Erodibility of frozen riverbed sediments

Cryosphere degradation due to climate change results in increased sediment availability, mobilization, and transport (Knight & Harrison, 2012; Lane et al., 2017; Micheletti & Lane, 2016; Beel et al., 2020; Li et al., 2021a; Zhang et al., 2021; Syvitski et al., 2022; Zhang, 2022). The change in magnitude and timing of water delivery and sediment transport alters river patterns (Lafrenière & Lamoureux, 2019; Fortier et al., 2007), such as via increasing channelization (Liljedahl et al., 2016; Li et al., 2021a; Chartrand et al., 2023). Despite the obvious signs of landscape transformation by rivers, the effects of climate change on High Arctic fluvial incision and sediment transport remain poorly quantified due to limited understanding of how thawing substrate, water availability, and erodibility interact.

Although frozen landscapes are expected to erode slower than their temperate counterparts, Eschenfelder et al. (in review) found that frozen riverbeds may erode faster. In their laboratory flume experiments, they observe injections of surface water into a frozen bed of uniform-sized spherical glass beads (D₅₀ of ~1.9 mm), delivering heat and momentum fluxes to the thaw front. This increased subsurface melting and drove development of pressure gradients which enhanced surface erosion. They argue that later in the thaw season, when the upper layer of the bed has thawed in a more homogeneous fashion, water injections into the bed are physically accommodated and hence do not contribute to elevated surface erosion.

The glass bead substrate of Eschenfelder’s experiments has a high porosity (p = 0.4) and permeability, supporting significant hyporheic flow. However, natural sediments vary in size and composition, altering porosity and permeability. Furthermore, sediment size variability affects thawing rate (Gatto, 1995; Costard et al., 2003, 2014) and erosion rate (Einstein, 1950, Mitchener & Torfs, 1996; McCarron et al., 2019; van Rijn, 2020; de Lange et al., 2024).

In the proposed follow-up experiments, we plan to use various mixtures of glass beads (D₅₀ = 0.88, 1.9 and 4.1 mm), and natural sediments, to alter porosity and permeability. The natural sediment distribution will be scaled to reflect substrate compositions observed during past field campaigns to the Canadian Arctic. We hypothesise that, if these water injections into the bed are still present in lower porosity sediments, fine grained beds will erode slower than courser grained beds, despite a lower threshold of motion. Furthermore, experiments with multiple freeze-thaw cycles will be performed, allowing assessment of a potential positive feedback loop where past-seasons’ thaw front undulations and surface topography can impact current season’s flow patterns – imparting “memory” onto the landscape.

With these experiments we will assess erosional mechanisms in a variety of grain sizes, allowing us to further explore the mechanisms of erosion in frozen riverbeds, ultimately aiding the understanding of spatial variability in channel incision in the field.