How Big Tech AI models nailed forecast for Hurricane Lee a week in advance
As Hurricane Lee tracked slowly westward in the middle of the Atlantic Ocean on Sept. 10, three new weather models developed in the private sector predicted the storm would make landfall in Nova Scotia about a week later. With the storm still thousands of miles from North America, the forecasts turned out to be an astonishingly accurate achievement for technology not long ago considered to be in its infancy.
Generated by artificial intelligence, the models are orders of magnitude faster and cheaper to operate than conventional, government-run weather models. While AI models don’t yet provide all the capabilities needed for operational forecasting, their emergence portends a potential sea change in how weather forecasts are made, and they could signal a new chapter in the weather forecasting rivalry between the United States and Europe.
“There’s an incredible story emerging about the role of AI weather forecasts,” Daniel Rothenberg, an atmospheric scientist with Google sister company Waymo, said in an email. “This is a glimpse into the future of meteorology, and probably on much faster timescales than most folks in the weather enterprise expect.”
AI weather models have made rapid advances in the past 18 months. Google, Microsoft, NVIDIAand China-based Huawei all published academic articles claiming their AI models perform at least as well as the “European model,” widely considered the gold standard in weather modeling. Those claims were recently corroborated by scientists at the European Centre for Medium-Range Weather Forecasts, which operates the European model. Start-ups, including Atmo, Excarta and Zurich-based Jua, are building AI weather models as well.
msn.com |
