From intraplate weakening to plate boundary: New diagnostics to quantify rheological controls on deformation localization in a simple extension set-up at lithospheric scale

Initiation of new plate boundary can be related to a spatio-temporal evolution of an intraplate vast diffuse deformation towards a narrow highly deforming boundary. It can also occur by reactivation of an inherited weak zone. In all cases, breaking a plate requires a weakening of the lithospheric cold mantle, whose rheological parameterisation often features a « yield strength » formulation that is not clearly related to actual deformation mechanisms. On the other hand, the bulk effective viscosity for mantle rocks, either cold lithospheric mantle or the hotter asthenosphere underneath, have multiple dependencies, that may co-evolve in geodynamic settings, e.g. temperature and strain rate increase during asthenosphere upwelling associated with plate extension.

In dynamic models, the (output) pattern of deformation localization cannot be directly predicted from the (input) flow laws governing material weakening, e.g. viscosity decrease when strain-rate or temperature increase. We currently lack diagnostics to quantify which rheological dependency weakens lithosphere through time.

Using finite-element Fluidity code, we designed 2-D upper-mantle thermomechanical models of plate extension. Simulations were run for various background strain rates (associated to various horizontal velocity profiles imposed along vertical sides) and for various rheological parameterizations featuring Newtonian diffusion creep, non-Newtonian low/high temperature dislocation creep, and/or yield stress. We propose diagnostics to quantify, through space and time, the weakening efficiency associated to thermomechanical parameters (here either strain-rate, or temperature) . The weakening efficiency is defined as the temporal variation of viscosity relative to only strain rate (or only temperature), normalized to the total viscosity variation. It is used to characterize the chronological sequence and feedbacks leading to deformation localization, and compare them for different rheological parameterizations. From these diagnostics, we discuss which deformation mechanisms are activated during plate extension and thinning, and the characteristic time-scale of successful or failed localization for various rheologies. We compare especially simulations featuring an ad hoc yield strength parameterization vs. low-temperature dislocation creep.