GNGTS 2019 - Atti del 38° Convegno Nazionale

104 GNGTS 2019 S essione 1.1 Scattering and absorption tomography. The aim of this analysis is to use earthquakes occurred in and around the area of the 2010-2014 Pollino seismic swarm to separate scattering and absorption contribution to the total attenuation. The choice of this technique is related to its higher sensitivity to heterogeneity and fluid variation if compared with velocity tomography. We mainly propose a physical validation of the presence of fluids and of the role that fluids may have had in driving the Pollino swarm (Passarelli et al. , 2015). We used 117 earthquakes selected among the events of the swarm and nearby, characterized by magnitude range between 1.8 and 4.3 and depth between 1 and 56 km (Fig.1C). We used all stations installed in the area between 2010 and 2014 to obtain a suitable ray coverage. Then we applied two techniques: the peak delay method to measure the scattering, and the Q-coda method to measure the absorption. The peak-delay, defined as the time lag from the direct-wave onset to the maximum amplitude of the signal envelope, is a direct measure of the multiple forward scattering (Calvet and Margerin, 2013). The greater the lag, the higher the scattering, that means presence of a significant amount of heterogeneities. We mapped the variation of the scattering in the medium using the regionalization approach. The Q-coda method is based on the equation by Aki and Chouet, 1975: (2) that describes the coda waves envelope. Q c -1 is the inverse coda quality factor and, in diffusive regime and in absence of leakage, represents the absorption quality factor Q i -1 . High Q c -1 is representative of high absorption, that usually corresponds to the presence of fluids. Assigning a Q i -1 value to each waveform, using the diffusive kernels by Del Pezzo et al. (2016) we inverted for the 2D spatial distribution of the variations of absorption in the area. Results and discussion. We performed the relative location of clusters of similar waveforms to obtain a detailed image some faults responsible for the 2010-2014 Pollino seismic swarm. One of the main results of the relative location analysis is an improved imaging of the fault size, position, direction and dipping. Relocated hypocenters depict a clear fault plane that we could only guess from the absolute locations. Fitting the hypocenter distribution obtained from the relative location we find a fault characterized by strike around 150°, dip 48°, and an overall normal fault kinematics. The seismogenic volume involved in the swarm is inferred to be 5x2x2 km 3 , at a depth between 4.5 and 6.5 km b.s.l., and its orientation in space is in agreement with the focal mechanism of the mainshock of M5.0 (s=166°, d=47°, r=−84°), that likely occurred at the southern end of the same structure. The high precision locations allows also to follow the evolution in time of the swarm, that nucleated from the central part of the seismogenic volume and migrated northward. Then, after few months, during which the seismicity migrated on the eastern fault the most of earthquakes occurred again in the main western fault filling the edges of the seismogenic volume. The lack of seismic stations during some periods is a serious issue because we could only have an incomplete picture of the swarm evolution. We mapped scattering and absorption taking into account part of the events of the swarm and surrounding area. The attenuation analysis highlights that the Pollino area is characterized by relative high scattering-high absorption at high frequency (12 Hz). We interpret this result as a fluid-filled connected fault network, mainly composed by medium size young structures that include the 2010-2014 volume imaged from the relative location (Fig.3). The results are in agreement with the slow strain regime highlighted by Cheloni et al. (2017), with the high v p /v s found by Piana Agostinetti and Amato (2009) and with the role the fluids that drove the sequence as suggested by Passarelli et al. (2015). In conclusion, the 2010-2014 Pollino seismicity was a swarm-like sequence occurred in an area of slow strain, filled by fluids that may have driven the sequence that started from the central portion of the main fault. The application of the techniques described here to other tectonic areas may contribute to better understand the physical mechanisms behind the nucleation and evolution of swarms. The investigation of areas where the exploitation of geo-