In addition to predicting extreme weather, GenCast was also more accurate in predicting wind power generated by groups of over 5,000 wind farms. The DeepMind team has made the GenCast code open source and the models available for non-commercial use. The team is also working on releasing an archive of historical and current weather forecasts, and the models are starting to power user experiences on Google Search and Maps.
Key takeaways:
- Google’s DeepMind team has released a new weather prediction model called GenCast, which outperforms traditional models in predicting extreme weather, tropical storm movement, and wind gust force.
- GenCast is a generative AI model that uses the most recent state of the weather to generate a probability distribution of future weather scenarios, unlike traditional models that solve physics equations.
- GenCast outperformed the leading traditional model, ENS, on 97.2% of targets using different weather variables, and was more accurate than ENS on 99.8% of targets with lead times greater than 36 hours.
- The DeepMind team has made the GenCast code open source and the models available for non-commercial use, and is working on releasing an archive of historical and current weather forecasts.