“It dawned on us that discussions of sea-level projections...in a warming world seemed to be based on simplistic notions of what would happen if the large polar ice sheets collapsed,” Jerry Mitrovica of the University of Toronto told environmentalresearchweb. “Such discussions, which generally assumed that meltwater would distribute evenly around the globe, include such authoritative documents as the recent fourth assessment report of the Intergovernmental Panel on Climate Change.”

Assuming a uniform distribution of meltwater indicates that melting of the West Antarctic Ice Sheet would cause sea level rise of around 5 m. Mitrovica and colleagues modelled sea-level rise more rigorously; they found that sea level could rise an additional 1.5 m above this figure in North America and the southern Indian Ocean if the whole ice sheet collapsed. On the other hand, South America and Asia could benefit from rises less than 5 m.

“We've been able to isolate and describe the physical processes that give rise to this amplification,” said Mitrovica. “These include rather subtle gravitational, deformational and rotational effects – and they are universal in the sense that they have to be accounted for regardless of which ice sheet is your focus.”

One factor to include is that an ice sheet exerts a gravitational pull on water around it, actually raising sea level locally. Once the ice sheet has gone, removal of this effect tends to reduce sea level over an area extending to about 2000 km away. Rebound of the Antarctic bedrock once the weight of the ice is removed also acts to increase sea level by expelling water from the West Antarctic “hole”. And finally, disappearance of the West Antarctic Ice Sheet would displace the south rotation pole of the Earth by around 500 m, shifting water northwards.

“The most important aspect of the work is to alert scientists and policy makers that to make accurate estimates of the sea level rise that would follow collapse of the Earth's reservoirs of ice will require state-of-the-art sea level models,” said Mitrovica. “It simply isn't good enough to make the simple assumption that things will redistribute evenly – you are bound to underestimate the problem.”

Now the researchers, who reported their work in Science, will extend their analysis to sea level effects of melting of the East Antarctic Ice Sheet, Greenland and mountain glacier systems.