When the developers of the popular OpenSignal Android app analyzed some of the data they’d gathered from their users, they were in for a surprise. They realized that the temperature sensors meant to keep the phones’ batteries from overheating could be used to predict the weather outside. The app doesn’t work on an individual level, instead it gathers data from a large number of users to estimate daily temperatures within 1.5 degrees Celsius.
Using data from phones in London, Los Angeles, Paris, Mexico City, Moscow, Rome, and Buenos Aires, they were able to estimate daily average temperatures in each area within 1.5 degrees Celsius. Their new app, WeatherSignal, takes greater advantage of this technology.
WeatherSignal is not accurate on an individual level. It takes a large number of users in a given area to account for the various temperatures and conditions different phones may be exposed to — air conditioned rooms or hot cars, for example. But when enough people send data through the app, clear weather trends start to emerge.
The American Geophysical Union believes this technology could revolutionize how weather predictions are made, allowing for much more accurate and timely information — tailored within a city block of the user. It could be especially helpful for people living farther away from their local weather station, who may not be getting the most accurate weather data available.