Space-time earthquake clustering: nearest-neighbor and stochastic declustering methods in comparison

Earthquakes do not occur randomly in space and time; rather, they tend to group into clusters that can be classified according to their different properties, presumably related to the specific geophysical properties of a seismic region. Two methods for detection of earthquake clusters are considered in order to take advantage of different descriptions of the seismic process and assess consistency with the obtained clusters: the former is based on "nearest-neighbor distances" between events in space-time-energy domain; the latter is a stochastic method based on a branching point process, named Epidemic-Type Aftershock-Sequence (ETAS) model, which provides different plausible clustering scenarios by simulation. Both methods allow for a robust data-driven identification of seismic clusters, and permit to disclose possible complex features in the internal structure of the identified clusters. We aim at exploring the spatio-temporal features of earthquake clusters in Northeastern Italy, an area recently affected by low-to-moderate magnitude events, despite its high seismic hazard attested by historical destructive earthquakes.