Clustering and classification of infrasonic events at Mount Etna using pattern recognition techniques

A novel pattern recognition system using infrasonic waveform features successfully clustered and classified explosive volcanic events at Mount Etna, enabling automatic vent identification without traditional location algorithms. The study extracted key features—peak frequency, quality factor, and peak-to-peak amplitude—to create a 3D 'feature space' where three distinct clusters were identified using the DBSCAN algorithm. These clusters correlated with specific source vents or volcanic activity types, validated by semblance location methods and volcanological data. A Support Vector Machine model trained on these clusters achieved accurate event classification, demonstrating potential for real-time monitoring advancements in volcanic hazard assessment.

Cannata, A., Montalto, P., Aliotta, M., Cassisi, C., Pulvirenti, A., Privitera, E., & Patanè, D. (2011). Clustering and classification of infrasonic events at Mount Etna using pattern recognition techniques. Geophysical Journal International, 185(1), 253–264. https://doi.org/10.1111/j.1365-246x.2011.04951.x