Submesoscale eddies and sea ice interaction

The submesoscale has been defined dynamically as those processes with Rossby and Richardson numbers approaching O(1). This scale is of emerging interest within oceanography due to the role it plays in surface layer nutrient and tracer transport. Submesoscale baroclinic eddies or mixed layer eddies (MLEs), if energised in the marginal ice zone (MIZ), have the potential to impact both the rate of ice melt/formation and the magnitude of air-sea heat fluxes in the vicinity of the ice edge.  

In this study, an MITgcm idealised high resolution simulation is used to quantify the impact of MLEs in the vicinity of the ice edge, focusing on the thermodynamic component. The domain (75 km by 75 km at 250 m resolution) is a zonally re-entrant channel with ice-free/ice-covered conditions in the South/North, representing a lead or the MIZ. To measure the eddy impact on both sea ice and air-sea heat fluxes, comparisons are made between a 3D simulation with eddies and a 2D simulation with no eddies (no zonal extension, but otherwise identical to the 3D version). Typical conditions (stratification, forcing) of the Arctic/Antarctic and summer/winter seasons are considered. 

When eddies are permitted to energize and develop within these simulations, their impacts are numerous and coupled: under summer Artic conditions, meridional heat transport to the ice-covered region is tripled with eddies present, which leads to a first order impact on the sea ice melt and a doubling of the average heat storage in the ice-covered ocean. Novel analysis into the direct impact of these eddies on air-sea heat fluxes also shows that - due the partial absorption of downwelling solar radiation by sea ice cover - the solar heat flux into the ice-covered mixed layer increases by 20% when eddies are present. Computing the residual overturning stream function, responsible for driving warmer waters under the ice, reveals the ocean dynamics behind these impacts. The overturning, weakly present in the 2D model due to frontogenesis, increases threefold in the 3D case with submesoscale eddies. Tests with the Fox-Kemper parameterization within the 2D set-up are also helping evaluate to which extent this parameterization can capture the influence of MLE eddies in these polar conditions.