{
  "id": "earth-mantle",
  "title": "Earth’s Mantle",
  "lang": "en",
  "source_lang": "en",
  "summary": "The mantle is Earth’s thick silicate interior, whose slow circulation drives plate tectonics, generates magma, and transfers heat and materials between planetary reservoirs.",
  "category": "science",
  "subcategory": "earth-science",
  "type": "concept",
  "aliases": [
    "Terrestrial mantle"
  ],
  "generated_at": "2026-11-02",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/earth-mantle/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/earth-mantle/",
      "json": "https://aiwiki.page/zh/earth-mantle.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "The mantle is the predominantly solid, rocky layer of Earth between its crust and metallic core. It begins beneath the crust at depths ranging from a few kilometres beneath parts of the ocean floor to more than 70 kilometres beneath major mountain belts, and extends to approximately 2,890 kilometres below the surface. It accounts for about 84% of Earth’s volume and roughly 67% of its mass. Although mantle rock can fracture during earthquakes and transmit elastic waves, it also deforms slowly over millions of years, allowing the interior to circulate.\n\nThis circulation connects the mantle to plate tectonics, volcanism, mountain building, and the long-term exchange of water and carbon between Earth’s surface and interior. The mantle is not a global ocean of molten rock: melting is generally restricted to particular environments, while most of its volume remains crystalline. Its composition, mineral structure, and mechanical behavior change with depth, producing distinct upper-mantle, transition-zone, and lower-mantle regions.\n\nKnowledge of the mantle comes from earthquake waves, rocks brought to the surface, high-pressure experiments, geochemical measurements, and numerical models. These approaches reveal a differentiated and evolving interior, but leave important questions about its circulation, chemical reservoirs, volatile inventory, and interaction with the core."
    },
    {
      "heading": "Boundaries and internal divisions",
      "text": "The mantle’s upper boundary is the Mohorovičić discontinuity, usually called the Moho. It is recognized primarily by an increase in seismic-wave speed between crustal rocks and the denser rocks below. Beneath typical oceanic crust, it lies about 6–7 kilometres below the seafloor, whereas beneath continents it commonly occurs at depths of 30–50 kilometres. It can be substantially deeper beneath thickened continental crust. The Moho is a compositional boundary, not the base of a tectonic plate: the rigid outer shell includes both the crust and some underlying mantle.\n\nThe upper mantle extends from the Moho to approximately 660 kilometres depth. Within it, the mantle transition zone occupies the interval between major seismic discontinuities near 410 and 660 kilometres. These boundaries arise mainly from pressure-induced changes in mineral structure rather than abrupt changes in bulk chemical composition. A less prominent discontinuity near 520 kilometres is observed in some regions. Their actual depths vary because temperature, composition, and water content affect mineral stability. The lower mantle extends from about 660 kilometres to the core–mantle boundary.\n\nA separate classification describes mechanical behavior. The lithosphere consists of the crust and a relatively strong, cool uppermost mantle. It forms the tectonic plates and ranges in thickness from very thin near spreading ridges to more than 200 kilometres beneath some ancient continental interiors. Beneath it lies the asthenosphere, a weaker part of the upper mantle that accommodates plate motion. Its boundaries are less sharply defined than the Moho, and its seismic expression varies geographically. Low seismic velocities can reflect high temperature, small amounts of melt, water-related effects, and the behavior of mineral grains; they do not imply that the whole asthenosphere is liquid.\n\nThe mantle’s lower boundary is much sharper. At approximately 2,890 kilometres depth, solid silicate rock meets the liquid, iron-rich outer core. Density increases abruptly, and shear waves cannot propagate through the liquid core. Pressure at this boundary is about 136 gigapascals, more than a million times atmospheric pressure. The lowermost few hundred kilometres of mantle are often designated the D″ region, a name inherited from an early seismic classification. This region is unusually heterogeneous and contains a thermal boundary layer, variable mineral structures, and localized zones of exceptionally low seismic velocity. Its upper boundary is not a single globally uniform surface."
    },
    {
      "heading": "Chemical composition and mineral structure",
      "text": "The mantle consists mainly of silicate minerals containing oxygen, magnesium, silicon, and iron, with smaller quantities of calcium, aluminium, and other elements. It is chemically distinct from both the silica-rich continental crust and the metallic core. A useful reference composition is a fertile upper-mantle rock broadly resembling peridotite, although no single sampled rock represents the entire mantle. Mantle regions differ because melting, melt migration, recycling, and early planetary differentiation redistribute chemical components.\n\nIn the shallow upper mantle, peridotite is dominated by olivine, accompanied by orthopyroxene, clinopyroxene, and an aluminium-bearing mineral. The latter is commonly plagioclase at low pressure, spinel at intermediate pressure, and garnet at greater depth. These assemblages depend on temperature and bulk composition as well as pressure, so their boundaries cannot be assigned one universal depth. Extraction of basaltic melt leaves residues depleted in relatively fusible components. Some mantle beneath old continents is strongly depleted in this way, making it both chemically distinctive and relatively buoyant.\n\nIncreasing pressure changes how atoms are packed into crystals. Near 410 kilometres, olivine transforms into wadsleyite; near 520 kilometres, wadsleyite transforms into ringwoodite. These minerals have the same ideal magnesium–iron silicate composition as olivine but denser crystal structures. At the base of the transition zone, ringwoodite breaks down into a magnesium silicate phase and a magnesium–iron oxide. Other upper-mantle minerals also transform, with pyroxene components increasingly incorporated into garnet-related structures before entering lower-mantle assemblages. These phase transitions explain much of the observed increase in seismic speed with depth.\n\nThe lower mantle is dominated by bridgmanite, a magnesium–iron silicate with a perovskite-type crystal structure. It is accompanied mainly by ferropericlase, a magnesium–iron oxide, and a calcium silicate perovskite phase. Bridgmanite is generally regarded as Earth’s most abundant mineral by volume. Near the mantle’s base, suitable pressure and temperature conditions stabilize post-perovskite, a different structure of magnesium silicate identified experimentally in 2004. Its elastic and deformation properties may explain some features of D″. Mineral proportions, iron oxidation state, and the distribution of elements among phases all influence density and wave speeds, making mineral physics essential to interpreting observations of the deep interior."
    },
    {
      "heading": "Evidence and methods of investigation",
      "text": "Most of the mantle is inaccessible to direct sampling. Scientific drilling has recovered rocks formed in the mantle where tectonic processes have exposed them near the seafloor, but no borehole has penetrated an intact, typical crustal section and continued far into the underlying mantle. Consequently, the principal global evidence comes from seismology. Earthquakes generate seismic waves whose travel times, reflections, conversions, amplitudes, and polarizations record the properties of the materials through which they pass.\n\nCompressional waves, or P waves, travel through both solids and liquids; shear waves, or S waves, require a material that supports shear stress. Their propagation through almost the entire mantle demonstrates its predominantly solid character. In an isotropic elastic material, the wave speeds are related to density and elastic moduli by\n\n\\[V_P=\\sqrt{\\frac{K+\\frac{4}{3}\\mu}{\\rho}}, \\qquad V_S=\\sqrt{\\frac{\\mu}{\\rho}},\\]\n\nwhere \\(K\\) is the bulk modulus, \\(\\mu\\) the shear modulus, and \\(\\rho\\) the density. Earth’s free oscillations after large earthquakes provide additional constraints on its radial structure. Reference models, including the Preliminary Reference Earth Model published in 1981, combine multiple observations to describe average density, elastic properties, and attenuation with depth.\n\nSeismic tomography maps departures from such averages. It commonly identifies fast anomalies associated with cold descending plates and slow anomalies associated with hotter material, but velocity is not a unique thermometer. Composition, crystal orientation, mineral transformations, and partial melt can also alter wave speeds. Resolution depends on earthquake locations, station coverage, wave types, and inversion methods. An apparent gap or connection between structures may therefore reflect limited resolving power rather than a genuine geological boundary.\n\nDirect geological evidence includes mantle fragments carried upward in volcanic rocks, called xenoliths, and tectonically exposed bodies of mantle rock. Diamonds can preserve mineral inclusions formed at great depth, including phases associated with the transition zone and lower mantle. Laboratory experiments using large-volume presses and diamond-anvil cells reproduce high pressures and temperatures, while X-ray diffraction identifies resulting crystal structures. Electrical conductivity measurements, gravity observations, and surface deformation supply complementary constraints. Each method samples different properties and scales, so reliable interpretations usually require agreement among several independent lines of evidence."
    },
    {
      "heading": "Temperature, heat sources, and physical behavior",
      "text": "Mantle temperature rises strongly from the surface downward, but the temperature gradient is not uniform. Across the lithosphere, heat is transferred mainly by conduction, producing a relatively steep gradient. Through much of the convecting mantle, rising and sinking material approximately follows an adiabatic temperature path: its temperature changes as it expands or compresses, even without exchanging much heat with its surroundings. Another steep thermal gradient occurs near the core–mantle boundary, where heat enters the mantle from the core.\n\nUpper-mantle temperatures are often compared using mantle potential temperature, the temperature a parcel would have if brought adiabatically to a reference pressure without melting. Estimates for ordinary mid-ocean-ridge source mantle commonly fall around 1,300–1,400 °C, although values depend on the model and on corrections for melting. Temperatures at the base of the mantle are much less certain and are generally estimated to be several thousand kelvin, commonly in the broad range of about 3,500–4,500 K. Such estimates depend on experimental melting relations, mineral properties, and models of heat transfer.\n\nEarth releases approximately 47 terawatts of heat through its surface, a global estimate with an uncertainty of several terawatts. This is the total heat loss of the planet, not a direct measurement of mantle heat production. Its sources include heat retained from accretion and differentiation, ongoing cooling of the mantle and core, and radioactive decay, especially of uranium, thorium, and potassium isotopes. These heat-producing elements are unevenly distributed and enriched in continental crust relative to the mantle. Their mantle abundance remains a major uncertainty in models of Earth’s thermal evolution.\n\nThe mantle remains solid because pressure raises mineral melting temperatures and because most mantle conditions lie below the relevant solidus, the temperature at which melting begins. Nevertheless, high temperature enables crystals to deform through defects and grain boundaries. Over geological time, this produces creep rather than purely elastic behavior. Effective viscosity depends on temperature, pressure, stress, grain size, mineral structure, and water content. Models of mantle flow and postglacial adjustment commonly involve values of roughly \\(10^{19}\\)–\\(10^{23}\\) pascal seconds across different regions and depths, rather than one universal viscosity. On the short timescale of a seismic wave, the same material responds mainly as an elastic solid."
    },
    {
      "heading": "Convection and the motion of tectonic plates",
      "text": "Mantle convection is the slow movement of mantle material driven by buoyancy differences and coupled to the motion of tectonic plates. Hot material generally has lower density than cooler material of the same composition, while cooling near the surface creates dense lithosphere that can sink. Chemical differences and mineral transitions can reinforce or oppose thermal buoyancy. Flow is extremely slow by everyday standards, commonly comparable in scale to plate velocities of centimetres per year, yet it transports enormous quantities of heat over geological time.\n\nA basic measure of the tendency to convect is the Rayleigh number. For a simplified layer heated from below,\n\n\\[Ra=\\frac{\\rho g\\alpha\\Delta T\\,d^3}{\\eta\\kappa},\\]\n\nwhere \\(g\\) is gravitational acceleration, \\(\\alpha\\) thermal expansivity, \\(\\Delta T\\) the temperature contrast, \\(d\\) layer thickness, \\(\\eta\\) dynamic viscosity, and \\(\\kappa\\) thermal diffusivity. Earth’s mantle has conditions strongly favorable to convection. However, its internal heating, spherical geometry, variable viscosity, moving plates, and phase transitions make it much more complex than the idealized layer represented by this expression.\n\nTectonic plates are part of the convecting system rather than passive rafts carried by an independent circulation beneath them. At subduction zones, cold oceanic lithosphere descends into the mantle. Its negative buoyancy provides a major driving contribution to plate motion. Gravitational forces associated with elevated oceanic ridges also contribute, while mantle traction and resistance at plate boundaries can assist or oppose movement. Ascending flow beneath spreading ridges is often largely a response to plate separation, rather than the upward limb of a fixed, symmetrical convection cell.\n\nThe 660-kilometre boundary influences circulation because mineral transformations and viscosity contrasts can impede descending or rising material. Tomographic images indicate that some slabs flatten within or near the transition zone, whereas others penetrate into the lower mantle. These observations argue against both a universally isolated upper mantle and an everywhere-unobstructed whole-mantle circulation. Slab geometry, descent history, phase-boundary behavior, and surrounding flow affect the outcome. Mixing is likewise incomplete: convection stretches and folds chemical heterogeneities into complex patterns without necessarily homogenizing the entire mantle. Surface motions, sinking slabs, broad return flow, and localized upwellings together form an evolving three-dimensional system."
    },
    {
      "heading": "Partial melting and the production of magma",
      "text": "Most mantle-derived magmas form through partial melting, in which only a fraction of a rock becomes liquid. Different minerals and chemical components enter the melt at different rates, so the liquid generally has a different composition from the source. Melt fractions vary considerably among tectonic settings. Extraction of melt leaves a more refractory residue, while migrating liquids can react with surrounding rock and modify its mineralogy and chemical composition. The result is a continuing chemical differentiation of the mantle and crust.\n\nAt mid-ocean ridges, mantle rises as plates separate. Pressure falls more rapidly than the rising material cools, allowing it to cross its solidus and undergo decompression melting. The resulting liquids aggregate, move through porous rock and channels, and eventually feed crustal magma systems. They produce predominantly basalt and related intrusive rocks that build new oceanic crust. Melting usually occurs over a depth interval rather than at one sharply defined level. Its extent depends on source temperature, composition, volatile content, and the rate and geometry of upwelling.\n\nSubduction zones create a different melting environment. Descending oceanic plates carry hydrated minerals, altered crust, and sediments into the interior. As pressure and temperature increase, reactions release water-rich fluids and, under some conditions, melts. These materials enter the overlying mantle wedge and lower its melting temperature. The resulting magmas contribute to volcanic arcs such as those around much of the Pacific. The sinking slab itself does not generally melt wholesale; heating, dehydration, fluid transport, mantle melting, and later crustal processing must be distinguished.\n\nSome intraplate volcanic regions are associated with mantle plumes, proposed buoyant upwellings that may rise from deep thermal boundary layers. Hawaii is a classic example used to develop plume models, supported by its volcanic age progression and other geological and geophysical evidence. Yet intraplate volcanism is not uniquely diagnostic of a deep plume: lithospheric extension, small-scale convection, source composition, and shallow flow can also generate or focus melting. The depth, temperature, and geometry of individual sources remain subjects of investigation.\n\nOnce formed, magma can separate from solid rock because of buoyancy and pressure gradients. During ascent and storage it may crystallize, mix with other melts, or assimilate crust. Surface lava therefore records both mantle conditions and later processing. Reconstructing the source requires mineral chemistry, trace-element patterns, and isotope measurements rather than treating erupted rock as an unchanged sample of the deep interior."
    },
    {
      "heading": "Water, carbon, and chemical reservoirs",
      "text": "The mantle stores volatile elements in forms very different from surface oceans or the atmosphere. Water is commonly incorporated as hydrogen-bearing defects in mineral crystals, not as free liquid. Even minerals described as nominally anhydrous can contain small but geologically important amounts of hydrogen. Wadsleyite and ringwoodite can incorporate considerably more water under suitable conditions than many minerals stable above or below the transition zone. This gives the transition zone a substantial potential storage capacity, but capacity is not equivalent to its actual global water inventory.\n\nA natural ringwoodite inclusion reported in 2014 provided direct evidence for locally water-rich material in the transition zone. Such observations demonstrate that significant water can occur at depth, but they do not show that the entire transition zone has the same hydration state. Electrical conductivity, seismic properties, and laboratory measurements provide broader constraints, each with interpretive ambiguities. Water also alters mantle behavior: it can lower melting temperatures, influence deformation, and affect electrical conductivity. Consequently, small concentrations can have effects disproportionate to their mass.\n\nCarbon occurs in several mantle forms, including carbonates, diamond, and carbon-bearing fluids or melts. Its stable form depends strongly on pressure, temperature, and oxidation state. Oxidized conditions favor carbonate-bearing assemblages, while more reducing conditions can stabilize diamond or other reduced forms. Carbon-rich melts may form at lower temperatures than ordinary silicate melts and can transport incompatible elements efficiently. Subduction introduces carbon in sediments and altered oceanic crust, whereas volcanic and diffuse degassing return some carbon to the surface, linking the mantle to the long-term carbon cycle.\n\nMantle geochemistry distinguishes reservoirs through element abundances and isotope ratios. Elements that preferentially enter melt are called incompatible; repeated melt extraction leaves the residual mantle depleted in them. Mid-ocean-ridge basalts commonly sample such depleted sources, while ocean-island basalts display a wider range of isotopic signatures. Ratios involving strontium, neodymium, lead, hafnium, helium, and other elements reveal histories of differentiation, radioactive decay, and recycling.\n\nThese reservoirs are not necessarily large, sharply bounded chambers. They may include dispersed recycled oceanic crust, ancient depleted residues, enriched veins, and domains that have experienced limited mixing. High helium-3 relative to helium-4 in some volcanic materials is consistent with contributions from sources that have preserved a comparatively primordial noble-gas signature. It does not, by itself, identify a unique source depth or prove that the material has remained completely isolated since Earth formed."
    },
    {
      "heading": "Formation and geological evolution",
      "text": "The mantle originated during Earth’s accretion approximately 4.54 billion years ago. Collisions, compression, radioactive heating, and the release of gravitational energy produced extensive melting during early growth. As metallic material separated from silicates and accumulated in the core, the residual silicate portion became the precursor of the mantle and crust. Core formation occurred under changing pressure, temperature, and chemical conditions, leaving evidence in the distribution of elements that preferentially partition into metal or silicate.\n\nLarge impacts, including the event associated with the Moon’s origin, probably generated extensive magma oceans. Their crystallization established early mineral and chemical stratification. Depending on crystallization sequence and the efficiency of mixing, dense late-forming material could have sunk, while lighter material rose. Such overturn could redistribute heat-producing elements and influence the formation of long-lived deep reservoirs. The precise number, depth, and duration of these melting episodes remain model-dependent because the earliest terrestrial rock record is extremely limited.\n\nOver billions of years the mantle has cooled, although its temperature history cannot be reconstructed from one simple trend. Ancient komatiites, magnesium-rich volcanic rocks especially common in the Archean geological record, provide evidence for unusually high melting temperatures in some early mantle environments. Their interpretation must account for source composition and water as well as temperature. Higher ancient heat production and different lithospheric conditions may have altered convection and melting, but modern plate behavior cannot simply be projected unchanged into Earth’s earliest history.\n\nThe onset and development of sustained plate tectonics remain debated. Geological indicators record different parts of the process: localized subduction-like activity, widespread recycling, continental assembly, and a global network of long-lived plate boundaries need not have begun simultaneously. Once established, subduction repeatedly returned altered oceanic lithosphere and sediments to the mantle, while melting generated new crust. Continental growth therefore changed both the surface and the residual interior by extracting incompatible elements and concentrating some heat-producing elements in the crust.\n\nAncient continental roots preserve another aspect of mantle evolution. Beneath many cratons, thick, cool, chemically depleted mantle has survived for billions of years. Its buoyancy and strength help explain the persistence of these continental interiors, although later heating and melt infiltration can modify or destroy parts of their roots. At greater depths, recycled material and possibly early differentiation products may preserve histories much older than individual surface plates."
    },
    {
      "heading": "Development of mantle science and unresolved structure",
      "text": "The mantle became a distinct object of investigation through the development of earthquake seismology. In 1909, Andrija Mohorovičić interpreted earthquake arrival times as evidence for a boundary separating the crust from faster material below. Work by Beno Gutenberg in the early twentieth century helped establish the depth of the core–mantle boundary. Subsequent observations refined the mantle’s radial structure and identified internal discontinuities. These discoveries transformed a largely inferred planetary interior into a structure that could be tested through wave propagation.\n\nThe connection between mantle motion and surface geology developed more slowly. Arthur Holmes argued in the late 1920s and early 1930s that internal convection could provide a mechanism for continental movement. The emergence of plate tectonics in the 1960s supplied a coherent framework linking seafloor spreading, earthquakes, ocean trenches, and the recycling of lithosphere. High-pressure mineral experiments then connected seismic discontinuities to specific phase transformations. From the late twentieth century onward, seismic tomography and increasingly powerful numerical models revealed lateral complexity that one-dimensional Earth models could not describe.\n\nOne major unresolved issue concerns the large low-shear-velocity provinces beneath Africa and the Pacific. These broad regions occupy substantial parts of the lowermost mantle and are characterized by slower-than-average shear-wave speeds. Their origin may involve elevated temperature, distinctive composition, or both. Some models treat them as accumulations of chemically dense material with considerable longevity; others emphasize evolving structures shaped by the distribution of subducted slabs. Their boundaries, density contrasts, and relationships to surface volcanism remain actively investigated.\n\nMuch smaller ultralow-velocity zones occur locally near the core–mantle boundary. Their strong reductions in seismic speed can be explained in different settings by partial melt, iron-rich compositions, or combinations of mineralogical and chemical effects. No single explanation has been established for all such zones. Understanding them matters because the boundary regulates heat loss from the core, which helps sustain the fluid motions responsible for Earth’s magnetic field.\n\nOther open problems include the amount of water and heat-producing elements stored at depth, the degree of chemical exchange between upper and lower mantle, and the longevity of ancient reservoirs. Progress depends on jointly interpreting observations rather than assigning every seismic anomaly to temperature or every chemical signature to a fixed layer. Dense seismic networks, improved high-pressure experiments, more precise isotope measurements, and physically constrained models continue to test competing explanations of the mantle’s hidden structure."
    }
  ],
  "links": [
    "earth",
    "plate-tectonics",
    "lithosphere",
    "asthenosphere",
    "oxygen",
    "silicon",
    "pressure",
    "phase-transition",
    "density",
    "diamond",
    "adiabatic-process",
    "radioactivity",
    "viscosity",
    "mantle-convection",
    "subduction",
    "water",
    "carbon",
    "oxidation-state",
    "carbon-cycle",
    "isotope",
    "magnetic-field"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}