Insurance
Home›Insurance›Property Insurance›AI forecasting aims to improve earthquake warnings up…
AI forecasting aims to improve earthquake warnings up to 25 days ahead
Munich Re estimates a repeat of the 1906 San Francisco earthquake could cause $200 billion or more in direct losses, while California earthquake coverage remains under 15%.
Risk & Insurance says a new AI approach using cosmic radiation data from deep space, combined with satellite imagery and machine learning, is designed to forecast earthquakes up to 25 days before they strike.
The piece cites a Munich Re estimate that a repeat of the 1906 earthquake in the modern San Francisco Bay Area would produce direct losses of $200 billion or more, noting the figure has roughly doubled since the last major estimate as technology and infrastructure in Silicon Valley have grown.
It also points to U.S. Geological Survey figures placing the probability of a magnitude 6.7 or higher earthquake hitting the San Francisco Bay Area within the next 30 years at 63 to 72%, and says less than 15% of California homeowners carry earthquake coverage.
Risk & Insurance argues that longer lead times could change preparedness compared with existing systems that issue warnings only seconds ahead, giving examples such as shutting down equipment or pipelines and evacuating non essential personnel before a major quake.