Oblique rifting under low effective stresses or orthogonal rifting from stress rotations: how do faults reactivate in southern Malawi?

In developing a fault map for the Malawi Rift, it is observed that active normal faults at its southern end follow an arcuate bend in the regional amphibolite to granulite grade Proterozoic fabrics, in common with elsewhere along the East African Rift System (EARS). Reactivation of faults across such bends means that either some faults accommodate oblique-extension under a uniform stress state, or that the stresses locally rotate so that all faults are optimally oriented for reactivation. To test these two hypotheses, we determined the reactivation potential of the faults in southern Malawi with respect to the regional stress state, which was previously derived from focal mechanism inversions. We find that faults in the SW trending Zomba Graben are unfavourably to severely misoriented to the regional orientation of the minimum principal stress (06/242), with stress ratios (s₃/s_1, where s₁and s₃ are the maximum and minimum principal stresses) <0.15 required for failure. Assuming that extension is parallel to the SW trending s₃, the angle of obliquity between extension and the rift trend normal in the Zomba Graben is 60 ± 22°. We demonstrate that reactivation of its normal faults under oblique extension requires either: (1) low frictional strengths (m <0.4), (2) high pore fluid pressures (P_f; s₃-P_f < tensile rock strength (T₀)), or (3) a location rotation of s₃ to a NW-SE trend, so that faults are well-oriented for normal reshear and thus accommodate orthogonal extension instead. If none of these conditions are met, new optimally-oriented NW-SE striking faults will form.

Given its near-uniform distribution of fault strike, and geomorphological and seismological evidence of pure normal fault slip, we suggest that the Zomba Graben is an orthogonal rift. This favours the stress rotation hypothesis. Furthermore, the oblique fault models are difficult to support as deformation experiments indicate that the high-grade metamorphic rocks of the Malawi Rift are not frictionally weak, and if s₃-P_f<T₀, the rock will hydrofracture and relieve P_f. If true, local rotations of s₃ may be linked to pre-existing fabrics, as has been proposed elsewhere along the EARS.