GTC’s CTO Jesse Williams recently chaired a panel at the 2022 Seismological Society of America Annual Meeting. The session, “Machine Learning Techniques for Sparse Regional and Teleseismic Monitoring”, centered on applying machine-learning applications to seismic processing problems on sparser global and continental scales. The scale of these analyses present extra challenges, such as the detection and identification of additional phase types (such as local, regional and teleseismic P and S arrivals), as overlap among arrival types and network sparsity can prevent the direct application of local and dense network approaches that rely on moveout patterns.
The session involved presentations on methods that can enhance the performance of steps in both retrospective and near-real-time network processing, and methods that combine multiple steps or imagine processing into novel pipelines enabled by data science methods or high performance computing capabilities.
During the panel, Georgia Tech student Lindsay Chuang presented on joint work between GTC and Georgia Tech, “Towards a Dynamic Multi-net Approach for Earthquake Association” (Chuang, L. Y., Williams, J., Barama, L., Peng, Z., Newman, A. V.).
Full details and abstracts from the event can be found here.
