Plain Language Summary
The Antarctic Ice Sheet is the largest source of uncertainty in
projecting future sea-level rise. This is due to our limited
understanding of drivers and mechanisms triggering tipping points in
ice-sheet instability, and knowledge gaps regarding the retreat and
disintegration of ice shelves. Understanding processes leading to ice
shelf destabilization is critical to improving estimates of future
Antarctic mass loss because of their important role in stabilizing ice
flow. We studied Larsen C Ice Shelf changes using satellite data
collected over 1963–2020, and conducted modeling experiments to
elucidate the observed linkages between front retreat, flow
acceleration, and rifts. We find the development of rifts near ice rises
to be the primary control on Larsen C front calving and flow
acceleration in the past six decades. Rifts can affect each other by
causing structural weakening and modifying stress fields. To predict
future dynamical changes, it is necessary to account for this feedback
and capture how ice rigidity changes over time in response to rift
growth. If Larsen C retreats such that the compressive arch ceases to
exist, it will resemble the pre-collapsed Larsen B ice shelf, producing
widespread flow accelerations in response to the backstress loss from
ice rises.
1 Introduction
Ice shelves, as floating extensions of the grounded ice sheet, fringe
~75% of Antarctica’s coastline (Rignot et al., 2013).
Ice shelves gain mass through ice flux from inland glaciers, local snow
accumulation, and basal marine ice accretion, while they lose mass
primarily via basal melting and iceberg calving (e.g., Depoorter et al.,
2013; Liu et al., 2015), together with minor runoff of surface meltwater
(Bell et al., 2017). Although the direct effect of ice shelf mass loss
on sea-level rise is negligible, ice shelves provide an important
buttressing force to stabilize the grounded ice sheet and regulate the
contribution from inland glaciers to global sea level (Dupont & Alley,
2005; Gagliardini et al., 2010; Gudmundsson, 2013; Pritchard et al.,
2012; Reese et al., 2017; Thomas, 1979). Understanding the processes
driving ice shelf retreat is therefore essential to generating realistic
projections of future Antarctic ice sheet mass loss in a warming climate
(DeConto et al., 2021; Noble et al., 2020; Pattyn & Morlighem, 2020).
Satellite data have indicated increasing trends in thinning and retreat
of Antarctic ice shelves (e.g., Depoorter et al., 2013; Paolo et al.,
2015), revealing the vulnerability of ice shelves to atmospheric warming
and oceanic forcing (Holland et al., 2015; Liu et al., 2015; Vaughan &
Doake, 1996). As a region sensitive to climate warming, the Antarctic
Peninsula has undergone pronounced regional warming (Vaughan et al.,
2003), along with rapid retreat of ice shelves and glaciers on both the
west and east coasts (Cook et al., 2005; Doake & Vaughan, 1991; Rott et
al., 1996; Scambos et al., 2009; Scambos et al., 2004). Successive
retreat has occurred on the Larsen Ice Shelf along the east coast. The
Larsen Ice Shelf was formerly composed of Larsen A, B, C, and D, four
sections from north to south (Figure 1). Larsen A disintegrated in 1995,
with an areal loss of ~1600 km2 (Rott
et al., 1996). Larsen B collapsed in February–March 2002, with an areal
loss of ~3250 km2 (Scambos et al.,
2004). The rapid disintegrations of Larsen A and Larsen B triggered
immediate flow accelerations of upstream glaciers (Rignot et al., 2004;
Scambos et al., 2004), which highlights the importance of studying how
an ice shelf evolves to an unstable condition and how ice shelf retreat
affects upstream flow dynamics. This has also led to an increasing
concern as to whether Larsen C will disintegrate in the future (e.g.,
Borstad et al., 2017; Glasser et al., 2009; Hogg & Gudmundsson, 2017;
Jansen et al., 2010, 2015; Kulessa et al., 2014), especially after a
giant iceberg (A68, ~5800 km2) broke
off from Larsen C in July 2017.
The disintegrations of Larsen A and Larsen B were preceded by warm air
temperature and intense surface melt (Banwell et al., 2013; Scambos et
al., 2003; Scambos et al., 2000). Meltwater-induced hydrofracturing has
been postulated to be the dominant mechanism in the ice shelf collapse
(van den Broeke, 2005; MacAyeal et al., 2003; Robel & Banwell, 2019),
likely preconditioned by increased basal melting (Shepherd et al., 2003;
Vieli et al., 2007). To predict the future stability of Larsen C, recent
research has focused on surface melt and associated atmospheric drivers
(Bevan et al., 2020; Luckman et al., 2014), and parallel efforts have
focused on ocean-driven basal processes (Holland et al., 2015; Jansen et
al., 2013; McGrath et al., 2014), such as basal melting and the
stabilizing role of suture zones. Despite the importance of atmospheric
and oceanic forcing, the stability of an ice shelf is also controlled by
its stress field (Doake et al., 1998; Fürst et al., 2016; Kulessa et
al., 2014; Lhermitte et al., 2020). The stress field, either compressive
or tensile, is closely related to fracture opening. The spatial
distribution and temporal evolution of fractures determine the ice shelf
resilience, and enhanced fracturing along shear margins has been
suggested to be critical for ice shelf stability and flow dynamics
(Lhermitte et al., 2020; Joughin et al. 2008). Larsen A and Larsen B
were already in an unstable stress state before their collapse as the
front retreat had passed a “compressive arch” (Doake et al., 1998).
Investigating the dynamic changes of ice shelves over the long term is
important to understanding the destabilization processes. Multidecadal
satellite observations on the retreat of Larsen C provide an excellent
opportunity to study the long-term dynamic behavior of ice shelves and
associated rifting and calving processes. Understanding these processes
is also the key to predicting the future changes of other
embayment-confined ice shelves, such as the Ross Ice Shelf and
Filchner–Ronne Ice Shelf.
Here, we use multi-source satellite images spanning six decades and ice
shelf modeling experiments to conduct a comprehensive analysis of the
changes of ice shelf front, flow velocities, and stress fields over
Larsen C. The earliest satellite imagery acquired in the 1960s allows us
to extend the flow velocity data back to the 1960s–1980s. We analyze in
detail the development of two rifts that led to the 2017 calving event,
and reveal the close linkages between rift propagation, flow
acceleration, and front retreat. We further carry out ice shelf modeling
experiments to examine how ice shelf flow velocities and stress fields
vary in response to the front geometry change due to ice shelf retreat
and the mechanical weakening due to rift development.
2 Study area
Larsen C is confined by Jason Peninsula in the north and Kenyon
Peninsula and Gipps Ice Rise in the south (Figure 1). Because of the
barrier effect of the high mountain ridge on westerly winds, the east
side of the Antarctic Peninsula is generally colder than the west side
(King et al., 2017; Morris & Vaughan, 2003). However, strong downslope
foehn winds occasionally bring warm and dry air masses over the northern
Larsen C during the austral summer, causing intense surface melt in the
northwest near the Cabinet Inlet (Elvidge et al., 2015; King et al.,
2017). The structure of Larsen C (Figure 1) consists of fast flow units
draining from the twelve inlets, and suture zones originating from the
Churchill Peninsula, Cole Peninsula, Marmelon Point, Francis Island,
Tonkin Island, and Joerg Peninsula (Borstad et al., 2017; Glasser et
al., 2009). Suture zones connect those fast flow units and have been
believed to play a significant role in stabilizing the ice shelf
(Kulessa et al., 2014; McGrath et al., 2014). Two pinning points, the
Bawden Ice Rise and Gipps Ice Rise, anchor the northern and southern ice
front, respectively. The ice thickness varies from ~1000
m near the grounding line to ~200 m near the ice front.