Skip to Main content Skip to Navigation
New interface

Surveillance sismologique des glaciers rocheux : pour mieux contraindre un risque émergent

Abstract : Within mountain permafrost regions, rock glaciers refer to periglacial formations composed of a mixture of rock debris of varying size and frozen materials at depth. Among them, active rock glaciers exhibit surface displacement velocities of the order of m/year, reflecting deep-seated deformation dynamics governed by thermo-hydro-mechanical processes. Thus, their contribution to the sediment transfer that can be mobilized in torrential flows may be substantial, as well as the emerging natural risks linked to their destabilization. In support of kinematic monitoring, geotechnical and geophysical methods can be used to investigate a rock glacier at depth, but remain punctual and discontinuous. On the other hand, environmental seismology methods, already applied to landslide monitoring, allow continuous seismic recordings, reporting the state of the subsurface; the aim of this thesis is therefore to apply these passive seismic techniques to the monitoring of rock glaciers.Two active sites have thus been instrumented for several years: the Gugla rock glacier (Valais, Switzerland) and the Laurichard rock glacier (Hautes-Alpes, France). Several methods were applied to these seismic data: ambient noise correlation, microseismicity and spectral analysis. In general, the relative variation of the surface wave velocity (dv/v) shows significant seasonal variations: it reaches a maximum at the end of winter, before dropping during spring melt and reaching a minimum in summer. Similarly, the resonance frequencies specific to rock glaciers, that have been picked within the spectral content of the seismic ambient noise, follow the same seasonal pattern. We interpret these observations as the effect of freeze-thawing processes on the averaged stiffness of the rock glacier.In order to model the variation in the elastic parameters of the medium due to this freezing process, we used a three-phase Biot-Gassmann poroelastic model, considering the rock glacier as a solid porous matrix with pores filled with liquid water or ice. The parameters of the model were constrained by the results of active geophysical surveys carried out in summer at each of the sites. The observed variables (dv/v and resonance frequencies) have then been modelled by comparison of dispersion curves and modal analysis, respectively. The results are consistent with the observations, thus improving the quantification and localization of seasonal freezing and thawing processes within rock glaciers.Similarly, the decorrelation (Kd) of the ambient noise cross-correlations, as well as the daily variability of the spectral power density, increase sharply during melting periods. These observations make it possible to better date and locate liquid water infiltration within the porous matrix.In parallel with modelling of physical processes, an empirical approach defining alert thresholds can also be adopted, towards the integration of this seismic monitoring of rock glaciers into early warning systems.
Document type :
Complete list of metadata
Contributor : ABES STAR :  Contact
Submitted on : Monday, January 4, 2021 - 2:37:26 PM
Last modification on : Sunday, May 1, 2022 - 3:41:32 AM
Long-term archiving on: : Monday, April 5, 2021 - 8:26:40 PM


Version validated by the jury (STAR)


  • HAL Id : tel-03094893, version 1



Antoine Guillemot. Surveillance sismologique des glaciers rocheux : pour mieux contraindre un risque émergent. Glaciologie. Université Grenoble Alpes [2020-..], 2020. Français. ⟨NNT : 2020GRALU020⟩. ⟨tel-03094893⟩



Record views


Files downloads