The impact of "assimilating" global, daily snow depth in seasonal climate simulations is tested in the Center for Ocean Land Atmosphere Studies (COLA) atmospheric general circulation model (AGCM) and the Goddard Earth Observing System 2 (GEOS-2) AGCM. The seasonal simulations are run for the northern hemisphere spring (March-June) when the typical widespread ablation of the ephemeral snow cover occurs, and span the years 1982-1998. A set of control simulations are performed in which snow cover is allowed to evolve interactively within the climate model, and the "assimilation" runs are performed in which global fields of daily snow-depth observations are prescribed. The global snow fields are taken from the U.S. Air Force (USAF) daily snow-depth analysis.

Given the framework of the numerical experiments and the influence of snow on the surface energy-budget (principally through albedo effects), the analysis of these impact studies will focus on near-surface air temperature skill. Diagnosis of the COLA AGCM results suggest that the most notable and widespread impact of the realistic prescription of snow depth occurs during the month of April, and a weak impact is seen in the remaining months of the simulations (i.e. March, May, and June). The result is consistent with the intuitive expectation that (realistic) snow-cover anomalies will support strong (and skillful) temperature responses when coincident with relatively high incident solar radiation (i.e., when the impact of albedo modulation from the snow anomalies would be maximal). Results from the GEOS-2 AGCM, using a similar land surface model as in the COLA AGCM but with an improved snow-physics scheme, will also be analyzed and compared to the COLA AGCM results. The relative strength between the two AGCMs' responses to the snow assimilation will be weighed accordingly against the errors of simulated snow in the control runs, as well as the representation of snow processes (and their functional link to the surface energy budget) between the AGCMs.