BERKELEY, CA —Studies have shown that global climate change can set-off positive feedback loops in nature which amplify warming and cooling trends. Now, researchers with the Lawrence Berkeley National Laboratory (Berkeley Lab) and the University of California at Berkeley have been able to quantify the feedback implied by past increases in natural carbon dioxide and methane gas levels. Their results point to global temperatures at the end of this century that may be significantly higher than current climate models are predicting.

Science image spacer image
Margaret Torn and John Harte estimated the carbon dioxide and methane released into the atmosphere in response to past global warming trends, applied the results to current climate change models, and found a much hotter future in store for Earth in the 21st century than has gnerally been predicted. (Photo by Roy Kaltschmidt, Berkeley Lab)

Using as a source the Vostok ice core, which provides information about glacial-interglacial cycles over hundreds of thousands of years, the researchers were able to estimate the amounts of carbon dioxide and methane, two of the principal greenhouse gases, that were released into the atmosphere in response to past global warming trends. Combining their estimates with standard climate model assumptions, they calculated how much these rising concentration levels caused global temperatures to climb, further increasing carbon dioxide and methane emissions, and so on.

“The results indicate a future that is going to be hotter than we think,” said Margaret Torn, who heads the Climate Change and Carbon Management program for Berkeley Lab’s Earth Sciences Division, and is an Associate Adjunct Professor in UC Berkeley’s Energy and Resources Group. She and John Harte, a UC Berkeley professor in the Energy and Resources Group and in the Ecosystem Sciences Division of the College of Natural Resources, have co-authored a paper entitled: Missing feedbacks, asymmetric uncertainties, and the underestimation of future warming, which appears in the May, 2006 issue of the journal Geophysical Research Letters (GRL).

In their GRL paper, Torn and Harte make the case that the current climate change models, which are predicting a global temperature increase of as much as 5.8 degrees Celsius by the end of the century, may be off by nearly 2.0 degrees Celsius because they only take into consideration the increased greenhouse gas concentrations that result from anthropogenic (human) activities.

“If the past is any guide, then when our anthropogenic greenhouse gas emissions cause global warming, it will alter earth system processes, resulting in additional atmospheric greenhouse gas loading and additional warming,” said Torn.

Torn is an authority on carbon and nutrient cycling in terrestrial ecosystems, and on the impacts of anthropogenic activities on terrestrial ecosystem processes. Harte has been a leading figure for the past two decades on climate-ecosystem interactions, and has authored or co-authored numerous books on environmental sciences, including the highly praised Consider a Spherical Cow: A Course in Environmental Problem Solving.

In their GRL paper, Torn and Harte provide an answer to those who have argued that uncertainties in climate change models make it equally possible that future temperature increases could as be smaller or larger than what is feared. This argument has been based on assumptions about the uncertainties in climate prediction.

However, in their GRL paper, Torn and Harte conclude that: “A rigorous investigation of the uncertainties in climate change prediction reveals that there is a higher risk that we will experience more severe, not less severe, climate change than is currently forecast.”

spacer image Science image
The world’s glaciers are shrinking as a result of rising temperatures and will eventually cause sea-levels to swell. Studies have shown that global warming triggers natural feedback loops that can amplify the warming’s magnitude. (Photo courtesy of NOAA)

Serious scientific debate about global warming has ended, but the process of refining and improving climate models – called general circulation models or GCMs – is ongoing. Current GCMs project temperature increases at the end of this century based on greenhouse gas emissions scenarios due to anthropogenic activities. Carbon dioxide in the atmosphere, for example, has already climbed from a pre-industrial 280 parts per million (ppm) to 380 ppm today, causing a rise in global temperature of 0.6 degrees Celsius. The expectations are for atmospheric carbon dioxide to soar beyond 550 ppm by 2100 unless major changes in energy supply and demand are implemented.

Concerning as these projection are, they do not take into account additional amounts of carbon dioxide and methane released when rising temperatures trigger ecological and chemical responses, such as warmer oceans giving off more carbon dioxide, or warmer soils decomposing faster, liberating ever increasing amounts of carbon dioxide and methane. The problem has been an inability to quantify the impact of Nature’s responses in the face of overwhelming anthropogenic input. Torn and Harte were able to provide this critical information by examining the paleo data stored in ancient ice cores.

“Paleo data can provide us with an estimate of the greenhouse gas increases that are a natural consequence of global warming,” said Torn. “In the absence of human activity, these greenhouse gas increases are the dominant feedback mechanism.”

In examining data recorded in the Vostok ice core, scientists have known that cyclic variations in the amount of sunlight reaching the earth trigger glacial-interglacial cycles. However, the magnitude of warming and cooling temperatures cannot be explained by variations in sunlight alone. Instead, large rises in temperatures are more the result of strong upsurges in atmospheric carbon dioxide and methane concentrations set-off by the initial warming.

Using deuterium-corrected temperature records for the ice cores, which yield hemispheric rather than local temperature conditions, GCM climate sensitivity, and a mathematical formula for quantifying feedback effects, Torn and Harte calculated the magnitude of the greenhouse gas-temperature feedback on temperature.

Science image
The ice core drilled at the Russian Vostok station in East Antarctica reaches a depth of more than two miles and provides information about climate cycles over hundreds of thousands of years. (Schematic courtesy of John Priscu research group, Montana State University)

“Our results reinforce the fact that every bit of greenhouse gas we put into the atmosphere now is committing us to higher global temperatures in the future and we are already near the highest temperatures of the past 700,000 years,” Torn said. “At this point, mitigation of greenhouse gas emissions is absolutely critical.”

The feedback loop from greenhouse gas concentrations also has a reverse effect, the authors state, in that reduced atmospheric levels can enhance the cooling of global temperatures. This presents at least the possibility of extra rewards if greenhouse gas levels in the atmosphere could be rolled back, but the challenge is great as Harte explained.

“If we reduce emissions so much that the atmospheric concentration of carbon dioxide actually starts to come down and the global temperature also starts to decrease, then the feedback would work for us and speed the recovery,” Harte said. “However, if we reduce emissions by an amount that greatly reduces the rate at which the carbon dioxide level in the atmosphere increases, but don’t cut emissions back to the point where the carbon dioxide level actually decreases, then the positive feedback still works against us.”

This research was supported by the U.S. Department of Energy’s Climate Change Research Division and by the National Science Foundation.

Berkeley Lab is a U.S. Department of Energy National laboratory located in Berkeley, California. It conducts unclassified scientific research and is managed by the University of California. Visit our Website at

Additional Information