Thunder helps fiber optics map the subsurface
DAS captured 458 events along 4 km of cable under Penn State and mapped velocities to 100 m. The work is a single-campus proof of concept.
A study published on August 21 used thunder recorded along 4 kilometers of optical fiber beneath Penn State’s campus to build an image of the shallow subsurface. The researchers analyzed roughly two and a half years of data, manually selected 458 events, and estimated shear-wave velocities down to about 100 meters. The result is a proof of concept for one campus on karst terrain, with performance elsewhere untested.¹
The measurement comes from distributed acoustic sensing, or DAS. An instrument called an interrogator sends laser pulses through a fiber and measures changes in the phase of backscattered light. Tiny strains caused by waves moving through the ground alter that signal. A preexisting telecommunications cable can provide densely spaced measurement points while it is connected to an interrogator and acquisition system; buried fiber does not record a storm by itself.¹ ²
Penn State’s FORESEE array comprised 2,137 channels at 2-meter intervals, with continuous data collected from 2019 through 2021. Earlier work on the same research line had characterized thunder-induced ground motion across a fiber-optic DAS array. The 2026 paper moves from detecting that phenomenon to producing shallow tomography.³
Thunder transfers acoustic energy into the ground
A lightning discharge heats the air rapidly and produces shock waves that travel as thunder. When they reach the surface, some of that acoustic energy couples into the ground. The authors call the resulting seismic motion a thunderquake.
The signal contains Rayleigh waves that travel near the surface. They are dispersive, meaning different frequencies move at different speeds depending on the structure and mechanical properties of shallow layers. That relationship can be inverted into a velocity model without opening the ground at every location.¹
The path from discharge to fiber is complicated. Lightning may create acoustic sources at different altitudes, and sound reaches the surface along different routes. Buildings and topographic or geological variation scatter the energy. The team identified 458 high-fidelity signals manually, then checked their timing and peak current against National Lightning Detection Network records.¹
After filtering noise and DAS artifacts, the researchers used cross-correlation and stacked information from many events. The processing reduced the sound’s travel history through the air and brought out coherent dispersion along the cable. Dispersion curves then fed inversions of shear-wave velocity, a property sensitive to the stiffness of underground material.
The event count does not represent 458 independent measurements of the entire structure. Signals were combined and segmented into profiles along the same installation. Manual selection also limits continuous use: another deployment would need to separate usable thunderquakes from noise and unfavorable recording conditions.
Tomography revealed four low-velocity zones
The final section contains four low-velocity regions, labeled WZ-1 through WZ-4. In the campus’s karst geology, such a signal can correspond to sediment or other weak material, weathered rock, fractures, or fluids. Velocity alone cannot choose among those explanations, and the result does not confirm a cavity or sinkhole.¹
The authors compared their profiles with engineering multichannel analysis of surface waves, or MASW, and borehole logs. Velocities and the depths of major stratigraphic changes were broadly consistent at nearby locations, even though some comparisons were separated by approximately 50 meters. An earlier ambient-noise survey also found WZ-4 and locally low velocities near WZ-2.¹
Satellite interferometric radar data provided a different check. From 2017 through 2025, areas near WZ-1 and WZ-2 showed line-of-sight subsidence. WZ-3 and WZ-4 lacked that signal. The overlap supports an interpretation involving fractured or altered material at the first two locations, without establishing the cause of deformation or demonstrating an active collapse hazard.¹
Resolution depends on recorded frequencies and choices in the inversion. The paper notes that extremely shallow sediment with shear-wave speeds below the acoustic wave speed in air cannot be resolved by the mechanism used here.
The 3D simulation is a semiquantitative check
A three-dimensional simulation tested whether atmospheric sound could generate the dispersive waves found in the observations. Computing limits held the model to a dominant frequency of about 4 hertz, while the observed mode near 20 hertz was most sensitive to the first 15 to 20 meters below the surface.¹
To bridge that mismatch, the authors used depth stretching: the real 0-to-20-meter interval was represented as 0 to 200 meters in the model. They describe the result as semiquantitative validation of air-to-ground coupling. It is not a quantitative prediction of the observed high-frequency amplitudes or every dispersion curve.¹
The study did not evaluate networks beyond the campus. Urban networks vary in cable construction, installation, access, noise, and geology. Reusing infrastructure still requires permission to connect an interrogator, suitable physical coupling, data storage, and processing. The supply of useful storms also changes by region and season.
FORESEE’s raw waveforms are available through Penn State Data Commons, while the paper identifies the software and supplementary materials used in processing.⁴ Those records make the published analysis testable. A deployment elsewhere would still need comparisons with boreholes, engineering surveys, or other independent observations before low-velocity zones informed decisions about the ground.
Sources
- Imaging Earth’s subsurface with thunderstorm-generated seismic waves · Science Advances · https://doi.org/10.1126/sciadv.aeg8096 · Aug. 21, 2026
- Thunderquakes: A new way to image the Earth’s subsurface · Penn State · https://www.psu.edu/news/earth-and-mineral-sciences/story/thunderquakes-new-way-image-earths-subsurface · Aug. 21, 2026
Show 2 more sourcesHide sources
- Characterizing Thunder-Induced Ground Motions Using Fiber-Optic Distributed Acoustic Sensing Array · Journal of Geophysical Research: Atmospheres · https://doi.org/10.1029/2019JD031453 · 2019
- Sensing Earth and environment dynamics by telecommunication fiber-optic sensors: an urban experiment in Pennsylvania, USA · Penn State Data Commons · Dataset 6290 · https://www.datacommons.psu.edu/commonswizard/MetadataDisplay.aspx?Dataset=6290 · accessed Aug. 26, 2026