2:30 PM - 2:45 PM
[SSS32-04] Tectonic and volcanic deformation at the Azores Triple Junction, observed by continuous and campaign GPS analysis
Keywords:crustal deformation, GPS, GNSS, plate tectonics, Azores, volcano geodesy
We analyze 9 continuous GPS (CGPS) stations and the campaign data of the island for the period of 2008-2013 using Bernese5.0 software (Dach et al. 2007). In order to tie the estimated coordinates to the global geodetic reference frame - ITRF2005, neighboring international IGS station data are simultaneously processed along with the local datasets. By comparing with the current plate angular velocities (DeMets et al. 2010), we find a high-strain-rate (0.28 ppm/yr of expansion) zone in the east of Fogo volcano, which accommodates about 50% of the Eurasian–Nubian plate spreading.
Fogo exhibited intense seismic swarm during 2011-2012. The analysis of detrended GPS time-series after subtracting regional plate velocities reveals the existence of two different types of ground deformation associated with the seismicity. One is the edifice-scale inflation of Fogo, which corresponds to the increase in volcano-tectonic events. Another is inflation–deflation reversal in the east of Fogo, which coincides with the sharp decrease in lower-frequency events in August 2012. A strong similarity to the Matsushiro, Japan, earthquake swarm (1965–66) and Campi Flegrei, Italy, volcanic episodes (1969–72 and 1982–85) may suggest importance of the hydrothermal system at Fogo volcano. We propose the following hypothesis for the Fogo unrest: (1) the primary inflation source beneath Fogo promotes lateral diffusion of fluids that is selectively guided by existing cracks/fissures formed from regional extension; (2) an influx of fluids increases pressure in cracks/fissures and generates lower-frequency earthquakes; and (3) discharge of fluids causes pressure decrease and dilatancy recovery (i.e. seismic quiescence). To estimate the source parameters, the result of GPS campaigns is modelled by an integrated inversion using a genetic algorithm. The best fit model agrees well with the regional/local tectonic feature.