Burgeoning interest in paleoclimatology has been spurred by growing awareness of the control of paleoclimates on the formation of economic deposits. tn past studies, paleoclimatic patterns were derived empirically from biogeographic patterns and, to a lesser extent, from the distributions of sedimentary paleoclimatic indicators, such as coals. The problems with this approach are numerous. In early studies, the paleoclimatic patterns appeared to make very little sense because they were reconstructed on modern continental positions. Even after the acceptance of continental drift, problems arose when the paleoclimatic indicators were poorly dated or when geologists chose paleoclimatic indicators from too long a time period, during which major paleoclimatic changes could have occurred. Biogeographic patterns commonly proved too coarse for understanding all but the broadest paleoclimatic controls, either because the data are insufficient to permit fine biogeographic divisions or because the paleoclimatic significance of extinct organisms (that is, almost all those in the fossil record) is unknown.
More recently, qualitative and quantitative models of paleoclimate have proved useful for understanding the distributions of climatically significant geologic data. These models are founded on basic principles of atmospheric and oceanic circulation as applied to global paleogeography, including reconstructed plate positions. Using geologic data to reconstruct paleoclimates leaves no way of independently testing the hypothesized paleoclimatic patterns. Withclimate models, geologists can formulate hypotheses about the paleoclimatic patterns that might be expected during various intervals in Earth history and test those hypotheses with the paleoclimatic indicators. Models also provide a framework into which scarce data can be fitted.