Exciting news! We're transitioning to the Statewide California Earthquake Center. Our new website is under construction, but we'll continue using this website for SCEC business in the meantime. We're also archiving the Southern Center site to preserve its rich history. A new and improved platform is coming soon!

Efficient blind search for small similar-waveform earthquakes in a decade of continuous seismic data (2007-2017) in coastal central California

Clara Yoon, Karianne Bergen, Kexin Rong, Hashem Elezabi, Peter Bailis, Philip Levis, & Gregory C. Beroza

Published August 15, 2017, SCEC Contribution #7749, 2017 SCEC Annual Meeting Poster #024 (PDF)

Poster Image: 
Template matching has proven to be a sensitive, discriminating detector of small events missing from earthquake catalogs, but a key limitation of this approach is that it requires prior knowledge of the earthquake signals we wish to detect.

The Fingerprint And Similarity Thresholding (FAST) detection method (Yoon et al., 2015) enables a comprehensive blind search for similar-waveform earthquakes in a fast, scalable manner by adapting data-mining techniques originally developed for audio and image search within massive databases. FAST converts seismic waveforms into compact “fingerprints”, which are efficiently organized and searched within a database. In this way, FAST avoids the unnecessary comparison of dissimilar waveforms that otherwise leads to quadratic scaling of compute time with data duration. To date, the longest duration of continuous data used for event detection with FAST was 3 months at a single station near the Guy-Greenbrier, Arkansas, earthquake sequence. The application revealed abundant microearthquakes closely correlated in time, location, and morphology with stages of hydraulic fracturing (Yoon et al., 2017).

We introduce an optimized, parallel version of the FAST software that includes improvements to the fingerprinting algorithm and the ability to detect events using continuous data from a network of stations (Bergen et al., 2016). We apply FAST to detect low-magnitude similar-waveform earthquakes within a decade of continuous seismic data near the Diablo Canyon Nuclear Power Plant in coastal central California. Although the rate and magnitude of seismicity near the power plant are low, the presence of recently discovered nearby faults (Hardebeck, 2010; Hardebeck, 2013) motivates further investigation of microearthquake activity that might illuminate potential sources.

Key Words
earthquake detection, earthquake monitoring, data mining, seismicity

Citation
Yoon, C., Bergen, K., Rong, K., Elezabi, H., Bailis, P., Levis, P., & Beroza, G. C. (2017, 08). Efficient blind search for small similar-waveform earthquakes in a decade of continuous seismic data (2007-2017) in coastal central California . Poster Presentation at 2017 SCEC Annual Meeting.


Related Projects & Working Groups
Special Projects