Ambient noise analysis for expeditious evaluations of slope susceptibility to co-seismic failures: potential and limitations

During an earthquake, mountainous and hilly areas can suffer severe additional damage as an effect of co-seismic landsliding favored by the presence of impedance contrasts (e.g. caused by soft slope materials resting on a rigid substratum) that cause seismic shaking amplifications. As a tool to guide actions for the mitigation of such effects, we are testing an expeditious approach to make a rapid reconnaissance of slopes susceptible to seismically induced landsliding. Our approach is based on the estimation of the slope resistance demand posed by seismic shaking with an expected Arias intensity. We exploit expeditious ambient noise analyses to assess the contribution of the dynamic response of slopes to their susceptibility to seismic failure. A technique of instantaneous polarization analysis is used to extract Rayleigh waves from noise recordings with the consequent possibility of (i) inverting the Rayleigh wave ellipticity curves as a function of frequency in terms of S-wave velocity vertical profiles and (ii) calculating the amplification factors in terms of Arias intensity through 1D site response modeling.

To verify the effectiveness of our approach, we use data from a local network of accelerometer stations installed on marginally stable slopes to compare amplification factors estimated from noise recordings with those derived from accelerometer recordings. The results show differences in the amplification factor estimates within 50%, an acceptable level of approximation for a preliminary regional-scale assessment of the slope dynamic response. However, the following limitations in the applicability of the approach should be considered: 1) a careful selection of the seismic inputs for the numerical modeling, with a well-balanced distribution of spectral energy, to avoid strong concentrations over limited frequency bands that may result in a significant underestimation or overestimation of the amplification factor; 2) particular attention to complex slope settings (e.g., the presence of dominant fracturing/fissuring systems resulting in strong anisotropies in the mechanical properties of the slope materials); in such cases, the amplification estimates could be greatly underestimated (up to 300%) by assuming the occurrence of purely stratigraphic site effects.