Can ridge axial deformation constrain mantle viscosity?

E. Choi¹* and W.R. Buck¹

Corresponding author: echoi@ldeo.columbia.edu
¹Lamont-Doherty Earth Observatory, P.O. Box 1000, 61 Rt.9W, Palisades, NY 10964

Abstract:
The range of possible sub-ridge mantle viscosities is uncertain to several orders of magnitude, depending particularly on its hydration state. Some models for melt migration require the sub-ridge viscosity to be quite high. We investigate whether ridge axial deformation and morphology can provide a constraint on the upper mantle viscosity. The width of tectonic deformation zone indicates the relative importance of diking and tectonic deformation for accommodating plate motions. If dike opening accommodates the entire spreading motion, no tectonic deformation would be expected. On the contrary, significant tectonic deformations would occur if the amount of dike opening is much less than the total divergent plate motions. Axial morphologies have also been shown to reflect the balance between diking and tectonic deformation. The major objective of our work is to use improved numerical approaches to learn how magmatic and tectonic processes interact to form the observed variety of mid-ocean ridges. Tectonic deformation can be simulated in a way that is internally consistent and which allows for simulation of model fault zones with some realistic properties. Although we have to specify the location and time of dike opening, we have developed methods to treat the amount of dike opening in a mechanically consistent way. Preliminary results indicate that very large mantle viscosities are not consistent with axial high morphology and deformation pattern at fast spreading ridges.

Keywords:
mantle viscosity, axial morphology, diking, numerical models.

Contributions to Integration and Synthesis:
Partial melting and dehydration have opposite effects on the mantle viscosity beneath the spreading centers. Partial melting and hydration reduces viscosity of olivine, which can be stronger by several orders of magnitude when dry. On the contrary, the dehydration due to the high solubility of water in the melt relative to olivine has been suggested as a viable mechanism for increasing mantle viscosity. Implications of partial melting and associated dehydration have been studied in various geological, geochemical and geodynamic contexts. However, the idea of highly viscous upper mantle due to dehydration has not been tested against the observed widths of tectonic deformation zones or the axial morphologies, for which mantle viscosity might have a first order control. The effect of high viscosity mantle on spreading center processes should be most evident at the fastest spreading ridges. Thus, the deformation pattern of EPR ISS can be used to constrain the mantle viscosity. Our study will also contribute to synthesize all of these spreading center processes in an internally consistent model.

Figures:
Figure 1. Preliminary results showing that deformation pattern for a fast-spreading ridge is not consistent with the observed highly focused one when the mantle viscosity is assumed to be high (≥ 10²⁰ Pa · s). (a) The domain is 100X20 km and the element size is 500 m. The full spreading rate is 10 cm/yr and the plate age varies from 0.01 My at the spreading center to 1.01 My at the boundaries. Plastic strain field developed in the models after 1 km of extension for different mantle viscosities: (b) 10¹⁹, (c) 10²⁰, and (d) 10²¹ Pa · s. Choi_fig1.jpg