Earthquake rupture in a strike slip experiment

Seismotectonic analogue models provide a valuable complement to seismological, geodetic and paleoseismological-geomorphological approaches for investigating the earthquake cycle. Physical models mainly made with elastic and frictional materials, allow the simulation of multiple seismic cycles under controlled laboratory conditions. Such physical models can reproduce interseismic, coseismic and postseismic deformation. However, all those models include a pre-existing fault, and thus do not necessarily model realistic fault geometries. As a consequence, the influence of geometric complexity—such as fault segmentation, bends, and step-overs—on rupture dynamics and seismic cycle behaviour remains poorly explored in those seismotectonic models.

Here, we present a new analogue seismotectonic model of a strike-slip fault system that allows complex fault geometries to emerge and evolve while producing multiple seismic cycle. The experimental setup consists of two juxtaposed horizontal PVC plates separated by a straight velocity discontinuity, with one plate fixed and the other moving at constant velocity, simulating a vertical basement fault. The overlying medium is composed of three granular layers:  a basal layer of rubber pellets that stores elastic strain, an intermediate rice layer that exhibits stick–slip behavior and represents the seismogenic crust, and an upper frictionally stable sand layer mimicking a non-seismogenic shallow crust.

Surface deformation is monitored with photographs acquired every 2 seconds, corresponding to 25 μm of displacement for the basal plate , and processed using image correlation and dense optical flow methods. Seismic events are detected when surface displacement exceed the imposed basal plate displacement. In addition, recordings made with a high-speed camera at 100 frames per second capture transient surface deformation during rupture propagation. A total of 23 high-speed sequences, each lasting 10 seconds, document coseismic surface deformation associated with earthquake propagation.

We explore the potential of this experimental setup to investigate how rupture characteristics—such as rupture velocity, nucleation and arrest processes— may depend on the evolution of fault geometry and associated off-fault deformation. By quantifying the spatiotemporal distribution of surface deformation and seismic events along evolving fault networks, this approach allows us to investigate how fault segments are activated, temporarily locked, or interact throughout successive stages of the seismic cycle. Moreover, we examine how interseismic deformation reflects the evolving mechanical state and geometry of the fault system, and how this state influences subsequent earthquakes.