Zhu, L.

In this paper, we use regional seismic waveforms recorded by the recently-installed Istituto Nazionale di Geofisica e Vulcanologia (INGV) national network and the Mediterranean Very Broadband Seismographic Network (MedNet) stations to develop one-dimensional (1-D) crustal velocity models for the Italian peninsula. About 55,000 P -wave and 35,000 S -wave arrival times from 4,727 events are used to derive average seismic parameters in the crust and uppermost mantle. We define four regions, according to geological constraints and recent travel-time tomography results. Based on the average seismic parameters, we combine broadband seismic waveforms and travel-times of regional phases to model crustal structures for the four regions by applying the genetic algorithm. Our results indicate smooth velocity gradients with depth beneath the Apennines, and a deep Moho beneath the central Alps. Green’s functions from the regionalized 1-D velocity models are used to determine source depths and focal mechanisms for 37 events with magnitude larger than 3.5 by a grid search technique. Our results show that normal and strike-slip faulting source mechanisms dominate the Apenninic belt and most thrust faulting events occur in the Adriatic sea and the outer margin of the northern Apennines.

The aim of this study is to test the stability of moment tensor solutions for crustal earthquakes in the Calabro-Peloritan area (southern Italy). We used waveforms recorded by the Italian National Seismic Network managed by the Istituto Nazionale di Geofisica e Vulcanologia and the CAT-SCAN (Calabria Apennine Tyrrhenian - Subduction Collision Accretion Network) project. We computed the moment tensor solutions using the Cut And Paste (CAP) method. The technique allows the determination of the source depth, moment magnitude and focal mechanisms using a grid search technique. For the earthquakes investigated, we tried different station distributions and different velocity models. Results were also checked by computing the moment tensor solutions using the SLUMT grid-search method. Both methods (CAP and SLUMT) allow time shifts between synthetic and observed data in order to reduce the dependence of the solution on the assumed velocity model and on earthquake location errors. Comparisons have been made with the available published solutions. The final focal mechanisms were robustly determined. We show that the application of the CAP and SLUMT methods can provide good-quality solutions in a magnitude range not properly represented in the Italian national earthquake catalogues, and where the solutions estimated from Ponset polarities are often poorly constrained.