Sea Ice and Climate. Part II: Model Climate Stability to Perturbations of the Hydrological Cycle

Abstract

The sensitivity of a simple climate model to variations in a global hydrological cycle is studied. The model consists of a zonally averaged single basin, two-hemisphere ocean model coupled to an atmospheric energy balance model and a thermodynamic sea ice model. Land processes are reduced to considerations of the oceanic catchment basins, which serve to amplify or decrease the oceanic freshwater forcing. The model typically admits both symmetric and asymmetric equilibria. For sufficiently strong global freshwater forcing, the symmetric circulation is replaced by asymmetric states. This is accomplished by a supercritical pitchfork bifurcation. It is shown that sea ice enhances climate sensitivity in that larger changes of the hydrological cycle are required to induce instability if sea ice is neglected. This behavior comes from the strong damping of SST anomalies under ice, which is essential for suppressing deep ocean temperature anomalies. Thus, ocean temperature plays an important role in climate stability. This is due to the interactive atmosphere, a feature absent from ocean-only models. The roots of transition from symmetric to asymmetric states are in the salinity balance. Differences in ocean temperature anomalies result in stronger anomalous overturning in the model with sea ice. Salt transport differences are proportional to those in overturning anomalies. The anomalous heat transports are similar due to compensating differences in the surface?abyss steady-state temperature contrasts, arising from sea ice phase transition. In steady state, the water column under ice is homogeneous and energetically passive, consistent with oceanic insulation by sea ice. Thus, the downwelling in the region covered by sea ice is controlled kinematically by the basic flow. The width of this region is restricted by the sea ice extent. In contrast, the narrow high-latitude open ocean sinking is most dependent on the relative importance of temperature advection versus vertical diffusion in the global oceanic heat budget. Sea ice albedo and brine rejection effects have negligible influence. Comparisons with highly truncated box model studies suggest that high horizontal resolution is essential to accurately determine model climate stability.