Northern North Atlantic climate variability controls on ocean carbon sinks in EC-Earth3-CC

Abstract

The northern North Atlantic (NA) is an important net sink of atmospheric CO2, though atmosphere–ocean CO2 fluxes exhibit substantial variability across different timescales. The underlying drivers of this variability remain poorly understood across both temporal and regional scales. Here, we investigate interannual to decadal CO2 flux variability in the northern NA using historical simulations from the EC-Earth3-CC model. We assess the role of key dynamical and physical processes in shaping CO2 flux variability across five regions: the Nordic Seas (NS), eastern NS, the eastern and western subpolar NA, and the full NA. Our analysis reveals that physical parameters – including sea ice concentration (SIC), sea surface temperature (SST), sea surface salinity (SSS), and wind stress – along with dynamical processes related to ocean mixing and circulation, play a central role in regulating CO2 flux variability. Using regression analysis, we demonstrate that these drivers exert regionally and temporally varying influences, with our models achieving high R2 values indicating a strong degree of explanation for CO2 flux variability. The regression models capture interannual variability more effectively than decadal variability, highlighting the dominant role of short-term fluctuations in shaping CO2 flux dynamics. Overall, our results demonstrate that the predictors of CO2 flux variability are both spatially and temporally dependent. We find that CO2 flux variability cannot be fully explained by simple linear correlations with individual predictors but instead arises from complex interactions among multiple physical and dynamical processes. Notably, CO2 flux variability is particularly sensitive to changes in certain predictors, such as wind stress, consistent with expectations based on the gas transfer equation used to compute atmosphere–ocean CO2 fluxes.