Toward a global synthesis of the oceanic carbon sink since the mid 1990s

Substantial efforts have been undertaken in the past 5 years by the international community with guidance by the IMBER/SOLAS carbon working groups and the IOCCP to assemble, harmonize, and interpret global ocean surface and interior carbon data sets, such as SOCAT (www.socat.info) and GLODAPv2. While the interpretation of these data sets is on-going, among the first emerging results are: (i) an estimate of the oceanic accumulation of anthropogenic CO2 between the 1990s and the mid-2000s (Gruber et al., in prep.) based on the ocean interior carbon data, and (ii) monthly resolved global air-sea CO2 fluxes for this entire period and extending back as far as 1982 developed on the basis of the surface ocean data (e.g., Landschützer et al., 2015). Here, we aim to synthesize and compare these two estimates in order to assess the consistency in time and space between the time-integrated air-sea CO2 fluxes and the changes in ocean interior storage. The results suggest a global increase in the inventory of anthropogenic CO2 of 32±6 Pg C between 1994 and 2007, while the cumulative air-sea CO2 flux over this period amounts to about 17±3 Pg C. Assuming a cumulative outgassing flux of ~6 Pg of “natural” carbon stemming from the carbon input by rivers, the surface ocean perspective suggests an anthropogenic CO2 uptake of about 23±4 Pg C over the 13 years, i.e., considerably smaller than the ocean interior estimate, although not statistically different. Thus the two perspectives suggest an ocean sink rate for anthropogenic CO2 between 1.8±0.3 Pg C yr-1 (surface ocean) and 2.5±0.5 Pg C yr-1 (interior ocean), which is consistent with the expected uptake based on the increase in atmospheric CO2, although at the lower end. Substantial variability emerges in both uptake flux and ocean interior storage in response to recent climate variations, suggesting an ocean carbon cycle that responds quite sensitively to changes in climate forcing.