Clustering regimes in a sandpile with targeted triggering

Physics Department, College of Science, De La Salle University - 2401 Taft Avenue, Manila, 0922, Philippines

^(a) rene.batac@dlsu.edu.ph (corresponding author)

Received: 24 November 2020
Accepted: 11 June 2021

Abstract

The sandpile is one of the earliest discrete models to have demonstrated the self-organisation believed to be at work in earthquakes and other nonlinear systems in nature. However, sandpile-based approaches are less utilised for modelling seismicity, due to the cascading behaviour of avalanches, which, in contrast to the bursty correlated earthquake dynamics, release all the accumulated energy in a single event. Here, the sandpile model is modified by introducing a preferential targeting of the most susceptible grid site, allowing it to generate simulated sequences that are statistically similar to those of earthquake events in space, time, and energy domains. In particular, the proposed model creates generalised space-time-magnitude “distance” between events akin to real aftershock sequences, and a comparison with real-world data from a catalogue of recent earthquakes is presented. The underlying mechanisms of earthquake generation are clearly very much complex, but discrete models such as this one provide easily reproducible results for statistical comparisons.