The evolution of ice fabrics: A continuum modelling approach validated against laboratory experiments

Modeling Ice‐Crystal Fabric as a Proxy for Ice‐Stream Stability

“…If migration recrystallization is active and rapid, then there may be little information about the flow preserved. Comparison of laboratory measurements of deformation with small‐scale modeling suggests that below C lattice rotation and migration recrystallization can happen at similar rates, though speeds diverge significantly by C (Richards et al., 2021); however, such modeling must make its own approximations, and it is possible that it does not accurately capture the rate of recrystallization at all temperatures or stresses. Moreover, adding new processes to the fabric evolution in large‐scale models can require fundamental changes to the modeling approach.…”

“…However, in effect is a decay timescale for the fabric; an infinite would instantly erase any fabric while = 0 implies that fabric lasts forever in stagnant ice. The effect of on the fabric is roughly akin to that of rotation recrystallization, which is thought to impede fabric development, and the dependence on temperature and was incorporated so that this process is more active in warmer and higher stressed areas; other models have used a similar diffusion term to represent rotation recrystallization (Gödert, 2003; Placidi et al., 2010; Richards et al., 2021). However, this term should not be seen as having a physical basis; besides allowing for regularization, it only allows for a sensitivity test to confirm that results are robust to a fabric decay timescale.…”

“…It cannot exclude the possibility that processes that create a preferred fabric different from that formed by lattice rotation would affect the results. Migration recrystallization tends to cause fabric to evolve to have ‐axes well aligned for the in‐situ stress rather than the strain history (e.g., Duval & Castelnau, 1995), and models of this process are more complex and not yet integrated into ice‐flow models (Placidi et al., 2010; Richards et al., 2021). Further details of parameter choices and the form of Equation 8, and demonstration of the insensitivity of results to , are presented in Text S1.…”

“…Different models have been developed to describe the effect of fabric on the large-scale flow (Castelnau et al, 1996;Meyssonnier & Philip, 1996;Svendsen & Hutter, 1996), the evolution of the fabric in response to the flow field (Thorsteinsson, 2002), or both (e.g., Azuma, 1994;Rathmann et al, 2021;Richards et al, 2021). Some microstructural models aim to capture many of the physical processes at the grain scale (Faria et al, 2002;van der Veen & Whillans, 1994), but it is computationally infeasible to do this in a largescale model.…”

“…Much work on ice‐crystal orientation fabric has focused on the fabric's effect upon deformation (e.g., Hruby et al., 2020; Ma et al., 2010; Martín & Gudmundsson, 2012; Minchew et al., 2018) since it can control flow by introducing local directional hardening (Duval et al., 1983). Though several models have been developed to describe the evolution of ice‐crystal fabric (Azuma & Higashi, 1985; Faria et al., 2002; Gillet‐Chaulet et al., 2005; Kennedy et al., 2013; Richards et al., 2021; Thorsteinsson, 2002), relatively little work applies these models to understand crystal fabric as a proxy for flow, and only a small subset of that work considers fast‐flowing glaciers (Thorsteinsson et al., 2003). However, ice‐stream velocities are a primary control on ice‐sheet mass balance on various timescales, and are thus of broad glaciological interest.…”