Physix Frontier · News Briefing Card (enbrief · Sep 4, 2026)
Google Launches WeatherNext 3 for Hourly, 5km AI Forecasts
KEY FACTS
- Google DeepMind has released the WeatherNext 3 model, enabling high-resolution global weather forecasts with hourly updates.
- The model achieves a 5-kilometer resolution for surface variables and 25 kilometers for atmospheric variables, trained on real-time satellite data.
- By leveraging sparse observational data to overcome computational limits, it delivers localized, precise forecasts for regions like Latin America and Africa.
- New features include renewable energy predictions, estimating wind speeds at 100 meters altitude and solar radiation levels.
- The tool is integrated immediately into core Google products including Search, Maps, Gemini apps, and Earth Engine.
KEY DATA
5 kmMax Spatial Resolution
HourlyForecast Update Frequency
100 mWind Forecast Altitude
PHYSIX OBSERVATION
WeatherNext 3 marks a shift in AI meteorology from general prediction to vertical scenario empowerment. Its core value lies not just in improved accuracy, but in filling data gaps in the Global South at low cost while directly addressing pain points in energy trading. When weather forecasts can precisely guide grid dispatch, the service upgrades from information provision to critical infrastructure, signaling that tech giants are using foundational data capabilities to reshape decision-making logic in traditional industries.
Source: enbrief original report ↗
Physix Frontier