Temperatures in Earth’s mantle range from roughly 1,000°C just beneath the crust to over 3,500°C at the boundary with the outer core, spanning nearly the entire 2,900-kilometer thickness of solid rock between the surface and the liquid iron below. That gradient is not smooth or uniform. Mantle temperature at any given depth depends on location, the presence of rising plumes or sinking tectonic plates, and the mineral phase transitions that reorganize rock structure under extreme pressure. The heat that sustains these temperatures comes from two broad sources, radioactive decay and leftover energy from Earth’s formation, and the balance between them remains an active area of research.
The Upper Mantle and the Concept of Potential Temperature
Right at the boundary between the crust and the mantle, temperatures vary enormously depending on whether you are beneath an ocean or a continent. Under mid-ocean ridges, where the crust is thin, mantle rock at shallow depths is hot enough to partially melt. Under old, thick continental interiors, temperatures at the same depth can be several hundred degrees cooler. To make meaningful comparisons across locations, geophysicists use a concept called mantle potential temperature. This is the temperature mantle rock would have if you brought it to the surface without letting it lose heat, essentially correcting for the way pressure heats rock at depth.
Estimates of potential temperature beneath mid-ocean ridges cluster around 1,400–1,475°C, though the exact number depends on the method used. Olivine crystals trapped in lavas provide one line of evidence, and that approach gives a range of roughly 1,453–1,475°C for ordinary ridge settings.1Geochemistry, Geophysics, Geosystems. Mantle potential temperatures at Hawaii, Iceland, and the mid‐ocean ridge system, as inferred from olivine phenocrysts: Evidence for thermally driven mantle plumes A separate analysis based on the chemistry of primitive magmas and the behavior of partially melting peridotite gives a somewhat lower figure of about 1,430°C.2Plates, plumes and paradigms. Primary magmas at mid-ocean ridges, “hotspots,” and other intraplate settings: Constraints on mantle potential temperature These two numbers are not in dramatic conflict; they reflect differences in the mineral systems being studied and the assumptions fed into the calculations. The honest range for mid-ocean ridge potential temperature is somewhere around 1,430–1,475°C.
A revised adiabatic temperature profile, which traces how temperature rises with depth along a path of convecting rock, puts the mantle temperature at 50 km depth at about 1,646 K (roughly 1,373°C).3Journal of Geophysical Research: Solid Earth. A Revised Adiabatic Temperature Profile for the Mantle From there, the temperature rises steadily because the weight of overlying rock compresses the material below, and compression generates heat. The rate of that increase in the upper mantle is roughly 0.4–0.5°C per kilometer of depth.4Physics of the Earth and Planetary Interiors. Adiabatic temperature profile in the mantle
The Transition Zone and Its Mineral Boundaries
Between about 410 and 660 km depth, the mantle passes through the transition zone, a layer where minerals undergo pressure-driven structural changes. At the 410-km boundary, the mineral olivine transforms into a denser crystal structure. This phase change is sensitive to temperature, so the actual depth at which it occurs shifts depending on how hot or cool the rock is locally. Observations show the 410-km boundary varies in depth from about 405 to 440 km across the globe, consistent with a temperature range of roughly 360 K in dry mantle rock.5Earth and Planetary Science Letters. Global variability of the composition and temperature at the 410-km discontinuity from receiver function analysis of dense arrays In other words, the boundary is deeper where the rock is warmer and shallower where it is cooler, giving geophysicists a natural thermometer embedded in the Earth.
A similar boundary at about 520 km depth has been mapped beneath the United States using vibrations from ambient seismic noise. It appears deeper in the warmer western U.S. and shallower in the cooler east, consistent with the temperature-dependent behavior seen at the 410.6Geophysical Research Letters. Internal Structure of the Mantle Transition Zone Beneath the Contiguous U.S.: Insights From the 520‐km Discontinuity Revealed by Ambient Noise Correlations At the base of the transition zone, around 660 km, another phase transformation occurs. This one is more complex. The dominant mineral reaction shifts character as temperature crosses roughly 1,920–2,020 K, which affects how the boundary interacts with convection currents passing through it.7Earth and Planetary Science Letters. Reconciling Pacific 410 and 660 km discontinuity topography, transition zone shear velocity patterns, and mantle phase transitions
The revised adiabatic profile puts the temperature just above the 660-km discontinuity at about 1,994 K (roughly 1,721°C) and just below it at about 1,960 K, a slight drop that reflects the energy absorbed by the phase transition.3Journal of Geophysical Research: Solid Earth. A Revised Adiabatic Temperature Profile for the Mantle The interplay between these phase boundaries and mantle convection matters because how easily material crosses the 660-km boundary affects how heat is transported from the deep Earth toward the surface.
The Lower Mantle and the Core-Mantle Boundary
Below 660 km, the lower mantle stretches for more than 2,200 km. Temperature continues to climb, but more slowly, at a gradient of roughly 0.3°C per kilometer.4Physics of the Earth and Planetary Interiors. Adiabatic temperature profile in the mantle At 2,800 km depth, close to the base of the mantle, the adiabatic profile gives a temperature of about 2,587 K (roughly 2,314°C).3Journal of Geophysical Research: Solid Earth. A Revised Adiabatic Temperature Profile for the Mantle The actual temperature gradient through the lower mantle is not perfectly adiabatic, though. Convection models consistently show that the lower mantle runs slightly cooler than a purely adiabatic path would predict, except in the bottommost few hundred kilometers, where temperatures spike upward again.8Earth and Planetary Science Letters. How flat is the lower-mantle temperature gradient?
That spike happens in a region called D″ (pronounced “D double-prime”), a layer roughly 200–300 km thick sitting directly above the liquid outer core. Here, the cold rock of the deep mantle meets the intensely hot iron alloy of the core. The temperature jump through D″ is dramatic. Laboratory measurements of the melting behavior of lower-mantle minerals suggest temperatures in this layer range from roughly 3,300 to 4,300 K, and the temperature at the outermost core is at least 4,300 K. That means the temperature rises by around 1,000 K across this relatively thin boundary layer, making D″ one of the most extreme thermal gradients on the planet. Some minerals partially melt in this zone, giving it unusual seismic properties.
Where the Heat Comes From
Earth’s total surface heat flow is about 46 terawatts.9Nature Geoscience. Core–mantle boundary heat flow That is the rate at which heat escapes through the surface, and it is maintained by two major internal sources working in tandem.
The first is radioactive decay. Unstable isotopes of uranium, thorium, and potassium are scattered through the mantle and crust, and their slow breakdown releases energy as heat.10Reviews of Geophysics. Geoneutrinos and the radioactive power of the Earth Measurements of geoneutrinos, the ghostly particles emitted during radioactive beta decay inside the Earth, have allowed physicists to directly estimate how much heat these elements produce. Combined analyses from two major geoneutrino detectors favor a present-day total radiogenic power of roughly 20 terawatts.11Earth and Planetary Science Letters. Quantifying Earth’s radiogenic heat budget That accounts for close to half of the planet’s total heat output, consistent with earlier geoneutrino-based estimates.12Nature Geoscience. Partial radiogenic heat model for Earth revealed by geoneutrino measurements
The second source is primordial heat, energy left over from when Earth formed. The violent collisions of the accretion process and the segregation of iron into the core released enormous amounts of energy, much of which the planet has not yet radiated away. The fact that radiogenic heating accounts for only about half of the surface heat budget means the other half is primordial heat still leaking out after 4.5 billion years.12Nature Geoscience. Partial radiogenic heat model for Earth revealed by geoneutrino measurements These two sources are not neatly separated by depth. Radioactive elements are concentrated in the crust and upper mantle, while primordial heat is stored throughout the deep interior and core. The core itself contributes little radiogenic heat; uranium and thorium ended up almost entirely in the silicate mantle during core formation, supplying negligible power to the core’s own energy budget.13Earth and Planetary Science Letters. Earth’s chondritic Th/U: Negligible fractionation during accretion, core formation, and crust–mantle differentiation
How Heat Moves Through the Mantle
Most of the mantle’s heat is transported by convection, the slow overturn of solid rock that behaves like an extremely viscous fluid on geological timescales. Hot rock near the core is buoyant and rises; cool rock near the surface is dense and sinks. This circulation is what drives plate tectonics. Subducting slabs of oceanic lithosphere carry cold material downward, while hot mantle plumes carry heat upward from the deep interior.
The relationship between temperature and viscosity makes convection self-regulating but also complicated. Hotter rock flows more easily, while cooler rock resists flow. When viscosity depends strongly on temperature, the convection pattern can change character near the transition zone, where latent heat released by phase transitions interacts with the viscosity structure. Modeling shows this feedback can produce very high local temperatures between the 410 and 660 km boundaries.14Physics of the Earth and Planetary Interiors. The effects of temperature-dependent viscosity on mantle convection with the two major phase transitions
At the core-mantle boundary, heat flux is far from uniform. Where cold subducted slabs reach the base of the mantle, they pull heat out of the core efficiently, creating regions of high heat flux.15Geophysical Journal International. Changes in core–mantle boundary heat flux patterns throughout the supercontinent cycle In other areas, particularly beneath large structures called large low-shear-velocity provinces (LLSVPs), heat flux from the core can actually become locally negative, meaning heat flows back into the core from the surrounding hot mantle material.16Nature Geoscience. Negative core–mantle boundary heat flux beneath low-shear-wave-velocity provinces This patchwork of heating and cooling at the bottom of the mantle influences how the core’s magnetic dynamo behaves.
Hot Spots and Cold Slabs
Not all parts of the mantle at the same depth are the same temperature. Mantle plumes, columns of unusually hot material rising from the deep mantle, create volcanic hotspots at the surface. Seismic imaging beneath Iceland reveals a narrow plume extending through the entire mantle, with excess temperatures of roughly 200–300 K in the upper mantle and around 200 K in the lower mantle compared to surrounding rock.17Earth and Planetary Science Letters. Tomographic evidence for a narrow whole mantle plume below Iceland Olivine-based estimates put the mantle potential temperature beneath Hawaii at about 1,688°C and beneath Iceland at about 1,637°C, both hundreds of degrees above the mid-ocean ridge average.1Geochemistry, Geophysics, Geosystems. Mantle potential temperatures at Hawaii, Iceland, and the mid‐ocean ridge system, as inferred from olivine phenocrysts: Evidence for thermally driven mantle plumes
However, a global assessment converting seismic velocities to temperatures found that only about 45% of plume-fed hotspots are genuinely hot, with excess temperatures of 155°C or more above ridge values. Roughly 40% are warm but not hot enough to drive active upwelling on their own, and about 15% are surprisingly cold.18PubMed. On the relative temperatures of Earth’s volcanic hotspots and mid-ocean ridges This challenges the simple picture of all hotspots as expressions of deep thermal plumes. Some may be driven more by chemical buoyancy or other processes than by extreme heat alone.
On the cold side, subducting oceanic plates introduce slabs of relatively frigid rock into the mantle. These slabs can sink all the way to the core-mantle boundary, where they accumulate and produce regions of lower-than-average temperature. Seismic observations show that these thermally defined cold slabs explain the discontinuity structure seen in the deep mantle.19Earth and Planetary Science Letters. Lower-mantle seismic discontinuities and the thermal morphology of subducted slabs Yet even within these cold, slab-dominated regions, geodynamic models predict the formation of localized hot ridges, thin linear anomalies where heat accumulates between the cold downwellings.20Journal of Geophysical Research: Solid Earth. The Formation of Hot Thermal Anomalies in Cold Subduction‐Influenced Regions of Earth’s Lowermost Mantle The deep mantle, in short, is thermally lumpy.
How Scientists Actually Measure Mantle Temperature
Nobody has drilled even a tenth of the way into the mantle. The deepest borehole ever drilled reached about 12 km, barely scratching the crust. Everything we know about mantle temperatures comes from indirect methods, each with its own strengths and blind spots.
Seismic tomography is the workhorse. Earthquake waves travel faster through cold, stiff rock and slower through hot, soft rock. By measuring tiny differences in wave travel times across thousands of earthquake-station pairs, geophysicists build three-dimensional maps of seismic velocity throughout the mantle. Converting those velocity maps to temperature requires knowing how rock composition and mineral phase affect wave speed. Researchers in the Canadian Shield, for instance, showed that upper-mantle P-wave velocities correlate well with temperatures derived from surface heat flow data, confirming that temperature is the dominant control on seismic speed in the shallow mantle.21Journal of Geophysical Research: Solid Earth. Upper mantle velocity‐temperature conversion and composition determined from seismic refraction and heat flow Similar inversions of shear-wave data have been used to map regional temperature profiles and lithosphere thickness beneath continents.22Geochemistry, Geophysics, Geosystems. Thermal and compositional structure of the subcontinental lithospheric mantle: Derivation from shear wave seismic tomography
Mineral physics provides the calibration. By squeezing mantle-like materials in diamond anvil cells and heating them with lasers, experimentalists determine how minerals behave under deep-Earth pressures. These experiments establish melting curves, phase boundaries, and the relationship between density, temperature, and seismic velocity. For example, laser-heated diamond anvil cell experiments on mantle-composition rock have been performed at pressures up to 179 gigapascals, spanning nearly the full pressure range of the mantle.23Journal of Geophysical Research: Solid Earth. Melting experiments on peridotite to lowermost mantle conditions Computational simulations complement the lab work. First-principles molecular dynamics calculations recently predicted that the mineral davemaoite has a melting temperature at core-mantle boundary pressure of about 7,700 K, roughly 2,000 K higher than the more common bridgmanite.24PubMed Central. Davemaoite as the mantle mineral with the highest melting temperature Because the mantle does not actually reach that temperature, davemaoite stays solid in the deep mantle, but its extreme melting point matters for understanding what would happen to mantle rock under different thermal scenarios.
Geoneutrino detection is a newer and entirely different approach. Particle physics detectors originally built to study neutrinos from reactors and the sun also pick up electron antineutrinos from the radioactive decay of uranium and thorium inside the Earth.11Earth and Planetary Science Letters. Quantifying Earth’s radiogenic heat budget These measurements directly constrain how much heat the planet produces from radioactive decay, bypassing the need to estimate element concentrations from rock samples alone. The technique cannot tell you the temperature at a given depth, but it pins down the power source that sustains those temperatures over geological time.
The Mantle Has Been Cooling for Billions of Years
Earth’s mantle was significantly hotter in the deep past. The strongest evidence for this comes from komatiites, a type of ultra-hot volcanic rock that requires mantle temperatures far above anything seen at modern mid-ocean ridges. The decline in komatiite eruptions from the Archean eon (more than 2.5 billion years ago) to the present has been interpreted as evidence for up to 500°C of secular cooling.25Earth and Planetary Science Letters. Thermal evolution of the Earth as recorded by komatiites Put differently, the mantle beneath mid-ocean ridges billions of years ago may have been close to 2,000°C, hot enough to produce lavas that no modern volcano can match.
This cooling is not steady or uniform. It depends on how efficiently plate tectonics removes heat from the interior and how much radiogenic fuel remains. Uranium, thorium, and potassium are slowly being consumed by their own decay, so the radiogenic contribution to Earth’s heat budget is gradually shrinking. The planet is losing heat faster than it generates it, which is why the mantle cools over time. But the rate is slow enough that the mantle will remain convectively active for billions of years to come.
How Earth’s Mantle Compares to Its Neighbors
Mars, Venus, and Earth all likely started in a similar hot, molten state after formation, but their mantles have diverged dramatically since then. Mars, being smaller, lost its internal heat faster. Its mantle cooled to the point where global volcanism and any magnetic dynamo shut down relatively early in its history. The Martian core may have avoided freezing entirely if it contains enough sulfur, which lowers the melting point of iron alloys.26Icarus. Magnetism and thermal evolution of the terrestrial planets
Venus presents the opposite problem. Without plate tectonics to export heat efficiently, Venus’s mantle is predicted to run hotter than Earth’s. Convection models with a rigid, immobile lithosphere, a reasonable approximation for Venus, produce higher interior temperatures than comparable models with a mobile lid.27Journal of Geophysical Research: Solid Earth. Mantle dynamics in Mars and Venus: Influence of an immobile lithosphere on three‐dimensional mantle convection Venus’s slightly higher temperature and lower central pressure compared to Earth may also explain why it lacks a significant inner core, since the conditions needed to freeze iron at the center of the planet may never have been met.26Icarus. Magnetism and thermal evolution of the terrestrial planets In this sense, Earth’s plate tectonics is not just a surface phenomenon. It is a cooling mechanism that has shaped the thermal state of the entire planet, from the mantle all the way down to the inner core.