2009 Portland GSA Annual Meeting (18-21 October 2009)

Paper No. 5
Presentation Time: 9:00 AM

SEISMO-ACOUSTIC ARRAY OBSERVATIONS OF SHALLOW CONDUIT PROCESSES AT FUEGO VOLCANO, GUATEMALA


WAITE, Gregory P., Geological and Mining Engineering and Sciences, Michigan Technological Univ, 1400 Townsend Dr, Houghton, MI 49931 and LYONS, John J., Geological and Mining Engineering and Sciences, Michigan Technological University, 1400 Townsend Dr, Houghton, MI 49931, gpwaite@mtu.edu

Fuego volcano, Guatemala, is a basaltic stratovolcano currently characterized by low-level strombolian eruptive activity. In January 2008, we deployed small antennas of six broadband seismic and five acoustic sensors 0.9 km north of the active vent to investigate the source of explosions and low-frequency seismicity. The seismic array had stations spaced 30 m apart and a total aperture of ~140 m. The infrasound sensors were deployed in a similar array, but with average station spacing of 50 m. There was no lava effusion during the deployment, but explosions were recorded approximately once per hour, with varied amounts of ash, and with durations from 20-150 s. In addition to the explosions, our seismic array recorded tremor with a dominant frequency of 1.9 Hz and discrete events that were not generally detected by the acoustic array. The dominant class of these events, which repeated approximately 10-15 times per hour, had an impulsive onset with first motion toward the vent, a short duration of <5 s, and dominant frequencies from 1-3 Hz. Their similarity suggests a nondestructive source process. While waveforms are similar from event to event when viewed on the same channel, the large variation in waveforms across the array yields a large uncertainty in slowness parameter estimates. We take advantage of the high degree of similarity between events to determine relative slowness estimates. After determining the best-fit slowness parameters for a master event, we measured the relative slowness parameters for 203 similar events. The results indicate a stationary source. The low-frequency events were then located by comparing them with slowness parameters computed from synthetic waveforms for a volume of sources beneath the summit region. The source is ~150 m directly beneath the active vent, but not associated with explosions. Full moment tensor source inversion is not possible due to the limited array geometries, but we use forward modeling of candidate source geometries to infer differences between the sources of the dominant seismic signals. The events are consistent with a near-vertical north-south oriented crack. Despite the stability of the source during our observation period in 2008, these events did not occur during a 2009 deployment during similar, albeit more energetic eruptive activity.