Crustal Response to Changes in the Magmatic System at the Soufrière Hills Volcano, Montserrat
Open Access
- Graduate Program:
- Geosciences
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- October 26, 2011
- Committee Members:
- Barry Voight, Dissertation Advisor
- Barry Voight, Committee Chair
- Charles James Ammon, Committee Chair
- Sridhar Anandakrishnan, Committee Member
- Derek Elsworth, Committee Member
- Keywords:
- locations
- earthquake
- volcano-tectonic
- Montserrat
- volcano
- Soufrière Hills
- focal mechanism
- stress
- VT-string
- Abstract:
- This dissertation addresses the crustal response to the magmatic system beneath the Soufrière Hills volcano (SHV), Montserrat, using various seismic techniques. To facilitate a more robust interpretation of these results and other studies at similar crustal environments, we have also analyzed the seismic techniques typically utilized in such settings. In the first article of this dissertation I explored volcano-tectonic (VT) earthquake activity recorded during the onset of volcanic activity, from July 1995 to October 1996, at SHV. I have determined earthquake locations for 6605 events and focal mechanism solutions for select events to reveal the stress axes. I identified several zones of temporally-confined seismic activity that comprise assemblages of small scale structural elements, >2-4 km distant from SHV, at depths 2-4 km b.s.l. I suggest that the clustered seismicity and relatively aseismic zones reflect a broad weakened tectonic zone of ESE trend that crosses Montserrat, which is overprinted by the ascent of a magmatic dike of NNE trend. This event, which altered the stress distribution, promoted localized fault movements and dilatation, with resultant changes in pore-fluid pressures. Ultimately, depending on the local polarity of strain these events may have weakened or strengthened the rock mass. In the second article I have explored the limitations of focal mechanism solutions and identified methods to fortify their robustness by using three subsets of focal mechanisms with the best-quality determinations. One subset contains only the largest magnitude events. Two additional subsets, which were chosen for their tight temporal and spatial occurrence are part of the distal earthquake swarms located NE of the summit (NE) and at St. Georges Hill (SGH). I assume these two subsets were generated under consistent stress conditions. I propose that inconsistencies in the determined mechanisms for each subset may be due to uncertainties in the depth estimates, or real differences in stress due to crustal variations. I have used two approaches to determine the compression (P), intermediate (B), and tension (T) stress axes for the solutions at a range of hypocentral depths. These methods are successful in amplifying the signals of the dominating stress axis orientations. The dominating stress in the two focused clusters is that of extension; the NE subset has a T-axis orientated NW-SE, whereas the SGH subset has a T-axis oriented NE-SW. The P- and B- axes for both subsets are more variable. On the basis of these results we suggest that a local stress regime (generated by magmatic pressurization) is overprinting the regional stress regime. Finally, I examine the significance of a seismic pattern recently identified on Montserrat (termed VT-strings). I compare VT-string earthquake characteristics to other VT events to derive how they relate to the evolution of the volcanic system. In late 2007, the first observation of a VT-string sequence was made (7 November) consisting of a short-lived intense swarm of VT events. Since this initial sequence, several other similar swarms have been identified and have been correlated to the onset of eruptive activity. To determine how these unusual VT earthquakes relate to magmatic processes at SHV, I have examined the 7 November swarm for collective properties and distinguishing characteristics, when compared to other VT earthquakes recorded during this time period. I have calculated earthquake locations and focal mechanisms, and have explored the event waveforms for their spectral characteristics and similarity among events. The 7 November swarm events were closely related in time (all within a 1 hour period) and space (~2 km b.s.l.); but lack a distinct pattern in waveform characteristics. Conversely, other earthquakes (mostly occurring on 16 December 2007), are also closely related in space (~3 km b.s.l.), but correlate highly with events that occurred one or two days later. The VT-string sequence earthquake mechanisms exhibit no dominant orientation of maximum stress, whereas the later clusters highlight a vertical T-axis and a NW-SE trending P-axis. The repeating earthquakes are interpreted as occurring on pre-existing fractures at the top of the well-established conduit from depth, whereas the 7 November (not-correlating) events represent new fractures produced by the upward migration of magma into previously unstressed host rock or fatigue fracturing from repeated periods of pressurization. Together, the articles in this dissertation increase our understanding of the crustal-magmatic system interactions at SHV. Exploring the processes at the onset of activity and later in the eruption, once the system has evolved.