Creep Parameterization for the 3-D Mantle: Implications for Geodetic Interpretations of GIA

Abstract:

Providing a volumetrically comprehensive characterization of the variations in the earth’s mantle rheology remains one of the grand challenges in geodynamics. An especially difficult task is to make quantitative and non-unique connection between laboratory experiments, seismological mapping and geophysical modeling of glacial isostasy and mantle flow. Advances in seismology and from theoretically modeled high pressure and temperature laboratory experiments make a new set of assessments now possible. We explore the most recent whole mantle seismic models for 3-D shear wave velocity ν_S( θ, φ, r ): S40RTS (Ritsema et al., 2011) and S362ANI+M (Moulik and Ekström, 2014) together with new data for grain boundary sliding and dislocation mechanics in the olivine/wadsleyite/ringwoodite portions of the upper mantle, connecting these to the well-constrained effective viscosity profiles known for Fennoscandia and Laurentia derived from glacial isostatic adjustment (GIA) data. The latter regions also provide information that connect the operative flow laws for perovskite (Mg,FeSiO₃) – periclase (Mg,FeO) mixtures appropriate to levels in the lower mantle corresponding to the high-to-low spin transition occurring at about 2000 km depth (Bardo et al., 2003). Three classic relations for effective viscosity, η_eff, proposed by Ranalli (1990) are employed and evaluated using the two seismic models. We discuss the implications for modeling both Würm-Wisconsin and Little Ice Age deglaciation-related rebound. In addition we discuss the parameter range η_eff, appropriate to modeling GIA in Antarctica, and provide reasonable bounds on errors in forward models. Finally, the implications for GIA removal from GRACE solutions for continent-wide ice sheet mass balance are also treated.