Decompression melting happens when hot rock from deep inside Earth rises toward the surface fast enough that it doesn’t cool down significantly, but the pressure squeezing it drops. That falling pressure lowers the temperature at which the rock begins to melt, so the rock crosses its melting threshold without any added heat. This single mechanism produces the vast majority of magma on Earth, feeding volcanism at mid-ocean ridges, hotspots, and continental rifts. The process is counterintuitive because most people picture melting as something caused by heating, but beneath your feet, the bigger driver is simply relieving the crushing weight of overlying rock.
Why Pressure Matters More Than Temperature
Rock deep in Earth’s mantle is extremely hot, often well above the temperature that would melt it at the surface. What keeps it solid is pressure. At depths of a hundred kilometers or more, the weight of everything above compresses the rock so intensely that its minerals hold together even at temperatures exceeding 1,300°C. Every mineral has a solidus, the temperature at which it first begins to melt, and that solidus rises steeply with pressure. Push the same rock deeper and it needs to be even hotter before any liquid appears.
When a parcel of mantle rock rises, the pressure drops. If the rock ascends quickly relative to how fast it loses heat to its surroundings, it stays nearly the same temperature while the solidus falls to meet it. At some critical depth, the rock’s temperature and its solidus intersect, and the first tiny fraction of liquid appears. From that point upward, continued ascent generates progressively more melt. The thermodynamic framework for this process treats the rising mantle as undergoing roughly adiabatic decompression, meaning the rock expands slightly and cools a little from the expansion itself, but not enough to prevent crossing the solidus.
The solidus curve for the mantle’s dominant rock type, peridotite, sets the depth where melting kicks in. Laboratory experiments have mapped this curve under volatile-free conditions and found that it defines the fundamental depth-temperature framework for magma generation beneath features like mid-ocean ridges and volcanic hotspots.1Journal of Geophysical Research: Solid Earth. Solidus curves, mantle plumes, and magma generation beneath Hawaii More recent high-pressure experiments using chondritic mantle compositions have refined these curves, reporting that at the deepest mantle pressures near Earth’s core-mantle boundary, the solidus reaches roughly 4,150 K and the liquidus sits a few hundred degrees higher.2Earth and Planetary Science Letters. Solidus and liquidus profiles of chondritic mantle: Implication for melting of the Earth across its history Those extreme depths aren’t where everyday decompression melting takes place, but the shape of the solidus curve at shallower depths, roughly 50 to 150 km, is what controls volcanism at ridges and hotspots.
Mid-Ocean Ridges, the Biggest Melting Factory
The most prolific site of decompression melting on Earth is the global mid-ocean ridge system, a network of underwater mountain chains stretching over 60,000 km where tectonic plates pull apart. As plates diverge, mantle rock wells up from below to fill the gap. That upwelling is mostly passive: the plates move aside and the mantle flows in, rather than being forcefully pushed from below. Electrical imaging beneath the East Pacific Rise has revealed a symmetric, high-conductivity zone at depths of roughly 20 to 90 km, consistent with partial melt forming in passively upwelling mantle.3PubMed. Electrical image of passive mantle upwelling beneath the northern East Pacific Rise The triangular shape of this melt zone matches what passive-flow models predict: a broad region at depth narrowing toward the ridge axis at the surface.
Seismic studies tell a similar story. The MELT Experiment beneath the East Pacific Rise showed that partial melt extends across a broad region several hundred kilometers wide and deeper than 100 km, rather than being confined to a narrow column directly beneath the ridge.4Science. Imaging the Deep Seismic Structure Beneath a Mid-Ocean Ridge: The MELT Experiment The melt begins forming over a wide area at depth as mantle rock crosses its solidus, then gradually focuses toward the ridge axis as it migrates upward. This focusing is what allows a thin strip of volcanic crust to form at the ridge even though melting happens over a much broader zone.
The speed at which plates spread apart matters. At fast-spreading ridges like the East Pacific Rise, mantle upwelling is vigorous and sustained, producing a relatively uniform layer of oceanic crust about 6 to 7 km thick. At very slow-spreading ridges, conductive heat loss from the slowly rising mantle cools it before it can melt as extensively, thinning the crust and changing the chemistry of the resulting basalt. Modeling of this behavior yields normal mantle potential temperatures of around 1,300°C.5Earth and Planetary Science Letters. Variation with spreading rate of oceanic crustal thickness and geochemistry Potential temperature is the temperature the mantle would have if you brought it to the surface without melting or heating it, and it serves as the baseline for estimating how much decompression melting will occur.
Hotspots and Mantle Plumes
Not all decompression melting is driven by plates pulling apart. Mantle plumes, columns of unusually hot rock rising from deep in the mantle, also decompress as they ascend and produce melt for the same reason. The key difference is temperature: plume material is hotter than the surrounding mantle, so it crosses the solidus at a greater depth and generates more melt overall. This is why hotspot volcanism, like Hawaii or Iceland, tends to build thick volcanic piles on top of already-existing plates rather than filling a gap between diverging ones.
Because plume rock starts melting deeper, the resulting magma carries distinct chemical fingerprints compared to mid-ocean ridge basalt. Certain trace elements and isotopic ratios record the depth and extent of melting, giving geochemists a way to distinguish plume-fed lavas from ridge-fed ones. The deeper onset of melting also means the melt interacts with a greater column of surrounding rock as it rises, picking up chemical signatures along the way.
Continental Rifting
When a continent begins to break apart, the stretching and thinning of its thick lithosphere can also trigger decompression melting. Continental lithosphere is typically much thicker than oceanic lithosphere, so the mantle beneath it normally sits at too great a depth (and too high a pressure) to melt. But as rifting thins the overlying plate, the underlying mantle rises and decompresses. Modeling of rifted margins like eastern Canada treats this melting as a direct consequence of upwelling caused by lithospheric stretching.6Earth and Planetary Science Letters. Decompression melting at rifted margins: comparison of model predictions with the distribution of igneous rocks on the eastern Canadian margin
The amount of melt produced during rifting depends on how fast the stretching occurs and how hot the underlying mantle is. A rift above average-temperature mantle may produce only modest volcanism. But if a rift happens to sit over a mantle plume, the combination of lithospheric thinning and anomalously hot rock can flood the surface with enormous volumes of basalt, forming what geologists call large igneous provinces. The Deccan Traps in India and the volcanic margins of the North Atlantic are classic examples of this double effect.
How Water and Carbon Dioxide Change the Picture
The simple story of dry peridotite crossing its solidus as it rises is only part of the picture. Small amounts of water and carbon dioxide dissolved in mantle minerals dramatically lower the solidus, meaning melting starts deeper and produces more liquid than the dry scenario predicts. Beneath mid-ocean ridges, the combined effects of water and CO₂ enhance deep silicate melting: CO₂ triggers the first melts at great depth, and water that was locked in nominally anhydrous minerals partitions into those early melts, amplifying the process.7Geology. Water follows carbon: CO2 incites deep silicate melting and dehydration beneath mid-ocean ridges
This volatile-assisted melting has consequences beyond just producing more magma. As water leaves the solid minerals and enters the melt, the remaining rock becomes stronger and more viscous. This “dehydration strengthening” is thought to help define the base of the rigid oceanic lithosphere, the stiff plate that sits on top of the softer asthenosphere. In other words, the same process that generates melt also helps build the mechanical boundary between a tectonic plate and the convecting mantle beneath it.
Mantle composition adds another layer of complexity. The mantle isn’t a uniform block of peridotite. It contains streaks and blobs of pyroxenite, a rock with a different mineral makeup and a lower solidus than peridotite. When a mixed mantle of peridotite plus pyroxenite rises and decompresses, the pyroxenite begins melting first and can contribute a disproportionate share of the total melt. Calculations of isentropic decompression melting in mixed sources show that both the mantle’s temperature and the proportion of pyroxenite exert strong controls on how much magma is produced.8Journal of Geophysical Research: Solid Earth. The role of pyroxenite in basalt genesis: Melt‐PX, a melting parameterization for mantle pyroxenites between 0.9 and 5 GPa
What Happens to the Melt Once It Forms
Melt produced by decompression doesn’t just sit where it formed. Because liquid rock is less dense than the surrounding solid, it migrates upward through the tiny pore spaces between mineral grains. At first, melt fractions are very small, perhaps a fraction of a percent. But as the melt moves, it can dissolve the minerals it flows past, particularly pyroxene, leaving behind channels of dunite, a rock made almost entirely of the mineral olivine. These dunite channels act as high-porosity highways, allowing melt to travel much faster than it could through the surrounding rock.9Geophysical Research Letters. High‐porosity channels for melt migration in the mantle: Top is the dunite and bottom is the harzburgite and lherzolite
This channelized flow is important because it determines how quickly melt reaches the surface and how much it reacts with surrounding rock along the way. Melt that travels through narrow, fast channels preserves more of its original chemistry. Melt that percolates slowly through a wide zone of porous rock has more time to re-equilibrate, changing its composition. The chemical diversity seen in erupted basalts partly reflects these different transport histories rather than differences in the source rock itself.
When Ice Melts, Rock Melts Too
One of the more surprising implications of decompression melting is that removing weight from Earth’s surface, even the relatively modest weight of ice sheets, can nudge the underlying mantle across its solidus. Iceland sits on both a mid-ocean ridge and a mantle plume, so its mantle is already close to melting. Three-dimensional models of glacial rebound in Iceland since 1890 show that ongoing ice loss increases magma production rates by roughly 100 to 135 percent, generating an additional 0.21 to 0.23 cubic kilometers of magma per year.10Journal of Geophysical Research: Solid Earth. Effects of present‐day deglaciation in Iceland on mantle melt production rates
The effect isn’t just theoretical. Studies of lava compositions erupted from Iceland’s Snæfellsjökull volcano during late Pleistocene deglaciation found a clear chemical shift corresponding to a transient increase of about half a percent in the degree of mantle melting, along with a decrease in the depth at which melting occurred.11Nature. Increased mantle melting beneath Snaefellsjökull volcano during Late Pleistocene deglaciation Half a percent may sound trivial, but in a system where small melt fractions control eruption frequency and magma chemistry, it’s a measurable change. As modern glaciers continue to shrink, this feedback loop between surface ice loss and deep magma production is drawing increased attention from volcanologists.
Decompression Melting on a Hotter, Younger Earth
Earth’s mantle was significantly hotter in the Archean eon, roughly 2.5 to 4 billion years ago. Higher temperatures meant that pressure-release melting started deeper and generated thicker crusts. Modeling of this hotter mantle indicates that Archean decompression melting produced thick layers of basaltic or komatiitic crust underlain by a depleted harzburgite residue, compared with the thinner oceanic crust we see today.12Earth and Planetary Science Letters. Cooling of the earth in the Archaean: Consequences of pressure-release melting in a hotter mantle Komatiites, lavas so hot they could flow almost like water, are essentially absent from modern volcanism but are common in Archean rock records, providing direct evidence that decompression melting was far more vigorous in Earth’s youth.
This thicker early crust had consequences for plate tectonics. A thicker, more buoyant crust is harder to subduct, which may have influenced whether and how plate tectonics operated in Earth’s first billion years. As the mantle gradually cooled, the depth at which decompression melting began crept shallower, the crusts grew thinner, and the style of tectonics shifted toward the subduction-dominated system we recognize today.
Beyond Earth
Decompression melting isn’t limited to our planet. On Mars, researchers have proposed that large impacts could have triggered it. When a massive impactor excavates a crater, it removes a huge column of crustal material, suddenly reducing the pressure on the mantle below. Modeling of infilled craters on Mars suggests that this impact-induced decompression was sufficient to generate magma from the Martian mantle, with the resulting volcanic material partially filling the crater from below.13Icarus. The formation of infilled craters on Mars: Evidence for widespread impact induced decompression of the early martian mantle? The early Martian mantle, like the early Earth’s, was hotter and therefore more susceptible to this effect. Some researchers argue that widespread impact-induced decompression melting could explain volcanic features on Mars that don’t obviously correspond to plate boundaries or plume sites, since Mars likely never developed Earth-style plate tectonics.
The same basic physics applies to any rocky body with a hot interior: remove pressure from above, and you can trigger melting below. Jupiter’s moon Io, the most volcanically active body in the solar system, experiences extreme tidal flexing that drives internal heating, and some models invoke decompression of locally heated mantle material as part of its eruptive process. Even on Earth’s Moon, ancient mare basalts may partly reflect decompression melting triggered by large basin-forming impacts during the Late Heavy Bombardment.
Slab Break-Off and Unusual Tectonic Settings
Decompression melting can also occur in tectonic scenarios that don’t fit neatly into the “ridge, plume, or rift” categories. One example is slab break-off, a process where a subducting plate detaches at depth. When the dense slab tears away, hot asthenospheric mantle rushes in to fill the gap beneath the overriding plate, rising rapidly and decompressing as it does. Numerical modeling shows that this inflow of asthenosphere causes partial melting of both the upwelling mantle itself and the base of the overriding lithosphere, with the whole melting episode lasting a few hundred thousand years.14Lithosphere. Numerical Modeling of Melting Processes During Slab Break-off: Insights Into Tectonic Setting for Massif-Type Anorthosites Under warmer mantle conditions, the slab weakens and narrows more quickly, allowing stronger asthenospheric upwelling and more significant melting.
This mechanism has been invoked to explain certain igneous rock bodies that appear in continental collision zones, far from any spreading ridge. Massif-type anorthosites, large bodies of pale, feldspar-rich rock found in ancient continental interiors, are one example. Their formation has long puzzled geologists because they seem to require large volumes of melt in a setting that doesn’t obviously favor it. Slab break-off provides a plausible trigger: a sudden removal of a dense downgoing plate, followed by rapid upwelling and decompression melting of hot mantle into a space where magma wouldn’t normally form.
Detecting Decompression Melting from the Surface
Geologists can’t watch decompression melting happen directly, so they rely on indirect methods. Seismic tomography maps variations in the speed of earthquake waves through the mantle; regions where waves slow down or change character often indicate the presence of partial melt. The MELT Experiment and similar campaigns have used this approach to image the melt zones beneath ridges.4Science. Imaging the Deep Seismic Structure Beneath a Mid-Ocean Ridge: The MELT Experiment Electromagnetic surveys add another perspective: because silicate melt is far more electrically conductive than solid rock, magnetotelluric measurements can pick out melt-bearing zones that seismic methods alone might miss.3PubMed. Electrical image of passive mantle upwelling beneath the northern East Pacific Rise
At the surface, the chemistry of erupted basalts serves as a record of what happened at depth. Trace elements and isotopic ratios in mid-ocean ridge basalts encode information about the depth at which melting began, the fraction of source rock that melted, and even whether the source contained garnet, a mineral stable only at high pressures. Recognizing a true garnet signature in pooled ridge basalts requires conditions that are relatively uncommon: enriched, hot, and slightly damp mantle sources, and even then the garnet signal is only detectable at ultraslow-spreading ridges where melting fractions are small enough that the deep signature isn’t diluted.15Journal of Geophysical Research: Solid Earth. Origins of Major Element, Trace Element, and Isotope Garnet Signatures in Mid‐Ocean Ridge Basalts In subduction zones, seismic properties combined with mineral elasticity data can reveal metasomatized mantle regions where melts have altered the composition of the surrounding rock, leaving behind distinctive velocity signatures that researchers can map from the surface.16Tectonophysics. Seismic signals induced by the Metasomatism of mantle wedge by siliceous melts: Insights from the elasticity of orthopyroxene at high pressure and temperature
Together, these tools, seismic waves, electrical conductivity, and erupted rock chemistry, give researchers converging lines of evidence about where and how decompression melting operates beneath us. The picture that has emerged over the past few decades is richer than a simple “rock goes up, rock melts” story. It’s a process shaped by volatile content, source rock heterogeneity, spreading rates, ice loading, and even asteroid impacts, all feeding back into one of the most fundamental processes shaping rocky planets.