Byline: MICHAEL HANLON
POOR old Michael Fish. Remembered for one thing: not predicting the Great Storm of October 1987. Back then, the weathermen (there were few women in those days) came in for a lot of stick.
Gales would come and go with seemingly no warning whatsoever.
We would be told 'sunshine and showers' and be drenched by a day's downpour.
Bank holiday wash- outs would be predicted, and the sun would come out at the last minute.
The lot of the weather forecaster was not a happy one.
But things have been getting better. Maybe you haven't noticed, but the forecasts are a good deal more accurate now. If weathermen say it's going to rain, it usually does. If frost is predicted, it's a safe bet that there will be a coating of white on the lawn in the morning.
As well as getting younger, and prettier, the Met Office forecasters who present the weather on the BBC have had a good deal more to smile about in the past ten years, even if the rest of us, in washed-out Britain, have not.
But wouldn't it be nice if they could be better still? Farmers would love to know whether to get their tractors out or leave their harvesters in the shed.
Daytrippers would like to know whether to bother getting in the car for the long trip to the coast, or stay at home to do some DIY.
REALLY long-range forecasts could help people plan weddings and holidays.
And, of course, there is the matter of saving lives and money if you know exactly when and where the storms/blizzards/thunder is going to strike: 'Rain tomorrow before lunch, lasting 17 minutes.
Then sunshine till 4 when it will cloud over.' Just imagine the heartache, not to mention carpets, that could have been saved if last autumn's floods could have been predicted accurately a few days or weeks in advance.
The Met Office thinks it would be nice, too, and it claims that from next year a revolutionary new way of forecasting should make this possible.
At the end of 2001, the forecasting rulebook is to be thrown away and a replacement, called the New Dynamics, put in its place. The rules consist of millions of lines of computer code, operating on trillions of bytes of weather data. Rewriting the rules has been a big headache but the rewards should be more than worth the effort.
'At present even the best computers blur things a bit. Forecasts are fuzzy.
The New Dynamics is like giving the forecasting machine a pair of glasses,' says Professor Julian Hunt, former chief executive of the Met Office, Labour peer and researcher at University College London.
With the new system, the computerised forecast, which began with a strange link to the 1940s Lyons Tea Houses (as I'll explain later), will be brought into the 21st century.
The seeds of proper, scientific weather forecasting were sown in the 1920s by eccentric British genius Lewis Fry Richardson.
He drew up the first mathematical equations to predict the movement of air masses and fronts that determine what the weather is going to do now, tomorrow and the day after that.
Richardson - who was Lord Hunt's great uncle - thought that if you had a powerful enough machine, a 'computer', you could predict the weather to almost any degree of accuracy, and to any point in the future.
In other words, Richardson treated the weather as a huge piece of clockwork: if you could only know the exact position and speed of every pocket of air in the atmosphere, you could tell what it was going to do …