Tempo and scale of late Paleocene and early Eocene carbon isotope cycles: Implications for the origin of hyperthermals

A 200,000-yr interval of extreme global warming marked the start of the Eocene epoch about 55 million years ago. Negative carbon- and oxygen-isotope excursions in marine and terrestrial sediments show that this event was linked to a massive and rapid (approximately 10,000 yr) input of isotopically depleted carbon. It has been suggested previously that extensive melting of gas hydrates buried in marine sediments may represent the carbon source and has caused the global climate change. Large-scale hydrate melting, however, requires a hitherto unknown triggering mechanism. Here we present evidence for the presence of thousands of hydrothermal vent complexes identified on seismic reflection profiles from the Vøring and Møre basins in the Norwegian Sea. We propose that intrusion of voluminous mantle-derived melts in carbon-rich sedimentary strata in the northeast Atlantic may have caused an explosive release of methane--transported to the ocean or atmosphere through the vent complexes--close to the Palaeocene/Eocene boundary. Similar volcanic and metamorphic processes may explain climate events associated with other large igneous provinces such as the Siberian Traps (approximately 250 million years ago) and the Karoo Igneous Province (approximately 183 million years ago).

Carbonate and organic matter deposited during the latest Paleocene thermal maximum is characterized by a remarkable -2.5% excursion in delta 13C that occurred over approximately 10(4) yr and returned to near initial values in an exponential pattern over approximately 2 x 10(5) yr. It has been hypothesized that this excursion signifies transfer of 1.4 to 2.8 x 10(18) g of CH4 from oceanic hydrates to the combined ocean-atmosphere inorganic carbon reservoir. A scenario with 1.12 x 10(18) g of CH4 is numerically simulated here within the framework of the present-day global carbon cycle to test the plausibility of the hypothesis. We find that (1) the delta 13C of the deep ocean, shallow ocean, and atmosphere decreases by -2.3% over 10(4) yr and returns to initial values in an exponential pattern over approximately 2 x 10(5) yr; (2) the depth of the lysocline shoals by up to 400 m over 10(4) yr, and this rise is most pronounced in one ocean region; and (3) global surface temperature increases by approximately 2 degrees C over 10(4) yr and returns to initial values over approximately 2 x 10(6) yr. The first effect is quantitatively consistent with the geologic record; the latter two effects are qualitatively consistent with observations. Thus, significant CH4 release from oceanic hydrates is a plausible explanation for observed carbon cycle perturbations during the thermal maximum. This conclusion is of broad interest because the flux of CH4 invoked during the maximum is of similar magnitude to that released to the atmosphere from present-day anthropogenic CH4 sources.