Long-Term Clustering, Scaling, and Universality in the Temporal Occurrence of Earthquakes

TL;DR

This paper uncovers universal statistical patterns in earthquake timing, showing that earthquake occurrences follow a consistent probability distribution across regions and magnitudes, revealing long-term clustering and self-similarity in seismic activity.

Contribution

It demonstrates the universality and self-similarity of earthquake occurrence distributions across different regions and magnitudes, highlighting long-term clustering beyond aftershock sequences.

Findings

Earthquake inter-event times follow a universal probability distribution.

Seismic activity exhibits self-similarity when time is rescaled by average activity.

Clustering of earthquakes persists beyond immediate aftershock sequences.

Abstract

Scaling analysis reveals striking regularities in earthquake occurrence. The time between any one earthquake and that following it is random, but it is described by the same universal-probability distribution for any spatial region and magnitude range considered. When time is expressed in rescaled units, set by the averaged seismic activity, the self-similar nature of the process becomes apparent. The form of the probability distribution reveals that earthquakes tend to cluster in time, beyond the duration of aftershock sequences. Furthermore, if aftershock sequences are analysed in an analogous way, yet taking into account the fact that seismic activity is not constant but decays in time, the same universal distribution is found for the rescaled time between events.