Regional Distribution and Seasonal Mechanisms of Carbon Uptake in the Global Oceans

Abstract

Using the methodology of Schuster et al. (2013) and the RECCAP archive, we evaluate the mean and seasonal cycles of 1990-2009 air-sea CO2 fluxes in 11 regions that cover the global ocean. Best estimates of the 1990-2009 mean flux are estimated for each region from the mean of the pCO2 climatology (Takahashi et al. 2009) and the ocean inversion (Gruber et al. 2009). We find that the Arctic, Atlantic, Pacific, Indian and Southern Oceans are responsible for 7%, 29%, 24%, 18% and 21%, respectively, of the 1990-2009 contemporary uptake of 1.7±0.3 PgC/yr. With an estimated river carbon input of 0.45 ±0.2 PgC/yr (Jacobson et al. 2007), our estimate for the global anthropogenic flux is 2.1±0.4 PgC/yr. In each region, estimates are compared to the medians of 11 atmospheric inversions and 7 ocean models. Overall, there is cross-methodological agreement with respect to the mean fluxes, though in some regions there is large uncertainty due to a wide range of estimates from ensemble members. Seasonal cycles of air-sea CO2 fluxes in 11 regions agree best in the subtropical gyres, due to temperature dominance of the ocean pCO2 cycle. At the high latitudes, correlations are less strong, but are generally still statistically significant. We assess how well ocean models are able to capture the biologically-driven component of the seasonal cycles here. One notable exception to the agreement of seasonal timated for each region from the mean of the pCO2 climatology (Takahashi et al. 2009) and the ocean inversion (Gruber et al. 2009). We find that the Arctic, Atlantic, Pacific, Indian and Southern Oceans are responsible for 7%, 29%, 24%, 18% and 21%, respectively, of the 1990-2009 contemporary uptake of 1.7±0.3 PgC/yr. With an estimated river carbon input of 0.45 ±0.2 PgC/yr (Jacobson et al. 2007), our estimate for the global anthropogenic flux is 2.1±0.4 PgC/yr. In each region, estimates are compared to the medians of 11 atmospheric inversions and 7 ocean models. Overall, there is cross-methodological agreement with respect to the mean fluxes, though in some regions there is large uncertainty due to a wide range of estimates from ensemble members. Seasonal cycles of air-sea CO2 fluxes in 11 regions agree best in the subtropical gyres, due to temperature dominance of the ocean pCO2 cycle. At the high latitudes, correlations are less strong, but are generally still statistically significant. We assess how well ocean models are able to capture the biologically-driven component of the seasonal cycles here. One notable exception to the agreement of seasonal