Google's AI-Powered Weather Model Aims for Earlier Hurricane Warnings

Google has made significant strides in developing an AI-powered weather prediction model called WeatherNext. This model can provide a 15-day forecast, predicting both the track and intensity of storms, including hurricanes and typhoons. WeatherNext was trained on nearly 20 terabytes of global atmospheric data and historic information collected by the International Best Track Archive for Climate Stewardship.

The National Hurricane Center, the Cooperative Institute for Research in the Atmosphere, the UK Met Office, and other weather agencies around the world contributed to the model's development. Both the code and the model weights behind WeatherNext are being made open source on GitHub, allowing other scientists to access and build upon this work.

A study about WeatherNext was published in the journal Nature, highlighting its ability to predict both a cyclone's track and intensity with a single model. This is significant because global atmospheric currents that determine a storm's path have traditionally been best analyzed by coarser global models, while a storm's intensity is best predicted by specialized local models.