Landslide detection and monitoring by integrating electrical, seismic and interferometric techniques in a multi-hazard perspective: the case of San Vito Romano (RM)

Multi-hazard approaches to mitigate landslide risk can play a crucial role in promoting the sustainable planning of vulnerable urban areas. These procedures are particularly effective for managing events that occur independently but are triggered by multiple factors, involving landslides with different mechanisms. The application of appropriate investigative techniques is essential for studying complex landslides (Mangifesta et al., 2024). In such heterogeneous structures, the combination of different geophysical techniques has proven to be indispensable for obtaining reliable results (Jongmans et al., 2007). In this work, we present an integrated approach for the characterization and monitoring of the San Vito Romano (RM) landslide, combining active electrical and seismic techniques, interferometry and continuous ambient noise recordings. The results also lay the groundwork for long-term objectives, such as developing a quantitative tool that correlates satellite-obtained displacements with proxies derived from passive seismic techniques. This correlation aims to better understand the role of preparatory and triggering factors, such as precipitation and earthquakes, in landslide activity The investigated active retrogressive landslide occurs within the Tortonian Flysch of the Frosinone Formation. On the landslide slope, the weathering of sandstone and clay layers serves as a predisposing factor, while soil moisture and freeze-thaw cycles act as preparatory factors. Triggering factors include cumulative precipitation and seismic events. Therefore, the complexity of the landslide calls for a preliminary characterization of the geophysical properties through integrated methods. To this aim, we investigated two areas located outside (Line 1) and inside (Line 2) the landslide. (Fig. 1) Line 1 was carried out by a long ERT line, made of 81 electrodes spaced 10 m apart, using a multiple-gradient array with the ABEM Terrameter LS2 (Guideline Geo). The ERT line was then complemented by two MASW lines, with 48 geophones spaced 1.5 m apart and 2 Geode (Geometrics) seismographs, performed along the ERT profile, for retrieving the shallower shear-wave velocity profile (DOI ~ 30 m). At Line 2 we combined ERT and seismic refraction tomography (SRT) on a profile with 48 sensors spaced 1.5 m apart, together with a MASW survey and a down-hole investigation for the near-surface geophysical characterization of the area within the landslide. ERT and SRT data were inverted using pyGIMLi (Rücker et al., 2017), while MASW data were inverted with the code after Cercato (2009). The resistivity model at Line 1 (Fig. 2) enables a deep characterization of the site, reaching a maximum depth of investigation (DOI) of ~ 150 m, with a three-layer model, displaying the transitions between the marl (conductive) and the arenaceous (resistive) facies of the flysch formation. Then, for a continuous monitoring of the ambient seismic noise at the study site, we installed four seismometers (SARA electronic instruments) with an eigenfrequency of 2.5 Hz (SVR1-2-3-4), and 0.1 Hz lower corner frequency, two inside and two outside the landslide, with a sampling rate of 250 Hz (Fig.1). Recorded data were processed using MSNoise, an open-source software that allows for the analysis of ambient noise (Lecocq et al., 2014), in terms of horizontal-to-vertical spectral ratio (HVSR) of the noise. In addition to these techniques, the present study also incorporates satellite interferometry data, adopting a DInSAR approach called SBAS (Small Baseline Subset). This interferometric technique makes it possible to generate maps of the average deformation velocity and to monitor the temporal evolution of displacements at individual points in the image (CNR-IREA). More specifically, the Sentinel-1 satellite images were used. The preliminary results of the HVSR analysis (Fig. 3) indicate an increase in amplitudes at higher frequencies during rainy periods (highlighted in the red box). Notably, this effect is observed only in the seismic noise measurements recorded by the instruments located within the landslide area (SVR3 and SVR4). Upon completion of this phase, the continuous recording of ambient seismic noise over the medium to long term will enable the monitoring of ground instability in the coming months. For this purpose, single-station landslide mobility indicators such as relative velocity change, natural period and resonance peak polarisation will be used. The surface wave velocity variation will be determined using seismic interferometry techniques applied to environmental noise. By cross-correlating the recorded signals, this method facilitates the reconstruction of the Green's function, which describes the wave propagation between stations. The analysis of the velocity of the surface seismic waves (Rayleigh waves) and their temporal variations will provide insights into changes in the mechanical properties of the affected soil. Additional parameters will be derived using the Python library HVSRpy, which allows for the calculation of the Horizontal-to-Vertical Spectral Ratio (HVSR) over time. The processing was carried out with the following parameters: a window length of 200s, Konno-Ohmachi smoothing, and the geometric mean for averaging the horizontal components. These indicators enable continuous monitoring of changes in the soil’s mechanical properties, serving as proxies for detecting changes in rigidity, amplification and non-linear elastic behavior. Acknowledgements This work was partially funded by the “Sapienza” Ateneo Research project 2021 prot. RM12117A8A245367. “Advanced Active and Passive Seismic Techniques for Landslide Characterization and Monitoring”. P.I. Michele Cercato. The PhD research grant is funded within the framework of the extended RETURN partnership and is financed by the European Union Next-Generation EU (National Recovery and Resilience Plan - PNRR, Mission 4, Component 2, Investment 1.3 - D.D. 1243 of 2/8/2022, PE0000005). References Cercato M.;2009): Addressing non‐uniqueness in linearized multichannel surface wave inversion. Geophysical Prospecting, 57(1), 27-47. Institute for Environmental Research and Elevation (IREA). SBAS. IREA CNR. http://www.irea.cnr.it Jongmans D., & Garambois S.;2007: Geophysical investigation of landslides: a review. Bulletin de la Société géologique de France, 178(2), 101-112. Mangifesta M., Aringoli D., Pambianchi G., Giannini L. M., Scalella G., & Sciarra N.; 2024: A Methodologic Approach to Study Large and Complex Landslides: An Application in Central Apennines. Geosciences, 14(10), 272. Lecocq T., Caudron C., Brenguier F.; 2014: MSNoise, a python package for monitoring seismic velocity changes using ambient seismic noise. Seismological Research Letters 85.3 715-726. Rücker C., Günther T., Wagner F.M., 2017: pyGIMLi: An open-source library for modelling and inversion in geophysics. Computers and Geosciences, 109, 106-123.