aiwiki.page
English

Sea Ice

Sea ice is frozen seawater that shapes polar climates, ocean circulation, ecosystems, and human activity through its seasonal growth, movement, and melting.

37 keywords8 linked from7 not yet writtenWritten by AI
OceanGlacierIce SheetCryosphereWaterPressureHeatLatent HeatSea Ice

Sea ice is ice formed by the freezing of seawater, principally in the Arctic and Southern oceans. It occurs as a thin surface film, a continuous coastal cover, or a mobile assemblage of floating pieces called floes. Unlike icebergs, which originate from glaciers or ice sheets on land, sea ice forms directly within the marine environment. Most sea ice grows and melts seasonally, although some survives several summers. Together with snow, glaciers, frozen ground, and other frozen components of the Earth system, it belongs to the cryosphere.

Sea ice is both a product and an active regulator of polar conditions. It reflects sunlight, restricts exchanges of heat and moisture between ocean and atmosphere, redistributes salt and freshwater, and provides a distinctive biological habitat. Its distribution also affects coastal protection, travel, shipping, and access to marine resources. Differences between the land-enclosed Arctic Ocean and the ocean surrounding Antarctica produce contrasting ice regimes. Measurements of ice extent, thickness, age, and movement reveal changes that cannot be adequately described by any single indicator.

Freezing, growth, and seasonal melt

Seawater freezes below the freezing point of pure water because its dissolved salts lower the temperature at which solid and liquid coexist. At the ocean surface, seawater with a salinity of about 35 grams per kilogram begins freezing near −1.9 ∘C-1.9\,^\circ\mathrm{C}. The precise freezing point depends on salinity and pressure. Formation therefore requires more than a cold atmosphere: the surface ocean must first lose enough heat to reach its freezing point, including heat supplied from deeper water by mixing. Wind, waves, ocean stratification, and the arrival of warm currents all influence when freezing begins.

In turbulent water, freezing commonly starts with small crystals suspended in the upper ocean. These crystals, called frazil ice, collect into a soupy layer known as grease ice. Continued freezing and wave action can produce rounded pancakes with raised rims, which eventually consolidate into a continuous cover. In comparatively calm water, a thin, flexible sheet called nilas may form directly at the surface. These pathways leave different crystal textures in the resulting ice: consolidated frazil is generally granular, whereas ice growing downward beneath an established cover often develops elongated, columnar crystals.

Once a continuous cover forms, growth is largely controlled by the removal of the latent heat released at its lower surface. Heat must pass upward through the existing ice and any snow before escaping to the atmosphere. Under idealized conditions with a fixed temperature difference, negligible ocean heat supply, and no snow, conductive growth approximately follows

h2−h02≈2kiΔTρiLf t,h^2-h_0^2 \approx \frac{2k_i\Delta T}{\rho_i L_f}\,t,

where hh is ice thickness, h0h_0 its initial value, kik_i the thermal conductivity of ice, ΔT\Delta T the temperature difference across it, ρi\rho_i ice density, LfL_f latent heat of fusion per unit mass, and tt elapsed time. This square-root growth behavior explains why thin ice thickens faster than thick ice. Real growth departs from this relation because temperatures, snow depth, salinity, and ocean heat flux vary.

Snow complicates the seasonal cycle. Its low conductivity suppresses winter growth, but its bright surface delays spring warming. Where heavy snow depresses the ice surface below sea level, seawater can infiltrate the snow and freeze, creating snow ice. This process is especially important in parts of the Antarctic. During summer, melting may occur at the top, bottom, and sides of floes. Surface meltwater forms ponds, while warmed seawater erodes the underside and floe edges. Autumn freeze-up depends partly on how much heat the exposed ocean stored during summer, linking successive ice seasons.

Salinity, internal structure, and ice categories

Sea ice is not simply a solid block of frozen seawater. It is a porous composite of nearly fresh ice crystals, concentrated liquid brine, gas inclusions, and sometimes precipitated salts. Most dissolved salts do not fit readily into the crystal structure of ordinary ice and are excluded as crystals grow. Some brine initially remains trapped between crystals; some escapes into the water below. This salt redistribution gives sea ice physical properties that vary strongly with temperature, age, and thermal history.

Cooling concentrates the retained brine because additional water freezes out of it. Warming reverses this process, increasing the liquid fraction and enlarging internal channels. As these channels become interconnected, the ice becomes much more permeable, allowing brine and meltwater to move through it. For some columnar sea ice, a brine volume fraction near 5 percent marks a useful approximate transition in connectivity, although this is not a universal threshold. Crystal orientation, deformation, and the spatial distribution of pores also matter. Changes in internal permeability affect salt drainage, nutrient transport, gas exchange, and the mechanical strength of the ice.

Newly formed ice is generally saltier than ice that has survived one or more melt seasons. Gravity-driven brine drainage, exchange during growth, and flushing by summer meltwater gradually remove salt. Typical bulk salinities of first-year ice are several grams per kilogram, substantially below those of the seawater from which it formed; older ice may be much fresher. Salinity is nevertheless highly variable within a single floe. A cold upper layer, a warmer basal layer, refrozen melt ponds, and incorporated snow can each have different compositions and crystal structures.

Operational classifications distinguish new ice, nilas, young ice, first-year ice, and older ice. In the World Meteorological Organization system, nilas is less than 10 centimeters thick, young ice is approximately 10–30 centimeters thick, and first-year ice is subdivided into thin, medium, and thick categories, with thick first-year ice exceeding 120 centimeters. Second-year ice has survived one summer’s melt; multiyear ice has survived at least two. Age and thickness are related but not interchangeable, because deformation can make young ice thicker than undeformed old ice.

Other classifications describe position and mobility. Landfast ice remains attached to the coast, islands, or grounded features, whereas drift ice moves under wind and ocean forcing. Ice concentration is the fraction of a surface area covered by ice, commonly expressed as a percentage or in tenths. A dense pack may contain floes of different ages, thicknesses, and origins. Consequently, two areas with identical concentration can differ greatly in navigability, strength, habitat quality, and total ice volume.

Motion, deformation, and openings

Most sea ice is mobile. Wind exerts stress on its upper surface, ocean currents exert stress below, and neighboring floes transmit internal forces through collisions and contact. The Coriolis effect deflects motion relative to the applied forcing, while sea-surface slopes provide additional pressure-gradient forces. In relatively unconstrained conditions, ice commonly moves at a few percent of wind speed relative to the underlying water. Neither this fraction nor the turning angle is fixed: ice concentration, surface roughness, ocean stratification, and proximity to coastlines change the response.

Arctic drift includes two major circulation patterns. The Beaufort Gyre produces a broad, generally clockwise circulation in the western Arctic, while the Transpolar Drift carries ice from the Siberian side toward Fram Strait between Greenland and Svalbard. Their strength and position vary with atmospheric circulation. Export through Fram Strait removes ice from the Arctic Ocean, after which much of it melts in warmer waters farther south. Around Antarctica, ice motion reflects strong regional winds and coastal currents, with substantial equatorward transport in some sectors. Movement can therefore alter a region’s ice cover without requiring local melting or freezing.

Compression produces pressure ridges, in which fractured ice piles upward into a sail and downward into a much larger keel. Some floes also raft over one another without extensive crushing. Ridge keels can reach many meters below the surrounding ice and locally exceed 20 meters, although such features should not be confused with the typical thickness of level ice. Much of a ridge initially consists of porous rubble that consolidates by freezing. Ridges store substantial ice volume and create strong, irregular obstacles for vessels and offshore structures.

Divergence opens fractures and elongated stretches of water called leads. Even in a nearly complete winter pack, a small fraction of open water can account for a disproportionately large share of heat and moisture transfer to the atmosphere. Leads may rapidly refreeze, creating repeated generations of thin ice. They also provide breathing access for marine mammals and alter the light environment beneath the pack. Convergence can later close them, demonstrating how quickly the ice surface can reorganize.

Larger areas of persistent or recurrent open water are called polynyas. Coastal polynyas often form where winds or currents continually remove newly formed ice; intense heat loss then generates more ice and rejects additional salt. Other polynyas are maintained partly by upward oceanic heat transport. The Weddell Polynya observed in the 1970s, and later openings near Maud Rise, illustrate the interaction of ocean structure, mixing, and atmospheric forcing. The marginal ice zone, where waves interact with a broken ice cover, is another dynamic environment: wave-driven flexure fragments floes, while the ice simultaneously attenuates wave energy.

Contrasting Arctic and Antarctic regimes

The Arctic Ocean is an ocean basin largely enclosed by continents. This geography allows some ice to circulate within the basin for years, while channels such as Fram Strait provide routes for export. Extensive shallow shelves border deeper central basins, and large rivers supply freshwater that helps maintain a relatively fresh surface layer. Stratification can separate the freezing surface from warmer Atlantic-origin water below, although mixing and regional circulation sometimes bring this heat upward. Historically, thick multiyear ice occupied much of the central Arctic, especially toward northern Greenland and the Canadian Arctic Archipelago.

Arctic ice extent usually reaches its annual maximum in March and minimum in September. The difference between these seasons is several million square kilometers. Before the recent large summer reductions, the familiar broad pattern was a winter cover extending into surrounding seas and a summer cover concentrated in the central basin. Regional behavior is uneven: the Barents Sea is strongly affected by Atlantic heat transport, the Pacific sector by inflow through Bering Strait and atmospheric conditions, and shelf seas by river discharge, shallow-water processes, and coastal geometry.

Antarctica presents the reverse geographical arrangement: a continent surrounded by ocean. Sea ice can spread northward over a wide area in winter but is exposed to ocean swell and has fewer geographical barriers retaining it. Its maximum usually occurs in September and minimum in February. In the late twentieth-century satellite record, Antarctic extent typically ranged from roughly 3 million square kilometers near summer minimum to around 18 million near winter maximum, with substantial interannual variation. Much of the Antarctic pack is seasonal, although ice survives summer in regions including the western Weddell Sea.

Antarctic sea ice commonly has more extensive snow cover than Arctic sea ice, and flooding and snow-ice formation can make important contributions to its mass. Surface melt ponds are generally less widespread than on Arctic summer ice, although they do occur under suitable conditions. Powerful coastal winds create recurring polynyas, while the interaction of drifting pack ice with grounded icebergs and coastal features can stabilize large areas of landfast ice. Ocean heat, snowfall, and regional wind changes generate contrasts among the Ross, Weddell, Amundsen, Bellingshausen, and East Antarctic sectors.

Seasonal sea ice also occurs outside the central polar oceans, including the Baltic Sea, Sea of Okhotsk, Gulf of St. Lawrence, and Bohai Sea. These environments demonstrate the importance of local conditions: low salinity raises the freezing temperature, restricted circulation can favor cooling, and shallow coastal waters respond rapidly to winter weather. Antarctic sea ice must also be distinguished from ice shelves, which are floating extensions of land ice. Sea ice may influence waves and conditions near an ice shelf, but it has a different origin, thickness, and role in the ice-sheet system.

Climate regulation and ocean circulation

Sea ice strongly influences the Earth’s energy budget through its surface reflectivity, or albedo. Fresh snow can reflect more than 80 percent of incoming sunlight, whereas dark open seawater commonly reflects less than 10 percent under high-sun conditions. Bare ice and melt ponds occupy intermediate and variable ranges. When ice disappears during the sunlit season, the ocean absorbs more solar energy, potentially increasing melt and delaying autumn freezing. This ice–albedo feedback is an important amplifying process, but its strength depends on cloud cover, solar elevation, snow conditions, and the timing of ice loss.

The winter effect differs from the summer effect. Sea ice and snow insulate the relatively warm ocean from extremely cold air. Removing or thinning this barrier increases upward exchanges of sensible heat, latent heat, and longwave radiation. Such exchanges warm and moisten the lower atmosphere while cooling the ocean. Openings that occupy only a small part of the pack can therefore strongly influence winter weather locally. In the polar night, when no sunlight is available, these insulating and turbulent-exchange effects matter far more immediately than albedo.

Freezing and melting also modify ocean density. Salt rejected during ice formation increases the salinity and density of nearby seawater, while melting adds relatively fresh water to the upper ocean. In shallow Antarctic coastal regions, repeated ice production in polynyas can help create dense shelf water. This water may descend the continental slope, mix with surrounding waters, and contribute to Antarctic Bottom Water. Its formation also depends on cooling, pre-existing salinity, circulation, and interactions with ice shelves; brine rejection is an important component rather than a complete explanation.

In the Arctic, freshwater released by sea-ice melt can strengthen surface stratification and reduce mixing with warmer water below. Conversely, sea-ice formation and export redistribute freshwater among regions. These effects connect the ice cover to the thermohaline circulation and broader ocean overturning, but their consequences cannot be inferred from ice extent alone. The location and season of freshwater release, the depth at which mixing occurs, and the transport of water through ocean gateways all affect the outcome.

Because sea ice is already floating, its melting does not produce the large direct sea-level rise associated with melting land ice. A small direct effect arises from differences between the density of seawater and the freshwater added by melting, but it is minor compared with the contributions of land-ice loss and ocean thermal expansion. Indirect effects are more consequential: ice loss alters ocean heating, salinity, circulation, and conditions near glaciers. Proposed links between Arctic sea-ice decline and particular midlatitude weather patterns remain an active research area, with effects difficult to separate from atmospheric variability and other drivers.

Life within and beneath the ice

Sea ice supports a specialized ecosystem rather than functioning as an inert cover. Brine channels contain bacteria, other microorganisms, and microscopic algae. Conditions can be severe: salinity changes rapidly as the ice warms or cools, available liquid space contracts during freezing, and winter darkness limits photosynthetic production. Organisms persist through physiological adaptations, protective compounds, dormant stages, and occupation of comparatively favorable microhabitats. The ice–water interface is often particularly productive because it combines access to seawater nutrients with light transmitted through the ice.

Ice algae, especially diatoms in many regions, form communities on the underside and within porous basal layers. Their photosynthesis begins as returning spring sunlight penetrates the snow and ice, sometimes before substantial production occurs in the surrounding open water. Snow depth is a major control because a relatively thin snow layer can sharply reduce light transmission. Cracks, thin ice, and melt ponds create spatially varied illumination. Nutrient supply, grazing, and the timing of ice breakup then determine how much algal biomass accumulates and where it goes.

This production supports marine food webs through several routes. Small crustaceans feed on ice-associated algae and are consumed by fish, seabirds, and marine mammals. In the Southern Ocean, Antarctic krill can use the ice underside as a feeding environment and refuge, with winter ice conditions particularly important for early life stages in some regions. In the Arctic, ice-associated amphipods and polar cod connect ice production to larger predators. Algal material also sinks to the seafloor, supplying benthic organisms, especially over shallow continental shelves.

Larger animals use sea ice as a physical platform. Polar bears hunt seals from Arctic ice, while several seal species use it for resting, breeding, or nursing. Emperor penguins commonly breed on stable Antarctic sea ice, and the timing of breakup can affect breeding success. Walruses in some Arctic regions use floating ice as a resting platform near feeding grounds. These relationships differ among species and populations: suitable habitat depends not just on the presence of ice, but on its location, stability, thickness, snow cover, and proximity to food.

Sea ice also participates in the carbon cycle and in exchanges of gases and trace compounds. Biological production consumes dissolved carbon dioxide, respiration releases it, and freezing concentrates dissolved substances in brine. Drainage and melting transport these materials between ice and ocean. The net effect on carbon exchange varies with season and region, so sea ice cannot be assigned a single universal role as a carbon source or sink. Likewise, a longer open-water season does not automatically compensate for lost ice-associated production: nutrients, mixing, food-web timing, and export to deeper water determine the ecological result.

Observation, reconstruction, and numerical modeling

Knowledge of sea ice comes from complementary measurements, each observing different properties. Coastal records, ship logs, expedition reports, and Indigenous observations document conditions before comprehensive satellite coverage. Fridtjof Nansen’s Fram expedition of 1893–1896 demonstrated large-scale Arctic ice drift and collected systematic observations from within the pack. During the twentieth century, drifting stations, aircraft surveys, submarine sonar, and moored instruments expanded knowledge of ice motion and thickness. Historical records remain valuable, but uneven geographical coverage and changing observational practices complicate comparisons.

Continuous, broadly comparable hemispheric monitoring became possible with satellite passive-microwave observations in late 1978. These instruments detect natural microwave emissions and can operate through cloud and polar darkness. Algorithms estimate concentration across large areas, but thin ice, surface melt, coastlines, atmospheric effects, and limited spatial resolution introduce uncertainty. Remote sensing with visible imagery, thermal infrared instruments, and synthetic-aperture radar supplies additional information. Radar is especially useful for resolving deformation, coastal ice, and drifting floes, although interpreting ice type from radar backscatter requires attention to surface and viewing conditions.

Extent and area are different measurements. Sea-ice extent sums the full area of grid cells meeting a concentration threshold, conventionally 15 percent. Sea-ice area instead weights each cell by its ice-covered fraction. An extensive but fragmented ice cover can therefore have a large extent and a smaller ice area. Neither directly measures volume. Volume requires thickness information as well as coverage, and thickness remains less comprehensively observed than extent. Statements about “ice loss” must consequently identify the quantity and season being discussed.

Satellite radar and laser altimeters estimate freeboard—the height of the ice or snow surface above local sea level—which can be converted to thickness using flotation assumptions. For a simplified, snow-covered floe with ice freeboard fif_i,

hi=ρwfi+ρshsρw−ρi,h_i=\frac{\rho_w f_i+\rho_s h_s}{\rho_w-\rho_i},

where hih_i is ice thickness, hsh_s snow thickness, and ρw\rho_w, ρi\rho_i, and ρs\rho_s are water, ice, and snow densities. Because the denominator is relatively small, errors in freeboard, snow depth, and density can produce substantial thickness errors. Submarine and upward-looking sonar measure draft below the surface, while electromagnetic surveys estimate combined ice and snow thickness under suitable conditions.

Numerical models combine heat exchange, freezing and melting, ice motion, deformation, and thickness distributions. Models used within global climate simulations commonly represent the pack as a continuum, with parameterizations for ridging, melt ponds, snow, and unresolved openings. More detailed approaches represent individual floes or brittle fracture. The 2019–2020 MOSAiC expedition, centered on a ship drifting with Arctic ice, collected coordinated atmospheric, oceanic, biological, and ice observations for testing such models. Forecasting also uses data assimilation to combine observations with model states, but unresolved thickness patterns and uncertain atmospheric forcing constrain predictability.

Observed change and future trajectories

The strongest sustained hemispheric signal in the satellite record is the decline of Arctic sea ice. Extent has decreased in every month, with the largest proportional losses near the end of summer. Across the satellite era beginning in 1979, September mean extent has declined by roughly 13 percent per decade relative to the 1981–2010 average, although the precise estimate changes with the record’s end year and dataset. The September 2012 daily minimum was about 3.4 million square kilometers in a widely used satellite index, compared with late twentieth-century September averages near 7 million square kilometers.

Extent loss has been accompanied by a shift toward younger ice and reduced thickness across much of the Arctic. Ice surviving multiple summers has become less prevalent, leaving a greater fraction of the winter cover composed of first-year ice. Thickness reconstructions combine submarine observations, satellites, aircraft, and models, so the exact magnitude of volume change depends on method and period. Nevertheless, the broad transition from a thicker, older cover toward a thinner, more seasonal one is supported by independent evidence. Thin ice can grow rapidly in winter but is also more vulnerable to summer melt and deformation.

Year-to-year variations remain large. Wind can compact ice into a smaller area, export it through gateways, or disperse it into warmer waters. Snow conditions influence both winter growth and summer melt, while ocean heat transport can maintain regional losses even when local air temperatures are not exceptional. The long-term Arctic decline is primarily driven by human-caused climate change, superimposed on these natural fluctuations. A single cold season or temporary increase in extent does not reverse the multidecadal change in the system’s energy balance.

Antarctic behavior has been less monotonic. Overall extent increased slightly over much of the early satellite record, despite strong regional differences, before abrupt losses beginning in 2016. Exceptionally low coverage occurred in 2022 and 2023. In 2023 the summer minimum fell below 2 million square kilometers, and the winter maximum was also exceptionally low. Winds, ocean heat, stratification, and large-scale atmospheric patterns all contribute to Antarctic variability. The recent losses are consistent with an increasing influence of ocean warming, but the relative roles of forced change and internal variability remain less firmly constrained than for the long-term Arctic decline.

Climate models project continued Arctic ice loss as global warming increases. “Practically ice-free” Arctic conditions are commonly defined as an area or extent below 1 million square kilometers, depending on the study; the definition does not mean the disappearance of every coastal floe. Such conditions are expected first near the end of summer, not throughout the year. Their first occurrence is uncertain because emissions, model differences, and natural variability affect timing. Seasonal ice loss is not generally understood as an irreversible switch independent of temperature: cooling would favor renewed ice formation, although biological, oceanic, and societal consequences need not reverse at the same rate.

Human use, hazards, and governance

For Arctic Indigenous peoples and other coastal residents, sea ice is part of a lived landscape connecting settlements, hunting areas, fishing grounds, and culturally important places. Knowledge of currents, cracks, snow, weather, and seasonal ice development supports movement and resource use. Stable landfast ice can serve as a travel surface, while drifting ice provides access to marine animals. Conditions that appear similar in a coarse satellite image may differ greatly in practical stability. Changes in freeze-up, breakup, and ice attachment can therefore disrupt established seasonal activities even where substantial ice remains.

Sea ice also modifies coastal exposure. A broad ice cover suppresses waves and can limit their access to shore, although moving ice itself can scour shallow seabeds, push material onto land, or damage structures. Longer open-water periods permit larger waves to develop and allow autumn storms to reach coasts that were formerly protected by ice. Along shorelines containing permafrost, wave action can interact with thaw to accelerate erosion. The outcome depends on sediment, coastal orientation, water depth, storm timing, and the presence of protective landforms.

For marine transport, reduced ice does not mean uniformly easy passage. The Northern Sea Route along the Russian Arctic and the Northwest Passage through the Canadian Arctic Archipelago have different geographical constraints, infrastructure, and regulatory settings. Mobile old floes, pressure ridges, narrow channels, fog, and rapidly changing winds remain hazards. Thin first-year ice may be easier to break than old ice, but it can still exert substantial forces on a vessel. Icebreakers, reinforced hulls, ice charts, satellite observations, and trained navigation personnel are central to operations in ice-covered waters.

Offshore engineering must account for both moving ice and grounded ridge keels. Loads arise from crushing, bending, rubble accumulation, and the interaction of floes with structures. Subsea installations and pipelines may be affected by seabed gouging, particularly on shallow shelves. Operational ice management can include surveillance, forecasting, and the use of vessels to influence approaching floes, but its feasibility depends on weather, ice conditions, and the installation. Search-and-rescue and pollution-response operations face additional constraints from remoteness, darkness, low temperatures, and limited port facilities.

International regulation addresses navigation and environmental protection without treating sea ice as a separate category of sovereign territory. The International Maritime Organization’s Polar Code, which entered into force in 2017, establishes requirements for ships operating in polar waters, including provisions concerning design, equipment, operations, training, and pollution prevention. The United Nations Convention on the Law of the Sea includes specific provisions concerning pollution regulation in certain ice-covered areas within exclusive economic zones. In the Antarctic, activities also operate within the Antarctic Treaty System. Ice information services and international scientific cooperation provide practical support for navigation, environmental monitoring, and the interpretation of rapidly changing polar conditions.