Assessing the hazard of unstable rock blocks using seismic noise


Robust estimation of rock mass volume is a key challenge for rockfall hazard assessment. However, this estimation remains particularly difficult due to the irregular geometry of blocks, the heterogeneity of material mechanical properties, and the poorly constrained boundary conditions, particularly at depth.


Fig. 1. Numerical simulation of rock-block resonance frequency as a function of block geometry and boundary conditions. The left column (a, c, e) shows a slender block (5 × 5 × 20 m) and the right column (b, d, f) a thick block (10 × 10 × 10 m). Panels (a, b) show a fixed base, (c, d) a block resting on a softer layer, and (e, f) a block embedded 10 m into the layer. Each panel shows the fundamental resonance frequency f₀ and normalized displacement. Adapted from [Bottelin et al., 2026b].


Fig. 3. Seismological measurements at the top of the column during drilling. (a) Recorded acceleration and strain proxy εproxy. (b) Power spectrum showing a decrease in the fundamental resonance frequency f₀ from 9.0 Hz outside drilling to 8.5 Hz during drilling. (c) Spectrogram highlighting the drops in f₀ at the onset of drilling and its recovery when vibrations cease. Adapted from [Bottelin et al., 2026b].

A recently proposed approach in the literature [Pazzi et al., 2025] consists of studying the fundamental natural frequency of rock blocks — estimated, for example, from a horizontal-to-vertical spectral ratio — and relating it to their volume using a chart derived from numerical simulations of parallelepipedal blocks fixed at their base [Pazzi et al., 2025]. However, Bottelin et al. [2026b] have shown that several limitations can affect the robustness of volume estimates based on the fundamental natural frequency, notably: (i) the representativeness of numerical models with respect to actual site conditions, (ii) the quality and processing of geophysical data, and (iii) the methodological assumptions and validation procedures used. These limitations must be carefully taken into account when interpreting the results in order to rigorously estimate the volume of rock masses from their fundamental resonance frequency.


In a second study, Bottelin et al. [2026a] investigate, through in situ seismological measurements, the evolution of the resonance frequency of an unstable rock column (Fig. 2). These measurements were performed during drilling operations carried out to reinforce the column using rock bolts, revealing nonlinear dynamics that provide insights into the internal cracking of rock structures.

Analysis of the measurements (Fig. 3) highlights the rapid nonlinear dynamics (anomalous nonlinear fast dynamics) observed at the beginning of drilling, as well as a clear correlation between the acceleration at the top of the column and decreases in resonance frequency, of the order of a few percent. Slow dynamics were also observed, manifested by a reproducible recovery of the fundamental resonance frequency after drilling was stopped. This recovery was also observed following seismic shaking generated by regional earthquakes. The magnitude of the nonlinear elastic effects observed in situ exceeds those measured in laboratory experiments on intact rock, revealing extensive, multiscale cracking within the studied rock column. These results show that passive seismic surveys conducted during periods of significant vibration can be used to probe material nonlinearity at the geophysical scale. This approach could have potential applications in geotechnical engineering and civil engineering, as well as in monitoring internal cracking within rock structures.
Fig. 2. View of an unstable 760 m³ limestone column undergoing reinforcement works, showing the location of the seismological sensor at the top of the column (S). The reinforcing rock bolts (5 to 12 m long) are shown as white dots topped with crosses. Adapted from [Bottelin et al., 2026a].


The full study is published in: NATURE


References :

Bottelin, P., Baillet, L., Larose, E., Guillemot, A. & Johnson, P. A. Nonlinear mesoscopic elasticity revealed by passive seismic monitoring of a rock column during drilling operations. J. Acoust. Soc. Am. 159, 2844–2856 (2026).

Bottelin, P., Baillet, L. & Larose, E. On the use of seismic noise studies for landslide rock block hazard assessment. Commun. Earth Environ. 7, 769 (2026).

Pazzi, V., Fornasari, S. F., Devoto, S., Costa, G. & Forte, E. Fast estimation of landslide blocks’ volume from seismic noise measurements. Commun. Earth Environ. (2025).

Scientific contacts :

  • Pierre Bottelin – ISTerre, Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, IRD, Univ. Gustave Eiffel, Grenoble, France
  • Laurent Baillet – ISTerre, Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, IRD, Univ. Gustave Eiffel, Grenoble, France
  • Eric Larose – ISTerre, Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, IRD, Univ. Gustave Eiffel, Grenoble, France