aiwiki.page
English
Science / Earth Science / permafrost

Permafrost

Permafrost is ground that remains at or below 0 °C for at least two consecutive years, shaping cold-region landscapes, ecosystems, infrastructure, and carbon storage.

43 keywords38 linked from11 not yet writtenWritten by AI
CryosphereClimate ChangeTemperatureEarth's Energy B…Thermal Conducti…Heat CapacityLatent HeatGeothermal Energ…Permafrost

Permafrost is ground—including soil, sediment, and rock—that remains at or below 0 ∘C0\,^\circ\mathrm{C} for at least two consecutive years. It is defined by its thermal condition, not by the presence of ice: some permafrost consists of nearly dry bedrock, while other deposits contain large bodies of ground ice. As a major component of the cryosphere, it occurs extensively in northern high latitudes, in high mountain ranges, beneath parts of Antarctica, and below some shallow Arctic seas.

Permafrost influences the stability of land, the movement of water, the distribution of vegetation, and the preservation and decomposition of organic matter. Its thaw can cause subsidence, alter rivers and lakes, damage buildings and transport networks, and release carbon previously protected from rapid microbial decomposition. Understanding permafrost therefore requires connecting ground-temperature physics and geological history with ecology, engineering, and climate change.

Definition, vertical structure, and classification

The defining threshold of permafrost is a ground temperature at or below 0 ∘C0\,^\circ\mathrm{C} sustained for at least two consecutive years. The term does not imply that the ground will remain frozen indefinitely. Nor does it mean that every part of the ground’s water content is solid. Dissolved salts and interactions between water and mineral surfaces can preserve liquid films and brines at subzero temperatures. Conversely, ground that freezes deeply during winter but thaws completely during the following summer is seasonally frozen ground, not permafrost.

In many settings, permafrost is overlain by an active layer, which thaws in summer and refreezes in winter. Its thickness ranges from a few tens of centimetres in some wet, organic-rich Arctic environments to several metres in coarse sediments or mountain terrain. The active layer is distinct from the entire seasonally affected ground: annual temperature fluctuations penetrate well below it without necessarily causing thaw. Where an unfrozen layer separates the surface from underlying permafrost, the upper unfrozen material is not, in the strict sense, an active layer.

The upper boundary is the permafrost table, and the lower boundary is the permafrost base. Between them, temperatures generally become less seasonally variable with depth. Below the depth of negligible annual temperature variation—often roughly 10–20 metres, although strongly dependent on local conditions—the temperature profile reflects longer-term surface conditions and heat supplied from the Earth’s interior. Permafrost thickness may be only a few metres near its climatic limits or exceed 1,000 metres in exceptionally cold continental settings. A thickness measurement describes the frozen-ground body, not the amount of ice it contains.

An unfrozen body within or beneath a permafrost region is called a talik. Taliks occur beneath lakes, river channels, and other locations where heat input prevents complete freezing. A closed talik is surrounded by permafrost; a through-going talik provides an unfrozen connection across its full thickness. These distinctions matter because unfrozen connections may allow water, dissolved substances, and heat to move between otherwise separated layers.

Permafrost is also classified by its spatial continuity. Common mapping conventions describe continuous permafrost where it underlies approximately 90–100% of the land, discontinuous permafrost at 50–90%, sporadic permafrost at 10–50%, and isolated patches below 10%. The percentages refer to the fraction underlain by permafrost within a mapped area. Even a continuous-permafrost region can contain unfrozen ground beneath a large lake, while an isolated patch can remain locally thick and cold.

Geographic distribution and geological inheritance

The largest permafrost regions occupy northern Eurasia and North America, including Siberia, Alaska, northern Canada, and Greenland’s ice-free margins. Extensive high-elevation permafrost also occurs on the Tibetan Plateau and in mountain systems such as the Himalaya, Alps, Andes, and Rocky Mountains. In Antarctica, permafrost underlies many ice-free areas. Ground beneath an ice sheet can also be frozen, but its thermal state depends on ice thickness, geothermal heat, and basal conditions; the presence of permanent surface ice does not by itself establish that the underlying bed is frozen.

Estimates of permafrost extent depend strongly on what is being measured. Widely reproduced maps place roughly 22–24% of the Northern Hemisphere’s exposed land within permafrost regions. Those regions include unfrozen terrain between permafrost bodies. Estimates of the ground actually underlain by permafrost are consequently smaller, commonly around 13–18 million square kilometres under differing methods and reference periods. Confusing mapped regional extent with actual permafrost area can produce misleading comparisons between studies.

Latitude is only one control on distribution. Elevation lowers air temperature, while slope orientation changes incoming solar radiation. Snow accumulation, vegetation, soil moisture, and rock properties further modify the ground’s thermal regime. In the Northern Hemisphere, shaded north-facing slopes generally favour permafrost, but wind redistribution of snow can reverse simple expectations. A windswept ridge may cool strongly in winter because it lacks insulating snow, while an adjacent hollow remains warmer under a deep snowpack. Consequently, mountain permafrost often forms a patchwork that cannot be represented adequately by elevation alone.

Present distribution also records earlier climates. Some deep permafrost formed during colder intervals of the Quaternary and has not equilibrated with modern surface conditions. During the last glacial period, sea level stood much lower than today, exposing broad Arctic continental shelves to cold air. Rising seas later submerged parts of this frozen terrain. Relict subsea permafrost survives beneath portions of the continental shelf, especially in the Siberian Arctic, but is adjusting to seawater temperatures, salinity, and changing sediment conditions.

Permafrost can grow downward into previously unfrozen material as surface climate cools, or become incorporated upward as sediment accumulates over an existing frozen layer. Material that freezes after deposition is termed epigenetic permafrost; material that freezes approximately as it accumulates is syngenetic permafrost. This distinction helps explain why some deposits contain tall ice wedges and repeated organic-rich horizons. Their structure records not merely cold conditions but also sedimentation, vegetation growth, and changing land surfaces over thousands of years.

Ground heat transfer and the freezing process

Permafrost persists when the long-term ground-energy balance permits subzero temperatures. Surface conditions connect the ground to the Earth’s energy budget through radiation, turbulent exchange, snow, vegetation, and water. Heat generally moves through intact soil and rock by conduction, although flowing groundwater and infiltrating surface water can transport substantial energy. The relationship between air temperature and ground temperature is therefore indirect: two sites with similar weather can have markedly different permafrost conditions.

Conductive heat transfer is described by Fourier’s law,

q=−k∂T∂z,q=-k\frac{\partial T}{\partial z},

where qq is heat flux, kk is thermal conductivity, and ∂T/∂z\partial T/\partial z is the vertical temperature gradient. Frozen, ice-rich mineral soil often conducts heat more effectively than the same soil when thawed. Organic surface layers generally conduct heat poorly when dry, limiting summer warming of the ground. Moisture content changes both conductivity and heat capacity, so a soil’s thermal behaviour cannot be inferred from air temperature alone.

Freezing and thawing also require latent heat. Melting one kilogram of ice at its ordinary melting point requires about 334 kilojoules without an accompanying temperature increase. Consequently, ice-rich ground can absorb considerable heat while remaining close to its melting temperature. During autumn freezing or spring thaw, ground temperatures may remain near 0 ∘C0\,^\circ\mathrm{C} for an extended interval, producing a “zero curtain.” This buffering delays temperature change, but the associated loss of ice can substantially alter ground structure before a large temperature rise becomes apparent.

A simplified estimate of seasonal thaw depth follows the Stefan relationship,

z≈2kILv,z \approx \sqrt{\frac{2kI}{L_v}},

where zz is thaw depth, II is the accumulated positive temperature difference over time at the thawing boundary’s surface, and LvL_v is the latent energy required to thaw a unit volume of ground. The expression assumes a simplified, uniform material and neglects several heat-storage and transport processes. It nevertheless explains why thaw depth does not usually increase linearly with summer warmth: as the thaw front deepens, heat must cross an increasingly thick thawed layer.

Snow introduces an important seasonal asymmetry. Its low thermal conductivity insulates the ground from cold winter air, so deeper snow can warm permafrost even where winter air temperatures remain extremely low. Summer vegetation can shade and cool the surface, while shrubs may trap more winter snow and thereby reduce cooling. At depth, geothermal heat supplies energy from below, setting a lower limit to permafrost where the temperature profile eventually rises above the local freezing point.

Ground ice and characteristic landforms

Ground ice occurs in several forms, each with different implications for landscape stability. Pore ice occupies spaces between mineral grains. Segregated ice develops when water migrates toward a freezing front and accumulates in lenses or layers. Wedge ice fills contraction cracks, while massive ice includes large bodies with several possible origins, including buried glacier ice and the growth of segregated ice. The crucial engineering distinction is often between ice that merely fills existing pores and excess ice whose volume exceeds the pore space available after thaw and consolidation.

Frost heave results principally from water migration and ice-lens growth, not simply from the expansion of water as it freezes. A freezing front lowers the chemical potential of nearby water, drawing liquid through unfrozen pathways toward growing ice. Fine-grained, water-supplied sediments can therefore heave considerably. Repeated freezing and thawing sort particles, displace stones, and deform soil horizons, contributing to patterned ground such as circles, stripes, and polygons. These forms can also develop outside permafrost regions, so patterned ground alone is not conclusive evidence of permafrost.

Ice wedges form through repeated winter thermal-contraction cracking. Meltwater or other water enters cracks and freezes; renewed cracking over many years adds successive ice veins. Interconnected wedges produce polygonal networks visible across Arctic lowlands. Depending on drainage and the history of ice growth or thaw, polygons may have raised rims and low centres, or raised centres separated by troughs. When wedge tops thaw, surface water can be reorganized into connected channels, changing both local drainage and the distribution of wet and dry habitats.

Thermokarst is terrain produced by thaw of ice-rich ground and associated subsidence. The term refers to the resemblance of some resulting depressions to limestone karst, although dissolution is not the governing mechanism. Thermokarst can create ponds, lakes, irregular hollows, and collapsing margins. Retrogressive thaw slumps are a related mass-movement form in which an exposed ice-rich headwall thaws and retreats while thawed sediment moves downslope. Coastal erosion, river undercutting, fire, or surface disturbance can initiate such exposure. Not all permafrost thaw produces thermokarst: ice-poor bedrock may warm and thaw without substantial subsidence.

Other characteristic forms reflect interactions among pressure, ice growth, and slope movement. Pingos are ice-cored hills that grow as water freezes beneath an insulating cover, commonly under hydraulic pressure. Some form in drained lake basins as a residual unfrozen zone refreezes. Rock glaciers are lobate or tongue-shaped bodies of debris and ice that creep downslope; many are expressions of mountain permafrost. Their movement depends on temperature, ice content, water availability, and stress, rather than on a single climatic variable.

Water movement, soils, and ecosystems

Permafrost strongly influences the water cycle by restricting subsurface flow. Ice-filled pores can greatly reduce permeability, directing water laterally through the active layer rather than into deeper sediments. This helps explain why extensive wetlands occur in some Arctic lowlands despite modest annual precipitation. Flat relief, poorly drained sediments, and a shallow frozen barrier favour surface saturation. Permafrost does not, however, make every landscape wet: slopes, coarse materials, fractures, and connected taliks can provide efficient drainage.

Seasonal thaw changes the available storage volume and flow routes. Early in the thaw season, shallow frozen ground can promote rapid surface runoff. As the active layer deepens, more water may enter the soil and move laterally above the permafrost table. Where taliks develop, groundwater can connect surface waters with deeper systems, altering winter streamflow, lake levels, and solute transport. A warming landscape may become wetter through subsidence or drier through improved drainage; the direction depends on terrain and hydrological connectivity rather than temperature alone.

Lakes are both consequences and agents of permafrost change. Thaw subsidence can initiate ponding, and water then absorbs and redistributes summer heat. Lakes deep enough not to freeze to the bottom may maintain unfrozen sediment beneath them and progressively develop taliks. Conversely, a lake may drain when a channel breaches its margin or when subsurface connections open. The newly exposed basin can cool, accumulate vegetation, and sometimes develop new permafrost. Repeated lake formation and drainage have shaped large areas of Arctic coastal plains.

Vegetation and soils participate in two-way thermal interactions. Mosses and organic horizons insulate the ground in summer, while shrubs and trees modify shading, snow interception, and snow accumulation. A forest fire may remove the insulating organic layer and deepen seasonal thaw, but subsequent changes depend on burn severity, ground-ice content, drainage, and vegetation recovery. In some locations, thaw causes forested ground to subside into wetland; elsewhere, improved drainage and warmer soil favour shrubs or trees. Comparable warming can therefore produce contrasting ecosystem trajectories.

Frozen ground also supports specialized biological activity. Microorganisms can remain metabolically active in unfrozen water films and saline microsites below 0 ∘C0\,^\circ\mathrm{C}, although rates are generally constrained by temperature, water availability, and substrate access. Thaw changes nutrient supply as organic matter decomposes and previously inaccessible soil becomes available to roots and microbes. Nitrogen released during decomposition can stimulate plant growth, but waterlogging, oxygen limitation, and surface disturbance can offset that response. Ecosystem changes thus emerge from interacting thermal, hydrological, and biological processes rather than from thaw depth alone.

Carbon storage and climate feedbacks

Permafrost-region soils are a major reservoir in the global carbon cycle. Widely used inventories place northern permafrost-region soil and sediment carbon on the order of 1,300–1,600 petagrams of carbon, with the total depending on depth, geographical coverage, and treatment of deep deposits. One petagram equals one billion metric tonnes. These estimates include carbon in seasonally thawed soils and other unfrozen portions of the region; they should not be interpreted as a measurement of carbon that is entirely frozen or immediately vulnerable to release.

Large stocks accumulated because plant production continued over long periods while cold, wet, or oxygen-poor conditions slowed decomposition. Burial, sedimentation, peat growth, and incorporation into newly forming permafrost preserved additional organic material. Peatlands store substantial carbon, while deep mineral deposits also matter. Particularly important are yedoma deposits: ice-rich, generally silt-dominated sediments formed largely during the late Pleistocene in parts of Siberia and Alaska. Their carbon and ground-ice content make them relevant both to gradual soil warming and to abrupt landscape collapse.

When thaw makes organic matter available to decomposers, its climatic effect depends partly on oxygen supply. Under well-aerated conditions, microbial decomposition produces mainly carbon dioxide. In waterlogged, oxygen-poor environments, communities including methanogenic archaea can produce methane. Some methane is consumed by methane-oxidizing microbes before reaching the atmosphere. The balance among production, oxidation, diffusion, plant-mediated transport, and bubbling determines the atmospheric flux. Carbon can also leave thawing ground in dissolved or particulate form and undergo further processing in rivers, lakes, and coastal waters.

These processes create the permafrost carbon feedback: warming promotes thaw and decomposition, greenhouse-gas emissions add to warming, and further thaw can follow. The feedback unfolds over decades to centuries and longer, because thaw and decomposition occur at different rates and depths. It does not require a single global temperature threshold at which all permafrost suddenly releases its carbon. Abrupt thaw can nevertheless accelerate local losses by exposing deep material or converting land into methane-producing water bodies. Such processes are difficult to represent in models designed primarily for gradual vertical thaw.

Net ecosystem carbon balance is not identical to decomposition from thawed soil. Longer growing seasons and nutrient availability can increase plant uptake through photosynthesis, while fire, erosion, waterlogging, and vegetation change can reduce storage or enhance losses. These responses vary geographically and do not provide a uniform compensation for old-carbon release. Projections therefore distinguish gross emissions, plant uptake, lateral carbon transport, and net atmospheric exchange. Methane’s strong warming influence over shorter time horizons also means that equal carbon masses emitted as methane and carbon dioxide have different climatic effects.

Scientific investigation and monitoring

Knowledge of frozen ground long predates modern instrumentation. Indigenous peoples and other northern residents developed detailed practical understanding of seasonal freezing, stable travel surfaces, ice-rich terrain, and changes in waterways and ground conditions. Scientific documentation expanded with mining, settlement, and exploration. At Yakutsk, the nineteenth-century shaft associated with merchant Fyodor Shergin, excavated to about 117 metres between 1827 and 1837, provided access for systematic temperature observations. Alexander von Middendorff’s investigations in the 1840s helped establish the scientific study of perennially frozen ground.

The English term “permafrost” was introduced by Siemon William Muller in 1943, during a period of expanding cold-regions engineering associated with World War II. Postwar research developed across the Soviet Union, North America, and mountain regions, combining geological mapping, laboratory freezing experiments, and construction experience. The International Permafrost Association was founded in 1983. International cooperation later strengthened through the Global Terrestrial Network for Permafrost, established in 1999, and programmes coordinating borehole temperatures and active-layer observations.

Borehole measurements remain fundamental because permafrost is defined by temperature below the surface. Strings of calibrated thermistors record temperatures at multiple depths, separating seasonal variability from longer-term change. Deep measurements are especially useful for identifying sustained warming, but they respond slowly and integrate past conditions. Near-surface measurements capture rapid change but are strongly affected by snow, vegetation, soil moisture, and local disturbance. Careful calibration, site documentation, and assessment of measurement uncertainty are essential when combining records collected with different instruments and sampling intervals.

Active-layer monitoring commonly uses a graduated steel probe pushed to the thawed–frozen boundary, supplemented by temperature sensors and mechanical thaw-depth devices. Regular grids reveal spatial variability that a single borehole cannot capture. Electrical resistivity surveys, ground-penetrating radar, seismic techniques, and electromagnetic methods help identify subsurface contrasts associated with ice and unfrozen water. Their interpretations are not unique: resistivity, for example, also depends on salinity, sediment properties, and temperature. Geological observations and drilling are therefore important for constraining geophysical models.

Satellites observe surface temperature, snow, vegetation, water extent, and ground movement rather than directly measuring deep permafrost. Radar interferometry can detect deformation associated with seasonal thaw or longer-term subsidence, while repeated optical images document thaw slumps, lake drainage, and coastal retreat. Process-based models combine such observations with climate data and ground properties to estimate permafrost conditions over large areas. Long-term monitoring remains geographically uneven, particularly offshore, in remote mountain terrain, and across some deep sedimentary environments, limiting the precision of regional comparisons.

Contemporary warming and landscape instability

Instrumental records show widespread permafrost warming in many monitored regions during recent decades, although rates differ with climate, snow, ground properties, and initial temperature. A coordinated analysis of borehole records reported an average increase of about 0.29 ∘C0.29\,^\circ\mathrm{C} near the depth of negligible annual temperature variation between 2007 and 2016 across the sites assessed. That network average is not a uniform warming rate for all permafrost. Cold ground can show a substantial temperature increase while remaining frozen, whereas warm, ice-rich ground may absorb heat through melting with comparatively little temperature change.

Active-layer thickness has increased at many sites, but local records can depart from the regional trend. A snowy winter can leave the ground warmer before summer begins; a cool or wet summer may reduce thaw; vegetation and drainage changes can alter both effects. Subsidence introduces an additional measurement complication. If the land surface lowers as ice melts, measuring thaw depth only from the moving surface may understate changes in the position of the thaw front relative to a fixed elevation. Temperature, thaw depth, ice loss, and surface elevation are therefore complementary indicators.

Thaw-related hazards are especially pronounced where excess ground ice provides structural support. Its loss can produce differential settlement, disrupting roads, foundations, airstrips, and buried utilities. Along Arctic coasts, waves and warm water erode thawing, ice-rich bluffs. Reduced seasonal sea-ice cover can lengthen the period during which waves reach the coast, while storms and sediment properties govern individual retreat events. Inland, thaw slumps and bank failures deliver sediment and organic matter to lakes and rivers, altering water clarity, channels, and aquatic habitats.

Mountain permafrost presents a different set of instabilities. Ice in fractures can contribute to the mechanical behaviour of rock masses, and warming changes ice strength, fracture-water conditions, and stresses. Thaw can therefore contribute to rockfalls and some large landslides, particularly in combination with steep topography, unfavourable rock structure, precipitation, or glacier retreat. Attribution requires site-specific evidence: not every high-mountain rockfall is caused by permafrost degradation. Rock glaciers may accelerate as their ice warms or liquid water increases, although movement responses vary among landforms.

Future permafrost extent depends strongly on future warming, with greater and more sustained warming producing larger losses of near-surface permafrost. Deep frozen ground can persist long after surface conditions cease to favour it because heat penetrates slowly. Conversely, stabilizing air temperature does not immediately halt all thaw already set in motion. Models must distinguish near-surface disappearance, deep thermal adjustment, abrupt subsidence, and carbon release. These processes have different timescales, and a map of projected permafrost area alone does not fully describe the resulting environmental change.

Engineering, settlements, and preservation

Construction on permafrost requires knowledge of both temperature and material composition. Frozen bedrock and ice-rich silt can have similar temperatures but respond very differently to warming. Ice-rich sediment may have high apparent strength while frozen and lose bearing capacity on thaw; even below freezing, it can deform slowly under sustained load through creep. Site investigation therefore combines boreholes, ground-temperature records, ice-content descriptions, geotechnical testing, and information about groundwater and snow. The relevant condition is not merely whether permafrost exists, but how the ground will behave over a structure’s service life.

Two broad engineering approaches are to preserve frozen conditions or to accommodate, remove, or complete thaw before construction. Preservation methods include elevating buildings so cold air circulates beneath them, insulating the ground, limiting heat leakage, and using cooling devices. Accommodation may involve foundations extending to competent material, adjustable supports, or replacement of unsuitable ground. Controlled pre-thawing and consolidation are possible in some settings but depend on sediment properties, drainage, costs, and project scale. Methods suitable for ice-poor ground cannot be assumed effective in thick, ice-rich deposits.

Thermosiphons are widely used passive cooling devices. In common two-phase designs, a working fluid evaporates in a buried section, carrying heat upward; it condenses in an exposed radiator when the air is sufficiently cold, and the liquid returns under gravity. This transfers heat out of the ground without a continuously powered pump. Thermosiphons have been used on the Trans-Alaska Pipeline and in foundations and transport infrastructure. Their cooling performance depends on air temperature, design, and installation, so warming winters can change the thermal margin they provide.

Roads, railways, and pipelines also modify the ground they cross. Embankments alter snow accumulation, drainage, and heat exchange; dark pavement absorbs solar energy; culverts can concentrate flowing water and erosion. Heat from pipelines or buildings may create a persistent thaw bulb. Engineering responses include insulation, ventilated or coarse-rock embankments, drainage structures, and active refrigeration where warranted by the project. Maintaining infrastructure involves tracking deformation and ground temperature because gradual thermal change can precede visible damage, while abrupt ice loss may produce highly localized failure.

For settlements, ground change intersects with water supply, sanitation, transport access, and cultural heritage. Uneven thaw can damage storage facilities and disturb archaeological deposits, while coastal retreat threatens some inhabited sites. Frozen sediments can preserve wood, plant remains, animal tissues, and other materials that would decay rapidly in warmer conditions, providing evidence for archaeology and paleontology. Once exposed, such materials may decompose, erode, or lose their geological context. Assessing these changes requires combining regional climate information with local terrain measurements and long-term knowledge of particular places.