Next generation ice-sheet bed measurements

Subglacial bed topography is a critical input to understand the stability of ice sheets and for computer models to provide reliable estimates of future seal level rise. These topography maps are constructed based on line measurements collected by airborne radars. These measurements generally provide information only directly underneath the aircraft, and the bed topography in-between measurements has to be interpolated. The first maps of bed topography were constructed based on little data, and using very simple interpolation methods. Since the 2010s, significantly more data have bene acquired by different international agencies, and novel, more advanced interpolation approaches have been proposed. One of them is the use of mass conservation, which combines the radar information with satellite-derived ice flow speed and the principle of conservation of mass. This approach makes it possible to estimate how the bed topography changes in between flight lines by using other datasets for which we have complete coverage. Other new methods involving machine learning, surface features, or stochastic modelling have also been proposed. In this talk, we review the evolution of mapping method over the last 20 years, since the first release of Bedmap, discuss the benefits and limitations of different techniques, and provide inputs for how these maps are continuing to improve our ability to model ice sheet flow and their future contributions to sea level rise. We also discuss how future field campaigns should be designed to optimise these new mapping techniques.

Professor Mathieu Morlighem

Dartmouth College, USA

Professor Mathieu Morlighem

Dartmouth College, USA

Mathieu Morlighem is a Professor in the Department of Earth Sciences at Dartmouth College. His research interests are focused on better understanding and explaining ongoing changes in the Cryosphere, as well as reducing uncertainties in the ice sheet contribution to sea level rise using numerical modelling. He is one of the founders and core developers of the Ice-sheet and Sea-level System Model, a new generation, high-resolution, higher-order physics Ice Sheet Model. He also developed and maintains BedMachine, a high-resolution subglacial bed topography dataset of the Greenland and Antarctic ice sheets.

It is tempting to think that the key challenge in generating new Bedmaps is to re-run an interpolation algorithm with an ever-enlarging dataset of ice sheet thickness, as new airborne surveys are completed. This does help, but other challenges emerge. Surveys often don't agree, large data gaps remain, the ice thickness changes, algorithms work well for one landscape but not another. Similar problems afflict other key Bedmap components: the coastline, the grounding line, the ice shelves, and the bathymetry. The process of merging the interpolated ice sheets and shelves, and the grounded bed with the sea floor, can also inject spurious cliffs and bumps in the grounding zone - exactly where ice sheet models are most dependent on accurate boundary conditions. In each case, the unintended errors and artefacts that arise require careful checking, correction, and sometimes bespoke, local approaches to the interpolation. Can Bedmap be automated? Probably, but each of these challenges must be understood and overcome in ways that are robust to the data thrown at them. And it might still be quicker to do it by hand.

Dr Hamish Pritchard

British Antarctic Survey, UK

Dr Hamish Pritchard

British Antarctic Survey, UK

Dr Hamish Pritchard is a glaciologist at the British Antarctic Survey specialising in mapping, monitoring and understanding the cryosphere through remote sensing, field surveys and instrumentation. His research targets the two outstanding icy issues facing the world as the climate changes: sea level rise, and the fate of mountain water resources. He was a lead author on the IPCC Special Report on Oceans and Cryosphere and currently leads Bedmap3, re-mapping the hidden bed of the Antarctic ice sheets, and the Natural Environment Research Council project ‘The Big Thaw’, filling gaps in our knowledge of mountain snow and ice.

Bed topography plays a pivotal role in shaping the evolution of the Antarctic Ice Sheet, controlling ice flow dynamics, grounding line retreat, subglacial hydrological processes, and ice shelf melt rates. Despite its importance, Antarctic bed topography remains sparsely sampled, resulting in significant uncertainties in bed datasets that propagate in ice sheet model simulations. Synthetic bed topography datasets, generated using mass conservation methods, geostatistical approaches, and other multi-method integrations, are increasingly utilised ice sheet modelling applications. Here, we explore the role of synthetic bed topographies in advancing ice sheet modelling. We review methodologies for generating synthetic topographies, their applications in modelling ice flow, and examine the impact of topographic uncertainties on simulated ice and subglacial water flow in the Aurora Subglacial Basin, East Antarctica.

Dr Felicity McCormack

Monash University, Australia

Dr Felicity McCormack

Monash University, Australia

Dr McCormack is a Senior Lecturer and Chief Investigator on the Australian Research Council Special Research Initiative Securing Antarctica's Environmental Future. Felicity's research focusses on understanding processes of ice flow, and the impact of climate variability and change on Antarctic ice loss. Her research uses ice sheet models, geophysical observations, and theory. In 2019, Felicity was awarded a Fulbright Postdoctoral Fellowship at the University of California, Irvine.

Accurate subglacial topography is essential for ice sheet models that project past and future ice-climate interactions. Widely used datasets of subglacial topography, such as those for the Greenland Ice Sheet (GrIS), rely heavily on interpolation, for which uncertainties remain poorly constrained. Improved knowledge of these uncertainties is necessary for the optimisation of the next generation of ice sheet bed measurements. We show that a novel quantile regression forest approach to radio echo sounding (RES) observations of the GrIS improves both estimates of derived bed elevations and uncertainty mapping. By providing improved uncertainty estimates, this method offers a way to optimise the design of future airborne RES surveys, ensuring more accurate subglacial topography mapping for ice sheet modelling efforts.

Dr Steven Palmer

University of Exeter, UK

Dr Steven Palmer

University of Exeter, UK

Steve has a PhD in Polar Remote Sensing from the University of Edinburgh and has previously held research positions at Scott Polar Research Institute, University of Cambridge and the University of Leeds. Since 2013, he has been based at the University of Exeter, where he uses a variety of Remote Sensing approaches and emerging machine learning techniques to investigate Earth surface processes. His main research focus has been quantifying changes in glacier hydrology and ice flow in the context of climate warming. Using airborne ice penetrating radar methods, Steve has contributed to our understanding of conditions at the base of the Greenland ice sheet - better understanding of which is key to making more accurate projections of how the ice sheets are likely to change under continued global heating.

The margin of the Antarctic ice sheet has multiple tipping mechanisms that could trigger large, rapid, and potentially irreversible changes in the coming centuries. While bed topography largely controls these mechanisms, it remains poorly known beneath the ice sheet and ocean. Antarctic RINGS, a collaborative initiative under the Scientific Committee on Antarctic Research (SCAR) and the Council of Managers of Antarctic National Programs (COMNAP), aims to address critical data gaps in the ice-sheet margin, including subglacial topography, ice discharge, surface mass balance, subglacial hydrology, geological settings, and ice-shelf mass balance. Using the Bedmap3 FAIR database, we generated stochastic ensemble bed topographies within 100km of the grounding line. Our analysis reveals a much rougher bed than previously suggested by gridded map products, introducing significant uncertainty into assessments of future grounding line retreat, including marine ice sheet instability scenarios. We further estimate the uncertainty in ice discharge caused by ice thickness uncertainty, demonstrating how targeted RINGS survey at the grounding line can significantly reduced these uncertainties. Additionally, we assess how bed elevation uncertainty relates to bed roughness and the proximity of survey profiles. This provides an empirical framework for determining optimal survey spacing to meet specific uncertainty tolerances. International collaboration is essential for conducting dense surveys along the ice sheet margin and inland. Ultimately, these efforts will enable the development high-resolution bed topography products with documented uncertainties for diverse scientific applications.

Dr Kenichi Matsuoka

Norwegian Polar Institute, Norway

Dr Kenichi Matsuoka

Norwegian Polar Institute, Norway

Dr Kenichi Matsuoka earned his PhD from Hokkaido University, Japan, in 2002. After completing postdoctoral training in Japan and the USA, he was appointed Research Assistant Professor at the University of Washington in 2005. In 2010, he joined the Norwegian Polar Institute, where he currently serves as a Senior Research Scientist specializing in Antarctic glaciology. Dr Matsuoka's research focuses on subglacial environments, ice dynamics, and ice-sheet evolution, with an emphasis on using ice-penetrating radar. Through extensive international collaborations, he has advanced our understanding of Antarctic ice processes. He is the chair of the SCAR Action Group Antarctic RINGS, which aims to address critical data gaps in the Antarctic ice-sheet margin to enhance sea-level rise projections.

An overarching objective of the glaciological community is to develop numerical models of ice-sheet behaviour that are capable of robustly projecting the future evolution of the Antarctic and Greeland ice sheets in a warming world. A robust understanding of bed topography, and how it has evolved through time, is central to this challenge because bed topography: (i) influences how the overlying ice sheet behaves and responds to climate change, and (ii) provides a record of long-term glacial history, for example under past warmer climates that serve as analogues for projected future change.

The aim of this presentation is twofold. First, we will demonstrate how changes in bed topography over geological timescales (via processes such as erosion, sedimentation, tectonic activity, and isostasy) have influenced the behaviour of the Antarctic ice sheet. For example, glacial erosion and crustal subsidence have caused the development of increasingly low-lying and reverse sloping beds over time, particularly within near-coastal subglacial basins, which has amplified the sensitivity of the ice sheet to atmosphere and ocean forcing. Second, we will explore how bed topography can encode valuable records of long-term ice-sheet behaviour. For example, subglacial landscapes in Greenland and Antarctica retain the signatures of surface processes that operated prior to glaciation and/or during times when the ice configuration was substantially different to today. These findings provide important opportunities as constraints for models of past and present-ice sheet behaviour, as well as potentially informing future geophysical survey and ice/bedrock drill planning.

Dr Guy Paxman

Durham University, UK

Dr Guy Paxman

Durham University, UK

Dr Guy Paxman is a polar geophysicist and geomorphologist with a particular interest in the long-term evolution of Earth's ice sheets and their sensitivity to climate change. His research focuses on the interactions between solid Earth processes, topography, and ice-sheet dynamics, particularly in Antarctica and Greenland. This primarily involves: (i) reconstructing the long-term landscape evolution of Antarctica and Greenland and understanding how this has influenced ice-sheet behaviour over millions of years, and (ii) using geophysical and geomorphological measurements of subglacial landscapes to constrain sub-ice geology and past ice-sheet behaviour. Guy's research draws on a range of methods including geophysical modelling, geomorphological analysis, machine learning and automated landscape analysis, and ice-sheet modelling. He is also a member of the 'Antarctic Geological Boundary Conditions' sub-committee within SCAR's INSTANT programme and acts as a Theme Editor for Geophysics in Quantarctica version 4.

Simulations of the Antarctic ice sheet evolution over the coming century continue to dominate uncertainties in sea level projections. Despite a decade of rapid progress in numerical modelling of ice sheet flow - including the addition of new processes and the validation with a rapidly growing number of remote-sensing observations - the ability of continental scale simulations to capture recent changes in Antarctic dynamics remains limited. While uncertainties arise from a number of factors, including the choice of climate forcing, greenhouse gas emission scenarios and processes captured, the choice of ice sheet model itself is the dominant source of uncertainty for both decadal and century time scales. 

In this presentation, we will assess the viability of using continental scale simulations and explore whether regional or basin-scale models may provide more robust insights. We will first review recent results on Antarctic ice sheet projections and the main sources of uncertainty in these projections. We will then compare results performed at continental scale with regional scale models of the Amundsen Sea Sector and Wilkes Land, as well as models simulating individual glaciers for these two regions. We will compare uncertainty coming from these three spatial scales with uncertainty stemming from the choice of climate model, scenario, as well as uncertainties in bedrock elevation. Our results will highlight differences in the model’s simulated ice mass loss and grounding line retreat, and provide a new perspective on how to refine Antarctic ice sheet projections and improve confidence in future sea level rise estimates.

Associate Professor Hélène Seroussi

Dartmouth College, USA

Associate Professor Hélène Seroussi

Dartmouth College, USA

Dr Hélène Seroussi is an Associate Professor of Engineering at Dartmouth College. Her research interests are focused on better understanding and explaining ongoing changes in the cryosphere, reducing uncertainties in sea level rise using numerical modelling, and understanding interactions of ice and climate by combining process studies and state-of-the-art numerical modelling with remote sensing and in situ data. She is one of the co-founders and core developers of the Ice-sheet and Sea-level System Model (ISSM), is a member of the Steering Committee of the Ice Sheet Model Intercomparison project for CMIP (ISMIP6 and ISMIP7), and serves on the editorial board of the Journal of Geophysical Research - Earth Surface. Professor Seroussi received her PhD and Masters’ degrees from École Centrale Paris in France.

The Antarctic ice sheet's grounding zone, the critical transition between grounded ice and floating ice shelves, exerts a profound influence on ice flow and therefore global sea-level rise. Its dynamics is highly sensitive to ocean temperature and bed topography and accurate sea-level projections hinge on understanding these processes. Rapid changes in grounding line position, driven by ocean warming or ice shelf buttressing, can trigger significant ice mass loss through accelerated ice flow.

However, the complex and often poorly constrained bed topography beneath the ice sheet grounding zone introduces substantial uncertainty into ice sheet models. Bed roughness, subglacial channels, saltwater intrusion, and bedrock slope directly impact ice flow velocity and basal melt rates. Current topographic datasets, relying on sparse radar and seismic measurements, often lack the resolution needed to capture crucial bedrock features in grounding zone. This uncertainty propagates into sea-level rise projections, hindering our ability to accurately predict future coastal inundation. In this presentation, we will review knowledge of the current understanding of grounding line and grounding zone dynamics and how improved measurement of bed topography in these regions is essential to refine ice sheet models and reduce the range of uncertainty in sea-level predictions.

Professor Sophie Nowicki

University at Buffalo, USA

Professor Sophie Nowicki

University at Buffalo, USA

Dr Sophie Nowicki is a Professor in the Department of Geology and RENEW Institute at the University at Buffalo. Through applied mathematics, remote sensing observations and numerical modelling, her work spans the spectrum of local processes, such as understanding the physics of ice sheet grounding lines, or the impact of bedrock topography on ice dynamics, to that of large-scale continental ice sheet models and their use in projections of sea level change. As sea level projections from ice sheet models require knowledge of atmospheric and oceanic conditions that drive ice sheet evolution, she is also interested in how to improve climate models in the polar regions, as well as the use of multiple models for projections. Prior to joining the University at Buffalo, Sophie was a civil servant in the Cryospheric Sciences Laboratory at NASA Goddard Space Flight Center, Greenbelt, MD. Sophie was a science team member for NASA's Operation IceBridge and is co-leading the NASA Sea Level Change Team. She is an executive committee member for the Ice Sheet Mass Balance Intercomparison Exercise, phase 2 and 3 (IMBIE2-3), and co-leads the Ice Sheet Model Intercomparison Project for CMIP6 (ISMIP6) and its follow on activity ISMIP7.

Ice sheet models require explicit knowledge of the underlying bed. However, much remains unknown regarding the subglacial environment due to the difficulty associated with measuring it. Extensive radar surveys have been conducted across Antarctica, but the requirement of full coverage bed topography for models necessitates the usage of interpolation methods over gaps between existing observations, which often span several kilometres or more. Uncertainties in bed topography contribute to the large range in projections of future sea level rise. Advances in modelling capabilities now allow for the application of dynamic coupling between subglacial hydrology and ice dynamics in models of the Antarctic ice sheet. While a bed resolution of around 500m is recommended for modelling Antarctic ice sheet, it has been suggested that finer spatial resolutions are needed to accurately resolve subglacial water flow. We use a coupled model configuration within ISSM to generate projections of ice sheet evolution, including the subglacial hydrologic system, for the Amundsen Sea drainage basin in Antarctica initiated with several different bed topographies. We examine the impacts of the spatial resolution of the bed on modelled future sea level rise. Bed topographics of varying spatial resolution are obtained by manipulating BedMachine Antarctica data. We compare model outputs obtained with the original dataset, which has a 500m spatial resolution, to simulations initiated with a high and low-res variation of the original dataset. We reduce BedMachine data to a 5km spatial resolution to generate the low-res variation, and to simulate potential impacts from including high-res bed topography we add randomly generated noise to the original dataset. We use climate and ocean forcing from the CESM SSP 5-8.5 contribution for ISMIP6 experiments to project ice mass change and analyse impacts in the evolution of the subglacial environment over the 21st century when bed topographies with varying spatial resolution are employed. We discuss implications for sea level rise projections and make recommendations where additional bed observations will be the most impactful with respect to reducing uncertainties associated with modelling the subglacial environment in particular.

Dr Shivani Ehrenfeucht

Potsdam Institute for Climate Impact Research, Germany

Dr Shivani Ehrenfeucht

Potsdam Institute for Climate Impact Research, Germany

Shivani Ehrenfeucht completed her PhD at UCI, California, in September 2023. She is currently a postdoc at the Potsdam Institute for Climate Impact Research (PIK), in Germany, where she uses numerical models to study ice sheet tipping behaviour on long timescales (+1000 years). Previously, her research was focused on interactions between ice dynamics and subglacial hydrology, with a focus on the impacts of evolving subglacial hydrology on future sea level rise. Specifically, she has worked with various coupled model configurations to better understand how changes in the subglacial environment can impact ice velocity and mass loss over multiple timescales (i.e. seasonal to centennial) in Greenland and Antarctica.