Weather Forecast Accuracy Statistics
Using composite indices and statistical methods, data from sites like Accuweather and Weather.com is analyzed and rated.
This is a good resource for an interesting group or individual project in statistics: how accurate are the various services? What is a “good” prediction? How valuable is this information? How can we use statistics to evaluate these questions?
Several years ago I read this post from the Freakonomics blog: it details an informal study conducted by a man and his daughter who looked at seven months of TV weather forecasts in Kansas City and evaluated their accuracy. The entire article is interesting, but the bottom line is best summed up in a quote from someone from one of the TV stations: ““We have no idea what’s going to happen [in the weather] beyond three days out.”