Variation in Rayleigh wave ellipticity as a possible indicator of earthflow mobility: a case study of Sobradinho landslide compared with pile load testing

Abstract

Rainfall-induced landslides pose a significant risk to communities and infrastructures. To improve the prediction of such events, it is imperative to adequately investigate the rainfall-dependent dynamics (leading to fluidization) and any associated internal sliding along shear planes within clayey slopes. Therefore, the present study adopted ambient noise analysis based on the Horizontal to Vertical Spectral Ratio (HVSR) method, to measure the seasonal variation of Rayleigh wave ellipticity as an indicator for the internal deformation and transition in the material state. The methodology was applied to an existing landslide, where variations in soil stiffness and internal sliding were expected to occur in response to rainfall. To improve the interpretation of the HVSR results (and hence the prediction of landslide' reactivation by rainfall), HVSR measurements were also conducted on a field-scale pile load test. The pile test allowed a comparison of the seismic data generated by the soil movement along shear planes. The HVSR curves of this field test showed two frequency peaks with no changes in the resonance. In comparison with the data obtained from the landslide, the resultant HVSR curves showed three frequency patterns: ubiquitous (2Hz), landslide (4-8Hz), and flat (no peak). However, the HVSR curves did not show any response to the expected seasonally induced variations in the landslide mass because of the relatively short data acquisition. Nevertheless, time-lapse HVSR is a promising technique that can complement other geophysical methods for improving landslide monitoring.