Fault Friction, Plate Rheology, and Mantle Torques From a Global Dynamic Model of Neotectonics

Improvements in software, parallel computing, global data sets, and laboratory flow-laws help to develop the global Earth5 thin-shell finite-element model of Bird et al. (2008, https://doi.org/10.1029/2007jb005460) into a benchmark study. All experiments confirm that modeled faults (other than megathrusts) have low effective friction of 0.085 ± 0.034. The average down-dip integral of shear traction on subduction megathrusts is 1.3 ± 0.2 × 1012 N/m. In plate interiors the dislocation-creep flow-law for continental crust is about twice as strong as granodiorite; close to diorite. Upper-mantle creep strength is close to olivine-rich peridotite in both oceanic lithosphere and continents. “Byerlee's law” friction of 0.85 applies between active faults in both oceanic lithosphere and stable continents. Computed net slab-pull on subducting plates is typically comparable to ridge-push, but ranges to 5× larger; slab-pull moves the plate toward the trench (6 cases) or is neutral (1 case). Under the 6 largest plates with no attached slabs, basal tractions are 0.1∼1.2 MPa. Implications for Earth mechanics include: (a) Plate-boundary faults are weakened by low-friction minerals in oceanic transform faults, temporary coseismic pore pressure elevation in many continental transforms, and permanent high pore pressure in subduction megathrusts; (b) Such weak faults are rare in plate interiors, which display Byerlee's Law friction and dislocation creep strengths matching laboratory results; (c) Generally “forward” net slab pull and basal shear tractions help gravitational potential energy to drive the plates. Implications for future modeling include: (d) Fault elements are required; (e) Laboratory flow-laws should replace Newtonian viscosity; and (f) Mohr-Anderson friction should replace isotropic plasticity.