GNGTS 2021 - Atti del 39° Convegno Nazionale

GNGTS 2021 S essione 3.2 424 AMBIENT SEISMIC NOISE AND MICROSEISMICITY MONITORING OF POTENTIALLY UNSTABLE ROCK MASSES: A NEW CASE STUDY FROM ORMEA (NW ITALY) C. Colombero 1 , A. Godio 1 , D. Jongmans 2 1 Politecnico di Torino, DIATI, Torino, Italy 2 Univ. Grenoble Alpes, ISTerre, CNRS, IRD, IFSTTAR, Grenoble, France Introduction Continuous passive seismic monitoring has reached a decade of applications on landslides of different type, geology and geometry (Colombero et al. , 2021). Spectral analysis and cross- correlation of ambient seismic noise simultaneously recorded at different stations can be applied to extract resonance frequency variations and seismic velocity changes within the investigated volumes. Both seismic parameters can show reversible fluctuations driven by external modification in air temperature and precipitation (Bottelin et al. , 2013; Colombero et al. , 2018a) and irreversible drops in their values while failure is approached (Mainsant et al. , 2012; Fiolleau et al. , 2020). In rock sites, the resonance frequency vibrations are spatially driven by the orientation of the fractures constraining the potentially unstable volume. In particular, for vertically elongated rock columns (i.e. separated from the stable rock mass by one or more near-vertical fractures), the first vibration mode is generally related to bending perpendicular to the direction of the rear fractures, thus indicating the potential direction of collapse (e.g. Valentin et al. , 2017; Bottelin et al. , 2018; Fiolleau et al. , 2020). The second vibration mode is usually bending perpendicular to the first mode, i.e. parallel to rear fracture orientation. More complex behaviors are detected at higher modes and on sites with 3D geometry (Colombero et al. , 2017). Beside ambient seismic noise analyses, continuous passive seismic monitoring also allow for the recording of microseismicity induced by the landslide movement and/or pre-failure seismic signals. These seismic events can be extracted from ambient seismic noise recordings by means of detection algorithms and their classification, source location and temporal trends can help to identify the most unstable compartments and track the evolution to failure (Lévy et al. , 2011). Here, we present the results of passive seismic monitoring on a potentially unstable quartzite tower located west of the town of Ormea (CN, NW Italy, Fig. 1a). The tower is located along a rock cliff with large near-vertical discontinuities, generating isolated rock pinnacles and towers. In recent years, a landslide occurred at the base of the tower, involving the shallow slope deposits and the fractured rock mass (Fig. 1b). The quartzite tower (approximately 3,000 m 3 ) is nearly 25-m high and 10x12 m wide at the top, bordered by three sets of fractures (dip direction/dip: K1 310/75, K2, 235/80, K3, 110/25, Fig. 1c and Fig. 1d). Both the active landslide and the quartzite tower are monitored by ARPA Piemonte by a set of extensometers located at the rear fracture of the active landslide and at the top of the quartzite tower (C1 to C3 in Fig. 1d). The increasing fracture opening and episodic material sliding at the base may progressively destabilize the tower and lead to its collapse, threating the riverbed, buildings and infrastructures downslope. A network of four wireless passive seismic stations was deployed on site (December 2019 – June 2020). Each station includes a 2-Hz high-sensitivity geophone, a digitizer and storage unit, GPS timing and GSM-GPRS module for network diagnostics transmission (GEA-GPS, developed by PASI s.r.l. and Iridium Italia s.a.s.). Two seismic stations were deployed at the tower top (S2 and S3, Fig. 1d), while two reference ones were located outside the tower, on the stable cliff (S1 and S4, Fig. 1d). Reversible variations of both ambient seismic noise and microseismicity parameters were investigated and compared with on-site displacement measurements and meteorological parameters to gain insight into the tower stability.