Evolution of temporal seismic velocity changes and earthquake source mechanisms during the 2021 Fagradalsfall dyke intrusion

Abstract

Late February 2021, the Reykjanes Peninsula in southwest Iceland experienced severe seismicity associated with the development of a 9 km long dyke. Eight earthquakes of magnitude M≥5  were registered in the vicinity of Fagradalsfjall from February 24 until the onset of the Fagradalsfjall eruption in mid-March, which lasted for six months. Here, we analyze the temporal variations in crustal seismic wave velocities and the source characteristics of earthquakes during the dyke formation phase (February-March 2021).We apply ambient-noise seismic interferometry and compute seismic noise cross-correlations using the MSNoise software. Cross-wavelet analysis, a powerful technique that allows us to obtain frequency-dependence of velocity change, is used to investigate relative variations in seismic wave velocities (dv/v). Along with our wavelet-based dv/v results, we also present the stretching-based dv/v time-series that were calculated in real-time for volcano monitoring during the unrest. The Fagradalsfjall dyke intrusion induced temporal variations in seismic velocities and strong decorrelation that were picked up by the entire network across the peninsula. Beginning abruptly with the increased seismic activity, velocities at nearby seismic stations decreased by 1.5 percent. The amount of dv/v change was noticeably less than 1 percent at distant stations (15-30 km). The regional time-domain moment tensor inversion method (TDMT_INVC) was also applied to obtain earthquake mechanism solutions. Source parameters of 50 moderate-sized events with magnitudes Mw≥4.0 revealed predominantly normal and strike-slip faulting. We compare these to the deformation, dv/v and modeled Coulomb stress changes and present a joint interpretation.We provide a summary of the complex spatial and temporal evolution of crustal seismic velocity changes in the weeks preceding the effusive eruption. The understanding of the pre-eruptive geophysical signatures of the Fagradalsfjall volcano will contribute to better predict future volcanic activity in the area.