High-resolution seismic event detection using local similarity for Large-N arrays

Area of Science:

• Geophysics
• Seismology
• Array Signal Processing

Background:

• Seismic event detection is crucial for understanding Earth's processes.
• Traditional methods struggle with low-amplitude signals and high noise levels in large seismic arrays.
• The need for advanced techniques to monitor micro-seismicity and unusual seismic events is growing.

Purpose of the Study:

• To introduce a novel seismic event detection method applicable to large-N arrays.
• To quantify signal consistency using a new 'local similarity' detection function.
• To demonstrate the method's efficacy in detecting weak seismic events in noisy environments.

Main Methods:

• Development of the local similarity detection function.
• Application to a 5200-station Long Beach nodal array.
• Analysis of continuous seismic data around the 2011 Tohoku-Oki earthquake.

Main Results:

• Stacked local similarity successfully detected seismic events with amplitudes near or below noise levels.
• Analysis in the 5-10 Hz range identified natural and anthropogenic events.
• Low-pass filtering (1 Hz) revealed numerous events, likely Tohoku-Oki aftershocks.

Conclusions:

• The local similarity method offers high-resolution seismic event detection.
• It is effective for monitoring ultra-weak micro-seismicity in noisy conditions.
• The technique is versatile for both ultra-dense and regular seismic array recordings.