Related Experiment Video
Updated: Dec 12, 2025
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
879
Earthquake transformer-an attentive deep-learning model for simultaneous earthquake detection and phase picking
S Mostafa Mousavi1, William L Ellsworth2, Weiqiang Zhu2
1Geophysics Department, Stanford University, 397 Panama Mall, Stanford, CA, 94305-2215, USA. mmousavi@stanford.edu.
Nature Communications
|August 10, 2020
Summary
A new global deep learning model simultaneously detects earthquakes and picks seismic phases, improving accuracy and efficiency. This advanced earthquake detection method identifies more microearthquakes, even with limited seismic station data.
Area of Science:
- Geophysics
- Seismology
- Artificial Intelligence
Background:
- Earthquake signal detection and seismic phase picking are complex, especially with noisy data and microearthquake monitoring.
- Existing methods often struggle with accuracy and efficiency in these challenging scenarios.
Purpose of the Study:
- To develop and evaluate a global deep-learning model for simultaneous earthquake detection and seismic phase picking.
- To enhance the performance of both detection and picking tasks by integrating waveform and phase information.
Main Methods:
- A novel global deep-learning model employing a hierarchical attention mechanism was developed.
- The model performs simultaneous earthquake detection and seismic phase picking, leveraging combined data insights.
Main Results:
- The deep-learning model significantly outperforms previous deep-learning and traditional algorithms for earthquake detection and phase picking.
- Application to the 2000 Tottori earthquake data revealed a twofold increase in detected earthquakes using less than a third of seismic stations.
- P and S phase picking precision approached that of human analysts, with enhanced sensitivity for smaller events.
Conclusions:
- The developed deep-learning model offers a highly efficient and sensitive solution for earthquake detection and phase picking.
- This approach improves the characterization of microearthquakes and enhances seismic monitoring capabilities.
