Structure and composition of the Earth

The Earth is a sphere with a mean radius of 6371 km, built of nested shells. Scientists first described the shells by composition. On the outside is a thin stony crust. Beneath it, down to a depth of 2891 km, is the mantle, made of denser silicate rocks (rocks built from minerals that contain silicon and oxygen). At the centre is the core, made of iron with nickel. Its outer part is liquid and its inner part is solid.

The second division is by mechanical properties, that is, by how strong the rock is. The crust together with the upper layer of the mantle forms the rigid lithosphere. It behaves as one elastic body and is broken into plates. Below it lies the asthenosphere — the hot, and therefore soft, part of the mantle over which the plates slowly move. The boundary between crust and mantle is drawn where the rocks change. The base of the lithosphere is drawn where the strength changes, and that depends mainly on temperature. So these are two different surfaces, and the base of the lithosphere lies much deeper than the base of the crust.

Temperature increases with depth. At the base of the continental crust it is usually 400–800 °C. At the base of the lithosphere it is about 1300 °C, at the core–mantle boundary about 3500–4000 °C, and at the edge of the inner core and at the centre about 5000–6000 °C. The deep values are estimates from laboratory experiments on materials at high pressure and from calculations. They are uncertain by hundreds of degrees. The pressure at the centre is about 364 GPa, which is 3.6 million times atmospheric pressure.

The crust is of two types, and they differ in almost everything. Oceanic crust is built of basalts and gabbros — dark rocks rich in iron, magnesium and calcium. It is thin, dense and young. Mid-ocean ridges constantly create it, and it is swallowed in zones where one plate dives beneath another. So almost no sea floor older than about 200 million years remains on Earth. Continental crust is much thicker. Its upper part is close to granite in composition, and its lower part is close to darker, denser rocks. It is lighter than oceanic crust and can be very old — up to 4 billion years.

Almost half the mass of the crust is oxygen, more than a quarter is silicon, and about 8% is aluminium. Then come iron, calcium, sodium, potassium and magnesium. These eight elements make up about 98% of the mass of the crust. That is why the main minerals of the crust are silicates, built from silicon and oxygen (Minerals and rocks). The Earth as a whole is built differently: about a third of its mass is iron, and almost all of that iron has gone into the core.

The base of the crust is called the Mohorovičić discontinuity (a sharp boundary across which rock properties change), or simply the Moho. Across it the velocity of compressional seismic waves jumps from about 6.5–7 to 8 km/s. The Croatian seismologist Andrija Mohorovičić discovered it in 1909 from earthquake records. At distances of hundreds of kilometres from the source of the earthquake, the waves that arrived first were the ones that had travelled through the faster rocks of the mantle (see the advanced section).

In places the crust is covered by a sedimentary cover: a thick sequence of sedimentary rocks (rocks formed from layers of sand, mud and other sediment) that built up in low areas of the Earth's surface. Beneath it lies the basement, made of igneous rocks (rocks that cooled from molten rock) and metamorphic rocks. On ancient shields there is no cover, and the basement is exposed at the surface. In deep sedimentary basins the cover reaches 10–20 km. Almost all oil and gas accumulations lie in the sedimentary cover. It contains organic-rich rocks (the source of the oil), porous reservoirs (which hold it) and seals (which trap it) (Petroleum system and the play concept).

The mantle makes up about 83% of the volume of the Earth and two thirds of its mass. Its upper part is made of peridotite, a dark, dense rock that consists mainly of the minerals olivine and pyroxene. You can see such rocks at the surface as fragments carried up by volcanoes and in slabs of mantle squeezed up when plates collide. With depth, pressure rearranges the crystal structure of the minerals into denser forms. At depths of about 410 and 660 km this causes jumps in density and wave velocity. These jumps identify the transition zone, the layer between those two depths.

The mantle is solid (shear waves pass through it), but hot rocks under load creep slowly, like ice in a glacier. Over tens of millions of years, convection stirs the mantle material at speeds of the order of centimetres per year. Convection is a slow circulation: hot material rises and cooler material sinks. The lithospheric plates move at the same pace (Plate tectonics and sedimentary basins).

The core is iron with some nickel (about 5%) and a few per cent of light elements — oxygen, silicon, sulfur and others. We know the light elements are there because pure iron at such pressures would be roughly 5–10% denser than the seismic data show. The outer core is liquid. Currents in it act like a huge dynamo and create the Earth's magnetic field. The inner core is solid, although it is hot: the melting temperature of iron rises with pressure, and at the centre it is higher than the actual temperature.

The simplest evidence for a heavy core is the mean density of the Earth. We find the mass of the planet from its gravitational pull (from the orbits of the Moon and of satellites) and its volume from its size. The ratio gives 5.51 g/cm³. The rocks at the surface are only half as dense, and even the compressed silicates of the lower mantle are no denser than 5.6 g/cm³. So the centre must contain material far denser than any rock.

Pressure in the interior comes from the weight of the overlying shells, in the same way as overburden pressure in a sedimentary section. A column of crust 40 km high, with a mean density of 2.8 g/cm³, presses on its base with a pressure $\rho g z \approx 1.1$ GPa — about 11 thousand atmospheres. At a depth of 660 km the pressure is about 23 GPa, at the core boundary 136 GPa, and at the centre about 364 GPa. The acceleration due to gravity is almost constant throughout the mantle, at about 10 m/s². It decreases to zero only inside the core: the mass lying above pulls in all directions and adds no weight.

The lithosphere is the crust plus the upper, comparatively cold part of the mantle. It is rigid: a load makes it bend or break along faults, not flow. Its base is usually put at a temperature of about 1300 °C. Deeper than that, the mantle rocks are so close to melting that over geological time they flow even under small stresses.

The thickness of the lithosphere varies more than the thickness of the crust. At mid-ocean ridges, where hot mantle material rises, it is only a few kilometres thick. With distance from the ridge the oceanic lithosphere cools and thickens, roughly in proportion to the square root of its age, up to 100 km under old sea floor. Under young rifts (zones where the crust is pulling apart) and fold belts (mountain belts built by folding), the continental lithosphere is thin and hot. Under ancient shields it is cold and reaches 200–250 km.

Beneath the lithosphere lies the asthenosphere. In this layer seismic waves travel more slowly and the rocks are weaker. Under the oceans the layer is most pronounced at depths of roughly 80–200 km. The cause is high temperature and, possibly, a small fraction of melt between the grains. But on the whole the asthenosphere is solid: it lets shear waves pass, and it flows only over thousands to millions of years. That is why the lithosphere can “float” on it (see the section on isostasy).

Nobody has been able to drill deeper than the first 12 km of the Earth. So we work out the structure of the Earth indirectly, and the main source of information is seismic waves. An earthquake or an explosion sends two types of elastic waves through the rocks. In a compressional wave (P-wave, from “primary”) the particles move back and forth along the direction the wave travels, so the rock is alternately squeezed and stretched. In a shear wave (S-wave, from “secondary”) the particles move sideways, at right angles to that direction, so the rock is sheared. The compressional wave is faster and arrives first.

At a boundary between rocks with different velocities, a wave is partly reflected and partly refracted (it changes direction, like a ray of light at the surface of water). Seismic stations around the world record when the waves from thousands of earthquakes arrive. From how the travel time depends on distance, scientists work out how velocity changes with depth. This is how the main boundaries were found:

• The Moho (1909) — from the bend in the time–distance graph. At distances of hundreds of kilometres from the source, the first arrivals are waves that have travelled through the fast mantle.
• The core boundary (1914, Beno Gutenberg) — at about 2900 km. In the core, compressional waves slow down sharply and bend, so at epicentral distances (angles at the Earth's centre, measured from the point above the source) of roughly 103° to 142° there are no direct P-waves. This is the shadow zone.
• The liquid outer core — shear waves do not pass through it, so beyond about 103° from the source there are no direct S-waves.
• The inner core (1936, Inge Lehmann) — from weak P-waves that nevertheless arrive in the shadow zone, because a boundary inside the core deflects them. Later, the free oscillations of the Earth after very large earthquakes (the whole planet rings like a bell) confirmed that the inner core is solid.

Why do the continents stand above sea level while the ocean floor lies 4–5 km deep? Why do mountains not punch down through the crust? The answer is isostasy: the balance of the light lithosphere floating on the denser, softer mantle. Think of blocks of wood floating in water. The higher a block rises above the water, the deeper it sinks into it.

The condition for balance is this. At a certain compensation depth — below the deepest roots, in the soft mantle — the pressure beneath every area is the same. This means that rock columns of equal cross-section weigh the same. If the load grows somewhere, the column sinks and pushes the mantle sideways until the balance returns. If the load is taken away, the column rises.

The same principle explains why sedimentary basins can be so deep. Suppose a basin is filled with water to a depth $w$ and sediments begin to arrive. Each metre of sediment is heavier than the water it pushes out, so the floor beneath it sinks. While sediments keep arriving, the basin keeps getting deeper. In the end the sediments are 2–3 times thicker than the initial water depth.

Isostasy is an approximation for large areas. The strength of the lithosphere holds up a single mountain, a volcanic island or a narrow fold. The lithosphere bends like a plank under a weight and spreads the load over hundreds of kilometres around. Departures from balance show up in gravity anomalies (places where gravity is stronger or weaker than expected). Where the balance was disturbed recently, it is being restored before our eyes. Since the ice sheets of the last glaciation melted, the shores of the Gulf of Bothnia and Hudson Bay have been rising by up to a centimetre a year. Exploration also uses gravity anomalies. For example, salt is lighter than the surrounding rocks, and over salt domes gravity is usually lower (Geophysical exploration methods).

The Earth is slowly cooling. In the lithosphere, heat travels to the surface by conduction (passing through solid rock without the rock moving). In the mantle and the liquid core it travels mainly by convection. There are two sources of heat: heat stored when the planet formed, and heat from the radioactive decay of uranium, thorium and potassium-40. These are the same elements whose radiation the gamma ray log records. By current estimates, the two sources give comparable shares of the total flow. The radioactive elements are especially concentrated in the continental crust.

The measure is heat flow $q$: the heat that passes through one square metre of surface in one second, expressed in milliwatts per square metre. Scientists find it from temperature measurements in wells and from the thermal conductivity of core (how well the rock conducts heat). Heat flow equals thermal conductivity multiplied by the geothermal gradient (the rise of temperature with depth). On average about 65 mW/m² passes through the continents and about 100 mW/m² through the ocean floor. In all, the Earth loses about 46 TW. This is several thousand times less than it receives from the Sun, so the interior hardly warms the weather and climate.

The temperature gradient in each layer equals the heat flow divided by the thermal conductivity of the rock, $G_T = q/\lambda$ (example with a layer of salt). At a heat flow of 65 mW/m² and a thermal conductivity of 2.5 W/(m·K), this is 26 °C/km, a typical near-surface gradient. But you cannot extend it downwards. The radioactive elements scattered through the crust release heat themselves, so less heat flows through the deep layers than through the near-surface ones. The gradient therefore decreases with depth.

• The sedimentary cover is where the accumulations are. Oil and gas form from the organic matter in sediments and collect in sedimentary reservoirs beneath seals. So exploration starts with a basin that has a thick enough cover (Petroleum system and the play concept). The basement is a reservoir only where it is fractured or weathered (Minerals and rocks).
• Heat flow sets maturity (how far organic matter has turned into oil). It decides the depth at which a source rock enters the oil window (the temperature range in which it makes oil): shallower in a hot basin, deeper in a cold one (Source rocks, maturation and migration). It also sets the reservoir temperature, and with it the properties of the fluids (Reservoir pressure and temperature).
• Subsidence (sinking) creates room for sediments, while uplift and erosion bring previously buried rocks closer to the surface. These rocks are more compacted and more heated than their present depth would suggest.
• The weight of the rocks and the stresses in the lithosphere decide the overburden pressure, the fracture pressure (the pressure at which the rock splits) and wellbore stability (Pore pressure and overburden pressure, Fracture gradient and leak-off testing, Rock mechanics for hydraulic fracturing). The rigid lithosphere carries stresses from plate boundaries over thousands of kilometres. These stresses set the preferred directions of natural fractures and hydraulic fractures.
• Seismic waves are the main exploration tool. A seismic survey probes the sedimentary cover with the same compressional and shear waves that probe the whole globe. It uses frequencies of tens of hertz and reaches depths of a few kilometres (Geophysical exploration methods).