Isotopic diffusion in ice enhanced by vein-water flow
Felix Ng
Corresponding author: Felix Ng
Corresponding author e-mail: f.ng@sheffield.ac.uk
Diffusive smoothing of signals on the water stable isotopes in ice sheets limits the climatic information retrievable from these ice-core proxies. Previous theories explained how, in polycrystalline ice below the firn, fast diffusion in the network of intergranular water veins shortcircuits the slow diffusion in crystal grains to cause ‘excess diffusion’, enhancing the signal smoothing rate above that implied by self-diffusion in ice monocrystals. However, the controls of excess diffusion remain poorly understood. I show that vein-water flow amplifies excess diffusion, by altering the three-dimensional field of isotope concentrations and isotope transfer between the veins and crystals. The rate of signal smoothing depends not only on temperature, vein and grain sizes, and signal wavelength, but also on vein-water flow velocity, which can increase the rate by 1–2 orders of magnitude. This modulation can significantly impact signal smoothing at ice-core sites in Greenland and Antarctica, as demonstrated by simulations for the GRIP and EPICA Dome C sites, which show sensitive modulation of their diffusion–length profiles when vein-water flow velocities reach ~101–102 m a–1. Thus vein–flow mediated excess diffusion may help explain the mismatch between modelled and spectrally-derived diffusion lengths in other ice cores. I also show that excess diffusion biases the spectral estimation of diffusion lengths from isotopic signals and the reconstruction of surface temperature from diffusion–length profiles. These findings caution against using the single-crystal isotopic diffusivity to represent the bulk-ice diffusivity. The need to predict excess diffusion in ice cores calls for extensive study of isotope records for its occurrence and better understanding of vein-scale hydrology in ice sheets.
Limits on the orbital sounding of terrestrial ice sheets and near-surface reflectors
Dustin Schroeder
Corresponding author: Dustin Schroeder
Corresponding author e-mail: dustin.m.schroeder@stanford.edu
Orbital radar sounding of terrestrial ice sheets is an area of increasing research interest with mission concepts at 45 MHz P-Band, and L-Band under development. The success of orbital radar sounding instruments at Mars and planned missions to the icy moons of Jupiter make the prospect of similar instruments observing Earth’s cryosphere (and adding a subsurface complement to satellite observations of surface velocity and altimetry) an appealing prospect. However, large uncertainties remain regarding the impact of glacial conditions, platform altitude and range sidelobes on instrument performance. Here, we present limits introduced by signal, system and glaciological properties on the range of targets, conditions and processes for which orbital and other high-altitude sounders can substitute current instruments on crewed aircraft.
Har Amrit Singh Sandhu, Hemendra Singh Gusain, Manoj Arora, Dhiraj Kumar Singh, Neeraj Tripathi
Corresponding author: Har Amrit Singh Sandhu
Corresponding author e-mail: hassandhu@pec.edu.in
In the recent past, it has been reported that glaciers in different regions of the Himalaya are receding at varying rates. Receding glaciers will affect the availability of water in different regions of north India. The Himalaya has the largest glaciated area and housed a total number of 5243 glaciers. Glacier mapping and monitoring at regular intervals of time can assist in providing a timely solution for the imposed problems. Remote sensing provides an alternative for mapping and monitoring of glaciers in mountainous terrain. Numerous satellites have been successfully used for mapping and monitoring glaciers’ facies and features. In this study, 29 glaciers of the Bhagirathi basin, Garhwali Himalaya, have been monitored using remote sensing (RS) images for more than a decade. Bhagirathi basin has four sub-basins. The glacier area >5 km2 has been considered apart from a few small glaciers to estimate glaciers retreat and advances. Landsat Images of the different periods over the study area have been used during 2000 and 2015. ASTER GDEM has been used in the study for topographic analysis. All the datasets were co-registered taking 2015 OLI as base images. The slope match method has been used to get topographically corrected reflectance. Various glacier parameters such as glacier area, glacier length, snout position, altitude, aspect etc. were extracted using images to evaluate the glacier health in terms of retreat and advance. The glacier snout position has been estimated using visual interpretation techniques employing various features such as shape, texture, tone and surroundings. Glacier boundaries have been demarcated manually using Landsat images with the help of the 3-D visualization technique. Orientation has been determined in eight cardinal directions. It is observed that Bhagirathi sub-basin has the greatest glaciated area of ~35% and Pilang has the least with ~3.2%. 25 glaciers have shown retreat; 4 glaciers have shown advancement, resulting in a total glacier area loss of ~0.5%, while the retreat rate varies from ~0.06 m a–1 to ~19.4 m a–1. Dokarni glacier has the greatest retreat rate (~19.4 m a–1), and Dehigad has the greatest advance rate (~10.1 m a–1). Dehigad glacier is oriented 92% towards the north and was observed to be the greatest advancing (152 m) glacier too. The study covers more than 65% of the total glaciated area and, to our knowledge based on the existing literature; this is one of the first studies to cover the highest number of glaciers in this basin.
Using big data to understand calving in Greenland
Ginny Catania, Sophie Goliber, Christopher Miele, Enze Zhang, Timothy Bartholomaus
Corresponding author: Ginny Catania
Corresponding author e-mail: gcatania@ig.utexas.edu
With the advent of multiple satellite sensors pointed at the poles, we are amid a data explosion that assists in examining ice sheet change. This is particularly true for outlet glacier termini, which represent some of the most changeable parts of the Greenland Ice Sheet. Most of the large change occurring at glacier termini is related to calving, with at least three different styles of calving identified in the literature: 1) collapse of seracs probably related to undercutting from submarine melt of the glacier terminus; 2) buoyant flexure where basal crevasse propagation through the glacier releases full-thickness icebergs; and 3) tabular rifting caused by horizontal stretching in floating ice. In an effort to improve calving representation in ice sheet models we present a new dataset of glacier termini derived from machine learning and novel analyses of these data to understand calving style. AutoTerm is a collection of nearly 300 000 glacier terminus delineations from 295 glaciers in Greenland extending back to 1985 produced through a semi-automated machine learning pipeline. AutoTerm provides increased temporal resolution in terminus data over prior datasets allowing for unprecedented detail in glacier calving, particularly after 2014. We use AutoTerm data to examine glacier calving style for a subset of glaciers using two independent methods – one that examines the terminus shape and one that makes use of terminus change over time. These methods are validated with a detailed examination of existing calving laws. We then extrapolate our classification to the remaining glaciers in Greenland allowing a designation of calving style over time and space. These data may help to reduce parameter space uncertainty when fitting existing calving laws in ice sheet models to terminus change observations.
Nicolas Jullien, Andrew Tedstone, Horst Machguth
Corresponding author: Nicolas Jullien
Corresponding author e-mail: nicolas.jullien@unifr.ch
The firn layer of the Greenland Ice Sheet holds the potential to trap and refreeze surface meltwater within its pore space. Acting as a buffer, it prevents meltwater from leaving the ice sheet. However, the last two decades saw the development of several meter-thick ice slabs in the firn, reducing firn permeability and inhibiting vertical meltwater percolation. Ice slabs are located above the long-term equilibrium line along the west, north and northeast coasts of the ice sheet. Through time, ice slabs have thickened while new ones have developed at higher elevations. Concomitantly, the area of the ice sheet drained by surface rivers has increased by 29% from 1985–2020. This new runoff area corresponds strongly with where ice slabs are located. Nowadays, 5–10% of surface losses through meltwater runoff originate from these newly drained areas. Here, we demonstrate that the highest elevation that is drained by surface rivers – termed the maximum visible runoff limit – is controlled by the ice content in the subsurface firn. Using ice slab thickness derived from Operation IceBridge’s accumulation radar and annual maximum visible limit retrievals from Landsat imagery from 2002–18, we show that a threshold in ice content exists beyond which the shift from a ‘firn deep percolation regime’ to a ‘firn runoff regime’ occurs. Under a firn runoff regime, the ice slab forces the meltwater to pond on top of it, and to run off. We find evidence that the feedback between ice slabs and surface hydrology is mutual. Relating rivers and slush fields locations with horizontal–vertical polarization’s signal from the synthetic aperture radar onboard ESA’s Sentinel-1 satellite, we demonstrate that ice slabs are thicker underneath active surface hydrological features.
Anisotropy of the Priestley Glacier lateral shear margin, Antarctica
David Prior, M. Hamish Bowman, Lisa Craw, Reinhard Drews, Reza Ershadi, Shang Fan, Martin Forbes, David Goldsby, Travis Hager, Bryn Hubbard, Christine Hulbe, Deayeong Kim, Franz Lutz, Carlos Martin, Robert Mulvaney, Wolfgang Rack
Corresponding author: David Prior
Corresponding author e-mail: david.prior@otago.ac.nz
Shear margins play a critical role in controlling the flow of outlet glaciers and ice streams and the mechanical properties of ice shelves. We have measured several different types of anisotropy in the lateral shear margin of the Priestley Glacier (Terra Nova Bay), to provide data for models that include shear margins. Deformation kinematics approximate sinistral simple shear, on a vertical plane with a horizontal shear direction. Shear strain rates decrease from 10–9 s–1 (300 a–1), 0.5 km from the glacier edge, to 5 × 10–10 s–1 0.5 m further into the glacier and 10–10 s–1 1.2 km further still. Layers (millimetres to metre-scale) are vertical and the strike (azimuth of horizontal layer traces) changes from 5–30° clockwise of the shear direction, corresponding to reducing strain rates moving away from the glacier edge. Isoclinal folds and shear zones have geometries consistent with shear zone kinematics. Crystallographic preferred orientations (CPOs = fabrics) are characterized by very strong horizontal c-axis maxima sub-perpendicular to the shear plane. Smears of c-axes or a secondary maximum lie clockwise of the main maximum, describing a monoclinic symmetry in the horizontal plane. Sound wave (elastic) anisotropy matches models from the CPO data. Ultrasonic measurements on samples show that p-wave velocity maxima and s-wave velocity minima correspond to c-axis maxima. A multi-azimuth, hammer-plate vertical seismic profile using a 60 m borehole and a multi-azimuth, kilometre-scale, explosion source survey show the same pattern. The p-wave velocity maximum lies 15° clockwise of the shear plane normal. Velocity anisotropy has monoclinic symmetry in the horizontal plane. pRES polarimetric radar measurements yield horizontal maximum dielectric permittivity 10–30° clockwise of the shear direction. This is at a high angle (60–90°) to c-axis maxima and close to parallel with the traces of vertical layers. Our data suggest that seismic methods provide an excellent proxy for bulk ice crystallographic preferred orientations in shear margins. Polarimetric radar data do not match CPO measurements and may relate to the orientations of vertical layers. Experiments to measure viscous anisotropy using samples from the Priestley Glacier illustrate the importance of such measurements. Cylinders cored parallel to the c-axis maximum are 5 times as strong in axial compression as those cored 45° to the c-axis maximum.
Ablation areas in Antarctica: a hyperspectral analysis by in-situ and remote sensing observations
Giacomo Traversa, Roberto Colombo, Roberto Garzonio, Micol Rossini, Biagio di Mauro
Corresponding author: Giacomo Traversa
Corresponding author e-mail: giacomo.traversa@isp.cnr.it
Ablation areas in Antarctica are of great interest for studying the interactions between ice, snow and the atmosphere, as well as for predicting future sea level rise, in consideration of their relevance on the continent surface mass balance. In this study, we analyse hyperspectral satellite imagery (e.g. PRISMA) and ground-based spectrometer measurements of snow and ice reflectance over the Nansen Ice Shelf in Northern Victoria Land, Antarctica, with a focus on the ablation areas. These ablation areas are the blue ice areas, which are characterized by exposed, ancient ice that has been ablated over time, and the glazed snow, characterized by snowpacks that have been compacted and melted by the sun, resulting in a smooth, glossy surface. Additionally, white ice and sea ice surfaces were also investigated. Hyperspectral satellite imagery was used to derive surface reflectance data for the region, and these values were compared to in situ measurements obtained using handheld spectroradiometers. These ground-based measurements were collected during the 38th Italian Antarctic Expedition in the 2022/23 austral summer. In addition, we also compared our results with those obtained from multispectral satellite data (e.g. Sentinel-2). Specific spectral indices have been evaluated and identified to characterize the spectral signature of the ice and snow surfaces of interest. Our preliminary results show that satellite-derived reflectance values were consistent with ground-based measurements. The comparison with multispectral data provided additional insights into the reliability and consistency of our findings. Different ice and snow targets showed distinct hyperspectral signatures, making the spatial and temporal monitoring of these components possible from remotely sensed observations. In conclusion, our preliminary findings suggest that satellite-derived reflectance data, in combination with ground-based spectrometer measurements and spectral indices, can provide a valuable tool for monitoring and identifying different ablation zones of Antarctica, particularly in the areas of blue ice, white ice, glazed snow and sea ice, supporting current studies focused on the continent surface mass balance and prediction of future sea level rise.
Francisco Navarro, Cayetana Recio-Blitz, Jaime Otero, Kaian Shahateet, Ricardo Rodríguez-Cielos, María Isabel de Corcuera, Francisco Machío
Corresponding author: Francisco Navarro
Corresponding author e-mail: francisco.navarro@upm.es
The geodetic mass balance (GMB) of glaciers is estimated by differencing two DEMs separated by a given period (say, 10 years) and converting volume change to mass change using an assumption for the density of the material lost/gained. We have computed the GMB of Hurd (land-terminating) and Johnsons (tidewater) glaciers, Livingston Island, South Shetland Islands, Antarctica, for the periods 1957–2000, 2000–13 and 2013–19. This has allowed us to monitor the long-term mass-balance evolution of such glaciers in the context of global and regional climate changes. The DEM for 1957 was built from aerial photogrammetry. The DEM for 2000 was obtained combining classical surveying techniques (theodolite plus laser distance ranger) and GNSS measurements on the glacier surface. The latter technique was also used to produce the DEM for 2013. Finally, the DEM for 2019 was constructed by photogrammetric techniques from very-high-resolution optical images taken by the Pléiades satellite constellation (two satellites with an offset of 180°). The resulting GMBs averaged over the whole glacier extent have been: a) for Hurd Glacier, of –0.27 ± 0.09 (1957–2000)–0.21 ± 0.08 (2000–13) and 0.012 ± 0.074 (2013–19) m w.e. a–1; and b) for Johnsons Glacier, of –0.16 ± 0.09 (1957–2000)–0.14 ± 0.10 (2000–13) and 0.001 ± 0.140 m w.e. a–1 (2013–19). A trend towards less negative, reaching around zero GMBs can be observed. This is attributed to the regional cooling period that occurred in the northern Antarctic Peninsula and the South Shetland Islands during the first two decades of the current century, following a sustained warming period during the second half of the 20th century. This regional cooling was particularly intense during the period 2010–16, and seems to have come to an end. Regarding the spatial distribution of the mass changes, the losses have concentrated at the lowest elevations, while the gains have dominated at the upper elevations, thus suggesting a slow build-up of dynamic instability.
Dynamics and steady-state stability of marine ice sheets with arbitrary basal and lateral shears
Olga Sergienko, Marianne Haseloff
Corresponding author: Olga Sergienko
Corresponding author e-mail: osergien@princeton.edu
The existing analyses of marine ice sheets have been performed for specific sliding laws and, for the case of buttressed ice sheets, specific lateral shear stress parameterizations. However, results of numerical studies show that different sliding laws produce different behaviours. Here, we develop a framework to investigate these behaviours with basal and lateral stresses described by general functional forms (i.e. the functional dependencies are not specified). This framework provides insights into how the basal and lateral shears affect steady state configurations and their stability. The derived expression for the rate of the grounding line migration indicate that it is determined by imbalance between stresses on the grounded side (the basal, lateral and form drag associated with the shape of the bed), the ice-shelf buttressing and the net accumulation/ablation at the grounding line. The fact that the rate of the grounding line migration is determined by the imbalance of several terms, i.e. it is controlled by the competing effects of several processes, indicates that the observed migration of present-day grounding lines cannot be attributed to a single cause (e.g. submarine melting), and all other processes have to be taken into account.
Glacier sliding modulated by cavitation
Ian Hewitt, Gonzalo Gonzalez de Diego
Corresponding author: Ian Hewitt
Corresponding author e-mail: hewitt@maths.ox.ac.uk
Classical theories of glacier and ice-sheet sliding on a hard bed assume that a very thin layer of water lubricates the ice–rock interface, and that the formation of water-filled cavities in the lee side of bedrock bumps results in a crucial role for subglacial water. Macroscopic friction laws generally incorporate this – if at all – via a dependence on the effective pressure, an often vaguely defined quantity that represents a difference between ice and water pressures averaged over some area. Two related issues with the classical theory are that the ability of water to move in and out of cavities is often limited by the (lack of) connections between them, and that the water pressure (and ice pressure) vary significantly in space and time. In this work we re-visit a detailed model of the cavitation process, and examine the influence on macroscopic friction (the relation between shear stress and sliding speed) of different forcing conditions constrained by subglacial water volume or pressure variations. We discuss the relation to other recent experimental and modelling studies, and conclude with some thoughts about the most appropriate sliding laws to use in large-scale computational models.
Sophie Nowicki, Mikayla Pascual, Ginny Catania, Beata Csatho, Isabel Nias, Hui Gao, Denis Felikson
Corresponding author: Denis Felikson
Corresponding author e-mail: denis.felikson@nasa.gov
Projecting ice sheet contribution to future sea-level change is a challenge, in part because simulations of historical ice sheet change do not always match observations. Much of this mismatch is due to uncertainties in processes at the ice sheet edges, namely outlet glacier frontal ablation and basal sliding. Uncertainties that arise from parameterizations of these processes lead to uncertainties in projections of future sea-level change from ice sheet model simulations, such as those from the Ice Sheet Model Intercomparison Project for CMIP6 (ISMIP6), used in Intergovernmental Panel on Climate Change (IPCC) Assessment Report 6 (AR6). Here, we use Bayesian calibration to refine uncertainties in calving and sliding parameterizations in simulations of the Greenland Ice Sheet (GrIS). In this approach, we create a GrIS model ensemble, initialized in 2007, by sampling uncertainty in ice flow, calving and sliding parameterizations. During the hindcast (2007–present), we use model-observation residuals of mass change, dynamic thickness change derived from surface elevation change, velocity change and terminus position change to calculate the likelihood of the parameterizations. We then use the likelihood to calculate the posterior (‘calibrated’) probability distribution for each parameterization. We investigate both regional and ice-sheet-wide model simulations to explore structural model uncertainty. The overarching goal of this work is to improve historical ice sheet model simulations that can inform the experimental design of projects that will build on ISMIP6 and, ultimately, be used in the next generation of IPCC sea-level projections.
Supratim Guha, Reet Kamal Tiwari
Corresponding author: Supratim Guha
Corresponding author e-mail: supratim.guha2010@gmail.com
A complex interplay of several dynamical processes and surface factors affects the heterogeneity of glacier response patterns at the basin scale. The current study has utilized a multivariate linear regression model to examine the causes of the heterogeneous thickness change in the Central Himalayan and Western Himalayan regions independently. The topographical and morphological parameters were initially chosen from existing literature, i.e. glacier area, terminus type, debris cover extent, max elevation, mean elevation, minimum elevation, slope and aspect of the glacier. The most important topographical and morphological factors for each region were then chosen using backward stepwise subset selection techniques. Finally, least square methods were used to calculate the contribution of the chosen parameters. Our analysis indicates that, in the Central Himalayas, the variation of mean slope and debris cover on the glaciers alters the glacier thickness change. A 10% increment of debris cover is associated with 0.37 m a–1 thickness gain for a given slope. On the other hand, if the steepness of the slope surges by 10%, then a 0.86 m a–1 thickness change can be seen for a given debris cover in the Central Himalayan region. In the case of the Western Himalayan region, glacier terminus type, debris cover, mean elevation and slope significantly impact the glacial thickness changes. The thinning of glaciers is approximately 0.25 m a–1 higher in the lake-terminating glaciers compared to the land-terminating ones. The glaciers coated with 10% more debris than its counterpart lose 0.015 m a–1 in thickness for a given slope and elevation. Moreover, glaciers located 1000 ’m higher in elevation experience 0.22 m a–1 less change in thickness and a glacier with a 10% stiffer slope is associated with 0.19 m a–1 less thickness change than its counterpart.
Stress-dependent rheology of warm, watery ice in tertiary creep
Collin Schohn, Neal Iverson, Lucas Zoet
Corresponding author: Collin Schohn
Corresponding author e-mail: cschohn@iastate.edu
Models indicate that zones of temperate ice in the West Antarctic and Greenland ice sheets can develop in response to high rates of shear heating either at depth in ice-stream margins or near the bed. Liquid water residing at grain edges is modeled to commonly exceed 1% by volume, in agreement with limited measurements of water content in temperate glaciers. Although such warm, watery ice is thought to play a major role in fast glacier flow, no laboratory experiments have been conducted into tertiary (steady) creep with temperate ice that contains more than 1% water. A single earlier study in which ice was sheared to tertiary creep but at water contents less than 0.8%, indicated ice softening by water and a stress exponent of n = 3 in Glen’s flow law. Here we use a large ring-shear device to shear, either at a constant shear stress or strain rate, lab-made polycrystalline ice at its pressure-melting temperature and in confined compression until a steady state is achieved in tertiary creep. The ice ring is 0.9 m in outside diameter and ~0.16 m thick, allowing realistic mean grain sizes to be studied (e.g. ~4 mm). Water content (1.2–2.0%) is controlled with the confining pressure and is measured calorimetrically. To date, results indicate nearly linear-viscous deformation (n ~1.2) at stresses less than 180 kPa and nonlinear flow at higher stresses up to 240 kPa with n ~3.5. Thus, two stress-dependent, micro-deformation mechanisms are inferred for watery ice: pressure-melting and refreezing at grain boundaries (n ~1) and dislocation creep (n ~4). We speculate that at lower water contents regelation is impeded by thinner water films at grain boundaries that limit water flow and thereby inhibit pressure melting and refreezing.
Quantifying ocean forcing on Greenland’s tidewater glacier termini
Dominik Fahrner, Aman KC, Ellyn Enderlin, David Sutherland, Till Wagner, Peter Nienow, Femke de Jong, Sophie Nowicki, Donald Slater, Michael Wood, Krisitan Kjeldsen, Claudia Cenedese
Corresponding author: Dominik Fahrner
Corresponding author e-mail: dfahrner@uoregon.edu
Tidewater glaciers in Greenland have been retreating, thinning and accelerating since the mid-1990s and contribute between 30% and 60% of the annual mass loss from the Greenland Ice Sheet (GrIS) through terminus ablation. Terminus ablation is thought to be significantly influenced by ocean forcing, yet both parameters are still poorly constrained for Greenlandic tidewater glaciers as previous studies have been spatio-temporally limited. Here the GRISO Ocean Forcing Ice working group presents a multi-decadal terminus ablation dataset for 56 tidewater glaciers that span all areas of the GrIS as well as fjord ocean temperature time series compiled from a number of observational datasets. The tidewater glaciers are selected based on the reliability of bathymetry data and use terminus delineations from extensive existing data. We account for surface elevation change by using elevation change rates when available, and linear interpolation for dates before and after. The difference between the volume flux towards the terminus and terminus volume change between consecutive observations is used to calculate monthly terminus ablation. The terminus ablation data are correlated to ocean temperature on an individual glacier basis, as well as in groupings based on bathymetry, to determine the ocean forcing on terminus ablation. The results deepen our understanding of the influence of changes in ocean temperature on tidewater glacier termini across the edges of the GrIS. This work highlights the benefits of international scientific cooperation focused on Greenland’s coastal margins and is aimed at improving parametrizations of ocean forcing on Greenlandic tidewater glaciers to lead to more accurate predictions of ice sheet change.
Constraining subglacial properties from time-dependent data
Marianne Haseloff
Corresponding author: Marianne Haseloff
Corresponding author e-mail: mhaseloff@wisc.edu
The basal motion of glaciers and ice sheets depends on the geological, thermal and hydrological conditions at the base of an ice sheet. Due to the complexity of the processes involved, temperate sliding on both hard and soft beds is commonly parameterized by sliding laws that link the basal shear stress to the ice velocity and subglacial conditions, for example the effective pressure. A large variety of basal shear stress parameterizations exist, and projections of ice sheet evolution show a large sensitivity to the chosen parameterization. As remotely sensed data is becoming increasingly spatially and temporally dense, it may provide a way to constrain sliding laws, thereby giving insights into the subglacial conditions. Using synthetic data sets for a range of different subglacial conditions, we show that time-dependent information can be used to gain additional information about the form of the sliding law, for example allowing us to determine two unknown spatially variable coefficients (m and C) in a Weertman-style sliding law (τ = C um). We apply this methodology to observed velocity data from Pine Island Glacier to determine the most probable form of the sliding law for this region.
Constraining subglacial till properties from acoustic techniques
Lucas Zoet, Dougal Hansen, Atsuhiro Muto, Neal Lord, Peter Sobol
Corresponding author: Lucas Zoet
Corresponding author e-mail: lzoet@wisc.edu
Active source seismic analyses have been used widely in glaciology to infer deformational properties of basal till beneath ice streams. Subglacial effective pressure, N, is one of the key parameters required for estimating till deformation and thus glacial motion but is notoriously hard to measure. Common techniques for estimating N have been the labor-intensive practice of measuring it directly from boreholes and connected moulins or in some cases inferring it from surface-velocity inversions. Acoustic wave theory indicates that the N of water-saturated granular sediments, like subglacial till, is related to the seismic-wave propagation velocity and thus the acoustic impedance and seismic reflection amplitude. This theory is mildly sensitive to the grain-size distribution of the sediment but quite sensitive to N and porosity of the sediment. The canonical view in glaciology is that the reflection amplitude of the seismic wave primarily conveys changes in the porosity of the substratum, and therefore has been used to infer deeply deforming till layers. However, the fundamental acoustic principals of wave travel through a saturated porous medium indicate that N can also have primary control on the wave speed and, by extension, reflection amplitude. In environments where porosity may not vary greatly, like critically strained tills, it is possible that effective pressure exerts the dominant control on seismic reflection amplitude. To constrain the relative contributions of effective pressure and porosity to the seismic reflection amplitude, we measure acoustic reflection from the ice–bed interface within a cryosphere ring shear device where temperate ice is slid atop a water saturated deformable till layer under a prescribed effective pressure. In this setup, effective pressure of the system can be precisely varied, and porosity can be precisely recorded, while a series of acoustic piezo-electric transducers continually transmit and receive acoustic waves that have been reflected from the ice–bed interface in order to record the reflection properties of the interface. Here we show initial results of the study and how they may affect interpretations of deformation within subglacial tills.
What controls iceberg abundance and distribution in Northwest Greenland?
Allie Berry, Kristin Schild, Siddharth Shankar, Leigh Stearns
Corresponding author: Allie Berry
Corresponding author e-mail: allison.berry1@maine.edu
Icebergs can account for up to 50% of freshwater flux from tidewater glaciers, but are not currently well constrained in global circulation models (GCMs) . As icebergs calve into the ocean, they not only contribute to global sea level rise but their localized freshwater injection can alter fjord circulation as they move and melt. The rate of mass loss from polar regions has been increasing, so determining the impact of icebergs on broader oceanic circulation is increasingly important. Inherently, their impact depends on their size and spatial distribution; however, despite this understanding, GCMs do not accurately account for the freshwater flux generated by icebergs. This knowledge gap is in part due to the challenge of predicting iceberg abundance and size distribution, which is needed to properly inform the models. Here we examine the environmental controls on iceberg abundance and distribution around Northwest Greenland, spanning the years 2017–21. Iceberg metrics are obtained by applying an automated detection algorithm to Sentinel-2 imagery, and comparing against a suite of environmental and glaciological variables through a random forest analysis. In this presentation, we will highlight the main environmental and glaciological controls on iceberg distribution in northwest Greenland. Being able to better predict iceberg abundance and distribution will help to constrain the impact of freshwater flux on global ocean circulation.
Investigating calving dynamics at Jakobshavn Isbræ using a 3D full-Stokes calving model
Iain Wheel, Anna Crawford, Doug Benn
Corresponding author: Iain Wheel
Corresponding author e-mail: iw43@st-andrews.ac.uk
Despite its importance, calving across Greenland remains poorly constrained. This is in part due to our lack of understanding of the physical processes involved along with our inability to model observations. We have developed a new 3-D full-Stokes calving algorithm in Elmer/Ice which overcomes previous technical hurdles. This new algorithm allows limitless advance and retreat of the glacier front along with fully 3-D non-projectible calving. Using this new algorithm with a position-based crevasse depth calving law we investigate the controls on calving at Jakobshavn Isbræ during 2016/17. This represents the most detailed implementation of a physically-based calving law in a continuum model. Buttressing from a proglacial ice mélange is key to winter readvancement but other factors such as basal conditions and submarine melting are considered. The model can accurately represent seasonal terminus positions despite the large uncertainties in melt and ice melange conditions.
Diverse glacier surge dynamics in Svalbard suggest complexities in underlying processes
Adrian Luckman, Doug Benn, Ian Hewitt
Corresponding author: Adrian Luckman
Corresponding author e-mail: a.luckman@swansea.ac.uk
The Sentinel-1 spaceborne imaging radar has, since its launch in 2014, allowed glacier and ice sheet ice velocity to be monitored in unprecedented temporal and spatial detail. Measurements every 12 days (6 days from 2016–21 while Sentinel-1B was operational) of ice speed greater than around 0.5 m d–1 have made it possible to detect even subtle changes in glacier speed. In Svalbard alone, dozens of glacier surges have been detected and monitored during this period, allowing comparisons to be made between patterns of surge initiation and development. The variety in timing and style of ice acceleration and deceleration, sensitivity to meltwater, and length of surge suggests a diversity of detail in the underlying processes. Here we will present velocity data for a number of surges and discuss potential reasons for the differences between them.
Can we use ApRES to calculate iceberg ablation rates?
Kristin Schild, Irena Vankova, David Sutherland
Corresponding author: Kristin Schild
Corresponding author e-mail: kristin.schild@maine.edu
The increasing input of freshwater to the subpolar North Atlantic, both through glacier runoff and the melting of calved icebergs, has significant implications for fjord stratification and circulation. However, constraining this meltwater input has been difficult due to complex iceberg geometry and subsequently complex melt behavior. While a number of in situ and remote sensing approaches have been developed to measure or compute iceberg deterioration, many are challenged by limited data availability, unknown subsurface iceberg geometry and few opportunities for in situ validation. Here we use a combination of surface and subsurface in situ field measurements, collected during an intensive 2019 summer field campaign in Sermilik Fjord, southeast Greenland, to test the applicability of using an on-iceberg ApRES to quantify depth-dependent and geometry-dependent iceberg melt rates. We calculate comparison ablation rates using repeat ship-based multibeam subsurface surveys, repeat drone surveys and tandem on-iceberg GPS units. For the same iceberg over the same period of time, we find an order of magnitude difference in calculated ablation rates across the five methods and a division in depth-variable rates dependent upon the inclusion of mechanical deterioration alongside melt in the ablation calculation. This study pushes the limits (is on the edge) of field instrumentation and highlights the importance of considering mechanical processes and method limitations, in calculating iceberg ablation rates.
Ice Shelf aquifers of melt and brine: a lot to explore
Ted Scambos, Julie Miller, Riley Culberg, David Long
Corresponding author: Ted Scambos
Corresponding author e-mail: tascambos@colorado.edu
We discuss two types of ice shelf firn aquifer present in coastal Antarctica: melt-derived and brine-infiltration, and explore questions about their characteristics, formation, recharge and impacts on ice shelf stability. Extensive melt aquifers have been identified and confirmed by field work on the Wilkins Ice Shelf and Mueller Ice Shelf, and satellite data indicate a further extensive melt aquifer on the northern George VI Ice Shelf. We also present an improved preliminary assessment of the extent of brine aquifers on Antarctic ice shelves based on their characteristics in airborne ice-penetrating radar profiles. Brine aquifers are relatively widespread and are found in shelf areas with porous firn at the waterline, as was previously recognized by Cook et al.(2018). We assess the relative risk of hydrofracture for ice shelves bearing melt and brine aquifers under various scenarios of firn density, aquifer depth and brine density. We find that ponded surface melt poses the greatest threat but that the increased density of brine may also be an issue if the brine layer is relatively shallow in the shelf. We also discuss the likely future expansion of melt aquifers on ice shelves under warming conditions and consider the process of transitioning from a brine aquifer to a melt aquifer.
How do suture zones allow ice shelves to grow so large? Introducing the RiPIce Project
Adrian Luckman, Suzanne Bevan, Steph Cornford, Bryn Hubbard, Glenn Jones, Bernd Kulessa, Katie Miles, Eduardo de Souza Neto, Sarah Thompson
Corresponding author: Adrian Luckman
Corresponding author e-mail: a.luckman@swansea.ac.uk
Modelled projections of the contribution of the Antarctic ice sheets to sea level rise over the coming century vary from a few centimetres to more than one metre – a substantial uncertainty that undermines the credibility of sea-level rise projections. The reasons for this uncertainty lie in the treatment of ice shelves: assuming that they will disintegrate leads to a much higher estimate of ice discharge than assuming they remain in place. No forecast to date has realistically included the processes that lead to ice shelf disintegration. This is because fracture and rift propagation disrupt the normal assumptions of continuity inherent in ice sheet models, while these processes depend critically on poorly known inhomogeneities in ice shelf structure, such as flow bands arising from suture zones (suture bands). The RiPIce project (Rift Propagation for Ice sheet models) aims to explicitly represent heterogeneity in ice shelves and pioneer the inclusion of rift processes in an ice sheet model. The project has collected new field data from a key suture band and made advances in modelling rift propagation in the Larsen C Ice Shelf. This presentation will introduce the project and highlight the role of the suture band in arresting fractures that would otherwise have significantly reduced the size of the ice shelf.
Elastica’s snowy edge: spontaneous roll-up of a ruck in a thin, sliding snow sheet
Douglas MacAyeal, Etienne Reyssat, Mathilde Reyssat, David McClung, Thomas Witten
Corresponding author: Douglas MacAyeal
Corresponding author e-mail: drm7@uchicago.edu
Euler’s Elastica has given rise to a class of challenging mechanical problems involving the large-scale deflections of rods and sheets, both elastic and viscous. The ideas of Elastica apparently apply to an otherwise obscure phenomena in snow-sheet sliding dynamics. This phenomena is the ‘snow roll’, a spontaneous roll-up of a single ruck that develops in a thin snow sheet sliding on a wet glass plate. We present observations, laboratory-scale experiments with elastomers designed to represent thin, wet snow sheets, and perform mathematical analysis using 1-D mechanics. A preliminary understanding of the phenomenon is at odds with some immediate expectations from snow science: the snow sheet must be capable of supporting compressive stress, the adhesive bond between the sheet and the wet glass must be capable of creating a sliding/no-sliding transition, and snow surfaces must self-adhere. The work presented exposes an ‘edge’ in both our physical understanding of snow and the philosophical view of what constitutes ‘understanding’. At large scale, snow is viewed as a granular material that undergoes only a single transition between being at rest and being in a state of chaotic motion: stationary vs avalanche. The snow-roll described here calls for snow to additionally behave in ways that enable intricate, large-amplitude bending deformations without fracture or significant change in local snow-sheet thickness. Philosophically, we cordially debate what constitute the appropriate assumptions and goals of an explanation of the snow-roll phenomenon.
Mikayla Pascual, Ginny Catania, Benjamin Keisling, Douglas Brinkerhoff, John Erich Christian, Sean Gulick
Corresponding author: Mikayla Pascual
Corresponding author e-mail: mikayla.pascual@utexas.edu
Moraines (also termed morainal banks) in marine-terminating outlet glacier settings can provide a feedback mechanism for glacier stability or retreat. However, sediment dynamics in Greenland are poorly understood. Due to limited understanding and observations, sediment dynamics are rarely included in ice flow models, contributing to the uncertainty in Greenland’s future sea level potential. Coupled ice–sediment models show important feedbacks between moraine-building processes and ice dynamics. To advance our understanding of sediment dynamics on ice flow, it is necessary to quantify the impact of sedimentation on Greenland outlet glacier ice volume on different timescales. Here we explore the sensitivity of Greenland outlet glaciers to sediment dynamics including diffusion (removing sediment from moraines) and glaciofluvial sedimentation (adding sediment to moraines) in a flowline model adapted from Brinkerhoff et al., 2017. We run an ensemble of simulations to investigate these processes on multi-millennial timescales, using different bed topography slopes and surface mass balance scenarios, and with sedimentation coupling included or omitted. We find that sedimentation has a strong control on ice volume and creates tidewater glacier cycles with variable amplitudes in time. In addition, we determine the sensitivity of the tidewater glacier cycle to variations in bed topography and surface mass balance. The coupling between sediment and ice dynamics could explain and contribute to the divergent glacier behavior presently seen in Greenland and other glaciated margins. This work provides insights applicable to ice sheet model development and field-observation efforts aimed at understanding the rates and processes driving sedimentation in Greenland and marine-terminating glaciers globally.
Wilson Sauthoff, Matthew Siegfried, Benjamin Smith
Corresponding author: Wilson Sauthoff
Corresponding author e-mail: sauthoff@mines.edu
Subglacial water systems and deeper groundwater networks beneath the Antarctic ice sheet connect glacial and oceanic systems, modulate ice dynamics and remain a major physical uncertainty of future ice sheet predictions. Subglacial water collects in subglacial lakes, some of which are ‘active’ lakes that episodically fill and drain inferred from ice-surface uplift and subsidence respectively. Active lake drain–fill cycles generate time-varying evolution of subglacial water distribution, transport mechanisms and freshwater flux into sub-ice-shelf cavities and the Southern Ocean. When these lakes were initially discovered in the 2000s, the coarse spatial resolution and short temporal duration of available Ice, Cloud, and Land Elevation Satellite (ICESat) observations limited comprehensive knowledge about how lake geometry changed through fill–drain cycles. This limited view resulted in uncertain and stationary lake boundaries representative of a brief observational window when lakes were initially discovered. We have developed and applied a surface-deformation delineation algorithm to dynamically define active subglacial lake ice-surface outlines using height anomalies in high spatial resolution and sampling density ICESat-2 data. We observe large variability in area at previously established active subglacial lakes with some exhibiting newly observed lobes. Compared to the typical approach of using static lake outlines for volume estimates, we demonstrate that subglacial lake volume fluxes are underestimated by up to 85%. Additionally, we detected 77 new ‘lake candidates’ (cf 140 previously documented active subglacial lakes), defined as ice-surface deformation patterns with spatially coherent repeating height change anomalies greater than surrounding surface height change. These height anomalies can be small-scale (<5 km) making attributing causation to subglacial water activity more difficult than large-scale height anomalies. We explore drivers of these height anomalies with datasets of ice velocity, wind redistribution of snow, basal traction changes, and precipitation. Refined estimates of water volume fluxes along with an expanded inventory of subglacial lakes will inform transient subglacial water models with more precise inferences of water flux through the subglacial system, across the land–ocean interface at the grounding zone, and into sub-ice-shelf ‘estuaries’ and the greater Southern Ocean.
The life and death of a subglacial lake in West Antarctica
Matthew Siegfried, Ryan Venturelli, Molly Patterson, William Arnuk, Timothy Campbell, Chloe Gustafson, Alexander Michaud, Ben Galton-Fenzi, Mark Hausner, Stephanie Holzschuh, Bruce Huber, Kenneth Mankoff, Dustin Schroeder, Paul Summers, Scott Tyler, Sasha Carter, Helen Fricker
Corresponding author: Matthew Siegfried
Corresponding author e-mail: siegfried@mines.edu
Over the past 50 years, the discovery and initial investigation of subglacial lakes in Antarctica have highlighted the paleoglaciological information that may be recorded in sediments at their beds. In December 2018, we accessed Mercer Subglacial Lake, West Antarctica, and recovered the first in situ subglacial lake-sediment record – 120 mm of finely laminated mud. We combined geophysical observations, image analysis and quantitative stratigraphy techniques to estimate long-term mean lake sedimentation rates between 0.49 ± 0.12 mm a–1 and 2.3 ± 0.2 mm a–1, with a most likely sedimentation rate of 0.68 ± 0.08 mm a–1. These estimates suggest that this lake formed between 53 and 260 :a before core recovery, with a most likely age of 180 ± 20 :a before core recovery – coincident with the stagnation of the nearby Kamb Ice Stream. Our work demonstrates that interconnected subglacial lake systems are fundamentally linked to larger-scale ice dynamics and highlights that subglacial sediment archives contain powerful, century-scale records of ice history and provide a modern process-based analogue for interpreting paleo–subglacial lake facies.
The relationship between the permeability and liquid water content of polycrystalline temperate ice
Jacob Fowler, Neal Iverson
Corresponding author: Jacob Fowler
Corresponding author e-mail: jrfowler@iastate.edu
To better constrain meltwater transport and ice viscosity in temperate ice, particularly in ice stream shear margins and near glacier beds, we use a custom permeameter to study the untested model relationship between the permeability of temperate ice and its liquid water content. The permeability of lab-made ice of two mean grain diameters (1.8 and 4.2 mm) is measured, and water content is controlled with the ice salinity and measured calorimetrically. Fluorescein dye is added to through-flowing, chilled water to highlight flow pathways through the ice after experiments. As predicted by a simple model, permeability increases with approximately the square of the water content and by about three orders of magnitude across water contents of 0.1–4.4%. However, permeability values are less than those of the model by average factors of 2.6 and 4.1 for the finer and coarser ice, respectively. This discrepancy is probably due to tortuous, truncated or air-clogged veins. The order-of-magnitude agreement between measured and modeled values may indicate that reduced permeability from these factors is nearly compensated by preferential flow in oversized veins that are isolated or arborescent. Both kinds of preferred flow pathway are observed, but arborescence is observed only in fine-grained ice at water contents greater than 2%.
The impact of frozen sediments on basal friction during glacier sliding
Dougal D. Hansen, Lucas Zoet, Aaron Stubblefield, Colin R. Meyer, Alan Rempel
Corresponding author: Dougal D. Hansen
Corresponding author e-mail: ddhansen3@wisc.edu
Frozen sediments are frequently observed attached to the base of glaciers and play a significant role in basal sliding by influencing friction at the ice–bed interface. However, the mechanisms governing this interaction remain poorly understood. Fundamental questions, such as whether the frictional response is attributable to the fringe or the deformable till beneath it, and how ploughing clasts influence this interaction, remain unresolved. To address these inquiries, we conducted a series of laboratory experiments to investigate the impact of frozen sediments on basal friction during glacier slip over till. Using a cryogenic ring shear apparatus, we conducted sliding experiments with rings of temperate ice over saturated granular beds at realistic temperatures, effective pressures and ice velocities. In the absence of fringe, the till underwent plastic deformation, and resistance to slip matched the Coulomb strength of the till past a threshold velocity, which has been demonstrated to mark the transition from form drag to skin friction. Similarly, when the fringe thickness matched or exceeded the diameter of the largest ploughing clasts at the interface, motion occurred in a thin shear band beneath the frozen sediments, which advected as a rigid block above the bed, and till deformation dominated the drag response. For cases in which ploughing clasts were significantly larger than the fringe thickness, preliminary results suggest a viscous response at low velocities. We examine how the magnitude of drag at different ice speeds relates to the rheology of the fringe and its relative thickness. Our results demonstrate the contribution of frozen sediments to basal friction, highlighting their importance in understanding glacier sliding and its implications for landscape evolution and sea-level rise.
Till under glaciers: dilation, pore-pressure feedback and rate-weakening friction
Neal Iverson
Corresponding author: Neal Iverson
Corresponding author e-mail: niverson@iastate.edu
Subglacial till can help activate dramatic styles of glacier flow, including ice streaming, stick–slip and surging. The Coulomb behavior of water-saturated tills sheared at steady porosities is well established experimentally, with till shear strength dependent simply on effective stress and a rate-independent friction coefficient. During transient deformation, however, till can potentially admit a wide range of behavior owing to porosity changes that drive non-hydrostatic pore pressure and associated changes in effective stress. These changes depend sensitively on boundary and initial conditions, till permeability, and the depth of shearing. Transient dilation of till pores is common during the early stages of till shear. In some recent models of basal motion on soft beds, including stick–slip and surging, till strengthens as it dilates because pore pressure decreases – a process called dilatant strengthening. Importantly, in these models hydrological conditions at the ice–till interface are imposed that are assumed to be independent of till dilation. When subglacial till dilates during shear, however, rather than lifting the glacier, till is expected to squeeze into water-filled voids at the bed surface where till is under zero effective stress. Field observations of the surfaces of former till beds indicate that this squeezing of till into voids is common. Resultant shrinkage of these voids will drive water into the bed. I solve a forced pore-pressure diffusion equation across a thickness of subglacial till to determine the time evolution of non-hydrostatic pore pressure that develops as bed-surface voids close during till shear and pores in underlying till expand. Pore pressure is reduced at depth in the bed, causing dilatant strengthening. However, near the bed surface pore pressure increases, so that till weakens and resistance to slip decreases. This till weakening near the bed surface increases with increasing rates of shear. Such rate-weakening friction at glacier beds is a requirement for stick–slip motion and could contribute to the onsets of surges.
Hunting for super slippery surfaces in subglacial environments
Rebecca McCerery, John Woodward, Glen McHale, Kate Winter, David Pearce
Corresponding author: Rebecca McCerery
Corresponding author e-mail: r.mccerery@northumbria.ac.uk
The theoretical concepts of super slippery surface formation in the presence of a roughness structure, hydrophobic chemistry, and/or a lubricating oil are increasingly well understood in the field of material physics. Here, we investigate the ability of sediments in the natural environment to display similar super water repellent (superhydrophobic) properties and form slippery liquid-infused porous surfaces (SLIPS), which we hypothesize will drive instability and fast flow conditions in glacial systems. We tested 22 model (oil-free) hydrophobic and oil impregnated hydrophobic sediment surfaces, measuring water contact angles and droplet sliding angles to classify their degree of hydrophobicity and water shedding ability. Our results show that clay-silt sized particles can support superhydrophobicity and SLIPS, and larger particle sizes can also display extreme water repellence and water shedding abilities. We then searched for evidence of these sediments in the subglacial environment. First, we looked for lubricating oils in 82 sediment samples from the former bed of the Central Alberta Ice Stream (CAIS) flow track in the Laurentide Ice Sheet using gas chromatography–mass spectrometry. The results showed glacial erosion and long-distance mobilization of Alberta Oil Sands deposits from northern Alberta. Second, we looked for hydrophobic biofilms in glacial sediments from Svalbard. In extreme environments microbial life typically exists in biofilms – a flexible matrix composed of extracellular polymeric substances (EPS). 27 sediment samples from 10 glaciers were analysed using 16s rRNA gene sequencing on MiSeq to identify microbial communities, and EPS representative of biofilm production were extracted and analysed using high-performance liquid chromatography, showing evidence of EPS surfactants at the glacier bed. We hypothesize that the presence of oil at the ice–bed interface of the CAIS will have resulted in fast flow dynamics and ice stream instability. Furthermore, we hypothesize that microbes found under Svalbard glaciers act as important components of particle adhesion and sediment stability, with the ability to alter surface physics, thereby influencing glacier flow. Thus, recent breakthroughs in the study of surface physics can help us understand processes occurring at the ice–bed interface, fast flow processes, and glacier instability.
John Woodward, Rebecca McCerery, Glen McHale, Kate Winter
Corresponding author: John Woodward
Corresponding author e-mail: john.woodward@northumbria.ac.uk
Some glacial systems are observed to demonstrate unpredictable flow behaviour, resulting in instabilities such as ice stream switching and surging. Current theoretical models suggest fast flow is driven by sliding on water at the ice–bed interface and/or flow over deforming sediments in the basal till layer. However despite considerable field experimentation and observation, flow models currently struggle to fully predict instability in glacier flow. Here, we present a new school of thought detailing the importance of the micro-scale physics of interfaces on the slipperiness and stability of sediments in subglacial systems. The fundamental concepts of slipperiness in interface physics allow the formation of super slippery surfaces through the combination of: (i) roughness structure, (ii) hydrophobic surface geochemistry, and/or (iii) the presence of a lubricating substance such as oil. These super slippery surfaces display extreme water repellence and water shedding properties. Using the theoretical physical principles of slipperiness and applying them to soft-bed subglacial environments, we propose three modes by which subglacial sediment at different particle sizes could become super slippery; (i) a hydrophobic surface chemistry from pre-existing organic material, (ii) the presence of microbial biofilms, or (iii) the incorporation of a lubricating natural oil. We then present theoretical models of ice flow on super slippery sediments, which we hypothesize enhances sliding and deformation towards infinite flow velocity. By understanding the physics occurring at the ice–bed interface it is possible to better predict glacier flow conditions and also to predict internal instability in the flow mechanism, explaining, for example, surging behaviour. It is therefore critical that properties affecting wettability and water shedding of sediments, such as sediment geochemistry and microbiology, are considered in our understanding of flow instability in glaciers and ice sheets.
Discovering the rheology of Antarctic ice shelves via physics-informed deep learning
Yongji Wang, Ching-Yao Lai
Corresponding author: Yongji Wang
Corresponding author e-mail: yw1705@princeton.edu
Ice flows in response to stresses according to the flow law that involves ice viscosity. An accurate description of effective ice viscosity is essential for predicting the mass loss of the Antarctic Ice Sheet, yet measurement of ice viscosity is challenging at a continental scale. Lab experiments of polycrystalline ice show that the effective viscosity of ice obeys a power-law relation with the strain rate, known as Glen’s flow law. However, it remains unclear how processes at ice-shelf scale impact the effective viscosity of glacial ice. Here, we leverage the availability of remote-sensing data and physics-informed deep learning to infer the effective ice viscosity and examine the rheology, i.e. flow lawof glacial ice in Antarctic ice shelves. We find that the rheology of ice shelves differs substantially between the compression and extension zones. In the compression zone near the grounding line the rheology of ice closely obeys power laws with exponents in the range 1 < n < 3, consistent with prior laboratory experiments. In the extension zone where fractures exist, which comprises most of the total ice-shelf area, ice performs complex rheological behaviors. Our result highlights a need to examine the impact of ice-shelf scale processes on glacial rheology.
How swell instigated the Wilkins and Voyeykov calvings in 2007/08
Nathan J. Teder, Luke G. Bennetts, Rob A. Massom, Jordan P.A. Pitt, Phil A. Reid, Theodore A. Scambos, Alexander D. Fraser
Corresponding author: Nathan J. Teder
Corresponding author e-mail: nathan.teder@adelaide.edu.au
One potential driver of global sea level rise is through the loss of ice shelves, as the decreased buttressing effect can cause the adjacent ice sheet to speed up its outward flow into the sea. A significant period of ice shelf loss occurred on the Wilkins Ice Shelf, which underwent three separate calving or disintegration events in 2008, leading to the removal of over 2000 km of ice from the continent. Of these three events, two were in a period during which the section of the shelf was exposed to ocean swell and the third occurred alongside the removal of a significant layer of perennial fast ice, in a manner akin to the calving of the Voyeykov Ice Shelf a year prior. Our investigation into these events showed that the length of pack ice protecting the shelves had contracted earlier in the season. This allows for damaging ocean strains to influence the outer margin of the shelf and – if present – the perennial fast ice layer for a longer period. If this occurs when there is a weakened perennial fast ice layer, or outer margin of the shelf, these strains can cause the ice to reach a point where it undergoes a cyclic failure and breaks down. When paired with a weakened shelf due to rifts/retreat/thinning or other forms of fatigue, this can instigate a significant calving event with the potential to disintegrate an ice shelf.
Peter Nekrasov, Douglas MacAyeal
Corresponding author: Peter Nekrasov
Corresponding author e-mail: pn3@uchicago.edu
The Ward Hunt and Milne ice shelves, remnants of the former Ellesmere Ice Shelf, feature an unusual roll morphology whose origin is unknown. A classic theory for the origin of the rolls involves the elongation of surface meltwater ponds by unidirectional winds, while a more recent theory involves the viscous buckling of a sheet of ice due to sea-ice pressure. Not only are these theories at odds with the climate record and field observations of Arctic sea ice, they both presuppose the prior existence of an intact ice shelf upon which to initiate roll-building. For this reason, we consider roll development as a process occurring simultaneously with the formation of the ice shelf from thickening multiyear landfast sea ice. We propose a mechanism for roll formation that differs from pre-existing theories by examining how ocean waves interact with an emerging roll morphology. Using numerical models, we demonstrate that surface rolls protect a nascent ice shelf by producing frequency band gaps which inhibit ocean waves from entering and damaging the ice shelf. As sea ice undergoes accumulation and ridging, ridges that are not aligned with impinging ocean waves may be destroyed, while ridges that are oriented correctly may be preferentially preserved as rolls. We utilize stochastic models to test whether surface features resembling rolls may stabilize the ice shelf as it grows and become the dominant characteristic of the ice shelf that is eventually built. We ultimately argue that this long-term preference for roll morphology must also be significant to the breakup of the ice shelf, and that the gradual loss of surface rolls will become a tipping point for ice shelf instability.
Doug Benn
Corresponding author: Doug Benn
Corresponding author e-mail: dib2@st-andrews.ac.uk
Calving laws fall into two broad categories: rate laws, which predict the rate at which ice is lost from the front, and position laws, which define the location of the calving front and treat the calving rate as a function of ice velocity and change in frontal position. Rate laws are often chosen for numerical convenience but in many circumstances they do not reflect how calving actually works. Here, I show the crevasse-depth (CD) calving law has both a sound theoretical basis and demonstrable skill in predicting ice-front positions. The CD calving law predicts ice-front position from the penetration of crevasses, calculated using the ‘zero-stress’ approach pioneered by John Nye and generalized to 3-D by Joe Todd. In its most widely used form, calving is predicted wherever crevasses isolate portions of the glacier terminus, either via the meeting of surface and basal crevasses or via surface crevasses reaching the waterline. A number of issues have been raised concerning the rationale behind the CD law and its implementation. The Nye–Todd crevasse-depth function neglects stress-concentration effects and inheritance of damage from upglacier, and several studies have shown the need to adopt an unphysical ‘crevasse water depth’ parameter for tuning. Recent work has addressed these problems, using a combination of observations and modelling in Elmer/Ice and HiDEM. (1) Predicted crevasse depths tested against UAV-mounted LiDAR data from Tunabreen, Svalbard, show an excellent match between observed maximum crevasse depths and those predicted by Elmer/Ice. Inheritance is not an issue where strain rates increase towards the terminus. (2) Relationships between stresses and ice-front position at Store Glacier show that the system exhibits self-organized criticality. The CD law locates the threshold between sub-critical and super-critical system behaviour at a compressive arch between lateral pinning points, supporting a position law approach. (3) Improved calving routines applied to Jakobshavn Isbræ indicate that full-depth calving criteria are too stringent. Assuming fractures become critical before full penetration removes the need for unphysical model tuning. It is important to ask what a calving law is expected to do. Calving is a stochastic process but regularities emerge at large enough spatial and temporal scales. The CD law predicts these emergent behaviours and should be considered a leading candidate for use in ice-sheet models.
Katie E. Miles, Bryn Hubbard, Adrian Luckman, Bernd Kulessa, Sarah Thompson, Glenn Jones, Stephen Cornford, Eduardo de Souza Neto, Suzanne Bevan
Corresponding author: Katie E. Miles
Corresponding author e-mail: kam64@aber.ac.uk
The response of land-based glaciers to the loss of their buttressing ice shelves is an important source of uncertainty in projections of Antarctic mass loss. Rift propagation across ice shelves leads to mass loss through iceberg calving that, in extreme cases, can lead to ice shelf disintegration (e.g. Larsen A in 1995 and Larsen B in 2002). Antarctic ice shelves comprise alternating flow-parallel bands of cold, hard meteoric ice supplied from inland glaciers, and warm, soft suture zone ice supplied from the sub-shelf cavity. Rifts propagate quickly through meteoric ice, but that progress is arrested in suture zone bands. Despite this important role, little is known about either the material properties or the mechanical processes involved in the arrest process. In the austral summer of 2022, we drilled, logged (by video, optical televiewer and sonic logger), and instrumented (for temperature) three boreholes in a meteoric and a suture band on Larsen C Ice Shelf, Antarctica, to characterize the subsurface properties and structure. Initial results reveal: i) that the suture band ice is 2–3°C warmer than the meteoric ice; ii) a large number of sub-horizontal layers (probably primary stratification and/or infiltration ice) throughout the uppermost sections of both bands; iii) steeply dipping ice layers at depth in both bands; and iv) ice formed from platelets at depth in the suture ice band, absent from the meteoric ice band. We combine these properties to propose a conceptual model of suture band formation and structure.
Continuum damage modelling of ice shelf rift propagation and arrest
Glenn Jones, Stephen Cornford, Eduardo De Souza Neto, Suzanne Bevan, Adrian Luckman, Bryn Hubbard, Bernd Kulessa, Katie Miles
Corresponding author: Stephen Cornford
Corresponding author e-mail: S.L.Cornford@bristol.ac.uk
Ice shelves affect the flow of upstream land ice and their evolution is a major source of uncertainty in projections of ice mass loss from Antarctica. Fracturing and rifting play a key role in the evolution of ice shelves, whether through iceberg calving or wholesale ice shelf disintegration (e.g. Larsen A in 1995 and Larsen B in 2002). Many ice shelves, including Larsen C Ice Shelf (LCIS), comprise alternating flow-parallel bands of relatively resistant meteoric ice and softer suture zone ice, with full thickness rifts propagating more rapidly through the former than through the latter. Here, we develop a continuum damage, finite element model of rift propagation where a damage variable represents the continuous degradation of the material until fracture is complete. The model successfully predicts that rifts propagate more slowly through softer ice, due to the diminished accumulation of damage. Applying the model to LCIS reveals that the presence of flowbands both inhibits rift propagation within the suture zones and enhances rift propagation in the adjacent meteoric ice.
Bernd Kulessa, Sarah Thompson, Adrian Luckman, Katie Miles, Bryn Hubbard, Suzanne Bevan, Alex Brisbourne, Stephen Cornford, Eduardo De Souza Neto, Glenn Jones, Siobhan Killingbeck, Rebecca Schlegel
Corresponding author: Bernd Kulessa
Corresponding author e-mail: b.kulessa@swansea.ac.uk
Suture zones are present in all large and numerous smaller Antarctic ice shelves, stabilizing them by containing rifts within meteoric ice units derived from tributary glaciers. The zones’ marine-derived constituents (mélange) contain seawater and are warmer than the surrounding meteoric ice, allowing them to arrest rifts by accommodating strain and preventing brittle fracture. To improve our understanding of the internal structure and physical properties, we acquired several new geophysical data sets on and around the Joerg Peninsula suture zone in the southern Larsen C Ice Shelf, Antarctic Peninsula, in the 2022/23 austral summer. Key sections of a ground-penetrating radar (GPR) grid acquired in 2009/10 were re-surveyed to delineate and map the melange and meteoric ice within the suture zone. Together with the GPR data, a seismic reflection survey conducted near the centre of a 10 km long profile across the suture zone, ~20 km downstream of the peninsula’s tip, revealed that the zone is ~ 270–280 m thick in the absence of meange, and a few tens of metres thinner where melange is present. With its upper surface detected at depths of ~70 ± 10 m along the profile, the melange is accordingly inferred to be greater than 100 m thick. Transient electromagnetic (TEM) soundings were acquired every 500 m along the profile to image the suture zone’s electrical conductivity structure so that, together, our TEM, GPR and seismic data reveal the suture zone’s anatomy along the 10 km long profile. Our new observations extend our understanding of the properties and 3-D spatial evolution of the Joerg Peninsula suture zone, as inferred previously from earlier seismic and radar geophysical surveys conducted in 2008/09 and 2009/10.
Valeria di Biase, Peter Kuipers Munneke, Bert Wouters
Corresponding author: Valeria di Biase
Corresponding author e-mail: v.dibiase@uu.nl
In the past decade, localized in-situ observation have revealed the presence of firn aquifers on the Antarctic Ice Sheet. The main objective of this large-scale study is, for the first time, to detect aquifer presence in the entire Antarctic region by using a combination of remotely sensed and climate modeling data. The existence of firn aquifers may have an impact on our understanding of the water and energy budget, as they are believed to be an important component of the hydrological system of the ice sheet. The presence of firn aquifers in Antarctica has been confirmed in observational studies on the Wilkins Ice Shelf and on the Muller Ice Shelf, and has been reproduced by regional firn models. Currently, no large-scale observational studies exist in the Antarctic region, hampering a detailed assessment of their contribution to runoff and sea-level rise. In this study, we present a probability map of firn-aquifer presence around the Antarctic Ice Sheet, based on a combination of remote sensing and climate modeling data. Radar imagery from the Sentinel-1 and Advanced Scatterometer (ASCAT) missions, together with climate data from the regional atmospheric climate model RACMO2.3p2, are combined to map the probability of detecting firn aquifers in the period 2017–20. Our method is based on Monte Carlo simulations: a dedicated algorithm predicts the probability of aquifer presence based on a set of fixed inputs, to which dedicated thresholds and weights are assigned. Sentinel-1A/B, ASCAT and Racmo2.3p2 (liquid water content; accumulation; melt; surface mass balance) data have been used as inputs to predict the presence of aquifers. The Monte Carlo methodology has been applied in the Antarctic Peninsula and, thanks to the large spatio-temporal availability of the input datasets, the presence of aquifers in selected regions of Antarctica has been estimated. In agreement with observations from previous studies, we find a high probability of firn aquifers presence in the north and northwest coast of the Antarctic Peninsula, and on the Wilkins and George VI ice shelves. Outside the Antarctic Peninsula, no regions with significant aquifer probability (>50%) have been found. The methodology has been successfully validated within Greenland, where detailed aquifers data retrieved by the Operation Ice Bridge (OIB) mission (2010–14) are available.
Andrew Fowler, Mike Chapwanya, Iain Moyles, James Fannon
Corresponding author: Andrew Fowler
Corresponding author e-mail: andrew.fowler@ul.ie
The instability theory of drumlin formation has its roots in the work of Richard Hindmarsh in the late 1990s. His basic idea was that drumlins form through an instability due to the shearing motion of ice flowing over a deformable subglacial till. The ingredients of the theory are thus ice flow and till flow. Later, water flow was added. The development of the theory beyond the basic linear instability result was initially hampered by a concatenation of difficulties: these include till rheology, two-dimensionality and cavitation.The numerical solution of the model is also fraught with complication. In this sequence of short presentations, four of the subsequent protagonists will describe and illustrate their efforts to resolve these and other difficulties as they arose over the course of the last 20 years, and a summary of the way in which the theory needs to progress will be outlined.
A reassessment of ice cliff dynamics
Geoffrey Evatt, Christoph Mayer, Anna Wirbel, I.D. Abrahams, Lindsey Nicholson
Corresponding author: Geoffrey Evatt
Corresponding author e-mail: geoffrey.evatt@manchester.ac.uk
We present a new model for understanding ice cliff dynamics. This model incorporates a moving frame of reference, fixed to the surface of the surrounding debris covered ice as it melts. Whilst this is a simple change, the inclusion of a moving frame of reference allows for several inconsistencies within the extant literature to be rectified. These changes allow a number of predictions to be made in regard to ice cliff dynamics, as well as suggesting adjustments to field measurement practises. Our predictions include showing: that ice cliffs can endogenously select their own slope angles; that there should be an indifference between illuminated north and south facing ice cliff slopes; that ice cliffs grow steeper with thicker debris layers; that ice cliffs cannot persist below a certain critical debris thickness; and that some previous studies may have systematically overestimated enhanced ice cliff melt rates and total mass losses. All of our results are given in relation to Baltoro glacier, Karakoram, where the overestimate in predicted mass loss using previous methodologies, in comparison with our methodology, is over 140%. We also discuss the implications of our work for land-based ice cliffs, such as those sometimes observed at the snout of a glacier.
The spatial flux of Earth’s meteorite falls found via Antarctic data
Geoffrey Evatt, Andrew Smedley, Katherine Joy, I.D. Abrahams, Laura Gerrish
Corresponding author: Geoffrey Evatt
Corresponding author e-mail: geoffrey.evatt@manchester.ac.uk
Contemporary calculations for the flux of extraterrestrial material falling to the Earth’s surface (each event referred to as a ‘fall’) rely upon either short-duration fireball monitoring networks or spatially limited ground-based meteorite searches. To date, making accurate fall flux estimates from the much-documented meteorite stranding zones of Antarctica has been prohibited due to complicating glacial ice dynamics and difficulties in pairing together distinct meteorite samples originating from the same fall. Through glaciological analysis and use of meteorite collection data, we demonstrate how to overcome these barriers to making flux estimates. Furthermore, by showing that a clear latitudinal variation in fall frequencies exists and then modeling its mathematical form, we are able to expand our Antarctic result to a global setting. In this way, we provide the most accurate contemporary fall flux estimates for anywhere on Earth. Inverting the methodology provides a valuable tool for planning new meteorite collection missions to unvisited regions of Antarctica. Our modeling also enables a reassessment of the risk to Earth from larger meteoroid impacts – now 12% higher at the equator and 27% lower at the poles than if the flux were globally uniform.
The instability theory of drumlin formation
Iain Moyles, James Fannon
Corresponding author: Iain Moyles
Corresponding author e-mail: imoyles@yorku.ca
The instability theory of drumlin formation originated with the work of Hindmarsh in 1998, and has been developing ever since. This poster presents the most recent version of this theory, and illustrates the kinds of result that can be obtained.
Andrew Fowler
Corresponding author: Andrew Fowler
Corresponding author e-mail: andrew.fowler@ul.ie
Ice sails, as seen on the €1.10 stamp on the conference poster and postcard logo, are dramatic supraglacial features consisting of bare ice rising above a debris-lain surface. We provide a simple theory to describe their formation.
Calving: a graph theoretic approach with quantum computing
C. A. Fillekes
Corresponding author: C. A. Fillekes
Corresponding author e-mail: cfillekes@ibm.com
Calving is modeled as the joining of fracture networks when crevasses and other cracks in the ice intersect. While surface cracks can be minimally joined in 2-D with standard flow algorithms (the graphs they form being planar) crevasses and basal cracks can join to calve in 3-D, and the graphs thus formed are not necessarily planar. These may be modeled and characterized from data in 3-D as non-planar graphs, the calving prediction formulated as a weighted cut problem, which can run polynomial time on quantum computers. The physics of fracture is incorporated using an Hamiltonian formulation. While the cross-over to quantum advantage will require hundreds of qubits representing hundreds of fractures, IBM systems will have over 1000 this year and nearly 5000 by 2025]. There are some problems that can be examined with smaller-scale formulations in the mean time, such as constraining the distribution of fractures and cavities not visible at the surface to generate a proof-of-concept calving model.
Supraglacial lake drainage by tidally induced hydrofracture in the Amery Ice Shelf grounding zone
Hanwen Zhang, Richard Katz, Laura Stevens
Corresponding author: Hanwen Zhang
Corresponding author e-mail: hanwen.zhang@earth.ox.ac.uk
Atmospheric warming is driving increasing meltwater production on ice-shelf surfaces. This meltwater may eventually deepen existing crevasses via hydrofracture. Hydrofracture beneath supraglacial lakes can create efficient pathways for transporting surface meltwater to the subglacial hydrological system in a short period of time. Recently, Trusel et al. (2022) reported a series of repeated drainage events of a supraglacial lake close to the grounding line (GL) of the Amery Ice Shelf, East Antarctica. Interestingly, these drainage events do not occur past a threshold in lake volume, but rather most drainage events coincide with times of high daily tidal amplitude. These observations raise the question: How do tidally induced stresses near the GL contribute to hydrofracture? The GL in a marine ice sheet serves as an internal boundary connecting the grounded ice sheet and floating ice shelf. Here, we extend the viscous rheology framework for a migrating GL by Stubblefield et al. (2021) to a viscoelastic Oldroyd-B fluid model in order to capture the viscoelastic responses expected given a tidal period (~12 h) that is close to the Maxwell time of ice (hours to days). Our results show that downward tidal flexure can lead to significant tensile stress near the ice surface in a migrating grounding zone. We then use a linear elastic fracture mechanics model to estimate quasi-static hydrofracturing in the grounding zone. By exploring the dependence of hydrofracture propagation on both lake depth and tidal amplitude, we identify a model-based lake-drainage criterion and compare this criterion with the Trusel et al. observations. We find that both tidal amplitude and lake depth control the initiation of supraglacial lake drainage via hydrofracture in the Amery Ice Shelf grounding zone.