Surface and Atmospheric Heating Responses to Spectrally Resolved Albedo of Frozen and Liquid Water Surfaces
Cryospheric surfaces play a critical role in modulating Earth's global energy budget, as they reflect large fractions of solar radiation. While the albedo of these surfaces varies significantly with wavelength, Fully coupled Earth System Models exchange the solar flux and surface albedo of the entire spectrum within only two bins. Here, we quantify the instantaneous bias due to this semi-broadband approximation.
Using a spectrally resolved albedo reduces atmospheric warming, most significantly over snow-covered surfaces but also over bare ice, relative to the semi-broadband albedo counterpart. For fresh snow, shortwave absorption can decrease by as much as 1% to 10% for clear and cloudy skies, respectively.
Snow, ice, and (to a lesser extent) liquid water reflectances depend on the wavelength of light. Some Earth System Models (ESMs) approximate this spectrally varying reflectance with two bands. This eases the computational burden yet fails to accurately capture the spectrally structured absorption of the surface and changes the distribution of flux that is reflected back to the atmosphere. Here we use a single‐column radiative transfer model to expand the spectral representation of snow, ice, and liquid water albedo to 14 bands (matching those used by the atmospheric model in ESMs). By implementing this moderately resolved albedo, we change the radiation absorbed by the surface, as well as the atmosphere. Depending on environmental conditions, implementing a 14-band albedo over snow changes surface absorption by over 10%, relative to the semi‐broadband approximation, which increases or reduces the susceptibility of snow to melting. Furthermore, changes to the spectral distribution of sunlight reduce atmospheric absorption by nearly 2%. These results highlight the importance of accurately representing the spectral albedo for snow and ice‐covered surfaces in ESMs that adopt the semi‐broadband albedo approximation.