Weather reports are perhaps the most commonly encountered examples of probability in our daily life. For instance, the chance of precipitation for tomorrow is 40%. We know that there is only one chance of tomorrow happening, and only two possibilities – it rains or doesn’t. Then what does this 40% mean?
Let us start with what it is not. 40% rain does not mean it will rain 40% of the time or on 40% of the area!
One interpretation is that it rained 40 out of 100 days of similar weather patterns like tomorrow in the past. This interpretation relates closely to the climatology method of weather prediction, where past weather statistics guide the future. But whether predictions of today are far more advanced.
These days, weather forecasters run advanced mathematical models that take into account wind velocity, humidity, temperature, pressure, density etc. Even tiny errors in some of these variables can make the prediction off by a mile. Therefore, different models with several modes of sensitivities are solved to get an ensemble of outcomes. In the end, the Meteorologist looks at how many of them predicted rain. Suppose 20 out of a total of 50 realisations (model outcomes) predicted rain; the forecast becomes 40%.
