1 Introduction
The West Antarctic Ice Sheet (WAIS) is known to be a geometrically
unstable system – thin near its modern margins but thick in the
interior, where the ice sheet sits in deep marine basins (Morlighem et
al., 2020). Because the flux of ice to the ocean by deformation and
sliding scales non-linearly with ice thickness, forced retreat into the
thick ice sheet interior triggers a positive feedback, increasing ice
discharge and driving a self-sustaining mass imbalance that will proceed
without further forcing (the Marine Ice Sheet Instability, or MISI;
Weertman, 1974).
At present, WAIS is bounded by floating ice shelves, which play a
critical role in buttressing ice stored in the continental interior
(Dupont & Alley, 2005; Weertman, 1974). While ice shelf change over the
early 21st century has been dominated by ocean driven
melt from below, the disintegration of the Larsen B Ice Shelf (LBIS) by
surface melt and ice-shelf hydrofracture has demonstrated that other
modes of failure are possible (Glasser & Scambos, 2008), leading to a
range of studies about calving dynamics and the future of Antarctic ice
shelves in a warmer atmosphere (Bell et al., 2018; Kingslake et al.,
2017; Lai et al., 2020).
Calving occurs most frequently by the concentration of longitudinal
stresses in ice shelves at crevasse tips (Benn et al., 2007). Should
WAIS experience widespread ice shelf break-up, thick glaciers could
experience stresses at their termini that exceed the fracture toughness
of ice. Such configurations lead to cliff failure, a specialized calving
process controlled by differential stresses at the free surface of
marine terminating ice fronts (Parizek et al., 2019). Cliff failure is
thought to proceed faster than traditional calving, but the
rate-limiting process is uncertain (DeConto & Pollard, 2016).
The “marine ice cliff instability (MICI)” is the idea that cliff
failure, once initiated, accelerates with increasing cliff height. Like
with MISI, forced retreat into the WAIS interior by fracture processes
would then be self-sustaining. The inclusion of MICI into ice sheet
models results in a sea level rise by 2100 equal to eight times the
values predicted by models that do not include it (Bulthuis et al.,
2019; DeConto & Pollard, 2016; Golledge et al., 2019). While it is not
the only source of spread in ice sheet projections, MICI does represent
the largest source of uncertainty (Fox-Kemper et al., 2021). As a
result, there is a community desire to evaluate the likelihood of MICI
and constrain its initiation thresholds, a focus of this work.
Recent models of cliff failure
built from physics first principles (Bassis et al., 2021; Bassis &
Walker, 2012; Clerc et al., 2019; Crawford et al., 2021; Parizek et al.,
2019; Schlemm & Levermann, 2021; Ultee & Bassis, 2016) support MICI’s
theoretical validity, but the modes of failure and controlling
parameters differ across them. Depending on the prescribed (but
uncertain) material strength of ice, cliff-failure thresholds in the
literature span from 60 to 540 m high cliffs (Clerc et al., 2019).
Ice-sheet wide models rely heavily on process models (and their
underlying assumptions about ice strength) to justify parameterizations
of ice-terminus dynamics.
There are few geological or geophysical observations capable of
improving our understanding of ice cliff dynamics. The paleo-sea-level
record may provide indirect evidence of MICI; during several periods in
Earth’s history, ice sheet models have difficulty reproducing the
observed rapid sea level rise without it (Wise et al., 2016), but there
is disagreement in the literature about whether MICI is required to
explain historical Antarctic ice loss (Edwards et al., 2019). The
Intergovernmental Panel on Climate Change (IPCC) names Crane Glacier’s
response to the LBIS collapse as the only potential evidence for MICI
behavior in the satellite record (Oppenheimer et al., 2019). Crane
Glacier has significantly influenced community thinking about MICI, with
its post-LBIS-collapse ice calving rate used to constrain some
parameterizations of cliff failure (DeConto & Pollard, 2016). But our
knowledge of Crane Glacier’s geometry and behavior following the LBIS
collapse is limited by sparse geophysical data, challenging evaluation
of whether MICI-style cliff failure occured.
By both identifying what we do know and what we cannot know based on
available geophysical data, we can more precisely use Crane Glacier’s
behavior following the LBIS collapse event to refine our understanding
of ice cliff failure. In this work, we evaluate Crane’s retreat behavior
and terminal characteristics against process models, to both identify
whether or not unstable retreat by cliff failure was theoretically
possible at Crane, and if so, what unknown model parameters must be for
failure to occur there.