Jump to content

Archived

This topic is now archived and is closed to further replies.

cookie

The Weather Project

Recommended Posts

http://news.bbc.co.uk/today/hi/today/newsid_9029000/9029232.stm

Who can we trust to forecast the weather?

People moan about the Met Office forecasts, but can other forecasters beat them?

You might think there would be a clear answer to this question - but there isn't. There is nowhere you can see how weather providers compare.

Knowing the weather is not just a matter of keeping your new hairdo dry. If local authorities run out of grit, people die on icy roads. If a hurricane is forecast you stay at home.

Shops' profits depend on having the right goods on the shelves at the right time. Major events can be wrecked by weather. UK tourism depends on a sunny forecast.

So BBC News is trying to bring you the answer to the question "who can we trust?".

We are launching a Weather Test where we'll study the predictions of a number of people using different forecasting methods and we'll attempt to conclude which is most accurate.

I say attempt, because there is a chance that the project will not succeed for a number of technical reasons, especially because it may be hard to compare forecasters using different ways of expressing results

To help us we've formed a steering group including Paul Hardaker, who runs the Royal Meteorological Society; Martin Dougherty, director of the Royal Statistical Society; and a member of the Royal Astronomical Society.

Other partners on the Weather Test steering group are the Willis Research Network of insurers, who take an obvious interest in weather matters; and Philip Eden, an independent weather historian. The group is completed by myself and my colleague Dominic Groves, Assistant Editor of the Today programme, who will manage the project within the BBC.

When methodology has been agreed, the comparison of forecasts will be done by the University of Leeds.

We want this to be an open project, so we're inviting members of the public to a public meeting in London on Tuesday 12 October, where we'll discuss our plans so far.

If you're interested in taking part, please get in touch using the form at the bottom of the page.

The Royal Society is not involved in the project, but one of their fellows Professor Tim Palmer, incoming president of the Royal Meteorological Society, has agreed to scrutinise the methodology.

Professor David Spiegelhalter from the Cambridge Statistical Laboratory will also offer comments.

So far the project looks like this - we'll aim to choose a number of weather providers and ask them to supply information in advance of their monthly and seasonal forecasts. Then we aim to publish the results of the first six months' test next summer when we have the results.

The weather data against which the forecasts will be compared will be provided by Philip Eden. (His own forecasts will not be judged in the Weather Test).

The number of forecasters has to be limited because of the cost of processing the comparison data. Cost also rules out a comparison of daily and weekly forecasts.

These plans are provisional and subject to change. The project is completely independent of the Met Office and of BBC Weather - it is being run through BBC News.

We hope it works. If not, I'll be back to tell you why it failed and show how difficult it is to do something as apparently simple as this.

Share this post


Link to post
Share on other sites

A very interesting project indeed...

Share this post


Link to post
Share on other sites

×
×
  • Create New...