Core, mantle, crust

Foundations

What is under your feet

Almost everyone carries a picture of the Earth’s interior involving a thin crust floating on a sea of liquid rock, with volcanoes as holes that let it out. The picture is wrong at the most important point. The mantle, which is eighty-four percent of the Earth’s volume, is solid.

It is solid and it flows, which sounds contradictory and is not. Glacier ice is solid; you can walk on it, and it shatters if you hit it. Over years it also flows downhill. The mantle does the same thing at a temperature of one to three thousand degrees and over millions of years, creeping at a few centimetres a year. That slow flow is the engine of the next lesson.

Magma is the exception, not the rule. It forms in limited places where solid mantle rock rises and the pressure on it drops far enough to let it partly melt. Volcanoes tap those pockets. They are not windows into a global ocean.

How anyone knows

The deepest hole ever drilled is on Russia’s Kola Peninsula and reached about 12.3 kilometres before the rock became too hot and plastic to continue. The centre of the Earth is 6,371 kilometres down. Every direct sample humans have ever taken comes from the outermost fifth of one percent.

The interior is known instead from earthquakes. A large quake sends vibrations through the whole planet, and there are two kinds. Compression waves, called P-waves, squeeze and stretch the rock along their direction of travel, and they can pass through anything, including liquids. Shear waves, S-waves, wobble the rock sideways, and a liquid cannot be sheared, so S-waves stop dead at any liquid boundary.

Put seismometers around the world and the arrival times, paths and absences map the inside of the planet. In 1909 Andrija Mohorovičić noticed waves speeding up abruptly at a certain depth and identified the boundary between crust and mantle, now named after him. In 1913 Beno Gutenberg mapped a huge shadow zone where S-waves never arrive, revealing a liquid core. And in 1936 Inge Lehmann noticed faint P-waves arriving inside that shadow, where nothing should have reached, and worked out that they had been bent by a solid ball at the very centre. She had found the inner core, from arrival times, on paper.

The four layers

The crust is thin: about six kilometres under the oceans, thirty to forty under most continents, and up to seventy under high mountains. It is the light scum that floated to the top and stayed there.

The mantle, 2,900 kilometres of silicate rock, mostly magnesium and iron silicates. Solid, hot, and slowly convecting.

The outer core, about 2,260 kilometres of liquid iron and nickel. This is where Earth’s magnetic field comes from.

The inner core, a solid ball about 1,220 kilometres in radius, at over 5,000 degrees, roughly the temperature of the Sun’s surface.

Why the inner core is solid

The obvious question is why the hottest part of the planet is the one that is solid. The answer is pressure. Melting is a competition between heat, which shakes atoms apart, and pressure, which pushes them together, so a substance’s melting point rises as you squeeze it. At the centre of the Earth the pressure is over three million times atmospheric, and that is enough to raise iron’s melting point above the local temperature. The iron freezes despite the heat, and it is still freezing: the inner core grows slowly as the planet cools.

How the layers came to be

The Earth did not form in layers; it sorted itself into them.

While the young planet was hot from accretion and from the giant impact of the last lesson, it was partly molten, and molten iron and molten rock do not mix. Iron is far denser, so it drained downward in sheets and drops, releasing gravitational energy that heated the planet further, which kept it molten and helped the rest of the iron follow. Tungsten isotopes let this be dated: most of the core had formed within about thirty to fifty million years of the solar system’s beginning.

The iron took company with it. Elements that dissolve readily in molten iron, the iron-loving ones, went down too, which is why the rocky part of the Earth is so poor in gold, platinum and nickel compared with the meteorites it was built from. Nearly all the Earth’s gold is in the core, thousands of kilometres beyond reach. What miners find near the surface came later, in the late veneer of the previous lesson.

Meanwhile lighter minerals floated, and when the magma ocean froze, the lightest material formed the first crust. Earth has been remaking that crust ever since.

The magnetic consequence

The liquid outer core does something no other layer can. It is an excellent electrical conductor, it is convecting because heat escapes from below, and it is being twisted by the planet’s rotation. Moving conductor plus rotation plus a seed field gives a self-sustaining dynamo: the motion generates a magnetic field, and the field organises the motion.

This was suspected for decades and could not be shown, because the equations are ferocious. In 1995 Gary Glatzmaier and Paul Roberts ran the first three-dimensional simulation that sustained a field like Earth’s, and it did something they had not asked for: after simulated millennia, the field reversed polarity on its own, exactly as the geological record says it has done hundreds of times.

That field matters for what comes later. It deflects much of the solar wind of lesson 2.6, protecting the atmosphere from being stripped away. Mars, whose core cooled and whose dynamo died, lost most of its air.

Deeper

Reading the interior

The Preliminary Reference Earth Model gives density, pressure, gravity and seismic velocities as functions of radius, derived from travel times of body waves, the frequencies of the planet’s free oscillations after great earthquakes, and the constraints of total mass and moment of inertia. Its principal discontinuities are the Moho, the 410- and 660-kilometre phase transitions in the mantle, the core-mantle boundary at 2,891 kilometres, and the inner core boundary at 5,150 kilometres. Density rises from about 2,600 kilograms per cubic metre in the crust to about 13,000 at the centre, and pressure to about 364 gigapascals.

The mantle’s phase changes

The 410-kilometre discontinuity marks olivine transforming to wadsleyite, and the 660 marks ringwoodite breaking down to bridgmanite and ferropericlase. These are changes of crystal structure, not of state; the rock remains solid throughout. The 660 boundary resists but does not prevent flow between upper and lower mantle, and whether convection is layered or whole-mantle was argued for decades; seismic tomography now shows slabs penetrating to the core-mantle boundary in some regions and stalling in others.

Core formation chronology

The hafnium-tungsten system dates core formation because hafnium is lithophile and tungsten siderophile, and hafnium-182 decays to tungsten-182 with a half-life of about nine million years. If the core formed while hafnium-182 was still live, the mantle acquired excess tungsten-182 relative to chondrites. The observed excess implies core segregation was largely complete within roughly 30 million years, extending to 50 or more depending on the model of how equilibration proceeded during successive impacts.

Light elements in the core

Seismically determined core density is about five to ten percent lower than that of pure iron at core pressures and temperatures, requiring one or more light elements. Candidates are silicon, oxygen, sulphur, carbon and hydrogen. Each has different implications for how the core formed and at what conditions metal and silicate equilibrated. Constraints come from high-pressure experiments in diamond anvil cells and shock compression, from first-principles calculations, and from the difference in density between inner and outer core. No combination is yet agreed.

The dynamo

The geodynamo requires a conducting fluid, a source of motion, and rotation to organise it into a coherent large-scale field. The energy comes from cooling of the core, from latent heat released as the inner core freezes, and from the buoyancy of light elements excluded from the freezing inner core, which is thought to be the most efficient contribution. The magnetic Reynolds number must exceed a threshold for the field to be sustained against ohmic decay, which it comfortably does. Numerical dynamos still cannot run at Earth’s true viscosity and rotation rate, so they operate in a parameter regime far from the real one and yet reproduce many features of the observed field, which is itself a matter of some discussion.

When did the inner core start to freeze

Estimates for the age of the inner core range from under a billion years to over two, depending on the thermal conductivity of iron at core conditions, which was revised sharply upward around 2012 and then partly back. A young inner core requires either an implausibly hot early core or an additional energy source. Palaeomagnetic evidence for a change in field behaviour is being sought as an independent constraint.

Frontier

The bottom of the mantle

Just above the core-mantle boundary lie two continent-sized regions where seismic waves slow markedly, beneath Africa and the Pacific. Whether they are piles of chemically distinct material, thermal features, or graveyards of subducted ocean floor is unresolved, and it matters because they appear to anchor the plumes that produce large volcanic provinces, including the Deccan Traps of the next lesson.

Iron at extreme conditions

Reproducing core pressures and temperatures in the laboratory remains at the edge of what is possible, and different techniques give different melting curves for iron. Since the temperature of the inner core boundary anchors the whole thermal profile of the planet, that uncertainty propagates into estimates of heat flow, the dynamo’s power supply and the age of the inner core.

Transcript and shot list

Target runtime about 7 minutes. Built to correct the picture first, then prove it from the only evidence there is, earthquakes, and follow the layers down to the one that makes the planet habitable. Shot types: ANIM is rendered from code with Manim, AI is a generated shot from Higgsfield with the shared style reference, REAL is agency imagery with credit, CAM is the narrator.

[00:00] S01 · ANIM · The familiar textbook cutaway: thin crust floating on glowing orange liquid; a volcano as a hole in it; then the orange dims to solid grey rock, still slowly flowing.

Here’s the picture almost everyone has of the inside of the Earth. Thin crust, floating on a sea of liquid rock, with volcanoes as holes that let it out. It’s wrong in the most important place. The mantle, which is eighty-four percent of the Earth’s volume, is solid.

[00:22] S02 · ANIM · Glacier ice: a person walks on it; it shatters when struck; time-lapse shows it flowing downhill; the same treatment applied to mantle rock at centimetres per year.

Solid, and it flows. That sounds contradictory, and it isn’t. Glacier ice is solid. You can walk on it. Hit it and it shatters. And over years, it flows downhill. The mantle does exactly that, at one to three thousand degrees, over millions of years, creeping a few centimetres a year. That slow flow is the engine of the next lesson. Magma is the exception, not the rule: it forms in limited places where rising rock loses pressure and partly melts. Volcanoes tap those pockets. They aren’t windows into a global ocean.

[01:05] S03 · ANIM · Cosmic calendar; early 2 September; the Kola borehole at 12.3 km beside a 6,371 km radius, almost invisible.

So how does anyone know? The deepest hole ever drilled is on Russia’s Kola Peninsula. Twelve point three kilometres, before the rock got too hot and soft to continue. The centre of the Earth is six thousand three hundred and seventy-one kilometres down. Every direct sample we have ever taken comes from the outermost fifth of one percent.

[01:31] S04 · ANIM · An earthquake sends P- and S-waves through the planet; P-waves pass through liquid; S-waves stop at the liquid boundary, creating a shadow zone.

We know the inside from earthquakes. A big quake rings the whole planet, and there are two kinds of wave. Compression waves squeeze and stretch the rock along their path, and they go through anything, including liquid. Shear waves wobble it sideways, and you can’t shear a liquid, so they stop dead at any liquid boundary. Put seismometers around the world, and the arrival times, the paths and the absences map the inside of the planet.

[02:06] S05 · ANIM · 1909, Mohorovičić spots a velocity jump; 1913, Gutenberg maps the S-wave shadow; 1936, Lehmann finds faint P-waves inside the shadow, revealing a solid centre.

In 1909 Andrija Mohorovičić noticed waves speeding up abruptly at a certain depth: the boundary between crust and mantle, now named after him. In 1913 Beno Gutenberg mapped a huge shadow where shear waves never arrive, which meant a liquid core. And in 1936 Inge Lehmann noticed faint compression waves arriving inside that shadow, where nothing should have reached, and worked out they’d been bent by a solid ball at the centre. She found the inner core. From arrival times. On paper.

[02:44] S06 · ANIM · The four layers with thicknesses and temperatures: crust 6–70 km, mantle 2,900 km, outer core 2,260 km liquid, inner core 1,220 km radius solid at 5,000 K+.

Four layers. The crust: six kilometres under the oceans, thirty to forty under most continents, up to seventy under high mountains. The light scum that floated to the top. The mantle: two thousand nine hundred kilometres of silicate rock, solid, hot, slowly convecting. The outer core: two thousand two hundred and sixty kilometres of liquid iron and nickel. And the inner core: a solid ball, over five thousand degrees, about as hot as the surface of the Sun.

[03:20] S07 · ANIM · A melting-point curve rising with pressure; at 364 GPa it crosses above the local temperature; iron freezes.

Which raises the obvious question: why is the hottest part of the planet the solid part? Pressure. Melting is a contest between heat shaking atoms apart and pressure pushing them together, so a substance’s melting point rises as you squeeze it. At the centre of the Earth, pressure is over three million times atmospheric, and that’s enough to push iron’s melting point above the local temperature. It freezes despite the heat. And it’s still freezing: the inner core is growing as the planet cools.

[03:59] S08 · ANIM · The molten young Earth; iron draining downward in sheets and drops; gravitational heating; a hafnium-tungsten clock reading 30–50 million years.

None of these layers were there at the start. The Earth sorted itself. While the young planet was molten from accretion and from the giant impact, iron and rock refused to mix, and iron is far denser, so it drained downward in sheets and drops. That release of gravitational energy heated the planet further, which kept it molten, which helped the rest of the iron follow. Tungsten isotopes date it: most of the core had formed within thirty to fifty million years of the solar system’s beginning.

[04:39] S09 · ANIM · Gold and platinum atoms following the iron down; the mantle left depleted; a small late veneer arriving afterwards and staying near the surface.

And the iron took company. Elements that dissolve easily in molten iron went down with it, which is why the rocky part of Earth is so poor in gold and platinum compared with the meteorites it was built from. Nearly all of Earth’s gold is in the core, thousands of kilometres out of reach. What miners find near the surface arrived later, in the late veneer from the last lesson.

[05:12] S10 · ANIM · The liquid outer core convecting, twisted by rotation; field lines organise; a self-sustaining dynamo; then the 1995 simulation reversing polarity on its own.

The liquid outer core does something no other layer can. It’s an excellent conductor, it’s convecting because heat escapes from below, and it’s being twisted by the planet’s spin. Moving conductor, plus rotation, plus a seed field, gives a self-sustaining dynamo. This was suspected for decades and couldn’t be shown, because the equations are ferocious. In 1995 Gary Glatzmaier and Paul Roberts ran the first three-dimensional simulation that sustained a field like Earth’s. And it did something they hadn’t asked for: after simulated millennia, the field flipped polarity on its own. Exactly as the rocks say it has done hundreds of times.

[05:59] S11 · ANIM · The magnetosphere deflecting the solar wind; then Mars with a dead dynamo losing its atmosphere to space.

That field matters for everything after this. It deflects much of the solar wind from lesson 2.6, and shields the atmosphere from being stripped away. Mars’s core cooled, its dynamo died, and Mars lost most of its air.

[06:16] S12 · ANIM · Labels: ESTABLISHED for the layers, the dynamo and core formation; ACTIVE RESEARCH for the core’s light elements and the inner core’s age.

The scorecard. The layers, their sizes, that the mantle is solid, that the core made the magnetic field, and when the core formed: established. What the light elements in the core actually are, and when the inner core started freezing: active research, because reproducing three million atmospheres in a laboratory is at the limit of what can be done.

[06:43] S13 · AI into ANIM · A hand on bare rock; the camera plunges through crust, mantle, outer core to the inner core, then back out; the calendar returns.

So: a planet that sorted itself into layers in its first fifty million years, and has been quietly convecting ever since. That convection has to go somewhere. It goes into moving the surface around, and one of the places it does that most dramatically is the ground under northern India. Next lesson: plate tectonics and the Himalaya.

Shot list

One AI shot, twelve code-rendered shots. Estimated Higgsfield use: 2 to 4 generations including retries.