Identification of earthquake clusters through a new space-time-magnitude metric

We consider a metric that exploits the statistical properties of seismicity to quantify the correlation between earthquakes in a given catalogue. The method is based on nearest-neighbour distance between pairs of events in a combined space-time-magnitude domain and allows us to identify and analyse seismic clusters. We exemplify results from selected major earthquakes (i.e. Colfiorito 1997, L'Aquila 2009 and Emilia 2012), showing that the method can deal with data of different quality. Moreover, we show that this data-driven approach permits to disclose possible correlations and complex features in the internal structure of the identified clusters.