This is a question that appears again and again for people who work with forecasting. These days forecast centres do not produce one forecast, they produce a set or ensemble of forecasts, which is supposed to reflect the uncertainty in the forecast.
This means that rather than giving a woolly statement such as “there will be scattered showers around tomorrow” or “it is likely there will be a frost tonight”, the Met Office could give an estimate of the probability of the event. In fact, in other European countries (for example Switzerland) a forecast of rain is accompanied by a probability of getting wet.
But here in Britain, somewhere in the upper echelons of the Met Office a committee has decided that the British public should not be given probabilities. The argument is that the public would not understand and not be able to use the information. It is not clear what basis exists for that statement. The counter-argument is that in a gambling-obsessed nation such as the British, where there is an abundance of betting shops, surely probabilities are well understood.
It is likely that if you are reading this blog, you are fairly numerate and would love to get some kind of probabilities with your forecast. Equally, I’m sure you can think of someone you know who would only be confused by this information. So clearly communicating probabilities needs to be done carefully and needs some thought to how the final implementation will look.
The obvious answer is of course that the forecast should be tailored to the user. In an exciting exclusive for this blog, I can reveal that this is exactly what is being planned at the Met Office! A senior forecaster has revealed to me that sometime in the near future the Met Office will launch a study on the public understanding of probabilities through an online game.
Using the information from this game, the Met Office will then develop ways to communicate the probabilities of the forecast to the public. The idea is that when you visit the Met Office site, you will get a forecast that suits your level of understanding. Of course, most people get their forecast from the TV and radio and these forecasts will not be affected, but if the website experiment is a success this could eventually change.
As soon as I know more about this experiment I will post another blog on this topic and then perhaps we will get the answer to the question at the top of this post!