Google says WeatherNext helped the National Hurricane Center predict Hurricane Melissa's landfall in Jamaica
Google DeepMind said its WeatherNext model was used by the National Hurricane Center as part of its forecasting suite for Hurricane Melissa, providing 50-member ensemble forecasts alongside traditional physics-based mode.
At a glance
- Google said the NHC used WeatherNext ensemble forecasts (50 scenarios) in predicting Hurricane Melissa's track and intensity.
- Google said WeatherNext called the storm's Category 5 intensification five days out at 80% confidence, near 100% three days before landfall.
- Google cited the NHC's 2025 verification report as ranking WeatherNext the top individual model for track and intensity in the Atlantic and Eastern Pacific.
VERDICT — CONFIRMED
Google DeepMind said its WeatherNext model was used by the National Hurricane Center as part of its forecasting suite for Hurricane Melissa, providing 50-member ensemble forecasts alongside traditional physics-based models. The company said WeatherNext predicted the storm would reach Category 5 strength five days in advance with 80% confidence, rising to near 100% three days before its Jamaica landfall, from an initial system at Category 1 wind speeds.
Google also cited the NHC's 2025 verification report as ranking WeatherNext the top-performing individual model for both track and intensity across the Atlantic and Eastern Pacific. This is a company case study, so the performance framing is Google's, though the NHC verification citation gives it an external anchor.
Key facts on file
- Google said the NHC used WeatherNext ensemble forecasts (50 scenarios) in predicting Hurricane Melissa's track and intensity.
- Google said WeatherNext called the storm's Category 5 intensification five days out at 80% confidence, near 100% three days before landfall.
- Google cited the NHC's 2025 verification report as ranking WeatherNext the top individual model for track and intensity in the Atlantic and Eastern Pacific.
