Google has just released version 3 of its WeatherNext model, a significant leap in its artificial intelligence weather prediction capabilities. The latest update now incorporates real-time satellite data, reducing the delay between current weather conditions and new forecasts. This improvement means the model can now generate hourly weather maps, providing more frequent and accurate predictions.

The new version also enhances the spatial resolution of the model, ensuring it captures more detailed weather patterns. Additionally, a new machine-learning model has been trained specifically for satellite-based precipitation estimates, giving users multiple, more precise precipitation forecasts. These changes mean that WeatherNext 3 can better simulate the atmosphere using a blend of reanalysis data and raw satellite information, offering a more comprehensive picture of current weather conditions.

The shift to including satellite data not only shortens the lag time between observation and forecast but also provides a more up-to-date snapshot of the atmosphere. While this update may seem small, it represents a significant step forward in the accuracy and timeliness of weather forecasts. In the long run, this could lead to better preparation for severe weather events, improved agricultural planning, and more efficient use of energy resources.