Carbon Dioxide Levels Key To Global Warming Predictions

Despite incomplete agreement of computer generated models and physical evidence, Crowley believes carbon dioxide levels still prove key to predicting future climate events based on past history.

"Recently, some researchers have suggested that carbon dioxide changes are not primarily responsible for past climate changes over millions of years," said Crowley, a professor of oceanography at Texas A&M University. "But my co-author, Robert Berner of Yale, and I demonstrate in our perspective piece for Science that variations in CO2 are in fact very important for explaining past ice ages.

"Conclusions to the contrary have been based on fragmentary data," he observed. "I think it would be hazardous to conclude, based on discrepancies between models and data from the past, that projections interpreting the negative impact of future anthropogenic greenhouse gas emissions are erroneous."

Paleoclimate experts agree that over millions of years, the Earth's climate has undergone massive changes. Glaciers sculpted the face of entire continents, followed by warm periods with virtually no ice. Over 65 million years ago, dinosaurs grazed in Alaska, and the ice-cold waters of the ocean depths ran lukewarm.

Experts estimate historic atmospheric CO2 content based on analysis of fossilized soils, marine sedimentary carbon, fossil leaves and boron isotopes in carbonate fossils. Such indicators have suggested links between natural variations in the amount of CO2 present in the atmosphere and long-term climate changes.

But not all the data agree. For example, during certain periods of high levels of CO2, reconstructed temperatures in the tropics were quite cold.

Crowley and Berner dispute the idea, however, that such discrepancies should entirely call into question the role of CO2 in global warming, both now and in the past.

"There are some legitimate reasons for believing the tropical chemical data may give flawed estimates of temperature," Crowley said. "If these estimates are, however, eventually proven correct, then perhaps climate models are not correctly simulating tropical ocean responses to carbon dioxide changes."

"But even if there are mistakes in our modeling approaches, they may still have limited application to future greenhouse warming," he continued. "The locations of the continents were so different in the past from where they are now that we cannot be sure of the effects of such continental changes on the ocean."

Crowley and Berner's analysis of models and supporting data led them to conclude that the CO2 model is valuable for periods of glaciation at high latitudes, but that in the tropics, predictive applications of prior climate change are complicated by movement of the continents on tectonic plates and problems with interpretation of climate indicators.

"The bottom line remains the same -- CO2 is still very important to the whole process of climate change. We just don't have all the story yet," Crowley said.