Carbon Dioxide Through the Ice Ages

Schematic diagram illustrating the mechanism by which air trapped between snow flakes falling in polar regions is slowly trapped in bubbles as the snow is compacted into ice. The ice, which extends for hundreds of feet down, can be drilled using special techniques and taken back to a laboratory. The air, as well as the ice itself, can be extracted from the ice core and chemically analyzed. (After: Delmas R.J., 1992)
The revolution of plate tectonics in the earth sciences was completed in the 1970’s. The next most important discovery concerned climate history, especially the reconstruction of the climate and the chemistry of the atmosphere during periods of extended cooling of the earth, called “ice ages”. The study of the Earth’s past climate, called “paleoclimatology,” can be done a number of ways, but the most direct way to determine the composition of the atmosphere in the past is to analyze “fossil” air trapped in ice cores. Such measurements are possible because polar ice has the unique ability to trap atmospheric gases in bubbles created during its formation, a process that has accumulated a record of climate change dating back hundreds of thousands of years. Early efforts to analyze the ancient air trapped in ice cores drilled in Greenland and Antarctica were international in scope, involving Danish, French, Swiss, Russian and American scientists, and the prominent leaders of this research were Willi Willi Dansgaard (Kopenhagen), Claude Lorius (Grenoble), and Hans Oeschger (Bern). The most comprehensive of this kind of climate data set has recently been published by Petit et al., in the science magazine, Nature. In this article, the climate in Antarctica of the last 400,000 years is documented in detail, along with fluctuations of carbon dioxide and methane.

The results from Petit et al. (as well as earlier ones based on shorter climate records) show that carbon dioxide closely follows the course of climate, as tracked in the deuterium content of the water that made the ice. (The temperature curve is based on deuterium measurements: the colder the snow, the less deuterium in the vapor that made it.) The large range in the fluctuations of carbon dioxide (and methane) is noteworthy, as well as the very rapid rise of these gases during warming when the climate moves from a glacial state to an interglacial state, a process called “deglaciation.” The rapid rise of carbon dioxide and of methane during the periods of deglaciation is of special interest in the present context, because we may find hints as to future developments on a warming Earth. The first statement that needs to be made about these rather abrupt changes in trace gas content is that we are not dealing here with long-term geochemical changes involving large carbon reservoirs of low reactivity. This kind of event calls for involvement of fast-reacting reservoirs with the potential for large impact on the atmosphere. There are two such reservoirs: the ocean's dissolved carbon content ("ocean reservoir") and the seafloor's methane hydrates ("methane ice"). The ocean can be made to give off carbon dioxide to the atmosphere in a number of ways. The first is through physics: changes in temperature and in ocean circulation. By warming the ocean, the solubility of the carbon dioxide is decreased, so the ocean yields gas to the air. Also, by opening a path to the deep ocean some of the excess carbon dioxide stored in deep waters becomes available. Removing sea ice and intensifying deep circulation would work in this direction.

Vostok Ice Core record of variations in air temperature (relative to the current average temperature of –55.5°C at Vostok) and CO2 concentrations from gas bubbles in the ice. (Data from Petit et al., 1999.)
Second, the ocean can alter its biology and chemistry. By changing the rate of precipitation of carbonate and organic carbon in the surface layers of the ocean the chemistry of the sea can be altered to yield additional carbon dioxide. For example, blooms of calcareous algae or increased growth of coral reefs can stimulate the release of carbon dioxide to the air.

Third, there is the general climate change, especially a marked decrease in winds and in the dust brought by winds. By ceasing to provide calcareous dust particles (called "loess") the chemistry of the ocean can become slightly more acidic, which favors the release of carbon dioxide. Also, by ceasing to provide iron or nutrients within the dust, the plankton population of the ocean is reduced, so that carbon is not moved out of the surface layer as readily to deeper layers that do not exchange with the atmosphere.

Was another fast-acting reservoir, methane hydrate, an important contributor to the rise of carbon dioxide during deglaciation? The evidence indicates it was not. If such methane were released from the sea floor, most of it would quickly oxidize into carbon dioxide. Since methane is greatly enriched in the carbon-12 isotope (that is, it has less of the carbon-13 isotope) a corresponding change in the ratio of C-13 to C-12 in the CO2 in the ice core records would be seen if this were true. However, such a significant change is not seen, indicating that methane hydrate was not important factor during deglaciations.

As it stands, the rapid rise of carbon dioxide during the deglaciation periods is unexplained — and not for want of trying by many geochemists. This means, in fact, that we cannot predict how the ocean will react to warming, with regard to emission of carbon dioxide from the sea to the air or a decrease in the uptake of industrial carbon dioxide. All we can say is that, over the last 400,000 years, there seems to have been a positive feedback at work: whenever the climate became warmer, carbon dioxide and methane rose and helped make the climate even warmer.

Some scientists go even further. They say that carbon dioxide rose first, before the warming, and that this is proof that carbon dioxide drives the warming. A rise in carbon dioxide might indeed be the first thing to happen at the beginning of deglaciation. But perhaps the initial rise of carbon dioxide is like the initial gathering of the clouds announcing a storm. The clouds do not make the storm; they show that the process has begun and that the system is ready to change.