The location of potential large earthquake sources in the Apennines: major constraints from minor seismicity

Strong efforts have been made over the last several years in Italy to both identify the major sources for large earthquakes and locate events of even the smallest magnitude. As of today at INGV, two databases were compiled to accomplish these tasks: 1) the database of seismogenic sources for the largest earthquakes (M>5.5) which includes 60 fully-parameterized seismogenic sources and 181 tectonic lineaments; and 2) the database of instrumental seismicity for the past 20 years which includes about 50000 accurately located earthquakes that allowed the selection of 285 seismic sequences. We use these two databases to search for possible associations between seismic sequences and known/supposed earthquake sources in the Apennines.\Clusters of minor earthquakes could be grouped according to both horizontal distribution and depth, with 5 groups falling within areas of undergoing tectonic deformation with different styles: 1) the Tyrrhenian side (horizontal extension) is characterized by shallow seismicity (0-5 km) which clusters in a limited number of sequences; 2) the Inner Apennines (major normal faulting) are characterized by a significant number of well clustered sequences at depths of 5-15 km; 3) the Outer Apennines (mainly undetermined style of deformation) exhibit sparse seismicity and diffuse sequences at depths of 10-20 km; 4) the Adriatic side (horizontal compression) has few shallow (0-10 km) sequences; 5) areas affected by large tectonic lineaments (undetermined kinematics), which mark discontinuities in the lithosphere or in the crust, exhibit wide depth range seismicity (0-30 km). A significant number of seismic sequences, especially those of group 2, also show clear association with known seismogenic sources for they frequently occur at fault-segment boundaries and activate different fault segments at short time distance.\All in all, the analysis low-magnitude seismicity yields interesting inferences on the occurrence of future large earthquakes in the Apennines because 1) seismic sequences delineate fault segmentation, 2) large normal faults seem to be bounded by weak zones, and 3) weak zones may either correspond to areas with a different rheology or to clearly-detectable, pre-existing transverse structures.