New insights about Bárðarbunga's seismic sources during and after the 2014-2015 caldera collapse events
Abstract
Between September 2014 and February 2015, a fissure eruption occurred in the Holuhraum lava field due to the collapse of the Bárðarbunga caldera in Iceland with a maximum of 70 m at the center (Gudmundsson et al., 2016).
The Seismic moment tensors inversions done for the largest events (Global CMT project) exhibit large non-double-couple components (NDCC). The physical mechanism that underlies the NDCC moment tensors, is under debate. Two of the main hypothesis suggest slip on a ring fault (Nettles & Ekström, 1998) or a closing crack or sill (Kanamori et al., 1993). The large magnitude of the earthquakes (70 earthquakes larger than 5.0) made it possible to observe seismic waves on a high rate GPS station deployed in the middle of the caldera. Additionally, for one of the largest events (a Mw. 5.3 on 18th September 2014) also a fault model was inverted based on the Interferometric synthetic aperture radar (InSAR, Gudmundsson et al., 2016). Currently, the largest earthquakes in Bárðarbunga are up to Mw. 4.8 and the waveforms are very similar to the observed ones in 2014 but with reversed polarities (Jónsdóttir et al., 2017); which suggests that the ring fault in the caldera is slipping upward according to an inflating caldera. Coherent with this finding, we have inverted moment tensors and found reverse focal mechanisms for some of the recent earthquakes, similar to the historical ones reported before the collapse process (1976-1996, Nettles and Ekström, 1998). In this work, we have used seismic and geodetic data (InSAR and high rate GPS) for better constraining the seismic sources during and after the caldera collapse process in Bárðarbunga. As a result, we observed that some of the NDCC moment tensors reported, do not excite considerable amplitudes in the transverse component of the synthetics seismograms, but are observed in the regional data. Conversely, a pure double-couple source is able to reproduce such waveforms, raising the question of why global seismic moment tensor inversions "prefer" NDCC solutions. Another finding comes from using the fault model based on InSAR, for calculating synthetic seismograms at regional distances. The simulations are able to reproduce the observed data fairly well. Since the InSAR inversion did not consider seismic data, these results are an independent confirmation of the InSAR fault model.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.S13A..07R
- Keywords:
-
- 7215 Earthquake source observations;
- SEISMOLOGYDE: 7230 Seismicity and tectonics;
- SEISMOLOGYDE: 8123 Dynamics: seismotectonics;
- TECTONOPHYSICSDE: 8164 Stresses: crust and lithosphere;
- TECTONOPHYSICS