HR: 17:55h
AN: C22B-08    [PDF]
TI: Oases on Snowball Earth: Confluence of Ice Dynamics Modeling and Geological Observations
AU: * Hoffman, P F
EM: hoffman@eps.harvard.edu
AF: Department of Earth and Planetary Sciences, Harvard University 20 Oxford Street, Cambridge, MA 02138 United States
AU: Maloof, A C
AF: Department of Earth and Planetary Sciences, Harvard University 20 Oxford Street, Cambridge, MA 02138 United States
AU: Halverson, G P
AF: Department of Geology, University of Namibia, Windhoek, 9000 Namibia
AU: Schrag, D P
AF: Department of Earth and Planetary Sciences, Harvard University 20 Oxford Street, Cambridge, MA 02138 United States
AB: Recent model experiments suggest that marine ice dynamics are important in the initiation and development of a snowball Earth. Ice-line advance is facilitated by the transport of latent heat and freshwater associated with Ekman forcing in the zone of westerlies (Lewis et al., 2002 this meeting) and with equatorward glacial flow of thick, multi-annual, marine ice (Goodman and Pierrehumbert, 2002 this meeting). After ice lines meet at the equator (snowball Earth), continued invasion by marine glaciers from higher latitudes maintains tropical marine ice $>$2.25 times thicker ($<$450 m) than static ice at thermal dynamic equilibrium (G \& P, 2002). The dynamics imply that the tropical ocean will remain ice covered even after rising atmospheric $p$CO$_2$ drives tropical sea surface temperatures to the melting point. However, landfast marine ice on low-latitude continental shelves and inland seas that are protected from invasion by marine and terrestrial glaciers will melt away, creating 'oases' on a snowball Earth. Snowball oases should occupy a small fraction ($<$5%) of the global surface area but they might significantly enhance water vapor transport and accumulation rates of adjacent terrestrial glaciers. Despite acidification by high $p$CO$_2$, warming may cause critical oversaturation with respect to calcite or dolomite in snowball oases that are buffered by carbonate-rich bedrock and glacial debris (Fairchild, 1994). As snowball oases are small pools of water in contact with a $p$CO$_2$-rich ($>$0.1 bar) atmosphere, the carbon isotopic composition of any carbonates they precipitate should evolve accordingly. Initially, oasis water may resemble evolved snowball seawater, which will be dominated by hydrothermal activity buffered by dissolution of sea-floor carbonate (Higgins and Schrag, G-cubed, 2003). Oasis water should evolve rapidly towards equilibrium with the atmospheric reservoir, whose isotopic composition is set by volcanic outgassing, not by the ocean. Oasis carbonate strata should exhibit a rapid rise in $\delta^{13}$C with time, and may well have values substantially greater than 0 per mil PDB. We believe there are numerous examples of snowball oases in the geologic record associated with the Neoproterozoic Sturtian and Marinoan glaciations. Examples in NE Svalbard, SW Oman, SE California, and central and South Australia are well documented. In all but the first, the lithologic and isotopic evidence has been cited as precluding a snowball Earth altogether. In contrast, we consider that oases are a natural and important part of the snowball Earth cycle. Our concept of snowball oases combines insights from geophysical modeling and geology.
DE: 1827 Glaciology (1863)
DE: 3022 Marine sediments--processes and transport
DE: 3344 Paleoclimatology
DE: 4540 Ice mechanics and air/sea/ice exchange processes
DE: 4870 Stable isotopes
SC: Cryosphere [C]
MN: 2003 Fall Meeting