Google has unveiled WeatherNext 2, an advanced artificial-intelligence weather-forecasting model designed to deliver faster and more accurate predictions for users and developers alike. The new system powers weather updates across Google Search, the Pixel Weather app, and Google Maps, and promises to generate 15-day forecasts in under one minute, marking a substantial leap from traditional forecasting methods.
In their announcing blog, Google explained how they want to roll out the feature to the world:
We’re now taking our research out of the lab and putting it into the hands of users. WeatherNext 2’s forecast data is now available in Earth Engine and BigQuery. We’re also launching an early access program on Google Cloud’s Vertex AI platform for custom model inference.
WeatherNext 2 replaces the previous system used by Google with a Functional Generative Network architecture. The last model was the one based on graph neural networks and conditional diffusion models. By leveraging Google’s Tensor Processing Unit chips, the company says the model can handle forecasts at scale across hundreds of thousands of data points in real time. Google also reported that WeatherNext 2 outperforms the previous version on key variables, such as wind speed, temperature, humidity, and pressure, capturing 99 percent of these more precisely.
WeatherNext 2 can generate forecasts 8x faster and with resolution up to 1-hour. This breakthrough is enabled by a new model that can provide hundreds of possible scenarios. Using this technology, we’ve supported weather agencies in making decisions based on a range of scenarios through our experimental cyclone predictions.
What sets the system apart is its integration across Google’s services: The model will soon be embedded into Google Maps and the full rollout is underway to replace older forecasting infrastructure.
Weather forecasting has traditionally relied on physics-based simulations running on supercomputers, which can take significant time to churn out results. Google’s shift to an AI-based approach signals a broader trend toward machine-learning-driven forecasts that reduce latency and hardware demands.
“Extreme weather events are becoming more common across the globe. So predicting weather rapidly and accurately is more critical than ever,” Google explains.
You can see more info about Google’s commitment to using AI for climate change monitoring here: