A research team from the University of Edinburgh, the British Geological Survey, and the University of Padua has created AI tools for predicting earthquake aftershocks almost instantly after a major quake. Utilizing extensive earthquake data from areas like California and Japan, these machine learning models can forecast aftershock locations and numbers within the first 24 hours post-earthquake. Unlike traditional models like ETAS, which take hours to simulate aftershock patterns, the new AI systems deliver forecasts in seconds, enhancing emergency response capabilities. They utilize high-resolution seismic data to identify invisible patterns in aftershock behavior, adapting continuously as new data arrives. The models achieved comparable accuracy to ETAS, promising significant improvements in public safety and resource mobilization during seismic events. By integrating these AI systems with real-time data feeds, researchers aim to establish a global earthquake monitoring network capable of providing rapid updates and insights into earthquake dynamics while complementing existing forecasting methods.
Source link
