GNGTS 2016 - Atti del 35° Convegno Nazionale

202 GNGTS 2016 S essione 1.2 most of the Italian region where the typical hypocentral depth is in the order of 5 to 20 km. A few recent papers have proposed the combination of geophysical and geodetic data to infer properties of the crust-mantle coupling (e.g. Palano, 2015) or to constrain finite element models where the deformation is driven by large-scale forces (e.g. Carafa et al. , 2015). Such regional models provide fundamental insight for investigating large-scale geodynamic processes but are too coarse to be used for determining the relative state of stress of adjacent faults. Over the past two decades several papers (e.g. King and Cocco, 2000; Console et al. , 2008; Caporali et al. , 2016, among others) have investigated the Coulomb Failure Function (CFF) or Coulomb stress change in specific examples of earthquake sequences which are supposed to be both spatially and temporally related. In this approach the stress field generated by an earthquake of known fault plane solution is mapped onto a nearby “receiver” fault. This approach reveals whether this fault has been loaded or unloaded by the occurrence of the parent earthquake. In this paper we focus on the stress field which is measured on a regional scale by geodetic methods as a source of load on the faults. In particular, GPS geodesy provides a 2D strain rate that can be converted into an elastic plane stress rate (Fig. 1). This can be assumed to approximate the horizontal stress rate field throughout a 15-20 km-thick brittle seismogenic layer. Validation with independent stress data is possible thanks to the new improved stress map of Italy (Montone and Mariucci, 2016). Owing to the dense distribution of geodetic GNSS sites, this stress rate tensor can now be calculated for the centroid of several of the seismogenic sources of the DISS database. The normal and tangential stress rates relative to the fault plane can be computed, and finally the Coulomb Failure Function (CFF) as the difference between the tangential stress rate and the normal stress rate multiplied by an assumed friction coefficient. The obtained CFF represents a measure of the loading/unloading rate on the fault plane. Hence this analysis can eventually indicate the rate at which each fault is loading or unloading elastic energy. As a possible inference, regions where the current strains explain well the known seismicity can be identified, and areas that are historically quiescent but where stress is consistently building up Fig. 1 – Eigenvectors of the strain rate tensor inferred from GPS velocities (green arrows) are interpolated to the center of those 87 ISSs (brown rectangles) sufficiently covered by GPS data. Extension is in blue and compression in red. Major tectonic lineaments are in orange.