Mar 06, 2025 12:50:00

Google announces that it will offer its AI-based weather forecasting model 'WeatherNext' to businesses

Google Cloud, Google's cloud computing service, has announced that it will begin offering the AI weather forecasting model '

WeatherNext ' developed by Google DeepMind and Google Research to businesses. WeatherNext is faster and more efficient than conventional physics-based weather forecasting models, and its high predictive performance is expected to help various industries prepare for abnormal weather.

WeatherNext - Google DeepMind
https://deepmind.google/technologies/weathernext/

Google Cloud unveils AI weather models for the energy industry
https://www.axios.com/2025/03/05/google-cloud-ai-weather-predictions

Weather forecasting requires processing huge amounts of data from complex weather systems, and not all variables can be explained or measured, making it difficult to make speedy and accurate forecasts. In particular, abnormal weather such as heat waves, cold waves, and hurricanes not only affect life and the environment, but also cause large economic losses. According to data from the National Center for Environmental Information in the United States, there have been 403 weather disasters in the United States with total damages of more than $1 billion (about 140 billion yen) in the 44 years from 1980 to 2024, and 27 of these will occur in 2024.

Weather forecasts become inaccurate after a few days, making it difficult to prepare for extreme weather. However, the weather forecasting model 'WeatherNext' uses AI to process data on a larger scale to provide accurate predictions.

WeatherNext has two models, 'WeatherNext Graph' and 'WeatherNext Gen,' which can be used according to the purpose. WeatherNext Graph is a model that provides a more accurate and efficient single forecast, and can efficiently and quickly produce accurate forecasts. WeatherNext Gen is a model used for 'ensemble forecasts' that predict various weather scenarios, and is suitable for preparing for longer-term weather events.

'WeatherNext is designed to change the way companies approach weather-impacted operations, including retail inventory, logistics disruptions, manufacturing production needs, and power line maintenance,' said Carrie Tharp, vice president of Global Solutions and Industry at Google Cloud. 'By delivering WeatherNext, companies can make more informed decisions as weather patterns change, better protect their infrastructure, and enable stronger business continuity.'

Regarding the significance of making WeatherNext available to customers, Pete Battaglia, Director of Sustainability Research at Google DeepMind, said, 'By opening up WeatherNext to businesses, we are expanding its range of applications from the lab to the real world. This will enable companies to proactively prepare for extreme weather and better serve their communities.'

Google hopes that WeatherNext will help predict and respond to extreme weather, as well as help decide where to build renewable energy infrastructure. According to Battaglia, the approach to AI weather forecasting models is still in its infancy, so it is expected that Google DeepMind will work with the National Oceanic and Atmospheric Administration (NOAA) , a US government agency specializing in atmospheric and oceanic research, to advance the research.