Letters to Nature

Nature 408, 698-701 (7 December 2000) | doi:10.1038/35047044; Received 7 July 2000; Accepted 2 October 2000

Evidence for decoupling of atmospheric CO2 and global climate during the Phanerozoic eon

Ján Veizer1, Yves Godderis2 & Louis M. François2

  1. Institut für Geologie, Mineralogie und Geophysik, Ruhr Universität, 44780 Bochum, Germany, and Ottawa-Carleton Geoscience Centre, University of Ottawa, Ottawa, ON K1N 6N5, Canada
  2. Laboratoire de Physique Atmosphérique et Planétaire, Université de Liège, 4000 Liège, Belgium

Correspondence to: Ján Veizer1 Correspondence and requests for materials should be addressed to J.V. (e-mail: Email: veizer@science.uottawa.ca).

Atmospheric carbon dioxide concentrations are believed to drive climate changes from glacial to interglacial modes1, although geological1, 2, 3 and astronomical4, 5, 6 mechanisms have been invoked as ultimate causes. Additionally, it is unclear7, 8 whether the changes between cold and warm modes should be regarded as a global phenomenon, affecting tropical and high-latitude temperatures alike9, 10, 11, 12, 13, or if they are better described as an expansion and contraction of the latitudinal climate zones, keeping equatorial temperatures approximately constant14, 15, 16. Here we present a reconstruction of tropical sea surface temperatures throughout the Phanerozoic eon (the past approx550 Myr) from our database17 of oxygen isotopes in calcite and aragonite shells. The data indicate large oscillations of tropical sea surface temperatures in phase with the cold–warm cycles, thus favouring the idea of climate variability as a global phenomenon. But our data conflict with a temperature reconstruction using an energy balance model that is forced by reconstructed atmospheric carbon dioxide concentrations18. The results can be reconciled if atmospheric carbon dioxide concentrations were not the principal driver of climate variability on geological timescales for at least one-third of the Phanerozoic eon, or if the reconstructed carbon dioxide concentrations are not reliable.