GNGTS 2017 - 36° Convegno Nazionale

306 GNGTS 2017 S essione 2.1 conservation principle (Field et al. , 1999) that allows the estimation of the seismic moment rate from long-term slip rate and geometry of the fault source. F2 assumes that the distribution of the on-fault earthquake magnitudes follows a truncated GR model from a threshold magnitude Mw 5.5. The seismic moment rate has been thus partitioned and balanced over this range of magnitudes. F2 estimates the seismicity rate of smaller magnitudes using the smoothed- seismicity approach proposed by Frankel (1995) applied to the historical and instrumental catalogues in a regular spatial grid. For cells of the grid that overlap the surface projection of an individual fault, the rates of occurrence evaluated by the smoothed seismicity approach from magnitude greater or equal than Mw5.5, have been replaced by the rates estimated for that individual source, taking into account the extension of the overlapping area. Smoothing seismicity models. Model G1. it is a smoothed seismicity model that has been built assuming that larger earthquakes occur at, or nearby, clusters of past smaller earthquakes. G1 merges two different smoothed seismicity models following the well-known and widely applied fixed (Frankel, 1995) and adaptive smoothing methods. These two models have been merged accounting for the outcomes of likelihood tests. G1 follows the same procedures of previous studies to optimize the correlation distance (fixed smoothing) as well as the neighbouring number (adaptive smoothing); these procedures are based on likelihood scores, which estimate the likelihood that the observed earthquake epicentres from the recent catalogue are derived from the smoothed rate models. Model G2 . the model follows the Woo (1996) approach and proposes a zone-free method solely based on the use of the earthquake catalogue. Proxy seismogenic sources have been “created” from the epicentral locations of the events that are smoothed according with their fractal distribution in space. A grid of point sources has been defined around the site of interest and the activity rates of each source have not been computed using a recurrence relationship, such as the GR equation, but they have been calculated from the density and proximity of events lying within that magnitude range. The contribution of each earthquake to the seismicity of the region has been smeared over all grid points falling within an epicentral distance that depends on the magnitude of the event itself. This magnitude-dependent relationship has been defined by a kernel function. Moreover, given the availability of some earthquake-fault associations in the DISS 3.2.1 database, an anisotropic kernel function has been used when there is a connection between faults and earthquakes. Geodetic models. Model G3. this model estimates the seismicity rates over the whole Italian territory using exclusively the GPS velocity field. In particular, G3 has been based on the analysis of 919 GPS derived horizontal velocities, after excluding stations in volcanic areas and velocities that are inconsistent with regional velocity field. The strain rate tensor field has been calculated on a regular 0.1° x 0.1° grid using the VISR software (Shen et al. , 2015) taking into account the variable station spacing for the optimal smoothing parameters and finally applying a Gaussian filter of 50 km to the scalar strain rate value. The model converts the strain rate in seismic moment rate and then to earthquake rate under the assumption (Ward, 1998) that earthquakes magnitudes follow a tapered GR distribution, where the b-value and corner magnitude are given. Model G4. the model follows the approach proposed by Bird and Liu (2007), Bird et al. (2010), and Bird and Kreemer (2015), with some adaptations to Italy. A probabilistic method to assign upper crustal earthquakes from the historical catalogue to their presumed causative faults has been used. Then, the so-collected events have been grouped into three kinematics sub-catalogues. Then, the parameters of their GR distributions have been computed. These distributions have been integrated to estimate the long-term seismic moment rate for each class. Hence to evaluate expected rate of seismicity: 1) the principal axes and principal values of the long-term strain-rate tensor have been computed by using GPS data; 2) based on the orientation and relative magnitude of the principal axes and values and rake of active faults, the given grid point has been characterized by a certain amount of strain-rate and slip-rate; 3) the tectonic