by Maochuan Zhang ← Back to Axial Seamount Earthquake Catalog
This page provides near-real-time focal mechanism estimates for earthquakes at Axial Seamount recorded by the Ocean Observatories Initiative (OOI) cabled array. P-wave first-motion polarities are predicted using a deep learning model — a convolutional neural network trained on manually verified polarities from the Axial Seamount catalog. For each event, the pipeline identifies the 6 most similar historical earthquakes from a base catalog of events with known focal mechanisms, using a composite similarity metric that combines hypocenter location, deep-learning-predicted polarities, and S/P amplitude ratios across seven OOI broadband stations (AS1, AS2, CC1, EC1, EC2, EC3, ID1). Focal mechanisms are then computed using the HASH algorithm applied to the composite polarity dataset. Color coding: blue = Normal, red = Reverse, green = Strike-slip, black = Unclassified. A manuscript describing this method is in preparation.
Maps are updated hourly. Last updated: 12-Sep-2026 11:32:50 UTC
Event selection criteria: Real-time focal mechanisms are computed only for events with a successful hypocenter solution and at least 5 P picks and 5 S picks.
| 2025: | 01 | | | 02 | | | 03 | | | 04 | | | 05 | | | 06 | | | 07 | | | 08 | | | 09 | | | 10 | | | 11 | | | 12 |
| 2026: | 01 | | | 02 | | | 03 | | | 04 | | | 05 | | | 06 | | | 07 | | | 08 | | | 09 | | | 10 | | | 11 | | | 12 |
Downloads:
ML real-time FM catalog, Jan 2022 – present (.txt)
ML real-time FM catalog, Jan 2022 – present (.csv)
ML-derived polarity FM catalog, Jan 2022 – Mar 2025 (.txt)
Cross-correlation-derived polarity FM catalog, Jan 2015 – Dec 2021 (.txt) ← Back to Axial Seamount Earthquake Catalog