麻豆传媒

A SNOW-CAPPED mountain on the equator? 鈥淚mpossible,鈥� declared the learned gentlemen of the Royal Geographical Society in London. In 1848, the Swiss missionary and explorer Johannes Rebmann had become the first European to gaze upon the snows of Mount Kilimanjaro. But when his account reached London, some of the society鈥檚 members refused to believe him. One even suggested that Rebmann鈥檚 eyesight must be deficient.

Today the mountain is once again the centre of controversy. Its ice cap and the glaciers that flow from it are disappearing fast. Global warming can be intangible, but melting ice: that we can see. Many climate change activists have seized upon Kilimanjaro as a striking symbol of global warming, a poster child for the shrinking ice caps and glaciers around the world.

The evidence for a worldwide meltdown is overwhelming. According to the World Glacier Monitoring Service in Zurich, Switzerland, of the 88 glaciers surveyed in 2002 and 2003, only four were growing and at least 79 were receding. The crucial question, however, is why.

Global warming just doesn鈥檛 explain it all. Many mountain glaciers around the world started shrinking in the 19th century, long before human-induced climate change could have had any impact. And then there is Kilimanjaro. Some climate change sceptics go so far as to claim its vanishing ice is nothing at all to do with global warming.

Cores from Kilimanjaro鈥檚 summit show the ice has been there at least 11,000 years. The ice cap is thought to have grown until around 1880, long after Rebmann saw it. Since the first survey in 1912, 80 per cent of the ice has disappeared, and the rest could be gone in under 20 years.

It seems like a clear-cut case of global warming. But in a paper published last year, Georg Kaser of the University of Innsbruck in Austria pointed out that the ice began to recede before there was any known warming on the mountain. Instead, the retreat coincided with a dramatic drying. This part of Africa is noted for sudden changes in precipitation, reflected in lower water levels in Lake Victoria. And while the 1880s were unusually wet, later decades were much drier.

Kaser鈥檚 arguments have been seized upon by climate-change sceptics such as Patrick Michaels. 鈥淜ilimanjaro turns out to be just another snow job,鈥� he declared last year.

The case that the ice began to retreat because of less snow seems sound. But why is it still retreating? The climate-change sceptics argue that it must be lower snowfall, because the air at this altitude in the tropics has not warmed. As Kaser pointed out in his paper, weather balloon readings suggest a slight cooling between 1979 and 1997.

But recent studies show that this apparent cooling is just an artefact of changes to weather balloon instruments (麻豆传媒, 20 August, p 10). And satellite studies suggest continual slow warming of the tropical troposphere, according to climate dynamist Raymond Pierrehumbert at the University of Chicago.

If the ice loss was entirely due to less snowfall, it should have slowed or reversed after 1960, when the weather became wetter again. But Kilimanjaro鈥檚 ice kept right on retreating. Lonnie Thompson of the Byrd Polar Research Center at Ohio State University, points out that the glacier survived a 300-year drought around 4000 years ago. So something must be different now.

One possibility is that changes in cloud cover mean more sunlight is reaching the ice. Changes in humidity and the timing of snowfalls might also play a role. But even if these factors turn out to be more important than air temperature, that does not mean that the sceptics are right. 鈥淭here is a strong link between conditions on Kilimanjaro and global warming,鈥� says Kaser. 鈥淛ust it is not in a simple way, such as rising air temperatures melt glaciers. Any interpretation of our findings toward 鈥榥o climate change鈥� are entirely wrong.鈥�

But until the case is ice solid, perhaps campaigners should choose a different poster child. There appear to be plenty to choose from. 鈥淭hroughout the tropics, glaciers are in retreat,鈥� says Pierrehumbert. Elsewhere in central Africa, Mount Kenya has lost seven of its 18 glaciers since 1900, and most of the ice on the Rwenzori mountains between Uganda and Congo has disappeared. Across the Indian Ocean on New Guinea, the West Meren glacier, which vanished altogether in the late 1990s, is just the latest of several glacial disappearances.

In many cases these tropical retreats seem to be recent and unprecedented. Take the ice cap round Quelccaya mountain, the largest in the Peruvian Andes. Back in 1976, a core of the ice taken by Thompson showed layers of ice laid down annually for 1500 years. In 1991, when he returned to update the record, he found that the annual accumulation had stopped and the top 20 metres of the ice had melted away.

In the valley below, Quelccaya鈥檚 largest glacier, the Qori Kalis, has lost a fifth of its area since 1963. Last year, Thompson found the remains of plants that had been frozen under the ice for at least 50,000 years.

Across the Peruvian Andes, a quarter of the ice cover has disappeared in the past 30 years. Dozens of other glaciers up and down the Andes show similar dramatic changes, from Bolivia鈥檚 Chacaltaya to Ecuador鈥檚 Antisana. Venezuela has lost four of its six glaciers since 1975. Thompson can find no sign of reduced snowfall or increased sunlight to explain the dramatic meltdown. But air temperatures have risen by around half a degree.

Yet again, though, we cannot leap from the local to the global. The temperatures in the tropical Andes are rising in part due to the increased frequency and intensity of El Ni帽o, the Pacific climate anomaly. How much the changes to El Ni帽o are tied to global warming and how much to natural climate oscillations is an open question. So the link between the dramatic loss of ice in the tropical Andes and climate change is not as clear as it appears.

Too much, too soon

In other regions, as on Kilimanjaro, glacial retreat began long before human activity altered the climate. In both the Himalayas and the Alps, glaciers have been in retreat ever since the Victorians explored them. From the mid-19th to mid-20th centuries, Alpine glaciers lost half their mass, according to the European Environment Agency. Europe鈥檚 biggest glacier, the Breidamerkurjokull in Iceland, has been shrinking for most of the past century. The ice fields of Patagonia started melting around 1880. In the US, ice in the Glacier National Park in Montana was already melting when the park was set up in 1910.

Hans Oerlesmans of Utrecht University in the Netherlands has studied records of 169 glaciers round the world, some going back to 1600. He reported in Science this year that most glaciers reached a peak sometime in the early 19th century and then began a steady retreat (see Graph). That is far too early to be a signature of man-made climate change.

Instead, the most compelling culprit is the world鈥檚 rebound from the little ice age. This was an era of mild cooling between the 14th and 19th centuries. Climatologists now agree that most of the global warming in the 19th century was a consequence of this recovery and there seems little doubt that it also explains the 19th-century glacial retreat in the Alps, Himalayas and elsewhere.

But does it explain the continued meltdown? After all, glaciologists agree that glaciers take time, often several decades, to respond to changes in climate.

It seems unlikely, says Pierrehumbert. 鈥淚t is a fact of life that climate changes associated with the end of the little ice age overlap with the beginning of the era of industrial warming.鈥� But the bounce-back should have tailed off during the 20th century. There is no way it should be increasing 鈥� yet it is.

Oerlemans鈥檚 study of 169 glaciers reveals glacial retreat has accelerated in recent years. In Patagonia, the pace of retreat has doubled in the past three decades. In the Alps, melting all but ceased in mid-century, but since 1980 it has resumed with a vengeance.

Elsewhere, the records do not go back so far, but the pace of change is almost certainly new. In China, an inventory of 46,000 glaciers this year reported a 7 per cent loss of ice since 1980. In the neighbouring Tien Shan mountains of central Asia, the surface area of glaciers shrank by a quarter between 1955 and 2000. The only known mechanism that could account for such near-universal changes is global warming.

A handful of glaciers do buck the trend. In the US鈥檚 Glacier National Park, there was less melting in the second half of the 20th century than in the first half, largely because of local cooling in mid-century and increased snowfall. And while most glaciers in the Sierra Nevada are retreating, seven around Mount Shasta in California are growing apace, due to a local increase in snowfall. Three have doubled in size since 1950.

Two of Patagonia鈥檚 70 glaciers, Pio XI and the massive Perito Moreno, have been reported to be growing. But according to Helmet Rott of the University of Innsbruck, Pio XI is losing mass in its upper reaches as it grows longer 鈥� so it may just be sliding downhill faster. And he says Perito Moreno is stable rather than growing. The glacier is exceptional: its snout dams a lake and is periodically broken off, so it is not as large as it might otherwise be. Its accumulation zone is also very high and steep, insulating it from warming.

The only glaciers in Europe that have advanced recently are on the west coast of Norway, where increased snowfall seems to have overwhelmed the effects of warmer air. Even here, the Nigardsbreen and Aalfotbreen glaciers, which advanced strongly from the mid-1980s, have been losing ice since 2001.

The story is similar in New Zealand鈥檚 Southern Alps. The Franz Josef glacier, a popular tourist destination, has periodic surges but has been in general retreat since it was first photographed in 1867.

Does it matter?

If, as virtually all climatologists now believe, the fingerprints of human activity can be seen all over the rise in global temperatures over the past half-century, then they are all over the retreat of glaciers, too. Glaciologists may still be arguing about the comings and goings of specific glaciers such as those on Kilimanjaro, but they are in absolutely no doubt about the big picture: the glaciers are shrinking ever faster because of global warming.

So does it matter if mountain glaciers disappear? In the short term, there is a risk of catastrophic floods. Rocky debris piles up at the foot of glaciers. When melting glaciers retreat, vast lakes can form behind this unstable moraine. If the moraine gives way, the water sweeps through the narrow valleys below. Such glacial outburst floods are likely to become far more common.

In the longer term, the problem is likely to be too little water. As glaciers vanish, summer river flows could fall in many parts of the world, from China to California.

All this molten ice has to go somewhere. The shrinking ice fields in Patagonia and Alaska are already contributing to a rise in sea levels. The 2001 report of the Intergovernmental Panel on Climate Change (IPCC) predicted that the melting of mountain glaciers and ice caps will raise sea levels by up to 0.23 metres by 2100. Total meltdown would raise sea levels by about half a metre. That is significant but far less than the sea level rise predicted as a result of the thermal expansion of the oceans.

The real action is on the great ice sheets of Greenland and Antarctica, which are more than two kilometres deep in places. Total meltdown would raise sea levels by as much as 80 metres, drowning most populated areas.

Reassuringly, satellite radar measurements published earlier this year show that the East Antarctic ice sheet, the world鈥檚 largest, has grown thicker by 2 centimetres a year during the past decade. The most likely reason is more snowfall. Some sceptics have heralded the finding as proof that concern about global warming is misguided, even though the IPCC forecast that Antarctic would likely gain a little ice over the next century.

Now that forecast is starting to look optimistic. What matters is the balance between the gain of ice in the high hearts of the sheets and its loss through melting and calving of icebergs into the sea on the edges. Across the continent on the West Antarctic ice sheet, ice loss appears to be gaining the upper hand. The glaciers draining the heart of the ice sheet are accelerating, it was reported earlier this year.

Recent estimates are that peripheral melting round Antarctica is raising sea levels by 0.16 millimetres a year, while the thickening at the ice sheets鈥� heart is reducing that by 0.12 millimetres a year. Thus, overall, Antarctica is probably already shedding ice. The same holds for Greenland. One recent estimate puts the melting there at enough to raise sea levels by 0.2 millimetres a year.

And there are worrying signs that the loss of ice could accelerate dramatically. The break-up of the floating ice shelves around the continent would make no direct difference to sea level but would allow the ice on land behind to slide into the sea unhindered. Since the collapse of the Larsen B ice shelf off the Antarctic peninsula three years ago, some glaciers have started flowing up to eight times as fast.

Glaciers also speed up when the surface ice melts, as happens in the Alps in summer. Some of the water pours down vertical shafts known as moulins to the base of the glacier, where it lubricates the glacier鈥檚 base, speeding its slide. Recent studies in Greenland show this can happen to ice sheets as well as glaciers, speeding up flow towards the margins. The phenomenon may explain the thinning of the Patagonian ice fields, which is more rapid than seems likely from melting alone.

Until recently, climatologists thought that it would take thousands of years of warming to melt the vast ice sheets of Greenland and Antarctica. Now many think they could collapse within centuries. When that happens, Kilimanjaro will be the least of our worries.