4.1 Evaluating the Retreat Behavior of Crane Glacier
Without observations during the rapid 8 month retreat window, there is no direct evidence of ice cliff failure at Crane Glacier. But there is evidence for a geometric instability in the system, and indirect evidence that cliff failure may have occurred. We examine these two ideas separately.
As we observed at Crane, fracture processes dominated following LBIS break-up, where a terminal ice cliff was observed from 2002-2004. But immediately following the forcing event (LBIS collapse), retreat by fracture was slow. As the terminus retreated into thicker ice, calving and retreat rates increased despite no new climatic forcing, consistent with a geometric instability. Retreat continued, with a rapid final phase occurring where the ice had been clearly grounded in 2002. The system eventually reached a configuration with a terminal stress state that inhibited further retreat, and the system restabilized.
Many conditions are known to inhibit cliff failure and stabilize glacier systems, including narrowing of the fjord (Schlemm & Levermann, 2021), development of sea-ice or mélange in the fjord which can apply sufficient backstress to stabilize the cliff (Bassis et al., 2021; Crawford et al., 2021; Schlemm & Levermann, 2021), retreat into thinner ice along a prograde slope (Bassis et al., 2021), or retreat into ice that is not preconditioned for failure by crevassing or other processes that induce damage (Clerc et al., 2019). As Crane’s terminus retreated, the fjord narrowed to a point where calving and ice flow were in balance. Then, viscous processes dominated, and the system thinned. In summer 2011, perennial land-fast sea ice formed in Crane Glacier’s embayment (Figure S4), which likely enabled the growth of an ice shelf, consistent with models of sea ice backstress (Robel, 2017) and documented sea-ice circulation patterns in the region (Christie et al., 2022). The reestablishment of a floating ice shelf facilitated Crane Glacier’s recent thickening and velocity decrease.
Cliff failure is thought to be one of the fastest calving processes. Crane’s final minimum calving rate of 8.50 km/a represented an increase of both its pre-collapse ice shelf calving rate (Alley et al., 2008) and its calving rate during the initial period following collapse, which could indicate a change in the mechanism of calving over that period. But rates greater than 11 km/a have been recently observed elsewhere on the WAIS (Milillo et al., 2022), and at Jakobshavn and Helheim Glaciers in Greenland, whose calving rates exceed 17 km/a (Joughin et al., 2014) and 11 km/a (Howat et al., 2005), respectively. The calving rates at Jakobshavn and Helheim, in part, accommodate their high terminus velocities (9.5 to 15.5 km/a faster than the terminus of Crane), which is an important qualitative difference between these systems and Crane. As such, the calving rate observed at Crane does not fall outside of the range experienced by floating systems elsewhere in Antarctica and Greenland, but is anamolous for a glacier flowing only 1.5 km/a at its terminus.
Ultimately, we see 10.05 km of terminus retreat, 3.28 ± 0.03 km of which could have occurred in grounded ice, with calving rates that accelerated over the course of retreat.