Ventilation by mesoscale eddies has a negligible impact on the rate at which anthropogenic carbon is sequestered within the global ocean

In response to increasing concentrations of carbon dioxide in the atmosphere, the ocean is estimated to take up ~2.3 Pg C yr^-1. Emerging evidence has shown that mesoscale eddies can act to significantly alter the rate of carbon uptake by the ocean; however, current model-based estimates of the anthropogenic carbon flux rely on empirically derived parameterisations of mesoscale eddies. Such parameterisations may affect modelled carbon fluxes differently to models which explicitly resolve mesoscale eddies. The rectified impact of explicitly resolved mesoscale eddies on the global anthropogenic carbon flux has not been quantified before.

We estimate how changes in ocean ventilation resulting from the explicit resolution of mesoscale eddies alter the global uptake of anthropogenic carbon by the ocean. We use the transit-time distribution approach to reconstruct the oceanic inventory of anthropogenic carbon in both an eddy-resolving (5 km resolution) and an eddy-parameterising (20 km resolution) configuration of the ICON-Ocean model. Each model is integrated using a perpetual year forcing and five boundary impulse response tracers, required for estimating the transit-time distribution.

The uptake of anthropogenic carbon in the eddy-resolving model exceeds that in the eddy-parameterising model by 0.1 Pg C yr^-1 over the period 2005–2015, which is smaller than typical inter-model differences of around ±0.5 Pg C yr^-1. The root mean square difference in column integrated inventories of anthropogenic carbon between the eddy-resolving and eddy-parameterising model is 4.3 mol m^-2, which is slightly larger than uncertainties in observational estimates of column integrated anthropogenic carbon of around ±2 mol m^-2.

Our results suggest that explicitly resolving mesoscale eddies is unlikely to produce large differences in globally integrated anthropogenic carbon inventories via ventilation changes alone. Further differences may arise from eddy-driven effects on the solubility of carbon dioxide, gas transfer velocities and the biological carbon pump — the transit-time distribution approach only describes the effects of ventilation in the physical carbon pump.