The Paleocene-Eocene Thermal Maximum and Deep Sea Carbonate Compensation: Implications for Long-term Effects of Anthropogenic Carbon Emissions

The release of thousands of Pg C during the Paleocene-Eocene Thermal Maximum (55.5 Mya) represents a natural carbon cycle experiment that should provide insight into the potential carbon cycle response to current and future anthropogenic emissions. Unabated emission of fossil fuel CO2 over the next 3 centuries will release as much as 4500 Pg C to the atmosphere. Much of this carbon will eventually be absorbed by ocean, and buffered by massive dissolution of seafloor carbonates. Numerical simulations of this scenario often show similar patterns. Over the first several centuries, pCO2 rapidly rises peaking between 1800 and 2000 ppm. The total dissolved carbon in the ocean also rises, as pH and [CO32-] decline. Following peak emissions, pCO2 begins to decline and within several thousand years stabilizes, but at levels (350-500 ppm) significantly higher than the pre-anthropogenic steady state (Archer, 2005). This latter feature appears to be a consequence of the primary buffering process by which [CO32-] is initially restored, that is through the dissolution of seafloor carbonates, which leaves both ocean DIC and alkalinity at levels significantly higher than pre-anthropogenic (Tyrrell et al., 2007). Pelagic sediment records that span the PETM appear to exhibit several features that are consistent with the above pattern. The initial release of carbon, as represented by the negative carbon isotope excursion (CIE), triggered shoaling of the CCD and massive carbonate dissolution, followed by a gradual recovery of deep-sea carbonate chemistry. Though a detailed proxy record of pCO2 across the PETM does not yet exist, an indirect record, sea surface temperatures, suggests pCO2 remained high for tens of thousands of years after the peak input of isotopically depleted carbon as inferred from the CIE. This feature of the PETM is consistent with the simulated anthropogenic perturbation, and therefore lends some credibility to the prediction of high pCO2 for thousands of years after anthropogenic emissions have ceased. One feature that is not always reproduced in simulations, however, is a subsequent phase of carbonate oversaturation, a feature that has been attributed to a slow negative feedback, the weathering of silicate rocks (Dickens et al., 1997).