New Seismic Pattern Shows Potential to Predict Earthquakes
Researchers identified ringlike signals that correlated with 12 major events, offering a potential shift in forecasting.
Updated on Oct. 15, 2026 in Geology
Live Poll
Do you believe scientists will eventually be able to provide reliable, short-term earthquake warnings?
Researchers have identified ringlike disturbance patterns, termed seismo-ripples, which may offer a new method for earthquake forecasting. The study, published in NPJ Natural Hazards, details a predictive model that achieved an 83 per cent success rate during tests against 12 major earthquakes.
Why it matters
This research seeks to challenge the long-standing scientific consensus that earthquakes are inherently unpredictable. By shifting the focus from long-term seismic hazard assessment to short-to-medium-term event prediction, the findings could potentially reshape disaster preparedness strategies.
The model demonstrates an 83 per cent success rate for earthquakes occurring within three months of signal detection. However, predictions exhibit an average spatial variance of 280km and magnitude estimates that average 0.5 units higher than final recorded seismic data.
The players
NPJ Natural Hazards
A peer-reviewed scientific journal that focuses on the dynamics and mitigation of natural geological and atmospheric risks.
The details
Researchers identify seismo-ripples by analyzing anomalies in seismic, magnetic, and ionospheric data. They use ground-based Global Navigation Satellite System (GNSS) receivers—devices that track signals from satellite constellations to measure precise crustal positioning—alongside traditional seismometers and magnetometers to map these ringlike patterns. The team cross-referenced these disturbances against historical earthquake records to establish a predictive correlation.
Timeline
2021: Researchers investigated two earthquakes occurring within one day of each other.
April 2022: The team submitted a formal report predicting a 6.3 magnitude event.
June 1, 2022: A 6.1 magnitude earthquake occurred in Sichuan province, China.
2022-2024: The predictive model was verified against 12 major earthquake sites.
October 13, 2026: The research findings were published in the journal NPJ Natural Hazards.
The Tech Race
This work challenges the traditional seismological consensus that earthquakes remain fundamentally unpredictable. It competes with existing long-term statistical hazard mapping by attempting to define a short-to-medium-term operational forecasting methodology.
Current earthquake monitoring and safety protocols remain unchanged by these research-stage findings. The team is now constructing an upgraded ground-based observation system in China to integrate AI and satellite data for improved monitoring.
The takeaway
While the 83 per cent success rate suggests promise, the significant spatial and magnitude variances indicate the model is not yet ready for operational use. Researchers now aim to refine the accuracy of these predictions by incorporating satellite-based observation data and AI-driven analysis.
Further reading
Explore the current state of seismic monitoring and geological research in our Geology section.
Live Poll
Do you believe scientists will eventually be able to provide reliable, short-term earthquake warnings?
