Could AI Models Forecast Extreme Weather Events? with Pedram Hassanzadeh

About this episode

This episode discusses how AI models could improve the forecasting of extreme weather events.

What if we could predict the world’s most dangerous weather events—not days, but weeks in advance? Extreme events like heat waves, hurricanes, and floods cause massive loss of life and billions in damage, but they’re also some of the hardest events for traditional weather forecasting to predict. In this episode, Assoc. Prof. Pedram Hassanzadeh of the University of Chicago explains why forecasting extreme weather has long pushed science to its limits—and how a new wave of AI models could transform the field at a time when climate change is making these events more common. By learning directly from decades of atmospheric data, these systems can generate forecasts faster, more cheaply, and in some cases more accurately than traditional models—even to predict freak ‘gray swan’ weather events no one has ever seen.

People in this episode

Guest: Pedram Hassanzadeh

Topics covered

• AI
• extreme weather
• forecasting
• climate change
• atmospheric data

Keywords

• AI models
• extreme weather
• forecasting
• climate change
• atmospheric data
• heat waves
• hurricanes
• floods

Mentioned in this episode

Organizations: University of Chicago, University of Chicago Podcast Network

More episodes of Big Brains

• How Websites Trick You Using ‘Dark Patterns’ · June 11, 2026 · 37 min
• Quantum Technology, Explained: A Big Brains Live Event · May 28, 2026 · 40 min
• Our Untapped Superpower: Connecting With Others, with Nicholas Epley · May 14, 2026 · 39 min
• Why the Fed Matters Now More Than Ever, with Douglas Diamond · April 30, 2026 · 37 min
• Anxious? Avoidant? How to Build More Secure Relationships · April 16, 2026 · 37 min
• Are Judges Too Powerful? The Rise of Universal Injunctions, with Samuel Bray · March 19, 2026 · 40 min