Constraints on Mantle Viscosity From Slab Dynamics

The radial viscosity of the mantle is generally thought to increase by similar to 10-100 times from the upper to lower mantle with a putative, abrupt increase at 660 km depth. Recently, a low viscosity channel (LVC) between 660 and 1,000 km has been suggested. We conduct a series of time-dependent flow models with viscosity either increasing or decreasing at 660 km depth while tracking slab structure, state-of-stress, and geoid. We find that a LVC will lower the amplitude of long wavelength (>5,000 km) geoid highs over slabs, with amplitudes <10 m in height, while increasing the slab dip angle and downdip tension in the upper 300 km of slabs. A viscosity increase at 660 km gives rise to strong downdip compression throughout a slab and this pattern will largely go away with the introduction of the LVC. In addition, the endothermic phase change at 660 km depth can substantially affect the stress distribution within slabs but has a minor influence on the geoid. Models that fit the observed long wavelength geoid and observed focal mechanism in the western Pacific favor models without the presence of the LVC between 660 km and 1,000 km depths.

Automated summary

The study indicates that the presence of a low viscosity channel (LVC) in the mantle will reduce the amplitude of long wavelength geoid highs over slabs, while affecting the dip angle and tension distribution of slabs. A viscosity increase at 660 km depth leads to strong downdip compression within slabs, a pattern that largely disappears with the introduction of the LVC.