellite observations, most of them contain albedo over land-surfaces only, and the albedo over ocean and sea-ice surfaces is usually left blank or estimated based on the lambertian-assumption. In this paper, we extend the BRDF-based direct-estimation algorithm for mapping the shortwave surface albedo of the Arctic sea-ice zone with Moderate Resolution Imaging Spectroradiometer (MODIS) data. Firstly, a bidirectional reflectance distribution function (BRDF) database for sea ice was generated based on simulations of the asymptotic radiative transfer (ART) and the three components ocean water albedo (TCOWA) model. Secondly, a linear regression relationship between top of atmosphere (TOA) reflectance and surface broadband albedo was developed for different angular bins. Finally, the direct-estimation coefficients derived by the angular bin regression were stored in a pre-calculated look up table (LUT). When the MODIS L1B swath data are available, the corresponding coefficients are searched based on the solar/view geometry. Therefore, the shortwave surface albedo can be estimated from single-angular observations. Compared with the in situ measurements from the Tara polar ocean expedition in 2007, the results indicate that the albedo estimated by the BRDF-based direct-estimation algorithm has an R2 value of 0.67 and root mean square error (RMSE) of 0.068. The validation results show that the BRDF-based direct-estimation algorithm is suitable for mapping shortwave surface albedo of the Arctic sea-ice zone, which corrects the reflectance anisotropic effect with prior physical BRDF database and makes it possible to generate long-term Arctic surface albedo products with higher spatial and temporal resolutions.