Bayesian inference to calibrate flow geometry in ice sheet modelling of the last Scandinavian Ice Sheet

Both the Greenland and Antarctic ice sheets are experiencing increased levels of melt, contributing to potentially devastating sea level rise. Therefore, quantifying future changes to the contemporary ice sheets is imperative to understand and mitigate the risks associated with their demise. Uncertainty within projections of ice sheets under different climate scenarios is large. Palaeo-ice sheets left behind a wealth of information on past ice extents, timing and flow directions. By looking to the past and using data to validate and constrain numerical model simulations, numerical models of present-day ice sheets can be improved, and the uncertainty within projections of ice mass loss and sea level rise can be reduced.

Here, I focus on and simulate the last Scandinavian Ice Sheet between 40 and 5 thousand years ago, as well as the surrounding ice sheets over Eurasia, to find a model input parameter space that is optimised to fit the available flow geometry. In this thesis, I present a new Bayesian framework that takes an initial perturbed parameter ensemble for the last Eurasian Ice Sheet Complex, compares it to past observed flow directions and identifies an updated parameter sampling routine on a reduced parameter space so as to improve the overall model-data match of future ensembles. To quantitatively compare and score observed flow geometry from glacial landforms to model simulations in a statistically rigorous way, a new model-data comparison tool is presented: the Likelihood of Accordant Lineations Analysis (LALA) tool. This work could be used further to develop a robust simulation of the Scandinavian Ice Sheet, as well as other palaeo-ice sheets, optimised to flow geometry and to simulate data-driven spin-ups for use in ice sheet projections.