Google says its new artificial intelligence model outperforms the best weather forecast system

The Google DeepMind team this week unveiled a weather prediction artificial intelligence model called GenCast.

IN article published in NatureDeepMind researchers said they found GenCast outperforms the European Center for Medium-Range Weather Forecasts’ ENS – apparently the world’s best operational forecasting system.

and blog postthe DeepMind team provided a more accessible explanation of the technology: while the previous weather model was “deterministic and provided a single best estimate of future weather,” GenCast “includes a collection of 50 or more forecasts, each representing a possible weather trajectory,” making a “complex probability distribution of future scenarios weather reports.

As for the comparison with ENS, the team said they trained GenCast on weather data through 2018, then compared their forecasts for 2019 and located that GenCast was more accurate 97.2% of the time.

Google says GenCast is an element of a collection of AI-powered weather models it’s starting to include into Google Search and Maps. It also plans to publish real-time and historical forecasts from GenCast that anyone can use in their very own research and models.