Effect of gas-transfer-velocity parameterization choice on CO<sub>2</sub> air–sea fluxes in the North Atlantic and European Arctic
Abstract. The ocean sink is an important part of the anthropogenic CO2 budget. Because the terrestrial biosphere is usually treated as a residual, understanding the uncertainties the net flux into the ocean sink is crucial for understanding the global carbon cycle. One of the sources of uncertainty is the parameterization of CO2 gas transfer velocity. We used a recently developed software tool, FluxEngine, to calculate monthly net carbon air–sea flux for the extratropical North Atlantic, European Arctic as well as global values (or comparison) using several available parameterizations of gas transfer velocity of different dependence of wind speed, both quadratic and cubic. The aim of the study is to constrain the uncertainty caused by the choice of parameterization in the North Atlantic, a large sink of CO2 and a region with good measurement coverage, characterized by strong winds. We show that this uncertainty is smaller in the North Atlantic and in the Arctic than globally, within 5 % in the North Atlantic and 4 % in the European Arctic, comparing to 9 % for the World Ocean when restricted to functions with quadratic wind dependence and respectively 42, 40 and 67 % for all studied parameterizations. We propose an explanation of this smaller uncertainty due to the combination of higher than global average wind speeds in the North Atlantic and lack of seasonal changes in the flux direction in most of the region. We also compare the available pCO2 climatologies (Takahashi and SOCAT) pCO2 discrepancy in annual flux values of 8 % in the North Atlantic and 19 % in the European Arctic. The seasonal flux changes in the Arctic have inverse seasonal change in both climatologies, caused most probably by insufficient data coverage, especially in winter.