Weather forecasters are curiously unloved. Maybe it is the smug certainty - too often unjustified - with which they deliver their predictions on radio and television; their irritating habit of telling you how warm it is in Barcelona before going on to say how wet it could be in Birmingham; the nannyish tone they adopt to deliver the meaningless advice that we should not venture out unless we have to. A plague of frogs on all of them. Maybe the old methods - hanging out seaweed or noting whether the cows are lying down - were best after all.
The move to cut the number of weather-watchers, which will be discussed with their union tomorrow, stems in part from a report last August by the Audit Commission that accused the Met Office of exaggerating its "success" rate. The report added that more than half its private clients - they include energy suppliers, aviation authorities and transport companies - were less than satisfied with the service they paid for. Ironically, it was issued towards the end of the long, hot summer, when for weeks all the forecasters had to do was recline in their deckchairs and predict more of the same.
The Ministry of Defence, which has overall authority for the Met Office, wants to reduce the pounds 91m that the Government contributed to it last year. Ministers are urging a two-pronged approach: cutting costs and increasing the pounds 57m earnings from commercial work - a sum which includes pounds 4m from broadcasters for regular forecasts plus the services of such storm-tossed stalwarts as Michael Fish and Suzanne Charlton.
The Institution of Professionals, Managers and Specialists, which represents the forecasters, knows which way the wind is blowing. It accepts that new technology - much of it invented by its members - means that fewer people are needed. Union leaders are anxious, though, that jobs should not be culled purely for ideological reasons.
After the Audit Commission report, Dr Julian Hunt, the Met Office chief executive, came in for rough handling from members of the Commons Public Accounts Committee. He admitted to the MPs that some of the Office's figures on the achievement of performance targets had been fudged, by rounding up decimals to turn near-misses into hits.
The committee was especially critical of the accuracy of the National Severe Weather Warning Service. This has the highest profile of any of the Met Office's activities and was responsible for its most famous debacle, in 1987, when Michael Fish laughed off rumours of the impending hurricane. (Not as bad as the South Wales tornado of 1913, where several cows were blown over a high hedge into the next field. It is not recorded whether they were lying down at the time.)
The figures show that last year the Met Office correctly forecast 83 per cent of freak gales and extreme frosts. This success rate increases year by year and is broadly in line with the accuracy of forecasts overall. A reasonable failure rate? What would we say about a hospital that lost 17 per cent of its patients through faulty diagnosis?
Worse, nearly a quarter of severe weather warnings last year turned out to be false alarms - down from the 37 per cent of needless panics in 1991- 92, when the hurricane incident was fresher in forecasters' minds and they were sending out red alerts by the score. As the Audit Commission pointed out: you can theoretically increase your hit rate by making more predictions of severe weather, but this also puts up your tally of false alarms. Weather forecasting is not an exact science, despite the satellites that track global cloud patterns and sophisticated computers that absorb thousands of readings, blink once or twice and then work out tomorrow's pressure in Rockall and Malin.
When it comes to selling its services in Britain, the Met Office does not enjoy a monopoly. In recent years several rival concerns have been established. Some of them buy forecasts from the Met Office, repackage them and sell them on to users such as ITV's Teletext service. Others take basic information from the US weather service, which charges only a nominal rate because of the requirements of the US Freedom of Information Act. The Met Office is competitive, however, and has recently won back the custom of GMTV, the ITV breakfast station, and BBC's Ceefax.
Because of its role in national defence, it is unlikely that the Met Office would be privatised fully. But it will soon move to "trading fund" status, where the money it receives from government will be related to the service it provides.
The question nobody has yet dared to address publicly is whether we need the Met Office at all. Although it is respected internationally as a leader in the field, its functions are largely duplicated by the American service in Washington and to some extent by the European Centre for Medium Range Weather Forecasting, an international body based at Reading, to which the Met Office contributes.
Every night around midnight, BBC2 broadcasts a Met Office forecast for the whole world, filling us in on the chances of a heatwave in Alaska and the prospects for play in the South African cricket. At about the same time CNN, the American international station, is forecasting Europe's weather, based on the Washington service. If cost-effectiveness is the goal, why not concentrate resources on one global forecasting centre, to which all nations could contribute?
There are probably too many vested interests at stake to allow that. If it did happen it would be a long way into the future. In the medium term, Met Office staff might reflect that, under the plan to be discussed tomorrow, they would have an 82 per cent chance of keeping their job - not far from the success rate of their forecasts. If that level of uncertainty makes them quake in their snow boots, they know just how the rest of us feel when we wonder whether to take our umbrella with us when they promise a cloudless day.Reuse content