GNGTS 2022 - Atti del 40° Convegno Nazionale

110 GNGTS 2022 Sessione 1.2 relocation, we obtained relative location for 3062 and 2603 events, respectively for the two time spans. The seismicity distribution of the 1980s is widely scattered throughout the entire volume of the caldera and the average depth of events is placed around 2km, below the hydrothermal system, with a significant percentage of events located 3-4 km deep. The seismicity of the last decades is more concentrated around the volume of Solfatara, with mean depth around 1.3 km. The relative relocation shows some near-vertical fracturing alignments under the Solfatara that bend deep toward east (Bagnoli), suggesting that some of the preferred paths of rising fluids across the hydrothermal system could have an origin at a depth of about 2-4 km (De Siena et al. , 2017), where events of more significant magnitude are located (Md>2). Unlike what was observed in the 1980’s, recent seismicity is almost completely absent west of the Solfatara and in particular below 1 km depth, suggesting the presence of an impermeable structural barrier, mainly oriented in the EW direction (in accord with Buono et al. , 2022, and Calò and Tramelli, 2018). Since 2005, the OV catalog reports 4 events of magnitude Md>3 located in the CF area, two of them occurring in March 2022 (Table 1). We focused our attention on these events aiming at the quantitative estimation of some seismological parameters derived from waveform and spectral analysis. Tab. 1 - Ipocentral parameters extracted from the Osservatorio Vesuviano (Istituto Nazionale di Geofisica e Vulcanologia) catalog for the events under study. Origin Time (UTC) Lat (°) Long (°) Depth (km) Md 2019/12/06 00:17:23 40.8302 14.1493 2.3 3.1 2020/04/26 02:59:02 40.8292 14.1500 2.6 3.3 2022/03/16 14:14:34 40.8272 14.1402 2.7 3.5 2022/03/29 17:45:32 40.8293 14.1507 2.5 3.6 We determined the focal mechanism solutions with the FPFIT package from the P-wave first motion polarities (Reasenberg and Oppenheimer, 1985). For each event, we used the first motion polarities at seismic stations in a distance of 20 km from the epicenter. The 90% confidence intervals for the mechanism strike, dip, and rake are determined by finding how much each parameter may change without exceeding a critical misfit level between observed and predicted polarities. We used the resulting fault plane solutions to calculate the predicted surface vertical displacement and we compared it with the observed static displacement measured at close seismic stations following the method proposed by Zhu (2003). The permanent vertical displacement at each station is extracted from the zero- frequency seismic signal. It is worth noting that for low-energy events, such as those under attention, the measurement of the static displacement from seismic waveforms is exceptional and very valuable, as it is capable of detecting subsidence/uplift of the order of hundreds or even tens of micron, thus proving to be complementary to GPS data, which hardly can achieve sub-millimeter accuracy. Moreover, in the near field significant co-seismic rotations can affect the horizontal components of the acceleration in seismic records. For the events under study, we deduced the tilt amplitude from the observed acceleration jumps at near-earthquake stations. The results show a good agreement with the measurements recorded at tiltmeter stations of the CF network and therefore can be useful for the definition of the near-field ground deformation and rotation arising from local earthquakes. Finally, assuming the observed displacement spectrum formulation according to Boatwright (1980), we deduced the source parameters (corner frequency, source radius, seismic moment)