Parameterizing the Dependence of Surface Albedo on Solar Zenith Angle Using Atmospheric Radiation Measurement Program Observations

This study evaluates the parameterization of the dependence of surface albedo on solar zenith angle (SZA) over snow-free land surface used by the National Centers for Environmental Prediction Global Forecast Systems (GFS) and those derived from the moderate resolution imaging spectroradiometer satellite observations as described in two recent studies. The influence of cloud cover on surface albedo is also examined. The intensive ground observations of surface shortwave fluxes and cloud covers made by the Atmospheric Radiation Measurement (ARM) Program at the United States Southern Great Plains (SGP) site and at the Tropical Western Pacific (TWP) locale’s Manus and Nauru Islands in 1997 – 2004 are used for the investigation. The albedos for the direct and diffuse solar beams are considered separately. The diffuse-beam albedo is first derived from a subset of the ARM Program samples under overcast conditions and with the downward total flux being all diffuse. Then, it is used to divide the measured surface upward flux into two parts, one associated with the downward direct beam and the other with the downward diffuse beam for all samples. The direct-beam albedo is finally obtained using the first part of the upward flux and the downward direct-beam flux. It is further normalized either by the diffuse-beam albedo or by the direct-beam albedo at SZA=60°. The normalized albedo only varies with the SZA and is independent of the solar spectrum. Results show that at the SGP site, while the diffuse-beam albedo prescribed in the GFS matches closely the ARM Program observed, the direct-beam albedo parameterized in the GFS largely underestimated at all SZA. The parameterizations derived from the moderate resolution imaging spectroradiometer products tend to underestimate the direct-beam albedo at large SZA and slightly overestimate at small SZA. The influence of cloud cover on surface albedo is negligible in comparison with the albedo variation with the SZA. This study recommends that in weather and climate models the direct-beam albedo over snow-free land surfaces be parameterized as a product of the direct-beam albedo at SZA=60° (or the diffuse-beam albedo), which can be obtained from satellite observations and be prescribed in the model differently for different surface types and/or seasons, and a universal function that depends only on the SZA.