The origin of bedrock mega-grooves in glaciated terrain

Abstract

Bedrock mega-grooves (BMGs) are large-scale bedrock corrugations reported from glaciated terrain across the world. They are remarkably straight, and closely parallel with ice-flow direction. BMGs are thought to have formed subglacially, but no consensus exists regarding their typical dimensions, formation or glaciological significance. This thesis aims to improve understanding of BMGs, focussing on their initiation and evolution, and on their relationship to bedrock geology, other subglacial landforms and ice-flow characteristics. The approach is both theoretical, through the compilation of an initial literature review and conceptualisation of several models of BMG initiation, and analytical, involving statistical analysis of morphometric data derived both from a global dataset. The data used in this study has been acquired through remote sensing as well as via fieldwork at selected locations.

Remote sensing measurements show that BMGs have lengths of 224–2269 m, widths of 21–210 m, depths of 2–15 m, elongation ratios of 5:1 to 42:1, and the spacing between adjacent individuals is 35-315 m (n = 1242 and ranges represent the 10th-90th percentiles). The variability in frequency distribution of BMG metrics between the sampled sites is likely due to variability in site geological characteristics, and to differences in palaeoglaciological history. However, all BMGs plot as a single landform population with a unimodal distribution and overlapping value ranges for all metrics, implying that they represent one landform type. BMGs and mega-scale glacial lineations (MSGLs) are similar in shape, but BMGs are on average approximately 4× shorter, 3.5× narrower, 3.5× more closely spaced and 2× deeper than MSGLs. These differences, alongside the essentially different processes inferred to be driving their evolution, support the conclusion that BMGs and MSGLs represent different types of landforms, rather than sitting on a continuum. Despite being sometimes situated in fast-flow onset zones within ice-stream landsystems, it is apparent that BMGs are unlikely to represent a diagnostic landform for ice-streaming. It is more likely that BMGs are a velocity–duration product, which can form beneath slow-flowing ice over long durations or more rapidly in ice-streaming settings. BMG initiation is hypothesised to have occurred either under the control of pre-existing structural features in the bedrock geology or through the agency of large erodents present in pre-Quaternary regolith. Future research could profitably focus on landform dating, modelling experiments, acquisition of empirical evidence to validate hypotheses of BMG initiation, and geophysical methods to explore on-going BMG formation in modern glacial environments.