The Quaternary stratigraphy of the Gods and Yakaw rivers area, northeastern Manitoba

Abstract

Reconstructing the spatiotemporal dynamics of past glaciations provides a long-term perspective that is essential to understanding how Earth systems respond to climate change over different timescales. While most ice sheet reconstructions focus on the last glaciation, incorporating stratigraphic records from older glaciations enhances our understanding of past glacial-interglacial cycles. The western Hudson Bay Lowland (HBL) in northeastern Manitoba preserves thick Quaternary sediment sequences (20–70+ m) spanning multiple glacial-interglacial cycles, influenced by two major ice spreading centers of the Laurentide Ice Sheet (the Keewatin dome to the northwest and the Quebec-Labrador dome to the east). This study revisits the Quaternary stratigraphy along the Gods and Yakaw rivers in the western HBL using contemporary techniques, including a detailed, multi-parameter approach to characterize tills, combining field observations, paleo-ice flow indicators (till fabric analysis, lodged boulder striations), and till composition (matrix geochemistry, clast-lithology counts). For this thesis, a hybrid lithostratigraphic-allostratigraphic approach was employed across ten sections to reconstruct local glacial dynamics and establish an updated stratigraphic framework. Results confirm a complex, laterally variable stratigraphy. The revised framework includes 21 delineated stratigraphic units: 15 tills correlated using primarily paleo-ice flow indicators and stratigraphic position, and 6 sorted sediment units defined by allostratigraphy. Several key findings emerge from this framework, including: (1) A more detailed reconstruction of local glacial dynamics, (2) recognition of at least 4 sub-till sorted sediment units interpreted as ice-free intervals, and (3) a relative stratigraphy that extends the Quaternary record in the region, possibly back to MIS 12 or even older. Sediment provenance and ice-flow indicator analysis suggest that older tills (units 1t–11t) were dominantly sourced from a Quebec-Labrador dome, while younger tills (units 13t–20t) exhibit more complex signatures. This shift between units 11t and 13t may reflect changes in paleo-ice sheet configuration, bedrock availability to glacial erosion, or entrainment of pre-existing sediment. Although some till units appear superficially similar, they are interpreted as discrete tills deposited during separate ice-flow phases rather than products of glaciotectonic stacking. An alternative explanation invoking glaciotectonism is considered; however, variability of paleo-ice flow indicators, the absence of pervasive deformation structures, and a regional context unfavourable to large-scale glaciotectonism opposes this interpretation. These findings have broad implications for both ice sheet modeling and mineral exploration. In ice sheet modeling, incorporating records from older glacial-interglacial cycles provides key constraints to long-term ice sheet reconstructions, improving understanding of ice sheet response to climate change, which leads to better predictions of future changes. In mineral exploration, understanding glacial dynamics in thick-drift regions helps establish dispersal directions and sediment provenance, which may either 1) improve the interpretation of indicator mineral source, or 2) confirm a long history of inheritance and overprinting that requires the development of new techniques.