Weather changes often happen quickly, making accurate predictions difficult. Therefore, timely weather information remains important for daily planning, farming, power generation, and disaster preparation.
Now, Google has introduced WeatherNext 3, a new AI model designed to deliver faster and more detailed weather forecasts. The model aims to improve prediction accuracy with better resolution.
WeatherNext 3 brings major upgrades in weather forecasting
Google DeepMind and Google Research launched WeatherNext 3 on September 3. According to Google, it is the company’s most advanced global weather AI model so far.
The model uses live geostationary satellite data and creates fresh weather forecasts every hour. Earlier, WeatherNext 2 produced forecasts every six hours using a 25-kilometre grid.
However, WeatherNext 3 provides surface weather details with a five-kilometre resolution. It uses 10-kilometre resolution for other surface data and 25-kilometre resolution for atmospheric information.
Moreover, Google says the new model creates a weather view almost five times clearer than the previous version.
AI model improves rainfall and energy predictions
WeatherNext 3 focuses on rapidly changing weather conditions, especially rainfall predictions. Google trained the model using NASA’s IMERG and satellite radar-based global precipitation data.
According to Brightband, WeatherNext 3 delivered up to 60 percent better results than IMERG for medium-range weather forecasts. Additionally, it showed a 30 percent improvement compared with MRMS.
The model also achieved 10 percent better performance than rain gauge data for short-term forecasts.
WeatherNext 3 expands beyond regular weather updates
Furthermore, WeatherNext 3 can support renewable energy planning. The model predicts wind speed at 100 metres altitude to estimate wind energy generation.
It also provides high-resolution cloud cover and sunlight level forecasts. Therefore, solar farms can better estimate their possible electricity production capacity.
Google adds WeatherNext 3 across multiple platforms
Google will integrate WeatherNext 3 into Search, Gemini app, Google Maps, Google Maps Platform Weather API, and Google Earth Engine.
Additionally, developers, researchers, and business users can access weather data through BigQuery, Earth Engine, and Google Cloud Storage.
According to Google, this model can be especially useful for regions like Latin America, Africa, and Asia-Pacific.
