Colorado State University Scientist to Spearhead International Effort in Search of Northern Hemisphere’s Missing Carbon

As nations of the world continue to wrangle over how to cut carbon dioxide and other greenhouse gas emissions under the Kyoto Protocol, Scott Denning of Colorado State University and an international group of scientists are planning an experiment aimed at resolving the ongoing controversy over the unexplained removal of a large amount of carbon dioxide from the Earth’s atmosphere, a process known as the "missing sink."

Carbon dioxide levels in the atmosphere are known to be higher in the Northern Hemisphere than in the Southern Hemisphere due to higher rates of man-made carbon emissions in the north. But scientists have been puzzled for several years about why measured CO2 levels in the north are so much lower than model predictions based on estimates of carbon emissions.

"Something in the Northern Hemisphere is sucking up the carbon," said Kevin Gurney, Colorado State research associate who will discuss the international project at the American Geophysical Union’s meeting here next week. "This ‘missing sink’ isn’t new, but where it is located and how it works are still unresolved."

In a recent paper in Science (S. Fan, et al., Oct. 16, 1998), researchers claimed to have located a large carbon "sink" in North America, a surprising result because the United States is the world’s biggest greenhouse gas producer. The modeling technique used to reach this conclusion – an "inversion calculation" – is the focus of the new international project called TransCom3 to be launched at a workshop Dec. 4-5 in San Francisco.

"The new Science paper is an intriguing result and shows the potential use of this tool," said Denning, assistant professor of atmospheric science at Colorado State, "but it is too early to draw definitive conclusions." Denning is leading the three-year project that will seek definitive answers.

Not only has the location of the missing carbon sink eluded researchers, Denning said, but finding out exactly what is absorbing the excess carbon also remains an area of active scientific debate. Current theories include increased CO2 uptake by plants caused by more CO2 in the atmosphere, a longer growing season in northern latitudes resulting from climate change and the regrowth of forests in Europe and the northeastern United States.

Twelve of the leading research teams in carbon transport modeling from the United States, Germany, France, the United Kingdom, Japan and Australia are participating in TransCom3. These researchers use a worldwide database of CO2 measurements at 75 sites and combine it with information about global winds in a numerical model to infer where CO2 comes from and where it is absorbed.

"We run the logic backwards," Denning said, "essentially figuring out where the sinks are from the CO2 patterns downwind" of them.

This type of calculation is very difficult, according to Denning, because the data on CO2 is sparse and regional variation in CO2 levels is so small – often less than 1 percent. Researchers currently are obtaining very different results from essentially the same data. For example, some recent studies suggest that the Northern Hemisphere’s CO2 sink is much more evenly distributed across the North American and Eurasian continents, Denning said.

Research teams participating in the TransCom3 experiment will apply their models to a standardized problem. The results will be collected and analyzed at Colorado State. Denning hopes the experiment will improve the ability of the models to calculate the global carbon budget of the atmosphere and narrow the differences in results by different modelers.

Initial work on the TransCom project has been supported by NASA’s Earth Observing System Project. The San Francisco workshop is sponsored by the International Geosphere Biosphere Program’s Global Analysis, Interpretation, and Modeling task force.

"What is ultimately needed to solve this problem is better sampling of CO2 in the atmosphere," said Denning, noting that the current sampling system of 75 stations has very few sites over the continents.