GNGTS 2016 - Atti del 35° Convegno Nazionale

GNGTS 2016 S essione A matrice 83 An intriguing perspective on the source geometry and slip distribution of the 2016 Amatrice Mw 6.2 earthquake (central Italy) from geological and satellite data P. Tizzani 1 , M. Bonano 1 , P. Boncio 2 , F. Brozzetti 2 , R. Castaldo 1 , F. Casu 1 , D. Cirillo 2 , C., De Luca 1 , R. de Nardis 2,3 , V. De Novellis 1 , F. Ferrarini 2 , R. Lanari 1 , G. Lavecchia 2 , M. Manunta 1 , M. Manzo 1 , A. Pepe 1 , S. Pepe 1 , G. Solaro 1 , I. Zinno 1 1 IREA-CNR, Napoli, Italy 2 DISPUTER, Università G. D’Annunzio, Chieti, Italy 3 Dipartimento della Protezione Civile, Roma, Italy Abstract. On August 24, 2016, at 01:36 UTC¸ the intra-Apennine extensional fault system of central Italy released a destructive earthquake (Amatrice 2016, M W 6.0 TDMT, M W 6.2, QRCMT). It produced widespread damage and fatalities, killing about 300 people and severely destroying the town of Amatrice and other small localities. The event dramatically recalls the April 6, 2009 normal fault earthquake (M W 6.3), which nucleated about 40 km southward of the Accumoli earthquake, causing about 300 casualties and destroying the town of L’Aquila. After few hours, the Accumoli earthquake was followed by a significant aftershock (M W 5.5, QRCMT), which nucleated ~15 km NW-ward. In the following days, five events having M W between 4.5 and 5.0 were released and the sequence mainly grew northward. At today (October 4, 2016), the epicentral area extends in the NNW-SSE direction, for a length of about 25-30 km. It is located at the hanging-wall of theWSW-dipping Vettore-Gorzano active extensional fault system. Relevant co-seismic deformations were highlighted soon after the main shock. This was located at a depth of about 8 km; its epicenters are located within the relay zone between the two en-echelon fault segments. The epicentral area well coincides with the pattern revealed by DInsAR measurements, which is characterized by a double-eyed co- seismic shape. In particular, we generated several interferograms by using ALOS and Sentinel 1-A and B constellation data acquired on both ascending and descending orbits to show that most displacement is characterized by two main subsiding lobes of about 20 cm on the fault hanging-wall; this is consistent with the calculated focal mechanism. By inverting the generated interferograms, following a classical Okada analytical approach, the modelling results account for two sources related to main shock and more energetic aftershock. The time interval between the ascending and descending (August 31, 2016) does not discriminate the effects derived from the main 24 August event and by its aftershock, but the low magnitude of the second event can only very marginally contribute to the overall deformation pattern. The reconstructed 3D fault model consists in two major interconnected fault segments, Vettore and Gorzano, which are individual at depth shallower than about 7-8 km and converge into a unified surface at higher depths. The Vettore-Gorzano unified surface has a length of 65 km, dips WSW-ward with an angle of about 45-50° and reaches a depth of 11 km, where, according to the proposed reconstruction, detaches on an east-dipping basal detachment. The stereo-projection highlights the almost perfect fit in attitude among the outcropping faults, the surfaces reconstructed at depths, the preferential seismic planes from focal mechanisms, as well as the attitude of primary co-seismic facture. Through Finite Element numerical modelling that jointly exploits DINSAR deformation measurements and structural-geological data, we reconstruct the 3D source of the Accumoli 2016 normal fault earthquake which well fit a M w 6.2, event. We evolved the model through two stages: during the first stage (pre-seismic), the model compacted under the weight of the rock successions (gravity loading) until it reached a stable equilibrium. At this level, we considered only the retrieved tensional field while maintaining the total displacements equal to zero. At the second stage (co-seismic), where the stresses were released through a non uniform slip along the faults, we used an iterative optimization procedure based on a trial-and-error approach,