Detecting millions of earthquakes in southern California with template matching

Zachary E. Ross, Egill Hauksson, Daniel T. Trugman, & Peter M. Shearer

Submitted August 6, 2018, SCEC Contribution #8287, 2018 SCEC Annual Meeting Talk on Tue 14:00

Over the last twenty years, earthquake detection rates in southern California have improved dramatically, resulting in the minimum magnitude of completeness decreasing from M~2.5 to M~1.5 today. It is believed, however, that these events still constitute less than 10% of all activity that is being recorded by the seismic network on a regular basis. To address these shortcomings, we applied a matched filter (template matching) algorithm to the entire continuous waveform archive of the Southern California Seismic Network using the seismograms of ~300,000 past events as templates. This GPU supercomputing effort resulted in a catalog of ~2.4 million earthquakes for the period 2008-2017, which is ~13 times as many events as the standard regional catalog, and has a completeness magnitude of ~0.5. The recent double-difference GrowClust algorithm was applied to the entire dataset and its 1.3 billion differential times, resulting in state-of-the-art hypocenter precision for the whole of southern California. I will first discuss basic summary information about the catalog and new regional-scale observations. Then, I will focus on the most active sequences that occurred during the period and use the seismicity to investigate connections between properties of fault zones and the earthquake rupture process. The unprecedented level of detail in this next-generation seismicity catalog is expected to facilitate important new analyses of earthquakes and faults in southern California.

Citation
Ross, Z. E., Hauksson, E., Trugman, D. T., & Shearer, P. M. (2018, 08). Detecting millions of earthquakes in southern California with template matching. Oral Presentation at 2018 SCEC Annual Meeting.


Related Projects & Working Groups
Mining Seismic Wavefields (MSW)