Thermal softening in the ductile and brittle lithosphere

Plate tectonics on Earth is characterized by a largely localized distribution of lithospheric strain, both in the ductile and in the brittle regimes. While deformation in brittle or elastic materials is generally localized, in homogenous ductile or viscous materials some strain localization requires a softening mechanism. Here we will focus on thermal softening, a consequence of the conversion of mechanical work into heat (i.e. shear heating) and the temperature dependence of rock viscosities.
First, I briefly list the fundamental features of ductile shear zone evolution due to thermal softening driven by steady-state background deformation. (1) After an initial transient period, the maximum temperature and stress in the shear zone converges to a (quasi-)constant value (temperature increase and stress drop). (2) The steady-state maximum temperature can be estimated using a scaling law. (3) With ongoing deformation, the high-temperature zone and consequently the shear zone widen, which indicates that shear zone width is controlled by conduction time scales.
Second, I demonstrate that thermal softening is a feasible mechanism of lithospheric scale ductile strain localization by comparing geological observations and model results of subduction initiation, ophiolite emplacement, and high-temperature metamorphic nappes.
Finally, I will investigate the possible occurrence and importance of thermal softening in the brittle, elastoplastic domain, often associated with much shorter time scales such as slow slip events and earthquakes. Data from rock deformation experiments indicate that steady-state friction angle becomes primarily velocity-dependent at high slip rates, which is consistent with thermal softening. Numerical models of Maxwell visco-elastic deformation show that thermal softening can be an efficient mechanism of limiting elastic stress build-up, often resulting in a rapid stress release, often referred to as thermal runaway.