For a wide variety of earthquake scenarios in Alaska, an earthquake early warning (EEW) system could provide at least 10 seconds of warning time for hazardous shaking, according to a new report.
Increasing the density and improving the spacing of seismic stations around the state could add 5 to 15 seconds to these estimated warning times, write Alexander Fozkos and Michael West at the University of Alaska Fairbanks. Alaska experiences tens of thousands of earthquakes each year, and has been the site some of the world's largest and most destructive seismic events.
The study's findings published in the Bulletin of the Seismological Society of America could help lay the groundwork for the expansion of the U.S. ShakeAlert earthquake early warning system, which now covers California, Oregon and Washington State.
"There were a lot of studies before EEW was widely available on the West Coast, where people were looking at different scenarios," said Fozkos. "So we wanted a similar kind of science up here with numbers that are Alaska specific."
For earthquakes along well-known faults in southcentral and southeast coastal Alaska, Fozkos and West estimated potential warning times of 10 to 120 seconds for magnitude 8.3 scenarios.
For magnitude 7.3 earthquake scenarios in crustal faults in interior and southcentral Alaska, the researchers estimated potential warning times ranging from 0 to 44 seconds.
And for a set of magnitude 7.8 earthquake scenarios along the dip of the subducting slab beneath Alaska, estimated warning times ranged from 0 to 73 seconds.
"I was expecting decent warning times along the coast and for most of the subduction zone events," said Fozkos, because there is dense seismic station coverage in these areas. "I was not expecting decent warning times for the shallow crustal events, so that was the biggest surprise to me."
The scenarios used in the study vary in earthquake magnitude, depth, location and fault style — all of which impacted warning times. The researchers' models estimated how many seconds after an earthquake's origin the quake could be detected, how many seconds after origin time an alert could be available, and minimum and maximum warning times at a location.
Warning times were defined as the time difference between the time of the alert and the time that peak ground motion from an earthquake arrived at a location. This definition differs from a more common definition used in EEW systems, which ties warning time to the arrival of the initial S-wave or shear wave of an earthquake.
The researchers wanted to use peak ground motion instead, to create a warning time measurement that might be more relevant to people as they respond to an earthquake. The initial S-wave may not always cause significant ground motion, and strong shaking can arrive tens of seconds after the initial S-wave in large earthquakes, they explain.
The study doesn't analyze "the time it takes to disseminate the alert—the time it actually takes to send the alert from a radio tower or from a satellite to somebody's phone and then for them to take out their phone and react to it," Fozkos noted.
The potential lag time in transmitting data and sharing an alert with the public "could be a big challenge for Alaska, but I don't think it's going to be insurmountable," he added.
The harsh Alaskan winters and wilderness locations of some seismic stations could also be challenging for an early warning system, if stations go down and can't be repaired quickly. "I think there is definitely a need for adding stations to cover redundancy for remote stations," Fozkos said.
Ocean-bottom seismometers (OBS) and more earthquake detection via distributed acoustic sensing or DAS would also be welcome additions to a warning system, he added. "Coupled with the fact that some of our biggest earthquakes are going to be offshore, tsunamigenic threats, I think OBS and DAS are probably big targets for the future."
