An AI with MIT participation improves weather forecasts 2 to 6 weeks ahead
A study published in Nature, with MIT participation, shows that an artificial intelligence improves weather forecasts between 2 and 6 weeks ahead, the interval scientists call the "predictability desert." The method, called Probabilistic Bias Correction (PBC), teaches the AI to identify and correct systematic errors in existing models. Applied to ECMWF models, it improved precipitation forecasts in 98% of the evaluated areas and doubled the predictive skill score of its AI system.
In 35 seconds
- The PBC technique corrects systematic errors in weather models instead of simulating the atmosphere from scratch.
- It improved precipitation forecasts in 98% of evaluated grid cells, temperature in about 92% and pressure in 91%.
- It doubled the average predictive skill score of ECMWF's AI system over the studied interval.
Read the full story at National Geographic España
IASummary generated by artificial intelligence from a story by National Geographic España. It may contain errors: the original source prevails.
Comments
Be respectful. Comments reported by several people are hidden until a moderator reviews them.
You do not need an account to read, only to comment. Sign in with a code sent to your email or with your Google account. We only show the name you choose, never your email.