Storms link vertical turbulent mixing to CO2 outgassing in the Southern Ocean

Sarah-Anne Nicholson1, Marcel du Plessis2, Alice Dolaine Lebehot2, Daniel B Whitt3, Ilker Fer4, Sebastian Swart5 and Pedro M. S. Monteiro1, (1)Council for Scientific and Industrial Research, Cape Town, South Africa, (2)University of Cape Town, Department of Oceanography, Cape Town, South Africa, (3)NASA, Mountain View, CA, United States, (4)University of Bergen, Bergen, Norway, (5)University of Gothenburg, Gothenburg, Sweden
The upwelling sector of the Southern Ocean is a globally critical region. There, atmospheric modes of variability (e.g., Southern Annular Mode) are thought to directly influence the outgassing CO2 fluxes and therefore modify interannual and decadal trends in the global air-sea fluxes of CO2. This region is also associated with the maximum frequency and intensity of atmospheric storms. Although the interannual storm characteristics are intimately connected to such atmospheric modes, the mechanisms through which storms modulate the CO2 outgassing remains unexplored. South of the Antarctic polar front, surface and profiling robotic platforms are coupled together to provide an unprecedented dataset that links two months of continuous high-resolution turbulent dissipation to air-sea CO2 fluxes. We show a large decoupling between the storm-driven mixing-layer extent and the mixed-layer depth (up to 70 m) during low wind phases (< 0.2 Nm-2), highlighting the sensitivity/insensitivity of the mixing-layer/mixed-layer to wind variability. During strong storms (> 0.5 N m-2) mixing extends beyond the mixed-layer, and is linked to a reversal of air-sea CO2 flux (from ingassing to outgassing), due to the entrainment of DIC-rich subsurface waters. This work provides for the first time direct observations of intra-seasonal storm-driven mixing and CO2 outgassing in the Southern Ocean and points to the need to resolve the variability of upper-ocean mixing in Earth System Models in order to better constrain carbon-climate interannual and decadal trends.