Earthquake sensors can now help identify planes flying overhead

Thunder from an earthquake and the roar of a jet feel worlds apart. Yet deep under your feet, both can leave nearly the same kind of trace. New research from the University of Alaska Fairbanks shows that earthquake instruments can not only hear planes flying overhead, they can help identify what kind of aircraft is in the sky.

Graduate student researcher Bella Seppi has shown that sensitive seismometers pick up tiny vibrations caused by aircraft noise as it shakes the ground. Those faint signals carry a clear acoustic fingerprint that can be used to tell a small piston plane from a turboprop or a jet.

“Aircraft signals are a lot higher frequency than anything else that’s prominent in the spectrum that seismometers are recording,” Seppi explains. “Earthquake signals and other signals that people are typically looking for are a lot lower frequency, so aircraft signals are pretty obvious most of the time.”

Her method, described in The Seismic Record, turns the ground itself into a listening device for the sky.

Seismometers are designed to record ground motion from earthquakes, landslides and other geologic events. They also register vibrations from sonic pressure waves, which are simply sound waves that reach the ground and cause it to move by tiny amounts.

If you think of an ambulance siren that rises in pitch as it approaches and drops as it moves away, you already know the Doppler effect. The same physics applies to planes. As an aircraft flies toward a seismometer, the pitch of its sound increases. As it flies away, the pitch falls.

By transforming raw seismic data into a spectrogram that shows frequency over time, Seppi can see that Doppler shift curve. Higher frequencies mark the approach, lower frequencies mark the retreat, and the moment of closest approach captures the aircraft’s true pitch.

For this study, she used nearly 1,200 recordings from 303 seismometers deployed along the Parks Highway between Nenana and Talkeetna. The instruments were originally installed to monitor aftershocks of the 2018 magnitude 7.1 Anchorage earthquake and to image subsurface structure. Because they sampled data 500 times per second, much faster than many permanent stations, they could capture the higher frequencies generated by aircraft noise.

Alaska’s broader seismic network would need similar high sampling rates before it could routinely identify aircraft types this way.

Recognizing that an aircraft flew overhead is only the first step. Seppi wanted to know whether she could determine exactly what type of plane had produced the signal.

To do that, she had to remove the Doppler effect and recover each plane’s base frequency, the pitch it would have if it hovered in place. Once that baseline is known, the recording reveals a full “frequency comb” composed of the base tone and its harmonics, the repeating pattern that gives each object its unique sound.

Most vibrating systems, from guitars to jet engines, produce a fundamental tone plus harmonics. The pattern is rarely perfectly smooth, but it is stable enough to recognize.

No catalog of aircraft frequency combs existed, so Seppi had to build one. She turned to Flightradar24, a public flight tracking site that lists each plane’s type, position, altitude, speed and route. By matching flight times and locations from that site with the times when seismic stations recorded aircraft signals, she could link each spectrogram curve to a specific plane.

Once those matches were made, she mathematically removed the Doppler shift to recover the true frequency comb for each aircraft. She then grouped those combs by engine type, including piston aircraft, turboprops and jets.

“What surprised me the most is how consistent a lot of the frequency signals are,” she says. That consistency is what makes a catalog useful.

With this technique, any seismic recording that contains an aircraft signal can be converted into a frequency comb. In the future, that comb could be compared against a catalog of known patterns to identify the aircraft type passing overhead.

The spectrogram curves themselves also hold more information than just pitch. Their shape reflects the plane’s speed, direction and distance from the seismometer. By analyzing how quickly the frequencies rise and fall, Seppi can estimate how fast the aircraft is moving and when it passes closest to the station.

Additional work will explore how far away an aircraft can be and still be detected, and how combining data from multiple seismometers might reveal full flight paths. Using an array, you could potentially infer heading, altitude trends and other details, much like a passive, ground based radar.

The current study focused on a short stretch of highway in winter, but the method should work anywhere that high sample rate seismic data and air traffic information both exist.

This approach opens new possibilities for both seismology and aviation monitoring. For earthquake scientists, being able to recognize aircraft signals means they can better filter those high frequency vibrations out of their data and see subtle tectonic signals more clearly.

Seppi also points to environmental uses. The same method could help estimate how often certain aircraft types fly over sensitive wildlife habitats or quiet protected lands. Instead of installing microphones on fragile terrain, researchers could read the sky through buried seismometers and predict noise impacts.

“This new method has many uses,” she says, including helping regulators and communities understand sound exposure from different planes.

There are practical limits. High sample rate instruments are more expensive. Atmospheric conditions such as wind and temperature can change how sound travels from air to ground. A truly global catalog of frequency combs will take time to build and maintain. Still, this first catalog shows the idea is realistic.

The work hints at wider security and safety applications as well. Passive ground sensors that need no active signal could supplement radar in remote regions or help verify air traffic patterns without broadcasting their presence.

At its core, though, the study is a story about curiosity and rethinking tools. Sensors planted in the frozen ground to listen for aftershocks turned out to be listening to something else as well, the distant rumble of aircraft high above Alaska’s winter sky.

This research gives scientists and decision makers a low cost, passive way to monitor aircraft activity using instruments already in the ground. For seismologists, it offers a better way to recognize and remove aircraft noise from earthquake data, which can sharpen early warning systems and improve hazard maps.

For environmental planners and wildlife managers, the method could help map which aircraft types pass over parks, refuges and quiet rural areas, and how often. That kind of information can guide flight corridors that reduce noise stress on people and animals.

For aviation regulators and security agencies, a mature catalog could provide a backup way to track planes in regions with limited radar or where transponders are unreliable. Because the approach does not transmit signals, it can operate quietly in the background.

The study also encourages more creative uses of existing networks, which can save money and speed up innovation. As researchers expand the aircraft frequency catalog and upgrade more stations to higher sampling rates, seismometers may become a standard tool for understanding both the restless ground beneath you and the busy airspace above.

Research findings are available online in the journal The Seismic Record.

Like these kind of feel good stories? Get The Brighter Side of News' newsletter.