GNGTS 2016 - Atti del 35° Convegno Nazionale

90 GNGTS 2016 S essione A matrice latter in particular proved fundamental for understanding the details of the source processes and - even more importantly - for unraveling the puzzle of their associated surface coseismic deformation, represented by limited or ambiguous surface breaks and by extensive subsidence for all three earthquakes. In summary, the 1997, 2009 and 2016 sequences share the following common characteristics (see Fig. 2): - faulting occurred along southwest-dipping faults striking parallel to the local trend of the Apennines between 3 and 10 km depth. Neither seismological methods nor the inversion of GPS and InSAR data detected any slip above 3 km depth; - based on seismological, InSAR and GPS evidence, the dip of the seismogenic fault falls in the range 39°-51°, much lower than the dip commonly reported for central Apennines active faults (e.g. Vannoli et al., 2012) but coherent with a recent elaboration on the fault dip distribution of global normal faulting earthquakes (Basili and Tiberti, 2016); - the three earthquakes exhibit a relatively small magnitude, resulting from the limited width of the rupture plane and from the limited coseismic slip (<10 km and <1 m, respectively). Assuming a standard fault aspect ratio, such limited width implies a limited length, hence limiting the magnitude, although this does not prevent earthquakes from coming in clusters of similarly-sized events, such as in 1997; - most importantly, in all three cases the rupture appeared to be vertically confined by the interference with thrust surfaces dating to the construction of the Apennines fold-and- thrust belt (Fig. 3). In our opinion these three earthquakes delineate a characteristic seismogenic style for this portion of the central Apennines, where the upward propagation of seismogenic faults is precluded by the presence of pre-existing regional thrusts or other inherited structural features [as shown by the results of the experiments in Bonini et al. (2015)]. This implies an effective decoupling between the deeper seismogenic portion of the upper crust and its uppermost portion, i.e. between the surface and 3-4 km of depth. The decoupling implies (a) that active Fig. 3 – Geological cross sections across the causative faults of the 2009 L’Aquila, 1997 Colfiorito [based on data from Chiaraluce et al. (2005, 2009)], Campotosto [based on data from Bigi et al. (2012)] earthquakes. The seismogenic sources are shown by a red line; black dots show aftershock locations. The sections illustrate the confinement of the seismogenic sources within pre-existing thrusts, locally illuminated by minor aftershocks (from Bonini et al. , 2014, modified).