By Tom Geoghegan
BBC News Magazine
We're all set for a summer heatwave, say the newspapers. But if the weather forecast for tomorrow can't be guaranteed, why should we believe predictions of a warm summer?
The government's heatwave advice issued this week may have us contemplating a long hot summer, but don't get your hopes too high.
The health warning was based on the latest seasonal forecast of the Met Office, which suggests that temperatures are likely to be above normal.
But while the newspapers were quick off the mark with predictions of heat waves and pictures of women in bikinis, forecasters at the Met are more cautious. So how are seasonal forecasts made?
Richard Graham, manager of long-range forecasting at the Met Office, says long-term predictions depend on ocean temperatures. Short-term ones are based on changes in the atmosphere.
The transfer of heat and water vapours between the two means the ocean has a regulating effect on the weather.
"The sea makes up a large expanse of the earth's surface and its temperatures don't change quickly," he says. "They persist for several months at a time and they have a modulating effect on weather regimes."
It's these temperatures which inform the Met's three-month average prediction. Thousands of readings are taken in the air and sea. The Met then draws 40 long-range forecasts, each with a slightly different but plausible starting point.
This tries to eliminate the "butterfly effect", that a small variation in the atmosphere can cause huge weather fluctuations elsewhere. The 40 are then boiled down to one overall reading, which is either average, above average or below average temperature.
The forecast for this summer is "above average", which means above 22C, and the higher average will be in August and September. But it can't be any more specific than that and the summer is harder to predict than winter, says Mr Graham.
There's no such hesitancy at the spread-betting company Sporting Index, which predicts record-breaking temperatures around 12 August, based on previous summers and seasonal forecasts.
But any talk of records is just marketing hype, says Jim Dale, senior meteorologist at British Weather Services.
His company uses historical data to make its predictions, which he says have a seven to eight-out-of-10 success rate. It uses 30 years of statistics and the balance of probability to make its forecasts.
"To talk about a specific week or day is absolute rubbish and to then moot that the extreme temperature might be beaten. Well it might be. There's a lot of supposition here and a little bit of hope maybe.
"Warm summer stories are always good for stores and people's buying attitudes. I rather feel it's glorified marketing and sensationalist reporting."
Mr Dale says their forecast is for mixed weather for the next 20 days and a worse start to the summer than 2004.
June will be equally mixed, perhaps getting better towards the end, but late July and August will see a two-week period of hot and sunny weather, although not necessarily record-breaking.
A range of opinions on what 2005 holds, then, but there is one thing most meteorologists are agreed on. Global warming means the UK summers are very slowly getting hotter.