A framework for temporal and spatial rockfall early warning using micro-seismic monitoring

Rockfall risk is usually characterized by a high frequency of occurrence, difficulty in prediction (given high velocity, lack of noticeable forerunners, abrupt collapse, and complex mechanism), and a relatively high potential vulnerability, especially against people and communication routes. Considering that larger rockfalls and rockslides are generally anticipated by an increased occurrence of events, in this study, a framework based on microseismic monitoring is introduced for a temporal and spatial rockfall early warning. This approach is realized through the detection, classification, and localization of all the rockfalls recorded during a 6-month-long microseismic monitoring performed in a limestone quarry in central Italy. Then, in order to provide a temporal warning, an observable quantity of accumulated energy, associated to the rockfall rolling and bouncing and function of the number and volume of events in a certain time window, has been defined. This concept is based on the material failure method developed by Fukuzono-Voight. As soon as the first predicted time of failure and relative warning time are declared, all the rockfalls occurred in a previous time window can be located in a topographic map to find the rockfall susceptible area and thus to complement the warning with spatial information. This methodology has been successfully validated in an ex post analysis performed in the aforementioned quarry, where a large rockfall was forecasted with a lead time of 3 min. This framework provides a novel way for rockfall spatiotemporal early warning, and it could be helpful for activating traffic lights and closing mountain roads or other transportation lines using the knowledge of the time and location of a failure. Since this approach is not based on the detection of the triggering events (like for early warnings based on rainfall thresholds), it can be used also for earthquake-induced failures.

Automated summary

This study introduces a framework for temporal and spatial rockfall early warning based on microseismic monitoring. By detecting and classifying rockfall events, the method provides warnings based on accumulated energy, as demonstrated by successfully predicting a large rockfall 3 minutes in advance in an experiment.