GSC Logan Club Series: Dr. Rebecca Harrington

"What can low-frequency volcanic earthquakes tell us about how a volcano will erupt?".

Date: Thursday, March 20, 2014

Time: 11:30 a.m.

Location:127 Marion Hall (MRN), University of Ottawa

Volcanoes generate a variety of pre-eruptive low-frequency seismic signals.  Volcanic hybrid earthquakes comprise a class of such low-frequency signals that start with high-frequency onsets which transition into low-frequency signals with extended duration relative to event size.  Hybrids are used empirically to predict eruptions, but their ambiguous physical origin limits their diagnostic use.  Complex models involving elastic failure, fluid flow, fluid shear and their interactions are often used to explain their unusual waveforms.  Here we examine the size-duration scaling of volcanic hybrid earthquakes simulated in the laboratory during rock deformation experiments, and occurring at Mount St. Helens volcano during the 2004-2008 eruption.  We find that the size-duration scaling resembles that of similarly sized tectonic earthquakes observed in other shallow fault zones, suggesting that the low-frequency portion of the hybrid waveform can be explained by path effects. 

Both laboratory conditions and physical fault features at Mount St. Helens coupled with the similarity of the size-duration scaling with that of tectonic earthquakes suggest that the mechanical failure processes of hybrids are consistent with that of typical earthquakes, i.e., elastic brittle failure.  However, lower than expected corner frequency values relative to event size suggest that the rupture velocities may be low compared to tectonic earthquakes (< 2.5-3 km/s).  The low corner frequencies likely result from low shear wave velocities in the volcanic edifice (assuming that the rupture velocity is 0.9*shear-wave velocity), rather than low static stress drop values. The combined laboratory and field observations are the first seismological evidence for brittle-failure as a major process in volcanic dome building, and suggest that hybrids should not be used as direct indicators of fluids.