Published August 1, 2020 | Version Supplemental Material
Journal Article Open

ShakeAlert Earthquake Early Warning System Performance during the 2019 Ridgecrest Earthquake Sequence

  • 1. ROR icon University of California, Berkeley
  • 2. ROR icon California Institute of Technology
  • 3. ROR icon ETH Zurich
  • 4. ROR icon University of Washington
  • 5. ROR icon United States Geological Survey

Abstract

During July 2019, a sequence of earthquakes, including an M_w 6.4 foreshock and an M_w 7.1 mainshock, occurred near Ridgecrest, California. ShakeAlert, the U.S. Geological Survey (USGS) earthquake early warning system being developed for the U.S. West Coast, was operational during this time, although public alerting was only available within Los Angeles County. ShakeAlert created alert messages for the two largest events and for many of the larger aftershocks. In this study, we dissect log files and replay data through the system to reconstruct the sequence of events and analyze the performance of the system during that period. Although the system performed reasonably well overall, the sequence also revealed various issues and short comings that will be addressed in impending and future system upgrades. ShakeAlert detected and characterized both the M_w 6.4 and M_w 7.1 earthquakes within 6.9 s of their origin times and created alert messages that were available to ShakeAlert's pilot users. No public alerts were sent out by the ShakeAlertLA cell phone app (the only publicly available alerting method at the time), because the predicted shaking for Los Angeles County was below the app's alerting threshold of modified Mercalli intensity 4.0. For the M_w 6.4 event, this was accurate. For the M_w 7.1 event, public alerts for Los Angeles County were warranted, but ShakeAlert underpredicted the shaking levels, because both the point‐source and finite‐fault algorithms underestimated the magnitude of the earthquake by 0.8 units. A number of software and hardware issues that were responsible for the magnitude underestimation of the mainshock have been identified and will be addressed in future ShakeAlert releases. We also analyze the hypothetical alerting performance of ShakeAlert had public alerting been available throughout southern California.

Additional Information

© 2020 Seismological Society of America. Manuscript received 17 January 2020; Published online 21 July 2020. The authors would like to thank Valerie Thomas, Elizabeth Cochran, Stephen Hickman, Sarah Minson, Associate Editor Susan Hough, and an anonymous reviewer for constructive comments that helped improve the article. The authors also thank Igor Stubailo for his valuable input. This material is based upon work supported by the U.S. Geological Survey under Grant Agreement Numbers G17AC000346, G17AC000345, G17AC000347, G19AC00252, and preceding grants. Data and Resources: Log files used in this study were obtained from the ShakeAlert production machines. Seismograms used in this study were obtained from the Southern California Earthquake Data Center (SCEDC). Data can be obtained from the SCEDC at https://scedc.caltech.edu/ (last accessed December 2019). California Integrated Seismic Network (CISN) ShakeMaps are obtained from https://earthquake.usgs.gov/earthquakes/eventpage/ci38443183/shakemap/intensity?source=ci&code=ci38443183 (last accessed April 2020). The U.S. Geological Survey (USGS) finite‐fault model is obtained from https://earthquake.usgs.gov/earthquakes/eventpage/ci38457511/finite-fault (last accessed April 2020). The SCEDC and Southern California Seismic Network (SCSN) are funded through USGS Grant G10AP00091, and the Southern California Earthquake Center, which is funded by National Science Foundation (NSF) Cooperative Agreement EAR‐0529922 and USGS Cooperative Agreement 07HQAG0008. QuakeAlertUSA app information is obtained from https://earlywarninglabs.com (last accessed March 2020). The supplemental material contains a table of individual station magnitude estimates from the final earthquake point‐source integrated code (EPIC) update for the M_w 7.1 earthquake, EPIC uncertainty calculation description, and also figures comparable to Figures 7–10 with different alerting (⁠MMI_(Alert)⁠) and ground‐motion (⁠MMI_(tw)⁠) thresholds. Also included is a text file containing all M 4+ alerts in southern California created by ShakeAlert from 4 July to 17 July 2019. The unpublished manuscript by Stubailo, I., M. Alvarez, G. Biasi, R. Bhadha, and E. Hauksson (2020). "Latency of waveform data delivery from the Southern California Seismic Network during the 2019 Ridgecrest earthquake sequence and its effect on ShakeAlert," submitted to Seismol. Res. Lett. The MATLAB is available at www.mathworks.com/products/matlab (last accessed June 2020).

Attached Files

Supplemental Material - bssa-2020032_supplement.docx

Supplemental Material - bssa-2020032_supplement.zip

Files

bssa-2020032_supplement.zip

Files (3.0 MB)

Name Size
md5:8ba1c57fea66ce51a549f83061fd0e0d
3.0 MB Download
md5:e756ea9d364fe168f84c9606786a3fa4
3.6 kB Preview Download

Additional details

Identifiers

Eprint ID
104596
DOI
10.1785/0120200032
Resolver ID
CaltechAUTHORS:20200727-125421047

Related works

Describes
10.1785/0120200032 (DOI)

Funding

USGS
G17AC000346
USGS
G17AC000345
USGS
G17AC000347
USGS
G19AC00252
USGS
G10AP00091
Southern California Earthquake Center (SCEC)
NSF
EAR‐0529922
USGS
07HQAG0008

Dates

Created
2020-07-28
Created from EPrint's datestamp field
Updated
2021-11-16
Created from EPrint's last_modified field

Caltech Custom Metadata