Reevaluating the 2008-2009 Yellowstone Lake Seismic Swarm with Deep Learning

Hayley G. Woodrich¹ , Ross R. Maguire¹, Patricia M. Gregg¹

• Affiliations: ¹Department of Earth Science & Environmental Change, University of Illinois Urbana Champaign, IL, USA

• Presentation type: Poster

• Presentation time: Monday 16:30 - 18:30, Room Poster Hall

• Poster Board Number: 137

• Programme No: 3.1.29

• Theme 3 > Session 1

---

Abstract

The 2008-09 Yellowstone Lake seismic swarm was one of the most energetic in the last several decades with >800 earthquakes recorded between December 26, 2008 and January 8, 2009 (ANSS Comprehensive Earthquake Catalog, ComCat). While the cause of the swarm is uncertain, its location and coinciding extensional surface deformation hints at magma migration. Recent tomographic imaging suggests that the hypocenters occurred within an area of shear wave speed anomalies of up to 25% slow, suggesting that the swarm was located within a warm magma mush zone. In this study, we investigate the spatial and temporal evolution of the 2008-09 Yellowstone Lake swarm, as well as the spatial relationship between swarm activity and magma storage. In particular, we produced an updated seismicity catalog for the Yellowstone region between December 1, 2008, and January 31, 2009, using the QuakeFlow deep-learning algorithm for event detection and location. Compared to ComCat, our catalog increases the number of events during the swarm period from 811 to 1031 and the number of events during the full cataloged period from 1217 to 3248. Hypocenter relocations were completed with HypoDD, a double-difference algorithm that utilizes the offset between observed and calculated travel times. Updated hypocenter locations confirm that the Yellowstone Lake swarm occurred within the seismic slow zone, primarily in a region where wave speeds are 15-20% slow, consistent with a magma mush or partial melt.