Moving mesh finite element modeling of ocean circulation beneath ice shelves
The West Antarctic Ice Sheet drains to the ocean through a collection of large ice streams that terminate in floating ice shelves. In many places, warm ocean water intrudes beneath these ice shelves, driving ocean circulation and melting the ice, ultimately resulting in ice shelf thinning and ground...
| Main Author: | |
|---|---|
| Format: | Text |
| Language: | unknown |
| Published: |
Imperial College London
2018
|
| Subjects: | |
| Online Access: | https://dx.doi.org/10.25560/73697 http://spiral.imperial.ac.uk/handle/10044/1/73697 |
| Summary: | The West Antarctic Ice Sheet drains to the ocean through a collection of large ice streams that terminate in floating ice shelves. In many places, warm ocean water intrudes beneath these ice shelves, driving ocean circulation and melting the ice, ultimately resulting in ice shelf thinning and grounding line retreat. The importance of this ice–ocean interaction to the stability of the West Antarctic Ice Sheet has inspired a number of modelling studies of ocean circulation beneath ice shelves. These studies often use models with static and/or structured grids, which are sufficient for capturing large-scale ocean circulation with simple ice geometries, but are less well-suited for modelling sub-shelf ocean dynamics on smaller spatial and temporal scales where ice geometry is often complex and responds dynamically to melting, freezing and tidal forces. In this thesis, a moving-mesh, nonhydrostatic, finite element ocean model is developed and applied to study relatively small-scale hydro- and thermo-dynamic processes beneath ice shelves. First, the model is applied to study the near-grounding-line-region of an ice shelf, where warm deep water accesses the ice and creates a buoyant meltwater plume that rises along the underside of the shelf. Tidal forces cause the ice shelf to flex and modify the velocity of the ocean circulation which, in turn, affects melt rates. For various ocean temperatures, melt rates from the finite element model are compared to melt rates obtained using a simple 1D plume model. The plume model is found to perform well in comparison to the more complex finite element model, closely matching modeled melt rate profiles with respect to distance from the grounding line and changing ocean temperature. In order to allow for more complex, dynamic ice deformation with two-way coupling to the hydrodynamics, partial-differential-equation-based mesh movement schemes are implemented and tested using the FEniCS automated finite element framework. The newly implemented mesh movement schemes are then used in conjunction with the finite element ocean model to simulate the thermo- and hydrodynamics of ocean circulation in an idealized crevasse on the underside of an ice shelf and predict how it might be expected to evolve in time. |
|---|