Science Help with Science Connections Snapshots Solid Earth Cryosphere Environmental & Hydrogeodesy Ocean Atmosphere Human Dimensions Technology

Testing Earthquake Early Warning Using Geodetic Data in the Pacific Northwest

Researchers: Brendan Crowell, David A. Schmidt, Paul Bodin, John E. Vidale, J. Renate Hartog and Victor C. Kress, University of Washington, Joan Gomberg and Sarah E. Minson, U.S. Geological Survey, Timothy I. Melbourne, Central Washington University and Dylan G. Jamison, University of Waterloo.

Written by Linda Rowan
16 August 2016


The addition of geodetic data into the Pacific Northwest Seismic Network earthquake early warning system shows important enhancement to the warning system. Tests show the Geodetic First Approximation of Size and Time (G-FAST) could determine the characteristics of the 2001 Mw 6.8 Nisqually earthquake with sufficient robustness to warn communities at risk. Some challenges such as integrating geodetic data into the U.S. Geological Survey’s ShakeAlert earthquake early warning system and ensuring a spatially and temporally sufficient geodetic network remain.


The Cascadia Subduction Zone, stretching from the Mendocino triple junction in northern California to north of Vancouver Island in British Columbia, has the potential for megathrust earthquakes that could generate large tsunamis. It could be an event like the 2011 Tohoku Mw 9.1 earthquake. The Pacific Northwest Seismic Network (PNSN) monitors earthquakes and provides the seismic data for the U.S. Geological Survey’s ShakeAlert earthquake early warning system.

Now PNSN is testing the addition of geodetic data into an earthquake early warning prototype system. Here they use the GPS sites that are part of the Pacific Northwest Geodetic Network (PANGA) and the Plate Boundary Observatory (PBO). They tested four different scenarios of geodetic data quality for the 2001 Mw 6.8 Nisqually earthquake and ran 1000 trials for each scenario.


G-FAST provides the first magnitude estimates from peak ground displacement (PGD) scaling at about 17 seconds after the Nisqually earthquake starts or about 6 seconds before strong shaking reaches Seattle. All four scenarios for determining the PGD magnitudes would provide stable estimates of magnitude about 30 seconds after the onset of the event.

The first estimates of the earthquake’s characteristics (i.e. magnitude, depth, strike, dip, rake, and slip) from the centroid moment tensor driven finite fault determinations would come about 38 seconds after the onset of the event. The distribution and low density of the GPS sites would have the most deleterious effect on determining the earthquake characteristics. As is the case for many subduction zones risk areas, the monitoring networks are primarily on the land and not in the ocean making the spatial distribution uneven.

The tests confirm that adding and integrating geodetic data into earthquake early warning systems would improve the warnings and help at-risk communities. The next steps involve enhancing and hardening geodetic networks and integrating geodetic data into earthquake and tsunami early warning systems.

Related Links


Demonstration of the Cascadia G-FAST Geodetic Earthquake Early Warning System for the Nisqually, Washington, Earthquake, Brendan W. Crowell, David A. Schmidt, Paul Bodin, John E. Vidale, Joan Gomberg, J. Renate Hartog, Victor C. Kress, Timothy I. Melbourne, Marcelo Santillan, Sarah E. Minson, Dylan G. Jamison, Seismological Research Letters, v. 87, no. 4 (2016), doi:10.1785/0220150255.


earthquake, tsunami, subduction zone, megathrust

Map Center
Seattle, Washington

Send questions or comments about this page to scienceunavco.org.

Last modified: Monday, 15-Aug-2016 22:28:00 UTC


Sponsored by

National Science Foundation Logo National Aeronautics and Space Administration Logo