The Earth is alive. Well, not really. At least it is not life as we know it. Appropriately, the origin of this expression – slightly rephrased from the original Star Trek series – is from the second-most-famous non-human in existence, Spock, referring to a living rock. It is only a small step for humankind to apply it to an entire planet.
(What Spock actually said was ‘there is no life, other than the accountable human residents of this colony beneath the surface. At least, no life as we know it.’ In another episode Spock comments about a strange creature ‘It is not life as we know or understand it. Yet it is obviously alive’. The version that is always quoted is in reality from the 1986 song Star Trekkin’. But I digress.)
To be considered ‘alive’ an organism needs to meet some requirements (as defined by the powers-that-be): the ability to adapt to the environment, to respond to stimuli, to actively regulate its internal conditions, to produce energy by metabolism, to grow and reproduce. The Earth doesn’t do any of this. It is big enough not to have to worry about it. But speed up time, and you can see the Earth come to life. Continents drift, collide and separate; oceans grow and wane; mountains rise and fall. Volcanoes sprout, grow and die. Plates subduct, refresh the Earth inside and the enriched mantle resurfaces as new oceanic crust. This is the Earth breathing – if you can imagine breathing solid rock. There is a rhythm to the Earth. It dances to an inner beat.
Where there is a beat, there is a heart. This story is about the search for the Earth’s heart.
To find the heart, we need to look below the surface. From the outside to the core, our surgical cut reveals the crust (and/or lithosphere, if you prefer), the upper mantle, the lower mantle, the outer core and the inner core. Each has a different role to play.
The iron core acts as the Earth’s skeleton. It consists of two parts, a solid, stable inner core and a liquid outer core in which the liquid moves around in huge convecting bubbles. As a byproduct these bubbles create the Earth’s magnetic field, which immunizes us from space weather. This protective role can perhaps be compared to that of bone marrow. Of course it consists of iron and is outside rather than inside the solid part, but you get the idea. Apart from breeding the magnetic field, the core does not affect us: like our skeleton, it stays inside to provide stability.
(The inner core is not entirely uniform. There are hints that the composition slowly changes with depth. This is sometimes described as a separate innermost core, but this seems to overstate things as there is no obvious transition layer.)
Beyond the core lies the mantle, the main body of the Earth which ferries the heat from the centre to the surface. It consists of rock rather than iron. At the bottom of the mantle lies a thin region (no more than 150 km thick) named the D″ (pronounced D-double-prime) layer which encases the core. Blame mathematics for the strange name. Originally the entire mantle was called the D layer. When seismology showed the mantle to have subdivisions, the geo-mathematicians called these D, D′ (D-prime) and D″ (D-double-prime). The first two are now known as the upper and lower mantle, but no one wanted the last (tiny) layer and so its name survived. The D″ layer belongs to the mantle but the precise composition is unknown. It has been suggested to be partly melted, or to have an iron-enriched composition, or to be a denser phase of perovskite (found in the lower mantle) called post-perovskite. Seismology shows that D″ layer is variable both in thickness and in properties; in a way it is almost similar in structure to the crust of the Earth with valleys and mountains and different rocks. This is the most mysterious, and least studied, of all the layers of the Earth. I am not sure which body part to compare it to! Perhaps the equally mysterious but controlling amygdala will do. Or may be this gives it too much credit, and it is more like our coccyx in looking interesting but not doing anything useful. We’ll come back the mantle, as it is the main topic of this post.
Beyond the mantle lies the crust, which separates into the asthenosphere (a thin lubrication layer on which the crust can move), the oceanic crust and the continental crust. The oceanic crust is slightly processed mantle material with some external pollution such as water, carbon and shipwrecks. The continental crust is low-density flotsam which resolutely stays on top and therefore is well-placed to carry our cities – you might call the continents the Earth’s warts. In some places the crust is stacked, with continental crust lying on top of an oceanic crust, but most locations have either oceanic plates (with oceans, obviously) or continental plates (which can still be partially submerged). Continental crust is a passive on-looker in the rhythm of the Earth and is of little interest to us apart from providing a good place to call home; only the oceanic crust is invited to the inner party.
The part of the mantle that isn’t the D″ layer (i.e. almost all of it) consists of two parts, the lower mantle and the upper mantle. They differ in the structure of their rock. Both consist mostly of magnesium-silicates (peridotites), but the elements are arranged in different mineral structures. At the top of the mantle it forms olivine, which becomes compressed into ringwoodite a bit deeper in the upper mantle, into bridgmanite (and related minerals) in the upper part of the lower mantle and to perovskite in the deeper part of the lower mantle. If you now feel a bit lost in the nomenclature, join the club. Geology is even more opaque than medicine in its language. The terminology of geology has turned the Earth into a veritable tower of Babel. (More on that in part II.)
The Babel fish (the guide says “the poor Babel fish, by effectively removing all barriers to communication between different races and cultures, has caused more and bloodier wars than anything else in the history of creation“) will tell you that ringwoodite (found in the deeper part of the upper mantle) is a spinel. If that doesn’t help much, spinels are minerals with a (rather complex) pyramid-like structure of their atoms. Spinels are named after one particular mineral which is also called spinel. There is a gem with the same name. The gem spinel is a variant from the mineral spinel which belongs to group of the spinel. What’s in a name? In geology, confusion. I actually like spinels because one of these minerals is called (rather inaccurately) galaxite – a name that promises much more than it delivers.
We now need to make a small step (for humankind) from geology to geometry. The picture below shows the structure of a spinel. The large orange dots are magnesium or silicon atoms, encased by four red dots (oxygen atoms) arranged as a four-sided pyramid, also known as a tetrahedral structure (tetra meaning four). (You can do the translation without even a babel fish: ‘tetrahedal’ just means four-sided. You just earned yourself a geology Babel fish diploma.) These pyramids are linked by blue dots (calcium or aluminium atoms) which are arranged in an eight-sided structure called an octahedral structure. (You can do the translation.) The two structures are shown separately in the second frame (right, with different colours – sorry). The oxygen atoms are shared in a complex way. There are some variations but this is the basic structure of a spinel. It requires four oxygen atoms for each magnesium or silicon atom.
Olivine (the upper part of the upper mantle, before it becomes pressurised into a spinel in the lower part of the upper mantle) (still with me?) is quite similar but the octahedes (I am not sure that that is an existing word) are linked to each other at only one corner rather than two. You can create it by taking the image of spinel and shifting one of the two blue octahedes (and the attached orange pyramids) one step towards the top right. This leaves one connecting point between the blue shapes, and the blue octapy (a word that should exist) instead now share one edge with an orange pyramid.
Bridgmanite (found in the lower mantle) is a perovskite. Instead of the tetrahedral (pyramid) structure, it has only 3 oxygen atoms for each silicon or magnesium atom. A quick try with the lego box shows that this should form flat structures. But by linking these, you can make a linear chain of pyramids attached at one corner, with the silicon or magnesium atom slightly above the triangle of oxygen atoms. This is more compact overall than the structure of a spinel, and therefore pyroxenes take up less volume and consequently have a slightly higher density. This form is called pyroxene. But it is not quite bridgmanite.
At the pressure of the lower mantle, the pyroxene structure changes to that of a perovskite, compressed even further. Now the big atoms (green in the depiction, for instance silicon) line up in a cube, surrounded by smaller atoms at the corners of the cube (blue, for instance magnesium or calcium) and oxygen atoms at the centre of each edge line. This is repeated in each direction. Some math shows that there are three oxygens for each silicon, the same as for pyroxenes. It is a bit more dense than that. This is the structure of bridgmanite, which is found throughout the lower mantle.
That is enough for geometrics. Play time is over, and we need to get back to work.
After this excursion to Legoland, let’s return to the mantle. The division between the upper and lower mantle is where ringwoodite (a spinel) compresses into pyroxene and subsequently a perovskite. This change happens at a depth of around 660 km. It requires energy and therefore the change occurs a bit deeper for colder material, and a bit less deep for hotter material. We now have the funny situation that at this depth, there is a mix between cold, lower density material and hot, higher density material. The colder material is more buoyant. This makes it harder for warmer material from the lower mantle to mix into the cooler upper mantle and vice versa. The two layers have a degree of separation: they are kept apart by a buoyancy barrier.
The mantle rock circulates in large convection currents, powered by the heat below: heated rock moves up, cools, and comes down again. It is different from the convection in the core: core convection happens in liquid matter, whilst the mantle is solid. Mantle convection is therefore very slow, typically less than 1 meter per year. (Our garden snails would be embarrassed.) Rock is a reluctant mover! But this convection runs into the 660 km transition zone and has difficulty crossing this. In effect there are two separate convection zones, one in the upper mantle which circulates oceanic crust, and one in the lower mantle which circulates material from near to the core-mantle boundary up through the lower mantle. Only the hottest rocks manage to break through the barrier on the way up, and only the coldest (oldest) subducting plates manage to pass it downward. Otherwise, rarely the two do meet.
The subduction of oceanic rock is a bit like the Earth breathing (taking breaths of rock), while the circulation of the lower mantle is like the blood being pumped around.
The picture shows a cross section of the Earth. This particular cut runs through both poles, Africa and the Pacific and is chosen to highlight the life of the mantle. Note that this is a cartoon, and may deviate from reality.
At the top left the cut cuts through the Aleutian subduction zone. Here oceanic crust is descending into the upper mantle as the Earth takes a breath. The subducting crust is shown in dark blue, reflecting its lower temperature. (It has spend the last 100 million years being cooled by the ocean.) The subducted ex-crust has collected near the transition zone between the upper and lower mantle. As mentioned above, it is held up by the denser and warmer material around it, which has phase shifted to bridgmanite while the subducting crust is still the lighter ringwoodite. However, the cartoon depicts that the subducting material has managed to break the barrier and enter the lower mantle. Once that happens, it will descend rapidly (rapid for a rock moving through rock – our well-fed snails would still have rocket speed in comparison), sinking all the way to the bottom of the mantle.At the bottom of the picture (the far south), the cross section cuts though another subduction zone. This is most likely meant to be the subduction zone at the South Sandwich Islands (although in that case the cross section does not go through the south pole itself). That region with its 7-km deep deep trench is shown here: click on the image for higher resolution. The cartoon shows this subduction zone to also have broken the barrier and sunk to the bottom of the mantle.
The red colours in the cartoon represent warmer matter that is rising from the deep mantle. There are two big warm plumes in the lower mantle, one underneath the tropical Pacific, and the other underneath Africa. Both are shown to have reached the transition region, but have pooled there as the plumes are unable to break through this barrier. But thin plumes are shown to reach the surface, some coming from those plume heads and some starting much deeper. These rise up underneath Afar, Reunion, Louisville and Hawai’i. The secret lies in the temperature: to break through the barrier, the rising matter needs to be some 100C warmer than the surrounding mantle. Such hotter plumes form thinner, faster rising columns.
However, this is just a cartoon! It shows the general idea but it is not meant to be accurate in details of the internal structure. There will be a deeper discussion (pun intended) in part II.
The dashed white lines show the movement in the lower mantle, downward around the subductions and upward at the large plumes. The blue dashed lines (rather harder to see) show a similar movement in the upper mantle. In the upper mantle, this comes up at the spreading ridges, flows outward below the oceanic crust and goes down and bends back where the oceanic crust subducts. The general movement in the upper and lower mantle do not connect to each other. They are separate circulations.
How does it work in practice? Subduction is a bit more complex than depicted above. Ocean floor is almost mantle material but it is slightly altered. The densest mantle minerals have more difficulty surfacing: the other materials melt a bit and trickle up, leaving the densest minerals behind. That is the reason why the ocean floor floats and doesn’t immediately sink back. But as the ocean ages, it slowly loses its mantle heat to the cold water. The slab that makes up the oceanic crust, perhaps 10 km thick, cools. As it cools it contracts a bit and becomes a bit more dense. Initially this is fine, but at some point it becomes as dense as the mantle below. The upper mantle thinks ‘gotcha’ and the oceanic crust sinks back down. Subduction begins.
(There are variations on this theme. Different parts of the crust can have different temperatures and densities, and when they collide the lowest density wins and ends up on top. The other one is caught between the upper crust and the mantle but is too young and warm to sink. This is called ‘flat slab subduction’ and it happens underneath part of North America and in the west Pacific. Eventually the slab will still begin to sink – you can’t avoid old age by hiding.)
As it sinks, the slab remains colder than the surrounding mantle. It takes a very long time to heat a 10-km thick stone slab. So it keeps sinking, until it hits the boundary with the lower mantle. The lower mantle is denser, so the slab stops in its tracks. The slabs may spend time hanging around this boundary, at the so-called slab graveyard. But many do manage to eventually cross into the Hades of the lower mantle and once they do, they will sink all the way to the bottom of the mantle, finally hitting the D″ boundary. This part is reasonably well understood.
The reverse process, where parts of the mantle rise into plumes and hot spots, is much less well understood. In fact, there is no particular reason why plumes should exist. If you drop a stone in a bucket of water the water level will rise in response, but it does so everywhere: it doesn’t form a plume in one location which channels the rising water. The fact that we see rising plumes in the mantle means that the mantle is not uniform. It is thought that the plumes begin when a part of the lower mantle heats up, expands a bit and becomes less dense. It is now buoyant and begins to rise. The rise again has difficulty crossing the boundary between the lower and the upper mantle, so it typically gets stuck there. Only some hot and narrow plumes overcome this barrier and make their way to the surface.
How do we know this? Even the upper mantle is tens of kilometers below the surface. We have never been able to drill that deep. The rock in the bottom half of the continental crust is too soft for a drill to gain purchase. The Mars lander ran into this very problem (‘running’ might be a bit overstating for a completely stationary robot), when it failed to get its probe into the unexpectedly sticky ground. The robot arm tried to push the probe into the ground, but in vain. It makes you realize what a human on Mars could have achieved: they could have forced it in, with the age-old motto that a good kick solves many mechanical (and electrical) problems. The Mars lander should have brought a human for the essential kick. And in fact, a kick is just what is needed to create a viewing window into the Earth.
The medical world has already solved the problem how to look inside a body without having to open it up. Ultrasound echoes have become an essential part of any pregnancy. The same technique can be used to map the inside of the Earth. Of course the Earth is rather bigger than a typical human body, so we need infrasound rather the ultrasound: infrasound consists much larger waves, needed for a larger body. The infrasound is kindly provided by the Earth itself, by giving itself a good kick. This causes an earthquake, and the earthquake produces the infrasound waves.
After a shake, the waves pass through the Earth and reach the surface somewhere elsewhere. From the arrival time of the wave and the distance, you know the speed of the wave inside the Earth. And it turns out, this speed is not always the same. Especially the shear waves often arrive late. These waves are late anyway, as they move slower, but sometimes they are even slower than normal. If you know what determines the speed of this wave (we are talking physics), you can have an idea of the conditions deep below your feet.
What is a shear wave? It is one of the types of sound waves created by an earthquake that travels through the body of the Earth. There are two: the P wave and the S wave, where ‘P’ stands for ‘primary’ (it travels faster and is the first to arrive) and ‘S’ is the ‘secondary’ wave. Still with me? The P wave is a compression wave: material moves forward, pushes the material ahead. This pushes the matter behind back, and itself in turn moves forward and pushes the material ahead of it. This is in essence a sound wave and it is a fast way to travel. The ‘S’ wave, in contrast, moves sideways, pulls the matter ahead to the side while being itself pulled back to starting point, etcetera. This is a shear wave (‘shear’ for sideways movement) and it is slower because it tries to go forward by moving in the wrong direction.
P waves travel through any type of material. But S waves can only travel through solids. In liquids or gasses, they don’t do anything. So if somewhere on Earth an earthquake is detected, and the P wave arrives but the S wave does not, it means there is a liquid somewhere along the line of travel. That is how we discovered that the core of the Earth is liquid but that the mantle is solid. It is also one way to detect a magma chamber.
Got it? You can easily do this experiment yourself. All you need is some cooperative students (ask the local school for some) and the instructions as in this video:
Waves which arrive in different locations have taken different paths through the Earth. The further from the source, the deeper the wave has traveled. Based on the arrival times close to the source it is easy to calculate when they should arrive at more distant locations. As so often, this gives the wrong answer. They arrive earlier than they should. This shows that deeper waves travel faster.
The figure below gives the velocity versus depth for the crust and upper mantle. The velocities increase with depth in the crust. The asthenosphere shows lower velocity, and at 400 km depth there is a jump in the speed of primary waves which comes from the formation of ringwoodite. Go deeper, and the velocities continue to increase. The secondary waves go about 40% slower, but this factor varies a bit.
For those used to equations, the two velocities are given by two fairly simple relations:
Here ρ is the local density of the rock, G is the shear modulus of the rock (the amount of pressure needed to deform the rock) and K is the bulk modulus (the amount of pressure needed to compress the rock into a smaller volume). A liquid offers no resistance to deformation, so G = 0 and the vs becomes zero: the secondary wave comes to a screeching halt when entering a liquid. A solid is more easily deformed at higher temperature, so G reduces when the temperature increases: both wave velocities decrease in hotter material. But the compressibility, K, is much harder to change. Cold rock, hot rock and liquid rock are all highly resistant to compression. The primary wave can keep a high velocity even where the secondary wave slows down, because K remains high and this term contributes to the primary wave (which is a compression wave) but not to the secondary wave (which is a shear or deformation wave). But don’t evaporate the rock: unlike a solid or liquid, a gas is compressible, and it has a much lower K. So the primary wave is much slower in the atmosphere than it is in water or rock. A material that is easily deformable is also known as ‘ductile’. Making a material hotter makes it more ductile, and melting it makes it completely ductile: the force needed to deform it becomes smaller (and for a liquid goes to zero): this is what the G above is. But is also depends on the structure of mineral. Some minerals (for example graphite) are much easier to deform than others (say diamond). It depends how the atoms are organised. A cubic lattice bends more easily than a less symmetric arrangement. This changes the shear wave velocity, and it gives us a way to study the inside Earth. It is not always possible to say exactly which mineral fits the data (and we don’t always know how each mineral behaves under very high pressure) but the shear waves can be used to show that the minerals are changing at a certain depth. In practice, olivine has the lowest G (easiest to deform), ringwoodite is about 50% stronger (higher G), and perovskite is again 50% stronger than ringwoodite. (The precise values depend on the exact compositions: adding iron makes it weaker.) Bridgmanite (and related minerals) are similar to ringwoodite, and therefore the 660-km transition is not easy to see in seismology.
Still with me? You reached the end of the physics lesson. Now we know enough to dive into the mantle, in search of the heart of the Earth. But that is for another day, in part II.
Albert, June 2022