Aurora, the foundational AI model developed by Microsoft researchers for weather forecasting, is now able to predict air quality with high speed and accuracy. This information has been given by the company in a recent blog post.

Going beyond the weather, predicting air pollution

Aurora has already been known to predict weather events such as hurricanes, and cyclones, but now it has been fine-tuned to predict beyond traditional weather boundaries like air pollution. Microsoft said that this model first acquires general knowledge from large and diverse data sets, and then it is refined with small and specific data.

Open-source and integration into the MSN Weather app

Aurora's source code and model weights have now been made publicly available. A special version of it, which provides hourly forecasts (such as cloud cover), is now also integrated into the MSN Weather app.

How does Aurora work?

Aurora is trained on over 1 million hours of satellite, radar, and weather station data, as well as historical weather simulations and forecasts. Its encoder architecture transforms huge amounts of data from different sources into a standard format, allowing the AI ​​to make accurate predictions.

PC Social media