No abstract

Rock slope failures can lead to huge human and economic loss depending on their size and exact location. Reasonable hazard mitigation requires thorough understanding of the underlying slope driving mechanisms and its rock mass properties. Measurements of seismic ambient vibrations could improve the characterization and detection of rock instabilities since there is a link between seismic response and internal structure of the unstable rock mass. An unstable slope near the village Gondo has been investigated. The unstable part shows strongly amplified ground motion with respect to the stable part of the rock slope. The amplification values reach maximum factors of 70. The seismic response on the instable part is highly directional and polarized. Re-measurements have been taken 1 year later showing exactly the same results as the original measurements. Neither the amplified frequencies nor the amplification values have changed. Therefore, ambient vibration measurements are repeatable and stay the same, if the rock mass has not undergone any significant change in structure or volume, respectively. Additionally, four new points have been measured during the re-measuring campaign in order to better map the border of the instability.

An experimental quantification of the strength and volume of real, heterogeneous, fractured rock masses is crucial when assessing rock slope stability. In order to quantitatively characterize the internal structure of fractured rock slopes, we present three‐dimensional numerical simulations of seismic wave propagation and compare with observations. We introduce a simple, effective model for fractured rock mass, which can easily be applied to simulate weak‐motion seismic wave propagation. The macroscopic compliant fractures cutting the rock mass are modeled as finite‐width zones of reduced elastic parameters characterized by shear and normal stiffness. The widths of such zones are not fixed and can be adjusted to fit the grid step in the numerical method. The proposed rock mass model is applied and tested for the Walkerschmatt site in southwest Switzerland. Synthetic ambient vibrations are generated using a finite‐difference method for the fractured rock mass, shaped by the real terrain geometry, and compared with the measurements. The observed seismic response is satisfactorily reproduced in a broad frequency range (0.5–10 Hz). The synthetized response is primarily controlled by the stiffness, depth, number of fractures, and inertial mass of the fractured rock. The simulated amplification and ground‐motion directionality correspond with the observed levels, unless (1) the simplified cracks reach depths of 200–300 m; and (2) the fracture network is larger with respect to the mapped network. This illustrates the potential of ambient vibration methods in combination with numerical simulations to infer depth, volume, and mechanical characteristics of slope instability.

In this study, the seismic response of two slope instabilities is investigated with seismic ambient vibration analysis. Two similar sites have been chosen: an active deep-seated slope instability at Cuolm da Vi and the geologically, structurally and morphologically similar, but presently not moving Alp Caschlè slope. Both slopes are located at the upper Vorderrheintal (Canton Graubünden, Switzerland). Ambient vibrations were recorded on both slopes and processed by time-frequency polarization and site-to-reference spectral ratio analysis. The data interpretation shows correlations between degree of disintegration of the rock mass and amplification. However, the ambient vibration analysis conducted, does not allow retrieving a resonance frequency that can be related to the total depth of the instability of Cuolm da Vi. Even though seismic waves can be hardly traced in rock instabilities containing open fractures, it was possible to retrieve a dispersion curve and a velocity profile from the array measurement at Cuolm da Vi due to the high level of disintegration of the rock material down to a depth of about 100 m. From the similar amplification pattern at the two sites, we expect a similar structure, indicating that also the slope at Alp Caschlè was active in the past in a similar manner as Cuolm da Vi. However, a smoother increase of amplification with frequency is observed at Alp Caschlè, which might indicate less disintegration of the rock mass in a particular depth range at this site, when comparing to Cuolm da Vi where a high level of disintegration is observed, resulting from the high activity at the slope.From the frequency-dependent amplification, we can distinguish between two parts within both instabilities, one part showing decreasing disintegration of the rock mass with increasing depth, for the other parts less-fractured blocks are observed. Since the block structures are found in the lower part of the instabilities, they might contribute to the stability of the slopes. Using the velocity profiles, it was possible to estimate the depth of the two largest open fractures (i.e. tension cracks) at Cuolm da Vi.

<p>Earthquake-induced landslides can have serious social impacts, causing many casualties and significant damage to infrastructure. They are the most destructive secondary hazards related to earthquakes. The impact of strong seismic events is not limited just to triggering of catastrophic slope failures, it also involves weakening of intact rock masses and reactivation of dormant slides. Hazard mitigation of potentially catastrophic landslides requires a thorough understanding of the mechanisms driving slope movements and seismic response.</p><p>We present an overview of the investigations on more than 25 instabilities. The results show that ambient vibration measurements allow for a rapid and objective characterization of potential slope instabilities. It is possible to distinguish unstable from stable areas, to identify slope eigen-frequencies, local amplification levels due to weak excitation, local deformation directions and properties of the internal slope structure. The ambient vibration techniques include single-station H/V ratios and polarization analyses, site-to-reference spectral ratios, array methods to identify surface-wave dispersion curves, and/or normal mode analysis using enhanced frequency domain decomposition. We analyse the seismic response of the rock slopes in different frequency bands together with its spatial and azimuthal variability, which is a fingerprint of the slope’s internal structure at different scales (tenth of meters to hundred meters). Normal mode behaviour is typically observed in structures with distinct sub-volumes, where the wave field at the resonance frequencies is oriented perpendicular to the deep persistent fractures. These structures show maximum amplification at their resonance frequency. Normal mode behaviour is also observed for rock towers, similar to what can be observed for buildings. In contrast, a highly fractured rock mass without dominant cracks is characterized by an S-wave velocity gradient with shear-wave velocity being significantly reduced close to the surface. Generally, normal modes do not develop, but surface waves propagate in such structures, which can be used for the determination of the S-wave profile. This is typical for large deep seated landslides with a layered structure. Without strong S-wave velocity contrast at depth, H/V spectral ratios show no clear peak and are not conclusive to characterize structures with highly fractured material. However, frequency-dependent ground-motion amplification from standard spectral ratios is directly related to the S-wave velocity profile and damping. Therefore, wave amplification can be a measure for the disintegration of the rock.</p><p>Repeated measurements on slopes allow for the detection of possible changes in their properties. Semi-permanent installations on instabilities of interest allow for a continuous assessment of the dynamic response in order to understand variations due to weather conditions and potential long-term changes. This includes the measurement of site-amplification during earthquakes derived from empirical spectral modelling. When measuring in the same season and weather condition, the seismic response of rock instabilities in general remains unchanged over years, as long a no external trigger affects the instability, including a strong earthquake, partial failure of the slope or permafrost degradation.</p>