Shear wave structure of Umbria and Marche, Italy, strong motion seismometer sites Affected by the 1997-98 Umbria-Marche, Italy, earthquake sequence

A long sequence of earthquakes, eight with magnitudes between 5 and 6, struck the Umbria and Marche regions of central Italy between September 26, 1997 and July 1998. The earthquake swarm caused severe structural damage, particularly to masonry buildings, and resulted in the loss of twelve lives and about 150 injuries. The source of the events was a single seismogenic structure that consists of several faults with a prevailing northwest-southeast strike and crosses the Umbria-Marche border. The focal mechanism of the largest shocks indicates that the events were the product of shallow extensional normal faulting along a NE-SW extension perpendicular to the trend of the Apennines.

The network of analog seismometer stations in the Umbria and Marche regions recorded motions of the main September and October 1997 events and a dense array of mobile digital stations, installed since September 29, recorded most of the swarm. The permanent national network Rete Accelerometrica Nazionale (RAN) is administered and maintained by Dipartimento delle Protezione Civile (DPC: Civil Protection Department); the temporary array was managed by Servizio Sismico Nazionale (SSN) in cooperation with small agencies and Universities. ENEA, the operator of many seismometer stations in Umbria, is the public Italian National Agency for New Technologies, Energy and the Environment.

Many of the temporary and permanent stations in the Italian seismic network have little or no characterization of seismic velocities. In this study, we investigated 17 Italian sites using an active-source approach that employs low frequency harmonic waves to measure the dispersive nature of surface waves in the ground. We used the Spectral Analysis of Surface Wave (SASW) approach, coupled with an array of harmonic-wave electro-mechanical sources that are driven in-phase to excite the ground. An inversion algorithm using a non-linear least-squares best-fit method is used to compute shear wave velocities for up to 100 meters of the soil column. A draft report was published in the summer of 2008, followed by a comment period, lengthy discussions with Italian colleagues, and improved knowledge of the subsurface at the sites from soil logs. Four of the sites were reprocessed in order to correct issues with phase unwrapping of the field dispersion curves that complicated the velocity profile calculations at the lowest velocity sites. This report presents the final results from the reprocessing effort.