The role of glaciers and ice sheets in sea level rise

Ice loss from glaciers and polar ice sheets is a major driver of global sea-level rise, which has doubled in pace since 2006. The melting of the West Antarctic Ice Sheet presents the greatest long-term risk, while regional impacts will depend on local land movements and coastal conditions.

Aim

The loss or retreat of ice as a result of global warming is a major cause of the rise in global mean sea level (GMSL). The aim of this piece is to set the scene, to discuss what’s happening to sea level, polar ice sheets and glaciers and to flag implications for wider society.

Sea level change

According to the Intergovernmental Panel on Climate Change (IPCC), global mean sea level has risen approximately 20 cm since 1900 and approximately 10 cm since 1971.  Since 2006, the rise has increased to 4 mm per year, double the average of the 20th Century. Approximately half the rise is due to thermal expansion as the ocean warms in response to a warming atmosphere. Melting from the Greenland and Antarctic Ice Sheets is responsible for about quarter of the total, a proportion that has been rising in recent decades. The melting of mountain glaciers and ice caps accounts for the remaining quarter. Whether or not greenhouse gas emissions are reduced, global mean sea level is predicted to rise by a further 10-25 cm by 2050, a range reflecting the slow response of the ocean to atmospheric change and uncertainty about our ability to curb carbon emissions.

At present the Greenland Ice Sheet is losing mass by surface melting and the speed up and retreat of marine terminating outlet glaciers. The West Antarctic Ice Sheet is losing mass from increased melting of ice by the warming ocean. Melting of ice around the marine margins of the East Antarctic Ice Sheet is partially balanced by an increase in surface snowfall due to the warmer atmosphere. There is real concern that a significant part of the West Antarctic Ice Sheet covering an ocean basin is close to a tipping point and could discharge into the ocean and thereby raise global mean sea level by an additional 1- 3 m.

Sea level is not flat and the term Global Mean Sea Level describes an average. There is considerable decimetre-scale variation in sea level due to factors such as ocean currents, winds, salinity and submarine topography.

The link between glacier melting and sea level is influenced by both the location of the melting glacier and the location of a particular coast. For example, the melting of an ice sheet not only contributes an extra volume of water due to the loss of ice mass but, as argued by the Scottish scientist James Croll in 1867, the loss of gravitational attraction of the ice mass also lowers sea level locally around the diminishing ice sheet. This effect can involve several metres of local sea level which is transferred to the far reaches of the ocean. The practical implication is that the perhaps expected increase in sea level in the UK as a result of the reduced volume of the Greenland ice sheet will instead be largely balanced by the loss of the gravitational attraction of the diminishing ice mass. Rather, it is important to keep our eyes on the fortunes of the West Antarctic Ice Sheet. The possible increase in sea level as a result of its demise will be aggravated in much of the northern hemisphere by the loss of the gravitational attraction of the ice sheet.

The impact of sea level rise on a specific coast depends on relative sea level, which is the balance between global sea level rise at that point and any vertical movement of the land. The vertical movement of the land varies with tectonics and other factors. For example, the land in formerly glaciated regions of the northern hemisphere is still rising, while river delta lowlands worldwide tend to be sinking. In Scotland the main factor is the continuing but slowing isostatic response to the ice sheet that covered the land and surrounding seas at the Last Glacial Maximum some 22,000 years ago. The ice sheet was more than 1 km thick at its centre in the western highlands and it depressed the earth’s crust beneath its weight. Isostatic recovery since deglaciation continues at a rate of 1 mm per year where the ice was thickest. In peripheral areas such as the Orkneys, the Outer Hebrides and the English border isostatic uplift has ceased. There is long-term tectonic subsidence in the central North Sea that has had little effect on Scotland but contributes to depression of the land in eastern England. Using a middling warming scenario and allowing for vertical land changes, relative sea level in the year 2100 is predicted to rise by ~0.45 m in central western Scotland and ~0.55 m in peripheral areas. For London the prediction is 0.67 m (NASA/IPCC Sea Level Projection Tool).

The effect of sea-level rise on wider society depends on geographical location and use of the coast by society. On coasts consisting of soft, glacial and sandy deposits wave erosion during storms that coincide with a high tide will cause coastal retreat and loss of land. Even on hard rock coasts and also on constructions such as harbour walls and dykes, deeper water will encourage larger waves and lead to more erosion during major storms. In estuaries the landward encroachment of the highest tides and the interaction with river estuaries will bring problems of flooding and changes in drainage and salinity affecting farming. One has only to appreciate that 2 billion people live near the world’s coasts and 900 million live on low-altitude coastal plains to realise that continued global sea-level rise brings huge social, economic and environmental problems both locally and globally.

Antarctic Ice Sheet

The Antarctic Ice Sheet is subdivided into the East Antarctic Ice Sheet and West Antarctic Ice Sheet, geographically misleading names that go back to the Cold War of the 1950s. The East Antarctic Ice Sheet, holding enough ice to raise global sea level by 58 m (!), is based on a continental landmass akin to the combined size of South Africa and Australia. Ice accumulates from snowfall and flows radially from a high centre at ~4000m to the coast. In topographic embayments and some deep submarine valleys near the coast it merges into floating ice shelves, where ice may be lost through iceberg calving or submarine melting by warm sea water. The East Antarctic ice sheet is sufficiently large to control its terrestrial climate and has persisted intact for some 15 million years. During warmer intervals, such as the Pliocene 3-4 million years ago, when global climate was similar to that of the present, the ice-sheet margin and peripheral mountain glaciers advanced as a result of increased snowfall in the warmer climate. This pattern of increased snowfall in a warming climate is seen in remote-sensing observations today. On this basis, many scientists consider that, with the possible exception of some ice flowing into overdeepened troughs, increased snow accumulation on the ice-sheet surface in East Antarctica is likely to balance increased melting of its peripheral ice shelves.

The West Antarctic Ice Sheet is a different matter. The ice sheet is grounded on an archipelago of mountain massifs with its centre over an ocean basin up to 2,500 m deep. In 1978 John Mercer suggested that the ice sheet was held intact by a peripheral ring of land and ice shelves. If this protective ring is breached, then the whole ice sheet could collapse, raising global sea level by ~5 m. Subsequent work has shown that ice in the mountain massifs of the archipelago has persisted for millions of years but that the ice in the marine basin is indeed fragile. Studies of the molecular genetics of octopuses in the Amundsen Sea south of the Pacific Ocean and the Weddell Sea south of the Atlantic Ocean demonstrate the existence of a close and direct ocean link across Antarctica. The biological similarity suggests that an interior Antarctic seawater channel could well have existed in the Last Interglacial ~120 thousand years ago, when the world’s climate was similar to that of today and global sea level was close to, or a few metres, higher than today.

Satellite imaging of the West Antarctic Ice Sheet in recent years shows significant surface lowering and speed up of ice flow in the hinterland of outlet glaciers flowing into the Amundsen Sea. First, Pine Island Glacier and ice in its catchment sped up, thinned and retreated. Now the largest outlet glacier, Thwaites Glacier, popularly known as the Doomsday Glacier, is the focus of intense international research to establish the rate and processes of submarine melting and to discover how close the ice sheet is to a threshold of collapse or tipping point. Thwaites Glacier is also the focus of highly controversial ideas of massive environmental interference, such as constructing a curtain above the seabed ~80 km long designed to slow down the flow of warm seawater towards the glacier and to restrict the return flow of cold meltwater into the Southern Ocean. Even if possible, the risk is one of unintended environmental consequences, for example potentially changing southern ocean circulation, climate and biology.

There are many scientists who see the thinning of the Thwaites and neighbouring Pine Island glaciers as an indication that the threshold of collapse has already been reached. However, the complex interplay between ice sheets and the major environmental systems of climate, oceans and lithosphere may suggest otherwise. The Ross and Weddell Sea ice shelves, each floating and the size of an average European country, may have been thinner and smaller than present 6-9,000 years ago in the warmth of the Holocene and recovered subsequently. The argument is based on the nature of sea floor sediments, patterns of sedimentary ice layers on ice islands, the dating of the sudden thinning of the ice sheet in adjacent mountains, and rates of isostatic recovery. One possibility is that the early warmth of the Holocene some 9,000 years ago saw the retreat of the grounded ice sheet in the Ross and Weddell embayments and its replacement by floating ice shelves.  Relieved of its ice load, isostatic recovery of the sea floor then caused the ice shelves to ground on high points of the submarine topography, creating pinning points that evolved into ice islands. The latter added friction to iceshelf flow leading to a new phase of overall ice thickening. If the above argument can be substantiated, then the West Antarctic Ice Sheet may be able to survive thinning in warm periods at least for a century or so.

Greenland Ice Sheet

For millennia and up to the 1970s-1990s the Greenland Ice Sheet remained in approximate balance between snowfall accumulating on its high centre and melting on the surface of its low elevation margin and marine-terminating outlet glaciers. In the years 1992-2018, the ice mass began to decline and the annual loss of ice rose at a rate proportional to high-end scenarios of the IPCC, amounting to a loss of ~140 billion tonnes of ice per year and an overall sea-level rise of ~11mm. Approximately half the loss in this period was due to surface melting and half to the speed up and retreat of marine-terminating outlet glaciers.

During the Last Glacial Maximum ~22,000 years ago the Greenland Ice Sheet expanded across the offshore marine shelf. Subsequently, in common with the Antarctic-sized North American Ice Sheet and the North European ice sheet, it thinned and the margins retreated, especially across the flatter topography of the west. Then in the early Holocene, ~9,000 years ago, and in contrast to the other large northern hemisphere ice sheets, it stabilised for millennia close to its present extent. Perhaps it stabilised at the point where it became largely land-based; marine melting was curtailed and yet the elevation of the centre of the ice sheet, reinforced by uplift due to isostatic recovery, was high enough to accumulate sufficient snow for stability.

Other glaciers and ice caps

It is estimated that there are ~200,000 glaciers in the world covering an area of ~706,000 km2. Many glaciers are in the high mountains of Asia, Europe, New Zealand, North and South America. Regional-scale ice caps can be found in the islands and mountains of the polar regions. If they all melted, the loss of ice mass to the ocean would raise global sea level by 32 cm. Typically, such glaciers display a seasonal rhythm with snow accumulation dominant in winter and melting in summer. Most accumulation is in the higher reaches of these glaciers and ice caps while most melting is at lower levels. The significance of seasonal melting makes them particularly sensitive to atmospheric warming. The accelerating warmth in recent decades explains the recent rise in melting and rapid retreat of such glaciers. Over the years 2006-2016 the loss of ice contributed 0.92 mm per year to global sea level rise.

The pattern of melting on a glacier or ice cap varies as it responds to warming. First, run-off due to atmospheric warming rises during ‘summer’ as the lower reaches of a glacier melt intensely and melting extends to higher than usual elevations. The more the glacier melts and its surface lowers, the faster it melts. As melt rates accelerate, a glacier shrinks in size and there is less ice to melt. There comes a time when the shrinking ice-surface area causes meltwater runoff to decline and this turning point is known as peak water. The date that peak water is reached varies within and between different regions and climates. The relatively high contribution to sea-level rise from glaciers and ice caps in recent years reflects the rise towards peak water.

What will happen?

Predictions of future glacier contributions to sea level rise are based on models of glacier behaviour, constrained by past records, and tuned to respond to different climate scenarios. Such estimates are based on different scenarios as outlined by the International Panel on Climate Change (IPCC). The main uncertainties in the future will be how successful the world is in reducing greenhouse gas emissions and associated global warming and whether we can reverse the warming trend.

In addition to the above, there are uncertainties due to the way environmental systems work and interact. Global mean sea level marks a balance between the global systems of the cryosphere, lithosphere, atmosphere, ocean and biosphere. Each system has interactions within and external to the system on a range of time scales. Thus, taking an example, suppose a warmer ocean in Antarctica melts more ice and this changes the salinity of the seawater; this in turn changes the circulation of ocean water and this changes the distribution of sea ice. Such complex changes then modify the winds, temperatures and snowfall. The loss of grounded glacier ice instigates isostatic uplift and this changes the submarine topography and the flow of ocean currents. Such changes then have the potential to amplify the initial change or subdue it. Such feedback links occur around any coast and can be initiated for example by a rise in sea level or changes in sediment volumes delivered to the coast by rivers because of changes to the biosphere caused by human agriculture or deforestation. The complexity of the links and feedbacks explains why it is so difficult to reach firm predictions and why there can be surprises.

Bearing in mind such reservations, what can we say about future global sea level?

• The West Antarctic Ice Sheet is a clear priority in order to understand how close we are to the loss of the marine portions of the ice sheet with the potential to raise global sea level by up to 3 m. Also relevant are the deeper outlet glaciers in East Antarctica subject to ocean warming, each of which could raise sea level by tens of mm if they accelerate and retreat.
• The history of the Greenland Ice Sheet highlights the importance of atmospheric controls on both snow accumulation and surface melting and also the sensitivity of marine-terminating glaciers to ocean melting. With continued ice-sheet mass loss, the role of marine melting will decline as more and more marine-terminating outlet glaciers retreat onto the land. This means that the future of the Greenland Ice Sheet will depend on the balance between surface accumulation and surface melting and this in turn depends directly on the rate of atmospheric warming. If the warming continues, then the Greenland Ice Sheet will continue to decline at a rate proportional to the amount of warming. More optimistically, if we manage to reverse the trend of warming, then the Greenland Ice Sheet is in a position to grow again.
• The contribution of regional glaciers and ice caps to sea-level rise will depend partly on the trajectory of warming but also the date of global peak water, after which their contribution to rising sea levels will decline. However, it is important to stress the crucial importance of meltwater from such glaciers to agricultural communities that depend on glacial meltwater in summer or the dry season, for example in arid regions of Asia and South America; even if atmospheric warming stabilises, they will face a declining resource in the short to medium term. On an optimistic note, if we reverse global warming, mountain glaciers could grow again in some of the world’s higher mountain ranges.
• Global sea level will continue to rise as the ocean and atmosphere continue to warm. The problem is existential for tropical coral island communities and those heavily populated low-lying river basins of the Indian continent and southeast Asia. The large proportion of the world’s population living in coastal cities will also incur high social and economic costs as they are forced to adapt to the change.
• Given the scale of the social, economic and environmental challenges throughout the world caused by sea-level rise and the enormous costs incurred, the overriding priority must be to direct all our resources to reducing carbon emissions and kickstarting an era of global cooling.

• This article is part of a series from the RSE’s Environment Community of Interest. The aim is to stimulate informed public debate on the many aspects and elements of this complex and vexed subject. We’d love to hear your thoughts about this piece, email us on publicpolicy@theRSE.org.uk.
• To be notified when the next article in this series is live, sign up to the RSE’s monthly policy advice newsletter.
• The RSE’s blog series offers personal views on a variety of issues. These views are not those of the RSE and are intended to offer different perspectives on a range of current issues.
• David Sugden, FRSE, Institute of Geography, School of GeoSciences, University of Edinburgh, EH8 9XP. With thanks to Peter Nienow FRSE and Dr Donald Slater for their helpful comments