Carbon dioxide indirectly causes up to 50 percent more global warming than originally thought, a finding that raises questions over targets for stabilising carbon emissions over the long term, a study said on Sunday.
In a paper published in the journal Nature Geoscience, British scientists said a tool commonly used in climate modelling may have badly underlooked the sensitivity of key natural processes to the warming caused by CO2.
As a result, calculations for man-made global warming on the basis of carbon emissions may be underpitched by between 30 and 50 percent, they said.
The study was coincidentally published on the eve of a 12-day UN conference in Copenhagen aimed at providing a durable solution to the greenhouse-gas problem.
The authors stressed that the more-than-expected warming would unfold over a matter of hundreds of years, rather than this century.
The findings do not mean that the predictions for temperature rise by 2100, established notably by the UN's Intergovernmental Panel on Climate Change (IPCC), should be rewritten, they said.
"We don't want to be overly alarmist here," said lead author Dan Lunt of Britain's University of Bristol.
"But if people are thinking about stabilising CO2 at a certain atmospheric level, or putting together a treaty, or having a debate about what the levels should be, it really is important to know what the long-term consequences of those emissions are going to be, because CO2 hangs around for so long."
Lunt and colleagues decided to test a widely-used climate model on an epoch called the mid-Pliocene warm period, about three million years ago, when Earth heated up in response to natural processes.
Cores drilled from ocean sediment provide a good idea about atmospheric carbon levels and temperature at the time.
What the team found, though, was that the CO2 levels in the Pliocene - around 400 parts per million (ppm) - were not consistent with the warming, which was around three degrees Celsius (5.4 degrees Fahrenheit) higher than today.
The difference could only be fully explained by the long-term loss of icesheets and and changes in vegetation, says the paper. These changes cause Earth's surface to absorb more solar radiation, which causes more warming, and so on.
When applied to what awaits us this century, the adjusted model suggests that nothing significantly different will happen compared to what has already been estimated.
"In that time scale, we don't think the Greenland icesheet is going to melt completely or that East Antarctica will melt. That was what we saw in the model for three million years ago, but it is unlikely to take place in the next century," said Lunt.
Where it poses a dilemma, though, is how to fix a target for stabilising CO2 emissions so that future generations, centuries from now, are not hit by this long-term warming mechanism.
A popular goal is to limit warming since pre-industrial times to 2 C (3.6 F), a figure that in mainstream climate models typically equates to about 450 ppm. At present, Earth's CO2 concentrations are at around 387 ppm.
Lunt says that today's level may already be too high in this context.
"Our work says that at 400 parts per million, you are looking at more than two degrees C [3.6 F].
"To stabilise at two degrees C, you would have to aim for something like 380 ppm. But remember, this is the sort of level that applies if you want a long-term commitment that goes on for centuries, for generations to come."Reuse content