Phase Association via MCMC and Deep Learning.” While recent novel data-driven association techniques have shown promise for seismic phase association of local and regional earthquakes, their applicability on the global scale is still unclear. Dr. Williams et al. attempt to solve this by leveraging two deep learning models and a Markov Chain Monte Carlo (MCMC) method on a 17-year subset of the International Seismological Centre (SC) Bulletin catalog, with promising results. Find the full abstract here.
Dr. Williams will be presenting as part of session S42C: Machine-Learning-Based Earthquake Monitoring and Seismic Analysis II on Dec 15th; 9:00 AM – 12:30 PM CST in McCormick Place, Hall A. Find more information on the conference and how to attend (in person or online) on the AGU website.