Impact of eddy parameterisations on the simulated response of Southern Ocean air-sea CO2 fluxes to wind stress changes in IPCC-type ocean models
Final Report Abstract
The Southern Ocean is a sink for anthropogenic carbon. As to how this sink is changing in a world of strengthening winds in the southern hemisphere, there is no quantitative consensus. Among the processes limiting our ability to forecast carbon sequestered in the Southern Ocean discussed recently, is the role of swirling circulation features on scales of the order of 10 km to 100 km, the socalled eddies. Eddies are, due to associated computational burden, generally not explicitly resolved by numerical models applied to project air-sea carbon exchange. Rather, they are parameterized and - this being the main topic of the study at hand - there are several different contemporary approaches to parameterize them, and: there is worry that a deficient formulation of the parameterization may imprint wrong sensitivities. To this end we set out to assess the uncertainty of Southern Ocean carbon uptake associated with contemporary eddy-parameterization. Results obtained, show that differing approaches to parameterize the effect of eddies yield surprisingly similar Southern Ocean carbon uptake as a consequence of strengthening winds. This was unexpected. Overall, however and unfortunately, our results do not suggest that model projections of carbon uptake do converge. Contrary to the small effects of the choice of eddy-parameterization we found significant uncertainty associated with (rather unknown) dust emissions to the Southern Ocean which may (or may not) fertilise phytoplankton thereby driving additional photosynthesis and, eventually, sinking organic particles which may ultimately (or may not) sequester significant amounts of carbon. In addition, as a side-product of our coding work, we report a remedy for an apparently endemic model artefact in the eastern equatorial Pacific.
