Forecasters predict hot August. Are they sure?

We all want to know what the weather's doing. But can we predict it months ahead? It's an issue clouded in controversy.

The TV weather forecaster may be chosen for their cheerful manner, but behind their words is the application of science and technology to an age-old question: what does the weather hold for me today?

Humans have attempted to predict the weather informally for millennia, and scientifically, since the 19th century.

In former times, for the farmer planting crops or the merchant setting sail, knowledge of tomorrow's conditions could be the difference between success and failure; even life and death.

And so old adages developed to guide, and many had good science behind them, like 'Red sky at night, shepherd's delight; red sky in the morning, shepherd's warning.'

But these days it's computers that guide. Put simply, weather is the condition of the atmosphere (or changes in the atmosphere) at a particular time and place. Computers are able to gather this data and then project forward, using forecast models, how atmospheric conditions will evolve in thousands of locations around the world.

But even with modern technology, it is not an exact science. In 1987, at 11.30pm, the BBC forecaster scotched rumours of a hurricane arriving.

Yet as dawn broke, 18 people had lost their lives and 15 million trees been uprooted by the worst storm to hit Britain since 1703.

And now the weather is back in the news with Positive Weather Solutions, claiming that June will be wet but August will be scorching.

They claim their weather predictions are accurate, based on long-term temperature cycles, weather patterns and a 30-year statistics database.

The Meteorological Office gave up long-range forecasts in 2009 after wrongly predicting a 'barbeque summer'. As Paul Knightley of MeteoGroup says, 'Sometimes these predictions are right,' he says, 'but it is more by luck than design.'

Why, with all our technology, is weather so hard to forecast? Various factors contribute to the problem.

There is the chaotic nature of the atmosphere; the huge computational power required to solve the equations that describe the atmosphere; error in measuring the initial conditions that go on to shape the weather, and our still incomplete understanding of atmospheric processes.

With such variables in the mix, weather forecasts inevitably become less accurate the further ahead they look.

Tomorrow's weather, we can know, and for next week's, we can make an intelligent guess. But three months ahead? The science suggests we can be no more sure than our ancestors.

You Decide

  1. 'I never trust the weather forecast.' Do you?
  2. The Met Office uses the NEC SX-6/8 super computer to provide data on over 4000 locations around the world. It's remarkable technology but long-term weather remains a mystery. Do you think this will always be so?


  1. Draw up a list of ten industries/businesses that might be interested in future weather. How would they be helped by knowing?
  2. Research the 'Kepler problem' in relation to the butterfly effect – and explain what it is.

Some People Say...

“Accurate weather forecasts would just spoil the surprise.”

What do you think?

Q & A

So there's good science behind some of the old adages?
Indeed. 'When windows won't open and salt clogs the shaker, the weather will favour the umbrella maker.' And it's true. Wood swells when moisture is in the air and salt absorbs it efficiently. And when moisture levels are high, rain is more likely.
I hadn't realised the atmosphere is so chaotic.
Yes, and at the heart of this Chaos theory is the fact that in a complex system such as ours, a small event in one place can have a big impact somewhere else. It's called the Butterfly Effect.
It's the idea that a butterfly's wings can create tiny changes in the atmosphere that ultimately become a tornado.
And businesses pay for forecasts?
Certainly. If you sell gas, electricity or water, you want to know what people will be using in the next few days.


PDF Download

Please click on "Print view" at the top of the page to see a print friendly version of the article.