Rayleigh waves in seismic signals of rockfalls

Clara Lévy (1) , Agnès Helmstetter (2) , David Amitrano (2) , Gaëlle Le Roy (2, 3) , Fabrice Guyoton (3)

1 BRGM - Bureau de Recherches Géologiques et Minières (BRGM)
2 ISTerre - Institut des Sciences de la Terre
3 Géolithe Alpes

Seismic signals of rockfalls are usually very complex, as they are the result of superimposed arrivals of wave trains generated by successive impacts and have a low signal-to-noise ratio (SNR). Thus, retrieving information from these signals is usually challenging. Our objective was to apply the signal processing method proposed by Meza-Fajardo et al. (2015) in order to isolate Rayleigh waves from the rest of the signals. We expect to retrieve: 1) a simplified output signal as only Rayleigh waves trains are extracted, 2) the azimuth of incoming Rayleigh waves (e.g., information on the rockfall localisation), 3) an output signal with a better SNR than the original signal as the decrease of energy due to geometrical attenuation is proportional to 1/r (with r the distance of propagation) for Rayleigh waves, and is proportional to 1/r² for body waves. Our work confirms the large presence of surface waves for this type of signals (a fact widely accepted in the literature, without ever being really discussed). Future studies will determine whether the use of Rayleigh waves (rather than the entire seismic signal) improves the localization of rockfalls, especially when using a method specific for surface waves (the Fast Marching Method).

Keywords: signal processing; Rayleigh waves; rockfall impacts; S-transform

Domains: Earth Sciences; Applied geology; Geophysics [physics.geo-ph]

Clara Lévy, Agnès Helmstetter, David Amitrano, Gaëlle Le Roy, Fabrice Guyoton. Rayleigh waves in seismic signals of rockfalls. 4th RSS Rock Slope Stability Symposium, Nov 2018, Chambéry, France. ⟨hal-01815453⟩