No one can reach the Earth's centre, so how do we know what's inside?
― Faint tremors were reaching a "shadow" where no waves should go
The deepest hole people have ever dug is just over 12 kilometres deep. The centre of the Earth is almost 6,400 kilometres down. Even so, we know there is a solid iron ball at the centre. The clues were a place where quake waves "don't arrive", and a small tremor that arrived there anyway.
When buying a watermelon, some people tap it with the palm of the hand and listen. Without cutting it open, you can get a rough idea of whether it is full or hollow.
What scientists did to study the inside of the Earth is actually quite similar. Earthquakes do the tapping. When a big quake happens, its tremors travel through the Earth and reach seismometers in countries on the other side.
But when they gathered records from around the world, something odd showed up. On the Earth's surface, there was a "band" where almost no tremors arrived.
There are two main clues
Looking around the globe from where a quake starts, one band of distances gets almost no tremors. Something large inside is bending the paths the waves take.
Weak tremors were recorded in the shadow, where none should arrive. The explanation was that there is another object deeper still, inside that "something".
The first clue showed that "the Earth has a liquid core". The second showed that "there is a solid ball in the middle of that core". Let's take them in turn.
The Earth's interior casts a "shadow" of tremors
Quake tremors travel through the Earth as waves. A wave changes direction where its speed changes. It is like a straw in water that looks bent, and the same mechanism that bends light as it enters water.
In the early 1900s, seismometer records from around the world slowly piled up. They showed that in a band roughly 104 to 140 degrees away from the quake, the first tremors that should arrive were barely recorded. Near the Earth's centre there is a large mass where waves suddenly slow down. The waves bend there and jump over this band.
Look at Figure 1. Suppose a quake happens at the top point. Use the slider to move the station and see where tremors arrive in a straight line and where the shadow falls.
Working out where the shadow band begins showed that the boundary of the core lies about 2,900 kilometres below the surface. Around 1913, the German seismologist Gutenberg is said to have worked out this depth.
There was one more decisive clue. Quake tremors come in two kinds: waves that squeeze in the direction they travel, and waves that shake sideways. Sideways waves cannot travel through liquid. These sideways waves did not reach the far side of the core at all. This led people to think that the outer part of the core is liquid.
Faint tremors in the shadow revealed a second core
Here the Danish seismologist Inge Lehmann enters the story. She was studying records of a large earthquake that struck New Zealand in 1929. She noticed that stations inside the shadow band had also received weak but clear tremors.
With a liquid core alone, no tremors should have arrived there. Lehmann reasoned that there was another mass at the very centre of the core, where waves travel faster. If waves bounce or bend there, a little of the shaking would fall into the shadow. The dotted line in Figure 1 shows that path.
In 1936 she set out this idea in a paper. It is well known for its extremely short title, just a single character for the symbol of this wave. She is also said to have copied the records collected from observatories onto cards one by one and sorted them in empty oatmeal boxes.
Figure 2 lines up the speed of squeezing waves at each depth, as we know it from current research. The speed drops sharply on entering the core, and rises again on entering the ball at the centre. These two "steps" created both the shadow band and the faint tremors inside it.
What Lehmann showed was that "at the centre, there is another mass where waves travel faster". Strong evidence that it is truly solid came in the 1970s. After a giant earthquake, the whole Earth keeps trembling faintly for days, like a bell. It is said that a detailed study of how it trembles provided the proof. Lehmann lived to 104 and saw the result.
A hospital ultrasound scan also sends sound into the body and looks inside using the echoes. The idea behind studying the Earth is the same. Even now, records from seismometers around the world are collected, and maps of the Earth's interior keep being made.
Summary
No one can travel to the Earth's centre. Yet by comparing worldwide where quake tremors do and don't arrive, we have come to know what is inside. The shadow band where tremors don't arrive revealed the liquid core. The faint tremors that reached inside the shadow revealed the solid ball at the centre.
What you can't see shows itself in the shape of a shadow.
And a small exception inside the shadow points to something deeper still.
Why the ball at the centre stays solid despite being so hot is explained in "The Earth's centre is as hot as the Sun's surface, so why doesn't it melt?". For the difference between the two kinds of tremor, see "Why does the small 'rattle' of an earthquake arrive before the big 'sway'?", and for how waves bend at a boundary, see "Why does a straw in water look bent?". For the story of how the Earth's age was found not underground but from lead in meteorites, see "How do we know the Earth is 4.6 billion years old?".
- In a dark room, shine a torch from the side at a round glass of water. Look at the wall on the far side. You will see a bright, focused band and a dark area.
- Empty the glass and shine the torch the same way. The way the light lands should change. When the contents change, the "shape of the shadow" changes.
- Next, ask a family member to put something in a box, and try to guess what it is just from the sound when you tap it and the feel when you shake it. It is a small taste of what scientists did with the Earth.
Seismic waves and light are not the same kind of wave, but they share one thing: they "bend where their speed changes". Do not shine a torch directly into anyone's eyes.
For those who want to know more ― terms, formulas and links to textbooksShows which level each topic belongs to, from middle-school science to university-level courses
- Middle schoolCovered in middle-school science
- High schoolCovered in high-school earth science and physics
- High school+Advanced high-school material, or textbook sidebars
- UniversityUniversity-level specialist courses (seismology, solid-earth geophysics) not taught in high school
- ResearchTopics researchers are still investigating, not yet taught as established fact even at university
Middle schoolTerms: this phenomenon has names
- P waves and S waves: P waves squeeze in the direction they travel (longitudinal waves). S waves shake sideways (transverse waves). S waves do not travel through liquid.
- Shadow zone: the range roughly 104 to 140 degrees from the epicentre where P waves do not arrive directly. S waves do not arrive at all beyond about 104 degrees.
- Outer core and inner core: the liquid iron layer at about 2,900 to 5,150 km depth is the outer core. The solid iron ball inside it, with a radius of about 1,220 km, is the inner core.
- Epicentral distance: the distance between the epicentre and a station, expressed as an angle seen from the Earth's centre. The slider angle in Figure 1 is this.
Middle schoolHigh schoolChecking with a formula: how many minutes does a wave take to pass through the Earth's centre?
Consider a wave that travels straight through the Earth's centre to the point exactly opposite the quake (epicentral distance 180 degrees). We work out the time spent in each of three layers and add them up.
| In symbols | t ＝ L1 ／ v1 ＋ L2 ／ v2 ＋ L3 ／ v3 |
| In words | Travel time ＝ the sum, over the three layers, of "distance travelled ÷ speed in that layer" |
| Where it comes from | It solves "speed ＝ distance ÷ time" for time, and splits the trip into sections of different speed (the same as the middle-school speed calculations where a journey is split into parts) |
| Symbol | Meaning and unit |
| t | Time for the wave to reach the far side (seconds) |
| L1, L2, L3 | Distance through the mantle, outer core and inner core (km). Because the path goes through the centre, each is twice the thickness |
| v1, v2, v3 | Approximate average P-wave speed in each layer (km/s) |
| Distance through the mantle (twice the thickness of about 2890 km) | 5780 km |
| Distance through the outer core (twice the thickness of about 2260 km) | 4520 km |
| Distance through the inner core (twice the radius of about 1220 km) | 2440 km |
| Average P-wave speed (mantle / outer core / inner core, approximate) | 11 km/s ／ 9 km/s ／ 11 km/s |
| Diameter of the Earth | 12742 km |
| Time in the mantle (s) | 5780 ÷ 11 ≒ 525 |
| Time in the outer core (s) | 4520 ÷ 9 ≒ 502 |
| Time in the inner core (s) | 2440 ÷ 11 ≒ 222 |
| Mantle and outer core total (s) | 525 ＋ 502 ＝ 1027 |
| Total t for three layers (s) | 1027 ＋ 222 ＝ 1249 |
| In minutes | 1249 ÷ 60 ≒ 20.8 |
| For comparison: sound in air (0.34 km/s) covering the same distance (s) | 12742 ÷ 0.34 ≒ 37476 |
| In hours | 37476 ÷ 3600 ≒ 10.4 |
By this calculation the tremor reaches the far side of the Earth in about 21 minutes. Actual observations also say that a wave passing through the centre reaches the opposite side in about 20 minutes. Sound in air would need over 10 hours to cover that distance. By comparing "how many minutes later it arrived" at observatories around the world, the speed and depth of each layer have been worked out backwards. The slower speed in the outer core alone corresponds to the step in Figure 2.
High schoolHigh school+Why do waves bend at a boundary, and why does a shadow form?
High schoolWhen a wave enters a region of different speed at an angle, it changes direction (refraction). It is the same as light going from air into water. A P wave travelling at about 14 km/s at the base of the mantle slows to about 8 km/s on entering the outer core. A wave entering the slower region bends toward the normal, the line perpendicular to the boundary. So waves entering the outer core bend toward the centre, jump over the 104 to 140 degree band on the surface, and emerge beyond it.
High school+In the mantle, speed increases with depth. So wave paths are not straight lines but curve like a bow bent downward (the thin lines in Figure 1). S waves, on the other hand, cannot pass through liquid, so they do not reach the far side of the outer core. For them, everything beyond about 104 degrees is shadow. Together, the P-wave shadow and the S-wave shadow led to the conclusion that the outer core is liquid.
UniversitySnell's law and travel-time curves: working backwards to the inside
The direction of refraction is set by Snell's law. In a spherical Earth, we follow each path using a quantity that stays constant along it (the ray parameter). A graph of travel time against epicentral distance is called a travel-time curve. Calculating the speed at each depth from the travel-time curves of many earthquakes is called seismic tomography, or an inverse problem. Waves passing through the inner core are given their own symbol names so they can be told apart from waves that pass only through the outer core. That the inner core is solid was confirmed by analysing the Earth's free oscillations after giant earthquakes.
📖 Derivations and further reading: Snell's law (Wikipedia, in Japanese) ／ Shadow zone (Wikipedia, in Japanese)
ResearchWhat is still not clear
- Does the inner core rotate differently from the layers outside it? There are reports that the inner core spins slightly faster than the mantle. Other studies say the drift has stopped or reversed, and the debate continues.
- Is there yet another core inside the inner core? There are reports that in the central part of the inner core, the way wave transmission varies with direction changes. Whether there is an "innermost inner core" is still being checked.
- When was the inner core born? As the Earth cools, iron is thought to solidify and the inner core to grow little by little. Estimates of when it formed range from several hundred million to more than a billion years ago.
So the content of this article too is "an explanation of what is known so far". The map of the Earth's interior is redrawn each time more seismometer records come in.
Links to textbooks (by level)
| Level | Subject / unit | Where in this article |
|---|---|---|
| Middle school | Science "Changes in the Earth (earthquakes)", Maths "Speed" | P and S waves, checking with a formula |
| High school | Basic Earth Science "Internal structure of the Earth", Physics "Refraction of waves" | The shadow band, waves bending at a boundary |
| High school+ | Earth Science "How seismic waves travel" | Bow-shaped paths, the S-wave shadow |
| University | Seismology, solid-earth geophysics | Snell's law, travel-time curves, free oscillations |
| Research | Inner-core rotation, structure and age | What is still not clear |
| ― | Links to daily life | The same idea as ultrasound scans and other technologies that "see inside by the echo" |
- Lehmann, I. (1936) "P′", Publications du Bureau Central Séismologique International, Série A, Travaux Scientifiques, 14, 87–115.
- Gutenberg, B. (1914) Über Erdbebenwellen VIIA. Nachrichten der Königlichen Gesellschaft der Wissenschaften zu Göttingen.
- Bolt, B. A. "Inside the Earth: Evidence from Earthquakes", W. H. Freeman.
- Dziewonski, A. M. and Gilbert, F. (1971) Solidity of the inner core of the Earth inferred from normal mode observations. Nature, 234, 465–466.
- Inge Lehmann (Wikipedia, in Japanese)
- Shadow zone (Wikipedia, in Japanese)
※This article is a general-audience science explainer. The figures given are rough guides to help you understand the mechanism. Estimates of the depths of the Earth's layers and of wave speeds differ slightly between studies.