Arctic carbon cycling

The marine Arctic is considered a net carbon sink, with large regional differences in uptake rates. More regional modelling and observational studies are required to reduce the uncertainty among current estimates. Robust projections for how the Arctic Ocean carbon sink may evolve in the future are currently lacking. Direct connections have been documented between sea-ice dynamics and carbon cycling in marine ecosystems and on land. Projections suggest further sea-ice decline may accelerate changes in carbon cycling dynamics at sea and on land. Although rivertransported organic and inorganic carbon plays a major role in the marine Arctic carbon cycle this is not well studied. Changes in terrestrial ecosystems may also affect sea-ice decline – at least in the long term. Permafrost underlies ~75% of the area draining into the Arctic Ocean but its hydrology is poorly understood, especially under global warming. Arctic tundra is a net sink for atmospheric carbon dioxide (CO2) in the growing season and the sink strength has more than doubled since 2000 in Eurasia. In contrast, the few winter data available show tundra ecosystems are a net source of atmospheric CO2 in winter. Small features below the resolution of current lake and wetland databases may be important controls on carbon transfer from permafrost soils to the atmosphere. Earth System Models (ESMs) are not yet able to reliably simulate the full dynamics of the Arctic carbon cycle. This is mainly because such models still address terrestrial and marine systems separately and because they vary widely in their representation of permafrost. Further development of ESMs should include a focus on improving the connections between ocean and land, especially in the representation of lateral fluxes.