How do we know the Earth’s interior is molten?

I’ll give you a hint:

Source: J. D. Griggs, USGS.

Source: J. D. Griggs, USGS.

Need another one?  Here you go:

Source: D. A. Swanson, USGS.

Source: D. A. Swanson, USGS.

The Earth is pockmarked all over with thousands of holes that are literally leaking molten rock.  It’s clear that at least some of the Earth’s interior is molten.  So that’s it, then.  We’ve solved the mystery of the Earth’s molten interior, right?  If only there was more to the story.

There’s totally more to the story.  We’re taught in science classes that the lava we see spewing from volcanoes on the Earth’s surface is just a tiny fraction of the molten rock underground.  Furthermore, we’re told that the ground we’re standing on, the Earth’s crust, is just a few miles thick, like the skin of an apple, with a much thicker mantle underneath that gives way to a molten outer core and a solid iron inner core.  But the deepest hole we’ve ever dug (the Kola Superdeep Borehole in Russia) is only about 7.6 miles deep.  This isn’t even deep enough to penetrate through the crust into the mantle.  So how do we know that the mantle’s even there at all?  For that matter, how do we know that the core of the Earth is solid iron surrounded by a liquid?  How do we figure out the structure of Earth’s interior without probing it directly?

I’ll give you a hint:

Source: Wolfgang Moroder (CC-BY)

Source: Wikipedia user Wolfgang Moroder (CC-BY)

This is an ultrasound image of someone’s kid (maybe Wolfgang Moroder’s?).  What does ultrasound imaging have to do with the structure of the Earth’s interior?  It turns out that scientists can use the same principle used in ultrasound to figure out what’s going on beneath our feet.  To explain how they do this, we first need to discuss the physics of acoustics.

The field of acoustics deals in general with vibrations–in air, in liquid, or in solid objects.  We’re most familiar with acoustics as it relates to sound.  Sound is made up of vibrations that travel through air or other material and are eventually detected by our ears.  We call these vibrations sound waves.  We’ve discussed waves several times before, most frequently in the context of light waves.  Most of the time, waves are waves; the physics is pretty similar, and sound waves have many of the same properties that light waves do.  In particular, sound waves can reflect off objects, they can be focused like a light wave, and they can refract, or bend through other objects, usually when there’s a change in density in the object, as shown below:

Source Wikipedia user Oleg Alexandrov (mercifully, because I'm pretty tired of making the gifs myself).

Source: Wikipedia user Oleg Alexandrov.

Another important property of sound waves is that they can be attenuated, or dampened, by passing through different objects.  This is why, for example, music that we hear through a wall tends to sound muffled.

If we take all of these properties together, we can paint a picture of how ultrasound works:  Basically, an ultrasound machine consists of two parts:   a probe which emits and detects ultrasound waves, and a computer which deciphers the waves that are detected.  The probe focuses the waves to a point.  When the waves encounter a change in density–tissue to bone, tissue to fluid, etc.–they reflect off the boundary, forming an echo which is detected by the probe.  A computer then tabulates the time it takes for the ultrasound to penetrate the tissue and echo off an internal structure and maps all this information out in picture form, giving an image like the one we saw earlier.  The more focused the sound waves, the sharper the picture.

Anyone who’s ever been pregnant or had an ultrasound done for other medical reasons will tell you:  the technician usually pushes that probe in pretty hard.  The reason is that the ultrasound waves are quickly attenuated in human flesh, and in order to see deeper, the technician has to push down harder to get closer to the object of interest.  But eventually, if there’s enough flesh in the way, the attenuation will be too much and the technician will lose the signal.  This is a fundamental limitation of ultrasound:  it turns out that higher pitches–or higher frequencies, to use wave jargon–are attenuated much more quickly than lower pitches.  This is why you can hear the bass in someone’s car 3 blocks away but you can’t hear the person singing:  the lower-pitched bass is hardly attenuated at all, whereas the rest of the music dies off pretty quickly.  But at the same time, lower frequencies correspond to longer wavelengths, and longer wavelengths can’t be focused as well as shorter wavelengths can.  So ultrasound imaging is a tradeoff:  the frequencies are low enough to penetrate deeply into the human body, but high enough to give a reasonably sharp focus.

An ultrasound of the Earth’s core?

So who’s going around giving the Earth an ultrasound?  Unfortunately, attenuation becomes a big factor here:  An ultrasound can barely penetrate a few inches into human flesh; there’s no way it’s going to be able to map out details miles below our feet.  In order to probe the Earth’s interior without attenuation completely scrambling the information encoded in the sound waves, we need to go to lower frequencies.  Much lower.  We have to use what scientists call infrasound.  If ultrasound is made up of frequencies that are too high to hear, then infrasound is made up of frequencies that are too low to hear.  But using infrasound is only half the battle.

The Earth is 12,742 km (7,918 miles) in diameter; that’s a long way for any type of sound to travel.  In order for us to be able to use sound to probe the Earth’s interior, we need for the sound to be really, really loud.  What makes a really loud, really low frequency vibration in the Earth?  I’ll give you a hint:

Source: H. D. Chadwick/National Archives

Source: H. D. Chadwick/National Archives

Ok, I’ll admit that’s a crappy hint.  The answer is earthquakes.  (I couldn’t find a picture of earthquake damage that didn’t look like general mayhem and destruction, and I thought that turning this picture of the 1906 San Francisco earthquake into an animated gif that shakes back and forth would be in poor taste.)  Earthquakes provide gigantic, low-frequency vibrations that reverberate far and wide.  In fact, using an instrument called a seismometer, a powerful enough earthquake can be detected on the opposite side of the globe.  For example, check out this measurement of the 1906 San Francisco earthquake, recorded nine time zones away in Göttingen, Germany:
Source: USGS

Source: USGS

You’ll notice that the waves in the seismograph above are labeled with P’s and S’s.  It turns out that these labels are really important to determining the structure of the interior of the Earth.  P stands for “primary” and S stands for “secondary,” simply because the primary waves are typically detected before the secondary (scientists are nothing if not practical).  But P and S waves are actually fundamentally different.  See, waves can travel through solid rock in two different ways.  P waves are an example of longitudinal waves:

S waves are an example of transverse waves:

(Both gifs above were made by Christophe Dang Ngoc Chan).  If we take a look at the seismograph from Germany, we can deduce that the P waves clearly travel faster than S waves (they get to the seismometer first).  But we also notice that some of the waves are marked PP, SS, PPP, or SSS.  These markings indicate reflected waves.  Earthquake waves naturally bend away from the center of the Earth.  Since the rock is more compressed–and therefore denser–as you go deeper into the Earth, the waves are refracted more and bend back toward the surface, just as light bends when passing from air to water.  But the surface of the Earth acts like a mirror, and the waves bounce off the surface and continue traveling through the Earth:

Based on the time difference between the direct waves and the reflected waves, we can make pretty detailed maps of the paths that these waves take through the Earth, just like with ultrasound.  It is painstaking work to tease out the details of all these different wave patterns, and it was even more difficult before the advent of powerful computer algorithms.  One of the brave souls who took a crack at the problem was a Croatian geophysicist named Andrija Mohorovičić.  In 1909, he showed that earthquake waves were reflected off a boundary about 50 km below the Earth’s surface.  Today, this is called the Mohorovičić discontinuity–or Moho for short–and it marks the boundary between the Earth’s brittle crust and the more pliable mantle underneath.  [Side note:  “Pliable” is a relative term.  Mantle material is every bit as hard as the rocks we encounter every day at the Earth’s surface.  But over geological timescales–hundreds of thousands to millions of years–the mantle bends and flows much more easily than the crust.]

Even before Mohorovičić, though, Richard Oldham (in 1906) had made an odd observation.  If an earthquake happens on one side of the planet, it takes much longer for the P waves to pass straight through the planet than it should if the Earth is all one big undifferentiated ball of rock.  Furthermore, at an angle of about 130°, the P waves are much weaker than expected.  In fact, it’s almost as if the P waves suddenly hit an odd boundary and are refracted deep within the Earth.  To make things even stranger, in 1926 Harold Jeffreys noticed that S waves don’t seem to penetrate the planet’s center at all:

This all seems really weird, but there’s a very simple solution to the problem.  Remember that S waves are transverse waves, while P waves are longitudinal.  It turns out that S waves can only pass through material that has a certain degree of rigidity.  In other words, in order for S waves to propagate, the material must be solid.  S waves can’t pass through liquid; only P waves can, but even they are heavily refracted.  The solution was obvious to the geologists looking at the problem:  Beneath the Earth’s mantle lies a core made up of liquid metal and rock.  The “wave shadow” was direct evidence that there was a huge body of liquid in the middle of the Earth.  The picture got even stranger a few years later when a Danish scientist named Inge Lehmann looked more closely at the P-wave refraction.  She realized that the wave had to be refracted by an additional boundary inside the core, meaning that the core was made up of a liquid outer layer and a solid inner layer.

The discoveries of Mohorovičić, Oldham, Jeffreys, and Lehmann give us the basis for the modern conception of an Earth with many layers, and also answer our question:  How do we know the Earth’s interior is molten?  And as seismometers have become more and more sensitive, our ability to map out the interior features of the Earth has grown explosively.  Like the increasingly accurate ultrasounds used in obstetrics, modern seismology allows us a peek into the hidden world underneath our feet.  The roiling molten rock at the center of the Earth drives the birth and death of whole continents and ecosystems.  It pushes up Hawaiian paradises and Himalayan wonders, while tearing down manmade artifices in violent, fiery spectacles.  It has given way to mass extinctions, but it also gives the planet its magnetic field, protecting us from the Sun’s radiation and making life on land possible.  On timescales we can scarcely imagine, it changes our world in ways we can hardly believe.  We live on top of a truly fascinating place.


This stuff boggles my mind, but unfortunately I’m no expert.  Most of the info in this post comes from a couple of books:

  • The Changing Earth: Exploring Geology and Evolution, by James Monroe, Reed Wicander, which can be found here, and
  • Geological Science, by Andrew McLeish, which can be found here.
  • This site gives a pretty good overview of how waves are reflected and refracted throughout the planet.

Leave a Reply

Your email address will not be published. Required fields are marked *