A method for locating rockfall impacts using signals recorded by a microseismic network

Abstract

Background

Rockfall events are one of the most dangerous phenomena that often cause several damages both to people and facilities. During recent years, the scientific community focused the attention at evaluating the effectiveness of seismological methods in monitoring these phenomena. In this work, we present a quick and practical method to locate the rebounds of some man-induced boulders falls from a landslides crown located in the Northern Apennines (Central Italy). The reconstruction of the trajectories was obtained by means of back analysis performed through a Matlab code that takes into account both the DEM (Digital Elevation Model) of the ground, the geotechnical-geophysical characteristics of the slope and the arrival times of the seismic signals generated by the rock impacts on the ground.

Results

The localization results have been compared with GPS coordinates of the points and videos footage acquired during the simulations, in order to assess the reliability of the method. In most cases, the retrieved impact points match with the real trajectories, showing a high reliability. Furthermore, four different cases have been identified as a function of the geomechanical, geophysical and morphological conditions. Due to the latter ones, in some case it was necessary to assume different values for the propagation velocity of the elastic waves in the ground, here assumed to be isotropic and homogeneous.

Conclusions

This work aims at evaluating the effectiveness of a quick and practical method to locate rockfall events using a small-aperture seismic network. The obtained results indicate that the technique can provide quantitative information about the area most prone to impact of detached blocks. The method still presents some uncertainty, but reducing some of the approximations (e.g. by better constraining the velocity model), it could lead to prompt and more accurate results, easily applicable to hazard estimates.