The-low frequency natural variability of the arctic climate system is modeled using a single-column, energy balance model of the atmosphere, sea ice, and upper-ocean system. Variability in the system is induced by forcing with realistic, random perturbations in the atmospheric energy transport and cloudiness. The model predicts that the volume of perennial sea ice varies predominantly on decadal time-scales, while other arctic climate variables vary mostly on intraannual and interannual time-scales. The variance of the simulated sea ice volume is most sensitive to perturbations of the atmospheric forcing in late spring, at the onset of melt. The variance of sea ice volume increases with the mean sea ice thickness and with the number of layers resolved in the sea ice model. This suggests that much of the simulated variance develops when the surface temperature decouples from the sea ice interior during the late spring, when melting snow abruptly exposes the sea ice surface and decreases the surface albedo.
The minimum model requirements to simulate the natural variability in the arctic climate are identified. The implications of the low-frequency, natural variability in sea ice volume for detecting a climate change are discussed. Finally, calculations suggest that the variability in the thermodynamic forcing of the polar cap could lead to a freshening in the North Atlantic that is comparable to the freshening associated with the Great Salinity Anomaly.
Full article available upon request email@example.com