Angelika Humbert

Oceanic controls on the mass balance of Wilkins Ice Shelf, Antarctica

� 0.8 m a � 1 , driven by a mean basal melt rate of 〈wb〉 = 1.3 � 0.4 m a � 1 . Interannual variability was large, associated with changes in both surface mass accumulation and 〈wb〉. Basal melt rate declined significantly around 2000 from 1.8 � 0.4 m a � 1 for 1992–2000 to � 0.75 � 0.55 m a � 1 for 2001–2008; the latter value corresponding to approximately steady-state ice-shelf mass. Observations of ocean temperature T obtained during 2007–2009 by instrumented seals reveal a cold, deep halo of Winter Water (WW; T ≈ � 1.6°C) surrounding WIS. The base of the WW in the halo is � 170 m, approximately the mean ice draft for WIS. We hypothesize that the transition in 〈wb〉 in 2000 was caused by a small perturbation (� 10–20 m) in the relative depths of the ice base and the bottom of the WW layer in the halo. We conclude that basal melting of thin ice shelves like WIS is very sensitive to upper-ocean and coastal processes that act on shorter time and space scales than those affecting basal melting of thicker West Antarctic ice shelves such as George VI and Pine Island Glacier.

Modeling of Store Gletscher's calving dynamics, West Greenland, in response to ocean thermal forcing

Glacier-front dynamics is an important control on Greenlands ice mass balance. Warmer ocean waters trigger ice-front retreats of marine-terminating glaciers, and the corresponding loss in resistive stress leads to glacier acceleration and thinning. Here we present an approach to quantify the sensitivity and vulnerability of marine-terminating glaciers to ocean-induced melt. We develop a plan view model of Store Gletscher that includes a level set-based moving boundary capability, a parameterized ocean-induced melt, and a calving law with complete and precise land and fjord topographies to model the response of the glacier to increased melt. We find that the glacier is stabilized by a sill at its terminus. The glacier is dislodged from the sill when ocean-induced melt quadruples, at which point the glacier retreats irreversibly for 27 km into a reverse bed. The model suggests that ice-ocean interactions are the triggering mechanism of glacier retreat, but the bed controls its magnitude.

Complex network of channels beneath an Antarctic ice shelf

Ice shelves play an important role in stabilizing the interior grounded ice of the large ice sheets. The thinning of major ice shelves observed in recent years, possibly in connection to warmer ocean waters coming into contact with the ice-shelf base, has focused attention on the ice-ocean interface. Here we reveal a complex network of sub ice-shelf channels under the Fimbul Ice Shelf, Antarctica, mapped using ground-penetrating radar over a 100 km2 grid. The channels are 300–500 m wide and 50 m high, among the narrowest of any reported. Observing narrow channels beneath an ice shelf that is mainly surrounded by cold ocean waters, with temperatures close to the surface freezing point, shows that channelized basal melting is not restricted to rapidly melting ice shelves, indicating that spatial melt patterns around Antarctica are likely to vary on scales that are not yet incorporated in ice-ocean models.

Dynamics of the ice cap on King George Island, Antarctica: field measurements and numerical simulations

Abstract King George Island is located at the northern tip of the Antarctic Peninsula, which is influenced by maritime climate conditions. The observed mean annual air temperature at sea level is –2.4˚C. Thus, the ice cap is regarded as sensitive to changing climatic conditions. Ground-penetrating radar surveys indicate a partly temperate ice cap with an extended water layer at the firn/ice transition of the up to 700 m high ice cap. Measured firn temperatures are close to 0˚C at the higher elevations, and they differ considerably from the measured mean annual air temperature. The aim of this paper is to present ice-flow dynamics by means of observations and simulations of the flow velocities. During several field campaigns from 1997/98 to 2008/09, ice surface velocities were derived with repeated differential GPS measurements. Ice velocities vary from 0.7 m a−1 at the dome to 112.1 m a−1 along steep slopes. For the western part of the ice cap a three-dimensional diagnostic full-Stokes model was applied to calculate ice flow. Parameters of the numerical model were identified with respect to measured ice surface velocities. The simulations indicate cold ice at higher elevations, while temperate ice at lower elevations is consistent with the observations.

The Wilkins Ice Shelf, Antarctica : break-up along failure zones

The Wilkins Ice Shelf, Antarctica, has experienced a series of break-up events during the first half of 2008, the first on 28/29 February, the second on 30/31 May and the most recent in June/July. The ice shelf (area 13 000 km) experienced the first two break-up events on the connection of the central part of the ice shelf with two confining islands, Latady and Charcot islands. The June/July break-up started in a region where an area of 1100 km had already broken off in 1998. After the 1998 event, the icebergs were not transported out of the bay; rather, an ice melange formed, composed of icebergs and sea ice. Figure 1a (dated 2 March) shows the northern part of the Wilkins Ice Shelf and the location of the melange. The June/July break-up occurred in two phases: in the first, the ice melange opened at its northwestern margin and pushed against the northernmost part of the northern ice front, which then broke off. The formation of an iceberg in the interior, followed by a disintegration caused by a capsize mechanism, pushed towards the north and initiated a second phase. A total of 1220 75 km, or 8.8–10% of its total size, was lost. In the first phase the melange exhibited an area of open water (Envisat Advanced Synthetic Aperture Radar (ASAR) image, 28 June), while it was still in good order on 26 June. We observe a movement of the melange towards the east, causing fracturing and displacement. There is no visible indication that parts of the melange have melted. In the subsequent 3 days, 70% of the ice melange disintegrated into pieces of thin ice and thicker iceberg fragments (originated from the 1998 break-up). The eastward movement of the melange acted as a force on the northern ice front. The northernmost 30 km of the ice shelf was already fractured during the 1998 break-up event, so that it consisted of rectangular blocks (2–4 km long, 10–30 km wide), all connected on the eastern side to an unfragmented ice-shelf area (Fig. 1a). The acting force on this mass presumably caused the blocks to fracture. During that phase, 540 75 km (only ice-shelf ice counted, no melange ice) were lost and an intermediate northern ice front was formed. South of this line, the ice shelf was, up to that date, unaffected. This phase was comparable to (fast) calving rather than to ice-shelf break-up (timescale of hours). The second phase started on 4 July, when the formation of an iceberg took place 15 km inward of the intermediate ice front. Prior to this (first detected on 24 June), a 6.25 km long rift (inset in Fig. 1b) in the connection between Charcot and Latady islands appeared. The rift was perpendicular to the northern ice front of this ice bridge. It occurred in an area where the estimated ice thickness (Braun and others, 2008) drops from 240 to 180m normal to the ice front, causing variations in buoyancy forces, which leads to the accumulation of bending stresses. The shape of the fracture, starting at the margin of the ice front, propagating inward and stopping, leads us to assume that tensile stresses are responsible for the rift formation. There were two potential sources for tensile stresses. The first was the predominant wind direction – north-northeast (visible wind erosion structures), which acted as a drag (movement of the sea ice along the northern ice front indicates strong wind at that time). Secondly, the ice melange might have already been moving and a strong contact between the melange and the ice-shelf front could then have produced a tensile stress along the ice-shelf front. The latter can be excluded, since German X-band high-resolution synthetic aperture radar (TerraSAR-X) strip-map mode images from 19 and 25 June show that there was no displacement of the constituents of the melange. For 9 days following 25 June, the ice-shelf geometry remained stable, although the melange had already started its disintegration. On 4 July an iceberg formation at the rift started (inset, Fig. 1c). Load rearrangement due to mass loss in the first phase and tidal interaction at the intermediate northern ice front may have supported this iceberg formation. A TerraSAR-X scansar image from 5 July (Fig. 2) indicates, only 12 hours later, the formation of a second, smaller iceberg and a rapid increase in area between the two icebergs by 20 km. Figure 2b reveals formation of ‘sliver’ icebergs and capsized icebergs, as well as a crushed ice Humbert and Braun: Correspondence

A comparative modeling study of the Brunt Ice Shelf/ Stancomb-Wills Ice Tongue system, East Antarctica

Two diagnostic, dynamic/thermodynamic ice-shelf models are applied to the Brunt Ice Shelf/Stancomb-Wills Ice Tongue system, located off Caird Coast, Coats Land, Antarctica. The Brunt Ice Shelf/Stancomb-Wilis Ice Tongue system is characterized as a thin, unbounded ice shelf with an atypical and highly heterogeneous structure. In contrast to other ice shelves, a composite mass of icebergs that calved at the grounding line and were then locked within fast (sea) ice exists between the fast-moving Stancomb-Wills Ice Stream and the slow-moving Brunt Ice Shelf. We simulate the present flow regime of the ice shelf that results from the ice-thickness distribution and the inflow at the grounding line with two different models, and compare the model results with feature tracking and InSAR flow velocities. We then incorporate two observed features, a rift and a shear margin, into the models with two different approaches, and demonstrate the effects of variations in numerical values for the shear strength and viscosity in these zones on the simulated velocity field. A major result is that both kinds of implementation of the rifts lead to similar effects on the entire velocity field, while there are discrepancies in the vicinity of the rifts.

The mechanisms behind Jakobshavn Isbræ's acceleration and mass loss: A 3‐D thermomechanical model study

The mechanisms causing widespread flow acceleration of Jakobshavn Isbrae, West Greenland, remain unclear despite an abundance of observations and modeling studies. Here we simulate the glaciers evolution from 1985 to 2016 using a three-dimensional thermomechanical ice flow model. The model captures the timing and 90% of the observed changes by forcing the calving front. Basal drag in the trough is low, and lateral drag balances the ice streams driving stress. The calving front position is the dominant control on changes of Jakobshavn Isbrae since the ice viscosity in the shear margins instantaneously drops in response to the stress perturbation caused by calving front retreat, which allows for widespread flow acceleration. Gradual shear margin warming contributes 5 to 10% to the total acceleration. Our simulations suggest that the glacier will contribute to eustatic sea level rise at a rate comparable to or higher than at present.

Subglacial roughness of the former Barents Sea ice sheet

[1] The roughness of a glacier bed has high importance for the estimation of the sliding velocity and can also provide valuable insights into the dynamics and history of ice sheets, depending on scale. Measurement of basal properties in present-day ice sheets is restricted to ground-penetrating radar and seismics, with surveys retrieving relatively coarse data sets. Deglaciated areas, like the Barents Sea, can be surveyed by shipborne 2-D and 3-D seismics and multibeam sonar and provide the possibility of studying the basal roughness of former ice sheets and ice streams with high resolution. Here, for the first time, we quantify the subglacial roughness of the former Barents Sea ice sheet by estimating the spectral roughness of the basal topography. We also make deductions about the past flow directions by investigating how the roughness varies along a 2-D line as the orientation of the line changes. Lastly, we investigate how the estimated basal roughness is affected by the resolution of the basal topography data set by comparing the spectral roughness along a cross section using various sampling intervals. We find that the roughness typically varies on a similar scale as for other previously marine-inundated areas in West Antarctica, with subglacial troughs having very low roughness, consistent with fast ice flow and high rates of basal erosion. The resolution of the data set seems to be of minor importance when comparing roughness indices calculated with a fixed profile length. A strong dependence on track orientation is shown for all wavelengths, with profiles having higher roughness across former flow directions than along them.

The temperature regime of Fimbulisen, Antarctica

Abstract Numerical simulations of the temperature regime of the ice shelf Fimbulisen, Antarctica, are presented. A vertical temperature profile (S1) of Fimbulisen has been measured at the extension of Jutulstraumen, in which the temperature decreases with depth. The three-dimensional steady-state temperature field was computed by a finite-element technique. Horizontal flow velocities and surface accumulation rates were derived from observations. The basal melt rate distribution arose from an assumption of balance in the mass continuity equation. The computed basal melt rate distribution (a b) indicates that the highest basal melt rates, up to 15 m a 1 occur at the inflow gate of Jutulstraumen, and low basal melt rates (<0.6ma 1) occur in the slower moving parts. Where the ice shelf overhangs the continental shelf, a b ~1.2ma−1 . The resulting temperature field indicates that Fimbulisen consists of a cold middle part, built up by the extension of Jutulstraumen, and warmer ice masses in slow-moving areas to the west and east. Furthermore, model runs were set up in which the atmospheric temperatures increased in +1 K steps. The results suggest that the warming effectively increases the temperatures throughout the ice column in the slower-moving parts, therefore enhancing shear at the margins of the extension of Jutulstraumen.

Numerical simulations of the ice flow dynamics of George VI Ice Shelf, Antarctica

A diagnostic, dynamic/thermodynamic ice-shelf model is applied to the George VI Ice Shelf, situated in the Bellinghausen Sea, Antarctica. The George VI Ice Shelf has a peculiar flow geometry which sets it apart from other ice shelves. Inflow occurs along the two longest, and almost parallel, sides, whereas outflow occurs on the two ice fronts that are relatively short and situated at opposite ends of the ice shelf. Two data sources were used to derive the ice thickness distribution: conventional radio-echo sounding from the British Antarctic Survey was combined with thickness inferred from surface elevation obtained by the NASA GLAS satellite system assuming hydrostatic equilibrium. We simulate the present ice flow over the ice shelf that results from the ice thickness distribution, the inflow at the grounding line and the flow rate factor. The high spatial resolution of the ice thickness distribution leads to very detailed simulations. The flow field has some extraordinary elements (e.g. the stagnation point characteristics resulting from the unusual ice-shelf geometry).