The power of statistical analysis as a tool for an integrated hazard assessment at Mt. Etna (Italy): how earthquakes and flank eruptions move together on a long-term perspective

Abstract

Flank eruptions at Mt. Etna volcano are often related to the occurrence of strong earthquakes. Nevertheless, this is not always the case, and the relationship has to be modeled in probabilistic terms using Data Science. The topic has been investigated in a few studies, starting in the 80s and the 90s. This statistical analysis produced a clear result - the objective identification of the correlation of the two phenomena. However, from a quantitative point of view, the results were strongly affected by the limitations in the historical records available at that time, and by the statistical approaches adopted. The research continued afterwards, thanks to the basic research projects funded by the Department of Civil Protection in 2000-2015. This provided more complete and extended knowledge of the eruptive and seismic phenomena, and of their relationships.

The implications for a multi-hazard assessment are significant. In particular, the quantitative analyses performed show that the probability of strong (i.e. damaging) earthquakes is influenced by the occurrence of flank eruptions for several weeks, both during the eruptive activity and after the end. The main target of the study is measuring and modeling how much is probable, if compared to the average pattern, to have an earthquake temporally related to a flank eruption, and how long this statistical influence can last for.

In particular, we perform a time series analysis of both the seismic and eruptive records in the last 145-170 years, choosing the longest available seismic record which we assume to be complete. We included the estimation of the seismic rate and investigated the evidence for rate changes in time, as well as the effect of choosing different sub-records as a function of the scale parameters (e.g. earthquake magnitude thresholds, strain release). The key step in the study is the statistical analysis of the inter-event times, either between events of the same type or between earthquakes and eruptions. This produces quantitative estimates of the earthquake rate under the influence of flank eruptions, describing the behavior of the coupled volcano-tectonic system with the final purpose of performing a multi-hazard assessment. This research is currently funded by the Dipartimento della Protezione Civile (Italy), as part of the INGV-DPC contract 2019-2021.