Ice, Ice, Maybe

The message from Nasa in the past few days is that melting ice sheets in Antarctica and Greenland will make a greater contribution to sea level rise, beyond that estimated by the IPCC, because these ice sheets are melting at an accelerating rate. There have been over fifty news items written as a result of the Nasa press release, but hardly any have done anything other than uncritically reproduce it and the majority of them are very similar, almost interchangeable and unquestioning. But, as is the case with most research papers concerning climate change, when one looks in detail the results are not as unequivocal as journalists make out. Is this surprising? No. These are ice sheets, complicated, remote, widespread, difficult to measure, and poorly understood.The Nasa press release refers to a paper in Geophysics Research Letters of a 20-year study that shows that in 2006 the combined loss from the Greenland and Antarctic ice sheets were 475 GT per year which would contribute 1.3 mm to sea level rise. The two ice sheets are loosing 36.3 GT more than in the previous year.The researchers combine two sets of data, one obtained by radar from satellites and the other from the GRACE satellites that fly as a pair over the ice sheets measuring their relative movement to determine minute gravity changes caused by the moving ice sheet beneath them. There are 17 years of radar data and 8 years of GRACE data.Consider one of their graphs. Click on the image to enlarge. The GRACE data is in red and for Greenland (a) matches a decline and rise seen since2003. For Antarctica (b) it follows interannual variability over the same period quite well. This successful overlap, according to the authors validates the radar data that is important because in both ice sheets there are clearly multi-decadal changes apparent over the past 20 years. However, I am a little more cautious in believing that just because two data sets agree over an eight-year period this means that the longer data set (another ten years) which is unsupported by such collaboration has been shown to be ‘validated.’The graphs themselves are fascinating. Any conclusions about what is a trend and what could be cyclic must be guided by the fact that it is only 20 years of data in which there are obvious changes going on.Looking at the Antarctic data the mass loss line is heavily dependent on two features – the peak in 1993 and the dip in 2008, especially the peak that is at the start of the dataset. I’m always rather worried when if you started the data a year later, as is the case here, the evidence for a negative gradient would be far less robust. The line drawn through the GRACE data alone seems to me to be rather premature.Then there is the question of extending such lines forty years into the future – twice as long as the original data sets – and commenting on what this would mean for sea level rise in 2050. Personally I would like to see twenty more years of data (so as to see more clearly interannual effects and if it is really a trend and not cyclical) to have any confidence in extrapolating what might happen in 2050.The authors also state that they believe that the Greenland ice sheet is in transition to a regime of higher loss for which they cite as evidence that GRACE data shows decreased mass loss in south east Greenland and increased mass loss in north west Greenland!Overall it is a fascinating paper but its conclusions transcends the uncertainties in the data and its conclusions are extrapolated too far. It will however enter the canon of climate change ‘folk law’ and be quoted again and again as a result of the uncritical rehashing of a press release by so many journalists.That ice sheets are complex and poorly understood was also demonstrated in another paper, published in the journal Science, showing that what happens at the base of an ice sheet is more important, and more fascinating, that anyone thought.Radar measurements from aircraft flown over Dome A in Antarctica reveal more dynamic basal activity under the ice with refreezing of water and subsequent uplift than had been imagined. The conventional view was that ice sheets grow and disappear from the top. Now it seems they do the same from the bottom as well.What this means for climate change is uncertain. Some scientists have said the effect will increase the stability of the ice sheets but nobody knows. One thing is for certain however. The models of the ice sheets that are themselves fed into climate models are wrong.The details of the discovery are surprising. That we have learnt that we did not understand ice sheets as well as we though is not.Feedback: david.whitehouse@netzerowatch.com