GNGTS 2018 - 37° Convegno Nazionale

58 GNGTS 2018 S essione 1.1 Conclusions. We use data from station IOCA to improve the epicentral location and depth of the August 21, 2017 Ischia earthquake and discuss the focal mechanisms, as a clear positive first motion onset is apparently in contrast with moment tensor solutions. P-phase particle motion (Fig. 1), the evaluation of the rotated spectra, and S-minus-P-wave travel time lead to a hypocentral depth of 2 km, in the same epicentral area as the devastating seismic event that struck Ischia in 1883. Our best DC and MT solutions, as well as the other proposed solutions (Fig. 2), predict a negative onset, in contrast to observation. We could reproduce the positive onset for our best DC model and a shallower source depth (1 km or less). The 36% and 26% of negative isotropic component and negative CLVD components of full MT solution do not represent a pure closing tensile crack but a complex process, which could indicate the activation of a fault accompanied by a rapid subsidence. As conclusion we suggest a hypothesis to explain the polarity mismatch, based on a complex model which combines active reverse and normal faulting as the general mechanism for the August 21, 2017 Ischia earthquake. This hypothesis can explain the resurgence mechanism of Ischia caldera as reported in Molin et al. (2003) and is also proposed in other geological contexts as a copper mine in Poland (Rudzinski et al. , 2016): seismic activity is triggered by reverse faults, inducing a subsequent collapse. This collapse is compatible with the 4 cm vertical subsidence revealed by differential SAR in the adjacent area southwest of Casamicciola Terme (after De Novellis et al. , 2018). Under this hypothesis, the first motion at IOCA is controlled by the triggering event, whereas the moment tensor solution depends on the radiation pattern of the complex rupture. Fig. 3 - DC and full MT solutions for the Ischia earthquake. (a) Comparison of amplitude spectra fit (L2 norm) for different source depths, assuming a DC (gray line) and full MT (black line) source model. (b) Hudson plot representation of the source type for full MT and DC solutions at different source depths: best fitting solutions show significant negative tensile component, a potential signature of a collapse process. (c) Best full MT solution and related source parameters.