<img alt=“nat%20geo%20retr.bmp” src=“http://blogs.nature.com/climatefeedback/nat%2520geo%2520retr.bmp” width=“200” height=“196” align=“right”//>Scientists have last week retracted a study which, based on simulations of the past 22,000 years, had projected a 21st century global sea level rise between 7 and 82 centimetres. The authors say they no longer have confidence in the projections owing to serious mistakes in their model approach.
The results, published last year in Nature Geoscience, had been roughly consistent with projections by the International Panel on Climate Change (IPCC) which in its 2007 report gave a range of 18-59 centimetres sea level rise by 2100. However, some scientists caution that the IPCC numbers, which exclude the effect of changing Greenland and Antarctic ice sheets, could prove too conservative.
Mark Siddall and his co-authors (including Thomas Stocker, co-chair of the IPCC’s working group on the physical basis of climate change) had used an empirical model linking sea-level rise to changes in global mean temperature. They had their model move forward in discreet 100-year time-steps which, although sufficient for simulating the past, they later found too coarse for reliably projecting future changes.
The study also failed to properly take into account uncertainties in temperature reconstructions of the last 2,000 years.
It is unclear if the resulting projections over- or underestimate future sea level rise, says Siddall.
The mistakes, he says, are too profound to be dealt with in a mere correction to the initial paper.
The retracted paper has been “totally independent” from studies reviewed by the IPCC for its last report, and will not affect future work by the IPCC, he says.