Technology News - October 6, 2004

Positive feedbacks shaping climate-change forecasts

Positive feedbacks between the climate and the carbon cycle are now being incorporated into general atmospheric circulation models, with surprising results. Rudimentary calculations show that positive feedbacks from events such as land-use changes could boost the concentration of atmospheric CO₂ to 980 parts per million (ppm) by 2100, which is 40% more than the 700 ppm predicted without feedbacks.

Until now, feedback mechanisms have not been well represented in the increasingly sophisticated circulation models. But recent breakthroughs are bringing scientists closer to an accurate estimate of future carbon emissions, according to research presented at the Ecological Society of America (ESA) meeting held in Portland, Ore., in August.

Most climate models assume that atmospheric CO₂ concentration is a function of emissions, mainly from fossil-fuel burning, minus the amount of CO₂ soaked up by the ocean and locked by photosynthesis into plant tissue on land, says Chris Field, director of global ecology at the Carnegie Institution of Washington at Stanford University and author of the ESA research. These models can accurately reproduce past and present climate. However, the warming that is predicted to accompany rising CO₂ concentrations is expected to lead to changes in ocean circulation that reduce the ocean’s capacity to absorb CO₂. Meanwhile, changes on land, such as increased soil respiration, are expected to lead to the release of more CO₂. These changes could increase warming and trigger more releases of CO₂ in an upward spiral, he says.

Wildfires in tropical and boreal forests, melting of the permafrost, and drying and decomposition of wetlands are just a few examples of mechanisms that could further amplify positive feedbacks to climate warming, Field says. He and his colleagues applied a risk analysis to these and other carbon pools, which store 2–5 times more carbon than is now in the atmosphere. They found that up to 100 petagrams (Pg) of carbon could be lost to the atmosphere over the next 20 years and up to 1350 Pg over the next century. The loss from the permafrost alone could equal half as much carbon as the amount in today’s atmosphere, he says.

Two leading atmospheric models have taken a stab at incorporating some of these feedbacks—a model from the Institut Pierre-Simon Laplace (IPSL) in France, which forecasts that feedbacks would elevate CO₂ levels to 780 ppm by 2100, and the model from the U.K.’s Hadley Centre for Climate Prediction and Research (HCCPR), which predicts CO₂ levels of 980 ppm by 2100, Field says. Because the 200-ppm discrepancy between the two predictions is significant, scientists are working to refine the models, and new research is pushing the field ahead, he says.

However, reaching consensus may not be easy. The IPSL model was recently improved to take account of vegetation dynamics, such as the northward migration of deciduous forests with a warming climate, says biogeochemist Pierre Friedlingstein at the Laboratory of Climate and Environmental Science in France. The vegetation dynamics failed to alter the IPSL model’s outcome, because the carbon lost by dying tropical forests is taken up in the mid- and high-latitudes where more carbon-efficient temperate forests are replacing the boreal forest, Friedlingstein says. The lack of vegetation dynamics in the IPSL model had once been thought to partially account for the difference between it and the HCCPR model, which did include vegetation dynamics.

Another major difference between the two models is their treatment of tropical forests: The HCCPR model predicts a large release of carbon from dying trees in a warmer and drier climate, and the IPSL model maintains tropical rainforests as a carbon sink, although a more modest sink than at present, Friedlingstein says.

“One of the main uncertainties we have right now in the feedback models is the role of temperature, drought, and fire,” says Jim Randerson, biogeochemist at the University of California, Irvine. New results suggest that fire, especially in the tropics, is one of the key feedbacks between the biosphere and the atmosphere, he says. Using carbon monoxide as a tracer, he found that two-thirds of the increase in CO₂ during the 1997–1998 El Niño event was caused by drought-driven fires. Randerson and his colleagues are now working on a model to forecast fires on the basis of trends in temperature, moisture, lightning strikes, and human population density.

Although interactions between the climate, drought, and fires have not yet been represented in the atmospheric models, scientists have begun to nail down some of these feedbacks, says Dan Nepstad, forest ecologist at the Woods Hole Research Center in Massachusetts. In the Amazon forest in Brazil, fire begets fire by opening the tree canopy, thereby drying the understory and making a lightly burned area more vulnerable to subsequent fires, he says. Humans also play an important role by clearing land. This, in turn, robs the atmosphere of moisture and reduces local rainfall, which increases the chance of fires spreading and creates more local drought. Inhabitants respond by clearing even more land.

Although smoke in sparse concentrations in remote locations can stimulate rainfall, new research shows that when dense clouds of smoke from forest fires hang over large areas for weeks, such as in the Mato Grosso region in the southeastern Amazon, rain is inhibited. The smoke contains excess condensation nuclei that form water drops that are too small to fall as rain, Nepstad explains. This then increases the likelihood of more fires, he notes.

“The tropics play a huge role in carbon, and we have signals that the feedbacks are significant,” Nepstad says. For instance, estimates reveal that more than 1 billion tons of carbon were emitted from Borneo’s peat forests during the drought-driven fires of 1998, double the 500 million tons of carbon per year targeted to be cut by the international Kyoto Protocol, Nepstad points out.

“The bottom line is that scientists are finding that most of the feedbacks are positive, especially if the warming becomes substantial. As a consequence, CO₂ emissions will have to be ratcheted back even more, if we are to avoid dangerous interference with the climate system,” Field concludes. —JANET PELLEY