The earthquakes that jolt us out of bed, the ones that trigger an instinctual urge to duck and cover, are powerful and potentially destructive.
But they’re also rare: For every magnitude 7.0 temblor, roughly a million magnitude 1.0 earthquakes occur.
Earthquake detection algorithms must be able to accurately disentangle these numerous, low-amplitude seismic signals from other sources of ground shaking such as aircraft passing overhead and waves crashing into coastlines. Now, researchers have analyzed the waveforms of wind-induced ground motion, work that will boost the performance of future earthquake detection algorithms, the scientists suggest.
Research on a Ranch
Christopher Johnson, a seismologist at the Scripps Institution of Oceanography in San Diego, Calif., and his colleagues collected data on a privately owned unused ranch situated atop the San Jacinto fault zone near Anza, Calif. Johnson and his team placed 40 geophones around the property’s stands of vegetation, old structures, abandoned machinery, and defunct airstrip. The coffee can–sized instruments, anchored to the ground with a roughly 12-centimeter-long metal spike, recorded ground velocity measurements 500 times per second from 9 February through 17 March 2018.
The scientists wanted to better understand ground movement caused by wind energy being transferred into Earth through, for example, plant roots or building foundations. That’s important, researchers said, because wind-induced ground shaking—which affects roughly the top kilometer of Earth’s crust—is a source of noise in seismic records.
“We need to have a really good understanding of what all of these noise signals are,” said Johnson.
Seismic records are increasingly being mined to find new fault lines and better understand how one earthquake triggers others.
To test how ground shaking was correlated with wind strength, the scientists also measured wind velocity using an anemometer mounted near the ranch’s unused airstrip. They recorded an average wind velocity of roughly 2 meters per second and a maximum wind velocity of about 15 meters per second.
Masquerading as an Earthquake
Johnson and his collaborators found that gusts of wind stronger than a few meters per second were linked to ground shaking characterized by earthquake-like waveforms. The team examined the vertical and horizontal components of the wind-triggered shaking and found a consistently larger signal in the horizontal direction. That’s because the wind is pushing aboveground objects—like structures, vegetation, and machinery—with a shearing motion with respect to the ground, said Johnson.
Next, the researchers investigated the strength of the wind-triggered shaking at different locations on the ranch. They recorded the weakest signal on a hill and the strongest signal near a covered parking structure. These results make sense, said Johnson, because human-built structures are coupled to the ground through their foundations, which means that wind energy is efficiently transferred into the ground.
The scientists also found that wind-induced ground movement was sometimes strong enough to mask seismic signals arising from slipping faults. “About 6% of the day there are wind-induced ground motions with amplitudes greater than what we’re anticipating for microseismicity,” said Johnson. That’s bad news because there’s a lot that can be learned from these tiny earthquakes. “These tell a lot about the dynamics of fault zones and foreshock and aftershock sequences.”
Shaking Deep Down, Too
The scientists recorded wind-induced ground motion not only in the geophones resting on the ground but also in a 148-meter borehole that had been previously drilled and instrumented with a seismometer.
“We can see that [the wind energy] is going into the ground,” said Johnson. He and his colleagues published their findings last month in the Journal of Geophysical Research: Solid Earth.
These results are important, said Adam Ringler, a geophysicist with the U.S. Geological Survey at the Albuquerque Seismological Laboratory on Kirtland Air Force Base, N.M., not involved in the study. “By improving our understanding of how non-earthquake signals get recorded by sensors, we can improve our ability to detect small events and characterize them.”
In the future, Johnson and his team want to place more geophones near the San Jacinto fault zone. “We’ve been focusing on small areas, but you could imagine putting them out over 20 kilometers,” said Johnson. “They’re quick and easy to deploy.”
—Katherine Kornei, Freelance Science Journalist