No one can predict earthquakes, but what if cities in seismically unstable regions could receive advance notice of a major temblor’s shock waves? Could a few seconds’ warning save lives or reduce damage?
The technology exists and is being deployed along the West Coast, where 50 million people inhabit a region that could get a magnitude 8 or 9 quake. With far few people and a smaller seismic risk, Utah is next in line to get ShakeAlert, an early warning system developed by the U.S. Geological Survey.
That’s if it works for the Golden State and millions of dollars become available to install thousands of seismic-measuring instruments, build and operate a data-processing center and develop the communications infrastructure to relay alerts automatically.
In the wake of Utah’s magnitude 5.7 quake last week, officials extolled ShakeAlert’s potential to reduce destruction, deaths and injuries in the event of a big temblor that is expected in the coming decades. Such a system could come on line within five to eight years along the Wasatch Front, where the vast majority of Utah’s population lives near the state’s most serious seismic threat.
‘Not a lot of warning, but it is enough’
“It’s not a lot of warning, but it is enough warning for people to get under a table,” said Keith Koper, director of the University of Utah Seismograph Stations. “It’s enough warning for trains to stop. It’s enough warning for, let’s say, medical procedures to be halted. So even though it’s on the order of tens of seconds, we feel it is valuable.”
Just a few seconds is all that’s needed to automatically shut down gas transmission lines or roll up the bay doors at fire stations to prevent emergency vehicles from getting stuck inside.
Last week’s quake damaged at least 126 historic buildings in Salt Lake City and Magna and displaced the residents of a Salt Lake City apartment building and a West Valley City mobile home park. It resulted in no major injuries, although the rocking rattled thousands of Utahs already stressed from mandated social distancing to slow the spread of the coronavirus. The crumbling masonry from this quake is but of a picnic compared with what would happen if the area got hit with a magnitude 6.7 event, releasing 32 times more energy than the March 18 quake.
The early warning technology exploits variations of speed among different seismic waves as they travel through the Earth. When a fault ruptures, it releases P, or primary, waves, which cover up to 4 miles per second, followed by S, or secondary, waves, moving at half that speed, according to Robert de Groot, the USGS outreach coordinator for ShakeAlert. Causing little shaking, P waves strike sensors first, providing clues to the earthquake’s size and location. Moving side to side with much higher amplitudes, the late-coming S waves are what knock down buildings.
When the system detects a large quake, it instantly calculates the intensity of shaking expected for particular places, factoring the magnitude and distance from the epicenter. That information is then used to determine whether alerts and shutdown signals are automatically sent.
Imagine Paul Revere galloping through Lexington in front of a seismic British army, shouting “a quake is coming, a quake is coming.” Be sure to “drop, roll and hold on.”
Such early warning systems are used in Japan, Mexico and Turkey. One came on line last year in California with plans to expand coverage to Oregon and Washington. USGS won’t make a commitment to build a system in Utah or anywhere else until it completes the West Coast project, according to de Groot. While Southern California is well covered with sensors, the network is less than two-thirds complete as USGS and its state partners fill in less-populated parts of Oregon and Washington.
“The goal is 1,675 seismic stations. We are only 63% there. That reduces accuracy, but we can still do an effective warning with what we have,” de Groot said. “It’s like having greater pixel density in a digital camera.”
Data-processing centers in Pasadena, Menlo Park and Seattle crunch the information collected and transmitted by these sensors.
‘Working out the kinks’
In the year since its installation, ShakeAlert has been activated several times in response to movements along the massive continental faults where the North American and Pacific plates meet.
“They’re still working out the kinks,” Koper said. “It does seem to be working.”
One big kink was exposed last July during the Ridgecrest earthquakes, 120 miles north of Los Angeles, when a magnitude 6.4 temblor struck on July 4, followed by a magnitude 7.1 the next day. ShakeAlert underestimated the amount of shaking that was to be associated with these quakes, especially the larger one, explained de Groot.
As a result, the predicted jerking did not meet the thresholds for triggering alerts for many places. Even though the system detected the quakes, the shock waves from the magnitude 7.1 hit Los Angeles at 8:19 p.m. without dispatching alerts to those who had the ShakeAlertLA app installed on their cellphones, de Groot said, prompting some to conclude the system failed.
The West Coast project has cost $30 million to build so far and runs another $16 million a year to operate, Koper noted. A Utah system would not be so pricey.
“We would benefit from the lessons they learned in developing it. We would have to adapt it to our region,” Koper said. “We would need many more sensors and fancier recorders.”
Threat from the Wasatch fault
One drawback for Utah is the fact that much of the state’s population and critical infrastructure are within the Wasatch fault zone. Salt Lake City would not get more than a second or two of warning if a destructive quake occurred on its leg of the fault.
The Wasatch fault system’s network of cracks in the earth stretches 230 miles from Malad, Idaho, south to Fayette, Utah, through the Beehive State’s major metropolitan areas. A rupture on this fault could affect up to 87% of Utahns, said Bob Carey, earthquake program manager for the state Division of Emergency Management.
A magnitude 6.75 quake on the main fault could unleash catastrophic carnage, depending on where it strikes. A 2016 report forecast a 43% chance of such a quake or stronger within the next 50 years. One report warned that a quake of that magnitude in Salt Lake City could do $33 billion in damage, displace 84,400 households, injure 9,300 people and kill 2,300.
Scientists do not believe last week’s temblor and its 600-plus aftershocks have reduced the chance of a major quake on the Wasatch fault.
The Wasatch fault is not one continuous fracture but rather a system of nine distinct segments. The five internal segments between Brigham City and Nephi are the most seismically active and cover the most densely populated region.
Large earthquakes occur on these segments in roughly 1,300-year intervals, Carey said, although he cautioned that the geological record indicates a randomness associated with their recurrence. The Weber and Provo segments have seen major seismic events 300 to 600 years ago, suggesting they don’t pose much of risk now, but others are long overdue.
“The last big event on the Brigham City segment was 2,200 years ago. We are treating that as a locked segment. It’s dangerous,” Carey said. “There is sufficient strain to have a big event today.”
In theory, the P waves from an earthquake near Brigham City would reach Salt Lake City at least 7.5 seconds before the destructive S waves strike. Advance notice of a few seconds could make a difference in reducing harm, officials say, if people can use the time to “drop, cover and hold on” before the main shock waves strike.
Even before this month’s wake-up shaker, the Utah Geological Survey had planned to host a multiagency workshop next February to explore whether ShakeAlert would be a smart investment for Utah.
“There are pros and cons. Why not spend the money to make roads better so they won’t collapse?” Koper said. “It’s not a simple answer, but we have to evaluate it because it could be cost-effective and something that people are going to want if it could actually save lives.”