Rheology of the aesthenosphere

The aesthenosphere, the relatively fluid region beneath the rigid lithosphere, is inferred to be the mechanical decoupling zone that allows plates to move relatively unimpeded over the surface of the Earth. Many factors may be responsible for introducing a lateral variation in aesthenosphere viscosity. An activation volume of 20 cc/mol for olivine flow will introduce as much as an order of magnitude in viscosity between 20 million year old ocean basin and 100 million year old material since the decoupling zone will be deeper and at higher pressure in the older region. This effect becomes several orders of magnitude in scale beneath stable continental shields. Variations in water content and partial melting give rise to possible lateral variations in the aesthenosphere viscosity. We have begun experimental investigations of the rheological properties of olivine at high pressure in order to address these issues. We performed deformation experiments in the D-DIA large-volume high pressure apparatus up to 9 GPa and 1873K with samples of pure olivine, varying amounts of water, and with up to 10 wt per cent of basalt. The experimental protocol varied from studies where stress and strain rate were both determined using a synchrotron x-ray probe to comparative studies where the strain rate of stacked samples were compared. All measurements are made in situ as the sample is deforming. Strains up to 50 per cent ensure steady state deformation. Stress measurements are limited in precision to about 100 MPa while strain rate measurements are accurate to about 10-7 /s. Our observations differ from lower pressure observations in the following manner. We find that the activation volume for creep for olivine is less than 5 cc/mol, the effect of water is small once the water level reaches some modest amount, the effect of partial melting is also small (less than 20 per cent) once steady state is achieved. Dynamic recrystallization is very active and probably controls the texture of the system very early in the experiment. These observations suggest that the lateral variation of the aesthenosphere viscosity is not age dependent and only weakly dependent on the effects of water and partial melting.