Discussion of climate change on a range of time scales has tended to focus on carbon dioxide and a changing greenhouse effect. Because carbon dioxide couples climate to ocean, land, and biota, it has appealed to scientists with an interest in the whole Earth system. Carbon dioxide has left a geological record in fossils, isotopes, and sediments, so we can reasonably expect to reconstruct its history. While important questions of detail remain to be resolved, many published applications of carbon cycle modelling suggest that we understand the biogeochemical cycles of carbon well enough to estimate carbon dioxide concentrations in the past and the future. Furthermore, we have an excellent instrumental record of recent changes in atmospheric carbon dioxide. While these considerations make carbon dioxide attractive to paleoclimatologists, they do not necessarily make it a major component of climate change. I shall argue in this paper that clouds deserve much more attention than they have been getting. Copyright 1995 by the American Geophysical Union.
