Novel Sea Ice Volume Tracers in an Earth System Model: Understanding the Thermodynamic and Dynamic Processes Governing Arctic Sea Ice Evolution at the Regional Scale

Abstract:

The future of the Arctic is inextricably tied to the fate of its sea ice. Strategic modeling of how sea ice volume evolves seasonally, annually, and inter-annually is one important approach for understanding (and anticipating) current and future changes in the Arctic ice pack. To this end, we implement a novel collection of 19 bulk Eulerian ice volume tracers, spanning the entire Arctic, in the state-of-the art Community Earth System Model version 1.2 (CESM1.2). For each tracer, thermodynamic growth occurs over a restricted local tagged region; the ice pack advects, converges, and diverges the tracer from its region of growth; and the tracer melts in proportion to the local melt rate. These volume tracers are passive and, therefore, do not impact any characteristics of the ice pack in the model. We ran a 150-year, fully-coupled, preindustrial control run with these tracers, where the tracers were reinitialized every ten years, to obtain 15 realizations of Arctic sea ice volume evolution. We find that each tracer contributes uniquely to the evolution of the Arctic ice pack, with distinctive sub-seasonal to interannual patterns of ice growth, spatial extent, and ice export from the Arctic. We analyzed each of the tracers for their respective contributions to the North Atlantic freshwater cycle, particularly over regions of ocean convection in the Labrador, Greenland, and Nordic Seas. Our analyses show that sea ice originating in the eastern central Arctic and Kara Sea contribute the most to freshwater buoyancy forcing in the Greenland and Nordic Seas, while most sea ice in the Labrador Sea originates in Baffin Bay, the central Arctic, and the Greenland Sea. We further find that because these ice volume tracers integrate both thermodynamic and dynamic processes over the course of the seasonal cycle, they can also be used to improve sub-seasonal to seasonal predictability of extreme Arctic climate events. The sea ice volume tracer module developed here will be made available to the wider research community in future releases of CESM.