"Reviewing the geologic records of carbon dioxide and glaciations, we found that carbon dioxide was low during periods of long-lived and widespread continental glaciations and high during other, warmer periods," says Dr. Dana L. Royer, research associate in geosciences at Penn State. "Previous suggestions that cosmic ray flux correlated better with ancient temperatures than carbon dioxides do not appear true. While cosmic ray flux may be of some climatic significance, it is likely of second-order importance on a multimillion year timescale."
The researchers looked at climate changes that occurred over the past 570 million years. A direct record of global temperature and carbon dioxide exists for the past 100 years and ice cores provide carbon dioxide information for the past 400,000 years. However, for the remainder of the years, there are no direct measurements.
"A close correspondence between carbon dioxide and temperature has generally been found for the past 570 million years," says Royer. Scientists typically use proxies to determine carbon dioxide and temperatures in the distant past. Oxygen isotope ratios in shallow marine carbonate fossils were used by some researchers to determine surface water temperatures, and this indicated that carbon dioxide and temperature were not correlated, but that cosmic ray fluxes were correlated to temperature. Other proxies can determine carbon dioxide concentrations in both the atmosphere and the oceans.
Royer, working with Robert A. Berner, The Alan M. Bateman professor of geology and geophysics, Yale University; Isabel P. Montanez, professor of geology, University of California Davis; Neil J. Tabor, research associate, Southern Methodist University; and David J. Beerling, professor of animal and plant sciences, University of Sheffield, U.K., compared the results of a variety of carbon dioxide proxies to a model, GEOCARB III, that predicts carbon dioxide over time by tracking carbon entering and leaving the atmosphere. "Proxy estimates of paleo carbon dioxide agree, within modeling errors with GEOCARB model results," the researchers reported in the March issue of GSA Today.
The researchers also found good correlation between low levels of carbon dioxide in the atmosphere and the presence of extensive continental glaciations.
However, the proxy for temperature obtained from shallow oceanic carbonate deposits did not correlate well with the other temperature proxies or the carbon dioxide estimates.
"The acidity of the oceans changes depending on the amount of carbon dioxide in the atmosphere and the amounts of calcium and calcium carbonate in the water," says Royer. "When corrected for acidity, the temperature curve matches the glacial record much better."
The researchers applied correction factors for changes in acidity due to changes in carbon dioxide alone, changes in calcium ions in the water and carbon dioxide in the atmosphere and also for changes in calcium ions, carbon dioxide and calcium carbonate saturation of the water. The corrected temperature curves correctly predicted two major glaciations, one around 300 million years ago and one 30 million years ago. The cosmic ray flux does predict these glaciations, but also predicts cold temperatures when there is no evidence for ice.
"The global temperatures inferred from the cosmic ray flux model do not correlate with the temperature record determined from oxygen isotopes in shallow marine carbonate fossils, when these estimates were corrected for past changes in oceanic acidity," says the Penn State researcher.
The U.S. Department of Energy and the National Science Foundation supported this research.