The Greenland and Antarctic Ice Sheets (see figures) contain about 80% of the Earth's freshwater, around 99.5% of the glacier ice on the planet and cover 10% of the land surface. If they were to melt completely they would raise global sea level by some 70 m. Two processes control the growth or decay of an ice sheet, otherwise known as its mass balance. The first is surface mass exchange via snowfall and, in the case of Greenland, melting and runoff of the meltwater. This process responds almost instantaneously to a change in factors such as precipitation or air temperature. The second parameter is the flow of the ice, or ice dynamics, which delivers ice from the interior to the oceans via iceberg calving. This is controlled, in the main, by processes deep within the ice sheet and at its base, where it is in contact with the bedrock below. Ice in Antarctica is up to 5 km thick, and up to 3 km thick in Greenland. So a change in conditions at the surface takes thousands of years to influence the ice flow at depth. As a result, the "conventional" view was that ice dynamics respond very slowly to external changes - such as the climate.
But results obtained from satellite observations over roughly the last fifteen years suggest that the ice sheets are responding to recent warming of the surrounding ocean and atmosphere at rates far exceeding expectations, and in ways that have surprised glaciologists. In one particularly noteworthy case one of the largest outlet glaciers in Greenland - Jakobshavn Isbrae - doubled in velocity from around 6 km/yr in 1997 to 12 km/yr just five years later. The increase was linked to the loss of part of the glacier tongue - or ice shelf - that floats on the surface of the ocean. The glacier tongue is susceptible to melting both from below by oceanic warming and from above by increased atmospheric temperatures. This one outlet glacier (labelled, JI in the Greenland figure) was responsible for increasing the flux of ice into the ocean by 25 cubic km/yr between 1997 and 2002. For comparison, this volume of ice would be enough to supply 400 million people in the developed world - more than the population of the US - with their domestic water needs and represents 4% of the 20th Century rate of sea level rise alone.
Was this event a "one off": a localised internal instability that was perhaps unique to this outlet glacier? Or should we expect to see this kind of behaviour elsewhere as the climate warms? Worryingly, it looks like the answer is the latter. Several other outlet glaciers in Greenland, not directly linked to Jakobshavn Isbrae, dramatically increased in speed at around the same time. Also concerning is that several glaciers in West Antarctica, at the other end of the planet, have also speeded up recently. In this case, the likely cause appears to be warming by the ocean of the floating ice shelves in front of the glaciers. Numerical models of ice sheet dynamics did not predict changes of this magnitude or this speed; these observations are resulting in a paradigm shift in how we view the interaction of ice sheets with the atmosphere and ocean.
The fourth assessment report of the Intergovernmental Panel on Climate Change (IPCC) proposed a range for sea level rise of 18-59 cm by 2100. This estimate did not include, however, any prediction for an alteration in the dynamics of the ice sheets in the future because the panel members could not assess what this might be. They were left, therefore, in the awkward position of having to provide a prediction while omitting a large potential source of change. Their estimate is, as a consequence, controversial, and a number of scientists believe that it is too low, partly because of the caveat concerning the ice sheets. The problem is that, without significant further research in this area, any estimate of the contribution of the ice sheets to sea level rise remains highly speculative and so difficult to present in an unambiguous way.
So what is being, and needs to be, done to fill our gap in understanding, and boost our ability to model the processes the IPCC could not include? Some of the key processes concerned take place at the bed of the ice sheet, kilometres below the surface, and others occur beneath the floating ice shelves fringing the ice sheets. These are both areas where observations are difficult to collect.
Hope may be at hand: 2007-2008 is designated the International Polar Year and a number of international, co-ordinated efforts are underway to collect more - and better - observations of ice flow on the ground and with satellites from space. At the same time we need to improve numerical models of ice flow so that we can adequately reproduce the behaviour already observed.
Some of this work is underway but glaciology has always been the "poor relation" when compared to, say, oceanography or atmospheric sciences and if we are to make progress on this important, and challenging, problem then much more investment is needed. One thing is for sure, the ice sheets have a few more surprises waiting for us. As the saying goes: "watch this space".