A 3-D numerical model to bridge long- and short-term approaches of deformation on a strike-slip fault

A strike-slip fault is subjected to earthquakes spanning seconds to minutes, separated by periods of hundreds to thousands of years. As the fault matures and undergoes multiple seismic events, its geometry and strength evolve, hence impacting in return the course of seismic cycles. Given the variability of timescales, two approaches are generally chosen in order to model the deformation of the lithosphere. On one hand, long-term modeling looks at the tectonic evolution of deformation, and is usually quasi-static and disregards the effects of dynamic events. On the other, short-term modeling respects well earthquake mechanics, but does not account for the impact of evolution of fault geometry on seismic cycles.

Here, we construct a numerical model of a continental strike-slip fault system, in a way that can effectively bridge together the different spatio-temporal scales of lithospheric deformation, and include the mutual influence of fault maturation and earthquakes upon one another. The developed approach uses the Discrete Element Modeling (DEM) method, which is based on the discretization of the medium in a finite number of rigid, spherical particles interacting via predefined contact laws. Using this method, we build a 3-D model of a portion of the crust. Initially, the material is homogeneous and intact. Then, shearing boundary conditions are applied, leading to the spontaneous emergence of a through-going, strike-slip fault showing complexities and evolving naturally as the shearing is maintained. On this evolving strike-slip fault, unstable sliding occurs, that we identify as earthquakes.

In order to validate our model, we first compare the long-term tectonic deformation with that of previous analog and numerical experiments described in the literature, and with natural observations. Second, we assess the physical validity of the recurrence behaviour of our created fault by comparing the frequence-magnitude distribution of our events with the Gutemberg-Richter law. Finally, we also provide tools able to characterize particular events by imaging the rupture geometry, the coseismic surface deformation as well as the coseismic displacement field on the fault.