Surface temperature sensitivity to increased atmospheric CO2

The atmospheric CO2 content has increased significantly over recent decades, mostly as a result of the burning of coal and oil; and it could double in less than a century1–4. Because the CO2 molecule absorbs radiation in the thermal long-wavelength domain, it contributes to the blanketing or greenhouse effect of the atmosphere, and an increase in atmospheric CO2 would be expected to lead to a rise in the mean surface temperature of the Earth1–6. Early estimates indicate that for doubled CO2, the surface temperature would be raised by 1–10 K, but recent numerical experiments with radiative-convective or general circulation models (RCM, GCM) of the atmosphere seemed to narrow this range to 2–4 K (see refs 2–6). However, Newell and Dopplick7, and Idso8, argue that these results are inconsistent with the observed or computed values of certain terms of the surface energy budget, which seem to indicate a far smaller sensitivity. Here, I examine all of the terms of the surface energy budget7–11 and find that there is no basic disagreement between the surface energy budget approach and the more elaborate model experiments, provided that there is a clear distinction between external perturbation and atmospheric feedback, and that equivalent representations of surface processes are used. However, the temperature and humidity dependence of the terms of the surface energy budget shows that the surface temperature sensitivity depends critically on the way in which the atmospheric humidity evolves, and on the degree of compensation in the perturbation of the latent and sensible heat fluxes.