When it comes to snowfall figures, ski resorts can be a bit slippery.

Here is the weather forecast for the 2001/2 ski season. Christmas week will see the most plentiful snow on beginners' slopes at Avoriaz in France.

Here is the weather forecast for the 2001/2 ski season. Christmas week will see the most plentiful snow on beginners' slopes at Avoriaz in France. Falls on intermediate and advanced areas will be heaviest at Park City and Deer Valley in Utah, and on the steeper terrain of Mt Batchelor in Oregon. For each of the subsequent peak periods, at the New Year, the February half-term holiday and Easter, La Plagne in France will have better beginners'-slope snow than any other major ski area in Europe or North America. At New Year, the best snow cover on intermediate pistes will be found in Big Sky, Montana; for half-term it will be at the Espace Killy in France, for Easter at Whistler/Blackcomb near Canada's Pacific coast. At the same periods, skiers in search of good snow on advanced terrain have the best chance of finding it at, respectively, Jackson Hole in Wyoming, Aspen in Colorado, and again at Aspen.

Where do these predictions come from? From a house called Tigh na' Tilleadh at Polbain near Achiltibuie, 85 miles north-west of Inverness. There, a 50-year-old former tour operator, Ian Campbell Whittle, collects and processes information for his WorldSki Guide. In what sounds like perfect isolation – there is nothing but sea between his house and Canada – he surveys the world's ski resorts, from the famous to the obscure. If they so wish, visitors to his website (which is essentially commercial) can pay $2.95 and download an analysis of the Yazawa ski area in Japan – although he does not believe anyone has ever done so. "But if you set out to do a world guide," he says, "you have to cover places like that." He is sure that someone once bought WorldSki information on another Japanese resort, and equally sure that "nobody has bought the Perisher Blue area in Australia".

The dense WorldSki site – it is not easy to navigate – sells guides to what it considers to be the top 120 resorts; they contain the usual facts and figures, plus an accompanying text more opinionated than one would find in the ski guidebooks. (Elsewhere, the site also offers opinions on those books to which their editors would probably take exception.) The ski-area information, however, is unusual in the stress it places on the suitability of the terrain for different skier skill-levels. But the snow reports are the most interesting area, thanks to Campbell Whittle's outspoken views on the accuracy of the statistics supplied by some resorts.

Everyone accepts that ski resorts have no incentive to underestimate the amount of snow on their slopes. Unfortunately – since, as Campbell Whittle says, "nobody else is going to go up the mountains every day at 6am to check snow-depths" – the resorts' own figures are commonly the only ones available. Even information supplied by "independent" reporters tends to bear an uncanny resemblance to the official figures, although judgements on the snow's skiability may differ. The data that WorldSki uses is, says Campbell Whittle, mainly sourced from national tourist boards and resorts; "and we take the view that the upper- and lower-slope snow-depths they publish are reasonably accurate. But there are several caveats, so we go through the figures with a fine-tooth comb to see if there's any cheating going on."

Many resorts also indicate whether it is possible to ski right down to the valley. Since this is an easily observable fact, reports are very likely to be accurate – and can therefore serve to confirm or deny doubts about the amount of snow on the lower slopes. Another control is the number of lifts and pistes that are open. "If a resort reports that it has fantastic snow, but is only operating, say 10 out of 50 lifts, that makes one suspicious," says Campbell Whittle. The local snowfall figures also provide a useful reference: although some resorts without snow-making equipment apparently provide evidence to the contrary, "You can't have more snow on the slopes than has fallen from the sky," as Campbell Whittle points out.

"Probably the best reporting system is the one in Spain: it records the percentage of lifts and trails that are open, snow-depths, and whether it is possible to ski to the bottom. Norway's is probably the worst. Its system was changed a couple of years ago, and now only a single figure for average snow-depth is published. As soon as people start quoting averages, you suspect they are trying to hide something."

But the biggest problems, he says, are posed by North America. The majority of the resorts are privately owned, so they have to make money – which increases the temptation to hide bare patches on the slopes. And different reporting methods are used in different areas. New England's reports are, he says, "pretty accurate" – partly because most resorts there have such efficient snow-making equipment. Nevertheless, Campbell Whittle has had disagreements with the area's biggest resort, Killington, over the location at which it measures its lower-slope depths.

He has also had a long-running dispute – of the same nature – with a Lake Tahoe resort. "It was quoting a lower-slope depth of 24 inches; but only seven out of the 27 lifts were operating. Looking at the other Tahoe resorts made it even more obvious that something was wrong, because they were reporting only 12 inches of snow at the bottom of their ski areas."

Using such detective work, Campbell Whittle produces "corrected" weekly figures sufficiently authoritative to be bought by the main Swedish snow-reporting organisation, and to which both the Austrian National Tourist Board in London and Virgin Ski refer their customers. The database is also used to generate resort snow-records over an annually updated, five-year period. And it is those statistics which WorldSki uses to select the resorts (listed above) as historically the most snow-sure for peak-period holidays. So one other caveat, before you book a holiday in one of those resorts: snowfall figures may go down as well as up.

The new, 2001/2 data is due to be installed on Monday on the WorldSki site, www.worldski.demon.co.uk