Towards dynamically constraining subduction zone parameters from surface-topography characteristics

Abstract:

The sticky-air method allows global models of mantle convection calculated on an Eulerian grid to realistically reproduce a free surface that is both physically accurate and computationally sensible (Matsumoto and Tomoda, 1983; Schmeling et al., 2008; Crameri et al., 2012a). Being able to model both accurate surface topography and, thanks to the free surface, realistic subduction (Crameri et al., 2012b), we are able to thoroughly investigate the link between individual topographic signals and subduction parameters.

For this, we use global mantle models with a free surface that are calculated by the finite-volume code StagYY (e.g., Tackley 2008) using a multi-grid method on a fully staggered grid. We apply the sticky-air method with carefully chosen parameters (Crameri et al., 2012b) to approximate the free surface. A weak hydrated crustal layer ensures stable, on-going subduction (Crameri and Tackley, 2015). The dynamically fully self-consistent model is calculated in either a 2-D or 3-D Cartesian geometry and has an initial subduction zone to produce plate convergence.

We systematically test the topography signal depending on (i) shallow slab-dip, (ii) slab buoyancy (iii) radial mantle viscosity, (iv) plate strength, and (v) 3-D mantle flow. Results from our detailed parameter study include general trends and highlight slab dip and mantle viscosity as major agents for upper plate deflection and the slab buoyancy for lower plate deflection.

REFERENCES

Crameri, F., et al. (2012a), Geophys. J. Int., 189(1).

Crameri, F., et al. (2012b), Geophys. Res. Lett., 39(3).

Crameri, F., and P. J. Tackley (2015), J. Geophys. Res., 120(5)

Matsumoto, T., and Y. Tomoda (1983), J. Phys. Earth, 31(3).

Schmeling, H., et al. (2008), Phys. Earth Planet. Int., 171(1-4).

Tackley, P. J. (2008), Phys. Earth Planet. Int., 171(1-4).