It is one of the most alarming and ubiquitous images of global warming and general environmental degradation: the television pictures of chunks of melting ice cap crashing into the sea. The truth is, however, that the ice cap has always had bits falling off; it is part of a natural process of ice erosion compensating for the accumulation of ice from snowfall over the Antarctic.
Moreover, while it is true to say that the expert consensus of the last decade is that the Arctic has been melting and thinning-providing early warning of global warming-it now seems that the evidence is much more mixed. This is even more true of the Antarctic, where a comprehensive recent survey has concluded that on average the sea ice surrounding the continent is thickening.
Measurement of average ice thickness is not easy, and there is much disagreement over the best method. But in the case of Antarctic sea ice, several different studies have reached the same conclusion. One of the latest was conducted by Nasa's Goddard Space Flight Centre, using satellite observations over the 21-year period 1979 to 1999 inclusive. The conclusion was that although sea ice had thinned over one third of the area, it had thickened over the other two thirds. Furthermore, over the marginal regions where sea ice melts during the summer, the length of the period of ice cover had increased by a small but statistically significant amount of about a day a year.
Over the Arctic the picture is more confused. A study by the same Nasa team revealed a thinning of ice during the last 20 years, in contrast to its Antarctic survey. But a study by Gothenburg University of observations made by submarines found that mean Arctic sea ice thickness had remained almost constant between 1986 and 1997. This Swedish study agreed with other surveys that there had been an overall thinning of Arctic sea ice since 1976, but suggested that all of this occurred before 1986. Significantly this corresponds with some Antarctic temperature surveys, which found that while there was some evidence of warming between 1976 and 1986, this trend definitely ceased then.
While these findings provide some ammunition for global warming sceptics, the fact of ice thickening does not settle the matter. There are several ways in which temporary Arctic sea ice thickening would be compatible with a longer term warming trend. Global warming could, for example, divert or block warm ocean currents that previously flowed into Arctic regions and kept them warm; the best-known example being the Gulf stream. There is also the possibility of sea ice thickening without any cooling at all. This could happen in the northern hemisphere if global warming caused an increased flow of Russian and Canadian rivers into the Arctic ocean. Fresh water freezes at a higher temperature than sea water and floats on top of it, so if there is more of it, the sea surface freezes more readily.
I should emphasise here that we are talking about ice over the sea and not the Antarctic ice cap itself. No climate expert has ever suggested that there is any appreciable inland melting of ice over Antarctica or Greenland-it is simply too cold for that.
Frozen sea water only accounts for a small proportion of the world's total ice, most of which is locked up in the Greenland and Antarctic ice caps. But being a sensitive indicator of climate change, it is watched closely by climatologists. Indeed, the recent findings about Antarctic sea ice thickness do correlate strongly with changes in temperature over the Antarctic land mass. According to the Nasa satellite survey, the one Antarctic sea area where ice cover had diminished over the last 20 years was the Bellingshausen-Amundsen and western Weddell seas, immediately to the west and east of the Antarctic peninsula, which is indeed the one part of the continent that has shown significant warming over the past 20 years.
But fewer and fewer climatologists now dispute that most of the Antarctic has cooled over the last 15 years. The question then arises: why until just a year or two ago was there wide agreement among climatologists and glaciologists that the Antarctic was warming?
Part of the explanation is that there have been significant improvements in the measurements of both sea ice and surface temperature. Until recently, global temperature changes were measured by taking observations from weather stations, then extrapolating to derive data for the whole earth's surface. This approach relied disproportionately on urban areas. Satellite data cannot resolve local differences in micro-climate, but provides a much more accurate large-scale picture because it covers the entire world evenly. Similarly for sea ice, satellite data is more even and does not rely on specific routes taken by submarines.
It will take time to assimilate the new information and the impact on the global warming debate is unclear. But we should not become too complacent, for there is still the danger of large parts of the Antarctic ice cap becoming top heavy, and breaking off in large chunks. This would have nothing to do with global warming, but would have the same outcome of raising sea levels, only it would happen more quickly with less time to adjust. It is, in fact, the glaciological equivalent of an asteroid impact.
