dr. Andrea Manconi, Swiss Federal Institute of Technology, Department of Earth Sciences, Zurich, Switzerland - ore 11.30 - aula 1G09 - Dipartimento di Fisica "Ettore Pancini", Complesso Universitario di Monte S.Angelo - Via Cintiia, Napoli.
ABSTRACT
We propose a new approach to detect, locate, and characterize rockslides in real-time, by considering waveforms acquired from broadband regional seismic networks, and adapting algorithms originally developed for earthquake early warning. The discrimination of signals generated by rockslides from other sources, such as natural and/or induced earthquakes, is performed by exploiting the ratio between local magnitudes (ML) and duration magnitudes (MD). Remarkably, we found that signals associated to rockslides show ML/MD<0.8, while by definition for earthquakes is ML/MD ≅ 1. In addition, we derived an empirical law between MD and rockslide volumes that allows obtaining a preliminary characterization of rockslide volumes within seconds from its occurrence. The key-points of this study are presented by testing the hypothesis on a recent rockslide event occurred in northern Italy, and discussing the potential evolution of the methodology for early warning and/or fast response purposes.