Modelling Antarctic ice shelf, ocean, and sea ice interactions under present-day and future climate scenarios
This thesis uses two ocean/sea-ice/ice-shelf models to advance our understanding of physical processes in Antarctic ice shelf cavities and over the surrounding continental shelf, and to project how these environments may change in the future. A new circumpolar Antarctic configuration of the MetROMS...
| Main Author: | |
|---|---|
| Format: | Doctoral or Postdoctoral Thesis |
| Language: | English |
| Published: |
UNSW, Sydney
2018
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/1959.4/60073 https://unsworks.unsw.edu.au/bitstreams/4bbf801d-d4a8-4a2e-b2f3-1566b2f868b1/download https://doi.org/10.26190/unsworks/20546 |
| Summary: | This thesis uses two ocean/sea-ice/ice-shelf models to advance our understanding of physical processes in Antarctic ice shelf cavities and over the surrounding continental shelf, and to project how these environments may change in the future. A new circumpolar Antarctic configuration of the MetROMS model (ROMS: Regional Ocean Modelling System coupled to CICE: Community Ice CodE) is developed, and is compared to the Finite Element Sea-ice/ice-shelf Ocean Model (FESOM). Future projections of ice shelf basal melt rates through the 21st century are also produced using FESOM. Both models exhibit reasonable agreement with available observations, but share many of the same biases, such as an underestimation of ice shelf melt rates in the Amundsen and Bellingshausen Seas, insufficient summer sea ice cover, and weakening transport of the Antarctic Circumpolar Current. The major limitations on model performance appear to be (i) excessive smoothing of the topography in FESOM, which is necessary to ensure numerical stability; (ii) spurious diapycnal mixing inherent in the terrain-following coordinates of MetROMS; and (iii) potential biases in the atmospheric reanalysis used to force the models. The development process of the new MetROMS configuration is chronicled, and particular attention is given to a mechanism of numerical error that was found to be producing excessive sea ice. In particular, oscillatory ocean tracer advection schemes cause spurious supercooling and consequently sea ice formation, leading to a range of dynamic and thermodynamic impacts which degrade the simulation. Careful choice of advection schemes, increased parameterised diffusion, or the application of flux limiters can avoid this problem. Future projections with FESOM, under four 21st-century atmospheric forcing scenarios, all exhibit increased ice shelf basal melting in every sector of Antarctica. Total ice shelf basal mass loss from the continent increases by between 41% and 129%. The main mechanism of melting is an increased presence of warm ... |
|---|