Physical environment

This report focuses mainly on the environment of areas administered by Australia (the Australian Antarctic Territory – AAT – and the Territory of Heard Island and McDonald Islands), subantarctic Macquarie Island (which is part of Tasmania) and the Southern Ocean adjacent to these areas (Figure 1). Macquarie Island is managed by the Tasmanian Government, while the wider Antarctic regions are managed cooperatively through the international agreements of the Antarctic Treaty System, in which Australia is a leading participant. Many aspects of Australia’s interests in the region relate to the environment of broader geographical areas, which are described where required.

Antarctica is Earth’s southernmost, coldest, highest, windiest and driest continent. Including all islands and ice shelves, it covers an area of about 14.2 million square kilometres (km²) – nearly twice that of Australia.

Although isolated from other continents, Antarctica and the surrounding Southern Ocean are major drivers of global weather and climate (Owens & Zawar-Reza 2015). Interactions between the atmosphere, ice and ocean in the Antarctic region set up patterns of weather and climate that extend across the Southern Hemisphere and northward across the equator. These patterns interact with the land and ocean at lower latitudes, which in turn influences the southward flow of heat and moisture to affect the Antarctic continental margin and interior (Marshall 2007, Noble et al. 2020).

The weather and climate of Australia feel the influence of the Antarctic region because Australia is geographically close to the region, and is affected by mutual interactions with large-scale modes of climate variability (van Ommen & Morgan 2010, Vance et al. 2015, Lim et al. 2019, Abram et al. 2021, Udy et al. 2021). Consequently, understanding the state of the physical environment in the Antarctic region is important for assessing the state of the Australian environment.

Atmosphere

Atmospheric change in the Antarctic region can be divided into changes at and near the surface (primarily in temperature, wind and precipitation), and changes in the upper levels (primarily in ozone concentration, temperature and wind).

Surface changes

Patterns of change in the Antarctic surface climate have been mixed in recent decades (Hartmann et al. 2013, Turner et al. 2014, Meredith et al. in press).

Significant regional surface warming occurred in parts of the Antarctic Peninsula and the West Antarctic Ice Sheet from the late 1950s to the end of the 20th century (Turner et al. 2005b, Steig et al. 2009, Hartmann et al. 2013, Turner et al. 2014). The rate of this warming was among the most rapid anywhere on the globe (Steig et al. 2009, Bromwich et al. 2013).

Since the beginning of the 21st century, the Antarctic Peninsula has generally cooled (Turner et al. 2016), although an indication of warming from the mid-2010s has been reported (Carrasco et al. 2021). Warming in spring across the Antarctic Peninsula and West Antarctica over 1979–2012 was also observed (Clem & Fogt 2015). In East Antarctica, temperatures since the 1950s have generally shown no significant trends, apart from indications of weak cooling during autumn in the interior (Nicolas & Bromwich 2014, Jones et al. 2016). These changes have recently been confirmed by Turner et al. (2020a), who also found that the interannual variability on the western Antarctic Peninsula has decreased since the late 1950s as sea ice in the region has declined.

Turner et al. (2020a) have provided the most recent detailed analysis of temperature trends for Australia’s 3 continental Antarctic stations. Over 1979–2018, no significant trend was found for Mawson and Davis, either annually or seasonally. For these stations, the 95% confidence limit for the annual trend is approximately 0.2 °C per decade. Over the same period, a statistically significant cooling at the 95% confidence limit was found for Casey, for the overall annual mean temperature (trend −0.26 ± 0.24 °C per decade) and for the winter mean (trend −0.58 ± 0.51 °C per decade). Similar trends over the modern observational record are indicated by BOM (2019). For Macquarie Island, similar analysis to that by Turner et al. (2020a) using data from BOM (2019) indicates that any trend in the annual mean temperature over 1979–2018 was not statistically significant (within the 95% confidence limit of 0.1 °C per decade). However, during this period, and indeed since around 1970, Macquarie Island has shown a marked increase in annual rainfall, particularly in winter.

The overall cause of climate shifts in the Antarctic region is likely to be a combination of variability due to the natural internal modes of climate variability, particularly the Southern Annular Mode (SAM) and the El Niño–Southern Oscillation (ENSO) and their interactions, and the influence of global climate change (IPCC 2013).

As noted in the 2016 state of the environment report, of particular significance for the climate of the Antarctic region has been the tendency of the SAM to shift towards a more positive state during summer in recent decades (Marshall 2003). A strengthening of the westerly wind belt over the Southern Ocean and a shift of its core towards Antarctica have accompanied this change, and caused warming of the Antarctic Peninsula and cooling in the interior of East Antarctica (Thompson & Solomon 2002, Gillett et al. 2006, Marshall 2007, McLandress et al. 2011, Polvani et al. 2011, Thompson et al. 2011). Previous observational and model studies have pointed to the springtime Antarctic ozone hole (see case study: The Antarctic ozone hole), which has been brought about by human-produced emissions of ozone-depleting substances, as the main driver of the summer shift in the SAM (see Upper-level changes). In the case of the cooling trend at Casey, this is largely consistent with the effect of the annual mean change in the SAM (Turner et al. 2020a). The lack of significant temperature trends at Mawson Station, Davis Station and Macquarie Island suggests that other climate processes are countering the effect of changes in the SAM.

Because of the relatively short length of the observational record and the inherent level of interannual climate variability for the Antarctic region, the Intergovernmental Panel on Climate Change (IPCC (2013) determined that only low confidence could be ascribed in linking greenhouse warming to the observed temperature changes in Antarctica. It is still unclear how the large-scale climate modes, including ENSO, are responding to anthropogenic (human) forcing (Hartmann et al. 2013) and affecting Antarctica. However, it is clear that natural climate variability that is unforced by human influences has played a role, together with the changes in the SAM, in recent patterns of climate variability in the Antarctic region (Turner et al. 2016, Smith & Polvani 2017).

Studies published since the 2016 state of the environment report have provided new insights into the causes of changes in the Antarctic atmosphere. Evidence continues to emerge that, since the turn of the 21st century, the ozone hole has been repairing as a result of the action of the Montreal Protocol (WMO 2018). This has led to a pause in surface wind changes over the Southern Ocean (Banerjee et al. 2020) associated with the summertime shift in the SAM (Fogt & Marshall 2020). Various studies have highlighted the influence of the tropical oceans on Antarctica. Turney et al. (2015) showed that specific regions of the southern mid-latitudes regulate the exchange of heat across the Southern Ocean. According to Clem et al. (2018), increases in the La Niña phase of ENSO have been responsible for cooler temperatures in western East Antarctica in the past 2 decades. Variability in the Pacific Ocean is also implicated as having an increasing influence on Antarctic surface temperatures (Rahaman et al. 2019). In addition, recent extremes in Antarctic temperatures have highlighted the large-scale climate connections that occur across the Southern Ocean.

New climate simulations have provided insights into possible future surface changes, and have shown that the effect of anthropogenic climate change is likely to increase. Bracegirdle et al. (2020) examined state-of-the-art climate simulations for 4 greenhouse gas emissions scenarios. Overall, they found increased surface temperatures and precipitation across the Antarctic continent for all future projections, with mitigating influences under low-emissions scenarios by heat and carbon dioxide (CO₂) uptake by the Southern Ocean, and by ozone recovery. Cai et al. (2021) examined the contributions to ‘polar amplification’, which is the increased warming of the polar regions associated with climate change, and found that, for the Antarctic region, the main contribution to warming occurs in winter due to uptake of heat by the global oceans.

Upper-level changes

The 2016 state of the environment report summarised the observational evidence for the overall cooling of the Antarctic upper atmosphere over recent decades. Subsequent studies have further characterised temperature changes in this region.

Ozone in the upper parts of the atmosphere absorbs solar energy and causes heating in the lower atmosphere and at Earth’s surface. Consequently, trends in ozone have influenced temperature trends in the lower stratosphere (10–30 km altitude) during spring. During the growth of the ozone hole over 1979–97, the lower atmosphere cooled; as ozone concentrations have been recovering, it has warmed (Randel et al. 2017, Solomon et al. 2017).

In the upper stratosphere (30–50 km altitude) and mesosphere (50–95 km altitude), global temperatures have cooled over several decades (Randel et al. 2017). This is consistent with the increasing greenhouse gases in the lower atmosphere blocking more of Earth’s heat before it reaches the upper levels of the atmosphere (Goessling & Bathiany 2016).

Above Australia’s Davis Station, the temperature trend between 1995 and 2018 in the upper mesosphere (near 87 km altitude) was −1.2 ± 0.5 °C per decade after accounting for influences from the solar activity cycle (French et al. 2020a). Related work (French et al. 2020b) identified a previously unrecognised temperature modulation in the upper mesosphere with a period of approximately 4–5 years. The interaction of ocean and atmospheric modes of climate variability at lower altitudes appears to force this variability.

Cryosphere

The cryosphere comprises the parts of Earth that have frozen water in the form of snow and ice, including sea ice, glaciers, ice sheets and icebergs. About 90% of Earth’s ice occurs in Antarctica. With a volume of approximately 26.9 million km³ (Fretwell et al. 2013), the Antarctic ice sheet contains 70% of the world’s fresh water. If it melted, sea levels would rise by 58 m (Fretwell et al. 2013, Vaughan et al. 2013).

The annual growth and retreat of the Antarctic sea ice represents one of nature’s most significant large-scale annual changes. Sea ice forms when the temperature of the ocean surface decreases below approximately −1.8 °C. At its maximum annual extent in September–October, Antarctic sea ice covers a total area of up to 20 million km² (Fretwell et al. 2013), which retreats to a minimum area of about 3 million km² in February (Parkinson 2019). This annual cycle and the associated processes of sea ice formation and melt are of immense importance for weather, climate, ecosystems and human activities.

The Antarctic continental ice also provides a wealth of information about past climate. The ice forms from atmospheric moisture, which reaches the surface in the form of snow and small ice particles called ‘diamond dust’ (Grieger et al. 2016, Thomas et al. 2017). This material, along with trapped air, provides a detailed record of past climate (see case study: Antarctic ice-core records of past climate and atmospheric composition). The ice also preserves dust of continental and extraterrestrial origin, and other transported particulates such as sea salt.

As interpreted from the marine sediment record, the complete glaciation of Antarctica began about 34 million years ago. This was triggered by a decline in CO₂ levels to below 600 parts per million (DeConto & Pollard 2003, Galeotti et al. 2016). The ongoing glaciation was reinforced when Antarctica became fully surrounded by ocean about 30 million years ago (Noble et al. 2020).

Antarctic ice sheet and glaciers

Melting of the Antarctic ice sheet and the icebergs it discharges adds freshwater to the Southern Ocean (Hammond & Jones 2016). The amount of melting has increased since the 1990s (Rignot et al. 2019), escalating Antarctica’s contribution to global sea level rise. Changes in the salinity of the Southern Ocean because of freshwater input have affected its structure and circulation, the availability of iron for ocean ecosystems, and the timing and extent of sea ice production (Bintanja et al. 2013, Meredith et al. in press). The effect of these influences on a variety of Antarctic taxa, such as krill, seals and penguins, is not well understood.

Overall changes to the ice sheet and glaciers

The Antarctic ice sheet gains mass by snow and ice deposition, and loses mass by the discharge of melt from ice shelves, basal ice and surface run-off; the sublimation of surface ice; and the formation (discharge) of icebergs at the coast.

There are 3 main methods for measuring total ice mass changes for Antarctica:

• The mass budget method (Favier et al. 2017) calculates total gain and losses over time as the difference between the estimates of snowfall (input) and glacier outflow across the periphery of the grounded ice sheet (output, from measured velocity and thickness).
• A second method monitors surface elevation changes, primarily using pulsed range-finding radars and lasers from satellites, to determine losses (lowering surfaces) or gains (rising surfaces) and infer mass changes.
• A third method uses gravity satellites to measure deviations in the gravitational pull as they pass over the ice sheet to ‘weigh’ changes in ice-sheet mass.

Each method has advantages and disadvantages, and relies on different data sources. Consequently, the magnitudes of estimated change vary; however, most studies now broadly agree within their uncertainty estimates.

The Sixth Assessment report by Working Group 1 of the Intergovernmental Panel on Climate Change (IPCC) (IPCC in press-a) synthesised several recent studies and methods to assess the net mass budget. From 1992 to 2017, Antarctica’s net mass loss was 2670 ± 530 gigatonnes (Gt) of ice equivalent (Figure 5a). Each 360 Gt of ice mass converts to approximately 1 millimetre (mm) of sea level rise, and the Antarctic ice sheet has contributed an estimated 7.4 ± 1.5 mm to mean sea level rise from 1992 to 2020.

The rate of mass loss from Antarctica has increased over recent decades. The average loss of ice in 1992–2001 was 51 ± 73 Gt per year (Gt/yr). This increased to 82 ± 27 Gt/yr in 2002–11 and again to 199 ± 26 Gt/yr in 2012–17 (Meredith et al. in press). These rates mark an increase on those reported in the 2016 state of the environment report. Much of the difference is due to accelerating mass changes (loss) in both West Antarctica and vulnerable areas in East Antarctica (Table 1).

Mass loss is concentrated around Antarctica’s coastal margins, particularly at the ice shelves that fringe the continent. Ice shelves consist of floating ice where the continental ice discharges to the ocean. Importantly, the ice shelves impede the flow of ice discharge from the grounded ice behind. Although removal of floating ice has little direct impact on sea level, the removal of ice shelves allows accelerated discharge from the continent, with a consequent impact on sea level. More than 80% of the continental ice drains through such floating ice shelves (Pritchard et al. 2012).

Assessment of the net mass budget is complex, because of large unknowns associated with the state and rate of change in ocean-driven melting, snowfall and iceberg discharge between regions (Pattyn & Morlighem 2020). Abrupt changes have been observed in some coastal regions, including the rapid disintegration of floating ice shelves (Scambos et al. 2003). This has raised questions about the potential for rapid ice discharge from Antarctica into the sea, particularly from areas that are below sea level. However, because mass loss from floating ice shelves has little effect on sea levels, mass budget calculations focus only on mass gain and loss from the continental ice sheet.

The dominant driver of mass loss from ice shelves in most cases is increased ocean melting (Pritchard et al. 2012, Gudmundsson et al. 2019, Rignot et al. 2019). However, the effects of heat transport by both the atmosphere and the ocean are also important. This is particularly the case on the Antarctic Peninsula (Cook et al. 2005), where atmosphere-driven surface melting and reduced sea ice conditions have caused the rapid disintegration of some ice shelves (Scambos et al. 2003, Massom et al. 2018).

Regional ice-sheet changes

The Antarctic ice sheet consists of 3 geographically different regions:

• the Antarctic Peninsula, which reaches further north than any other area in Antarctica
• the West Antarctic Ice Sheet
• the East Antarctic Ice Sheet, which is by far the largest component, extending from about 30°W to about 165°E.

All recent studies agree that ice losses from the Antarctic Peninsula and the West Antarctic Ice Sheet have increased since the mid-2000s (Bamber et al. 2018, Gardner et al. 2018, IMBIE team 2018, Rignot et al. 2019). These losses represent the main Antarctic contribution to sea level rise.

During the latter part of the 20th century, the Antarctic Peninsula experienced one of the highest regional temperature increases on the planet (2.8 °C in 50 years), although this trend has since decreased (Turner et al. 2016). Several floating ice shelves in that region collapsed abruptly – for example, the Larsen B Ice Shelf collapsed in March 2002, and collapse and disintegration of large parts of the Wilkins Ice Shelf occurred in 2008 and 2009 (Steig et al. 2009, Humbert et al. 2010). By 2009, the Antarctic Peninsula had lost about 18% of the area of ice shelves present in the 1950s (Cook & Vaughan 2010), a decline of 28,100 km². With the buttressing effect of grounded ice shelves gone, glaciers adjacent to the collapsing ice shelves accelerated to flow approximately 3–4 times faster into the ocean (Scambos et al. 2003, Rintoul 2007). In contrast, other important ice shelves, such as the Amery and Ross ice shelves, have remained comparatively stable (Porter et al. 2019, Zhou et al. 2019).

Changes in West Antarctica have dominated the mass loss from Antarctica to date (Figure 5a). The West Antarctic Ice Sheet has a much larger volume than that of the combined glaciers on the Antarctic Peninsula, and stores enough ice to raise sea levels by 5.3 m (Morlighem et al. 2020). Areas of West Antarctica in contact with warm ocean waters have experienced ice-shelf thinning, and retreat of the grounding line (where the ice shelf starts to float) (Gudmundsson et al. 2019). This change has also spread inland, with grounded ice experiencing thinning and accelerated retreat of the grounding line (Smith et al. 2020).

The grounding line retreat may become irreversible in the near future, as much of West Antarctica sits on a retrograde bed slope (i.e. it slopes downward towards the interior of Antarctica), meaning that grounding line retreats cause a positive feedback (see case study: Vulnerability of the Antarctic ice sheet to future climate change). However, recent evidence suggests that rapid bedrock uplift due to recent ice melt may slow this retreat (Larour et al. 2019).

The situation for the East Antarctic Ice Sheet is more complicated. The East Antarctic Ice Sheet is the largest of the Antarctic ice sheets, and contains enough stored ice to raise sea levels by 52 m. The East Antarctic Ice Sheet appears to be generally close to being in balance or potentially in gain, with mass losses from ocean-driven melt compensated by increased snowfall (Martin-Español et al. 2017, Bamber et al. 2018, Gardner et al. 2018, IMBIE team 2018). The estimates of mass gain are smaller than in the 2016 state of the environment report because several key areas are losing mass at an accelerating rate. Rignot et al. (2019) estimated that the East Antarctic Ice Sheet contributed overall 4.4 ± 0.9 mm to sea level rise over 1979–2017 (approximately 20% of Antarctica’s total contribution, or roughly 10% of the global increase over this period).

There are also regional differences in the response of the ice sheet within East Antarctica to climate change (Edwards et al. 2021). Increases in snowfall are concentrated in Queen Maud Land (Velicogna et al. 2014, Smith et al. 2020). Several studies have measured mass loss from Wilkes Land since the mid-2000s (Velicogna et al. 2014, Gardner et al. 2018, Smith et al. 2020). This is of particular concern, as the Aurora and Wilkes subglacial basins are susceptible to irreversible grounding line retreat and rapid destabilisation of the margins of the ice sheet (Mengel & Levermann 2014, Gwyther et al. 2018).

Heard Island glaciers

On Heard Island, the glaciers have retreated since the first measurements in the 1940s, and have continued to retreat over the past decade. The first aerial survey of the glaciers on Heard Island was in the 1940s, and the glaciated area of the island was 288 km². The glaciated area had decreased to 256 km² by the 1980s (Ruddell 2006), to 236 km² in 2012 and to 235 km² in 2014. The most recent survey from 2016 of the larger Big Ben glaciers (Baudissin, Challenger, Downes, Ealey, Compton, Brown, Fifty-One, Gotley and Lied glaciers) indicated that there has been a further decrease in the area of these larger glaciers since the 1940s (Donoghue 2021).

Examination of more recent satellite images of the eastern glaciers indicates that these leeward glaciers continue to retreat at greater rates than those on the windward side of the island. For example, the 2011 and 2016 state of the environment reports stated that Brown Glacier had retreated from an area of 6.2 km² in 1947 to 4.4 km² in 2004, 3.6 km² in 2008 (Lucieer et al. 2009, Harris 2018), 3.5 km² in 2014 (Donoghue & Harris 2017) and 3.4 km² in 2016. This represents a total loss of 45% since 1947 (Donoghue 2021). As of January 2019, Brown Glacier had continued to retreat. The overall retreat of glaciers over 2000–19 was reported by Hugonnet et al. (2021) for Heard Island, as well as in other subantarctic regions, including the Kerguelen Islands, South Georgia and nearby islands.

Big Ben, a stratovolcano that rises to 2,745 m above sea level, has been frequently active since the 1910s. Because of the remoteness of the island and infrequent visits, most of the recent reports of volcanic activity have been from Middle InfraRed Observation of Volcanic Activity (MIROVA) analysis of satellite imagery. The most recent eruptive activity, reported by the Global Volcanism Program et al. (2020), occurred from October 2019 to March 2020. As is typical of Heard Island, cloud cover persisted over this period, but a possible lava flow – apparent as a hotspot – appeared to have extended south-west from the summit.

Sea ice

Antarctic sea ice plays a crucial role in the climate system, as well as the structure and function of high-latitude Southern Ocean ecosystems. Sea ice around Antarctica has 2 main components. The most extensive is ‘pack ice’, which is made up of individual pieces called ‘floes’ that are in constant motion in response to winds and ocean currents. The other main component is stationary (nondrifting) ‘fast ice’, which forms as a narrow band of compact sea ice generally tens of kilometres, but up to 250 km, wide. Fast ice is confined to Antarctica’s coastal margins, where it is held in place by icebergs grounded in waters shallower than about 450 m, coastal promontories, islands and sheltered embayments (Fraser et al. 2020). This form of floating ice is of key significance in coastal and ice-sheet processes and ecosystems (Massom et al. 2010, Massom & Stammerjohn 2010). It is also of importance to human operations where, for example, it can aid or hinder the discharge of heavy cargo and personnel from ships to Antarctic stations.

Sea ice changes

Sea ice in the Arctic and Southern oceans has different characteristics and patterns of large-scale change because of differences in geographical settings and the processes affecting them. Whereas landmasses largely enclose the sea ice in the Arctic Ocean, Antarctic sea ice surrounds the continent and is constantly exposed to Southern Ocean storms and waves.

Based on a consistent satellite record since 1979, overall Antarctic sea ice extent increased by 1.0 ± 0.5% per decade (about 11,300 km² per year) over 1979–2018 (Parkinson 2019), although the trend in sea ice area over 1979–2020 is not considered significant (IPCC in press-b). The past few years of the record have shown considerable interannual fluctuations (Figure 7; see case study: An extreme decline in Antarctic sea ice in 2016, for a discussion of recent annual records). Much of this increase occurred from 1979 to 2014. The years 2012, 2013 and 2014 set successive records for high sea ice extent (Reid & Massom 2015); this was followed by a sharp switch to record low extents for particular months during 2017–19. In contrast, Arctic sea ice over 1979–2020 decreased by 13.1% per decade (82,700 km² per year) relative to the 1981–2010 average (Perovich et al. 2020).

The overall trend in Antarctic sea ice extent masks strong regional (Parkinson 2019) and seasonal (Holland 2014) trends that are influenced by large-scale modes of climate variability (Yang et al. in press). Whereas sea ice extent decreased by 2.5 ± 1.2% per decade in the Amundsen and Bellingshausen seas sector for 1979–2018, other sectors displayed increases, ranging from +1.0 ± 0.8% per decade in the Weddell Sea to +2.3 ± 1.2% per decade in the south-west Pacific Ocean (Parkinson 2019).

Contrasting regional patterns of change are also apparent in the annual seasonality of sea ice coverage around Antarctica, compared with more uniform change around the Arctic (Stammerjohn et al. 2012). In the Amundsen, Bellingshausen and north-western Weddell seas, annual sea ice duration lengthened by about 2–3 months from 1979 to 2016 (Stammerjohn & Maksym 2017), due to a 1–2-month later ice advance in autumn and a 1-month earlier spring–summer retreat. In the western Ross Sea, the annual ice season lengthened by about 2–3 months due to earlier advance and later retreat. Across East Antarctica and within the AAT, the changes in the annual duration of sea ice are generally smaller and vary across zones (Hobbs et al. 2016a), with localised positive and negative trends (Massom et al. 2013).

Unexpected high variability has occurred in the overall Antarctic sea ice extent since 2012. Persistent daily positive anomalies and record highs of total sea ice extent in 2012 to mid-2015 (Reid & Massom 2015) were followed by an abrupt change to daily negatives and record lows from 2016 to early 2020 (Meehl et al. 2019, Parkinson 2019). The Weddell Sea made the largest contribution to the post-2016 decrease (Turner et al. 2020b).

In addition to sea ice extent, sea ice thickness and volume are important variables to monitor. However, available information is inadequate to determine whether large-scale sea ice thickness and therefore volume are changing around Antarctica (IPCC in press-a). Surface, air and under-ice observations of Antarctic sea ice thickness are sparse (Worby et al. 2008), as are measurements of snow cover depth (Webster et al. 2018). Estimates of Antarctic sea ice thickness and associated snow depth are emerging from analysis of satellite altimeter datasets (Paul et al. 2018, Kacimi & Kwok 2020), but definite trends are yet to emerge. Validation of the satellite-derived thickness estimates is a challenge (Newman et al. 2019). This is a critical knowledge gap because the global climatic importance of Antarctic sea ice stems in large part from the volume of ice that freezes and melts each year, rather than from the coverage alone.

In the 2016 state of the environment report, trends in fast ice extent in East Antarctica could not be identified with confidence because of the short (8.8-year) satellite timeseries of fast ice available (Fraser et al. 2012). Recent work on mapping the distribution and extent of fast ice around Antarctica has shown evidence of a small decline in the total coverage from 2000 to 2018 (Fraser et al. 2020).

Causes of sea ice changes

Several factors may contribute to the observed increase in overall Antarctic sea ice coverage during 1979–2014, and its pronounced regional and seasonal variability (Matear et al. 2015, Hobbs et al. 2016b). These include:

• strengthening of the Southern Ocean westerly winds encircling the Antarctic continent and associated cooling of surface waters (Armour & Bitz 2015)
• changes in atmospheric pressure patterns around Antarctica, and associated meridional winds and temperature, as they drive sea ice drift, formation and melt (Holland & Kwok 2012, Haumann et al. 2014).

Uncertainty remains as to causes of the dramatic, unanticipated decrease in overall Antarctic sea ice extent after 2016, which has been attributed to a complex combination of atmospheric and oceanic forcing (Kusahara et al. 2019, Meehl et al. 2019, Turner et al. 2020b), with differing regional contributions (Reid et al. 2020).

Because of large interannual variability and the relatively short duration of observations, changes in sea ice cover are too small to be separated from natural variability in the climate system (Hobbs et al. 2016a). Current climate models do not adequately simulate the observed patterns of change and variability in sea ice extent and seasonality since 1979 (Hobbs et al. 2016b, National Academies of Sciences 2017). This also confounds our ability to detect and attribute changes in sea ice cover. The model deficiencies relate to inaccurate understanding and parameterisation of the complex interactive processes and feedbacks within the southern sea ice–atmosphere–ocean–ice-sheet system and the various anthropogenic (human) forcings involved (i.e. greenhouse gases and ozone), together with decadal-scale natural variability (Notz & Bitz 2017).

Impact of sea ice changes

Given the pivotal role of sea ice in driving water mass transformations in the Southern Ocean (Abernathey et al. 2016, Pellichero et al. 2018), changes in sea ice distribution, properties and processes have strong potential to affect overturning ocean circulation and climate on decadal and longer timescales (Bindoff & Hobbs 2016, IPCC in press-a). Sea ice changes are also influencing the physical state of the Southern Ocean. An increase in wind-driven sea ice export of 20 ± 10% from 1982 to 2008 has produced ocean freshening (reduction in salt content per mass of seawater) of 0.002 ± 0.001 grams per kilogram per year in intermediate and surface waters (Haumann et al. 2016).

Regional sea ice changes also have indirect consequences for sea level rise. Sea ice loss to the west and north-west of the Antarctic Peninsula has been implicated as a trigger mechanism for the catastrophic rapid disintegration of 3 major ice shelves (Larsen A, Larsen B and Wilkins) since 1995, by increasing the exposure of damaged (highly crevassed) ice-shelf margins to destructive ocean swells (Massom et al. 2018).

There is growing evidence from the Arctic that sea ice loss has influenced mid-latitude seasonal climate in the Northern Hemisphere (IPCC in press-a). For the Southern Hemisphere, modelling suggests that changes in Antarctic sea ice appear to have had a comparatively smaller effect on climate (England et al. 2018). Projected future Antarctic sea ice loss will potentially increase warming and precipitation changes in the tropics (England et al. 2020).

Although relatively little is known about the overall environmental, ecological and climatic impacts of observed Antarctic sea ice change and variability compared with the Arctic (Meredith et al. in press), evidence is emerging that these impacts can be significant on both local and global scales. As detected by the Palmer Long-Term Ecological Research program, declining sea ice extent and annual duration west of the Antarctic Peninsula have had substantial and cascading impacts on ecosystem food-web structure and function, and biodiversity (Ducklow et al. 2013). Sea ice changes have been linked to:

• changes in primary production (phytoplankton species)
• a poleward migration of ice-dependent penguin and seal species due to habitat contraction and impacts on foraging success (prey distributions)
• the incursion (poleward migration) of subantarctic and warmer-water species.

Less clear are potential influences of sea ice change on the distribution of Antarctic krill (Euphausia superba), which is a major fishery, and a keystone prey species for fish, penguins, seals and whales (Atkinson et al. 2019).

These changes and their impacts are projected to continue and increase. Net marine primary production around Antarctica is predicted to increase in response to sea ice loss, warming and changing nutrient supply, due to shifts in ocean stratification and upwelling (IPCC in press-a). The habitat of krill in the Southern Ocean is projected to contract southward (IPCC in press-a). Current large uncertainty in future Antarctic sea ice conditions, however, poses challenges to the accurate assessment of potential impacts on logistical operations in support of Antarctic stations (involving both ships and aircraft), fishing, marine research vessels and tourism activities (Chown 2017).

Southern Ocean

The Southern Ocean surrounds Antarctica and covers 14% of the planet’s surface; it plays a key role in global climate and the global carbon cycle (Fogwill et al. 2020). It connects the 3 main ocean basins (Atlantic, Pacific and Indian) and strongly influences the global oceanic circulation system through the Antarctic Circumpolar Current (ACC), the world’s largest oceanic current (Rintoul 2018). The ACC flows from west to east around Antarctica and generates an overturning circulation that transports vast amounts of heat. The ACC also takes up a significant amount of carbon dioxide (CO₂) from the atmosphere (Rintoul et al. 2001, Rintoul 2018). The formation and circulation of Southern Ocean water masses provide a key link in the global ‘conveyor belt’ of ocean currents that controls climate by transporting heat and other properties (see Sea ice). The Southern Ocean influences the mass balance of Antarctica by influencing air temperatures, winds and precipitation, particularly near the coast, and the melting of ice sheets and coastal ice (Holland et al. 2020).

The Southern Ocean is changing in ways that are likely to affect regional and global climate (Rintoul et al. 2018, Meredith et al. in press), and marine productivity (Rhein et al. 2013, Deppeler & Davidson 2017, Boyd 2019). The changes include warming and acidification (see Pressures), as well as changes to ocean properties, circulation and sea level.

Ocean properties

At Southern Hemisphere latitudes north of about 50°S, the surface temperature of the Southern Ocean has shown a warming trend in recent decades (Sallée 2018). However, south of about 50°S towards the Antarctic coast, the surface waters have, on average, cooled and freshened (Durack et al. 2012, Swart et al. 2018, Bronselaer et al. 2020). These high-latitude changes are likely to reflect mainly increases in precipitation and Antarctic ice melt (Böning et al. 2008, Helm et al. 2010), and ocean upwelling (Armour et al. 2016, Hogg et al. 2017, Tamsitt et al. 2017). Antarctic bottom water in some locations has warmed, freshened and decreased in volume (van Wijk & Rintoul 2014, Sallée 2018); however, there has been a reversal of the freshening trend in recent years in the Australian–Antarctic basin, with near-bottom salinities in 2018–19 higher than during 2011–15 (Aoki et al. 2020).

Ozone depletion and increases in atmospheric greenhouse gases caused by human activities have resulted in changes in wind patterns in southern subpolar latitudes, which have influenced the heat content, salinity and dissolved oxygen content of the Southern Ocean (Turner et al. 2014, Rintoul et al. 2018, Swart et al. 2018).

Sea level rise

Over much of the global ocean, sea level is rising, and the rise is accelerating. The global mean sea level change obtained from tide gauges and altimetry observations increased from 3.2 mm/yr over 1993–2015 to 3.7 ± 0.5 mm/yr over 2006–18 (IMBIE team 2018). Glacier and ice-sheet contributions are now the main source of the rise, and these are associated with anthropogenic (human) forcing (Meredith et al. in press).

The rate of change of sea level has been regionally and globally variable in recent decades because of influences from natural climate variability and volcanic events. The influence of natural variability is noticeable particularly where the ocean is cooling, and hence, contracting; this can be caused by the ENSO mode of large-scale climate variability.

The rate of sea level rise is expected to increase because of continuing global warming (Fasullo et al. 2016). The fate and response of the ice sheets to global warming are the greatest source of uncertainty impeding constrained projections of future sea level estimates; the likely upper bound is 25 m of sea level for a climate that is 2–4 °C warmer than pre-industrial temperatures (Carson et al. 2019, Meredith et al. in press).

In the Southern Ocean, sea level is difficult to measure because of the presence of sea ice around the continent, which impedes altimetry observations of the surface, and influences from the ongoing unloading of forces on bedrock after the last ice age. From 1970 to 2018, most of the Southern Ocean increased in height; the region between 30°S and 60°S showed accelerating trends that were among the largest anywhere on the globe (Church et al. 2013, Wang et al. 2021). Over this period, parts of the Pacific sector fell modestly, which is ascribed to upwelling of cold waters near the Antarctic coast (Armour et al. 2016).

Marine heatwaves

Marine heatwaves (MHWs) are events occurring in the global oceans during which water temperatures are anomalously warm (up to several degrees Celsius above average) for extended periods (typically days to a few weeks). These events can have disruptive influences on marine ecosystems and their biodiversity, and regional fisheries (Holbrook et al. 2020, Samuels et al. 2021, Su et al. 2021). Several physical processes can produce MHWs. These can be broadly categorised as alteration of the oceanic heat transport, influences from slowly changing atmospheric patterns, and effects from large-scale modes of climate variability associated with ocean–atmosphere coupling (Holbrook et al. 2019, Sen Gupta et al. 2020).

Oliver et al. (2018) provided a global assessment of MHWs using a range of ocean temperature data. Over 1982–2016, MHWs have increased in frequency in northern parts of the Southern Ocean, particularly in the central Pacific sector, where the trends have been consistent with ocean warming. However, at latitudes poleward of 50°S, MHWs have decreased in frequency in recent decades (Oliver et al. 2018), particularly in the central Pacific sector, where ocean cooling has been observed (Haumann et al. 2020). Although none of the major climate modes dominantly result in MWHs in the Southern Ocean, some regions are preferentially affected by the SAM, ENSO and the Interdecadal Pacific Oscillation (Holbrook et al. 2019).

The impact of potential changes in the characteristics of MHWs in the Southern Ocean and coastal Antarctica is of concern, because food webs are adapted to a specific range of conditions. Recent model assessments suggest that MHWs will generally increase in frequency and duration over coming decades across all oceans, and lead to greater environmental effects. However, an exception is the coastal region of East Antarctica, where changes in the intensity of heatwaves are not projected to be significant (Oliver et al. 2019, Hayashida et al. 2020) because ocean warming in this region is relatively slow.