US EditionPublished • loading... • Updated
AI Uncovers 86,000 Hidden Earthquakes Under Yellowstone
Machine learning analysis expanded Yellowstone's earthquake catalog by 10 times, detecting over 86,000 small events from 2008 to 2022, improving seismic monitoring and understanding.
- Bing Li, Western engineering professor, and collaborators used machine learning to re-examine Yellowstone caldera records, detecting roughly 10 times more earthquakes, study published July 18 in Science Advances.
- Machine learning has sparked a data‑mining rush that revisits historical waveform data stored in datacenters, replacing slow, costly manual inspection by trained experts in recent years.
- Researchers found that more than half of events were part of earthquake swarms on immature, rougher fault structures characterized by fractal-based models linked to underground water and fluid bursts.
- The enlarged dataset lets researchers apply statistical methods to detect new swarms, improving safety measures, public information, and guidance for geothermal energy developers in Yellowstone's seismically active network.
- The authors suggest similar seismic-driven processes may operate across Earth and on Mars, and a related study in PNAS Nexus on November 25, 2025, documented changes after Yellowstone earthquake swarms.
Insights by Ground AI
17 Articles
17 Articles
AI uncovers 86,000 hidden earthquakes beneath Yellowstone’s surface
Beneath Yellowstone’s stunning surface lies a hyperactive seismic world, now better understood thanks to machine learning. Researchers have uncovered over 86,000 earthquakes—10 times more than previously known—revealing chaotic swarms moving along rough, young fault lines. With these new insights, we’re getting closer to decoding Earth’s volcanic heartbeat and improving how we predict and manage volcanic and geothermal hazards.
·United States
Read Full Article
AI finds signs of life in ancient rocks
A specially trained algorithm could aid the search for biological activity both on the early Earth and on other worlds. A specially trained algorithm could aid the search for biological activity both on the early Earth and on other worlds.
·United Kingdom
Read Full Article
Coverage Details
Total News Sources17
Leaning Left1Leaning Right0Center5Last UpdatedBias Distribution83% Center
Bias Distribution
- 83% of the sources are Center
83% Center
L 17%
C 83%
Factuality
To view factuality data please Upgrade to Premium
