Cosmic conclusions from climatic models: Can they be justified?

Abstract

Climatic models are increasingly being used to answer “cosmic questions” such as the possibility of an ice-covered Earth or a runaway greenhouse effect, or to examine the coevolution of climate and life. Conclusions from these models on such issues, of course, rest on the physical parameterizations of the models. Some of the basic parameterizations are reexamined quantitatively, and it is concluded that presently believed uncertainties in these parameterizations lead to an order-of-magnitude uncertainty in estimates of the sensitivity of the present Earth's climate to external forcings (like a change in solar constant). However, seasonal simulations with present Earth models suggest that estimates of the overall sensitivity of the climate to external forcing may be narrowed (over decadal time scales) to, perhaps, a factor of 2. But the effects of glaciers, continental locations, and atmospheric composition, all of which can change on geological time scales, further enhance the uncertainties in long-term climatic sensitivity estimates from state-of-the-art models. But it is precisely these long-term estimates of climatic sensitivity which support quantitative conclusions on, for example, the possible existence of continuously habitable zones around main-sequence stars. We believe that those who draw cosmic conclusions from climatic models should at least attempt to bracket the final results by repeating their calculations over a plausible range of uncertainty in basic model parameterizations.