Abstract
Abstract. We investigate the impact of three different parameterizations of ice–ocean heat exchange on modeled sea ice thickness, sea ice concentration, and water masses. These three parameterizations are (1) an ice bath assumption with the ocean temperature fixed at the freezing temperature; (2) a two-equation turbulent heat flux parameterization with ice–ocean heat exchange depending linearly on the temperature difference between the underlying ocean and the ice–ocean interface, whose temperature is kept at the freezing point of the seawater; and (3) a three-equation turbulent heat flux approach in which the ice–ocean heat flux depends on the temperature difference between the underlying ocean and the ice–ocean interface, whose temperature is calculated based on the local salinity set by the ice ablation rate. Based on model simulations with the stand-alone sea ice model CICE, the ice–ocean model MPIOM, and the climate model COSMOS, we find that compared to the most complex parameterization (3), the approaches (1) and (2) result in thinner Arctic sea ice, cooler water beneath high-concentration ice and warmer water towards the ice edge, and a lower salinity in the Arctic Ocean mixed layer. In particular, parameterization (1) results in the smallest sea ice thickness among the three parameterizations, as in this parameterization all potential heat in the underlying ocean is used for the melting of the sea ice above. For the same reason, the upper ocean layer of the central Arctic is cooler when using parameterization (1) compared to (2) and (3). Finally, in the fully coupled climate model COSMOS, parameterizations (1) and (2) result in a fairly similar oceanic or atmospheric circulation. In contrast, the most realistic parameterization (3) leads to an enhanced Atlantic meridional overturning circulation (AMOC), a more positive North Atlantic Oscillation (NAO) mode and a weakened Aleutian Low.
Highlights
The growth and decay of sea ice at the ice–ocean interface are determined by the local imbalance between the conductive heat flux within the ice and the oceanic heat flux from below the ice
The ice–ocean system is treated as an ice bath: The temperature of the uppermost ocean grid cells is fixed at the freezing point temperature, and any excess energy that enters these grid cells via advection, convection, or heat exchange with the atmosphere is instantaneously applied to the ice through lateral and bottom melting
The presented results focus on the Arctic Ocean, as we only find a small response of Southern Ocean properties to the change of ice–ocean heat flux parameterizations, in particular in Max Planck Institute Global Ocean/Sea Ice Model (MPIOM) and COSMOS; the stand-alone sea ice model CICE simulates an unrealistic distribution of sea ice in warm months in the Southern Ocean, as it fails to capture the heat release from the relatively deep mixed layer
Summary
The growth and decay of sea ice at the ice–ocean interface are determined by the local imbalance between the conductive heat flux within the ice and the oceanic heat flux from below the ice. It is necessary to investigate the ice–ocean heat flux formulations in a more complex system, including an interactive ocean or even the atmosphere Based on this motivation, in the present study we examine how different physical realism, represented by the three discussed parameterizations, impacts the resulting ice cover, large-scale oceanic circulation, and atmosphere properties in different numerical models including an idealized columnar model, a stand-alone sea ice model, an ice–ocean coupled model, and a complex climate system model. In the present study we examine how different physical realism, represented by the three discussed parameterizations, impacts the resulting ice cover, large-scale oceanic circulation, and atmosphere properties in different numerical models including an idealized columnar model, a stand-alone sea ice model, an ice–ocean coupled model, and a complex climate system model Another motivation of our study is to help improve the formulation describing ice–ocean heat flux in various models.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
