Science: Deep-frozen clues to a warmer world: John Wright explains how Greenland's glaciers could help to reveal secrets of climatic change

Greenland's icy mountains may hold important clues to the way in which our climate is changing. Earlier this year, the United Nations Environment Programme (Unep) published figures showing that mountain glaciers around the world were in retreat. But there was a problem with this apparent evidence of global warming: the data covered only the period 1965 to 1989.

However, surveys of some high polar glaciers go back a good deal further, and some were remarkably accurate. As long ago as the Thirties I surveyed two glaciers in Iceland and north-west Greenland, and several in Arctic Canada. Over the past few years, I have resurveyed the first two, thus extending the period available to Unep data.

In north-west Greenland there is a long and accurate record of the Moltke glacier, 20 miles long and five miles wide, which comes down from the main Greenland ice-cap, only a few miles away from the huge Thule Air Base set up by the United States in the Fifties. It had been mapped as early as 1915 by the Danish explorer Lauge Koch. So its changes have been recorded for nearly 80 years.

I spent the winter of 1937-38 in Thule on a three-man expedition organised by the explorer David Haig Thomas. Our main objective was Ellesmere Island, north Canada. In the autumn it was inaccessible by boat, so for our winter expedition we concentrated on the area round Thule and mapped the valley now occupied by the air base, including the Moltke glacier to the east.

Last year, as the only surviving member of that expedition, and aged 79, I revisited the area and by the kindness of the American Base commander Colonel Robert Cox, I was able to camp for two days at a point east of the base and survey the glacier front. It had retreated four miles from its 1937 position, but otherwise it had hardly changed since the last air photograph, taken by the Danes in 1985.

In 1938 our main objective had been southern Ellesmere Island; and in April and May the third member of the Haig Thomas expedition, Richard Hamilton, and I surveyed some 3,000 square miles of this island, including parts of the interior. Thanks to good weather and the photographic survey technique that we used, I have yet to hear of a high polar survey of comparable detail, accuracy and extent achieved by two men in so short a time.

What made the field survey so quick, precise and detailed? The answer is the special photographs. Before the days of air photographs, others had relied on sketches made in the field in temperatures of down to -20C. Detail was then sketched in back home from hand-held photographs. In contrast we relied on precise photographs taken with an ordinary but good camera set up over a theodolite. With the camera axis horizontal and parallel to that of the theodolite, angles could then be measured from the photographs.

The two keys to this technique, invented by the British surveyor and photogrammetrist Michael Spender for the 1935 Everest reconnaissance, were a simple bridge to fit the camera over the theodolite; and the realisation that a 35mm or similar small camera was capable of much higher consistency and therefore accuracy than had been thought.

The advantage of these photographs was that one could identify far more points of detail, including crevasses and dirt cones on glaciers and snow or rock patches on mountain sides, by comparing photographs taken from two survey stations in comfort at home, than one could ever do from sketches.

But with the analytical techniques of the Nineties, we can now do even better. Using microscopes and computers to measure and analyse more accurate distances off the negatives, Peter Dahl at City University has shown that our 1992 survey photographs in Iceland can achieve much higher accuracy - up to a minute of arc in direction. This gives heights accurate to 2m on a glacier surface up to 5km distant.

Similarly accurate maps could be made of the glaciers in Ellesmere Island from the photographs taken in 1938. I cannot, given my age, repeat our observations there. Nor can I undertake modern analysis of our photographs. But the records, including negatives, and my original map and angle books, are safely stored at the Scott Polar Research Institute in Cambridge.

The institute was set up by survivors of Scott's second and fatal south pole expedition using money left over after the dependants of those who died had been helped. Its primary purpose was to save early records of expeditions. Previously, these had often been left with explorers' families and subsequently lost. I left all my records with them in 1939; they are still in perfect condition. With the aid of computers, these could be used to produce large- scale, closely contoured maps of the glaciers and icecaps of south-east Ellesmere Island as they were 56 years ago, less than 1,000 miles from the north pole.

Several series of air photographs of the area were taken in the Fortiesand afterwards. By locating them, using unchanging rock features on these maps, it might be possible to show what changes in the glaciers have occurred. Records similar to those for the Iceland and Greenland glaciers could be produced, showing how these have behaved in this 56-year period of global warming.

(Photograph omitted)