Although we reside on Earth, we cannot easily probe our planet's interior. Drilling gear can penetrate rock only so far before breaking. No substance used for drilling—even diamond, the hardest known material—can withstand the pressure below a depth of about 10 km. That's rather shallow compared with Earth's 6400 = km radius. Fortunately, geologists have developed other techniques that indirectly probe the deep recesses of our planet.


A sudden dislocation of rocky material near Earth's surface—an earthquake—causes the entire planet to vibrate a little. Earth literally rings like a bell. These vibrations are not random, however. They are systematic waves, called seismic waves (after the Greek word for "earthquake"), that move outward from the site of the quake. Like all waves, they carry information. This information can be detected and recorded using sensitive equipment—a seismograph—designed to monitor Earth tremors.

Decades of earthquake research have demonstrated the existence of many kinds of seismic waves. Two are of particular importance to the study of Earth's internal structure. First to arrive at a monitoring site after a distant earthquake are the primary waves, or P-waves. These are pressure waves, a little like ordinary sound waves in air, that alternately expand and compress the material medium through which they move. Seismic P-waves usually travel at speeds ranging from 5 to 6 km/s and can travel through both liquids and solids. Some time later (the actual delay depends on the distance from the earthquake site), secondary waves, or S-waves, arrive. These are shear waves. Unlike P-waves, which vibrate the material through which they pass back and forth along the direction of travel of the wave, S-waves cause side-to-side motion, more like waves in a guitar string. The two types of waves are illustrated in Figure 7.5. S-waves normally travel through Earth's interior at 3 to 4 km/s; they cannot travel through liquid, which absorbs them.

Figure 7.5 (a) A pressure (P-) wave traveling through Earth's interior causes material to vibrate in a direction parallel to the direction of motion of the wave. Material is alternately compressed and expanded. (b) A shear (S-) wave produces motion perpendicular to the direction in which the wave travels, pushing material from side to side. Also shown is the motion of one typical particle. In case (a) the particle oscillates forward and backward about its initial position. In (b) the particle moves from side to side.

The speeds of both P- and S-waves depend on the density of the matter through which the waves are traveling. Consequently, if we can measure the time taken for the waves to move from the site of an earthquake to one or more monitoring stations on Earth's surface, we can determine the density of matter in the interior. Figure 7.6 illustrates some P- and S-wave paths away from the site of an earthquake. Seismographs located around the world measure the times of arrival as well as the strengths of the seismic waves. Both observations contain much useful information—both about the earthquake itself and about Earth's interior through which the waves pass. Notice that the waves do not travel in straight lines through the planet. Because the wave velocity varies with depth, the waves bend as they move through the interior.

Figure 7.6 Earthquakes generate pressure (P, or primary) and shear (S, or secondary) waves that can be detected at seismographic stations around the world. S-waves are not detected by stations "shadowed" by the liquid core of Earth. P-waves do reach the side of Earth opposite the earthquake, but their interaction with Earth's core produces another shadow zone, where no P-waves are seen.

A particularly important result emerged after numerous quakes were monitored several decades ago: seismic stations on the side of Earth opposite a quake never detect S-waves—these waves are blocked by material within Earth's interior. Further, while P-waves always arrive at stations diametrically opposite the quake, there are parts of Earth's surface that they cannot reach (see Figure 7.6). Most geologists believe that S-waves are absorbed by a liquid core at Earth's center and that P-waves are "refracted" at the core boundary, much as light is refracted by a lens. The result is the S- and P-wave "shadow zones" we observe. The fact that every earthquake exhibits these shadow zones is the best evidence that the core of our planet is hot enough to be in the liquid state. Interlude 7-1 presents some more seismic data on the structure of Earth's core.

The radius of the core, as determined from seismic data, is about 3500 km. In fact, very faint P-waves are observed in the P-wave shadow zone indicated in Figure 7.6. These are believed to be reflected off the surface of a solid inner core, of radius 1300 km, lying at the center of the liquid outer core.


Because earthquakes occur often and at widespread places across the globe, geologists have accumulated a large amount of data about shadow zones and seismic-wave properties. They have used these data, along with direct knowledge of surface rocks, to build mathematical models of Earth's interior.

Figure 7.7 presents a model that most scientists accept. Earth's outer core is surrounded by a thick mantle and topped with a thin crust. The mantle is about 3000 km thick and accounts for the bulk (80 percent) of our planet's volume. The crust has an average thickness of only 15 km—a little less (around 8 km) under the oceans and somewhat more (20 to 50 km) under the continents. The average density of crust material is around 3000 kg/m3. Density and temperature both increase with depth. Specifically, from Earth's surface to its very center, the density increases from roughly 3000 kg/m3 to a little more than 12,000 kg/m3, while the temperature rises from just under 300 K to well over 5000 K. Much of the mantle has a density midway between the densities of the core and crust: about 5000 kg/m3.

Figure 7.7 Computer models of Earth's interior imply that the density and temperature vary considerably through the mantle and the core. Note the sharp density discontinuity between Earth's core and mantle.

The high central density suggests to geologists that the inner parts of Earth must be rich in nickel and iron. Under the heavy pressure of the overlying layers, these metals (whose densities under surface conditions are around 8000 kg/m3) can be compressed to the high densities predicted by the model. The sharp density increase at the mantle—core boundary results from the difference in composition between the two regions. The mantle is composed of dense but rocky material, compounds of silicon and oxygen. The core consists primarily of even denser metallic elements. There is no similar jump in density or temperature at the inner core boundary—the material there simply changes from the liquid to the solid state.

The model suggests that the core must be a mixture of nickel, iron, and some other lighter element, possibly sulfur. Without direct observations, it is difficult to be absolutely certain of the light component's identity. All geologists agree that much of the core must be liquid. The existence of the shadow zone demands that (and, as we will see, our current explanation of Earth's magnetic field relies on it). However, despite the high temperature, the pressure near the center—about 4 million times atmospheric pressure at Earth's surface—is high enough to force the material there into the solid state.

Despite the fact that no experiment has yet succeeded in piercing Earth's crust to recover a sample of the mantle, we are not entirely ignorant of the mantle's properties. In a volcano, hot lava upwells from below the crust, bringing a little of the mantle to us and providing some inkling of Earth's interior. The chemical makeup and physical state of newly emerged lava are generally consistent with predictions based on the model sketched in Figure 7.7.

The composition of the upper mantle is probably quite similar to the iron—magnesium—silicate mixtures known as basalt. You may have seen some dark gray basaltic rocks scattered across Earth's surface, especially near volcanoes. Basalt is formed as mantle material upwells from Earth's interior as lava, then cools and solidifies. With a density between 3000 and 3300 kg/m3, basalt contrasts with the lighter granite (density 2700—3000 kg/m3) that constitutes much of the rest of Earth's crust. Granite is richer than basalt in the light elements silicon and aluminum, which explains why the surface continents do not sink into the interior. Their low-density composition lets the crust "float" atop the denser matter of the mantle and core below.


Earth, then, is not a homogeneous ball of rock. Instead, it has a layered structure, with a low-density crust at the surface, intermediate-density material in the mantle, and a high-density core. Such variation in density and composition is known as differentiation.

Why isn't our planet just one big, rocky ball of uniform density? The answer appears to be that much of Earth was molten at some time in the past. As a result, the higher-density matter sank to the core, and the lower-density material was displaced toward the surface. A remnant of this ancient heating exists today: Earth's central temperature is nearly equal to the surface temperature of the Sun. What processes were responsible for heating the entire planet to this extent? To answer this question, we must try to visualize the past.

We will see in Chapter 15 that when Earth formed, it did so by capturing material from its surroundings, growing in mass as it swept up "preplanetary" chunks of matter in its vicinity. As the young planet grew, its gravitational field strengthened and the speed with which newly captured matter struck its surface increased. This process generated a lot of heat—so much, in fact, that Earth may already have been partially or wholly molten by the time it reached its present size. As Earth began to differentiate and heavy material sank to the center, even more gravitational energy was released, and the interior temperature must have increased still further.

Later, Earth continued to be bombarded with debris left over from the formation process. At its peak some 4 billion years ago, this secondary bombardment was probably intense enough to keep the surface molten, but only down to a depth of a few tens of kilometers. Erosion by wind and water has long since removed all trace of this early period from the surface of Earth, but the Moon still bears visible scars of the onslaught.

A second important process for heating Earth soon after its formation was radioactivity—the release of energy by certain rare, heavy elements, such as uranium, thorium, and plutonium (see More Precisely 7-2). These elements emit energy as their complex, heavy nuclei decay into simpler, lighter ones. While the energy produced by the decay of a single radioactive atom is tiny, Earth contained a lot of radioactive atoms, and a lot of time was available. Rock is such a poor conductor of heat that the energy would have taken a very long time to reach the surface and leak away into space, so the heat built up in the interior, adding to the energy left there by Earth's formation.

Provided that enough radioactive elements were originally spread throughout the primitive Earth, rather like raisins in a cake, the entire planet—from crust to core—could have melted and remained molten for about a billion years. That's a long time by human standards, but not so long in the cosmic scheme of things. Measurements of the ages of some surface rocks indicate that Earth finally began to solidify roughly a billion years after it originally formed. Radioactive heating did not stop after the first billion years, however. It continued even after Earth's surface cooled and solidified. But radioactive decay works in only one direction, always producing lighter elements from heavier ones. Once gone, the heavy and rare radioactive elements cannot be replenished.

The early source of heat diminished with time, allowing the planet to cool over the past 3.5 billion years. In so doing, it cooled from the outside in, much like a hot potato, since regions closest to the surface could most easily unload their excess heat into space. In this way, the surface developed a solid crust, and the differentiated interior attained the layered structure now implied by seismic studies.