EDIT
By 2100, spring and summer temperatures in the Arctic could reach levels that trigger an unstoppable repeat performance, they say. Over several centuries, the melt could raise sea levels by as much as 20 feet, submerging major cities worldwide as well as chains of islands, such as the present-day Bahamas. The US would lose the lower quarter of Florida, southern Louisiana up to Baton Rouge, and North Carolina's Outer Banks. The ocean would even flood a significant patch of California's Central Valley, lapping at the front porches of Sacramento.
These estimates may understate the potential rise. The teams say their studies provide the first hints that during the last interglacial period, ice sheets in both hemispheres worked together to raise sea levels, rather than the Northern Hemisphere's ice alone. This raises concerns that Antarctic melting might be more severe this time, because additional melt mechanisms may be at work.
"It sounds bad," acknowledges Jonathan Overpeck, a University of Arizona researcher who led one of the two studies. He notes that rising temperatures are approaching a threshold. But "we know about it far enough in advance to avoid crossing it." The challenge, he and others say, is to take advantage of the remaining window by reducing emissions of greenhouse gases substantially.
EDIT
The new results aren't the end of the story. The researchers will refine the models, improve the measurements, and find other sources of data to verify the modeling. Coral data pointing to sea-level changes in the last warm period remain controversial, the team acknowledges. And the team's assumption that the amount of carbon dioxide would triple by 2100, although moderate among climate forecasts, is not a done deal. It depends on how quickly industrial and developing countries adopt low-emission technologies and take long-term steps to reduce greenhouse gases. But the window for action is relatively short, Dr. Overpeck says. CO2 remains in the atmosphere for more than a century after it's first emitted. And it takes time to implement policies and adopt technologies. Thus for all practical purposes, the tipping point may come sooner than atmospheric chemistry would suggest.
EDIT
http://www.csmonitor.com/2006/0324/p01s03-sten.html