Changing ice flow geometries of the British and Irish Ice Sheet: Applications of model-data comparisons

• Main Author: Veness, Remy Lawrence John
• Format: Thesis
• Language: unknown
• Published: 2023
• Subjects: Ailsa; Ailsa Craig; Ice Sheet
• Online Access: https://etheses.whiterose.ac.uk/34149/

Over time, changes in climate and subglacial processes result in shifts in ice sheet flow direction and extent, collectively referred to here as ice sheet geometry. For palaeo-ice sheets, changes in flow geometry are recorded in the form of glacial landforms and sediments, and can be simulated using numerical models. However, the modelled and observed flow geometries of palaeo-ice sheets have rarely been compared. In this thesis, I combine modelled ice flow geometry with observations of glacial landforms and sediments of the last British Irish Ice Sheet, to gain insight into model fidelity, the conditions of drumlin formation, and erratic transport. In Chapter 3, I use observations of ice flow direction and indicators of ice extent to evaluate an ensemble of ice sheet model simulations, curating a list of ‘not ruled out yet’ simulations. In Chapter 4, I use the ‘best-fit’ simulation as an ‘observational environment’ to investigate conditions associated with drumlin formation. I compare 85,834 drumlins against 15,000 years of modelled ice sheet conditions, observing an association between drumlins and a distinct velocity (>50-100 to <350 ma-1) and thickness ranges. I interpret these as the conditions of drumlin formation. In Chapter 5, I develop and implement a simple flowline model of erratic transport which I test by reconstructing the dispersal of 2295 erratics from six sources (Shap, Ailsa Craig, Galway Granites, Glen Fyne, Cheviots and the Sperrins). I find the best-fit model matches 82% of erratics, and that the transport model can be used to assess model fidelity with regards to ice flow geometry. Overall, this thesis contributes to the growing field of palaeo-ice sheet model-data comparison, demonstrating both how data can help constrain models, and models can help us understand processes.