Is ice loss from West Antarctica driven by ocean forcing or ice and ocean feedbacks?

Lead Research Organisation: University of Edinburgh
Department Name: Sch of Geosciences


The Antarctic Ice Sheet is a mass of ice larger than Europe, in some places several kilometres from top to bottom. Fed by snowfall over its interior, it spreads out under its own weight, going afloat at its edge in the form of enormous ice shelves with areas ranging from that of Greater London to that of France. The ice shelves are then melted from below by waters from the Southern Ocean. The inputs and outputs of the system are so massive that even very small imbalances can have catastrophic effects on global sea level: the portion of Antarctica known as the West Antarctic Ice Sheet (WAIS), suspected unstable due to the shape of its underlying bedrock, would contribute 3-5 m of sea level rise were it to collapse completely.

Satellite observations have shown that some of the fast-flowing outlet glaciers that carry ice out of Antarctica have sped up dramatically. Pine Island Glacier, which drains a significant portion of WAIS, has nearly doubled its speed in the last several decades, creating a large negative imbalance for the ice sheet. The acceleration is thought to be connected to the high under-ice shelf melt rates observed in the region. This melting reduces the ability of the Pine Island Ice Shelf to hold back the glacier feeding it.

Increased ice-shelf melt rates are possibly due to warming oceans; but recent studies suggest that melting could actually be strongly dependent on ice shelf and ice sheet behaviour as well. Additionally, a recent glaciological modelling study suggests a "tipping point" may have been crossed, and that ice retreat, though triggered by oceans, is now self-perpetuating regardless of melting. Determining whether the observed retreat is due to ongoing climate forcing, or to feedbacks of the coupled ice-ocean system, is of utmost importance to predicting (and if possible mitigating) future sea level contributions from WAIS.

In the proposed work we will address this question through the development of a sophisticated computer model of interacting ice sheet and oceans, and by investigation of the processes involved in ice retreat through controlled modelling experiments. Idealized experiments of ice-ocean interactions will lead up to a realistic modelling study of Pine Island Glacier, designed to assess the relative importance of forcing and feedback in its observed retreat.

This study will be unprecedented in terms of the tools developed, the experiments undertaken, and the knowledge gained. Presently no numerical model exists that can fully represent the close interaction between ice sheets, ice shelves, and the ocean circulating beneath them. Furthermore the ice and ocean codes, as well as being ideally suited for coupling together, share properties that will allow for in-depth investigation of model sensitivity and controls, and for the incorporation of ice-sheet observations in a physically consistent manner, vastly improving the reliability of results.

Planned Impact

We will gain new knowledge regarding the physical processes involved with West Antarctic ice loss, a significant source of uncertainty regarding sea level contribution in coming centuries. We will also determine the extent to which climate variability is responsible for observed ice loss.

Our findings will therefore be of great import to groups responsible for communicating science to policymakers (Met Office, IPCC) and for directing environmental science research (NERC). Additionally the importance of our research outputs needs to be communicated to the general public, on both local and international levels.

Government and Science Policy

Our results will be directly relevant to the science and summary reports produced by the International Panel on Climate Change by lessening uncertainties associated with future ice loss and with the impacts of human emissions on sea levels. This will benefit world governments and other organizations in shaping carbon policy and in anticipating detrimental effects of rising seas.

Through collaborative partnership with other UK efforts, our work will aid the improvement of the climate model run by the Met Office Hadley Centre by improving representation of ice sheets. This will have impact through the information and advice provided by the Hadley Center to the Department of Energy and Climate Change (DECC) and the UK Environment Agency.

NERC plays a leading role in funding, and determining future directions, of environmental and climate research in the UK. Therefore it is vital that our results be visible to the relevant decision makers within NERC, such as the Climate System and Earth System Science theme leaders and the Living with Environmental Change programme. One avenue is through the Joint Weather and Climate Research Programme (JWCRP) between NERC and the Met Office. Co-I Adrian Jenkins' membership in the JWCRP will provide a route of communication with key NERC stakeholders. Additionally we will draft a white paper detailing our results and their significance for policymakers and their relevance in directing future Antarctic research.

General Public

We will work with Corporate Communications at the University of Edinburgh and with BAS' award-winning press office to communicate to the general public any newsworthy findings resulting from our work. In addition, we will take advantage of the leading outreach efforts at the Massachusetts Institute of Technology (MIT), with expert science journalists dedicated to communicating scientific outputs involving MITgcm to a general (international) audience. (Funding for US-based outreach efforts will be provided by MIT.)

On a local level, our results will be shown to the public as the Southern Hemisphere component of an exhibition on ice-ocean interaction to be hosted by Dynamic Earth, Edinburgh. The exhibit will educate the public on our work, explaining the societal importance of rapidly changing ice flow in the West Antarctic Ice Sheet, and links to the oceans and the climate system.


10 25 50
Description The fastest-thinning glaciers in West Antarctica are in the Amundsen embayment, and at the same time their floating extensions (ice shelves) are being melted at very high rates, and it has been proposed this is due to warm circumpolar deep water. The melting does lead to further ice loss, as it reduces the amount of buttressing provided to the glaciers. Furthermore the heat provided by the ocean is thought to depend on large-scale climate variability. There seems to be two conflicting lines of thought -- that (1) the glaciers are simply at the "mercy" of the ocean, and knowing what happens to ocean circulation is sufficient for predicting ice loss, and (2) the eventual effect of ocean forcing depends on a complex interaction between ocean, ice shelf, and glacier, and the eventual result is not knowable without understanding these interactions. 2 recent publications supported by this grant, "Ocean-forced ice-shelf thinning in a synchronously coupled ice-ocean model" (Jordan and others, 2017) and "The response of ice sheets to climate variability" (Snow and others, 2017) make use of a model coupling framework developed under this grant to address these questions. The first paper shows that the pattern of melting, which is introduced by the flow of the ice shelf itself, can be as important as the overall heat forcing from the ocean in terms of the effect on grounded ice. The second shows that the effect of ocean warming depends on its patterns of variability, whereby short-scale (decadal or less) variability is expected to have little effect on ice loss, whereas long-term change might have very strong effects.

More recently, coupled ice and ocean modelling has been applied to a realistic setting: Dotson and Crosson Ice shelves, a key thinning region of the Amundsen. Using adjoint tools related to ice- and ocean-modelling coupled with new remote sensing observations of melt rates, we have been able to show the impact that bathymetric uncertainties have on ice-ocean interactions, as well as gauging the accuracy of coupled ice-ocean models in reproducing observed melt rates. Our adjoint-modelling portion of the study identifies key regions of ice shelves where this is important. Our results have been published in "How Accurately Do We Need to Measure Ice-shelf Melt Rates?" (Goldberg et al, 2018). In addition, in this paper we partially carry out initialisation of a dynamic ice-ocean system. further work with our coupled model will develop this even further.
Exploitation Route The coupled ice-ocean model developed in the publications above is open-source code, and is available to other modelling groups to enable improved coupling between ice and ocean models, to enable more comprehensive study of ice sheet response to ocean warming. These will include both academic efforts and government-funded climate-modelling institutions.
Sectors Environment