Forecasters at the National Hurricane Center (NHC) in Miami relied on artificial intelligence (AI) last year to accurately predict the rapid intensification of Hurricane Melissa ahead of its arrival in Jamaica. The storm, which strengthened quickly from a relatively mild condition to a highly dangerous hurricane, underscored the potential of AI-enhanced forecasting methods introduced by the center earlier in 2023.

This advancement was developed in collaboration with DeepMind, a unit of Google, which designed the AI model credited for providing earlier and more precise forecasts than traditional systems. In a recent study published in the journal Nature, DeepMind researchers evaluated hundreds of hurricane forecasts spanning from 2023 to 2025. Their findings suggest that the AI model can predict hurricane behavior with greater accuracy and provide earlier warnings—up to a day or more in advance—compared to widely used conventional models.

Amy McGovern, a meteorology professor at the University of Oklahoma and director of an AI weather institute, praised the development, noting that extended lead times from accurate forecasts could allow for better preparation and evacuation efforts, potentially saving lives and reducing property damage. “If you can tell people a day in advance about a major hurricane, they’ll have more time to evacuate and prepare, in theory saving lives and property,” McGovern said.

However, experts caution that human expertise remains essential in interpreting AI-generated predictions. Kerry Emanuel, an emeritus professor of atmospheric science at the Massachusetts Institute of Technology, acknowledged the progress but emphasized that meteorologists will continue to play a critical role in making sense of complex data and ensuring accurate communication to the public.

The integration of AI into hurricane forecasting marks a significant step forward in weather prediction technology, promising to enhance the timeliness and reliability of warnings during increasingly volatile storm seasons. While challenges remain in bridging AI capabilities with traditional meteorological analysis, experts agree that this approach has the potential to improve disaster preparedness and response on a broad scale.