A team comprising researchers from Edinburgh University, the British Geological Survey, and the University of Padua in Italy has developed advanced machine learning tools aimed at enhancing earthquake aftershock predictions. The technology utilizes seismic data from known earthquake-prone regions, including California, New Zealand, Italy, Japan, and Greece, and is designed to predict the number of aftershocks that may occur within 24 hours following an earthquake of magnitude four or higher.
The new artificial intelligence model demonstrates performance comparable to that of the Epidemic-Type Aftershock Sequence (ETAS) model, widely used in countries such as Italy, New Zealand, and the United States. However, while both systems yield similar results, the AI model provides predictions almost instantly, in contrast to the more computationally demanding ETAS, which can take several hours to generate forecasts. This research has been published in the journal Earth, Planets and Space.
“This study shows that machine learning models can produce aftershock forecasts within seconds, showing comparable quality to that of ETAS forecasts,” said Foteini Dervisi, a PhD student at Edinburgh University’s School of GeoSciences and a lead researcher on the study. Dervisi emphasized the significance of the AI model’s speed and low computational cost, which present substantial advantages for operational use. “Coupled with the near real-time development of machine learning-based high-resolution earthquake catalogues, these models will enhance our ability to monitor and understand seismic crises as they evolve,” she added.
By leveraging historical earthquake data from various tectonic landscapes, the research team believes that their AI models could offer valuable aftershock risk forecasts across most regions that regularly experience seismic activity. The rapid delivery of aftershock predictions is particularly crucial, as it can aid authorities in making informed decisions regarding public safety measures and resource allocation in areas affected by disasters.
The study received support from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie SPIN Innovative Training Network. As the world grapples with the impact of natural disasters, advancements such as this one underscore the potential role of machine learning in disaster preparedness and response.
See also
New York City Council Passes GUARD Act for AI Transparency and Accountability
Gao et al. Transform Lung Cancer Inflammation Index into AI Tool for Colorectal Cancer Risk Assessment
Vine Relaunches as DiVine with 100K Archived Videos, Bans AI-Generated Content
Dtonic Launches AI-Driven ‘Pharmkeeper’ for Enhanced Pharmacy Security and Efficiency
Colombian Students Win NASA Challenge with AI Tool Detecting Exoplanets in 48 Hours
