Characterizing the seasonal cycle of upper-ocean flows under multi-year sea ice

•Idealized numerical simulation employed to study ocean ows under multi-year sea ice.•Submesoscale ows (SMs) are generally weak in mixed layer under multi-year sea ice.•Ice-ocean shear is the dominant surface-ocean forcing in both winter and summer.•Sporadic energetic SMs are associated with winter convection during weak shear.

Observations in the Arctic Ocean suggest that upper-ocean dynamics under sea ice might be significantly weaker than in the temperate oceans. In particular, observational evidence suggests that currents developing under sea ice present weak or absent submesoscale (O(1) Rossby number) dynamics, in contrast with midlatitude oceans typically characterized by more energetic dynamics at these scales. Idealized numerical model results of the upper ocean under multi-year sea ice, subject to realistic forcing, are employed to describe the evolution of the submesoscale flow field. During both summer and winter under multi-year sea ice, the simulated submesoscale flow field is typically much less energetic than in the midlatitude ice-free oceans. Rossby numbers under sea ice are generally consistent with geostrophic dynamics (Ro∼O(10−3)). During summer, ice melt generates a shallow mixed layer (O(1) m) which isolates the surface from deeper, warmer and saltier waters. The Ekman balance generally dominates the mixed layer, although inertial waves are present in the simulations during weakening and reversals of the ice-ocean stress. During winter, mixed-layer deepening (to about 40 m depth), is associated with convection driven by sea-ice growth, as well as ice-ocean shear-driven entrainment at the base of the mixed layer. Submesoscale activity is observed to develop only rarely, when winter convective mixing is laterally inhomogeneous (i.e., in the presence of sea-ice leads or spatially inhomogeneous sea-ice thickness) and when this coincides with weak ice-ocean shear-driven mixing. These submesoscale features are diagnosed with particular focus on their implications for ocean-to-ice heat fluxes.