Some of Earth’s most persistent volcanic regions occur far from the boundaries between tectonic plates. Hawaii is the classic example: a chain of volcanoes crosses the middle of the Pacific Plate, with progressively older islands and seamounts extending away from the currently active end. One explanation is a mantle plume, a long-lived upwelling of unusually hot, buoyant mantle that rises beneath the moving plate.
For the deepest plumes, evidence increasingly points toward origins near the core–mantle boundary, almost 2,900 kilometers below the surface. That does not mean liquid iron from the core simply rises through the mantle. A plume is primarily an upwelling on the mantle side of the boundary, where extreme heat, chemical variations, and slow convection can make solid rock buoyant enough to begin a journey through much of Earth’s interior.
A Thermal Boundary Beneath the Mantle
The core–mantle boundary separates the liquid metallic outer core from the rocky lower mantle. Across this interface, density, composition, and physical properties change dramatically. Heat flowing out of the core enters the base of the mantle, while the mantle itself also carries heat left from planetary formation and produced by radioactive decay.
When a hot boundary layer becomes sufficiently unstable, buoyant material can rise. In the classic plume model, an initial bulbous plume head ascends ahead of a narrower tail. The head can spread beneath the lithosphere and, in some cases, is linked with exceptionally large episodes of volcanism. A persistent tail can continue delivering hot mantle toward the surface for tens of millions of years or longer.
The rock in a plume remains mostly solid through much of its ascent. Mantle rock flows because high temperatures allow it to deform over geological time, not because it is a column of liquid magma. Extensive melting generally begins much shallower, where rising material encounters lower pressure and undergoes decompression melting.
Seismic Tomography Can Trace Deep Upwellings
Scientists search for mantle plumes by measuring how seismic waves move through Earth. Hotter mantle usually transmits shear waves more slowly than cooler surroundings, allowing seismic tomography to image broad low-velocity structures beneath some hotspots.
Tracing a narrow conduit through nearly 3,000 kilometers of mantle is difficult. Seismic coverage is uneven, resolution decreases with depth, and mantle flow can tilt or broaden an upwelling. Plumes are therefore not expected to appear as perfectly vertical cylinders.
Even so, several major hotspots have evidence consistent with lower-mantle or very deep sources. Hawaii has long been associated with a plume extending into the deep mantle, and studies of Iceland also support a hot upwelling that reaches below the mantle transition zone. Global tomography has revealed broad conduits in the lower mantle beneath a number of hotspot regions, helping connect surface volcanism with structures near the base of the mantle.
Plume Roots Cluster Around Deep Mantle Structures
The core–mantle boundary is not uniform. Two enormous low-seismic-velocity regions, beneath Africa and the Pacific, dominate much of the lowermost mantle. These large low-velocity provinces, or LLSVPs, appear to influence where some deep upwellings form.
Plume reconstructions frequently place their roots near LLSVP margins rather than randomly across the core–mantle boundary. One explanation is that the edge of a dense thermochemical province creates strong lateral temperature contrasts. Hot mantle can be deflected and concentrated there, encouraging instability and upward flow.
Very small but extreme seismic anomalies called ultra-low-velocity zones also occur along parts of the boundary. Some sit near suspected plume roots. They may contain iron-rich material, partial melt, recycled crust, or products of reactions between core and mantle. Their exact origin remains unresolved, but they reinforce the picture of a chemically and thermally complicated plume source region.
Chemistry Can Preserve a Memory of the Deep Source
Volcanic rocks provide another way to investigate plume depth. Ocean-island basalts contain isotopic combinations that differ from typical mid-ocean-ridge basalts, indicating that plumes sample mantle reservoirs with distinct histories. Some signatures point to ancient material that has remained isolated for very long periods; others record recycled crust that once passed through subduction zones.
A striking development came from measurements of ruthenium and tungsten isotopes in Hawaiian basalts. Research published in 2025 found a combination consistent with a small contribution ultimately derived from the core. The interpretation is still being tested, but it offers unusually direct geochemical support for interaction across the core–mantle boundary and for the ability of a deep plume to carry that signal upward.
This would not require bulk liquid metal to travel to the surface. Chemical components could cross the boundary, become incorporated into mantle minerals or oxides, and later be entrained in rising mantle. By the time plume material melts beneath the lithosphere, the resulting magma can preserve only a diluted chemical trace of events that occurred thousands of kilometers deeper.
Plumes Change as They Rise
A plume beginning near the base of the mantle does not travel through a motionless background. The mantle itself circulates. Descending slabs, broad return flow, viscosity changes with depth, and phase transitions around the mantle transition zone can bend or reshape an upwelling.
Lower-mantle conduits may be hundreds of kilometers wide, becoming narrower or more complicated higher up. Near the lithosphere, hot plume material can spread laterally. If a tectonic plate moves over the melting region, successive volcanoes can form a track that records plate motion relative to the deeper source.
Not every hotspot must have the same kind of plume, and not every intraplate volcano requires a structure extending to the core–mantle boundary. Some volcanism can be generated by shallower mantle processes, lithospheric extension, or small-scale convection. The strongest deep-plume cases are those where multiple lines of evidence—seismic imaging, volcanic age patterns, geochemistry, and geodynamic modeling—point in the same direction.
A Connection Across Nearly the Whole Mantle
Mantle plumes provide one of the clearest possible links between Earth’s deepest rocky boundary and volcanic processes near the surface. Their roots form in a region where core heat, recycled tectonic material, ancient mantle reservoirs, and chemically unusual structures meet.
As a plume rises, it can transport heat and distinctive material through most of the mantle before decompression produces melt in the uppermost hundreds of kilometers. The resulting volcano may be the visible end of a circulation system whose source lies close to the metallic core. The details differ from plume to plume, but the broader picture has transformed hotspots from isolated surface features into clues about dynamics operating across almost the full depth of the mantle.