In October 2025, an AI model developed by Google’s DeepMind and Google Research predicted Hurricane Melissa would strike Jamaica as a Category 5 hurricane five days before landfall, with 80 percent confidence. The model’s early warning allowed communities to prepare for the storm, which caused catastrophic flooding and landslides. The AI model, called WeatherNext, was tested on historical data and showed it can predict cyclones with greater accuracy than existing models.
According to the researchers, WeatherNext provides forecasters with an average of one additional day of lead time compared to current models. This means its predictions three days in advance are as accurate as previous models’ two-day forecasts. Mike Brennan, director of the US National Hurricane Center, emphasized that even a few hours can make a significant difference in organizing evacuations and managing resources. The model’s ability to push forecast accuracy out by a day is described as valuable for time-sensitive decisions.
Modeling extreme weather events like hurricanes is challenging for AI due to the rarity of such events, which limits training data. Ferran Alet, a research scientist at DeepMind and lead author of the study, explained that the team trained the model to be effective at both weather prediction and cyclone forecasting. Kate Musgrave, a tropical cyclone expert, noted that predicting a storm’s intensity requires detailed local data, which traditional global models often lack. WeatherNext, however, appears to perform well at both track and intensity predictions, offering a significant advancement in hurricane forecasting.
Source: wired
