Is the Southern Ocean a net source of CO2 to the atmosphere or a net sink? Of what magnitude?
Studies of delta-pCO2 between the atmosphere and oceanic surface waters have previously indicated the
Southern Ocean (defined rather broadly in those studies as the zonal band between 40 and 70°S) to be
an area of net CO2 flux into the ocean, with an estimated magnitude of 2.7 GT C y-1 (Takahashi et al.,
1986). That flux is approximately equal to the global net uptake of CO2 by the oceans (ibid.),
indicating that CO2 fluxes are approximately in balance for the rest of the ocean. In other words, the
Southern Ocean may dominate the global net uptake of CO2 by the oceans. That view can be
disputed, however, as studies of meridional CO2 gradients in the atmosphere suggest little or no net
uptake of CO2 by the ocean in the Southern Hemisphere (e.g., Tans et al., 1990). With known areas
of net uptake in the vicinity of the Subtropical Convergence (Takahashi et al., 1986), a net Southern
Hemisphere flux of approximately zero would imply net outgassing from the waters to the south. The
large-scale circulation of the Southern Ocean is characterized by zonal bands of convergence and
divergence, suggesting a reasonably complex pattern of alternating sources and sinks, with the result
that the overall net balance is difficult to evaluate. The spatial coverage of existing delta-pCO2 data is not
particularly good in the Southern Ocean and seasonal information is almost completely lacking,
especially in the Pacific sector. Thus the net CO2 flux is difficult to constrain using the available data
to no better than -3 to +1 GT C y-1, with negative values denoting uptake.
Implications for field measurements: Field measurements of delta-pCO2 should be given a high
priority within the Southern Ocean JGOFS programs of all nations, and should seek to maximize both
spatial (especially meridional) resolution and seasonal coverage. The overall purpose of this work
should be to constrain the net CO2 exchange within much tighter limits than it is now possible to do.
Implications for modeling: It would be helpful to develop satellite-observable proxies for oceanic
pCO2 (based on surface temperature and perhaps ocean color). Once developed, these could be
merged with wind data to model CO2 fluxes after 1998.