Effects of nonlinear rheology and anisotropy on the relationship between age and depth at ice divides

Through numerical modelling using a full-system Stokes thermomechanical model, the effects of nonlinear rheology and strain-induced anisotropy on the age versus depth relation at ice divides are investigated. We compare our numerical results with field examples and analytical approximations commonly employed in age-depth prediction. We show that both the rheological index and strain-induced anisotropy profoundly affect the age distribution with depth, and caution must be exercised when estimating age of ice from ice cores with an isotropic age-depth model. Our main findings are: First, once the ice has developed a significant single maximum or vertical girdle fabric, the analytical approximations tend to underestimate the age of ice. Second, Bedrock topography and divide migration have a strong influence on the orientation of the ice fabric. They can force the development of single maximum and vertical girdle fabrics that are not aligned in the vertical. The orientation of the ice fabric can show sharp horizontal gradients and it has a significant effect on the age-depth relationship. We also study the coupling between anisotropic viscosity and internal heating. It does produce a warm spot and softer ice at the base of the divide when compared with surrounding areas. Finally we study the age-depth distribution in divides that show double-peaked Raymond bump in their radar stratigraphy and concavities in the surface parallel to and at both sides of the ridge. They provide ideal locations fore ice-core drilling as they have been stable for a long time when compared with their characteristic time (ice thickness divide by accumulation). Our model shows that the ice in these areas can be up to one order of magnitude older that ice at the same depth both at the flanks of the divide area or on similar divides that have not been stable for that long.