Why Is Everest Still Rising?: Going Further 9–12

Inside the Earth

In Grade 6 you met the Earth’s layers: a thin crust, a thick rocky mantle, and a core of iron and nickel, liquid outside and solid inside. Here they are with their sizes:

Layer	Depth (km)	State	Made of (mostly)	Density (g/cm³)
crust	0 to 5–70	solid rock	granite-like rock under continents; basalt under oceans	2.7–3.0
mantle	to 2,890	solid rock (that flows very slowly)	peridotite: rock rich in iron and magnesium	3.3–5.6
outer core	2,890–5,150	liquid	molten iron and nickel	9.9–12.2
inner core	5,150–6,371	solid	iron and nickel	12.8–13.1

The crust is thin: about 7 km under the oceans and about 35 km under most continents, but about 70 km under the Himalaya and Tibet. If the Earth were an apple, the crust would be about as thin as its skin.

Figure 36.1 The Earth cut open, to scale (left), and how dense it is at each depth (right). The density rises with depth, and jumps where one layer meets the next. This is a one-dimensional model of the Earth: it changes only with depth, the same in every direction.

But how can anyone know this? Nobody has seen the mantle, let alone the core. The first clue comes from simply weighing the Earth.

Worked example

How dense is the Earth?

The Earth’s mass is 5.97 × 10²⁴ kg and its radius is 6,371 km. (a) What is its average density? (b) The rocks of the crust have a density of about 2,700 kg/m³. What does that tell you about the inside?

Answer

a. Volume of a sphere:
$V = \tfrac{4}{3}\pi r^{3} = \tfrac{4}{3} \times 3.14 \times (6.371 \times 10^{6}\ \text{m})^{3} = 1.08 \times 10^{21}\ \text{m}^{3}$
$\rho = \frac{m}{V} = \frac{5.97 \times 10^{24}\ \text{kg}}{1.08 \times 10^{21}\ \text{m}^{3}} = 5{,}500\ \text{kg/m}^{3}$
b. The Earth as a whole is about twice as dense as the rock at its surface. So the inside must be made of something much denser than surface rock, or squeezed much tighter, or both. (It is both: dense iron at the center, and every layer squeezed by the weight above it.)

The Earth’s average density was first measured in 1798, by the English scientist Henry Cavendish, who weighed the Earth by measuring the tiny pull between lead balls in a shed. His answer, 5.45 times the density of water, was within about 1% of today’s 5.51. It proved that the Earth’s heart is heavy. To find out what it is like, we need to look inside, and the only way to look is to listen.

The Earth’s average density is about 5.5 g/cm³, but the rocks at its surface are about 2.7 g/cm³. Which conclusion does this support best?

Pick an answer to see why.

Listening to earthquakes

You learned in Grade 9 that sound and other mechanical waves need a medium to travel through, and that their speed depends on the medium Grade 9: Mechanical waves. When an earthquake strikes, the energy released spreads through the whole Earth as seismic waves, and instruments called seismometers, all over the world, record them. Two kinds travel through the inside of the Earth:

P waves (primary waves) are longitudinal, like sound: the rock is squeezed and stretched along the direction the wave travels Grade 9: Longitudinal and transverse waves. They are the fastest, about 6 km/s in the crust, so they arrive first. They can travel through solids, liquids and gases.
S waves (secondary waves) are transverse: the rock is shaken from side to side, across the direction the wave travels. They are slower, about 3.5 km/s in the crust, so they arrive second. They can travel only through solids.

Why only solids? A sideways shake only travels if each bit of material, pushed sideways, pulls its neighbors along with it and is pulled back. A solid resists being bent out of shape, so it does. A liquid doesn’t resist a change of shape at all: it just flows. Push a spoon sideways through water and the water slides out of the way; nothing springs back. So an S wave dies at the edge of any liquid.

Figure 36.2 A seismometer’s record of an earthquake. The P wave arrives first; the S wave, slower, arrives later and usually shakes harder. The farther away the earthquake, the longer the gap between them.

Because the two waves travel at different speeds, the gap between them tells you how far away the earthquake was, just as the gap between lightning and thunder tells you how far away a storm is.

Worked example

How far away was the earthquake?

A seismometer in Kathmandu records the P wave of an earthquake, and the S wave 10 s later. Taking P waves at 6.0 km/s and S waves at 3.5 km/s, how far away was the earthquake?

Answer

Let the distance be d. The time each wave takes is distance ÷ speed, so the gap is
$\Delta t = \frac{d}{3.5} - \frac{d}{6.0} = d \times (0.2857 - 0.1667) = 0.119\,d$
(with d in km and Δt in s). So
$d = \frac{\Delta t}{0.119} = \frac{10\ \text{s}}{0.119\ \text{s/km}} = 84\ \text{km}$
The earthquake was about 84 km away. (A handy rule for nearby earthquakes: distance in km ≈ 8 × the gap in seconds.)

The shadow on the far side

Here is how waves let us see inside. A seismic wave travels faster in deeper rock, which is squeezed harder and stiffer. So the waves don’t go in straight lines: they bend, the way light bends when it passes from air into glass Grade 10: Refraction of light, and most curve back up to the surface. Where they come up, and when, depends on what they have passed through. Every large earthquake is an X-ray of the whole planet.

Early in the 1900s, seismologists noticed something strange. After a big earthquake, stations within about 104° of it (measured round the Earth from the center, 1° being 111 km along the surface) recorded both P and S waves. But stations farther away than 104° recorded no S waves at all. And between about 104° and 140° there were hardly any P waves either: a ring round the far side of the Earth, a shadow zone, where the earthquake was strangely quiet. Beyond 140°, the P waves came back, but late.

Try it. Pick a model of the Earth, send out waves, and see if it gives the shadow zones real earthquakes give.

An earthquake at the top of the Earth sends out waves in every direction. Each line is one path; the dots show where the waves have reached. The rings outside show where waves arrive: the inner ring for your model, the outer ring for real earthquakes. Choose a model and a kind of wave, and press Send.

inner ring: where this model’s waves arrive · outer ring: where real earthquakes’ waves arrive · pale: only faint waves

Model of the Earth:

Waves:

P waves arrive (this model): everywhere
S waves arrive (this model): everywhere

With the same kind of rock all the way down, both kinds of wave reach every part of the surface. Real earthquakes leave shadows on the far side, so the Earth can’t be like this.

The only model that fits is one with a liquid outer core. S waves can’t cross a liquid, so the core casts a huge S-wave shadow over the whole far side of the Earth. P waves can cross it, but they slow down from about 13.7 km/s to about 8 km/s as they enter it, and so bend sharply. That bending leaves a ring with no P waves, and focuses them farther round, beyond 140°. The size of the shadow tells you the size of the core: it starts 2,890 km down. (Strictly, an S wave that hits the core can set off a P wave in the liquid, which can turn back into an S wave as it comes out on the far side; seismologists do record these. But no S wave crosses the core as an S wave.)

A story from science

Four discoveries

In 1906 the Irish geologist Richard Oldham, who had studied earthquakes in India, noticed that waves which passed near the Earth’s center arrived late: the Earth must have a core in which they travel slowly. In 1909 the Croatian seismologist Andrija Mohorovičić, studying an earthquake near Zagreb, found waves that had sped up beneath a certain depth, about 50 km down there, and so discovered the bottom of the crust. That boundary is called the Mohorovičić discontinuity, or the Moho. In 1913 Beno Gutenberg worked out from the shadow zone that the core begins about 2,900 km down, close to today’s 2,890 km. And in 1926 Harold Jeffreys showed that the core is liquid: no S waves ever cross it, and tides show the Earth is too flexible for a solid core.

Then, in 1936, the Danish seismologist Inge Lehmann looked closely at records of a big earthquake in New Zealand in 1929. In the P-wave shadow zone, where there should have been nothing, there were faint P waves. She worked out that they had been bent and bounced back by something inside the liquid core: a smaller, solid inner core. Her paper has one of the shortest titles in science: “P′”. Lehmann went on working into her nineties, and died in 1993, aged 104.

A black-and-white photograph of a young woman with short hair, in a dark dress, looking at the camera — **Figure 36.3** Inge Lehmann in 1932, four years before she discovered the Earth’s inner core. Photo: Even Neuhaus (6.2.1863-20.4.1946), Public domain

Seismometers more than about 104° from an earthquake never record its S waves. What does this show?

Pick an answer to see why.

More clues from the inside

Seismic waves give the layers’ depths, and their speeds in each layer. Other clues say what the layers are made of.

Meteorites. Some meteorites are rock; others, the iron meteorites, are almost pure iron and nickel. They are pieces of small planets, broken up long ago, that had cores of their own. The Earth is thought to be made of the same stuff, and iron is the only common element dense enough to make the core.
Squeezing rock in the laboratory. In a diamond anvil cell, a tiny sample is squeezed between the points of two diamonds and heated with a laser. Such experiments have reached the pressure at the Earth’s center, about 3.6 million times the pressure of the air. They show how fast seismic waves would travel through iron, or through mantle rock, at each depth, so the waves’ speeds can be matched to the materials. They show too that iron melts at a higher temperature when it is squeezed harder. The center of the Earth, at about 5,000–6,000 °C, is hotter than the outer core, but it is solid because the pressure there is so great.
The magnetic field. You learned in Grade 8 that the Earth’s magnetism comes from molten iron moving in the outer core Grade 8: Why the Earth is a magnet. The field is always changing, and its changes tell us how fast the iron flows. The magnetic pole in the north has moved over 2,200 km since it was first found in 1831; in the early 2000s it was moving 50–60 km a year, and as of 2025 about 35 km a year. To change the field that fast, the iron in the outer core must flow at roughly 10–20 km a year: a liquid, and a runny one.

A solid that flows

Now for the most surprising thing about the mantle. S waves pass right through it, so it is solid. Yet over millions of years, it flows.

You may have seen a solid flow without knowing it. A glacier is solid ice: you can walk on it, and cut steps in it. Yet it flows downhill like a very thick river. The Khumbu Icefall, which climbers cross on the way up Everest from the Nepal side, moves about a meter a day. The ice crystals slowly slide and change shape under the glacier’s own weight. Hot rock under great pressure can do the same, just far more slowly.

A wide, jumbled river of white ice blocks and crevasses pouring down between dark rocky mountainsides — **Figure 36.4** The Khumbu Icefall, seen from Kala Patthar. Solid ice, flowing downhill about a meter a day: a solid can flow if it is pushed long enough. Photo: Goutam1962, CC BY-SA 4.0

The mantle’s rock is hot, from about 1,300 °C just under the plates to over 3,000 °C at the bottom, close to its melting point, though not melted. Strike it and it rings like a bell, carrying seismic waves. But push on it for thousands of years, and it slowly creeps and flows, at a few centimeters a year. It is about 10²⁰ times as stiff as honey, but it moves.

How do we know the mantle flows? One clue is the ground in the far north. Twenty thousand years ago, ice sheets up to 3 km thick covered Scandinavia and Canada. Their weight pushed the crust down into the mantle, and the mantle rock slowly flowed out of the way. The ice had melted away by about 8,000 years ago, and the land is still rising back, as the mantle flows back in underneath: around the Gulf of Bothnia, between Sweden and Finland, by nearly 1 cm a year. Harbors built there a few hundred years ago are now high and dry, and old beaches are found as steps, hundreds of meters above the sea. From how fast the land rises, we can work out how stiff the mantle is.

The top of the mantle and the crust together make a cold, stiff, brittle shell about 100 km thick, the lithosphere (from the Greek for “stone sphere”). It is cracked into about a dozen large pieces and many small ones, the tectonic plates. Below it, the hotter mantle flows more easily: this layer is the asthenosphere (“weak sphere”), and the plates slide over it.

The mantle carries S waves, and S waves can’t cross a liquid. Yet the mantle flows. How can both be true?

Pick an answer to see why.

Where the heat comes from

The inside of the Earth is hot, and the heat is slowly leaking out through the surface. Measured in mines and drill holes all over the world, the ground gets warmer as you go down, by about 25 °C for every kilometer near the surface, and heat flows out, all the time, everywhere. Altogether, about 44 terawatts (44 × 10¹² W) flows out of the Earth: more than twice all the energy people use (about 20 TW, as of 2023).

It comes from two sources, in roughly equal parts:

Heat left over from the Earth’s birth. As the Earth formed, 4.5 billion years ago, rocks crashing together heated it, and iron sinking to the core released still more energy Grade 8: A young, violent Earth. The Earth has been cooling ever since, but a planet this size cools very slowly.
Radioactive decay. Rocks contain a little uranium, thorium and a kind of potassium, potassium-40, whose atoms are radioactive: now and then one breaks apart, giving out energy that ends up as heat. Their half-lives are long, 1.3 billion years for potassium-40, 4.5 billion for uranium-238, 14 billion for thorium-232, so they are still heating the Earth today.

A story from science

The age of the Earth, and a mistake about the inside

In 1862 the great physicist William Thomson, later Lord Kelvin, used the Earth’s heat to work out its age. If the Earth began as a molten ball and has been cooling ever since, then from how fast heat flows out of the ground today, he could work out how long it had been cooling. His answer was somewhere between 20 and 400 million years, and by 1897 he had narrowed it to 20–40 million. Geologists, who could see how long rocks took to form, thought it must be far longer, but nobody could find the mistake in his physics.

There were two. In 1895 Kelvin’s own former assistant, John Perry, pointed out that if the inside of the Earth could flow, carrying heat up by convection, it would stay hot inside much longer, and the Earth could be billions of years old. And in 1896 radioactivity was discovered: a source of heat inside the Earth that Kelvin couldn’t have known about. The Earth is about 4.5 billion years old, a hundred times Kelvin’s answer.

The mantle’s slow boil

In Grade 8 you watched water go round in a beaker heated from below: warm water rises, cools at the top, and sinks again, a convection current Grade 8: Convection. The mantle is heated from below by the hot core, and by its own radioactivity, and cooled from above at the surface. So, though it is solid, it convects, millions of times more slowly than the water.

But the mantle’s convection doesn’t look like neat loops in a beaker. The plates themselves are part of it: they are its cold, stiff top layer.

At a mid-ocean ridge, plates move apart. Hot mantle rock rises to fill the gap; as it rises, the pressure on it falls, and some of it melts. The melt comes up as new sea floor.
The new sea floor moves away from the ridge, cooling as it goes, and gets thicker and denser. After tens of millions of years it is denser than the hot mantle under it.
Where an old, cold plate meets another plate, it sinks back into the mantle. This sinking is called subduction. The sinking slab pulls the rest of its plate along behind it, like a tablecloth sliding off a table: this slab pull is the strongest of the forces moving the plates.
Hot columns of rock, mantle plumes, rise from deep in the mantle, some perhaps from right above the core, and make volcanoes in the middle of plates, as in Hawaii.

Figure 36.5 A slice through the Earth: the plates are the cold top of the mantle’s convection. Old, cold sea floor sinks into the mantle at trenches, pulling its plate along; hot rock rises under the ridges, where plates move apart, and in plumes. The liquid outer core convects too, much faster, and makes the Earth’s magnetic field. This is a three-dimensional model: the mantle is hotter in some places than others at the same depth.

This is not only a picture. Seismic waves travel a little faster through cold rock and a little slower through hot rock. Using millions of records from thousands of earthquakes, seismologists build three-dimensional maps of the mantle, the way a CT scanner builds a picture of your body from X-rays. This is called seismic tomography. The maps show cold slabs of old sea floor sinking deep into the mantle under Japan and South America, and under India and the Himalaya, the cold remains of the ocean that once lay between India and Asia. And they show hot, slow rock rising under Iceland, Hawaii and Africa.

Worked example

Crossing an ocean

The Atlantic Ocean is about 5,000 km wide, and is still widening at about 2.5 cm a year. If it has always widened at that rate, how long ago did it begin to open?

Answer

Use the same units: 5,000 km = 5,000,000 m = 5.0 × 10⁸ cm.
$\text{time} = \frac{\text{distance}}{\text{speed}} = \frac{5.0 \times 10^{8}\ \text{cm}}{2.5\ \text{cm/year}} = 2.0 \times 10^{8}\ \text{years}$
About 200 million years ago. The rocks agree: North America and Africa began to split apart roughly 180 million years ago, and the oldest Atlantic sea floor, off their coasts, is about that age. (The South Atlantic, between Africa and South America, opened later, about 130 million years ago.)

What is the strongest force driving the plates?

Pick an answer to see why.

Where plates meet

The plates move a few centimeters a year, but they are thousands of kilometers across, so where they meet, they build some of the biggest features on the planet. Most of them are hidden under the sea. The mid-ocean ridges wind about 65,000 km round the globe, the longest mountain chain on Earth. The ocean trenches are the deepest places on Earth: the Challenger Deep in the Mariana Trench, in the Pacific, is about 10,935 m deep (as measured in 2021), deeper than Everest is high. And the sea floor is dotted with tens of thousands of seamounts, underwater volcanoes, many in long lines.

Each drawing is a slice through the Earth’s surface, with the plates and the mantle below. The arrows show how the plates move. Choose a kind of meeting.

Plates moving apart:

Plates moving together:

Other:

Two ocean plates move apart, 2–15 cm a year. Hot mantle rock rises to fill the gap; as the pressure on it falls, some of it melts, and the melt erupts as new sea floor. The new rock is hot and stands high, making a mid-ocean ridge, the longest mountain chain on Earth. Example: the Mid-Atlantic Ridge, which comes above the sea in Iceland.

A narrow, straight crack between two walls of dark lava rock with snow on top, filled with still, dark water — **Figure 36.6** A crack in the lava at Þingvellir in Iceland, where the Mid-Atlantic Ridge rises above the sea. The North American and Eurasian plates are moving apart across Iceland by about 2 cm a year, and the ground between them is splitting and sinking into a rift valley. Photo: Debivort at en.wikipedia, CC BY-SA 3.0

Why is there almost no sea floor older than about 200 million years, when some continental rock is over 4 billion years old? Because sea floor is made at ridges and destroyed at trenches, while continents, too light to sink, stay on top. That is a topic of its own: Why Is the Sea Floor So Young?

Under which kind of plate meeting do you expect a deep trench in the sea floor and a line of volcanoes beside it?

Pick an answer to see why.

India meets Asia

About 120 million years ago, India was part of a huge southern continent, Gondwana, next to Africa, Antarctica and Australia. It broke away and moved north across an ocean called the Tethys, at times faster than 15 cm a year, faster than any continent moves today, pulled along by the Tethys sea floor sinking under Asia ahead of it. Long before, on the northern edge of India, under shallow seas, the shells of sea animals had piled up and become limestone, the rock now at the top of Everest Grade 8: The sea at the top of the world.

About 50 million years ago the last of the Tethys sea floor sank, and India hit Asia. A continent is too thick and light to sink far into the mantle, so India slowed down, but it didn’t stop. It is still moving north at about 4 cm a year. About half of that, nearly 2 cm a year, is taken up across Nepal; the rest squeezes Tibet and the land beyond.

Figure 36.7 A slice through Nepal from south to north (heights stretched 4 times). India slides north under Nepal and Tibet along a great fault, the Main Himalayan Thrust. Rock scraped off the top of India is stacked up into the Himalaya. Rivers carry the worn-down rock south, to build the plains of the Terai.

India doesn’t slide under Nepal smoothly. The upper part of the fault between them, the Main Himalayan Thrust, is stuck. Below about 20 km, the rock is hot enough to creep, and India slides on north. But above that, the two sides are locked together, so the creeping drags the edge of Nepal along, and the rock above the stuck part bends and squeezes like a spring being wound up. As it is squeezed, the high Himalaya is pushed up. This is what the GPS stations measure: Everest rises a few millimeters every year.

Every so often, after centuries, the stuck part can’t hold, and the fault slips in seconds: a great earthquake. On 25 April 2015, the Gorkha earthquake began 10–15 km under Barpak in Gorkha district. The fault unzipped eastwards for about 150 km, under Kathmandu, and Nepal’s side slid south over India by several meters. In those seconds, the Kathmandu valley was lifted by roughly a meter and moved 1.5–2 m south, while the land to its north, where the bent rock sprang back, sank by more than half a meter. Nearly 9,000 people died.

Mount Everest from the air: a snowy pyramid standing above a sea of white peaks and clouds, under a blue sky — **Figure 36.8** Mount Everest from the south, seen from a plane. The rock at its very top, the Qomolangma limestone, formed on the floor of a shallow sea on India’s northern edge, about 450 million years ago. Photo: shrimpo1967, edited by Papa Lima Whiskey 2, CC BY-SA 2.0

Between earthquakes, why does the high Himalaya rise?

Pick an answer to see why.

Building up, wearing down

A mountain range is a contest between two kinds of process. The heat inside the Earth builds up: plates collide and lift the land, volcanoes pile up lava, ridges make new sea floor. Water, ice, wind and gravity wear down: frost splits the rock, landslides drop it into the valleys, and rivers carry it away. (How water does that is a topic of its own: What Does Water Do to Rock?)

Nepal is one of the fastest-wearing places on Earth. The monsoon pours rain on steep, broken slopes, and every year the Ganga and Brahmaputra, with all their Himalayan rivers, carry about a billion tonnes of sand and mud to the Bay of Bengal. There it has piled into the Bengal Fan, the biggest heap of sediment on the planet, up to 16 km thick and 3,000 km long. Nearer home, the worn-down Himalaya built the flat plains of the Terai, and the Churia hills (the Siwaliks) are made of sand and gravel worn from the Himalaya over the last 20 million years, then caught up in the push and folded into hills themselves.

But wearing a mountain down doesn’t lower it as much as you would think, because mountains float.

Mountains have roots

The crust is less dense than the mantle under it: about 2.8 g/cm³ against 3.3. So the crust floats on the mantle (slowly, since the mantle flows slowly), like a block of wood floating in water. A thick block floats with its top higher, but also with its bottom deeper. The same goes for the crust: where it stands high, it also has a deep root below, pushing down into the mantle. This balance is called isostasy (Greek for “equal standing”).

Figure 36.9 Left: blocks of wood of different thickness, floating in water. The thicker the block, the higher its top and the deeper its bottom. Right: the crust floats on the mantle the same way. Under the Himalaya and Tibet, the crust is about 70 km thick, twice as thick as under the plains.

Worked example

How deep is the root?

The crust (density 2.8 g/cm³) floats on the mantle (3.3 g/cm³). A plateau stands 5 km higher than the plains. How deep a root must it have under it, below the bottom of the plains’ crust?

Answer

Take a column of crust under the plateau and one under the plains. Floating means that, down to a level below the deepest root, every column weighs the same. The plateau’s column has 5 km of extra crust on top and a root of depth r where the plains’ column has mantle instead. So the extra crust, 5 + r km of it, must weigh the same as r km of mantle:
$2.8 \times (5 + r) = 3.3 \times r$
$14 + 2.8\,r = 3.3\,r \quad\Rightarrow\quad 0.5\,r = 14 \quad\Rightarrow\quad r = 28\ \text{km}$
The root is 28 km deep: for every kilometer of height, 5.6 km of root. Add the 35 km of ordinary crust, and the crust under the plateau is about 35 + 5 + 28 ≈ 68 km thick. Seismic waves show that the crust under Tibet, about 5 km high, is indeed about 70 km thick.

Try it

How much does wearing lower a mountain?

Rivers wear 1 km of rock off the top of a whole mountain range. The range floats on the mantle as in the example. (a) As the rock is removed, what happens to the range’s root? (b) The surface ends up lower by only about 1 km × (1 − 2.8 ÷ 3.3). How much is that?

Answer

a.The range is lighter, so it floats up: the mantle flows back in under it and the root gets shallower.
b.1 − 2.8 ÷ 3.3 = 1 − 0.85 = 0.15, so the surface is lower by only about 0.15 km, 150 m. For every kilometer of rock worn off, the range rises back about 850 m. To wear a mountain range flat, rivers must remove about six times its height in rock. That is why old mountain ranges take hundreds of millions of years to wear away, and why rock that was once 20 km deep is now found on the surface in the Himalaya.

A story from science

Everest, and the plumb line that didn’t swing enough

In the 1850s, the Great Trigonometrical Survey of India, measuring the country with telescopes and chains, found that Peak XV, later named Everest, was the highest mountain on Earth. Its measurements held a puzzle too. A plumb line (a weight on a string) hangs straight down, but next to a mountain, the mountain’s gravity pulls it a little sideways. In 1855 the English mathematician John Pratt, working in Calcutta, calculated how far the Himalaya should pull the surveyors’ plumb lines towards them. They had been pulled, but only about a third as much as the rock of the mountains should.

The same year the astronomer George Airy explained it: the mountains are made of lighter rock that floats on denser rock below, with deep roots. (Pratt later offered another explanation, that the rock under mountains is lighter all the way down; both ideas are used today, but Airy’s roots are what seismic waves find under the Himalaya.) Where the root is, there is light crust in place of dense mantle, so the extra pull of the mountains is canceled out by the missing pull of the root. A century before anyone could see the roots in seismic waves, a plumb line had found them.

The contest over time

So does a mountain grow forever while the push goes on? No: the higher and steeper it gets, the faster rain and rivers wear it down. Eventually wearing down catches up with building up, and the height of the range levels off, while rock keeps moving up through it and being carried away, like water flowing through a full bucket with a hole in it. When the push stops, the range is slowly worn down. The Aravalli hills of India were once a great mountain range, more than a billion years ago.

A simple model. The push from the plates thickens the crust; the thick crust floats higher; rain and rivers wear away the top, faster the higher the range stands. Set how fast the push thickens the crust and how wet the climate is, and when the push stops. The graph shows the average height of the range over time.

Push thickens the crust by: 3.0 mm a year

Climate:

Push stops after: 60 million years

Levels off at: 4.5 km
Time to get 90% of the way: 15 million years
Height after 150 million years: 0.4 km

The range levels off where wearing down balances building up. When the push stops, it is worn down, but slowly: as rock is removed, the crust floats up again, and the lower the range gets, the more slowly it wears. Old ranges last hundreds of millions of years as low hills.

The two contests happen on very different scales, of size and of time:

Process	How fast	How long it lasts	How big
an earthquake	a fault slips meters in seconds	seconds to minutes	a fault up to hundreds of km long
a landslide	meters a second	minutes	meters to kilometers
a volcano erupting	lava builds a cone meters a day	days to years	a cone up to a few km high
a river cutting its valley	millimeters a year	thousands to millions of years	a valley kilometers deep
plates moving, mountains rising	centimeters a year	tens of millions of years	thousands of km
mantle convection	centimeters a year	hundreds of millions of years	the whole mantle, 2,890 km deep

Make a mantle in a pan

Activity

Convection and plates in a pan

You need: a wide, shallow steel pan or a big frying pan, water, a stove, and a few small, thin, light things that float: dry tea leaves, small pieces of thin paper or puffed rice. Ask an adult to help with the stove.

Put about 2 cm of water in the pan and let it go still.
Sprinkle the floating bits thinly over the water.
Put the pan on the stove so that only one side of it is over the flame, on a low flame.
Watch the floating bits closely, from above, for a few minutes, without stirring. Then turn off the stove.
Draw what you see: where the bits move apart, and where they collect.

Don’t let the water boil, and keep your face away from the steam.

What you should see

Warm water rises above the flame, spreads across the surface, cools, and sinks again at the far side: a convection current, as in Grade 8. The floating bits ride on the surface flow: they move apart over the place where warm water rises, like plates at a ridge, and drift towards where the water sinks, where they crowd together, like plates at a trench or a collision. The pan is much simpler than the Earth: the water is heated only from below, and the floating bits don’t sink. In the Earth the plates are part of the flow, and their sinking helps drive it.

In short

The Earth has a crust, a solid rock mantle, a liquid iron outer core and a solid iron inner core. Its average density, 5.5 g/cm³, twice that of surface rock, shows its inside is dense.
P waves (longitudinal) cross solids and liquids; S waves (transverse) only solids. Their shadow zones show the outer core is liquid and where it begins, 2,890 km down; faint P waves in the shadow revealed the solid inner core (Lehmann, 1936). Meteorites, laboratory squeezing and the changing magnetic field add more clues.
The mantle is solid, but over thousands of years it flows. It is heated from below and by radioactive decay, and cooled at the top, so it convects. The plates are the cold top of this convection; sinking slabs pull them along.
Plates moving apart make ridges and rift valleys; plates sinking make trenches and lines of volcanoes; continents colliding make mountain ranges like the Himalaya.
India still pushes north into Asia, about 4 cm a year. The stuck fault under Nepal squeezes the rock and lifts the high Himalaya by a few millimeters a year, until a great earthquake releases some of the strain.
Mountains rise until erosion wears them down as fast as they are built up. Because the crust floats on the mantle (isostasy), mountains have deep roots, and wearing them down takes a very long time.

Questions

Which of these is evidence that the Earth’s outer core is liquid?

Pick an answer to see why.

Glacier ice and the mantle are alike because both

Pick an answer to see why.

a.

A seismometer records the P wave of an earthquake and, 30 s later, the S wave. Taking P waves at 6.0 km/s and S waves at 3.5 km/s, how far away was the earthquake? Why does one station not tell you where it was?

Answer

The gap is Δt = d/3.5 − d/6.0 = 0.119 d, so d = 30 ÷ 0.119 = 252 km (about 250 km). One station gives only the distance, so the earthquake could be anywhere on a circle of radius 250 km round it. With three stations, the three circles meet at only one place: the earthquake.

b.

Explain how seismic waves show that the Earth has a liquid outer core, and how they show where it begins.

Answer

S waves are transverse, and can only travel through solids. Stations more than about 104° from an earthquake never record S waves, so something in the Earth’s middle stops them: a liquid. P waves cross the liquid, but they slow down sharply as they enter it, from about 13.7 to about 8 km/s, and bend, so they leave a ring between about 104° and 140° where hardly any P waves arrive. The size of the shadow zones depends on the size of the core: working back from them, the core begins 2,890 km down.

c.

The mantle is solid, yet it convects. Explain (i) how this is possible, (ii) what heats it, and (iii) what evidence shows that the mantle can flow.

Answer

i.Hot rock under great pressure, close to its melting point, creeps very slowly when pushed for a long time, as glacier ice does: over seconds it is a stiff solid (it carries S waves), but over thousands of years it flows, a few centimeters a year.
ii.It is heated from below by the hot core, which still holds heat from the Earth’s formation, and from within by the decay of radioactive uranium, thorium and potassium-40. It is cooled at the top, through the plates.
iii.Where thick ice sheets melted away 8,000–10,000 years ago, in Scandinavia and Canada, the land is still rising, by up to about 1 cm a year, as mantle rock flows back under it. Seismic tomography also shows cold slabs of sea floor that have sunk hundreds of kilometers into the mantle.

d.

Draw a model, a labeled cross-section, of the Himalaya: show India, Nepal and Tibet, the fault between them, the direction of movement, and where rock is being raised and where it is being worn away. Use it to explain why Everest is still rising, and why it isn’t far higher than it is.

Answer

The drawing should show the Indian plate sliding north under Nepal and Tibet along a gently sloping fault (the Main Himalayan Thrust), with arrows: India moving north at about 4 cm a year, about 2 cm a year of it taken up under Nepal. The Himalaya and Tibet stand high, with a thick crust and a deep root. Rivers flow south, carrying sediment to the Terai and the Bay of Bengal (see Figure 36.7).
Explanation: India keeps moving north; the deep part of the fault creeps, but the shallow part is stuck, so the rock above it is squeezed and the high Himalaya is pushed up, a few millimeters a year; great earthquakes release some of the strain every few centuries. At the same time, frost, landslides and monsoon rivers wear the mountains down, faster the higher and steeper they are. Over millions of years the wearing down nearly balances the building up, so Everest stays at about 9 km instead of growing for ever. Because the crust floats on the mantle, wearing rock away also lets the range float up a little, which slows the lowering.

e.

Explain, with an example of each, how internal processes build up features of the Earth’s surface, and how surface processes wear them down. Why do these happen on such different time scales?

Answer

Building up, powered by the Earth’s internal heat: plates colliding raise mountain ranges (the Himalaya, a few mm a year over tens of millions of years); plates moving apart make mid-ocean ridges and rift valleys (the Mid-Atlantic Ridge, Iceland’s rift); volcanoes build cones and seamounts (Hawaii, the Andes). Wearing down, powered by the Sun (which drives the water cycle) and by gravity: frost splitting rock, landslides in the monsoon, rivers carrying sediment to the Terai and the Bengal Fan, waves wearing coasts.
The time scales differ because the processes differ in speed and size: an earthquake or a landslide happens in seconds or minutes, but each moves only a little rock; plates move only centimeters a year, but over tens of millions of years that is thousands of kilometers. A feature like a mountain range is the result of both, added up over millions of years.

f.

Under the Tibetan Plateau, the crust is about 70 km thick; under the Ganga plains, about 35 km. Taking the crust’s density as 2.8 g/cm³ and the mantle’s as 3.3 g/cm³, how high would you expect the plateau to stand above the plains? (Tibet is about 5 km high.)

Hint

The extra crust, 35 km of it, is the plateau’s height h above the plains plus its root r below. Floating means the extra crust weighs the same as the mantle it replaces: 2.8 × 35 = 3.3 × r.

Answer

The extra 35 km of crust must weigh the same as the mantle its root replaces: 2.8 × 35 = 3.3 × r, so r = 98 ÷ 3.3 = 29.7 km. The height above the plains is h = 35 − 29.7 ≈ 5.3 km, close to Tibet’s real height of about 5 km. The plateau stands as high as a crust that thick should float.

Think about it

Seeing without looking

Comte said in 1835 that we could never know what the stars are made of; we found out from their light (What Are Stars Made Of?). In 1835, could anyone have imagined how we would find out about the Earth’s core? What do the two discoveries have in common?

Answer

In both cases, nobody could go there or bring a sample back, and the answer came from waves that had passed through the place: light from the stars, seismic waves through the Earth. Both needed laboratory measurements to compare with (the wavelengths of elements in a flame; the speeds of waves in iron and rock under pressure), and both needed many careful records, collected and compared. A good question to ask about anything “impossible to know” is: what passes through it, and could we measure that?

Where to go next

Why Is the Sea Floor So Young? No sea floor is much more than 200 million years old, yet some continental rock is 4 billion. The magnetic stripes on the ocean floor, and the evidence for plate tectonics.
What Does Water Do to Rock? Frost splitting rock, limestone dissolving into caves, and water helping rock melt under volcanoes: the wearing-down side of the contest.
How Old Is the Earth? The radioactive elements that heat the Earth are also its clocks.