Contents Participants Local website Program

International Symposium on
Maritime Glaciers

Juneau, Alaska, USA, 19–24 June 2022



Access the inaccessible: ploughing subglacial sediments of a surging glacier

Coline Bouchayer, Kjetil Thøgersen, Francois Renard, Thomas V. Schuler, Pierre-Marie Lefeuvre, John Hulth, Ugo Nanni

Corresponding author: Coline Bouchayer

Corresponding author e-mail:

The basal motion of soft-bedded glaciers consists of basal slip and sediment deformation, and changes at the ice–bed interface can lead to glacier instabilities such as surges. Variations of subglacial mechanical conditions have been investigated using ploughmeters at several glaciers around the world. A ploughmeter provides in situ information about transient sediment strength over the typically wide range of seasonal variations that naturally occur below glaciers. We deployed a ploughmeter and water pressure sensors in April 2021 at the base of the marine terminating Konsgvegen glacier in Svalbard. This surge-type glacier shows strong indications for an imminent new event. The ploughmeter precisely measures the magnitude and direction of forces that it experiences while ploughing through the sediment, which allows estimating the effective viscosity of the granular material at the base of the glacier and basal sliding rate. We estimate rheological and hydrological properties of the subglacial material and changes thereof as subglacial water pressure undergoes wide variations. Together with simultaneous measurements of water pressure, glacier flow and ice seismicity, these data shed new light on the basal dynamics of Kongsvegen glacier and eventually contribute to understanding the mechanisms of destabilization of Arctic glaciers.


90 years of glacier change: is it possible to measure glacier volume change from historical images?

William Smith, David Rippin, Edwin Hancock, Julian Dowdeswell, Paulina Lewińska

Corresponding author: Paulina Lewińska

Corresponding author e-mail:

We present a method for extracting quantifiable volumetric information from archival aerial photographs to extend the temporal record of glacier volume change in the central eastern Greenland region. We create 3D models of glaciers at four time steps, derived from three datasets that differ greatly in quality, capture technology and state of preservation. This allows us to measure volumetric change of glaciers over the 1930s, 1960s and 1980s and against modern digital elevation models as provided by ArcticDEM. The 1930s dataset consists of oblique images from BAARE (British Arctic Air Route Expedition) 1930/31. The 1960s dataset comes from CORONA (1959–1972) – a spy satellite mission. The 1980s dataset consists of vertical aerial images taken 1978–87. In this presentation we describe the methods used for reconstruction of 3D models, specific modifications we make in order to deal with glass-plate imagery and analogue historical datasets and our methods for aligning, evaluating the accuracy of and measuring volume change between the models. We also present a preliminary analysis of glacial volume changes over the central eastern Greenland region and its relation to glacier terminus movement over the span of 90 years.


Bayesian inference for data synthesis at Malaspina Glacier

Douglas Brinkerhoff, Martin Truffer, Mark Fahnestock, Brandon Tober, Victor Devaux-Chupin, Chris Larsen, Jack Holt, Michael Christofferson

Corresponding author: Douglas Brinkerhoff

Corresponding author e-mail:

Malaspina Glacier in the St Elias Mountains has the potential to induce significant changes in the coastal landscape through tidewater retreat and has recently been the focus of an intensive field and remote sensing campaign. In this work, we integrate time-dependent observations of surface velocity, surface elevation, bedrock elevation, and climate into a numerical model of glacier flow through stochastic variational Bayesian inference, yielding a probability distribution over bed elevation, basal traction, and mass balance that we use to explore an ensemble of Malaspina’s potential futures.


A new fast thickness inversion approach and its application to Kronebreen, Svalbard

Thomas Frank, Ward van Pelt

Corresponding author: Thomas Frank

Corresponding author e-mail:

As glaciers are melting and retreating worldwide, subglacial topography plays a crucial role in determining their future evolution. Although recent progress has been made in producing large-scale products of ice thickness, substantial uncertainties for individual glaciers still exist. Here, we present a fast thickness inversion approach that is capable of using advanced physics of state-of-the-art ice flow models to produce maps of subglacial topography that are consistent with external forcing and ice dynamics. We harness the vast amount of available observations, such as surface elevation change and ice velocity, to constrain both bed elevation and subglacial friction. This allows a seamless transition from bed recovery to prognostic simulations. Using the example of Kronebreen, a fast-flowing tidewater glacier on Svalbard, and the ice-flow model PISM, we demonstrate that our approach is capable of dealing with complex settings in a time-efficient manner. Ultimately, our method is of interest for prognostic studies where bed topography needs to be constrained first, as well as for applications on a larger scale.


Calving at Helheim Glacier dominated by large flexure events: observations from two autonomous terrestrial laser scanners

Michael Shahin, Leigh Stearns, David Finnegan, Adam LeWinter, Howard Butler, C.J. van der Veen, Sarah Child, Shad O’Neel

Corresponding author: Michael Shahin

Corresponding author e-mail:

Mass loss from the Greenland Ice Sheet has increased sixfold since the 1980s and has the potential to increase global sea level by 7.2 m. The continued increase of Greenland’s contribution to sea level rise is approximately split between increased surface melt and acceleration of marine-terminating glaciers. In order to project ice sheet evolution, the accurate tracking of outlet glacier behavior is essential. Here, we focus on iceberg calving, a key component of ice dynamical mass loss. The causes and impacts of calving occur at a range of spatial (m – km) and temporal (minutes – years) scales, which current observational strategies fail to capture.. To fill this observational gap, we installed two autonomous terrestrial laser scanners (ATLAS) near the terminus of Helheim Glacier, East Greenland – one in 2015, and the second one in 2018. Each ATLAS system scans every 6 hours, except in the winter, when scans are obtained once per day; together, the ATLAS systems provide year-round coverage of Helheim Glacier’s terminus. These year-round, high-resolution, ATLAS observations show that ~75% of calving at Helheim Glacier is from large events that span nearly the whole width (~6 km) of the glacier from 2018. These large calving events initiate upflow, where we observe large surface depressions form; the surface depressions grow in width and depth as they advect downflow, ultimately producing a large calving event. Here, we use ATLAS point cloud generated velocity, strain rates, elevation, and height above buoyancy to quantify the physical conditions around flexure zones in order to understand how these features form and how they vary seasonally.


Seasonal land ice-flow variability in the Antarctic Peninsula

Karla Boxall, Frazer Christie, Ian Willis, Jan Wuite, Thomas Nagler

Corresponding author: Karla Boxall

Corresponding author e-mail:

Three decades of routine Earth observation have revealed the progressive demise of the Antarctic Ice Sheet, evinced by accelerated rates of ice thinning, retreat and flow. These phenomena, and those pertaining to ice-flow acceleration especially, are predominantly constrained from temporally limited observations acquired over interannual timescales or longer. Whereas ice-flow variability over intra-annual timescales is now well documented across, for example, the Greenland Ice Sheet, little to no information exists surrounding seasonal ice-flow variability in Antarctica. Such information is critical towards understanding short-term glacier dynamics and, ultimately, the ongoing and future imbalance of the Antarctic Ice Sheet in a changing climate. Here, we use high spatial- and temporal- (6/12-daily) resolution Copernicus Sentinel-1a/b synthetic aperture radar (SAR) observations spanning 2014–20 to provide evidence for seasonal flow variability of land ice feeding the climatically vulnerable George VI Ice Shelf (GVIIS), Antarctic Peninsula. Between 2014 and 2020, the flow of glaciers draining to GVIIS from Palmer Land and Alexander Island increased during the austral summer (December – February) by ~0.06 m d–1 (22  m a–1). These observations exceed mean standard error bounds of ~0.005 m d–1 (1.8 m a–1) at the grounding line. This speedup is corroborated by independent observations of ice flow as imaged by the Landsat 8 Operational Land Imager that are not impacted by firn penetration and other effects known to potentially bias SAR-derived velocity retrievals over monthly timescales or shorter. Alongside an anomalous reduction in summertime surface temperatures across the Antarctic Peninsula since c.2000, differences in the timing of ice-flow speedup we observe between Palmer Land and Alexander Island implicate oceanic forcing as the primary control on this seasonal signal.


Glacier mass and surface speed variations on the Kenai Peninsula, Alaska

Ruitang Yang, Regine Hock, Shichang Kang, Wanqin Guo

Corresponding author: Regine Hock

Corresponding author e-mail:

Glaciers on the Kenai Peninsula in Alaska cover approx. 4000 km2, span across diverse climate environments, and include land-, marine- and lake-terminating glaciers. We find a 12% reduction in glacier area between 1986 and 2016, and a region-wide mass balance of –0.94 ± 0.12 m w.e. a–1 between 2005 and 2014, which is considerably more negative than found for earlier periods. Mass changes were most negative for lake-terminating glaciers (–1.37 ± 0.13 m w.e. a–1), followed by land-terminating glaciers (–1.02 ± 0.13m w.e. a–1) and tidewater glaciers (–0.45 ± 0.14 m w.e. a–1). We use intensity offset tracking on Sentinel-1 data to characterize the spatiotemporal variations of glacier surface speed between October 2014 and December 2019. Speeds are 50% greater in spring (March–May) than the annual mean, while winter speeds are close to the annual mean. On average the lake-terminating and tidewater glaciers flow 1.7 and 2.3 times faster than the land-terminating glaciers, respectively. Monthly variations over the 5-year period are strikingly synchronous regardless of terminus-type suggesting that regional-scale meteorological drivers govern the temporal variability. We find that winter and spring ice speeds increase with increased glacier runoff as calculated from a mass-balance model, while summer and fall speeds decrease with increased runoff. In addition, winter speeds are negatively correlated with the same year´s summer to fall runoff. Our results highlight the complex nature of the impact of glacier type and water input on spatial and temporal variations in ice speed, the latter likely linked to distinct changes in subglacial drainage systems.


Automated delineation of accumulation areas on Alaskan glaciers from satellite imagery, 1984–2021

Lucas Zeller, Daniel McGrath, Caitlyn Florentine, Louis Sass

Corresponding author: Lucas Zeller

Corresponding author e-mail:

Decadal observations of regional mountain glacier change are crucial to understanding how glaciers respond to changing climate forcings. Satellite multi-spectral imagery offers a wealth of data from the 1980s to present day which can be used to complement in-situ glaciological observations for regional- to global-scale analyses. Here, we present an automated pixel-based approach for delineating accumulation areas of mountain glaciers detected at the annual mass minimum in Landsat and Sentinel-2 imagery. A random forest algorithm is implemented in Google Earth Engine to classify images of glacier surfaces as snow, firn or ice, from which the total accumulation area is determined. Our approach shows high accuracy (>90%) in comparison to a manually produced training dataset using cross validation approaches. The greatest error occurred in differentiating snow from slightly darker (young) firn. A strong correlation is observed between our remotely sensed accumulation areas and independent in-situ observations of equilibrium line altitudes from glaciological observations. Preliminary results from the application of this method across all glaciers in Alaska and Northwest Canada for the entire Landsat 8 and Sentinel-2 image archives are promising, showing physically realistic distributions across individual glaciers and regional scales, with negative trends in accumulation areas observed throughout the study period (2013–present). We discuss the limitations of our approach and potential difficulties with accurately extrapolating this automated technique to greater spatial and temporal scales. Finally, we present plans for comparisons with glacier-specific thinning rates, coincident climatic data, and predicted accumulation area changes under future climate scenarios.


Weakening of a soft bed during glacier surging: Pore-pressure response to till dilation and shrinkage of bed-surface voids

Neal Iverson

Corresponding author: Neal Iverson

Corresponding author e-mail:

Surging glaciers usually have deformable beds. Commonly, water-saturated subglacial till will dilate during the early stages of shear deformation, potentially causing reductions in pore-water pressure that increase the till’s frictional strength. This dilatant strengthening is a central element of some models that seek to describe how bed deformation affects surging, stick-slip of ice streams, and sliding laws. For example, in the model of Minchew and Meyer, dilatant strengthening inhibits the rate of basal motion sufficiently during incipient surging to provide time for dynamic glacier thinning, which in their model promotes surging. Effective stress in their model is depth-averaged in the bed, and water pressure at the bed surface is held constant and independent of bed shear. Herein, motivated by the contrasting hypothesis that bed shear during incipient surging destroys drainage pathways at the bed surface, I explore how shear-induced dilation causes changes at the ice–bed interface that affect the time evolution of effective stress, strain distribution and shearing resistance in the bed. This approach is guided by geological, experimental and hydrological observations that indicate slipping ice locally separates from the surface of a soft bed. By solving a forced, pore-pressure diffusion equation, I show that shrinkage of bed-surface voids during dilative bed shear drives up pore pressure near the bed surface while at depth pore pressure is reduced. Thus, early during a surge or a more generic velocity increase, bed shear strain collapses to a thin, weak zone at the glacier sole, despite the dilatant strengthening of the bed at depth. These results generally agree with measurements of transient deformation in instrumented subglacial till layers. Therefore, the likely effect of till shearing and transient dilation is a decrease in basal drag.


Zeroing in on tidal impact on glacial cove hydrodynamics

Maria Osińska

Corresponding author: Maria Osińska

Corresponding author e-mail:

The importance of glacial meltwater injections on surrounding ocean hydrodynamics has been proven beyond discussion. However due to complexity of those ice–ocean interactions affected by glaciological, oceanographical and meteorological factors, it is still difficult to distinguish actual role of the relevant variables. The goal of this research is to concentrate on and extract the particular influence of tidal forcing on meltwater transport and mixing in near tidewater glacial waters. To achieve this objective in austral summer of 2021/22 repeated measurements have been carried out in three coves adjacent to various sized tidewater glaciers of Admiralty Bay, King George Island, Antarctica: Lange glacier (active ice front length: 1662 m, max height: 93 m, cove area: 1.41 km2), Krak glacier (active ice front length: 785 m, max height: 46 m, cove area: 0.44 km2), and Zalewski Icefall (active ice front length: 445 m, height: 33 m, cove area: 0.18 km2). These measurements consisted of dense profiling with an ADCP (Acoustic Doppler Current Profiler) meter, along transects parallel to the glacial fronts, combined with water quality investigations done with multiparameter sonde YSI Exo2. This methodology revealed occurring currents and allowed tracking of glacial meltwater through water properties’ variability. Measurement campaigns have been conducted in consecutive ebb and flood tide stages and repeated at least three times under varied meteorological conditions. Afterwards, using Deltares Delft3D 4 Flow hydrodynamic model, simulations have been carried out using high-resolution, local curvilinear grids nested in a larger-scale model of Admiralty Bay. Measured in situ data have been used in part to define boundary conditions for the model and in part for validation of its results. Modelling and further analysis exposed differences in glacial meltwater distribution and spreading determined solely by tidal phase, showing its previously underappreciated impact, correlated as well with glacial meltwater quantities in examined areas. Moreover, research demonstrated this mechanism’s properties dependent on glacier and glacial cove sizes.


Which marine-terminating, Northern Hemisphere glaciers have retreated the most since 2000?

Will Kochtitzky, Luke Copland

Corresponding author: Will Kochtitzky

Corresponding author e-mail:

To map and better understand marine-terminating glaciers we manually digitized the terminus position in 2000, 2010 and 2020 of all 1704 glaciers that reached the ocean in the Northern Hemisphere in 2000 using satellite imagery. We found that 85% of glaciers retreated and only 2.5% advanced, while the remaining glaciers didn’t change outside of uncertainty limits. By 2020, 123 glaciers ceased to reach the ocean compared to 2000. We found that most of the area losses occurred in Greenland, which experienced 62% of the total area loss. To understand what causes variations in the rate of retreat we examined a range of environmental and other variables. We found no correlation between air temperature, ocean temperature, or sea ice with retreat rates over any time period. Instead, we found that most retreat occurred on glaciers that have ice shelves, reverse bed slopes, or particularly wide calving margins, or are surge-type. Of the few glaciers that advanced, most of them are surge type. In all, these glaciers lost 7527.3 ± 30.7 km2 (389.7 ± 1.6 km2 a–1) from 2000 to 2020, or an average of just over 1 km2 per day.


Change in firn thickness across the Juneau Icefield between 2012 and 2021

Mikaila Mannello, Jonathan Maurer, Seth Campbell, Kristin Schild, Jordan Farnsworth, Emily Holt, Ian Nesbitt, Sabrina Jones, Rachel Meyne

Corresponding author: Mikaila Mannello

Corresponding author e-mail:

400 MHz ground-penetrating radar was collected along 150 km of repeat transects across the Juneau Icefield, Alaska, USA, in 2012 and 2021. For each year, we determined the depth to the annual accumulation horizon and to the firn–ice transition horizon then differenced these depths to calculate the thickness of the firn. To determine the change in firn thickness, we calculated the difference between the 2021 and 2012 firn-thickness values. Relative permittivity values for both the snow and firn were determined by linking ground truth observations to the corresponding internal reflection horizon. We observed an average decrease in firn thickness of 5.6 m (± 3.0 m) across the Juneau Icefield over the past 9 years. Among those areas which experienced the greatest decrease in firn thickness were Matthes-Taku, Echo Branch-Center, and the Southern-Northwest Branch. This result is in line with an observed significant increase of the equilibrium line altitude over the previous decade. We observed spatial variability of the meltwater content in the snowpack evidenced in the range of calculated relative permittivity values (2.1–4.7) which we attribute to topographic influences. This decreased firn thickness and subsequent decrease in meltwater storage capacity has the potential to impact the downstream hydrology of the local region.


Continuous passive seismic observations of the most recent surge of Sít’ Kusá (Turner Glacier), Alaska, USA

Yoram Terleth, Timothy Bartholomaus, Flavien Beaud, Dylan Mikesell, Jukes Liu, Ellyn Enderlin

Corresponding author: Yoram Terleth

Corresponding author e-mail:

Past studies of glacier surges in maritime climates have identified the critical role of the subglacial hydrological system in transitions to and from the active surge phase, and in ice flow acceleration in general. Despite this consequential progress, many questions surrounding the exact nature of phase transitions within the surge cycle remain unanswered. While recent theoretical work and numerical modelling have made new suggestions towards the interplay between subglacial hydrology and ice velocities, the difficulties inherent to observing the subglacial environment of a surging glacier mean there are few methods to assess the validity of these new theories. In an effort to build a comprehensive dataset of glacier behavior throughout the surge cycle, our ongoing field campaign monitors the 2020/21 surge and subsequent quiescent phase of Sít’ Kusá (Turner Glacier), a 177 km2 glacier located in Disenchantment Bay on the southwest flank of the St Elias range, Alaska. USA. An extensive network of geophysical instruments was deployed, including seismometers, GPS receivers, weather stations, ice-penetrating radar, and time-lapse cameras. Here we present a continuous record of glacio-hydraulic tremor observed through the network of 13 seismometers. This record spans much of the latest surge of Sít’ Kusá, including a temporary slowdown in September 2020 and the suspected termination in August 2021. We discuss the use of the recorded seismic tremor signal as a proxy for the state of the subglacial drainage system and highlight changes in hydrology that coincide with changes in ice velocities. Many of our observations are reminiscent of founding work towards glacier surges in maritime climates, and our early interpretations broadly fit within existing theories of glacier surging. However, we find considerable spatial and temporal variability in the state of the subglacial drainage system. While a relation between subglacial pressurization and ice velocity seems present, our data point towards a subglacial system that is out of equilibrium throughout the surge. We suggest the level of complexity reflected by our data is, as of yet, poorly accounted for in the theoretical understanding of glacier surging.


Applied deep learning techniques in temperate glacial environments for the prediction of snow and firn distributions from ground-penetrating radar data

Jonathan Maurer, Mikaila Mannello, Seth Campbell, Kristin Schild, Yifeng Zhu, Ian Nesbitt

Corresponding author: Jonathan Maurer

Corresponding author e-mail:

High-elevation mountain glaciers are referred to as the ‘water towers of the world’ due to their ability to store water internally and release it later through melt and runoff. Much of the work to understand these complex environments has focused on cold-based polar glaciers in high-atitude regions, neglecting temperate glaciers and associated snowpack, which are more responsive than polar systems to minor climatic changes. Advancements in the availability of open source data, deep learning theory, and computational efficiency has opened up new methods of ‘data hungry’ modeling that may be well suited to the type of complex terrain–climate interactions common in these environments. Through the use of regression-based deep neural networks (DNNs), this study explores the applicability of this method on a 400 MHz ground-penetrating radar dataset collected on the Juneau Icefield between 2012 and 2021. The primary task of these networks is to find the statistical relationship between firn thickness and a collection of topographic and climatological data with the goal of upscaling measurements from radar transects to the entire region. Models are validated using the standard ‘leave on glacier out’ (LOGO) and ‘leave one year out’ (LOYO) holdout methods. Analysis of relative feature importance within each model reveals slope, elevation, and annual accumulation thickness are the most importance predictors for firn thickness on the Juneau Icefield. This method generates realistic simulations of icefield-scale firn and snowpack distributions and can be paired with existing models to output estimations of key environmental characteristics.


Changes in the dynamics of lake-terminating glaciers after the outburst of Lago Greve in Patagonia

Shuntaro Hata, Shin Sugiyama

Corresponding author: Shuntaro Hata

Corresponding author e-mail:

The proglacial lake plays an important role in the evolution of lake-terminating glaciers, by affecting ice flow speed and frontal ablation. Sudden change in the water level due to lake outburst affects the ice dynamics and ablation near the terminus, which give impacts on the glacier front and thickness variations. Despite a number of previous studies focused on glacial lake outburst, the impact of the outburst on the glacier dynamics is poorly understood. During a drainage event in 2020, the lake level of Lago Greve dropped by 18 m. To study the impact of the outburst on Glaciar Pío XI and Greve, the largest glaciers flowing into the lake, we analyzed satellite remote sensing data from 2016–2021 for ice-front position, velocity, and surface elevation. Here, we report the dynamic change of the lake-terminating glaciers in response to the drainage event of Lago Greve. Glaciar Pío XI and Greve are outlet glaciers of the Southern Patagonia Icefield. The focus of this study is the northern front of Glaciar Pío XI and two ice-fronts of Glaciar Greve, which are flowing into Lago Greve. The positions of the ice fronts were measured with optical satellite imageries (Sentinel-2, Landsat 8 and PlanetScope). Glacier velocity was measured by applying a feature tracking method to the Sentinel-2 and Landsat 8 imagery. We also utilized monthly-mean and image-pair velocity products of the RETREAT project. Glacier surface elevation change was measured by comparing 5 m resolution digital elevation models generated from stereo-pair images of WorldView-1 and SPOT-6/7. The velocity of the northern front of Glaciar Pío XI decreased after the event by 50% as compared to the mean speed from 2016–19. Because of the deceleration of the northern front, the main flow of the glacier switched to the southern front. The western front of Glaciar Greve advanced by 113 m from April 2020 to December 2021 in contrast to a retreating trend since 1985. The glacier surface elevation in the terminus region decreased by 12.7 m in February–July 2020. The sudden ice-front advance and surface lowering implied that the glacier terminus acceleration upon the drop in the lake level and the change in the speed enhanced longitudinal stretching flow regime. These results indicated significant influence of the lake water level on the glacier dynamics. Further investigation for a longer period is needed to investigate lasting impact of the water-level change on the lake-terminating glacier.


Characterization of meltwater discharging channels in Hansbreen and their role on the intraseasonal front behaviour

Eva De Andrés, Kaian Shahateet, José M Muñoz-Hermosilla, Jaime Otero, Francisco J. Navarro, Michał Ciepły, Dariusz Ignatiuk, Waldemar Walczowski

Corresponding author: Eva De Andrés

Corresponding author e-mail:

Mass loss of tidewater glaciers in Svalbard has increased over the last decades, with frontal ablation (calving plus submarine melting) accounting for up to 30% of the total loss. Under this scenario, Hansbreen front has retreated more than 1 km over the period 1999–2016. The limited size of this tidewater glacier makes it highly vulnerable to changing ambient conditions, with PDD and seawater temperature as main controls on glacier-front retreat. Coupled glacier–fjord models have also confirmed the tight dependency of intraseasonal glacier retreat with both subglacial meltwater discharges and seawater temperature. However, some aspects concerning Hansbreen front behaviour, such as the actual meltwater production and its subglacial distribution, still remain unclear. In this study, we aim to constrain some of the key aspects affecting Hansbreen front retreat by combining subglacial hydrology models along with downscaled surface melting (at 350 m spatial resolution) and time-lapse-camera front observations. We compare the location of modelled vs observed discharging channels at Hansbreen front from Jan 2010 to Sep 2011. Submarine melting within each subglacial channel is estimated by using the line-plume model with their corresponding subglacial discharges and observed seawater properties (temperature and salinity profiles). Preliminary results show three subglacial discharging channels along Hansbreen front. The fastest-flowing one is located at the deepest zone, in the central-west side of the front. The two other channels are shallower and closer to the margins. Submarine melting by itself is not sufficient to account for all of the observed front retreat. Nevertheless, we think that this channel configuration might be responsible for the front destabilization. 3D glacier–fjord coupled models could be used to test this hypothesis. This project was funded by grant PID2020-113051RB-C31 from MCIN/AEI and RD 289/2021 under Next Generation EU.


The relationship between submarine melt and subglacial discharge from observations at LeConte Glacier, Alaska

Rebecca Jackson, Roman Motyka, Jason Amundson, Nicole Abib, David Sutherland, Jonathan Nash, Christian Kienholz

Corresponding author: Rebecca Jackson

Corresponding author e-mail:

At tidewater glacier termini, ocean–glacier interactions hinge on two sources of freshwater – submarine melt and subglacial discharge – yet these freshwater fluxes are often unconstrained in their magnitude, seasonality, and relationship. With measurements of ocean velocity, temperature and salinity, fjord budgets can be evaluated to partition the freshwater flux into submarine melt and subglacial discharge. We apply these methods to calculate the freshwater fluxes at LeConte Glacier, Alaska, across a wide range of oceanic and atmospheric conditions during six surveys in 2016–2018. We compare these ocean-derived fluxes with an estimate of subglacial discharge from a surface mass balance model and with estimates of submarine melt from multibeam sonar and autonomous kayaks, finding relatively good agreement between these independent estimates. Across spring, summer and fall, the relationship between subglacial discharge and submarine melt follows a scaling law predicted by standard theory (melt~discharge1/3), although the total magnitude of melt is an order of magnitude larger than theory. Subglacial discharge is the dominant driver of variability in melt, while the dependence of melt on fjord properties is not discernible. A comparison of oceanic budgets with glacier records indicates that submarine melt removes 33–49% of the ice flux into the terminus across spring, summer and fall periods. However, while submarine melting is an important component of the glacier’s mass balance, it does not appear to directly amplify calving; instead melt may indirectly promote calving and affect flow dynamics by forcing the terminus to retreat into deeper water during periods of exceptionally high melt.


Maritime glacier change in Kenai Fjords National Park between 1984 and 2021

Taryn Black, Deborah Kurtz

Corresponding author: Taryn Black

Corresponding author e-mail:

Glaciers are an important component of the Alaska landscape, influencing terrestrial and marine ecosystems and statewide ecotourism. Glaciers are sensitive to climate change and their presence and extent are ephemeral. We used Landsat imagery from 1984 through 2021 to map glacier ice margins for 19 maritime glaciers in Kenai Fjords National Park in southcentral Alaska, to quantify seasonal terminus position and areal change, including distinguishing between ice loss at glacier termini and ice loss along glacier margins. Of these glaciers, 13 lost substantial length, 14 lost substantial area, and only two (Holgate and Paguna glaciers) underwent both net advance and net area gain over the course of the study period; the glaciers that had insubstantial length and area changes were predominantly tidewater. Cumulatively, these 19 glaciers retreated 21 km and lost 42 km2 of ice. The ice area loss was nearly evenly distributed between loss at the terminus and loss along the lateral margins. Ice loss affects terrestrial, freshwater, and marine ecosystems and physical properties, and glacier retreat and associated ecosystem changes in Kenai Fjords National Park will likely impact ecotourism associated with the park.


Detecting decadal-scale trends in hydropower availability in a coastal glacierized watershed, Cordova, Alaska, USA

Joanna Young, Daisy Huang, Jeremy VanderMeer

Corresponding author: Joanna Young

Corresponding author e-mail:

The coastal town of Cordova in southeast Alaska, USA, is characterized climatologically by extreme mean annual precipitation (~3.8 m w.e. a–1) and temperatures frequently fluctuating near the freezing point (annual mean of 4.5°C at sea level). Geographically, the surrounding Chugach Mountains rise steeply from sea level to 2100+ m. As a result, this environment supports mountain glaciers and extreme snowfall and rainfall that contribute to high volumes of freshwater runoff. In turn, this abundant freshwater flux and steep topography also yield ideal sites for hydropower generation. The Cordova Electric Cooperative, a remote microgrid utility, relies on hydropower from the glacierized Power Creek watershed for up to 60% of the electrical demand for Cordova’s 2200 inhabitants and several large-scale commercial fish processing plants. Yet despite nearly 20 years of operation and an anticipated lifetime of several more decades, little is known about how water availability to the dam site may be changing in a warming climate. This data assimilation and modeling study examines recent changes to the hydrological regime of the Power Creek watershed from 1980 to present. We leverage the glacio-hydrological modeling tool SnowModel and a novel calibration data set derived from power availability at the dam site. We assess whether runoff in this watershed is increasing or decreasing overall under glacier loss and changing precipitation patterns, which translates to changes in potential power availability at the dam site. We also identify any seasonal trends in runoff, either in terms of timing or water volume, information that is useful given the community’s increased power demand at certain times of year in association with, for example, commercial fish processing. Finally, we provide information on trends over time in magnitude and likelihood of extreme precipitation events and droughts, which translate to periods of risk for hydroelectric infrastructure and periods of low power availability at the dam. Overall, in partnership with the electric cooperative, this study aims to build community resilience to future changes in water availability and thus power availability in this hydrologically complex environment.


Optimizing radar surveys to accurately measure winter mass balance on mountain glaciers

Louis Sass, Daniel McGrath, Shad O’Neel, Caitlyn Florentine

Corresponding author: Louis Sass

Corresponding author e-mail:

Measuring the magnitude and spatial distribution of snow accumulation across glaciers is fundamentally important to understanding how mountain glaciers respond to a changing climate. Ground penetrating radar (GPR) is an efficient tool to increase the quantity and distribution of snow depth estimates relative to snow-pits and probes, but even a comprehensive GPR survey still requires extrapolation to unmeasured sites. The uncertainty from extrapolation is challenging to quantify, which hinders optimal survey design and limits real-world interpretation. In 2015 we collected 174 km of high-frequency (500 MHz) GPR profiles at Wolverine Glacier (15 km2) in the Prince William Sound region of Alaska, USA. These include a regular grid with 250 m spacing collected from a helicopter, complemented by ground-based ‘opportunistic’ surveys. We compare three common extrapolation methods on jackknifed subsets of data (n =  1024) to quantify how extrapolation method and data distribution affect uncertainty in glacier-wide (mean) accumulation and in distributed accumulation. Glacier-wide accumulation can be within 10% of the complete data even with very sparse subsets (down to 5% of the total) using a simple elevation-based (profile) extrapolation method, if the subset of data encapsulates the complete elevation range and avoids major outliers (drifts or scours) in the higher elevations. This is fundamentally different from the uncertainties in the distributed accumulation, which are higher even with larger data sets and more complicated models. Regression-tree (RT) statistical models incorporating terrain parameterizations of elevation, slope, aspect, curvature, shelter, and slope breaks, can result in root-mean-squared errors (RMSEs) as low as 10% of the mean point value, but only for large subsets (>25% of the total) that are well distributed over the variation in terrain parameters (i.e. that include the outliers in higher elevations). Multi-variable regression (MVR) models on the same terrain parameters have higher RMSEs than RT models on large subsets but show better stability on smaller subsets of data. Subset distributions that are limited to one region of the glacier, such as the ablation zone, have the highest uncertainties in all applications regardless of the type of model used, indicating that they should not be used to extrapolate glacier-wide accumulation or infer distributions outside of the sampled domain.


Understanding the underwater noise of melting glacier ice

Hayden Johnson, Grant Deane, Dale Stokes, Oskar Głowacki, Mateusz Moskalik, Mandar Chitre, Hari Vishnu

Corresponding author: Grant Deane

Corresponding author e-mail:

Ocean-forced melting of tidewater glaciers and ice shelves has the potential to drive grounding line retreat and ice shelf thinning that can, in turn, increase the mass flux from marine-terminating ice sheets into the global oceans, causing sea level to rise. Efforts to improve understanding of ice sheet–ocean interactions are key to reducing uncertainty in long-term estimates of global sea level rise. The question of how quickly ice melts under given far-field ocean conditions remains difficult to answer, though, in part because measurements of submarine melting rates are difficult to make. Some recent work suggests that conventional approaches may be significantly underestimating ambient rates of submarine melting. In recent years, progress has been made towards using passive acoustics to observe submarine melting. This method makes use of the sound produced by the release of air bubbles from the ice into the water. The bubbles are compressed by the overburden pressure in the ice, and are released rapidly into the water if their internal pressure is greater than the hydrostatic pressure in the water. So far, conclusions drawn from the use of passive acoustics to study submarine melt have been largely qualitative and comparative. Quantitative estimates of melting rates would require sufficient knowledge of the properties of the ice and the air bubbles trapped in it, and sufficient understanding of the physical processes involved in the sound production during release of the bubbles, to predict the total sound energy radiated per unit volume of ice that melts. We will present the results of experimental work carried out with glacier ice collected in Hornsund Fjord, Svalbard. Pieces of floating ice, calved from several glaciers in Hornsund Fjord, were collected and analyzed to determine bubble sizes and average gas pressure in the bubbles. Some blocks were melted in tanks and the sound produced during the melting was recorded. We have demonstrated a correlation between estimates of the potential energy stored by compressed air bubbles in the ice and the measurements of the acoustic energy released during melting. This represents a first step towards making quantitative estimates of submarine melting from passive acoustic measurements. Additionally, we will discuss implications of this relationship for potential applications of passive acoustic measurements of submarine melting of glacier ice.


Surge evolution of Sít’ Kusá (Turner Glacier), southeast Alaska

Jukes Liu, Ellyn Enderlin, Timothy Bartholomaus, Yoram Terleth, T. Dylan Mikesell, Flavien Beaud

Corresponding author: Jukes Liu

Corresponding author e-mail:

The physical mechanisms that drive regular oscillations between modes of slow and fast flow for surge-type glaciers remains one of glaciology’s greatest unsolved mysteries. The activation of fast flow during the surge phase is attributed to a reduction in resistance at the glacier bed that allows accelerated basal motion. However, the relative importance of the wide variety of internal and external mechanisms that influence basal resistance remains unclear, in part due to the scarcity of detailed surge observations. We use the recent surges of Sít’ Kusá (Turner Glacier) in southeast Alaska to investigate controls on surge evolution. Records of Sít’ Kusá from Nolan and others indicate that it has surged 6 times since 1980, exhibiting the shortest surge cycle known in the world (~8 years). Here we describe detailed records of surface velocity, thickness, and terminus position evolution throughout its most recent surge cycle, from 2013–2021, produced from analysis of optical and SAR images. We focus on kinematics during the most recent surge, from 2020–2021, when the satellite record is complemented by in situ GPS measurements. The thickness and velocity records suggest that the surge originates from the northern tributary of the glacier. During the quiescent phase, the glacier undergoes gradual but unsteady acceleration leading up to the surge initiation. We resolve changes to the speed of down-glacier propagation of the surge front throughout 2020 with the high temporal resolution and coverage of these records. Afterwards, we observe a sudden drop in speeds (from 30+ m d–1 down to ~1 m d–1) associated with the surge termination in 2021. Future work will focus on attributing the kinematic changes throughout Sít’ Kusá’s surge to changes in glacier hydrology and the internal balance of forces.


Investigating recent and historical change in tidewater glaciers in Glacier Bay, Alaska, USA

James Sanders, Timothy Barrows

Corresponding author: James Sanders

Corresponding author e-mail:

Alaska is the largest source of ice loss outside of Greenland and Antarctica, and sensitive to climate change due to steep mass balance gradients. Marine-terminating glaciers are further prone to collapse due to rapid dynamic losses through calving. Glacier Bay in southeast Alaska, USA experienced the largest known post-Little Ice Age (LIA) deglaciation; 120 km of ice up to 1 km thick was lost from 1750 to ~1920, resulting in possibly 8 mm of sea level change. West of Glacier Bay, Columbia Glacier has retreated >20 km since 1980, documented by satellite imagery. Processes involved in the tidewater glacier cycle are poorly understood due to limited full-cycle observations. Many prognostic climate models thus neglect dynamic loss, despite significant importance for marine-terminating glaciers. We will improve modelling of dynamic retreat by testing several calving parameterizations at Glacier Bay and Columbia Bay with the Ice Sheet System Model (ISSM). Although modelling tidewater glaciers will greatly enhance understanding of long-term tidewater glacier collapse before the satellite era, recent tidewater glacier change can be investigated using modern remote sensing techniques. Most tidewater glaciers in Glacier Bay have retreated on to land. There are five fully tidewater glaciers remaining, and all but one have demonstrated unusual stability or advance in previous studies. Through a holistic study of remotely sensed glacier attributes, we have determined that this period of stability is ending. Terminus retreat is accelerating, including at the previously advancing Johns Hopkins Glacier. Thinning is propagating from the terminus up-glacier in nearly all cases, along with acceleration of surface velocities. Documenting the change from the previously observed stability (from the 1950s onwards) to the resumption of retreat in the 2010s will help identify the causes of the heterogeneity of tidewater glacier response. This will be combined with the planned modelling to methodically determine the accuracy and generalizability of common calving parameterizations when applied at the centennial scale over a large, 100 km long domain. Therefore, the project as a whole aims to increase practical decision-making for future modelling and contribute to analysis of similar Alaskan tidewater glacier behaviour in models or observation.


Subglacial discharge controls seasonal variations in the thermal structure of a glacial lake in Patagonia, Chile

Shin Sugiyama, Masahiro Minowa, Yasushi Fukamachi, Shuntaro Hata, Yoshihio Yamamoto, Tobias Sauter, Christoph Schneider, Marius Schaefer

Corresponding author: Shin Sugiyama

Corresponding author e-mail:

Glaciers terminating in the ocean and lakes are changing more rapidly than those terminating on land. Mass loss of freshwater calving glaciers is particularly important in regions under maritime climate, including Patagonia, Alaska and New Zealand. Despite recent progress in the research on tidewater glaciers, studies on freshwater calving glaciers and glacial lakes are sparse. For example, water temperature in glacial lakes affects underwater melting, but seasonal lake temperature variations are poorly understood because taking long-term measurements near the front of calving glaciers is challenging. To investigate the thermal structure and its seasonal variations, we performed year-around temperature and current measurement in the glacial lake Lago Grey in Patagonia, Chile. Lago Grey is situated in southeastern Patagonia with an area of 38 km2 and a length of 16 km. The lake is fed by Glaciar Grey, an outlet glacier in the Southern Patagonia Icefield. From March 2017 to November 2018, we installed a mooring in the lake at 1.4 km from the glacier front, where the water depth was 410 m. The mooring consisted of three temperature loggers (Seabird, SBE56) at depths of 139, 224, and 309 m from the surface, and two current/temperature loggers (JFE Advantech, INFINITY-EM) at 57.5 and 391.5 m. The measurement was performed with intervals of 1 min (temperature loggers) and 1 h (current/temperature loggers) for 611 days. The measurement revealed critical impacts of subglacial discharge on the lake thermal condition. Water below a depth of ~100 m showed the coldest temperature in mid-summer (~2°C), under the influence of glacial discharge, whereas temperature in the upper layer followed a seasonal variation of air temperature. The temperature below 200 m is colder than 3°C throughout the year, suggesting relatively small rates of underwater melting. The upper-most 50 m is kept above 4°C from January to March, implying that the most significant melting occurs near the lake surface during this period. The boundary of the lower and upper layers was controlled by a ~150 m deep sill situated at 4 km from the glacier front, which blocks outflow of dense and cold glacial meltwater. Our data from Lago Grey implies that subglacial discharge and bathymetry play key roles in the frontal ablation of lake-terminating glaciers. The cold lakewater hinders underwater melting, leading to formation of a floating terminus and buoyancy driven calving.


Historic imagery reveals glacier response to ice shelf debutressing on the Antarctic Peninsula

Ryan North, Timothy Barrows

Corresponding author: Ryan North

Corresponding author e-mail:

The Antarctic Peninsula is one of the fastest warming regions on Earth. Since the 1950s, ~87% of its tidewater glaciers have retreated, many after ice-shelf debutressing. The dramatic collapse of the Larsen B ice shelf in 2002 has caused the thinning and acceleration of the former ice-shelf tributaries. However, these glaciers were unmonitored prior to collapse, which restricts assessments of the magnitude of recent thinning. Here, we calculate the total amount of ice drawdown since the collapse. To do this we determined the pre-collapse elevation of the glaciers and the adjoining ice shelf using archival aerial photography from 1968. We assume that the 1968 configuration is the same as before the collapse because this is the only complete set of photography available. Structure-from-motion photogrammetry was used to reconstruct a historic digital surface model (DSM), which was then differenced from a modern reference DSM (‘Reference Elevation Model of Antarctica’) to compare the surface elevation and volume of former Larsen B ice shelf tributaries ~34 years pre-collapse to ~15 years after their transition to tidewater glacier. Results indicate that the surface of the Crane glacier, the main tributary of the former Larsen B ice shelf, has lowered by ~58 m and by a maximum of ~170 m. Approximately 34 km3 of glacier volume has been lost since 1968, which represents ~21% of its volume. Most of this mass was likely lost shortly after the ice-shelf collapse in 2002. This study highlights the rapid contributions to sea level from buttressed glaciers after ice-shelf collapse.


The glaciers of Glacier Bay

Andrew Bliss

Corresponding author: Andrew Bliss

Corresponding author e-mail:

The glaciers of Glacier Bay National Park and Preserve have retreated dramatically over the last 250 years (>100 km). We give an overview of these changes derived from historical glacier observations, repeat photography, bathymetric surveys, and remote sensing. We present new results showing the current stability of Johns Hopkins Glacier’s terminus (<100 m net change in the last 10 years), incipient retreat of Margerie Glacier (1 km retreat over the last 5 years after 65+ years of stability), and ongoing rapid retreat of McBride Glacier (>3 km in 10 years). Oceanographic and climate measurements from Glacier Bay allow us to investigate forces driving glacier change. We also explore the consequences of glacier change for flora, fauna, and ocean temperature and salinity.


A 3D glacier dynamics-line plume model to estimate the frontal ablation of Hansbreen, Svalbard

José M. Muñoz-Hermosilla, Eva De Andrés, Kaian Shahateet, Jaime Otero, Francisco J. Navarro

Corresponding author: José M. Muñoz-Hermosilla

Corresponding author e-mail:

Frontal ablation is responsible for a large fraction of the mass loss from marine-terminating glaciers. The main contributors to frontal ablation are iceberg calving and submarine melting, being calving the largest contributor. However, submarine melting, in addition to its direct contribution to mass loss, also promotes calving through the changes induced in the stress field at the glacier terminus, so both processes should be jointly analysed. Among the factors influencing submarine melting, the formation of a buoyant plume due to the emergence of fresh subglacial water at the glacier grounding line plays a key role. In this study we use Elmer/Ice, an open-source, parallel, finite-element software which solves the full-Stokes system, to develop a 3D glacier dynamics model including calving and subglacial hydrology coupled with a line-plume model that accounts for the submarine melting at the calving front. The ice flow model provides the calving front position and the subglacial runoff at every time step. This runoff is used as input to the line-plume model for estimating the subglacial melting at the glacier front. The latter, in turn, is used as input to the next step of the ice flow model. We apply this model to the Hansbreen–Hansbukta glacier–fjord system in Southern Spitsbergen, Svalbard, where a large set of data are available for both glacier and fjord. The modelled front positions are in agreement with those observed from time-lapse images of the glacier front. This research was funded by grants PID2020-113051RB-C31 and PRE2018-084318 from MCIN/AEI /10.13039/501100011033 and the FSE ‘El FSE invierte en tu futuro’.


Geometric and geodetic mass balance changes of Jostedalsbreen ice cap

Liss Marie Andreassen, Hallgeir Elvehøy, Bjarne Kjøllmoen, Joathan Carrivick, Kjetil Melvold, Benjamin A. Robson, Kamilla H. Sjursen

Corresponding author: Liss Marie Andreassen

Corresponding author e-mail:

The Jostedalsbreen ice cap is mainland Norway’s and Europe’s largest ice cap with present maximum ice thicknesses of 600 metres. It is a maritime glacier with high mass turnover and snow accumulations of up to 7 m in a regular winter. The glacier had its maximum Little Ice Age (LIA) extent about 1740–1860 with an area of 568 km2 and an ice volume of between 61 km3 and 91 km3. The major outlet glaciers had lost at least 110 km2 or 19% of their LIA area and 14 km3 or 18% of their LIA volume by 2006. The glacier area has been further reduced by 3% from 2006–19. Jostedalsbreen therefore now accounts for 458 km2, 20% of the total glacier area of mainland Norway. Mass balance is presently measured on two outlet glaciers and length change on six outlet glaciers. Many of the outlet glaciers had advances that culminated around 2000. Since 2000 the glaciers have reduced in both length and mass. Geodetic mass balance results for three outlet glaciers, which in combination cover 20% of the total Jostedalsbreen area, have revealed large differences in surface elevation change and geodetic mass balance. Whereas Nigardsbreen had a small mean surface elevation change of –2.2 m from 1964 to 2013, Tunsbergdalsbreen had a mean surface lowering of 40 m in the same time period. The surface elevation change of Austdalsbreen was –17.4 m for the shorter time period 1966–2009. The differences in surface elevation changes between these three outlet glaciers can partly be attributed to very different hypsometry distributions. In this paper we will present results of geodetic mass balance for the entire ice cap from 1966–2020 with a reconstructed digital terrain model based on aerial photographs from 1966 and laser scanning from 2020.


Seasonal snow line mapping using Planet imagery

Rainey Aberle, Ellyn Enderlin

Corresponding author: Rainey Aberle

Corresponding author e-mail:

The seasonal snow line and snow-covered area (SCA) are critical parameters for estimates of when and where the snowpack completely melts (i.e., snow melt out). Changes in these parameters over time are important indicators of changes in glacier surface mass balance and can be mapped on global scales using space-borne observations. However, the accuracy and applications of remote seasonal snow line and SCA measurements have been limited by the temporal and spatial resolution of the imagery, particularly in maritime and/or mountain regions where cloud cover is abundant. Additionally, well-established SCA classification methods such as that used for the Theia Snow collection have not been calibrated or well assessed in glacierized environments, limiting their use for detecting the glacier equilibrium line (i.e. the snow line at the end of the melt season). Here we develop a new classification workflow for detecting the SCA and seasonal snow line using Planet imagery, which has a spatial resolution of ~3 m, daily to sub-weekly temporal coverage, and contains bands in the visible and near infrared spectrum. We compare results to Landsat- and Sentinel-derived SCA products and apply the method to glaciers in the western USA to assess temporal and spatial patterns in snow line and glacier equilibrium line altitudes since 2016.


Bayesian estimation of mass balance model parameters for glaciers along a maritime–continental climate gradient

Kamilla Hauknes Sjursen, Thorben Dunse, Antoine Tambue, Liss Marie Andreassen, Thomas Vikhamar Schuler

Corresponding author: Kamilla Hauknes Sjursen

Corresponding author e-mail:

Mass balance models are tools to assess past and future glacier and runoff evolution. Empirical models are commonly used in large-scale applications or where meteorological data for energy-balance modelling is lacking. Such models require calibration of parameters using observations of glacier mass change. Limited by the low availability of in-situ glaciological observations, models have generally relied on transferring parameters between glaciers or calibrating regional parameters based on sparse observations. In this work we aim to evaluate the variability in mass-balance model parameters across glaciers in different climatic settings. We apply a Bayesian approach to parameter estimation in a temperature-index model and use Markov chain Monte Carlo simulations to estimate posterior parameter distributions based on in-situ mass balance observations. We focus our analysis on seven glaciers situated along a maritime–continental climate gradient in southern Norway, from the maritime Ålfotbreen glacier in the west to the continental Gråsubreen glacier in the east. We evaluate parameter estimates with respect to continentality and assess the transferability of parameters across glaciers. We provide estimates of glacier mass balance from 1960–2020 and evaluate model performance over the period of available in-situ mass-balance observations.


NOAA’s climate data records, model output and services for Alaska, USA

Jessica Cherry

Corresponding author: Jessica Cherry

Corresponding author e-mail:

This overview by NOAA’s Regional Climate Services Director for Alaska provides researchers with a better understanding of how longterm data are collected by NOAA and agency partners and processed into climate data records (CDRs) by the National Centers for Environmental Information. New and emerging products from surface sources and remote sensing are attempting to close the information and quality gaps for cold regions precipitation. Model analyses from NOAA are increasingly used in operational weather forecasting and remote-sensing for the monitoring and diagnosing of glacial dam outburst hazards. Some researchers may be unaware of data resources available to the operational community and other services from NOAA. The talk will highlight these resources.


Seasonal strain rate and force balance evolution from terrestrial radar data at LeConte Glacier, Alaska

Dominik Fahrner, David Sutherland, Erin Pettit, Jason Amundson, Nicole Abib, Christian Kienholz, Roman Motyka

Corresponding author: Dominik Fahrner

Corresponding author e-mail:

LeConte Glacier, a tidewater glacier in southeast Alaska, USA, has experienced rapid retreat and thinning in the mid-1990s, serving as an analogue to Greenland’s many tidewater glaciers. Although its retreat and thinning has continued, the glacier has stabilized somewhat, possibly due to a constriction in the fjord geometry. However, the influence of fjord geometry on ice dynamics and the drivers of flow variability at LeConte Glacier have not yet been determined. Here we present high-temporal resolution (sub-daily) surface strain rates derived from terrestrial radar velocimetry products acquired at LeConte Glacier in April, May and August 2017. Radar data have been widely used to determine changes in seasonal or annual ice dynamics of tidewater glaciers. Ice surface velocities proximal to the terminus were determined from 24 h time-separated radar images using the Lucas–Kanade method within OpenCV, and upstream from interferometry. Shear, longitudinal and transverse surface strain rates were derived using velocity fields rotated to the local flow direction (left to right) and smoothed to the input posting. Longitudinal strain rates at LeConte Glacier show contrasting, but spatio-temporally variable, regimes of extensional and compressional flow at the terminus, similar to spatial patterns observed at Narsap Sermia, a tidewater glacier in SW Greenland. The results show that fjord geometry does impact velocity gradients and thus strain rates, but also suggest that other factors affect the observed variability in ice dynamics. By utilizing force balance mapping we hope to provide further insights into the environmental and geometric factors which influence the stress distribution at LeConte Glacier. The results could improve our understanding of the drivers of tidewater glacier retreat in Alaska and might have broad implications for analogous Greenlandic systems.


Analysis of terminus ablation time series for Helheim Gletsjer and Kangiata Nunaata Sermia

Aman KC, Ellyn M. Enderlin

Corresponding author: Aman KC

Corresponding author e-mail:

In recent decades, Greenland’s marine-terminating glaciers have retreated, accelerated and thinned as a result of climate change. Although mass loss driven by flow acceleration (i.e., dynamic mass loss) can be approximated as the flux of ice towards the glacier terminus, the precise timing of the mass loss deviates from flux estimates due to irregular iceberg calving. Here we focus on terminus ablation from Helheim Gletsjer and Kangiata Nunaata Sermia, located in southeast and southwest Greenland, respectively. The terminus ablation rate is estimated as the difference between surface speed and the terminus position change rate along each glacier centerline. We make use of the tremendous increase in data availability since the launch of Landsat 8 and Sentinel-2, using the velocity data from NASA ITS_LIVE project and detailed glacier termini datasets produced from Termpicks for our analysis. Comparison to potential driving mechanisms yields insights into controls on terminus position and will be especially valuable for ice-sheet mass flux estimates and/or process-based studies of iceberg calving.


The cryosphere and climate change communication: a systematic review of messengers, messages, and outcomes of framing climate change as thawing ice

Kristin Timm, Julian Dann, Joanna Young

Corresponding author: Joanna Young

Corresponding author e-mail:

Whether it’s a calving glacier or other imagery of disappearing ice, dramatic changes and dynamic processes in the cryosphere have had a prominent place in how scientists, the news media, film makers, and others talk about the effects of climate change. This strong, sometimes emotional imagery, and the relatively easy-to-understand relationship between heat and ice makes glaciers and other cryosphere features helpful for communicating about climate change. However, their position on the globe, away from major population centers, could also reinforce the perception that climate change is a problem that predominantly affects the polar regions and alpine areas– and not mid-latitudes and the people who live there. Through a systematic review of the academic literature on climate change communication, we synthesize the use of the cryosphere in climate change messaging, the types of communicators and messengers who use this framing, and the effects on different audiences of emphasizing the cryosphere in climate change communication. We find that glaciers and glacier-related elements (icebergs, ice sheets, etc.) dominate the cryosphere topics engaged in the literature, featuring more than twice as often as the next common cryosphere feature (sea ice). We synthesize the benefits of invoking affective imagery like melting glaciers, such as eliciting emotional rather than cognitive responses, as well as the disadvantages, such as reinforcing the notion that climate change must be ‘seen to be believed.’ Overall, this synthesis identifies areas of agreement and gaps in our understanding that can inform future areas of communication research. It also proposes evidence-based, practical guidance for science and climate-change communicators, including scientists, educators, and tourism operators, so that their efforts to highlight melting glaciers and the changing cryosphere lead to better and more productive communication with their audiences.


Hydrographic trends in glacier influenced fjord environments across the Gulf of Alaska

Anna Simpson, Journey Berry, Peter Kirchner, Erin Pettit

Corresponding author: Anna Simpson

Corresponding author e-mail:

The nearshore environments in the Gulf of Alaska are strongly influenced by both large-scale oceanographic patterns and runoff from tidewater and land terminating glaciers. As Gulf watershed glaciers lose volume and marine heat waves continue to increase in frequency and intensity as forecasted, nearshore physical properties will also change, directly impacting the marine ecosystems that thrive in these environments. Fjord geometry, particularly sills, influence how the heat and freshwater are exchanged between the outer ocean and inner fjords as well as their distribution in the water column. We use a harmonized dataset of previously collected physical oceanographic data from 50 fjords over the past 60 years to compare and examine temperature and salinity trends in the inner fjord nearshore environments across the Gulf of Alaska. Overall, data in these fjord environments is limited and disparate, with the exception of a few well studied locations. We describe the spatial and temporal variability of temperature and salinity data in these environments across the region. We use this dataset to describe temperature and salinity trends across these fjords and compare these across different regions within the Gulf of Alaska. Factors that influence the differences in physical properties both spatially and temporally, include sill depth, local climate, winds, and watershed characteristics. We discuss how all these factors, but particularly sill depth, influence the observed differences of temperature and salinity in fjords across the region. Lastly, we relate the trends to regional warming and cooling trends across this time period.


Spatiotemporal variability in basal melt derived from sequential radar echograms for ice shelves in northern Greenland and Antarctica

Cody Barnett, Leigh Stearns, John Paden, C. J. van der Veen

Corresponding author: Cody Barnett

Corresponding author e-mail:

Submarine melt, enhanced from the intrusion of warm ocean water or a subglacial plume, can impact iceberg calving and grounding line retreat of tidewater glaciers. In addition, basal crevasses, which can grow through fracture or submarine melt, modulate the location and occurrence of discrete calving and rift formation. While submarine melting and basal crevassing can occur coincidentally, the relationship between the two processes is unconstrained, in part because of observational limitations in sub-shelf environments. To address these knowledge gaps, we use NASA Operation IceBridge (OIB) radar echograms to compile a high-resolution multi-annual record of basal melt and localized grounding line migration for glaciers in northern Greenland and Antarctica. Our novel methodology isolates basal melt between consecutive epochs by correcting for spatiotemporal variability of ice advection, dynamic thinning, surface mass balance, and tidal states between echograms. Our results show no significant change in submarine melt rate magnitudes over the observation periods for each glacier in Greenland (roughly 2010–2019). Conversely, the spatial variability in basal melt both within and outside of basal crevasses varies greatly between glaciers (from 5 m a–1 at 79N Glacier, to greater than 50 m a–1 at Petermann Glacier), with the largest melt rates proximal to the grounding line and along the inner walls of basal crevasses. Additionally, our findings highlight a distinct pattern of basal crevasse initiation and vertical growth near the grounding line, crevasse widening (no vertical growth) as individual crevasses advect down flow, and re-initiation of new crack-tips atop the apex of the pre-existing crevasses closer to the terminus. Lastly, while melt rates near the grounding zone varied highly, basal melt proximal to the terminus was consistently 5–15 m a–1 for all glaciers. Our work highlights the importance of collecting repeat, long-term, observations of sub-shelf environments needed to parameterize poorly constrained processes.


Taku Glacier’s swan song?

Jason Amundson, Ian Joughin, Chris McNeil, Roman Motyka

Corresponding author: Jason Amundson

Corresponding author e-mail:

Taku Glacier, Alaska, USA, was in the advance phase of the tidewater glacier cycle during the 20th century and early part of the 21st century. The glacier’s advance was facilitated by the rapid excavation of soft marine sediments and the development of a large proglacial moraine that protected the glacier from the ocean and allowed the glacier to thicken. Eventually the moraine rose above sea level, which created an excellent platform from which to investigate the dynamics of an advancing tidewater glacier. However, climate change appears to have caught up with the glacier. Around 2015 the glacier began to thin and retreat. It has since thinned by several meters, retreated about 50–100 m, and a moat has started to form around its terminus. As one of the small number of tidewater glaciers worldwide that is not already in a calving-induced retreat, the ongoing changes at Taku Glacier provide a rare opportunity to investigate the processes leading to tidewater glacier destabilization, particularly in a sediment-rich environment. Here, we use observations from satellite and UAV imagery to take an early look at what appear to be the first stages of tidewater glacier retreat, with a focus on elucidating the timescales over which a tidewater glacier transitions into unstable retreat.


Investigating feedbacks between time-varying terminus morphology and spatial patterns in frontal ablation at a tidewater glacier

Nicole Abib, David Sutherland, Jason Amundson, Christian Kienholz, Dan Duncan, Emily Eidam, Rebecca Jackson, Roman Motyka, Jonathan Nash

Corresponding author: Nicole Abib

Corresponding author e-mail:

Ongoing changes in mass loss from tidewater glaciers and the input of fresh water into glacial fjords are influenced by both submarine melting and iceberg calving, collectively termed frontal ablation. Frontal ablation can change the geometry of a glacier’s terminus, influencing glacier dynamics, the fate of upwelling subglacial discharge plumes, and the distribution of submarine melt input into the ocean. Historically, directly observing time-varying terminus morphology has been difficult, limiting our understanding of subsurface ice–ocean processes. Here we use a novel dataset from LeConte Glacier, Alaska, USA, to investigate the temporal evolution of the subsurface terminus and relate it to the spatial patterns and drivers of frontal ablation. We combine high-resolution maps of the glacier’s submarine terminus from repeat multibeam sonar imaging with concurrent observations of subaerial geometry derived from terrestrial radar interferometry and time-lapse imagery collected during three field campaigns between 2016 and 2018. In each season, we observe a terminus morphology that is distinctly three-dimensional and varies spatially across the subsurface terminus. Unique geometries exist seasonally, with a 100 m undercut subglacial discharge outlet in summer and an ice ramp that protrudes 150 m into the fjord at the same location during low subglacial discharge periods in spring and fall. In addition, we find that the highest frontal ablation rates in summer occur directly above the subglacial discharge outlet, but during low subglacial discharge, frontal ablation rates reach a maximum on either side of the ice ramp. Finally, our data allow us to disentangle the spatial patterns in subaerial iceberg calving from the subsurface frontal ablation, improving our ability to identify the ice–ocean processes that control the evolution of the glacier terminus.


Terminus position change of Greenland’s peripheral marine-terminating glaciers 2013–21

Phoebe Kinzelman, Ellyn Enderlin

Corresponding author: Phoebe Kinzelman

Corresponding author e-mail:

Mass loss from marine-terminating glaciers is strongly controlled by terminus position change, and terminus positions are typically manually delineated and time series are therefore sparse in space and/or time. Detailed records of terminus position change have revealed important insights into controls on terminus stability, but these datasets have been limited to a small number of Greenland’s large outlet glaciers. Here we apply an automated terminus mapping method to the Landsat 8 record for marine-terminating glaciers spanning Greenland’s periphery. With this extensive record, we analyze spatial variations in seasonal to multi-year trends in variability of terminus position. Further expansion of this analysis to Greenland’s ~600 peripheral glaciers over the full Landsat time series will inform our understanding of what controls terminus change and help predict future glacier stability.


Subglacial discharge plume at LeConte Glacier, Alaska, USA: evolution from upwelling to outflowing

Bridget Ovall, Rebecca Jackson, Jonathan Nash, Nicole Abib, Dylan Winters, David Sutherland, Jason Amundson, Roman Motyka, Erin Pettit

Corresponding author: Bridget Ovall

Corresponding author e-mail:

Tidewater glaciers can have a strong influence on fjord systems through their freshwater contribution. In particular, the input of freshwater at depth via subglacial discharge introduces a buoyancy force and mixing as the subglacial discharge plume rises. The mixing of deep waters and associated upwelling provides heat to the glacier face, brings nutrients towards the surface, and induces an overturning circulation within the fjord. Many previous studies have attempted to quantify the influence of subglacial discharge plumes on fjord processes, but have been challenged by a dearth of near-glacier data. In this study, we utilize an extensive dataset of velocity, temperature, and salinity obtained via autonomous kayak near the terminus of LeConte Glacier in southeast Alaska, USA. This intensive sampling took place within 400 m of the glacier terminus over a 1-week period in September 2018, including observations less than 50 m from the ice face and near the subglacial discharge outlet. We observe both stable and unstable water parcels with distinct influences of subglacial discharge and submarine melt. Combining the near-terminus data with concurrent downstream ocean measurements and repeat multibeam scans of the submarine glacier terminus, we explore the evolution of the subglacial discharge plume from upwelling to outflowing in the context of evolving terminus morphology. Additionally, comparison with buoyant plume theory provides insight into how well such models are able to represent the upwelling phase of the subglacial discharge plume and the significance of subsequent modification of water properties as the plume transitions from upwelling to outflowing.


Wet firn cores and alpine weather station data constrain Alaska maritime glacier firn densification

C. Max Stevens, Louis Sass, Caitlyn Florentine, Emily Baker, Christopher McNeil, Shad O’Neel

Corresponding author: C. Max Stevens

Corresponding author e-mail:

Firn meltwater percolation, refreezing, and runoff are key physical processes in the hydrology of glacierized basins and glacier mass balance. A warming climate causes more meltwater and less firn porosity; quantifying those changes is essential for mass balance studies and prognostic simulations. However, uncertainties in model physics limit the real-world fidelity of wet firn simulations, and observational constraints on the firn of mountain glaciers are sparse. The USGS Benchmark Glacier Project has recovered firn cores in spring and fall from a fixed location on the upper bench of Wolverine Glacier in Alaska’s Kenai Mountains, USA, since 2016. We use firn data from the site and local meteorological data collected from a glacier-adjacent weather station to observationally constrain Community Firn Model simulations of firn densification and hydrology. Measurements for each firn core include density, ice-lens stratigraphy, and annual layer identification. The 2009 eruption of Redoubt Volcano left an easily identifiable ash layer on the glacier, and we use this well dated layer along with other identified layers (typically the previous year’s summer surface) to track how the firn has evolved on seasonal and annual timescales. For example, in the spring and fall 2018 cores the distance and mass between the 2009 and 2017 annual layers was constant, indicating that the firn experienced little vertical strain and little to no water mass change over the 2018 summer melt season. This suggests that there was minimal meltwater percolation into the deep firn, or that meltwater percolation into the deep firn (inflow) was equal to liquid water outflow. We use discrepancies between field observations and firn model outputs to quantify model uncertainty and to demonstrate the challenges in modeling wet firn.


Reproducible investigations of maritime glaciers using open-source tools

Jessica Scheick

Corresponding author: Jessica Scheick

Corresponding author e-mail:

Multiple open-source software (OSS) packages developed by and for the glaciological community enable rapid investigations of maritime glaciers. Focusing on Alaskan maritime glaciers, we illustrate how icepyx and other community-built software packages can be leveraged to quickly explore ICESat-2 data in combination with data from other sensors for a given glacier. The first tool showcased, the Python package icepyx, was created by the author in response to challenges faced by the glaciology community in accessing ICESat-2 data programmatically. With icepyx, we query and quickly visualize ICESat-2 data of the glacier. Then, we construct a time series of elevations spanning the ICESat, IceBridge, and ICESat-2 sensors using the NSIDC IceFlow package. Last, we customize our ICESat-2 data analyses with in-cloud processing using SlideRule. The workflow, encapsulated within an executable Jupyter Notebook, showcases the tools’ ease of use for data access, analysis, and visualization while demonstrating the application of FAIR (Findable, Accessible, Interoperable, Reusable) principles and collaborative development in glaciological research.


Fjord circulation in Greenland: freshwater sources, mixing, and the estuarine exchange flow

Lois Anderson, Rebecca Jackson, Robert Chant, Elias Hunter, John Wilkin

Corresponding author: Lois Anderson

Corresponding author e-mail:

Fjords exert controls over both the delivery of oceanic heat to Greenland’s tidewater glaciers and the outflow of glacial freshwater to the ocean. Fjord conditions and fjord-scale circulation likewise modulate the submarine melt rates at glacier termini, as well as the final properties of glacially modified water entering the continental shelf ocean, the details of which have implications for larger-scale ocean circulation. Upwelling subglacial discharge plumes entrain warm, saline Atlantic waters residing deep in these fjords. These sites of vigorous mixing and buoyancy input modify fjord-scale stratification and set up an estuarine-type exchange flow. At present, the impacts of fjords’ exchange flow on both glaciers and the ocean remain poorly constrained, and represent an important gap in our ability to model changes in both systems. In this work, we aim to understand the impact of two freshwater inputs on the fjord exchange flow: subglacial discharge plumes originating at depth, and terrestrial runoff entering at the surface. We have developed a quasi-realistic numerical model of Kangerlussup Sermia Fjord (KS) in Greenland’s Uummannaq fjord system using the Regional Ocean Modelling System (ROMS) to explore these dynamics. In conjunction with realistic bathymetry and stratification, we implement a new parameterization for subglacial discharge plumes that improves upon the common approach for representing plumes in fjord-scale models, and we compare our model with near-glacier observations of plume structure and properties. Building on frameworks from the estuarine literature, we calculate the total exchange flow (TEF), quantify salt and heat transports, and identify spatial patterns of mixing in the fjord. While the exchange flow is predominantly driven by the subglacial discharge, as expected, we find that a relatively small input of surface runoff can play an important role in modifying the stratification and thereby the structure of the subglacial discharge-driven flow. We also explore how various drivers of mixing within the fjord result in adjustments to the total exchange flow. Consideration of these interactions at the fjord-scale is critical for understanding the exchanges of heat and freshwater between glaciers and the ocean.


Early 21st century evolution of Johns Hopkins Glacier and Inlet

Lynn Kaluzienski, Jason Amundson, Andrew Bliss, Jaime Womble

Corresponding author: Lynn Kaluzienski

Corresponding author e-mail:

Despite the clear impact of glaciers on fjord environments, little work addresses glacier–fjord–ecosystem interactions. We characterize the evolution of the glacier–fjord environment at Johns Hopkins Glacier and Inlet, which hosts one of the largest seasonal aggregations of harbor seals in southeast Alaska, USA. Harbor seals use icebergs in the fjord as critical habitat for pupping, molting, and foraging. Johns Hopkins has been advancing since the 1950s; we focus on changes that have occurred in the last 20 years. Satellite-derived velocity measurements and digital elevation models indicate the glacier has continued to advance and thicken and has slowed down in recent years. Analysis of aerial photogrammetric surveys indicates that concurrent with the slowdown was a decrease in the iceberg abundance and changes in the iceberg size distributions, trending toward a higher percentage of small icebergs. Moreover, satellite and aerial observations indicate the surfacing of a terminal moraine in 2019. We see the influence of this growing moraine throughout many of our datasets over the past two decades, such as an overall decrease in iceberg discharge as the terminus became more grounded. Furthermore, in contrast to most tidewater glaciers, peak velocities occur ~2 km upstream of the terminus, and seasonal variations in velocity appear to be strongly associated with glacier runoff (as opposed to fluctuations in terminus position). Preliminary observations indicate the recent reduction in iceberg concentration has caused the seals to migrate to the nearby McBride Inlet.


The role of ice as habitat for harbor seals in tidewater glacier fjords

Jamie Womble, Perry Williams, Linnea Pearson

Corresponding author: Jamie Womble

Corresponding author e-mail:

Tidewater glaciers play an important role in landscape and ecosystem processes and undergo substantial seasonal variation which can influence freshwater flux, vertical mixing, nutrient input, and biological productivity, and create important habitats for invertebrates, fish, seabirds, and marine mammals during critical phases of the annual cycle. Ice that is discharged from the terminus of tidewater glaciers produces icebergs that serve as resting, pupping, and molting habitat for some of the largest seasonal aggregations of harbor seals (Phoca vitulina richardii) in the world. Harbor seals exhibit a high degree of seasonal fidelity to tidewater glacier fjords during energetically demanding life-history phases of the annual cycle, including pupping and molting. Although tidewater glaciers are naturally dynamic, advancing and retreating in response to local climatic and fjord conditions, most of the ice sheets that feed tidewater glaciers in Alaska are thinning and, as a result, many of the tidewater glaciers are retreating. However, the potential impacts of changes in ice availability and characteristics and the relationship with the distribution and abundance of harbor seals are unknown. To understand the impacts of changing ice availability on harbor seals, we quantified annual and seasonal changes in the availability of ice by developing multivariate spatial models to jointly model stage-structured seal location data and ice in Johns Hopkins Inlet, Glacier Bay National Park in southeastern Alaska, USA. There was substantial interannual and seasonal variation in the amount and spatial distribution of ice, with 2011 and 2012 standing out as particularly icy years. Ice habitat was substantially reduced in 2018 and 2019. On average there was 7.8 times more ice during June than August, which was likely driven by seasonal variation in physical processes that influence the calving dynamics of tidewater glaciers. Nonpup seals and ice were correlated during the pupping season (λ1,3 = 0.27; 0.23–0.31), but this correlation was reduced during the molting season (λ1,3 = 0.22; 0.17–0.27) suggesting that seals may respond to changes in habitat differently depending upon trade-offs associated with life-history events, such as pupping and molting, and energetic costs and constraints associated with the events.


Mass continuity constraints on outlet glacier surface mass balance in Greenland

Dakota Pyles, Timothy Bartholomaus

Corresponding author: Timothy Bartholomaus

Corresponding author e-mail:

Measurements of glacier mass balance in low-elevation, fast-moving, rapidly ablating regions are sparse. In large part due to logistical challenges, measurements from traditional weather stations are rare. These low elevations are also the areas with the greatest disagreement between modeled and observed surface mass balance, and among models. To fill the need for improved understanding of surface mass balance in these observation-poor regions, we are developing and testing a new method to estimate summer ablation at low elevations of the Greenland Ice Sheet. Our method, based on the mass continuity equation, combines ice thickness and ice flow velocity fields with measurements of elevation change to estimate the ablation from the ice surface. Data for these needs comes from BedMachine, MEaSUREs, and ICESat-2 products respectively. In this presentation, we present our method along with early results, validated where possible at observation sites from GC-Net and PROMICE automated weather stations.


Community perceptions of the impact of glacier loss on salmon and people in coastal British Columbia and Alaska

Annika Ord, Michele Koppes

Corresponding author: Michele Koppes

Corresponding author e-mail:

Glaciers along coastal British Columbia, Canada, and southeast Alaska, USA, are part of a dynamic system that connects mountains to the sea, drives local food webs, and supports fishing-based livelihoods and cultures. As the climate warms and glaciers recede, shifting runoff patterns and warmer temperatures are expected to impact salmon runs, coastal food webs, and livelihoods. Salmon are central to coastal and indigenous peoples’ way of life; they support coastal communities and livelihoods, shape cultural values, and provide essential food and nutrients to species ranging from the riverbed to the tops of mountains. Glaciers, too, are a powerful presence in this region. Ice has shaped the landscape through countless cycles of advance and retreat, delivers important nutrients and minerals from the land to the sea that help drive rich marine ecosystems, and modulates the effects of climate variability on stream temperature and flow. Glaciers have both displaced and provided new habitat for people and salmon for millennia. This study seeks to learn from indigenous and place-based knowledge to better understand how glacier loss is impacting coastal ecosystems, hazards, local food resources, and culture in communities in the region. There is growing concern about the impact of accelerating glacier retreat due to climate change on salmon runs. Much attention has focused on the biophysical relationship between glaciers and salmon; however, most studies are not engaging with local communities to understand local concerns and observations of glacier change and its impacts. Through talking with indigenous communities, subsistence and commercial fishermen, resource managers and scientists, we seek to integrate local knowledge of these system linkages and identify community-based questions that can help guide future research. By exploring the relationship between glaciers and coastal systems from the people who know these lands best, most especially from those who have adapted to climate change and glacier advance/retreat for millennia, there is a unique opportunity to understand how the loss of glaciers has and will continue to impact communities and essential ways of life along this glaciated coastline.


Radar sounding of Alaskan glaciers by NASA’s Operation IceBridge

Jack Holt, Brandon Tober, Martin Truffer, Michael Christoffersen, Christopher Larsen

Corresponding author: Jack Holt

Corresponding author e-mail:

Airborne radar sounding data were acquired over many glaciers in the Gulf of Alaska region during the period 2014–2021 as part of NASA’s Operation IceBridge, revealing new information about ice thickness and subglacial bed topography. We used radars at 2.5 MHz and 5 MHz, both impulse and chirped, the latter employing pulse compression and Doppler filtering to improve the signal-to-noise ratio and resolution. Scanning lidar data acquired simultaneously allowed for verification of the surface elevation in order to register the bed echo time delays. Bed data were interpreted using the Radar Analysis Graphical Utility (RAGU) and validated with surface clutter analysis to discern bed echoes from those arising from nearby surface topography. This employed a regional DTM and a facet-based geometric optics algorithm to simulate surface clutter, which is a serious obstacle to sounding mountain glaciers from an airborne platform. Useful new data were acquired over portions of larger glaciers including Malaspina, Hubbard, Bering, and the Bagley Icefield, in addition to more limited data for many other glaciers in the region. Malaspina Glacier exhibits deep troughs that reach 350 m below sea level approaching the terminus where ice-cored forelands are degrading rapidly via thermokarst processes. The Hubbard terminal lobe likewise exhibits an overdeepening 500 m below sea level, while Bering Glacier exhibits a distinct trough extending its full length, from the Bagley to its terminal lobe where it splits into multiple channels. The western Bagley contains ice ~1.4 km thick, with the bed well below sea level there. This is the thickest temperate ice ever sounded with an airborne radar. Aspects of subglacial trough location and morphology beneath Malaspina, Bering and Bagley correlate well with regional faulting in the context of the St Elias tectonic regime. Results from this radar sounding effort show significant discrepancies when compared with recent ice thickness estimates in the region based on mass conservation constrained only by ice velocities and surface elevations. Data and clutter simulations have been archived with NSIDC.


Snow accumulation results from airborne radar in the Gulf of Alaska

Michael Daniel, Brandon Tober, Jack Holt, Michael Christoffersen, Jilu Li, Fernando Rodriguez-Morales, Chris Larsen

Corresponding author: Michael Daniel

Corresponding author e-mail:

Snow accumulation is a critical parameter to constrain for estimating glacier mass balance and modeling glacier dynamics to predict future response to climate change. The rugged terrain and vast area of glaciers in the Gulf of Alaska region inherently leads to difficulty in measuring snow accumulation on glaciers from the surface, and satellite-based methods are generally inaccurate and need validation. Data acquired as part of NASA’s Operation IceBridge by the airborne snow radar developed and operated by the University of Kansas Center for Remote Sensing of Ice Sheets (KU CReSIS) shows great promise in quantifying snow accumulation over the catchment areas of Alaska’s maritime glaciers. The KU CReSIS snow radar is a frequency-modulated, continuous-wave system that operates between 2–18 GHz that was flown over glaciers in southern Alaska, USA, in May 2018 and May 2021. Key areas of interest surveyed include the upper portion of Hubbard Glacier, the upper portion of Seward Glacier (catchment area for Malaspina Glacier), and the Bagley Ice Valley (catchment area for Bering Glacier). The KU CReSIS data exhibit reflectors meters below the surface, sometimes with multiple reflectors, and we have digitized these reflectors in an effort to map snow depths within these catchments. A few high-altitude sites such as Mt Bona were also flown and exhibit many layers that may be correlatable with data from cores. With an adequate understanding of the data, we may be able to develop improved algorithms for satellite-based methods to measure snow accumulation.


Flow and ice flux history of the Malaspina Glacier, Southeast Alaska, USA, from Landsat 5 to 8 and Sentinel 1–2 images

Victor Devaux-Chupin, Martin Truffer, Mark Fahnestock

Corresponding author: Victor Devaux-Chupin

Corresponding author e-mail:

Malaspina Glacier is fed by the St Elias mountains of southeast Alaska, USA, and spills onto a coastal plain into the planet’s largest piedmont glacier. It has been rapidly thinning and the degradation of debris- and vegetation-covered ice along its perimeters opens the possibility of exposing the glacier to sea water and accelerating its retreat, particularly in view of a greatly overdeepened glacier bed. A multidisciplinary effort is under way to characterize the forelands and describe the climatic forcing to help build a numerical model with the goal of assessing its future behavior. Here, we analyze a 37-year record of ice velocities to better understand the dynamics of the glacier. Malaspina Glacier is fed by three tributaries, all of which show surging behavior. Seward Glacier supplies the majority of the ice to the Malaspina lobe, and its surges have the largest effect on it. The surges emanate from Seward Glacier mostly, but not exclusively, and follow a deep subglacial channel, all the way to the glacier front. From the velocity time series and ice thickness data, we are also able to reconstruct a history of ice flux into the Malaspina lobe, an important constraint on its mass balance. To further constrain the mass balance history we digitize snow lines using optical and radar satellite imagery.


Exploring the role of frontal ablation parameterization in projected 21st-century mass change for Northern Hemisphere glaciers

Jan-Hendrik Malles, Fabien Maussion, Will Kochtitzky, Lizz Ultee, Luke Copland, Ben Marzeion

Corresponding author: Jan-Hendrik Malles

Corresponding author e-mail:

Marine-terminating glaciers cover roughly one-third of the Northern Hemisphere’s glacierized area (outside the Greenland ice sheet) and constitute a relevant fraction of the mass turnover in some regions with maritime climate. Their direct freshwater export to the oceans has the potential to not only change global mean sea level, but local and regional ocean circulation patterns as well. Due to the interrelation of surface and frontal mass budget, such glaciers are subject to different dynamics than land-terminating ones, which are only forced by the atmosphere. Thus, if frontal ablation is not explicitly treated in a mass-balance model, calibration might lead to an overestimation of marine-terminating glaciers’ sensitivity to atmospheric temperatures. However, most large-scale glacier models are not yet able to account for this and marine-terminating glaciers’ frontal processes remain elusive. We explore this issue by implementing a simple frontal ablation parameterization in the Open Global Glacier Model (OGGM) and calibrating it with satellite-derived estimates. One of the major changes this entails is the lowering of marine-terminating glaciers’ sensitivities to atmospheric temperatures in the surface mass-balance model, which is calibrated with independent satellite-derived mass change data. Additionally, we account for mass changes below the water level, which are often neglected in observational mass-balance estimates. We then use the calibrated model, forced with an ensemble of atmospheric temperature and precipitation projections from the Climate Model Intercomparison Project’s sixth phase (CMIP6), to project Northern Hemisphere glacier mass change until 2100. The aim of this work is to investigate the influence of including frontal processes on these projections. To better represent dynamical processes occurring at marine-terminating glaciers’ fronts, we also incorporated water-depth dependent sliding and the hydrostatic stress-balance at the terminal boundary into OGGM. We find that including frontal processes, but ignoring changes in ocean climate, reduces the spread between different emission scenarios at the end of the 21st century. Total mass loss projected for Northern Hemisphere glaciers is reduced by ~7% in 2100, while the reduction we find for marine-terminating glaciers is ~23%. Moreover, the projected amount of frontal ablation is almost independent of the climate scenario applied.


Monitoring spring meltwater infiltration at Wolverine Glacier, Alaska, USA

Verenis Lucas, Timothy Bartholomaus

Corresponding author: Verenis Lucas

Corresponding author e-mail:

Maritime glaciers are known for receiving high snowfall during winter. During the transition from winter to the summer melting season, snowmelt is initially absorbed by this snowpack and ultimately delivered to the subglacial hydrologic system. However, the impact of a glacier’s late-winter snowpack on the hydrological system within glaciers is not well understood and the buffering and absorption of early meltwater creates the potential for the subglacial system to be ‘shock’ by the abrupt water delivery later in the summer. Under ongoing climate change, higher meltwater discharge earlier in the melt season may alter glacier flow, mass loss, and ecosystems in unappreciated ways. In large part, the limited understanding of winter to spring glacier hydrology results from a profound dearth of observations. To understand how the spring snowpack impacts the delivery of meltwater to the glacier bed, four temperature strings were installed at a single site at Wolverine Glacier, Alaska, USA. These temperature strings were installed near the equilibrium line on 13 May 2021, with thermistors spaced vertically every 0.7 m within a 5.1 m thick snowpack. We observe that the infiltration of liquid water is the main driver of warming in the snowpack in addition to the heat released during refreezing. At installation, the deepest portion of the snowpack started near –5°C and warmed through a series of steps associated with warm air temperatures until the snowpack became completely isothermal on 6 June, giving a total of ~3 weeks of melting. The water infiltration measurements will be compared with co-located measurements of glacier motion from a GNSS station. The snow temperature dataset from this project will provide insight into snow’s role in the beginning stages of glacial hydrology.


Sediment dynamics and erosion in the subglacial environment; sedimentological analysis of ancient processes in Kennicott glacier, Alaska, USA

Bruno Belotti, Bartholomaus Timothy, Elizabeth Cassel

Corresponding author: Bruno Belotti

Corresponding author e-mail:

The established view of the subglacial environment is that modern glaciers are devoid of thick, widespread layers of subglacial sediments, and are therefore erosive across most of their basal surface. Yet the occurrence, distribution, residence time, and controls on subglacial till dynamics remain unknown, as sediments can only be identified through the ice locally in boreholes, or at very low resolution in geophysical profiles. Sediments are produced in the subglacial environment as a result of glacial erosion, mainly by quarrying and abrasion of the underlying bedrock, and while the presence of clasts at the ice–bed interface is necessary for abrasion, a layer of subglacial till just a few decimeters thick might be enough to substantially (if not completely) inhibit glacial erosion. The presence of thick layers of till beneath certain glaciers has been acknowledged for decades, along with the importance of till deformation on sediment dynamics and glacier motion. However, subglacial till is considered to be confined to limited areas or ‘pockets’, that are not representative of the general subglacial environment; the suppressing effect of the till would, otherwise, make large-scale erosion impossible. Yet, observations in Kennicott valley near the town of McCarthy, Alaska, USA, seem to suggest that sediments from before the Last Glacial Maximum (LGM) are still present at relatively shallow depths (<100 m), and are interbedded with lacustrine sediments (probably deposited within glacial lake Atna), glacio-fluvial sediments, and possibly glacial till, as exposed in the bluffs cut by the Kennicott proglacial stream. Such a setting suggests that LGM glaciers were not universally erosive as they were unable to reach the underlying bedrock, even in up-valley positions that are close to modern glaciated areas. Consequently, subglacial sediment re-mobilization might be way less efficient than previously thought, and glacial erosion rates might be overestimated for ancient glaciers. During summer 2022, we will investigate those exposed sediments, create stratigraphic columns, and date the strata in order to unravel their subglacial history and quantify erosion rates. Better constraints on sediment occurrence and dynamics in the subglacial environment will enhance our understanding of many related glacial processes, such as erosion, calving, and sediment yield, allowing for better reconstructions and predictions of glaciated landscapes.


Using distributed acoustic sensing (DAS) to constrain glacier surface melt

John-Morgan Manos, Brad Lipovsky, Andreas Fichtner, Dominik Gräff, Eileen Martin, Patrick Paitz, Fabian Walter

Corresponding author: Brad Lipovsky

Corresponding author e-mail:

Constraining glacier surface melt and supraglacial meltwater flow is crucial in quantifying meltwater contributions to freshwater resources and understanding spatiotemporal variations in glacier surface hydrology. Distributed Acoustic Sensing (DAS) is a relatively new geophysical research technology that senses the acoustic wavefield as strain rate at set intervals along a fiber optic cable, effectively turning the cable into a large, distributed seismic array with a range of up to 100 km. With a high sampling rate, we are able to observe the ambient surface acoustic noise generated by supraglacial turbulent water flow that contributes to total glacier meltwater discharge. Here, we use machine learning to both quantify and predict glacier meltwater discharge from the observed surface acoustic wavefield as sensed by the fiber optic cable deployed longitudinal to glacier flow. We utilize a Long Short-Term Memory (LSTM) approach to analyze 20 TB of DAS strain rate measurements on a 9 km cable through a 1 month time period during the 2020 summer melt season on the Rhonegletscher in Switzerland. The data is highpass filtered at 100 Hz and decimated to 30 s RMS measurements to isolate the surface hydrological noise from other environmental seismic sources. We find that our approach is able to predict glacier meltwater to ~0.1 unit of standard deviation. It is evident that diurnal oscillations in high-frequency acoustic noise correlates with the proglacial meltwater discharge. We show that DAS is a robust method for observing supraglacial hydrological phenomena and machine learning is capable of discharge interpretation of the glacier acoustic wavefield.


Characterization of relict ice in the Malaspina Glacier forelands using time domain electromagnetic induction and seismic methods

Michael Christoffersen, Tyler Kuehn, Natalie Wagner, Brandon Tober, Douglas Brinkerhoff, John Holt, Martin Truffer

Corresponding author: Michael Christoffersen

Corresponding author e-mail:

The forested forelands of the Malaspina Glacier, Alaska, USA, are pockmarked by thermokarst lakes and depressions, evidence of stagnant ice left behind as the glacier’s terminus has retreated over the past century. Continued degradation of the ice-cored forelands through thermokarst processes could allow for increased interaction between warm water from the Gulf of Alaska and Malaspina’s terminus, so an understanding of the stability of the forelands is critical to predicting the future of the Malaspina Glacier. Therefore, we seek to characterize the spatial distribution and character of relict ice in Malaspina’s forelands using ground-based geophysical measurements that probe the subsurface, in conjunction with remote sensing techniques. Towards this end, we developed methods to constrain the extent and character of ice in the subsurface using time domain electromagnetic induction (TDEM) soundings, seismic refraction tomography, and multichannel analysis of surface waves (MASW). The seismic methods we have applied are sensitive to structure in the upper 20 m of the subsurface, a zone with high ambiguity in our TDEM measurements. The TDEM measurements explore deeper than the seismic refraction and MASW measurements. Individual deterministic inversions of the TDEM and seismic measurements do not agree well at some measurement sites due to this mismatch in depth sensitivity. Inversion methodologies that reliably capture uncertainty are necessary to confidently interpret subsurface structure. To achieve this, we turn to Bayesian frameworks to individually and jointly invert the TDEM and seismic measurements. A Bayesian approach yields a robust assessment of the uncertainties involved and allows for a simultaneous joint inversion of the TDEM and seismic refraction data.


Radar mapping of Malaspina (Sít’ Tlein) Glacier’s bed reveals potential for instability

Brandon S. Tober, John W. Holt, Michael S. Christoffersen, Martin Truffer, Chris F. Larsen, Doug J. Brinkerhoff

Corresponding author: Brandon S. Tober

Corresponding author e-mail:

To address the role of bed geometry in the fate of Malaspina Glacier, we compiled measurements from 3250 line-km of airborne radar sounding profiles acquired by NASA’s Operation IceBridge (OIB) into the glacier’s first comprehensive bed map. Discrete radar-derived measurements were gridded through constrained Gaussian process regression to provide a detailed understanding of Malaspina’s bed geometry and ice thickness distribution. We find that two-thirds of Malaspina’s bed is grounded below sea level. Elevation exceeds 350 m below sea level along several deep channels dissecting the glacier’s bed. The most prominent of these channels extends ~35 km from the throat of Seward Glacier towards Fountain Stream at the glacier’s terminus, deepening by 300 m within ~15 km of the glacier’s toe. The locations and morphology of Malaspina’s subglacial channels indicate that they are at least partly controlled by regional faulting. Modeling the expected water routing across the glacier’s bed based on the subglacial hydraulic potential gradient, we find that, despite flow being generally well-distributed, there is a clear connection between subglacial drainages and topography. Altogether, this work determines an ice volume of nearly 700 km3 for Malaspina Glacier. Should the world’s largest piedmont glacier evolve into a tidal system, rapid retreat has the potential to contribute up to 1.4 mm to sea level rise.


Calving from tidewater glaciers as a source of kinetic energy to fuel enhanced submarine melt

Marguerite Shaya, Erin Pettit, Jonathan Nash, Jason Amundson, Dylan Winters, Rebecca Jackson, Dave Sutherland, Jasmine Nahorniak, June Marion

Corresponding author: Marguerite Shaya

Corresponding author e-mail:

A driver of tidewater glacier retreat and sea level rise, submarine melting at glacier termini depends on the flow regime of the near-terminus fjord environment. The kinetic energy released by iceberg calving can change or accelerate the fjord flow near a glacier terminus, suggesting that calving has the potential to increase submarine melt rates. Until now, the effect of calving on submarine melt rates has been unknown because of the limited availability of in situ observations. Here we present an analysis of the energetics of a submarine calving event and its effect on near-terminus dynamics, using direct observations of calving from LeConte Glacier/Xeitl Sít, Alaska, USA, coupled with simple physical models. Using an upward looking acoustic Doppler current profiler located directly beneath the calving event along with time-lapse photography of the fjord, we find that the calving event released 6.5 e+10 J of buoyant potential energy over 1 minute. We partition this energy into surface gravity waves, internal waves, mixing, turbulent dissipation, and the movement of water. The iceberg entrains water upwards during its rise and accelerates the horizontal flow of water towards the glacier at depth. These results suggest that iceberg calving could significantly enhance submarine melt rates, and that more research is needed into the feedbacks between calving and the process of melting.


Salty lakes, crevassed forests, and the non-retreat of a shrinking glacier: patterns of landscape development and decay along the Malaspina Glacier foreland

Michael G. Loso, Anna Thompson, Tahzay Jones, Martin Truffer, John Holt, Sydney Mooneyham, Brandon Tober, Mark Fahnestock, Chris Larsen

Corresponding author: Michael G Loso

Corresponding author e-mail:

Malaspina Glacier, the largest piedmont glacier in the world, covers nearly 5000 km2 (over a million acres) of the coastal landscape of Wrangell–St Elias National Park & Preserve in southern Alaska, USA. Many neighboring glaciers have already succumbed to tidewater glacier retreat, exposing large fjords like Columbia Bay and Glacier Bay, but in recent decades the terminus of Malaspina Glacier has barely moved. Protected from subaerial melt by a nearly continuous mantle of debris, the terminus of Malaspina has developed into a glacial jungle of extensive alder thickets, mature spruce forests, collapsing thermokarst pits, and growing ice-walled lakes. Occasional surges have periodically displaced discrete portions of the terminus as recently as 2021, but even those dramatic accelerations appear to have had little lasting impact on the foreland. In the first part of this presentation, we discuss the gradual evolution of this remarkable landscape, which has for over two centuries been developing on the surface and margins of stagnant, slowly melting ice. We use digital elevation models to quantify the gradual downwasting of this historically stable foreland, while documenting the concurrent processes of lake growth, vegetation succession, fluvial migration, and sediment deposition with satellite imagery, historic photos, and dendrochronology. In the second part of this presentation, we consider evidence for more dramatic changes on the horizon. Radar soundings show that much of the Malaspina Glacier bed sits well below sea level, protected from potentially catastrophic tidewater glacier retreat by only a narrow strip of vulnerable shoreline. In the summer of 2021, we mapped bathymetry and collected temperature and salinity profiles in the most prominent proglacial lakes at Malaspina Glacier. One of them, the rapidly expanding Sitkagi Lagoon, which is centrally located on the Malaspina lobe, is bordered by calving ice cliffs and is separated from the ocean only by a narrow boulder spit. We observed tidally controlled water flow in both directions through this outlet and detected clear evidence of saltwater intrusion into the lagoon. The other lakes showed no conclusive evidence of saltwater intrusion, but our early results from Sitkagi Lagoon confirm that Malaspina Glacier is now in direct contact with marine water and vulnerable to processes – including submarine melt and calving – that may lead to rapid disintegration of its hitherto persistent foreland.


Diffuse debris entrainment in glacier, lab, and model environments

Alan Rempel, Dougal Hansen, Colin Meyer, Lucas Zoet

Corresponding author: Alan Rempel

Corresponding author e-mail:

The small quantities of liquid water that line triple junctions in polycrystalline glacier ice form connected vein networks that can facilitate material exchange with underlying basal environments. For example, under conditions that cause liquid water to be sucked up into the ice, small debris particles can be mobilized and deposited along these vein networks. Diffuse debris concentrations that are commonly observed in ice marginal regions and along the exposed flanks of overturned icebergs might be attributed to this mechanism. However, the relative importance of vein transport to several proposed competing explanations has not been established. During recent cryogenic ring shear experiments that we conducted under controlled conditions over a variety of unconsolidated substrates, we observed significant diffuse debris entrainment in some cases and negligible diffuse debris entrainment in others. Accordingly, we developed an idealized model of vein transport that builds upon the formulations and experiments described decades ago in a series of papers by Nye and Mader, and is augmented to include advective transport and predict how vein liquid flow responds to changes in conditions along exterior surfaces of the polycrystal (e.g. the glacier base). The equilibrium ice–liquid phase behavior within a polycrystal tends to favor liquid transport towards colder temperatures and lower solutal concentrations, with deviations of the ice stress state from hydrostatic balance tending to produce additional suction towards anomalously low ice pressures. Even for idealized cases with hydrostatic ice stress states, differences in the rates of heat and solute transport can lead to changes in the relative importance of thermal and solutal gradients in controlling the magnitude and direction of Darcy transport. We apply our model framework to examine the extent to which observed diffuse debris concentrations are consistent with a vein entrainment mechanism, and highlight circumstances where a different mechanism must be invoked (i.e.where liquid is predicted to squirt out of the ice instead of getting sucked in).


Do we know enough about glacier tsunamis? A review

Efthymia Koliokosta

Corresponding author: Efthymia Koliokosta

Corresponding author e-mail:

Natural hazards have been widely discussed. Among the most significant are those associated with climate change,volcanic eruptions, earthquakes and tsunamis, which cover the majority of the literature findings. Although each of them is adequately discussed, literature has paid little attention on the combination of climate and tsunami hazards, despite that fact that glacier tsunamis are the most frequent and catastrophic tsunamis generated in Arctic regions. This research reviews the literature findings discussing glacier tsunamis in order to identify the gaps in the study of glacier tsunamis’ risk assessment and management, adaptation, mitigation and resilience building.


Determining the drivers of recent retreat at Sermeq Kujalleq (Jakobshavn), Greenland through large ensemble simulations

Ziad Rashed, Alexander Robel, Helene Seroussi

Corresponding author: Ziad Rashed

Corresponding author e-mail:

Sermeq Kujalleq (SK, aka Jakobshavn Isbræ) is the fastest flowing outlet glacier in western Greenland, draining more than 10% of the ice sheet surface. In the late 1990s, SK begun a rapid phase of retreat and mass loss, which led to ~0.9 mm of global mean sea-level rise between 2000 and 2010. High retreat and thinning rates at SK’s front have been attributed to an influx of warmer deep ocean waters to West Greenland, leading to collapse of SK’s floating ice tongue, and an acceleration in retreat and thinning. Since 2017, SK’s terminus has stopped retreating and begun to thicken and readvance as regional ocean temperatures have cooled. To investigate the primary drivers in the retreat of SK from 1985–2018, we use the Ice-sheet and Sea-level System Model (ISSM) to simulate large ensembles of SK’s evolution in response to warming ocean temperatures. We focus on three parametrized ocean-forced processes that could be driving retreat: submarine melt under floating ice, undercutting melt at the calving front, and calving modulated by mélange rigidity. We ran a large ensemble of more than 1000 simulations in which the sensitivity of each of these three processes to ocean temperature wass systematically varied over a realistic range. Each simulation was scored based off its ability to match observed position and geometry of the glacier calving front. There are multiple regions of parameter space where simulations match observations well, indicating potential compensation between different processes in driving retreat. Before the ice tongue collapse, SK was quite sensitive to submarine melting and responded rapidly to change in ocean temperature. Following the loss of floating ice, frontal ice loss via both frontal melt and calving (and the interaction between the two) takes precedence in driving thinning and retreat. We find that it is not possible to explain the rate of SK retreat without a role for calving modulated by ocean temperature, likely through changes in melange rigidity. We note a hysteresis effect in glacier front evolution as temperatures cool and comment on the significance of retreat mechanisms and their various response times to oceanic forcing.


Modeling of maritime ice cap Vatnajökull, Iceland

Guðfinna Aðalgeirsdóttir, Andy Aschwanden, Finnur Pálsson

Corresponding author: Guðfinna Aðalgeirsdóttir

Corresponding author e-mail:

Vatnajökull ice cap, on the southeast coast of Iceland, is by far the largest ice body on the island (~7700 km2, ~3000 km3). It is located in a maritime climate with high mass turnover and is covering several volcanically active regions, causing regular jökulhlaups to flow from it. Several of the largest outlet glaciers are surge-type glaciers that have had regular surges in the 20th century, but surge activity has not been observed in the last two decades. All the outlet glaciers of Vatnajökull have losing mass since 1995 at variable rates, closely related to the ocean temperature surrounding Iceland. Simulations that hindcast the evolution of Vatnajökull during the past few decades using the SIA model PISM and forcing from various regional climate model sources are presented. The simulations are validated with a wealth of observations that have been gathered about the flow field, both GPS point measurements and satellite-derived velocity fields. Several other sources of data such as aerial extent, mass balance measurements, elevation changes and geodetic mass balance are also used for validation.


Optimum multi-parameter analysis of glacially modified water in Sermilik Fjord, Greenland

Aurora Roth, Fiamma Straneo, James Holte

Corresponding author: Aurora Roth

Corresponding author e-mail:

Meltwater from the Greenland Ice Sheet is a critical freshwater source for the North Atlantic Ocean, impacting regional and global ocean circulation and local fjord ecosystems, and it is projected to increase in the coming decades. Freshwater input at the head of fjords from runoff, subglacial discharge, and submarine melting of tidewater glaciers and icebergs is diluted through mixing with other water masses present in the fjord and originating from the continental shelf. This dilution and mixing of the freshwater results in a new water mass termed ‘glacially modified water’ (GMW), which is then exported from the fjord. This transformation of glacial freshwater is driven by fjord circulation and fjord–shelf exchange processes that are not currently represented in large-scale ocean or ice sheet models despite their importance as a boundary condition. The goal of this study is to define characteristics of the GMW exported from Sermilik Fjord, Greenland by using multiple annual surveys of physical (temperature, salinity) and biogeochemical (oxygen, δ 18 O, nutrients) properties in the fjord. The methodology, building on earlier work, uses an optimum multi-parameter (OMP) analysis to estimate the water mass fractions (percentage of subglacial discharge, submarine melt, Atlantic Water, and Polar Water) constituting the GMW. The results of the OMP analysis will be compared to previous noble gas measurements which can provide an accurate quantification of GMW but are not always available. Identification of GMW and estimates of water mass fraction in this study will allow us to assess the interannual variability of GMW in Sermilik Fjord. The long-term objective of this work is to inform a first-order parameterization for GMW composition and export from glacial fjords needed in large-scale models.


Impacts of glacier change on aquatic food webs in southeast Alaska, USA

Eran Hood, Jason Fellman, Kara Pitman, Ryan Bellmore

Corresponding author: Eran Hood

Corresponding author e-mail:

Glacier volume loss can have profound impacts on the physical, chemical, and ecological properties of pro-glacial rivers and estuaries. For example, glacier runoff is a strong control on discharge and streamwater turbidity, both of which influence the abundance and diversity of aquatic organisms in receiving waters. This talk will explore the downstream impacts of glacier runoff on the chemistry and ecology of rivers and estuaries in southeast Alaska, USA. The export of biogeochemically important elements such as carbon, nitrogen, and phosphorus (C, N, and P, respectively) across the land–ocean interface will shift as glaciers retreat and contribute less to streamflow at the watershed scale. These changes in biogeochemistry have implications for aquatic food webs in freshwater and marine ecosystems. The loss of glaciers from coastal watersheds also influences habitat quality for keystone species like Pacific salmon. In this context, glacier loss presents both challenges and opportunities for Pacific salmon, which are an important cultural and economic resource in southeast Alaska. Challenges to salmon associated with glacier loss include the diminished availability of cold water during late summer spawning months. However, glacier recession also exposes new watersheds that contain streams available for colonization by salmon. Overall, developing a better understanding of the links between runoff from rapidly changing glaciers and the ecology of downstream ecosystems will inform future management of aquatic ecosystems in southeast Alaska, including the important resources contained within them.


The influence of glacier cover on iron cycling in Patagonian fjords

Jon Hawkings, Robert Sherrell, Tim Conway, Elizabeth Shoenfelt Troein, Katharine Hendry, Mattias Sieber, Alexander Beaton, Rodrigo Torres, Giovanni Daneri

Corresponding author: Jon Hawkings

Corresponding author e-mail:

Fjords are dynamic interface zones between fresh and marine waters and are hotspots of carbon burial. However, the importance of fjords as biogeochemical reactors remains uncertain, in part because they are relatively understudied compared to other aquatic critical zones. Specifically, few fjords are studied in the context of trace element cycling. Iron (Fe) plays an important role in the carbon cycle due to their importance as micronutrients for marine biota, complexation and association with macronutrients (e.g. organic C and P), and their influence for carbon burial in sediments (e.g. the ‘rusty carbon sink’). Glaciers are major contributors to fjord freshwater and sediment budgets, and have been postulated to be a significant source of Fe to downstream environments. Turbid glacial meltwaters carry elevated concentrations of labile (nano)particulate and dissolved (<0.45 μm) Fe that may be directly or indirectly available to biota, and tidewater glaciers drive a ‘meltwater pump’, upwelling fjord bottom waters to the surface. However, the impact of these meltwater Fe inputs downstream is debated due to rapid removal from surface water at low salinities. Furthermore, no information is available on concentrations, speciation and cycling in glacial meltwaters and downstream fjords in Patagonia, Chile. Fjords dominate the coastline of Chilean Patagonia, spanning over 14° of latitude, and receive freshwater inputs from pristine rivers draining hydrological catchments of variable glacial cover. The region therefore provides a natural laboratory to test hypotheses relating the importance of glacial cover on fjord biogeochemical cycles. Here we combine a complementary suite of surface water measurements (including size fractionated concentrations, reactive particulate phase concentration, Fe isotope measurements, electron microscopy, organic matter composition and synchrotron X-ray absorption spectroscopy) from four Patagonian fjord systems and 33 rivers, spanning a large latitudinal range and with catchments of no to high glacial cover, to elucidate the importance of glacial inputs in fjord Fe cycling and export further offshore. Here we will highlight the importance of particulate and organic carbon complexed phases in sustaining high surface water concentrations of Fe, and discuss the importance of glacier Fe input for ocean inventories around Patagonia.


Contrasting mercury sources and fluxes from adjacent glacierized and forested watersheds in Juneau, Alaska, USA

Sonia Nagorski, Andrew Vermiyea

Corresponding author: Sonia Nagorski

Corresponding author e-mail:

Mercury (Hg) export from glacierized watersheds to downstream freshwater and coastal ecosystems is poorly understood. Glacial erosion and seasonal glacier ice melt may release both geogenic sources of Hg and anthropogenically sourced Hg trapped in glacier ice. Here we present data on Hg concentrations, speciation, and yields out of two glacierized watersheds in the Juneau area, Alaska, USA, and compare them to an adjacent forested watershed. We found strong differences between outflows from the two glaciers, and we deduce that the contrast is driven by contrasting underlying lithology. The total export of Hg (780 g km–2 a–1) from the glacierized Mendenhall River (Áakw) was higher than that of any other stream in the literature outside of the Greenland ice sheet, exemplifying the power of glaciers in releasing trace elements by way of erosion, particularly if the bedrock acted upon is enriched in metals. The two glacierized rivers were similar in that they both carried Hg dominantly in the particulate fraction, whereas the non-glacierized stream carried it largely in the filtered fraction, at ~20-fold higher filtered concentration, and with a higher percent in its methylated form. However, the flux of total and methyl-Hg in the glacierized streams dwarfed that of the forested stream, as they are dominated by strong yields of water and sediment. Differences in speciation between the glacierized and forested streams are likely accounted for by glacier and watershed geochemical conditions that variably promote mercury methylation.


Permeability of temperate ice: sensitivity to ice-crystal size and liquid water content

Jacob Fowler, Neal Iverson

Corresponding author: Jacob Fowler

Corresponding author e-mail:

Results of glacier flow models that treat temperate ice deformation as a two-phase flow are sensitive to the value of ice permeability. We have constructed and begun using a custom, falling-head permeameter for measuring the permeability of temperate, polycrystalline ice as a function of its grain diameter, d, and liquid water content, φ. Chilled water is passed through an ice disk kept at the pressure-melting temperature while the rate of head decrease indicates permeability. Fluorescein dye in the water allows water-vein geometry to be studied using fluorescence microscopy. In these experiments, water flow over durations of seconds to hours is Darcian, and for grain diameters increasing from 1.7 to 8.9 mm, average permeability decreases from 2 × 10–12 to 4 ×  10–15 m2. In tests with dye on fine (d =  2 mm) and coarse (d =  7 mm) ice, average area-weighted vein radii are nearly equal: 41 and 34 μm, respectively. If included in a theory slightly modified from Nye and Frank, these average radii yield permeability values within a factor of 2 of best-fit values based on regression of the data. If vein radii are indeed generally independent of d, permeability values depend on d–3.4 rather than, as in theory, d–2. In another set of experiments, grain size is kept constant while water content is controlled by varying ice salinity and measured by inducing a freezing front at the ice-disk edges and tracking its propagation inward with thermistors in the ice. Fitting freezing-front arrival times with solutions of the relevant diffusion-controlled, moving-boundary problem provides the liquid water content. Experimental results, to date, indicate that permeability values increase by a factor of 30 as water content increases from 0.55 to 3.71%. The data obey a power-law relation, with permeability k = Cφa, where C is a geometric constant and a = 1.7, close to the value of 2.0 indicated by theory.


The past decades of glacier change at Malaspina Glacier and lessons for the future

Martin Truffer, Mark Fahnestock, Christopher Larsen, Victor Devaux-Chupin, Mike Loso, Michael Christofferson, John Holt, Anna Thompson, Brandon Tober

Corresponding author: Martin Truffer

Corresponding author e-mail:

Malaspina Glacier, fed by the large amounts of snowfall in the coastal mountains of southeast Alaska, USA, spreads into an iconic piedmont reminiscent of a river delta. Aptly named Sit T’lein, meaning Big Glacier, by the local Tlingit population, it is the biggest remaining glacier reaching the Gulf of Alaska. The neighboring Hubbard Glacier and Icy Bay glaciers have left open ocean bays after their retreats, which begs the question whether a similar fate is in store for Malaspina. A recently acquired comprehensive set of radar measurements shows that much of the glacier bed is indeed located well below sea level. However, the glacier has not yet significantly retreated from its Little Ice Age extent, instead, much of the peripheral ice has become stagnant as it is first covered by debris and later by vegetation. However, despite the lack of large-scale retreat, the glacier has thinned substantially. Here we analyze elevation information from digital elevation models and repeat altimetry profiles over the glacier lobe. We find persistent thinning of several meters per year over the clean ice areas of the lobe over the past two decades. The elevation change signal is complicated by the signature of several surges, which led to local areas of thickening and thinning rates up to 10 m a–1 in ice deposited by the surge. In the stagnant ice along the glacier’s perimeter the elevation change can be directly interpreted as surface mass balance and provides a comprehensive look at ice melt under debris and vegetation cover. Thinning rates of several meters per year are apparent under the debris with spatial patterns mimicking those of morainal bands in imagery. Vegetation covered ice, however, shows much more reduced thinning and a patchy pattern that reflects the formation of thermokarst ponds and the inland progression of proglacial water bodies. While future scenarios will be explored in more detailed flow and mass balance models, the evolution over recent decades suggests that with continued thinning and a deep-bedded glacier, Malaspina is prone to become more susceptible to retreat, as areas near the front will thin to near the floatation limit. A potentially stabilizing feedback is the rapid colonization of stagnant ice by vegetation, which appears to reduce melt significantly.


The human ecology of glacial retreat in a subarctic Alaskan fjord since 800 CE

Aron L. Crowell

Corresponding author: Aron L. Crowell

Corresponding author e-mail:

Coastal glaciers carved Yakutat fiord in southeastern Alaska, USA, and continue to influence its marine and terrestrial ecosystems. Glacial discharges support a thriving marine food web, ice floes are platforms for a harbor seal rookery, and biodiverse coastal forests grow near the ice. Na-Dene peoples (Eyak, Ahtna, and Tlingit) settled in this highly productive biome as the glaciers underwent climate-contrary retreat during the Little Ice Age. Collaborative, community-based research in archaeology, oral tradition, and traditional ecological knowledge demonstrate cultural niche construction at Yakutat since 800 CE, including the correlation of human settlement and foraging patterns with the complex biogeography of the fiord.


Glaciology, oceanography and early Americans

Thomas Royer

Corresponding author: Thomas Royer

Corresponding author e-mail:

The glaciers of southeast Alaska, USA, influenced the historical southward movement of the earliest Alaskan settlers in at least two ways. First, there was an impenetrable wall of glacial ice across North America that made overland travel southward impossible. Second, prior to the Last Glacial Maximum (LGM), seafaring migrants were able to use skin boats to detour around the glaciers along the western coastline of North America. As those glaciers began to melt after the LGM, the meltwater accelerated a northward coastal freshwater flow that greatly hindered or eliminated further southward human migration for centuries, beginning about 15 000 years ago. Evidence for changes in ocean conditions at that time has been found in deep-sea sediment records. Deep-sea archeological records that could support these changes in migration are not yet available.


Tidewater glaciers and dynamic landscape changes: three examples from southeast Alaska

Roman Motyka

Corresponding author: Roman Motyka

Corresponding author e-mail:

Glaciers are agents of landscape change and tidewater glaciers (TWG) in particular, can drive dramatic changes to landscapes and surrounding terrain. Here, we examine three examples of such systems found in southeast Alaska, USA. The first, Glacier Bay (Sít’ Eeti Geiyí) experienced an expansion of glaciers during the Little Ice Age (LIA) that reach its maximum ~1750 AD. The advancing ice filled the bay, blanketed the surrounding terrain, overran indigenous villages, and caused widespread land subsidence due to glacier isostatic adjustment (GIA). The subsequent post-LIA calving retreat, 120 km in less than 160 years, reopened the bay and surrounding terrain, exposing deep fjords and providing new habitat for flora and fauna. The cumulative post-LIA ice loss triggered regional isostatic rebound, with 6 m of total uplift in the Gustavus area and rates as high as ~3 cm a–1 in the upper bay. The second is Taku Glacier (T’aaḵú Ḵwáan Sít’i), a 55 km long and nearly 1500 m thick glacier located 30 km northeast of Juneau. It also reached its LIA maximum ~1750 AD, blocking the Taku River and creating a glacier-dammed lake that extended more than 70 km up-valley into British Columbia, Canada. The post-LIA collapse of the dam and subsequent outburst led to a calving retreat and the formation of Taku Inlet. Glacier re-advance was under way by 1890 AD with ice and sediments progressively infilling the inlet over the next several decades. Remobilization of sediments overridden by the advancing glacier resulted in excavation rates of 1– 3 m a–1. A warming climate appears to have finally caught up with Taku Glacier and a retreat is currently under way. The third is LeConte Glacier (Xeitl S´ıt’), which drains the Stikine Icefield into LeConte Bay (Xeitl Geeyi’). The glacier underwent a 2 km long calving retreat in the late 20th century with the terminus then temporarily stabilizing at a fjord constriction. LeConte Glacier has become well-known because of numerous studies on identifying and documenting the effects that a warm ocean has on submarine melting of TWG termini. Perhaps the most interesting current landscape change is the progressive buildup of submarine morainal banks. It is not clear at this point whether LeConte will further stabilize, and possibly re-advance (à la Taku Glacier), or retreat into an over-deepened basin where ice depths are well-below sea level for a considerable distance up-glacier.


Of subshelf melt, calving, and the break-up of Jakobshavn Isbræ’s floating tongue

Andy Aschwanden, Douglas C. Brinkerhoff

Corresponding author: Andy Aschwanden

Corresponding author e-mail:

The break-up of Jakobshavn Isbræ’s floating tongue in the early 2000s was triggered by an increase in ocean temperature in Ilullisat Fjord, Greenland, and thus increased sub-shelf melt which caused the floating to to thin until it eventually broke up. What appears to be conceptually simple remains a big challenge for ice flow models and to date, no model has shown skill in hindcasting the evolution of Jakobshavn Isbræ from the mid-1980s to 2020. Here we pair a new reconstruction of fjord temperature and salinity with an ice flow model that encompasses all the relevant process (calving, frontal- and sub-shelf melt, fracture dynamics) to generate a large ensemble of simulations. We then weigh ensemble members based on the skill to reproduce the wealth of direct and indirect observations such as surface speeds, terminus position, elevation change and sub-shelf melt rates. While the ensemble is able to reproduce the bigger picture of increasing ocean temperatures leading to increased sub-shelf melt and thus thinning of the floating tongue which results in an increase in surface speeds, it lacks the skill to reproduce the actual break up. This could either due to the challenge of calibrating the damage mechanics model or due to a missing process.


Sit’tu Kwaani (Spirit of the Glacier): Lingit beliefs and oral history of glaciers

Judy Dax̱ootsú Ramos

Corresponding author: Judy Dax̱ootsú Ramos

Corresponding author e-mail:

Glaciers are viewed by Lingit people as powerful beings. Many stories relate to the surging and retreat of glaciers because of some human violation of a taboo. Yakutat Tlingit who hunted seals near glaciers accumulated indigenous knowledge related to ice conditions and types of iceberg.


Tlingit oral histories and place names can provide a better understanding of the human history of Glacier Bay

Daniel Monteith

Corresponding author: Daniel Monteith

Corresponding author e-mail:

The archaeological record verifies that humans have been in the Glacier Bay area for over 10 000 years. Tlingit oral narratives and place names can give a better understanding how the Huna Kaawu responded to major glacial events and changes. The Tlingit place names in particular can give us a more specific human perceptions of the ecology and landscape during different time periods.


Evolving outburst flood hazards and impacts from glacier-dammed lakes: A case study of Mendenhall Glacier (Áakʼw Tʼáak Sítʼ), Alaska, USA

Amy Jenson, Jason Amundson, Christian Kienholz, Jonathan Kingslake, Eran Hood

Corresponding author: Amy Jenson

Corresponding author e-mail:

Glacier outburst floods are a common hazard in glacierized landscapes. These floods can threaten infrastructure and cause semi-regular but short-lived perturbations to downstream ecosystems. Outburst flood theory dictates that flood characteristics, such as event timing and peak discharge, depend on glacier and basin geometry, both of which evolve as glaciers advance or retreat. Consequently, outburst floods can be viewed as semi-periodic disturbances to glaciated landscapes that switch on/off and evolve in response to climate change. We use observations from Mendenhall Glacier, a maritime glacier that drains from the Juneau Icefield and terminates in Mendenhall Lake, Alaska, USA, to motivate a theoretical investigation of decadal-scale variations in outburst floods from glacier-dammed basins. Mendenhall Glacier has retreated over 4 km from its Little Ice Age maximum and has thinned by tens of meters. A significant impact of this retreat was the formation of Suicide Basin along the glacier’s margin, which resulted from the thinning and detachment of a tributary glacier. The geometry of the basin continues to evolve as remnant ice from the tributary glacier melts, icebergs calve into the basin, and the adjacent glacier continues to thin. Outburst floods originating in Suicide Basin have occurred annually since 2011. We have monitored the basin since 2012, and since 2018 we have conducted several UAV surveys each summer in order to quantify changes in basin storage capacity and volume of remnant ice in the basin. Our observations have allowed us to develop a framework for outburst flood evolution, which suggests that outburst flood magnitude will increase as long as there is remnant ice in a basin that is wasting away faster than the adjacent trunk glacier. Numerous other marginal basins appear to be forming around the Juneau Icefield as the icefield continues to thin and retreat, motivating further research into the impacts of these floods on landscape and ecosystem evolution.


Historical evolution and possible triggers for the retreat of Bear Glacier Lagoon, Alaska, USA

Journey Berry, Erin Pettit, Taryn Black, Deborah Kurtz

Corresponding author: Journey Berry

Corresponding author e-mail:

Bear Glacier Lagoon within Kenai Fjords National Park, Alaska, USA, is a fast-changing lake at the terminus of Bear Glacier. The lake may only be 50–60 years old, but there is very little information on the history and evolution of this lake and the ecosystem it supports. Given limited existing data, what can we say about historical evolution and modern dynamics of this glacier–lake system with tidewater attributes? We present our initial findings towards two interrelated goals. The first is to visually describe the evolution of Bear Glacier Lake over the last half century using historical maps and satellite imagery in the context of historical climate data. The second is to describe the modern physical and environmental factors that may begin to explain the glacier’s fast retreat and the lagoon’s overall circulation. These factors include the shape and bathymetry of the lake, the magnitude of outburst flood events from an upstream ice-marginal lake, and the local climate and connections to the ocean. Landsat imagery starting from 1984 allows for the creation of yearly lake outlines to better observe the retreat of the glacier, and to determine if the changes to lake levels and structure of the morainal spit are related to periodic outburst floods that have been identified by Park Service monitoring. The creation of this history and the initial analysis of lake dynamics will ultimately give insight into how Bear Glacier Lagoon was formed from retreat of Bear Glacier and the effect of potential future retreat on the entire Harding Icefield. This rapidly changing glacier–lake system nearby Seward, Alaska, will significantly impact the local ecosystem as well as the safety and extent of recreational opportunities.