Hidden Deep Earth Blobs May Explain Why Our Planet Came Alive

Far beneath our feet, two gigantic blobs of strange rock and splashes of molten material near Earth’s core may help explain why this planet, and not its rocky neighbors, ended up with oceans, a breathable atmosphere, and life. New research argues that these mysterious deep-mantle structures are fossil traces of Earth’s molten youth and quiet conduits linking the planet’s core, mantle, volcanoes, and long-term habitability.

In a Nature Geoscience study published on 12 September 2025, an international team led by Jie Deng of Princeton University and geodynamicist Yoshinori Miyazaki of Rutgers University used thermodynamic and geodynamic modeling to show how a “basal magma ocean” at the base of the mantle, continuously contaminated by material leaking from the core, could crystallize into the odd, dense regions seismologists see today. Their models reproduce continent-sized large low-shear-velocity provinces (LLSVPs) under Africa and the Pacific and ultra-low-velocity zones (ULVZs) hugging the core, and they argue that these structures carry chemical fingerprints that help make Earth uniquely habitable.

For decades, seismologists have known that Earth’s lowermost mantle is not uniform. Instead, it hosts two vast LLSVPs, each the size of a continent, perched some 1,800 miles down beneath Africa and the Pacific Ocean. These regions are hotter, slightly slower for seismic waves, and denser than the surrounding mantle. Embedded within or near their margins are ULVZs, thin patches of partially molten material that slow seismic waves dramatically, almost like lava puddles clinging to the core’s surface.

Classic models of planet formation struggle to explain this landscape. In the standard picture, early Earth was wrapped in a global magma ocean produced by colossal impacts. As that ocean cooled and crystallized, scientists expected the mantle to separate into relatively clean, chemically layered shells, not into two gigantic piles of odd, dense rock plus scattered melt. Yet global seismic imaging shows no strong layering and instead reveals these irregular piles at the base of the mantle, raising an uncomfortable question: what went wrong in the calculations?

Deep Mantle Blobs As Ancient Fingerprints

Miyazaki and colleagues argue that the missing ingredient is the core itself. In their scenario, the basal magma ocean at the bottom of the early mantle did not crystallize in isolation. Instead, as the metallic core slowly cooled, tiny crystals of magnesium oxide (MgO) and silicon dioxide (SiO₂) exsolved out of the liquid iron, rose into the overlying magma, and dissolved. This steady chemical leak, they propose, transformed the deep magma into what they call a “basal exsolution contaminated magma ocean,” or BECMO.

Using a self-consistent thermodynamic model in a simplified MgO-FeO-SiO₂ system, the team tracked how this contaminated magma ocean would crystallize as temperatures at the core-mantle boundary fell from about 5,000 to 3,700 Kelvin. The exsolved SiO₂ from the core continually replenished silicon in the melt and strongly suppressed the formation of ferropericlase, an iron-rich mineral that would otherwise create an excessively dense, thick basal layer inconsistent with seismic data. Instead, the solidified BECMO produced a moderately dense bridgmanite-dominated layer enriched in iron and silicon, capped in places by thin, extra-dense patches rich in iron that naturally resemble ULVZs.

“These are not random oddities,” said Miyazaki, an assistant professor in the Department of Earth and Planetary Sciences in the Rutgers School of Arts and Sciences. “They are fingerprints of Earth’s earliest history. If we can understand why they exist, we can understand how our planet formed and why it became habitable.”

High-resolution geodynamic simulations then followed the evolution of this dense basal layer over billions of years of mantle convection. The models show the material being swept into thick, thermochemical piles above the core in regions of converging mantle flow, naturally forming LLSVP-like structures that occupy about 2 to 3 percent of the mantle volume. Smaller, very dense patches at the base survive as ULVZ analogues with heights of roughly 20 to 50 kilometers, especially around the edges of the larger piles, a pattern that broadly matches how ULVZs are observed today.

This framework also helps reconcile the density and seismic velocity contrasts implied by observations. Because the BECMO-derived layer spreads iron more evenly and introduces large amounts of silica, the resulting piles are intrinsically denser than the surrounding mantle yet not so dense that they must form a continuous shell. They can persist for billions of years, but remain dynamic enough that portions are entrained into mantle plumes, potentially feeding volcanic hotspots such as Hawaii and Iceland and carrying deep-Earth chemical signatures to the surface.

How Earth’s Deep Memory Ties To A Living Surface

The modeling suggests that the basal exsolution contaminated magma ocean inherited silicon, tungsten, and helium isotopic characteristics from the core while retaining trace element enrichments from the original magma ocean. When bits of this material are later tapped by plumes, they could help explain why ocean island basalts, such as those erupting at hotspots, show isotopic signatures distinct from mid-ocean ridge basalts sourced in the upper mantle, including lighter silicon isotopes in some cases and anomalous helium and tungsten ratios.

Those chemical quirks may be more than a curiosity. Core-mantle exchange affects how fast Earth cools, how heat and material move between layers, and how vigorously the mantle convects over time. That, in turn, shapes volcanic activity and the evolution of the atmosphere. By preventing strong, simple layering in the mantle and maintaining complex, long-lived reservoirs at depth, the deep Earth may have helped sustain a style of volcanism and outgassing that kept oceans stable and the atmosphere relatively temperate, in contrast to Venus’s crushing carbon dioxide shroud and Mars’s thin, freezing air.

“Even with very few clues, we’re starting to build a story that makes sense,” Miyazaki said. “This study gives us a little more certainty about how Earth evolved, and why it’s so special.”

At the same time, the authors are cautious about the limits of their work. The thermodynamic calculations rely on a simplified three-component mantle composition and assume that MgO and SiO₂ are the dominant exsolved oxides. The geodynamic simulations are two-dimensional and tuned to plausible, but still uncertain, values for mantle viscosity, Rayleigh number, and exsolution rates. Future three-dimensional models and additional high-pressure experiments will be needed to fully test whether BECMO-derived piles can match the detailed shapes and seismic textures of real LLSVPs and ULVZs.

Still, by explicitly coupling core exsolution to magma ocean crystallization and tracking the consequences for both mantle dynamics and geochemistry, the study offers a unified way to understand deep mantle heterogeneities that were previously treated as separate puzzles. The same process that etched continent-sized “blobs” and small molten patches into Earth’s lowermost mantle may also have helped to regulate the planet’s cooling, feed hotspots, and preserve long-lived reservoirs that record conditions from the planet’s fiery beginnings.

If that picture holds up, the strange structures sitting just above Earth’s core are not geological curiosities. They are the planet’s deep memory, carrying a chemical record from a time when the surface was an incandescent ocean of magma, and quietly influencing the conditions that made a watery, life-supporting world possible.

Nature Geoscience: 10.1038/s41561-025-01797-y