AI Loses Edge In Predicting Extreme Weather Events — For Now
Low frequency, high-impact phenomena are the most destructive and the hardest to forecast, making greater accuracy and earlier warnings crucial to saving lives.
A military firefighter works to extinguish a wildfire in Ourense, Spain, on Aug. 19.
Photographer: Pablo Blazquez Dominguez/Getty Images EuropeThis article is for subscribers only.
Welcome to Weather Watch, our weekly newsletter on how the planet’s ever wilder weather patterns are impacting the global economy. Got feedback and forecasts? Write us at weatherteam@bloomberg.net. And sign up here if you’re not on the list already.
AI models are setting new benchmarks for weather forecasting, but their traditional, physics-based counterparts appear to still have an edge in one critical area: predicting extreme events.