The response of phanerozoic surface temperature to variations in atmospheric oxygen concentration

Recently, Poulsen et al. (2015) suggested that O₂ has played a major role in climate forcing during the Phanerozoic. Specifically, they argued that decreased O₂ levels during the Cenomanian stage of the middle Cretaceous (94-100 Ma) could help explain the extremely warm climate during that time. The postulated warming mechanism involves decreased Rayleigh scattering by a thinner atmosphere, which reduces the planetary albedo and allows greater surface warming. This warming effect is then amplified by cloud feedbacks within their 3-D climate model. This increase in shortwave surface forcing, in their calculations, exceeds any decrease in the greenhouse effect caused by decreased O₂. Here we use a 1-D radiative-convective climate model (with no cloud feedback) to check their results. We also include a self-consistent calculation of the change in atmospheric ozone and its effect on climate. Our results are opposite to those of Poulsen et al.: we find that the climate warms by 1.4 K at 35% O₂ concentrations as a result of increased pressure broadening of CO₂ and H₂O absorption lines and cools by 0.8 K at 10% O₂ as a result of decreased pressure broadening. The surface temperature changes are only about 1 K either way, though, for reasonable variations in Phanerozoic O₂ concentrations (10%-35% by volume). Hence, it seems unlikely that changes in atmospheric O₂ account for the warm climate of the Cenomanian. Other factors, such as a higher-than-expected sensitivity of climate to increased CO₂ concentrations, may be required to obtain agreement with the paleoclimate data.