Transient deformation of the upper mantle from the crystal to the plate scales

Viscous deformation of the upper mantle controls surface motions following melting of large ice masses and during the seismic cycle; deformation that we can observe in both Global Navigation Satellite System and time-variable gravity measurements. Such deformation typically decays over time (i.e., is transient) and involves a component of time-dependent recoverable strain (i.e., is anelastic). Transient anelastic behaviour has been recognised as a key component in models of postseismic deformation and has been reported in laboratory rock deformation experiments. However, the microphysical mechanisms by which this deformation occurs have remained poorly constrained, and consequently it remains unknown how exactly geodetically observed deformation at the plate scale results from deformation at the crystal scale. Recently, we conducted innovative deformation experiments resulting in a new flow law incorporating the microphysics of anelasticity in olivine, the main constituent of the upper mantle. We will apply this novel flow law to geodynamic problems where geodetic measurements capture transient deformation: glacial isostatic adjustment due to recent ice mass changes in Antarctica and Greenland, and postseismic relaxation after megathrust earthquakes. 

Recently, several studies have observed unexpectedly fast uplift due to recent/ongoing ice melt in Antarctica or Greenland, attributing these fast uplifts to steady-state creep with anomalously low asthenosphere viscosity. We hypothesise that anelastic creep provides a more likely and general explanation for rapid uplift. For the extrapolation of uplift rates, it makes a large difference whether steady-state or anelastic creep is responsible for the observed surface motions, which in turn affects the future stability of ice sheets close to the grounding line. Concerning postseismic relaxation, the inclusion of the new anelastic creep laws in models allows us to study the temporal variability of a stress-dependent mantle viscosity, shedding new light on the problem of discriminating between distributed and localized deformation.