Thunder + fiber-optic cabling used for seismic imaging
Penn State researchers developed a seismic modeling approach using thunderquakes (seismic signals from thunderstorms) to map subsurface terrain The model, built on SPECFEM3D Cartesian software, overcomes the extreme complexity of thunderquake signals through strategic approximations Two years of data from 458 thunderquakes detected via a 4km fiber-optic cable on campus successfully identified four subsurface weak zones The technique proves viable for near-surface seismic imaging in thunderstorm-
Analysis
TL;DR
- Penn State researchers developed a seismic modeling approach using thunderquakes (seismic signals from thunderstorms) to map subsurface terrain
- The model, built on SPECFEM3D Cartesian software, overcomes the extreme complexity of thunderquake signals through strategic approximations
- Two years of data from 458 thunderquakes detected via a 4km fiber-optic cable on campus successfully identified four subsurface weak zones
- The technique proves viable for near-surface seismic imaging in thunderstorm-prone regions, offering a free, frequent alternative to earthquakes or controlled explosions
- Weak zones corresponded to karst formations with sediments, fractured rock, and high water content, later confirmed by radar, boreholes, and independent seismic surveys
Why It Matters
This research opens a novel, cost-effective pathway for near-surface seismic imaging by repurposing naturally occurring thunderstorms as seismic sources, eliminating the need for expensive controlled explosions or reliance on unpredictable earthquakes. For AI and geophysics practitioners, it demonstrates how machine-readable infrastructure (fiber-optic cables) combined with sophisticated modeling can extract valuable subsurface data from previously dismissed "noisy" signals.
Technical Details
- Modeling framework: Built on SPECFEM3D Cartesian, a 3D seismic wave reconstruction software, with key approximations including treating the atmosphere as a 3.6 km homogeneous layer and stretching the top 20 meters of Earth to 200 meters to compensate for slower update frequencies
- Data source: 458 well-resolved thunderquakes captured over two years using a 4 km fiber-optic cable repurposed as distributed acoustic sensing (DAS) seismometers on the Penn State campus
- Signal characteristics: Thunderquakes produce multiple signals arriving from different altitudes (from the "string of beads" plasma structure along lightning paths), generating high-energy impulsive wavelets followed by decaying surface-wave trains lasting 1-2 seconds
- Validation: Four identified weak zones were confirmed through ground-penetrating radar (surface deformation), engineering surveys, borehole measurements, and independent seismic data
- Target geology: The campus sits on a karst formation where water-altered limestone bedrock creates variable subsurface rigidity, making it an ideal test case for detecting weak zones
Industry Insight
- Distributed Acoustic Sensing (DAS) expansion: The successful use of existing fiber-optic infrastructure as seismometers signals a broader trend—telecom networks can double as passive seismic monitoring arrays, dramatically reducing instrumentation costs for geophysical surveys.
- Natural-source seismology as a viable alternative: Thunderquakes, along with other ambient seismic sources (ocean waves, wind, traffic), represent an untapped resource for near-surface imaging. Regions with frequent thunderstorms now have a free, recurring seismic source for mapping infrastructure-relevant subsurface features.
- Modeling approximations can yield practical results: The researchers acknowledged significant simplifications in their model yet achieved confirmed, actionable results. This validates a pragmatic approach in geophysical AI—high-fidelity models aren't always necessary when approximations can be empirically validated against ground truth data.
Disclaimer: The above content is generated by AI and is for reference only.