SEISMICITY AND SEISMOGENIC STRUCTURES IN CENTRAL APENNINES FROM TWO RECENT PASSIVE SEISMIC EXPERIMENTS: THE SLAM PROJECT (2009-2013)

Abstract

As its long historical record shows, the portion of the Central Apennines which extends to the Tyrrhenian Sea margin and includes the Avezzano-Sulmona area to the north and the Isernia-Cassino area to the south, is affected by moderate to strong seismicity. In the last millennia three highly energetic events occurred: Venafro 1349, Boiano 1805 and Marsica 1915 with effect as large as I = X-XI Mercalli - Cancani - Sieberg Scale (MCS). More recently the study area was hit by the Val Comino seismic sequence (May 1984, Mw 5.9) at the border between Lazio and Abruzzo. Two passive seismic experiments were carried out in the period between 2009 and 2013: the first one with 4 seismic stations (October 2009-January 2010) deployed in the Marsica-Sora area and the second one with 17 stations (November 2011-October 2013) deployed in the whole Central Apennines study area. These two arrays, together with the 20 stations of the Abruzzo regional seismic network (RSA) and the stations of the Italian seismic network (RSNC), have recorded in the study period 6742 earthquakes with local magnitude (ML) ranging from 0.5 to 4.8. Temporary arrays data were integrated with the data of the two IESN (Italian Experimental Seismic Network) stations of southern Lazio and with those of the 5 stations of the Molise regional seismic network. The data considered in this study were recorded by a total of 74 seismic stations. We re-picked the arrival times of the events recorded by the RSNC and RSA and picked those recorded by the two temporary arrays. Our dataset consists of 39.322 P- and 32.600 S-arrival times. We read also P-wave first motion polarities to compute focal mechanisms. During the study period several seismic sequences occurred. Among these the most important are the Campoli Appennino sequence (September-October 2009) with 1299 events (ML max 3.6) and the Sora sequence with 606 events (ML max 4.8). With this reviewed dataset we computed the best minimum 1-D velocity model for both Vp and Vs applying the Velest code. We compare and discuss the improved earthquake locations with the known seismogenic structures of the study area. The future planned activities will include the generation of high resolution 3-D tomographic models (Vp, Vs, Qp, Qs) of the fault systems, the relocation of all the studied seismicity with 3-D velocity models, the computation of the focal mechanisms, a review of the ML and the mapping of b-value. Further studies could be on the temporal behaviour of observable seismic parameters, for instance the change in time of Vp/Vs during the seismic sequences.