Google’s AI weather model now uses more raw satellite data
Google launched WeatherNext 3, an AI weather model using raw satellite data that beats ECMWF and now powers Search, Gemini, and Maps.
Google released WeatherNext 3, an AI weather forecasting model that incorporates physical surface information (land/ocean type and elevation) to improve surface temperature and dewpoint calculations, improving point location temperature accuracy by up to 30 percent. Its white paper reports roughly 5 percent better upper-atmosphere accuracy than WeatherNext 2, equating to about six additional hours of forecast lead time, outperforming the ECMWF AI model on these metrics. The model now supplies forecast information across Google Search, Gemini, and Maps, though the paper notes unexplained short-lead degraded results and grid-shaped artifacts in some predictions.
- WeatherNext 3 adds physical surface data to ML forecasting, improving surface temperature accuracy up to 30%.
- Upper-atmosphere accuracy improved about 5% over WeatherNext 2, adding roughly six hours of lead time.
- Reported metrics beat the ECMWF AI weather model across a 15-day horizon.
- Model now powers forecasts in Google Search, Gemini, and Maps.
Full article355 words · extracted from arstechnica.com · click to collapse
Unlike traditional models that use physical properties in a location to simulate physical processes, machine-learning models are largely black boxes that train on past patterns and spit out predictions of future patterns. But WeatherNext 3 is adding a tiny bit of physical information to calculate surface temperature and dew point at any specific location you want to pull up. It checks whether that point is land or ocean and uses its surface elevation. By training on past weather station data tagged with that information, the team says they get better forecast predictions.
Some oddities
The white paper shows some results to document forecast performance improvements over WeatherNext 2, as well as the European Centre for Medium-Range Weather Forecasts (ECMWF) AI model.
They note a roughly 5 percent improvement in upper atmosphere condition accuracy over their previous model, for example, which they say equates to about six more hours of accurate forecast lead time. And their change to calculating surface temperature for a specific location improved accuracy by up to 30 percent. They’re generally beating the ECWMF model on these metrics as well.
There is one curious exception that the paper doesn’t even guess at the cause of. For a number of variables, their comparison to the initial six-hours-ahead forecast from the other models shows WeatherNext 3 doing worse before pulling ahead for the rest of a 15-day forecast.
Larger-scale patterns are also not without some weirdness. You can see the shape of the model’s grid in some predictions, like the map of precipitation showing some distinctly hexagonal blobs. Their method of generating multiple surface temperature forecasts to represent the range of possible outcomes also has a habit of producing snapshots where the global average temperature is higher or lower. Normally, you would want to see the average be consistent, with local-scale variability that averages out across the globe.
Overall, the team says their new model “represents a major step forward for AI-based weather predictions by going beyond relying purely on analysis and utilizing information-dense, low-latency observation data.”
WeatherNext 3 is now the source of forecast information across Google services, including Search, Gemini, and Maps.
Text extracted automatically; images, tables and formatting may be missing. Original: https://arstechnica.com/science/2026/09/googles-ai-weather-model-now-uses-more-raw-satellite-data/