Influence of pre-existing fabrics on fault kinematics and rift geometry of interacting segments: Analogue models based on the Albertine Rift (Uganda), Western Branch-East African Rift System

This study aims at showing how far pre-existing crustal weaknesses left behind by Proterozoic mobile belts, that pass around cratonic Archean shields (Tanzania Craton to the southeast and Congo Craton to the northwest), control the geometry of the Albertine Rift. Focus is laid on the development of the Lake Albert and Lake Edward/George sub-segments and between them the greatly uplifted Rwenzori Mountains, a horst block located within the rift and whose highest peak rises to >5000 m above mean sea level. In particular we study how the southward propagating Lake Albert sub-segment to the north interacts with the northward propagating Lake Edward/George sub-segment south of it, and how this interaction produces the structures and geometry observed in this section of the western branch of the East African Rift, especially within and around the Rwenzori horst. We simulate behaviour of the upper crust by conducting sandbox analogue experiments in which pre-cut rubber strips of varying overstep/overlap connected to a basal sheet and oriented oblique and/or orthogonal to the extension vector, are placed below the sand-pack. The points of connection present velocity discontinuities to localise deformation, while the rubber strips represent ductile domain affected by older mobile belts. From fault geometry of developing rift segments in plan view and section cuts, we study kinematics resulting from a given set of boundary conditions, and results are compared with the natural scenario. Three different basal model-configurations are used to simulate two parallel rifts that propagate towards each other and interact. Wider overstep (model SbR3) produces an oblique transfer zone with deep grabens (max. 7.0 km) in the adjoining segments. Smaller overlap (model SbR4) ends in offset rift segments without oblique transfer faults to join the two, and produces moderately deep grabens (max. 4.6 km). When overlap doubles the overstep (model SbR5), rifts propagate sub-orthogonal to the extension direction and form shallow valleys (max. 2.9 km). Relative ratios of overlap/overstep between rift segments dictate the kind of transition zone that develops and whether or not a block (like the Rwenzoris) is captured and rotates; hence determining the end-member geometry. Rotation direction is controlled by pre-existing fabrics. Fault orientation, fault kinematics, and block rotation (once in play) reinforce each other; and depending on the local kinematics, different parts of a captured block may rotate with variable velocities but in the same general direction. Mechanical strength anisotropy of pre-structured crust only initially centres fault nucleation and propagation parallel to the grain of weakness of the basement, but at later stages of a protracted period of crustal extension, such boundaries are locally defied.

Research highlights
► Fault orientation and kinematics change cause block rotation and vice versa; the three reinforce each other and are controlled by pre-existing fabrics.
► Extending pre-structured crust initially centres fault locus parallel to weakness grain but at later stages such boundaries are defied.
► Relative overlap/overstep ratio between interacting rift segments dictates end-member transition zone geometry, presence of captured block, block rotation, etc.
► A rotation-dominated model approximates to the scenario in the Rwenzoris – block capture, varying rotation rates for different parts of the block.
► Rotation of the Rwenzoris plays is significant for (re-)activation of transfer faults.