Plate-boundary maintenance – formation, preservation, and reactivation in global plate-like mantle convection models

   Plate boundary dynamics remain incompletely understood in the context of thermo-chemical convection. Strain-localization is affected by weakening in ductile shear zones, and a change from dislocation to diffusion creep caused by grain-size reduction is one of the mechanisms that has been discussed. However, the causes and consequences of strain localization remain debated, even though tectonic inheritance and strain localization appear to be critical features in plate tectonics.

   Frictional-plastic faults in nature and brittle shear zones in the lithosphere may be weakened by high transient, or static, fluid pressures, or mechanically by gouge, or mineral transformations. Weakening in ductile shear zones in the viscous domain may be governed by a change from dislocation to diffusion creep caused by grain-size reduction. In mechanical models, strain weakening and localization in the shallow parts of the lithosphere has mainly been modeled by an approximation of brittle behavior using a pseudo visco plastic rheology in combination with a linear decrease of the yield strength of the lithosphere with increasing deformation (plastic-strain (PSS) softening). Strain weakening in viscous shear zones, on the other hand, may be described by a linear dependence of the effective viscosity on the accumulated deformation (viscous-strain (VSS) softening). These different types of strain weakening are further explored and compared to the predictions from different laboratory-based models of grain-size evolution for a range of temperatures and a step-like variation of total strain rate with time. Such a parameterized, apparent-strain, or “damage”, dependent weakening (SDW) rheology (mainly PSS) can successfully mimic more complex weakening processes in global mantle convection computations. While we focus on GSS rheology to constrain the parameters of SDW, the analysis is not limited to grain-size evolution as the only possible microphysical mechanism.

   The SDW weakening rheology allows for memory of deformation, which weakens the fault zone as well as the lithosphere for a longer period and allows for a self-consistent formation and reactivation of inherited weak zones. In addition, the memory effect and weakening of the fault zone allows for a more frequent reactivation at smaller strain rates, depending on the strain-weakening parameter combination. Reactivation within the models occurs in two different ways: a), as a guide for laterally propagating convergent and divergent plate boundaries, and b), formation of a new subduction zone by reactivation of weak zones. A longer rheological memory results in a decrease in the dominant period of the reorganization of plates due to less frequently formed new plate boundaries. In addition, the low frequency content of velocity and heat transport spectra decreases with a decreasing dominant period. This indicates a more sluggish reorganization of plates due to the weaker and thus more persistent active plate boundaries. These results show the importance of a rheological memory for the reorganization of plates, potentially even for the Wilson cycle.