Cryogenian Snowball Earth: a 25th anniversary assessment of models and data

Abstract:

In 1989, Joe Kirschvink applied a nascent planetary climate theory—the self-reversing ice-albedo instability—to Neoproterozoic geology, citing paleomagnetic evidence that ancient Australian glaciers reached tidewater near the paleo-equator. He attributed the “snowball Earth” to a unique preponderance of low-latitude continents, and inferred that complex life including metazoa were spawned in the nutrient-rich glacial aftermath. There have been major advances in models and data since 1989, and gratifying convergences between them. In models, frozen oceans coexist with dynamic ice sheets, as demanded by geological data. Geochronological data show that snowball glaciations were prolonged and their terminations globally synchronous, as required by geochemical models for snowball deglaciation. In models, zonal average sources and sinks of water vapor are reversed in snowball atmospheres relative to climate states with ice lines (at low or high-latitude). Ice-covered oceans are well mixed, by true thermohaline meridional overturning and related jets, with implications for data linking iconic snowball iron-formations with silled basins. Triple oxygen isotope data from sulfate robustly supports model predictions of atmospheric CO₂ required for snowball deglaciation, while promising results have been obtained from a variety of pH, weathering, productivity and redox sensitive proxies. The known first appearances of sponge biomarkers and metazoan embryos bracket the terminal Cryogenian snowball glaciation. Existing model-data conflicts are good targets for future research. Boron isotopes in syndeglacial cap dolomites consistently imply catastrophic ocean acidification upon deglaciation, while models suggest that the existence of cap dolomites implies that seawater acidified gradually and was buffered by the dissolution of carbonate, delivered by glaciers. Above all, the newly emerging Cryogenian chronology (highlighted in this session) presents new puzzles. Two low-latitude glaciations of grossly unequal longevity occurred in rapid succession. The Sturtian was 55-60 Myr long, followed a 10⁹-yr glacial hiatus, and was distinguished by Fe₂O₃-rich sedimentary deposits; the Marinoan was 5-20 Myr long, followed a 10⁷-yr glacial hiatus, and was distinguished by BaSO₄-rich cap dolomite.