Reactivation of inherited faults in rift basins: insight from analogue modeling

Continental rifts accommodate shallow extensional stresses both by brittle deformation (normal faulting) and volcanism (i.e. dykes and lava flows). Lava flows, together with clastic sedimentation reshape the topography of the rift floor, forming fresh new layers of rock that cover ancient faults. Therefore, the influence of the inherited buried faults on the development of the new faults and the processes of linkage at depth between them remain difficult to investigate. Here we use analogue brittle-ductile modeling with orthogonal extension to elucidate fault growth and reactivation modes, and then compare the results with data from natural rift systems.

In our models, deformation is produced above an elastic band placed between a fixed and a moving wall controlled by a stepper motor. A  layer of viscous material distributes the deformation within the model. On top of the viscous material we use a  layer of sand mixture to simulate the brittle properties of the upper crust. A first phase of extension develops an entire normal fault network, which is then carefully buried under a variable thickness of sand, simulating a cover of sedimentary or volcanic deposits. A second phase of extension allows us to study the mode of reactivation of the inherited faults.The progress of the deformation is tracked using top view images and digital elevation models interpolated from perspective images. At the very end of the model, cross sections cut at regular intervals show the faults at depth by overlaying coloured brittle layers. The high quality of the images allow us to map and analyze the network semi-automatically. We derive displacement/length profiles to characterize the style of fault growth and propagation mode.

Model results show the development of normal faults creating systems of fault-bounded basins, horst-graben structures and conjugated faults. The setup creates a gradient of deformation from the moving wall, where the faults nucleate first near the fixed wall. We thus observe the coexistence of faults of slightly different ages on the same model, as would occur in nature over time. The cross-section shows an upward propagation and the propagation of faults from depth to surface. The preliminary results indicate different styles of reactivation depending on the stage of fault development: reactivation according to a propagating fault mode where faults still have space at tips to develop and a constant-length fault mode where the network is already fully developed. In addition, we find that the surface overlying the inherited structures first bends, then fractures (without observable vertical displacement), and finally develops from the fracture into a proper fault before it finally propagates to connect laterally within the network. This latter growth mode is consistent with the process observed in Iceland by Braham et al. (2021). Understanding the processes of fault network inheritance holds broader applications to many areas where lava or sediments cover faults, layer after layer, such as magma rich rifts like the Eastern Africa Rift or Iceland.