S33B-2759
ElarmS Earthquake Early Warning System Updates and Performance

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Angela I Chung1, Richard M Allen2, Margaret Hellweg1, Ivan H Henson2 and Douglas S Neuhauser3, (1)Berkeley Seismological Lab, Berkeley, CA, United States, (2)University of California Berkeley, Berkeley, CA, United States, (3)Univ California Berkeley, Berkeley, CA, United States
Abstract:
The ElarmS earthquake early warning algorithm has been detecting earthquakes throughout California since 2007. It is one of the algorithms that contributes to CISN’s ShakeAlert, a prototype earthquake early warning system being developed for California.

Overall, ElarmS performance has been excellent. Over the past year (July 1, 2014 – July 1, 2015), ElarmS successfully detected all but three of the significant earthquakes (M4+) that occurred within California. Of the 24 events that were detected, the most notable was the M6.0 South Napa earthquake that occurred on August 24, 2014. The first alert for this event was sent in 5.1 seconds with an initial magnitude estimate of M5.7. This alert provided approximately 8 seconds of warning of the impending S-wave arrival to the city of San Francisco. The magnitude estimate increased to the final value of M6.0 within 15 seconds of the initial alert. One of the two events that were not detected by ElarmS occurred within 30 seconds of the M6.0 Napa mainshock. The two other missed events occurred offshore in a region with sparse station coverage in the Eureka area.

Since its inception, ElarmS has evolved and adapted to meet new challenges. On May 30, 2015, an extraordinarily deep (678km) M7.8 teleseism in Japan generated 5 false event detections for earthquakes greater than M4 within a minute due to the simultaneous arrival of the P-waves at stations throughout California. In order to improve the speed and accuracy of the ElarmS detections, we are currently exploring new methodologies to quickly evaluate incoming triggers from individual stations. Rapidly determining whether or not a trigger at a given station is due to a local earthquake or some other source (such as a distant teleseism) could dramatically increase the confidence in individual triggers and reduce false alerts.