It would take 7400 years to melt Greenland at currently observed ice melting rates, and many hundreds of years even with a non-linear increase in melting rates.
Numerous studies and direct observation show that Greenland’s ice sheet is melting, and the latest studies show that snowfall on Greenland is not sufficient to offset this loss. If the entirety of Greenland’s ice sheet were to melt, sea levels worldwide would rise by roughly 20 feet. This rise would severely damage or wipe out many of the world’s major cities, and accordingly the threat of catastrophic sea level increase is listed among the primary threats of global warming.
This brings to light an important question: how long do we have until Greenland’s ice sheet melts, or until its melting causes a significant amount of sea level rise?
Greenland’s ice sheet has a total volume of 2.6 million kilometers cubed, and according to a recent study, lost an average of 222 km cubed of ice in the last couple of years. In 2007, the year that Arctic ice cover set a new summertime minimum, 350 km cubed of ice was lost. Assuming this extreme rate continues, Greenland will be fully melted in 7400 years. Of course, ice melt rates are non-linear, and so this rate could increase considerably as temperatures rise.
Assume for a moment that the rate of ice melt rises by an order of magnitude (10x) beyond the 2007 record. 740 years of melting would still be required to melt all of Greenland, and even at that rate sea levels would rise only 2 feet by 2100! If the ice immediately begins melting at 100 times the record rate, then a real catastrophe would ensue, as sea levels would rise by the full twenty feet before the end of the century.
What would it take to trigger a catastrophic increase in Greenland’s ice melt rate? Unfortunately, this isn’t yet well understood, with some research showing that Greenland’s ice sheet could tolerate significantly higher temperatures. This is based in part on research showing that during the last interglacial period, 125,000 years ago, about half of the Greenland ice sheet persisted despite temperatures 5C higher than today. It’s a good idea to keep this and the melt rate calculations in mind when weighing the threat of AGW-induced sea level rise against potential solutions.
True Cost - Analyzing our economy, government policy, and society through the lens of cost-benefit 
Thanks for this look at the numbers. However, you don’t support this claim:
“It would take 7400 years to melt Greenland at currently observed ice melting rates, and many hundreds of years even with a non-linear increase in melting rates.”
Instead of looking at a linear melt and a non-linear melt, you look at the present-day melting and a faster melting. Both are linear.
Compare this graph: http://www.columbia.edu/~mhs119/IceSheet/Milankovic_Fig8.gif
Reasonable non-linear estimates, (if the melting proves to be non-linear, which it may not be, or not extremely) based on the recent melt rate, include doubling times of five, seven, or ten years.
With a doubling time of five years, Greenland melts at a thousand times the current rate in 2060, and the whole ice sheet is gone a few years after that. At seven years, the ice sheet is gone by 2080 or so. With a ten year doubling, you’re looking at maybe 2110.
I don’t think the paleoclimate record supports the idea that the whole ice sheet will behave nonlinearly. But if, say, half of it is responding non-linearly, while the other half is isolated in the interior, melting much more slowly, that could still be a huge problem — 10ft of sea level rise in a matter of decades.
How fast sea levels might rise remains an active area of debate, as the models of ice sheet behavior have so far proven inadequate.
TheTracker, thanks for the comment. If indeed the melt rate is non-linear and follows short doubling times measured in years or decades, then that’s potentially troubling.
I’m curious as to whether there’s consensus in the literature on rate of mass change (in both Greenland and Antarctica) – I recall that some studies were inconclusive, showing melting at the edges but increased snow accumulation on top of the ice sheets.
What the author forgets to mention, is that an ice age is due before the Greenland ice sheet could melt.
It survived periods with temperatures approximately 4ºC higher than the current temperature and I imagine that much of it will still be around when the next ice age gets underway.
As for the Antarctic ice sheet, an Australian would say, “No worries, mate.”
That would be a pretty dumb Australian. The WAIS is highly unstable and already contributing significantly to sea level rise.
This is a good point, it will take the Greenland Ice Sheet a long time to melt. The key for an ice sheet to lose mass fast is via calving. The marine terminating outlet glaciers of the GIS have accelerated and are delivering more icebergs to the ocean, this allows the melting to take place in the ocean. The http://glacierchange.wordpress.com/2009/06/28/recession-of-jakobshavn-isbrae/ Jakboshvan Isbrae is the most significant outlet glacier and it has accelerated and retreated considerably since 2001, and is constantly calving icebergs. Solarkent the moulin method of meltwater enhanced flow is not very powerful and not the reason for recent glacier acceleration of GIS glaciers. However, the key mechanism a reduction in backforces at the calving front is much more powerful. http://www.realclimate.org/index.php/archives/2008/04/moulins-calving-fronts-and-greenland-outlet-glacier-acceleration/#more-550
I can’t disagree with your comments. As to the statement “catastrophic increase” , I would have thought the risk there would be associated with meltwater feeding through cracks in the ice, accumulating and lubricating leading to the risk of large sections of ice slipping into the sea. Possibly a very very small fraction of the overall ice sheet, but imagine a slippage that was made up of 1% of the ice sheet!
Perhaps a possibility best reserved for science fiction …