Estimates and models of the statistical dependence between local earthquakes and flank eruptions at Mt. Etna volcano (Italy): an old topic revised through new (historical) data

Abstract

Our study focuses on the estimation of the correlation between flank eruptions at Mt. Etna volcano and the pre-eruptive, syn-eruptive, and post-eruptive local seismicity, with the final purpose of improving the quantitative multi-hazard assessments. This research was motivated by the lateral eruption started at Etna on Dec. 24th 2018, that was followed by an intense seismic swarm producing a destructive earthquake (Mw 4.9, Dec. 26th) along the nearby Fiandaca fault. Indeed, earthquakes and eruptions at Etna are well known to be often occurring close in time, and a full probabilistic assessment would be very useful for civil protection purposes. The study relies on a few investigations developed in the 80s and 90s, which tested various statistical methods on the datasets available at that time, and the more recent research projects which provided more complete and extended databases on the eruptive and seismic phenomena. In particular, first we performed a time series analysis of both the seismic and eruptive records in the last 130-150 years, which are not affected by significant underrecording issues. We included the estimation of event rate and rate changes in time, and the description of spatial location and scale parameters (e.g. earthquake magnitude, erupted volume). Then, we accomplished a statistical analysis of the inter-event times, either between events of the same type, or between earthquakes and eruptions. Finally, we implemented a series of statistical tests aimed at the quantitative estimation of the correlation strength and its duration. The analyses were then conditioned on specific magnitude thresholds or spatial constraints (epicenters’ distribution vs eruptive fissure/vent location), with the purpose of further exploring the behavior of the coupled volcano-tectonic system.