Laurentide Ice Sheet dynamics evolution in the inner-ice sheet region of northeastern Quebec, Canada

Abstract

The study of past ice sheets, such as the Laurentide Ice Sheet (LIS), which covered most of northern North America during the last glaciation (last ~100-kyr), provides critical knowledge about the long-term behaviour of ice sheets, how they modified landscapes and the sedimentary record, and how they responded to external climate forcings. The LIS had a complex evolution including several changes in its configuration (e.g., ice sheet extent, its surface topography and its thickness). This is essentially known through multiple studies of ice-flow indicators and glacial landforms and sediments. However, conflicting ice flow reconstructions exist for many regions of the LIS, with implications for the overall understanding of the long-term evolution of ice sheets. One of these regions, within the Quebec-Labrador (Q-L) sector of the LIS, is located in northern Quebec and western Labrador. The Q-L sector is one of the largest ‘inner-regions’ of the LIS where the main ice divides formed with an unresolved ice-flow history. Specifically, a major landform boundary, commonly referred to as the horseshoe unconformity, has been the source of conflicting paleoglaciological reconstructions. Changes in the thermal regime at the base of the ice sheet have been invoked to explain this landscape, whereby relict fragments of the ice sheet imprint are partially preserved following a switch from warm-based to cold-based conditions. However, previous research has disagreed on the sequence of events, the relative timing/duration, and significance. Large areas within the former Q-L sector also lack field-based evidence and detailed data regarding ice-flow indicators, sediment transport, and other characteristics (e.g., weathering), which contributed to the uncertainties in ice sheet reconstructions. Therefore, a study area within the inner-region of the Q-L sector was selected to investigate its glacial evolution. The overall goal of this thesis is to characterize the glacial sediments and their distribution across the study area and develop a better understanding of the evolution of this inner-ice sheet region throughout glaciation. A specific focus is placed on establishing the ice-flow chronology, subglacial conditions associated with these ice-flow phases, and detailing the resulting dispersal of subglacial materials. To achieve these research objectives, a multi-faceted approach is used, whereby ice-flow phases are reconstructed based on the glacial geomorphological evidence (e.g., landform and outcrop-scale ice-flow indicators), with new geochronology constraints on deglaciation (e.g., optical dating and cosmogenic isotope ‘exposure’ dating). Additionally, the evolution of subglacial dynamics was reconstructed using a number of different proxies (e.g., landform density and elongation, lake abundance and area, and the presence of a till blanket). These proxies were then used to derive a subglacial dynamics index map to infer the mobility of the ice sheets bed/sediment layer, which is tested against a chemical index of alteration (from till matrix composition) as well as cosmogenic isotope (10Be) inheritance from till and bedrock outcrops. Results from this research have revealed a complex ice-flow history, detailing four ice-flow phases associated with complex spatiotemporal fluctuations in subglacial conditions that led to a fragmentary glacial landscape, as well as amoeboid type sediment dispersal patterns (i.e., dispersed in multiple directions around the source). Six distinct glacial terrain zones (GTZs) with varying degrees of overprinting and preservation are recognized. Together, these GTZs form a mosaic landscape with remarkable landscape preservation of older glacial terrains in some GTZs, and stronger overprinting by younger phases in others. Dispersal patterns across the study area provide additional insights into sediment entrainment and deposition in relation to the different ice-flow phases identified, allowing for a more holistic understanding of the glacial processes and history for the inner-regions of the Q-L sector. The oldest flow recognized in the study area (Flow 1) was a regional northeast ice-flow event with predominantly warm-based conditions across the majority of the ice-bed interface; it is best preserved in the eastern portion of the study area (GTZ1). Flow 1 had a significant impact on the dispersal patterns across the study area, with evidence for long transport distance (>100 km) towards a general northeast direction. Following Flow 1, subglacial conditions transitioned to more cold-based conditions, likely relating to the formation of an ice-divide in the eastern portion of the study area, except in the northwest corner (GTZ2), where ice-flow indicators almost completely overprint the evidence of Flow 1. Flow 2 is associated with ice streaming events in Ungava Bay and its landscape imprint has produced high subglacial index, indicating highly dynamic basal ice. Following Flow 2 the ice stream catchment shifted west in Ungava Bay and evidence within the study area suggests the ice divide within the study area also shifted west, as indicated by multiple sites where an almost complete ice-flow reversal is recorded. Most of these sites are located along the edge of a central upland terrain, consisting of resistant bedrock (De Pas Batholith), characterized by an overall lower subglacial index (GTZ4) than elsewhere within the study area. The high angle crosscutting of Flow 2 and Flow 3 (near opposite flows) provides compelling evidence for westward ice divide migration across the study area, which probably happened in response to the changing configuration of the Ungava Bay ice stream. Evidence of Flow 3 is discontinuous, but it is abundant on the eastern flank (GTZ3) of the central uplands (GTZ4), where Flow 3 indicators overprint Flow 1 indicators at a few key sites. Considering the broader glacial landscape from outside the study area, it appears that this flow phase was also influenced by ice streaming, specifically by a few ice streams operating on the eastern margin of the Quebec-Labrador sector. Again, high proxy values associated with mobile bed conditions are abundant within the inner regions of the two GTZs correlated to this ice-flow phase (GTZ3 and GTZ3b). Both GTZ3 and GTZ3b are surrounded by areas interpreted as inter-ice stream regions, which preserved most of the Flow 1 evidence (forming GTZ1), while producing undulating till blanket areas lacking streamlined landforms (GTZ5) due perhaps to basal meltout. East-trending eskers also crosscut Flow 1 features, suggesting Flow 3 occurred just prior to channelized meltwater drainage and final deglaciation. However, a fourth and final deglaciation ice-flow phase (Flow 4) is recognized in the striation record. This ice-flow phase was largely topographically controlled as the larger ice sheet began to fragment into smaller and thinner ice caps during deglaciation, which probably occupied the central upland area (GTZ4). Flow 4 had little if any impact on the landscape and has no discernable dispersal patterns. Geochronological results suggest the study area was deglaciated by about 8 ka. This deglacial timing is consistent with other regional deglacial ages, but it provides new additional constraints for the position of the retreating ice margin within the study area during this time. Although this thesis focuses on a specific region of the Q-L sector, evidence indicates that the sector was characterized by transient polythermal conditions, similar to the other large inner region with ice divides of the LIS (the Keewatin sector over northcentral Canada). Specifically, the spatiotemporal subglacial fluctuations occurred predominantly at the onset of deglaciation, after spatially extensive warm-based conditions had covered the region. As the ice sheet thinned, a greater proportion of the bed became cold-based; however, ice streaming events began to drain ice further inland, keeping warm-based conditions along narrow corridors. Subglacial conditions thus changed relatively rapidly, highly influenced by both the thinning of the LIS and regional ice streaming. This suggests the Q-L sector of the ice sheet had more transient polythermal conditions than previously reconstructed and modelled, which highlight the need to further improve our knowledge of bed thermal properties of inner ice sheet regions and the effect of ice stream catchment processes far into ice sheets, close to ice divides.