Quantifying the ocean carbon sink for 1994-2007: Combined evidence from multiple methods

By means of a variety of international observing and modeling efforts, the ocean carbon community has developed several independent estimates for ocean carbon uptake. In this presentation, we report on the synthesis effort we are undertaking under the auspices of an Ocean Carbon and Biogeochemistry Working Group. Our initial goal for this working group is to determine the best estimate for the net and anthropogenic carbon sink from 1994-2007, and then to infer the total magnitude of the poorly quantified fluxes that constitute their difference. Estimates for the net, or contemporary, ocean carbon uptake are derived from surface ocean pCO2 data interpolated to global coverage. From 4 of these products, we find Fnet = -1.7 PgC/yr for 1994-2007. Estimates for uptake of anthropogenic carbon comes from (1) interior observations of dissolved inorganic carbon and other tracers, (2) an ocean model constrained with observations, and (3) a suite of nine free-running ocean hindcast models in which the natural carbon cycle is assumed to be in a long-term steady state. Fant = -2.3 PgC/yr from the mean of these approaches. The difference between these two estimates is -0.6 PgC/yr, and acts as a quantitative constraint on the sum of the additional fluxes. As coastal zones and the Arctic are additional net carbon sinks, the sum of outgassed river-derived carbon, skin temperature effects on air-sea CO2 exchange, and non-steady state natural carbon fluxes in the open ocean can be no larger than a few tenths of PgC/yr. Our presentation details the uncertainties and assumptions made in deriving these estimates, and suggests paths forward to further reduce uncertainties.