正在加载视频...

视频加载失败

Artificial intelligence (AI) is making waves in hurricane forecasting. 🌊🤖 NHC’s Science and Operations Officer Dr. Wallace Hogsett explains how forecasters use their expertise to guide & apply AI to make forecasts even better. Watch here 👇 🎥

12,474 次观看 • 10 个月前 •via X (Twitter)

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

AI JUST OUTPERFORMED TRADITIONAL HURRICANE MODELS - AND IT’S ABOUT TO REWRITE WEATHER FORECASTING This year’s hurricane season quietly delivered one of the biggest breakthroughs in climate science: AI didn’t just compete with the world’s best forecasting models - it outperformed almost all of them. When the National Hurricane Center agreed to test Google DeepMind’s experimental forecast engine, it wasn’t clear whether the AI model could handle something as chaotic and physics-driven as tropical cyclone behavior. Then Hurricane Melissa happened - and everything changed. DeepMind saw a Cat-5 coming before anyone else. Forecasters say the AI’s rapid-intensification prediction for Melissa was so precise, so early, that it fundamentally shaped the NHC’s risk messaging. And unlike the classic supercomputers that need hours of crunch time, DeepMind generated hundreds of weather scenarios in minutes. Even veteran hurricane specialists who’ve seen every model in the book said the same thing: “Unquestionably critical… especially on rapid intensification.” Rapid intensification has always been the Achilles’ heel of hurricane modeling. AI nailed it. Here’s what makes this different: Traditional models use massive physics simulations - accurate but slow. AI models “learn” from 40+ years of atmospheric data - fast, adaptive, insanely scalable. DeepMind became the most accurate model of the season, beaten only by the NHC’s human-crafted official forecast. Let that sink in: A machine-learning system just outperformed the U.S. and European legacy models that governments have poured billions into. But the experts are divided - and here’s the warning. AI is extraordinary at recognizing patterns that existed before. But climate change is producing storms outside historical patterns. That’s the catch. AI might see the past perfectly… but the future isn’t guaranteed to look like the past. This is why meteorologists stress: AI can augment the forecast stack - it cannot replace physics-based modeling. Not yet. Still, the writing is on the wall. The same storm that devastated Jamaica may end up rewriting the science of prediction itself. And if Melissa was the test case? The future of hurricane forecasting isn’t just faster. It’s fundamentally smarter. Source: abc News

Mario Nawfal

23,019 次观看 • 7 个月前