Kinematic block modeling of GPS velocities in Italy and seismic potential

“Ettore Majorana” Foundation and Centre for Scientific Culture, Erice, Sicily, IT

We use a dense GPS velocity field, from the analysis of >1000 continuous stations, and
elastic block modeling to study the interseismic strain accumulation along the Alpine and
Apennines active tectonic belts in Italy. We consider available fault catalogues, instrumental and
historical seismicity to determine the blocks boundaries geometry, parameterized as uniformly
slipping rectangular planes. We invert horizontal velocities to estimate Euler vectors of tectonic
blocks together with slip-rates at block-bounding faults. When allowed by density of GPS data, we
optimize faults dip and locking-depth by searching the parameters that provide the best fit to local
GPS data. Overall we obtain a good fit of the horizontal velocities and geodetic slip rates that are
kinematically consistent with available geological and seismotectonic information.
We use the best-fit geometric and kinematic model parameters to compute the expected GPS
velocities over a dense regular grid. Denser model velocities are used to estimate the velocity
gradient field on a regular grid, made by cell elements of 0.25°x0.25°. Geodetic strain-rates at each
cell are converted into seismic moment accumulation rates, following the Kostrov formulation,
considering as seismogenic thickness values obtained from a crustal (EPcrust) model and
earthquake hypocentral distribution. Geodetic moment accumulation rates are compared with
seismic moment rates released by earthquakes, obtained from the analysis of a seismic catalogue
realized by merging several instrumental and historical catalogues covering the 1600-2012 timespan,
and uniformly defined moment magnitudes. The comparison between geodetic moment
accumulation rates and seismic moment release rates highlights regions with significant moment
deficits but also areas with a surplus of the seismic moment released, with important implications
for seismic hazard evaluations and assumptions behind the approach used in this work.