Researchers eye AI revolution in natural disaster forecasts
Summary
Researchers in Switzerland are training AI models on vast NASA climate datasets and one of the world’s most powerful supercomputers to speed up weather and natural disaster forecasting. The work could improve early warning systems for events like landslides and glacial collapses, potentially saving lives in places such as Nepal.
Key Points
- Researchers at ETH Zurich are copying around 100 petabytes of NASA climate data onto servers next to the Alps supercomputer in Lugano.
- AI models can produce global weather forecasts in about a minute and may detect disaster warning signs earlier than traditional methods.
- The approach could strengthen early warning systems for hazards such as landslides and glacier collapses.
- Scientists cite Nepal’s recent glacial disaster as an example of an event that might have been flagged sooner with better monitoring.
