GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 3.2 ___ GNGTS 2023 terminal part of the tongue. The rock glacier thickness, estimated from the height of the rock glacier front, is approximately 20 to 30 m. Since 2012, the site is periodically surveyed with UAV (Unmanned Aerial Vehicle) photogrammetry by ARPA Valle d’Aosta for the evaluation of the rock glacier flow rates in the framework of climate change impacts on high-mountain environments and infrastructures, such as the ski resort of Cervinia. The front of the landform is indeed flowing on the ski run (Fig. 1c), thus requiring constant maintenance and attention. A big artificial lake for hydropower generation (Goillet Lake, Fig. 1b) is also present downward the rock glacier flow, posing additional concerns in the case of rapid movements or fluidification of the material. The highest flow rates were depicted on the black lobe, with a displacement of 3.8 m in the frontal sector between 2012 and 2015 occurring especially in the summer months (Dell’Asta et al., 2017). The site is also regularly surveyed with active geophysical measurements, including seismic refraction and electrical resistivity tomography (Bearzot et al. 2022). Methods Due to these considerations, the passive seismic network installed on site in late July 2020 is aimed at continuously recording ambient seismic noise for the understanding of the ongoing internal processes. Two wireless seismic stations were installed in the frontal portion of the black lobe (S2 and S3 in Fig. 1c), one station was deployed on the white lobe (S4 in Fig. 1c) and a fourth one was placed in a stable area outside the rock glacier (S1 in Fig. 1c) to be used as a reference station. Each station comprises a 2-Hz triaxial high-sensitivity geophone and an on-purpose designed digitizer/recorder (GEA–GPS, developed by PASI s.r.l. and Iridium Italia s.a.s.), ensuring continuous seismic noise recording at 250-Hz sampling frequency, low power consumption in the absence of an external power supply (approximately 30 days of autonomy) and daily remote information about the system state of health by a GSM–GPRS module. Synchronization between the different stations is provided by GPS timing. Data from each station are stored in 1-hour files in an internal memory card. Data processing includes ambient seismic noise spectral analyses and cross-correlation between station pairs, together with microseismicity analyses. In particular, the power spectral density of each station and/or the computation of single-station (e.g. H2/V2) and site-reference (e.g. E2/E1 or N2/N1) may highlight noise amplification in specific frequency bands related to internal glacial processes. In addition, cross-correlation of noise recorded between the same component of two stations (e.g. V1 and V2) may potentially disclose seismic velocity changes within the rock glacier due to modifications in the material properties and flow rate increase or decrease. Microseismic events related to glacial processes (e.g. icequakes, rock glacier flow, water seepage, rock falls) were complementary extracted from continuous noise recordings through a STA/LTA detection algorithm and classified through k-means cluster analysis of selected time- and