Antarctic Sea Ice #1: Physical Role and Function

Sea ice is formed by freezing the ocean surface (at about -1.8° C) and is a major and highly-dynamic component of the Earth’s cryosphere. Each year, sea ice transforms a vast area of the Southern Ocean south of ~55° S (Fig. 1), with its coverage increasing more than five-fold from 2-4 million km² in February to 19-20 million km² in September (almost 1.5 times the area of the Antarctic continent) (Parkinson 2019). In so doing, sea ice has a major influence on the ocean and atmosphere and on key physical, biological and chemical processes. This influence is not only local-to-regional but also global in its reach through its impact on global heat transport through  ocean and atmospheric circulation (Rintoul et al. 2018; England et al. 2020).

Antarctic sea ice differs from its Arctic counterpart, due in large part to the different geographical settings of the two polar regions (Maksym, 2019). Whereas Arctic sea ice extends to the pole and is largely enclosed by land masses, Antarctic sea ice occurs around a continent that straddles the South Pole. As such, Antarctic sea ice extends no further south than approximately 78°S, but its equatorward extent (to a maximum of about 55°S in winter) is not constrained by land. Rather, it remains in contact with – and is strongly influenced by – the stormy Southern Ocean, home to some of the strongest winds and largest waves on the planet. Distinct differences also exist in the mean overall annual cycles and seasonality. While the Arctic sea-ice cycle is symmetrical (with virtually equal-length growth and melt seasons), Antarctic sea ice has a relatively short (5-month) melt-back season but a relatively long (7-month) growth season (Parkinson 2019). Factors driving Antarctic sea-ice seasonality are given in Stammerjohn and Maksym (2017), Eayrs et al. (2019) and Roach et al. (2022).

Antarctic sea ice comprises two main components:

• dynamic “pack ice” made up of ice parcels known as “floes” that are in constant motion under the influence of winds, currents, tides and waves; and
• a relatively narrow coastal zone of stationary and consolidated “fast ice” that is attached (in places) to the Antarctic coastline, islands and/or icebergs that are grounded in waters less than ~500m deep (Massom and Stammerjohn 2010; Fraser et al. 2020).

In both cases, the sea ice intercepts snowfall that would otherwise fall directly into the ocean, with the resultant snow cover substantially modifying the physical, biogeochemical and optical properties of the sea ice (Sturm and Massom 2017).

Sea ice is influenced by an interplay of thermodynamic processes (freeze and melt) and dynamic processes (ice motion and deformation from wind, ocean currents, tides and waves) (Weeks, 2010). The thermodynamic processes of freeze and melt are finely balanced between ocean salinity (salt) content and ocean temperature. For example, increased upper-ocean stratification due to enhanced freshwater input from ice-shelf melt may increase regional sea-ice formation and compete against or counter ocean warming (Bintanja et al. 2013; Pellichero et al. 2018). Teleconnections from the tropics also influence the position and strength of cyclones and anticyclones, changing the regional growth/advance and melt/retreat of sea ice (Stammerjohn and Maksym 2017; Yuan et al. 2018). Moreover, and in addition to local growth/retreat, the amount and conditions of sea ice in any given area are influenced by the zonal transport (advection) of sea ice both within the Antarctic Coastal Current (westward) and the Antarctic Circumpolar Current (eastward); this contribution is approximately an order of magnitude smaller than local processes (Kimura and Wakatsuchi 2011).

Important characteristics of sea ice include its overall spatial extent; concentration of coverage in any given area; seasonality (timings of annual ice-edge advance and retreat and resultant duration of coverage); rates of formation and melt; movement and deformation; and the thickness of both the ice and its snow cover.

Covering up to ~4% of the planet’s surface (6.5% including the Antarctic continent), Antarctic sea ice has a profound influence on high-latitude atmosphere-ocean interaction and processes, and is key to keeping the southern polar region cold. On the one hand, sea ice substantially increases the reflectivity (albedo) of the surface ocean, with ~80-90% of incoming solar radiation being reflected for snow-covered sea ice compared to ~7% for ice-free ocean (Brandt et al. 2005) – meaning that snow-covered sea ice reflects nearly all incoming sunlight to shield the much darker ocean surface from direct heating by the Sun.

During the winter, when little to no sunlight reaches the high latitudes, the presence of sea ice greatly reduces ocean-atmosphere exchanges of heat, greenhouse gases (e.g., water vapour and CO₂) and wind energy (Weeks 2010). Any open ocean area within the winter pack ice becomes a relative ‘hot spot’ for ocean heat loss and near-surface atmospheric warming (Turner et al. 2013), along with increased water vapour and CO₂ release to the atmosphere. In summer when rapid melting/retreat occurs, the presence of sea ice prevents the ocean’s absorption of incoming solar radiation, thus also maintaining colder surface air temperatures. Because sea ice also decreases momentum (wind energy) transfer, it also reduces wind mixing of the upper ocean (McPhee 1991).

The accumulation of a snow cover substantially modifies the physical and optical properties of the sea ice and influences atmosphere-ocean interactions and processes (Sturm and Massom 2017). Snow not only has one of the highest albedos of any natural substance on the planet (Perovich 1990), but it is also a strong thermal insulator, thus decreasing heat loss from the ice and slowing its thickening (Sturm and Massom 2017). Simultaneously, and somewhat paradoxically, snow also contributes to increased thickening by weighing and depressing the ice surface below the sea surface, leading to flooding and the creation of a slush layer that forms “snow ice” when it freezes on to the existing ice surface (Massom et al. 2001). Snow-ice formation is a widespread but poorly quantified phenomenon around Antarctica (Maksym and Markus, 2008), and one that will likely become increasingly important in a warming climate as atmospheric moisture content and snowfall are likely to increase (Massom and Stammerjohn 2010). In late-spring through summer, the presence of a thick snow cover can also extend the ice season in certain regions, including producing areas of multi-year or perennial sea ice (Eicken et al. 1995).

The formation and decay of sea ice strongly influences the ocean. Each winter when seawater freezes and crystalizes into ice, it expels most of the salts, producing a dense brine that mixes downward in the ocean. In this way, the production of sea ice in autumn/winter removes freshwater from the ocean, but in spring/summer that freshwater is returned when sea ice melts. This seasonal cycle redistributes salt and freshwater, and is a key player in the seasonal structure, stratification and properties of the Southern Ocean and its circulation (Pellichero et al. 2018). Particularly high rates of sea-ice production and resultant salt/brine rejection occur in coastal polynyas (Tamura et al. 2016), where recurrent and persistent openings in the sea-ice cover are maintained by strong winds and ocean currents (Barber and Massom 2007). In these regions of extremely high sea-ice production, full vertical mixing of rejected brine can occur. In a limited number of important  polynyas, (i.e. Ross Sea, Weddell Sea, Adélie Land and Mac. Robertson Land) this creates a very cold, dense and salty water mass known as Antarctic Bottom Water (AABW). In turn, AABW is a major component of the global ocean overturning (thermohaline) circulation that is central to, and helps regulate, Earth’s climate (Meredith and Brandon 2017). Any substantial long-term decrease in polynya size or duration and associated sea-ice production will result in reduced production of AABW, slowing the global ocean thermohaline circulation and altering its effect on global climate on decadal to centennial time scales (Broecker 1991). The formation of AABW is also susceptible to increased freshwater input from enhanced basal melt of coastal ice-shelves and glaciers as well as surface meltwater runoff (Williams et al. 2016).

Sea ice also mediates the uptake of anthropogenic heat and CO₂ from the atmosphere (Bitz et al. 2006; Fogwill et al. 2020), via its influence on Southern Ocean circulation, stratification and mixing, temperature gradients, and biogeochemical processes. This is of major climatic importance in a warming world, because the high-latitude Southern Ocean serves as a crucial heat and carbon sink for the planet and is therefore a moderator of anthropogenic climate change (Frölicher et al. 2015).

In addition, sea ice plays multiple important roles around the floating margins of the Antarctic Ice Sheet. Sea ice helps maintain the stability and buttressing capacity of ice shelves (Massom et al., 2010), thereby contributing to regulating glacial flow into the ocean and resultant sea-level rise. It can slow glacier flow/reduce mass loss by locking onto the face of and mechanically bonding ice shelves (in the form of fast ice), thus protecting fractured ice-shelf outer margins from potentially-destructive ocean waves (both pack and fast ice) (Massom et al. 2018). As a result, break-out of fast ice (Arthur et al. 2021) or an increase in sea-ice-free conditions (Reid and Massom, 2022) can increase the susceptibility of ice shelf margins to calving, and can even trigger wider-scale disintegration of ice shelves that are already weakened by fracture, melt and thinning, e.g., the Larsen A and B and Wilkins ice shelves on the Antarctic Peninsula since 1995 (Massom et al. 2018). At the same time and in certain regions, sea ice processes modulate the incursion of warm ocean waters into ice shelf cavities (Khazendar et al. 2013), to enhance or suppress ice-shelf basal melt and thinning (Rignot et al. 2013).

Sea ice also plays an important role in polar biogeochemical cycles that link to ecosystems and weather/climate by acting as an active biogeochemical reactor at the interface between the atmosphere and ocean (Vancoppenolle et al. 2013; BEPSII). In this way, coupled physical, biological, and chemical processes occurring within the sea ice substrate result in ocean-atmosphere fluxes of biogenic gases (including CO₂ and dimethyl sulphide) that are active constituents of the climate system. For example, atmospheric aerosols related to dimethyl sulphide emitted from ice algae and ice-edge phytoplankton blooms form nuclei for the formation of clouds (Charlson et al. 1987), which then regulate the surface radiation budget. This represents a potentially important feedback mechanism that requires further research (Goosse et al. 2018).

Finally, sea ice has a major impact on shipping and human activities around Antarctica, including ship-based scientific research, logistical operations such as base resupply (COMNAP 2015) and commercial activities such as fisheries (CCAMLR) and tourism (IAATO). Such activities require accurate sea-ice information and (model) forecasting. The next generation of coupled (atmosphere-ocean-sea ice) models will step towards filling this critical void (Jung et al., 2016).

Our understanding of the coupled Antarctic sea ice system – and change and variability therein – is currently held back by a lack of adequate observations of key aspects of the sea-ice environment (notably ice and snow thickness, and under-ice ocean conditions) and the processes involved (Polar Research Board et al. 2017). This critical gap needs addressing with a coordinated programme of integrated and sustained cross-disciplinary observations of the coupled sea ice-Southern Ocean-atmosphere-ice sheet system – as proposed by (Smith et al. 2019) and the Southern Ocean Observing System or SOOS (Newman et al. 2019) – together with dedicated cross-disciplinary experimental and process-based studies, combined with Earth System modelling and satellite remote sensing.

Autonomous technologies operating on, above and below the ice are crucial to advancing our understanding of the Antarctic sea ice environment, and also to bridging the time/space scale gap between detailed in situ observations (on the floe scale) and coarse-resolution (but wider-coverage) satellite observations (Smith et al. 2019). Calibration and validation of satellite-derived products are still largely lacking, yet are crucial given the invaluable role of satellites in providing systematic measurement and monitoring of Antarctic sea ice on the large-scale and over prolonged periods (Lubin and Massom 2006).

While the large-scale horizontal distribution of Antarctic sea ice (concentration, area and extent) has been measured reliably and continuously from space since 1979, this is not the case with the thickness of the sea ice and its snow cover (Maksym et al., 2012; Webster et al. 2018). As a result, there is no accurate baseline information on the large-scale thickness distributions of Antarctic sea ice and its snow cover, let alone whether these important quantities are changing. Current thickness knowledge is largely limited to surface and airborne observations that, while detailed and invaluable, are sparse in both space and time. Satellite radar and laser altimeters hold the key to filling this critical gap (e.g., ICESat-2), and important new datasets are emerging (Paul et al. 2018; Kacimi and Kwok 2020), but challenges remain regarding calibration and validation and in quantifying uncertainties (Kern and Spreen 2015).

Lack of accurate information on precipitation rates, and crucially the type of precipitation, over the Antarctic sea ice zone is also a critical knowledge gap (Webster et al. 2018), particularly for understanding its influence on sea ice. Precipitation over the Southern Ocean is difficult to model (Boisvert et al., 2020; Lang et al., 2021), yet its relationship with sea ice is crucial to understanding broader context environmental variables such as ice sheet snow accumulation and mass balance (Wang et al., 2020).

The present generation of climate models exhibits relatively low skill in reproducing observed patterns of variability and change in Antarctic sea ice extent and seasonality, at least as observed by satellites since 1979 (Hobbs et al. 2016). This is due to a combination of: (i) the complexity of the highly-coupled processes (Polar Research Board et al. 2017); (ii) incomplete understanding and parameterization of the interactive processes and feedbacks (Notz and Bitz 2017); and (iii) deficiencies in the ability of climate models to correctly simulate large-scale atmospheric circulation patterns that influence the sea ice (Liu et al. 2002). This in turn substantially limits our ability to predict the likely future trajectory of the Antarctic sea-ice environment and the local to global effects of its change (see Antarctic Sea Ice #3).