- Ground-motion model is: where Y is in cm∕s2, Mref = 5.0, Rref = 1.0, Mh = 6.75, a = 3.863, b1 = 0.004, b2 = -0.070, c1 = -2.039,
c2 = 0.222, H = 11.91km, fNF = 0.036, fSS = -0.036, sGR = 0.479, sST = 0.475, sSO = 0.617, ϕ = 0.107
(intra-event), τ = 0.322 (inter-event) and σ = 0.339 (total).
- Distribution w.r.t. Eurocode 8 site classes shows vast majority of records from classes A or B based on
surface geological information (few measurements of V s,30.
Identify 22 reference rock (RR) sites based on the lowest misfit between recorded response spectra and
predicted response spectra from Bindi et al. (2011a) GMPE for Italy. Find a lower amplification for these
stations than those used to derive the Bindi et al. (2011a) model. Find a number of Eurocode A sites in
Sicily are not classed as RR sites and a number of Eurocode B and C sites in Calabria are.
Use 4 site classes:
- Reference rock. About 160 records.
- Eurocode 8 class A sites but not RR. Use SGR. About 300 records.
- Stiff soil: Eurocode 8 class B and E sites. Use SST . About 290 records.
- Soft soil: Eurocode 8 class C and D sites. Use SSO. About 90 records.
- Use 3 mechanisms (exclude data from reverse events because poorly represented in dataset):
- Normal. Use FNF
- Strike-slip. Use FSS
Distribution of complete database w.r.t. mechanism is: normal (43 events), strike-slip (22 events), thrust
(13 events) and unknown (98 events).
- Data from southern Calabria and Sicily (Italy) from 1978 to 2016 from about 230 different stations
of 4 networks: Rete Accelerometrica Nazionale, Italian National Seismic Network, Mediterranean Very
Broadband Seismographic Network and Calabria-Appennine-Tyrrhenian/Subduction-Collision-Accretion
Network (2003–2006). About 1200 records from accelerometric stations and about 2000 from velocimetric
- Selection criteria are: 35.6-40.2∘ N and 12-18∘ E; repi ≤ 200km and M > 3.5.
- Uniformly process all records using the tool of the Engineering Strong Motion database.
- More than 90% of records from M ≤ 5.
- Roughly uniform distribution w.r.t. distance.
- Select a calibration subset by excluding records from M ≤ 4.0, focal depth ≥ 25km and volcanic events of
Mount Etna and the Aeolian Islands with depth < 5km.
- Data from 194 different stations in calibration subset.
- Examine residuals w.r.t. magnitude and distance and find no trends.
- Model poorly constrained for < 10km because of lack of data.
- Compare predictions and observations for M4.0 and M5.0 and find good fit.