MIT Doubles Weather Forecast Reliability

MIT researchers use AI to identify systematic errors in existing weather models, doubling forecast reliability for the two-to-six-week period. Published in Nature, the technique aims to improve agriculture and disaster preparedness by correcting predictions that struggle beyond the standard ten-day window.