摘要

Multiple short-lived global warming events, known as hyperthermals, occurred during the early Eocene (56-52 Ma). Five of these events - the Paleocene-Eocene Thermal Maximum (PETM or ETM1), H1 (or ETM2), H2, I1, and I2 - are marked by a carbon isotope excursion (CIE) within both marine and terrestrial sediments. The magnitude of CIE, which is a function of the amount and isotopic composition of carbon added to the ocean-atmosphere system, varies significantly between marine versus terrestrial substrates. Here we use the increase in carbon isotope fractionation by C-3 land plants in response to increased pCO(2) to reconcile this difference and reconstruct a range of background pCO(2) and peak pCO(2) for each CIE, provided two potential carbon sources: methane hydrate destabilization and permafrost-thawing/organic matter oxidation. Although the uncertainty on each pCO(2) estimate using this approach is low (e.g., median uncertainty = +23%/-18%), this work highlights the potential for significant systematic bias in the pCO(2) estimate resulting from sampling resolution, substrate type, diagenesis, and environmental change. Careful consideration of each of these factors is required especially when applying this approach to a single marine-terrestrial CIE pair. Given these limitations, we provide an upper estimate for background early Eocene pCO(2) of 463 +248/-131 ppmv (methane hydrate scenario) to 806 +127/-104 ppmv (permafrost-thawing/organic matter oxidation scenario). These results, which represent the first pCO(2) proxy estimates directly tied to the Eocene hyperthermals, demonstrate that early Eocene warmth was supported by background pCO(2) less than similar to 3.5x preindustrial levels and that pCO(2) > 1000 ppmv may have occurred only briefly, during hyperthermal events.

  • 出版日期2017-11-15