Characterization of the Vajont landslide (North-Eastern Italy) by means of reflection and surface wave seismics

• We present the use of both reflection and surface waves for landslide study.
• We test the robustness of comparing borehole data and lab data.
• We propose an interpretation of the Vajont landslide body basing on the prospecting results.

The mechanisms of the disastrous Vajont rockslide (North-Eastern Italy, October 9, 1963) have been studied in great detail over the past five decades. Nevertheless, the reconstruction of the rockslide dynamics still presents several uncertainties, including those related to the accurate estimation of the actual landslide mass. This work presents the results of a geophysical characterization of the Vajont landslide body in terms of material properties and buried geometry. Both aspects add new information to the existing dataset and will help a better understanding of the rockslide failure mechanisms and dynamics. In addition, some general considerations concerning the intricacies of landslide characterization can be drawn, with due attention to potential pitfalls. The employed techniques are: (i) high resolution P-wave reflection, (ii) high resolution SH-wave reflection, (iii) controlled source surface wave analysis. We adopted as a seismic source a vibrator both for P waves and SH waves, using vertical and horizontal geophones respectively. For the surface wave seismic survey we used a heavy drop-weight source and low frequency receivers. Despite the high noise level caused by the fractured conditions of the large rock body, a common situation in landslide studies, we managed to achieve a satisfying imaging quality of the landslide structure thanks to the large number of active channels, the short receiver interval and the test of appropriate seismic sources. The joint use of different seismic techniques help focus the investigation on the rock mass mechanical properties. Results are in good agreement with the available borehole data, the geological sections and the mechanical properties of the rockmass estimated by other studies. In general the proposed approach is likely to be applicable successfully to similar situations where scattering and other noise sources are a typical bottleneck to geophysical data acquisition on landslide bodies.