How WeatherNext helped the National Hurricane Center better predict Hurricane Melissa’s historic landfall in Jamaica
DeepMind Blog · 2026-05-16
Google DeepMind's WeatherNext AI model helped the National Hurricane Center predict Hurricane Melissa's Category 5 landfall in Jamaica five days in advance — the first such prediction from a Category 1 starting point — enabling early evacuations for the historically strongest hurricane to strike the island.
Appears in
Extraction
Topics: ai-weather-forecastingtropical-cyclone-predictionrapid-intensificationgoogle-deepminddisaster-preparedness
Claims
- WeatherNext predicted Hurricane Melissa would reach Category 5 intensity from Category 1 wind speed five days before landfall, with 80% confidence rising to near 100% three days out.
- This was the first time any model successfully predicted a storm reaching Category 5 from such a low initial wind speed.
- WeatherNext was the top-performing individual model for both track and intensity across the full 2025 Atlantic hurricane season, per NHC's annual verification report.
- Traditional models face a trade-off between global track accuracy and local intensity resolution; WeatherNext bridges this gap by training on both global weather patterns and specialized tropical cyclone datasets.
- Google is expanding WeatherNext collaboration to meteorological agencies in the Philippines, Taiwan, Indonesia, and Vietnam.
Key quotes
WeatherNext, however, predicted a Category 5 strength landfall in Jamaica five days in advance with 80% confidence, which increased to near 100% three days in advance.
With early evacuation and better preparation, that reduction in harm really does make a difference to our people. [...] It does actually save their lives, and it saves the livelihoods that they want to secure.
By combining the speed and accuracy of AI with the irreplaceable experience of expert forecasters, we aim to reduce the human and economic toll of natural disasters.