Abstract. During the past 1000 years, eruptions of Vesuvius have often been accompanied by large earthquakes in the Apennines 50-60 km to the northeast. Statistical investigations had shown that earthquakes often preceded eruptions, typically by less than a decade, but did not provide a physical explanation for the correlation. Here, we explore elastic stress interaction between earthquakes and eruptions under the hypothesis that small stress changes can promote events when the Apennine normal faults and the Vesuvius magma body are close to failure.We show that earthquakes can promote eruptions by compressing the magma body at depth and opening suitably oriented near-surface conduits. Voiding the magma body in turns brings these same normal faults closer to Coulomb failure, promoting earthquakes. Such a coupling is strongest if the magma reservoir is a dike oriented normal to the regional extension axis, parallel to the Apennines, and the near-surface conduits and fissures are oriented normal to the Apennines. This preferred orientation suggests that the eruptions issuing from such fissures should be most closely linked in time to Apennine earthquakes. Large Apennine earthquakes since 1400 are calculated to have transferred more stress to Vesuvius than all but the largest eruptions have transferred to Apennine faults, which may explain why earthquakes more commonly lead than follow eruptions. A two-way coupling may thus link earthquakes and Vesuvius eruptions along a 100-km-long set of faults. We test the statistical significance of the earthquake-eruption correlation in the two-way coupling zone, and find a correlation significant at the 95% confidence level.

[1] In this study we aim to understand the effect of static and dynamic stress changes in promoting earthquake failures on secondary faults. Toward this goal we solve the equation of motion of a spring-slider dynamic system including inertia and using rate-and state-dependent constitutive laws. We separately investigate the dynamic response of this fault analog system to a sudden stress change represented either as a stress step or as a stress pulse, which are used to model permanent (static) and transient (dynamic) stress perturbations. The induced earthquake failure does not occur immediately at the application of the coseismic stress change, but it is delayed in time: we define this time interval as the triggering delay. For a given stress perturbation, we analyze the dependence of triggering delays on different system conditions and constitutive parameters. Our results clearly show that the effects of static and dynamic stress changes are quite different. While a static stress change is able to advance as well as to delay an induced instability depending on its sign, a dynamic stress pulse is only able to promote a nearly instantaneous failure, provided its amplitude is positive and large enough with respect to the direct effect of friction. In other words, dynamic stress changes can only cause nearly instantaneous failures, without any relevant triggering delay. These results should be considered in interpreting the seismicity rate changes caused by large earthquakes at least as long as seismic events are interpreted as sliding instabilities obeying rate-and state-dependent friction laws.INDEX TERMS: 7209 Seismology: Earthquake dynamics and mechanics; 7260 Seismology: Theory and modeling; 7299 Seismology: General or miscellaneous; KEYWORDS: stress trigger, fault interaction, rate-and state-dependent friction, spring-slider Citation: Belardinelli, M. E., A. Bizzarri, and M. Cocco, Earthquake triggering by static and dynamic stress changes,

A critical point in the analysis of ground displacement time series, as those recorded by space geodetic techniques, is the development of data-driven methods that allow the different sources of deformation to be discerned and characterized in the space and time domains. Multivariate statistic includes several approaches that can be considered as a part of data-driven methods. A widely used technique is the principal component analysis (PCA), which allows us to reduce the dimensionality of the data space while maintaining most of the variance of the dataset explained. However, PCA does not perform well in finding the solution to the so-called blind source separation (BSS) problem, i.e., in recovering and separating the original sources that generate the observed data. This is mainly due to the fact that PCA minimizes the misfit calculated using an L 2 norm (χ 2 ), looking for a new Euclidean space where the projected data are uncorrelated. The independent component analysis (ICA) is a popular technique adopted to approach the BSS problem. However, the independence condition is not easy to impose, and it is often necessary to introduce some approximations. To work around this problem, we test the use of a modified variational Bayesian ICA (vbICA) method to recover the multiple sources of ground deformation even in the presence Electronic supplementary material The online version of this article

Analyzing the displacement time series from continuous GPS (cGPS) with an Independent Component Analysis, we detect a transient deformation signal that correlates both in space and time with a seismic swarm activity (maximum Mw=3.69 ± 0.09) occurred in the hanging wall of the Altotiberina normal fault (Northern Apennines, Italy) in 2013–2014. The geodetic transient lasted ∼6 months and produced a NW‐SE trending extension of ∼5.3 mm, consistent with the regional tectonic regime. The seismicity and the geodetic signal are consistent with slip on two splay faults in the Altotiberina fault (ATF) hanging wall. Comparing the seismic moment associated with the geodetic transient and the seismic events, we observe that seismicity accounts for only a fraction of the measured geodetic deformation. The combined seismic and aseismic slip decreased the Coulomb stress on the locked shallow portion of the ATF, while the transition region to the creeping section has been loaded.