Numerical ice sheet models and the theory that underpins them provide the most promising tools for forecasting future sea level contributions from shrinking Antarctic and Greenland ice sheets. The beds of palaeo ice sheets capture ice sheet behaviour through time far beyond the brief snapshot observable by satellite records. This makes the paleo record the ideal scenario for studying lagged and nonlinear responses in the ice sheet system that a numerical ice sheet models must capture to make robust forecasts.
The Eurasian Ice Sheet Complex with its four constituent ice sheets and marine and terrestrial components could be viewed as the ‘final exam’ for numerical ice sheet models. Within this complex the Scandinavian Ice Sheet’s land-terminating peripheral sector (in Denmark, Germany, Poland, the Baltic States and Russia) is among the most well-studied anywhere. Despite this, the growing body of hard-won ages in this land terminating periphery have proven increasingly difficult to integrate into regionally consistent models of ice sheet behaviour. This PhD thesis, makes use of digital elevation models and re-examines the landform record of the region with the aim that a new sector-scale approach may provide a geometric framework that could help integrate regional geomorphological and chronological data and observations. The detailed landform record encountered in the region was found to be surprisingly complex, fragmentary and ambiguous. It proved problematic to relate this record to existing sector-scale syntheses of prominent ice margins, or to untangle using well-established reconstruction methods that have been successful elsewhere. A new heuristic method was developed, where speculative landform mapping was conducted and subsequently, with the use of the wider regional context, evaluated for its fidelity and modified. This heuristic approach, blurring the boundaries between landform interpretation and subsequent reconstruction, was found to be necessary to make progress. The method developed here will likely serve future paleoglaciological studies in regions with complex and fragmentary records, with this method outlining an approach that although commonly applied in the field has generally not been transparently documented.
This thesis presents the first sector-wide map of ice flow and margins, compiled as flow-sets and margin-sets, that has been made by applying a consistent approach across the entire land terminating periphery. Additionally, glaciotectonic complexes were found to be pervasive throughout the study region, with these allowing the reconstruction of the deformation push direction, highlighting a complex record of tectonic shortening. This suite of glacial map products revealed a locally dynamic ice sheet and margin largely driven by instability rather than simple synchronous sector-scale responses to climate fluctuation suggested by prior investigations. Such instability provides a likely explanation to why the sector has proved difficult to reconstruct and constrain by chronological tools. Noise introduced by unstable behaviour impeded reconstruction of a generalised retreat pattern, with too many ambiguous cross-cutting relationships and spatially discontinues landform evidence. Beyond pervasive signs of unstable margins, this PhD documented signals of ice flow not radiating from the Scandinavian mountains across much of the study area, that until now was only identified locally. Two process-driven hunches, or hypotheses, are presented that could explain how these non-radial flow and margin patterns may have come about. One explores the role of large depressions in steering ice away from a strictly radial pattern. The second suggests this pattern is the function of the wider ice sheet configuration bought about by the confluence and separation of the SIS with adjacent ice sheets. The insights of a complex and oscillating margin, as well as the dynamic hunches that may explain the non radial dynamics, motivate a map which identifies a fruitful chronological sampling strategy for testing hypotheses while mitigating the noise of instability.
Motivated by difficulties in reconciling conflicting age determinations with ice dynamics at various parts of the margin, this thesis further introduces a conceptual model using a grain history lens in luminescence dating that may resolve contradictory ages at some sites.
It is anticipated that the mapped products and insights gained may serve as a useful framework for planned reconstructions of the Scandinavian Ice Sheet, and use by the wider modelling and empirical community of palaeo-glaciologists. With the transparency of the method in producing the framework presented, it hopefully stands as a useful model of ice marginal fluctuations that can be further tested and modified when new data or insights emerge.