Abstract

Earthquakes are a destructive natural phenomenon with high societal costs and impact. A better understanding and characterisation of their underlying governing equations can help to better estimate the hazard associated with them. Slow slip events (SSEs) show many similarities with regular earthquakes but are characterised by a much shorter recurrence time (months/years instead of decades/millennia), offering us the possibility to study multiple cycles. Being able to model their dynamics can answer questions concerning the complexity of the frictional failure phenomenon at natural scale and its predictability. Studying slow earthquakes in nature and in the lab, I will show how a system of Stochastic Differential Equations can help us better characterise the seismic cycle, offering an alternative way to the classical two end members (purely deterministic or purely stochastic) used to describe seismicity. Blending the deterministic and stochastic approaches shows that friction is highly sensitive to small perturbations, suggesting that the macroscopic dynamics is influenced by small scale interactions. The so-called fast degrees of freedom active at the small spatio-temporal scales can be taken into account via a stochastic framework, while the underlying low-dimensional deterministic dynamics is used as a support to describe the evolution of the system.