It’s actually very easy to predict the weather. You can look out of the window, see what the weather is doing, and say “tomorrow will be like today”. In much of the world – southern California, the Mediterranean – that will be pretty reliable. It’s called the “persistence method” and works fairly well in places where the weather only changes slowly throughout the year.
The persistence method is significantly less useful in places like Britain, though, where we have a great landmass on one side and a huge ocean on the other, and two major convection systems – the Gulf Stream in the sea and the Jet Stream in the air – pushing vast amounts of energy around us. Here it changes regularly – within the limits of “pretty boring, probably not going to kill you”, of course, we don’t get many hurricanes or famine-causing droughts. But still, farmers and sailors and people haven’t been able to rely on a simple steady weather system. That’s probably why Britain has been at the forefront of efforts to find new and better methods of predicting the weather.
It’s not been that long since we had no useful methods whatsoever, really, beyond the persistence method and various forms of folk science of differing value (“red sky in the morning” actually does have some predictive value, as might the suggestion that cows lie down before rain and swallows fly low. Pains in your bones, not so much). The spread of the telegraph in the early 19th century led to the ability to send gale warnings over vast distances; later the barometer, and the realisation that low pressures often preceded bad weather, improved local predictions.
Less than 200 years later we’re complaining if the Met Office – which consistently manages to predict not just gross measures like “will it be clear tomorrow” but fine details, such as the temperature to within 2C, a day in advance, with 85 per cent success – can’t tell us whether an entire summer will be barbecue-appropriate. We really are a bunch of entitled sods. (The Met Office, incidentally, said they were “odds on” for a barbecue summer. As in, it’s likely but not certain.)
Anyway. Given the chaotic nature of the weather system, it is impossible to make long-term predictions with all that much accuracy – tiny changes in starting conditions, far below the ability of any satellite to measure, can have huge effects on outcomes, in the tradition of the proverbial butterfly flapping its wings in Brazil and causing a thunderstorm in China. But the Met Office has started using a new computer system, lumbered with the name ENDgame, which it claims will improve accuracy, through more detailed mapping of the atmosphere and more sophisticated models.
No doubt it will work, and weather forecasts will improve somewhat, from “really pretty good” to “even better” (it’s been getting better for a while: their four-day prediction now is as reliable as their one-day prediction in 1980). But still there will be times when they predict sunshine and we get rain, or vice versa, and we will complain, because we don’t remember the many other times when they said “rain” and it rained, and we don’t understand that if they predict a 75 per cent chance of rain, there’s still a 25 per cent chance it won’t. The poor old Met Office, it just can’t win.