David E. Archer, Pamela A. Martin, and Carrie Ashendel, Department of
Geophysical Sciences, University of Chicago, Chicago, IL, USA
Jose Milovich, Lawrence Livermore National Laboratory, Livermore, CA, USA
Viktor Brovkin, Potsdam Institute for Climate Impact Research, Potsdam, Germany
Gian-Kasper Plattner1, Climate and
Environmental Physics, Physics Institute, University of Bern,
Switzerland
1Now at Institute of Geophysics and
Planetary Physics, University of California, Los Angeles, Los Angeles,
California, USA
Several recent papers have demonstrated a decrease in atmospheric pCO2 resulting from barriers to communication between the deep sea and the atmosphere in the Southern Ocean. Stephens and Keeling [2000] decreased pCO2 by increasing Antarctic sea ice in a seven- box model of the world ocean, and Toggweiler [1999] showed a similar response to Southern Ocean stratification. In box models, the pCO2 of the atmosphere is controlled by the region of the surface ocean that fills the deep sea [Archer et al., 2000a]. By severing the Southern Ocean link between the deep sea and the atmosphere, atmospheric pCO2 in these models is controlled elsewhere, and typically declines, although the models range widely in their responses. "Continuum models", such as 3-D and 2-D general circulation models, control pCO2 in a more distributed way, and do not exhibit box model sensitivity to high latitude sea ice or presumably stratification. There is still uncertainty about the high-latitude sensitivity of the real ocean. Until these model sensitivities can be resolved, glacial pCO2 hypotheses and interpretations based on Southern Ocean barrier mechanisms (above plus [Elderfield and Rickaby, 2000; Francois et al., 1998; Gildor and Tziperman, 2001; Sigman and Boyle, 2000; Watson et al., 2000]) are walking on thin ice.