GNGTS 2017 - 36° Convegno Nazionale

GNGTS 2017 S essione 3.1 573 The 6 April 2009 Mw 6.1 earthquake and the 2016-2017 Amatrice-Visso-Norcia seismic sequence (culminated with the Mw 6.5 earthquake on 30 October 2016) in central Italy are an expression of all that we discussed above. Both seismic sequences were caused by the activation of complex and segmented normal fault-systems characterized by the presence of Quaternary basins in their hangingwall (the MiddleAterno Valley basin – MAV, for the L’Aquila earthquake; the Norcia and Pian Grande di Castelluccio basin for the Norcia earthquake). Prior to these two catastrophic events, the overall deep geometry of the basins, the structural setting, and the long- term activity of the earthquake causative fault-systems were only roughly defined. Onshore settings may pose severe limitations to the acquisition of good quality seismic data, which usually provide very high-resolution images of the subsurface. Moreover, such kind of surveys are expensive and time-consuming. For the abovementioned reasons, we adopted a different geophysical approach based on the integration of costly-effective, fast and accurate geophysical methods. Time domain electromagnetic (TDEM) soundings, single-station ambient vibration (horizontal to vertical spectral ratio - H/V) and electrical resistivity tomography (ERT) data have been applied in these two seismic areas in order to: 1) image the 3-D basin geometry, 2) characterize the subsurface of active normal faults and 3) gain insights into the long-term basin evolution integrating the geophysical data with existing geophysical, geological and paleoseismological information. The 3-D reconstruction of the MAV. The MAV is a >100 km 2 Quaternary basin surrounding L’Aquila and generated by the long-term activity of the Paganica-S. �������� ����� ������ Demetrio Fault System (PSDFS, Civico et al. , 2012). ����� ����� � ��� ������� �������� �� ��� ���������� �������� �� There exist a few studies focusing on the subsurface geometry of the MAV through geophysical investigations, mostly consisting of 2-D seismic and electrical surveys acquired over limited portions of the MAV (Improta et al. , 2012). Deep ERT profiles (Pucci et al. , 2016) and gravimetric map (Cesi et al. , 2010) provide hints of the existence of some deep depocenters (up to a depth of 500-600 m) filled with moderately resistive deposits, however a thorough 3-D image of the basin was not achieved. We performed an extensive geophysical survey made up of 86 TDEM soundings and 155 H/ V recordings in order to accurately image the buried interface between the basin infill deposits and the top-bedrock over the entire area (Civico et al. , 2017; Fig. 1A). Both methods owe their popularity to their relatively easy deployment, cost efficiency, and reliability in detecting the depth of large lithological changes (i.e., resistivity variations and impedance contrast between soft soils and bedrock, respectively), obtaining at the same time significant penetration depths. We first calibrated our geophysical data with an existing well log which hit the bedrock at relatively shallow depth (Fig.1B-a). Afterwards, both methods were tested over areas with different bedrock depths (as suggested from existing ERT and gravity models) in order to check the consistency between the two methods in recovering the depth to bedrock (Fig.1B). The presence of infrastructures, buried and/or surficial lifelines, posed severe limitations to the survey design representing a source of disturbance for TDEM and H/V methods. In addition, given the complex 3-D geological architecture of the survey area, other limitation occurred from 3-D effects as due to the presence of strong lateral resistivity contrast which may vanishing the 1-D assumption for TDEM data modelling. For the TDEM survey, we used a calibrated Geonics digital Protem equipped with both high-frequency 1-D and 3-D receiver induction coils, coupled with a transmitter square loop of 50 m and 100 m size (Sapia et al. , 2015). TDEM measurements were performed either in central loop and offset receiver configurations, covering the 7 μs to 28 ms time interval to accurately measure both the early and late time transient amplitudes. To choose the optimal loop size and recording time in order to provide a sufficient depth of investigation, we made use of the available gravity map for the MAV area. The existence of long-wavelength gravity anomalies (attributable to deep-seated geological substratum), suggested the 100 m loop size as more suitable to recover deeper depths to bedrock, while the 50 m loop size better reproduced the shallower ones.