Tidewater Glacier Surges Initiated at the Terminus

Abstract

There have been numerous reports that surges of tidewater glaciers in Svalbard were initiated at the terminus and propagated up-glacier, in contrast with downglacier-propagating surges of land-terminating glaciers. Most of these surges were poorly documented, and the cause of this behavior was unknown. We present detailed data on the recent surges of two tidewater glaciers, Aavatsmarkbreen and Wahlenbergbreen, in Svalbard. High-resolution time series of glacier velocities and evolution of crevasse patterns show that both surges propagated up-glacier in abrupt steps. Prior to the surges, both glaciers underwent retreat and steepening, and in the case of Aavatsmarkbreen, we demonstrate that this was accompanied by a large increase in driving stress in the terminal zone. The surges developed in response to two distinct processes. (1) During the late quiescent phase, internal thermodynamic processes and/or retreat from a pinning point caused acceleration of the glacier front, leading to the development of terminal crevasse fields.(2) Crevasses allowed surface meltwater and rainwater to access the bed, causing flow acceleration and development of new crevasses up-glacier. Upward migration of the surge coincided with stepwise expansion of the crevasse field. Geometric changes near the terminus of these glaciers appear to have led to greater strain heating, water production, and storage at the glacier bed. Water routing via crevasses likely plays an important role in the evolution of surges. The distinction between internally triggered surges and externally triggered speedups may not be straightforward. The behavior of these glaciers can be understood in terms of the enthalpy cycle model.Many aspects of this class of surges remain poorly known. In particular, it is unclear whether surges of Svalbard tidewater glaciers are driven by similar cyclic internal processes to land-terminating glaciers SEVESTRE ET AL.

The Possible Transition From Glacial Surge to Ice Stream on Vavilov Ice Cap

“…When the major collapse initiated in 2014/2015, diffusivity was low (mostly < 1 × 10 5 m 2 /day, Figure 4e) and the perturbation at the thinning center (8-10 km distance from 1984 terminus) was dominated by advection. A similar process has been observed in many tidewater glaciers of Svalbard (e.g., McMillan et al, 2014;Murray et al, 2012;Nuth et al, 2019;Sevestre et al, 2018;. In 2016, the downstream region (0-15 km distance from 1984 terminus) reached a high diffusivity (> 2 × 10 5 m 2 /day) and a low |P e | while the upstream region still underwent advection (P e ≈ −6).…”

“…The analysis of P e also shows that once a glacier begins to collapse, P e is likely to decrease, causing thinning or thickening perturbations to diffuse inland (or advect inland if P e is negative enough). A similar process has been observed in many tidewater glaciers of Svalbard (e.g., McMillan et al, 2014;Murray et al, 2012;Nuth et al, 2019;Sevestre et al, 2018;. Fowler and Johnson (1996) point out mathematically that this process can also take place on an ice sheet and subsequently form an ice stream, although an ice sheet surge has never been observed in the modern data records.…”

“…Many tidewater glaciers and ice streams show periodic summer speedups (e.g., Dunse et al, 2015;Kehrl et al, 2017;Neckel et al, 2016;Schellenberger et al, 2017;Sevestre et al, 2018). This is due to seasonal variation in subglacial hydrology that greatly affects the glacier sliding speeds (Sevestre et al, 2018). During summer, melt can penetrate to a glacier bed and lubricate the interface, causing higher glacier speeds (Dunse et al, 2015).…”

Seven decades of uninterrupted advance of Good Friday Glacier, Axel Heiberg Island, Arctic Canada

“…This behaviour has been observed on several marine-terminating glaciers on Svalbard (e.g. Murray and others, 2003;Sevestre and others, 2018), as well as in the Canadian Arctic (Van Wychen and others, 2016), but is rather unlikely to have occurred on GF at the time since the terminus was then grounded well above sea level, behind a ∼1 km wide and 30 m high moraine (Müller, 1969). That being said, it would appear that the ice front reached fjord waters at some time in the mid-1960s, and the terminal moraine became progressively submerged as the terminus advanced past the nunatak and into a wider section of the valley (Fig.…”

Hagen Bræ: A Surging Glacier in North Greenland—35 Years of Observations

“…We suggest that surge conditions were primed by thickening of the ice, increase in slope, and late season high melt in 2001 but that the timing was determined by the large amount of meltwater produced in summer 2002. A study of two tidewater glaciers in Svalbard also concluded that the observed surges were preconditioned by internal mechanisms but set off by external processes (Sevestre et al, ).…”