Sudden intrusion of corrosive bottom water into the South Atlantic during the Paleocene-Eocene Thermal Maximum

Thursday, 17 December 2015: 11:05
2010 (Moscone West)
Katrin J Meissner, University of New South Wales, Sydney, NSW, Australia, Kaitlin Alexander, University of New South Wales, Sydney, Australia and Timothy J Bralower, Pennsylvania State University Main Campus, University Park, PA, United States
The Paleocene-Eocene Thermal Maximum, ∼55 million years before present, was a period of rapid warming marked by a negative carbon isotope excursion and widespread dissolution of seafloor carbonate. These changes have been attributed to a massive release of carbon into the exogenic carbon cycle, and thus, the event provides an analog for future climate and environmental changes given the current anthropogenic CO2 emissions. Previous attempts to constrain the amount of carbon released have produced widely diverging results, between 2000 and 10,000 gigatons carbon (GtC). Sediment records indicate that acidification of deep waters was generally more extensive and severe in the Atlantic and Caribbean regions, with more modest changes in the Southern and Pacific Oceans. Here we compare simulations integrated with the UVic Earth System Climate Model with reconstructions of temperature and dissolution to present a mechanism that might explain the observed spatial differences and to constrain the total mass of carbon released. Due to the late Paleocene topography, highly corrosive waters accumulate in the deep North Atlantic before the PETM in our simulations. Several thousand years into the event, deep ocean warming destabilizes the North Atlantic water column and triggers deep water formation. This causes the corrosive bottom water to spill over an equatorial sill into the South Atlantic and through the Southern and Pacific Oceans, progressively gaining alkalinity. The simulated pattern of sediment dissolution along the path taken by this corrosive water is consistent with most dissolution estimates made from CaCO3 measurements in the Paleocene-Eocene sediment record. We find two scenarios that agree best with proxy data: a carbon release of 7000 GtC in combination with pre-event atmospheric carbon dioxide concentrations of 840 ppm and a carbon release of 7000-10,000 GtC with pre-event CO2 concentrations of 1680 ppm.