Previous estimates of how much the world's sea level will rise as a result of global warming may have seriously underestimated the problem, according to new research.
The study, published in Science, uses a new "semi-empirical" method instead of relying purely on computer modelling. While some modelling significantly underestimates the amount of sea-level rise that has already been seen over the last century, the new method matches the observed rise very closely, says Stefan Rahmstorf, at the Potsdam Institute for Climate Impact Research, Germany, who conducted the new study.
The existing computer model deviates even more from the actual observations built into the new estimates included in a draft of the next report of the Intergovernmental Panel on Climate Change, due to be released in February 2007.
The draft report says newer climate models now suggest a rise only half as great as projected in the previous IPCC report. But that draft may be revised before its release to reflect the new research that suggests the rise will be greater than the IPCC's previous estimate, Rahmstorf told New Scientist.Capital cities
For a given amount of warming, Rahmstorf says, the rise in sea level "could well be twice as much as was so far expected, based on the last IPCC report".
At the top of the range of possible temperature rises estimated by the last IPCC report, the rise could be as great as 140 centimetres by 2100. That would be bad new for cities like London and New York, which lie close to sea level, and would leave them facing an increased risk of devastating storm surges. Even the lowest predicted temperature rises would cause a 50 cm rise, Rahmstorf says.
The predictions in the previous IPCC report - its third - ranged from 9 cm to 88 cm by 2100, and the initial draft of the next report was to cut those figures in half. But Rahmstorf, who is a lead author of the paleoclimate section of the upcoming report, says he hopes his new results will be incorporated before IPCC 4 is officially released in February 2007.Search for meaning
Rahmstorf says there are so many possible factors and feedback mechanisms that affect sea level that it is almost impossible to derive a meaningful model of future rises from purely physical modelling. Instead, he uses a method similar to that used for calculating tide tables.
The method relies on actual observations of past changes in sea level, and their correlation with temperature changes, to derive an estimate of the amount of increase expected for a given temperature change.
Rahmstorf acknowledges that the simple linear extrapolation derived using the new semi-empirical method will not hold good over a timescale of millennia, but he argues that it is a good approximation for the next century. However, the strongest conclusion of the new work, he says, is that uncertainties in sea level rise predictions are far greater than expected.
"We should not take this risk," Rahmstorf says. "We should start with very effective emission reduction measures. The global temperature increase should be kept to under 2๏ฟฝC."
"We still have some work to do to improve our comprehensive physical models, especially for ice sheets," says Richard Alley, at Penn State University, who specialises in ice sheets and glaciers. "But given the difficulties with modelling ice sheets etc., Rahmstorf's approach is clever and useful."
Journal reference: Science (DOI: 10.1126/science.1135456)
Reader Comments
to our Newsletter