How The Earth Sheds Its Warmth

The Earth's Interior Is Cooling

The earth is constantly losing its thermal energy by all three of the mechanisms that transport heat from one place to another: conduction, convection, and radiation. Heat is conducted through the solid material—the inner core and the lithosphere (including the crust). It is carried upward in convection currents in the outer core, the mantle, the oceans, and the atmosphere, and it is radiated away into space.

Convection not only transports heat, it brings about the generation of more heat because of the inevitable drag within moving fluids. The resulting feedback makes convection the most complicated of the mechanisms of heat transport. Not surprisingly, the deepest of the convection currents, those in the outer core, are the least understood. The heat causing them comes, in unknown proportions, from several sources: radioactivity, gravitational energy, and the latent heat of freezing released by liquid iron as it solidifies. Heat from the core almost certainly assists the flow of convection currents in the mantle,

Convection Currents The Mantle

Figure 15.1. Tectonic plates separating (on the right), and one subducting below another (on the left). The open arrows show the direction of plate movement. The curved arrows represent the flow of currents in the mantle; the descending current may divide, as shown. Approximate distances, in kilometers, from the earth's center are shown on the right of the figure, which is not to scale. The same figure with added detail appears in figure 15.2.

Figure 15.1. Tectonic plates separating (on the right), and one subducting below another (on the left). The open arrows show the direction of plate movement. The curved arrows represent the flow of currents in the mantle; the descending current may divide, as shown. Approximate distances, in kilometers, from the earth's center are shown on the right of the figure, which is not to scale. The same figure with added detail appears in figure 15.2.

which appear to be aligned with the currents in the core. The connection between core currents and mantle currents is at present unclear, however, because data from the deep inner mantle and the core are hard to obtain and difficult to interpret. Geophysicists and seismologists at work on the problem are testing a number of models to help them discover exactly what is happening thousands of kilometers down in the depths of the earth.

More is known about the convection currents in the outer mantle, which are coupled to the movement of the lithosphere's tectonic plates. The discovery of continental drift, as it used to be called, was one of the great advances of the twentieth century. With it came the realization that the pattern of the world's continents and oceans, far from being fixed for all time, is constantly changing; our mental image of the world was altered profoundly.

The way plates move is familiar nowadays from countless books and articles. Bear in mind that our focus here is on the energy associated with the movement. Figure 15.1 diagrams what happens. The curved arrows show the near-surface parts of the convection currents that bring about an overturn of the whole mantle, allowing heat to be transported outward, toward the earth's surface; the lower halves of the convection currents are left to the imagination, because their routes are enigmatic.

It is uncertain what keeps the convection currents going. Possibly they start as long, narrow plumes of hot material rising from the depths that, on reaching the base of the lithosphere, spread out and flow horizontally.1 Possibly they start as upwellings sucked up to fill the gaps opening up as the tectonic plates separate. It is unclear whether currents in the mantle drag the floating plates or whether the drift of the plates keeps the mantle currents flowing.2 In any case, the convection currents flow. A complete overturn of the mantle is believed to take hundreds of millions of years.3 The power needed to drive plate motion is probably on the order of 1 trillion watts, or 1 terawatt.

The plates are believed to have been moving ever since the lithosphere first hardened. In Archean time, more than 2.5 billion years ago, when the mantle was hundreds of degrees hotter than it is now and its outermost layer (the as-thenosphere) correspondingly less viscous, the plates must have moved faster, perhaps ten times faster, than at present.4 The speed of the motion differs from plate to plate; the maximum speed at present is thought to be between 15 and 20 cm per year.

Hot, soft rock oozes up from the asthenosphere to fill the spreading gap wherever plates pull apart, notably at the midocean ridges. The upwelling molten rock (magma) cools and congeals and becomes part of the trailing edges of the retreating plates on both sides of the gap. It becomes new seafloor; the process is known as seafloor spreading. Because of the way new seafloor is continuously created, the seafloor near each midocean ridge on each side of it is younger and hotter than that farther away. The oldest and coolest part of a drifting plate is at its leading edge. The cool rock is dense, ready to sink into the mantle.

When two moving plates meet, one is driven downward and subducts below the other. If this happens in the ocean, a deep trench forms in the ocean floor where the edge of the subducting plate slopes steeply downward. The subducting slab travels with the descending part of the convection current. Studies of the speed of seismic waves traveling through the earth by different routes show that subducting slabs do not all follow the same path: some level off quite soon, about 700 km down, where the mantle rock becomes more viscous; others sink to much greater depths, down to the core-mantle boundary. A slab has been detected in the mantle below Siberia, at a depth of 2,800 km, that is estimated to be 200 million years old; it must have been in existence since very early in the time of the dinosaurs.5

It appears that the tectonic plates are pushed away from the midocean ridges where molten rock wells up; at the same time they are pulled toward the ocean trenches by the sinking of the cool, dense rock of the plates' leading edges—the subducting slabs. Two forces seem to be shifting the plates, which raises the question, Which is the stronger force? Are the plates pushed or pulled?

Research strongly suggests that the chief force is the pull of the subducting slabs.6 It follows that "plate tectonics is a primary result of a cooling earth,"7 because it is cooling that makes the slabs dense enough to sink, exerting a strong pull as they do so. The temperature difference between the newly formed, young crust close to the active parts of the midocean ridges and old crust about to subduct is sometimes 1,000°C.8 A huge quantity of heat is evidently dissipated by the plates as they drift across the earth. It is heat they received from the mantle below them.

The surface of the earth is about 29 percent land and 71 percent ocean. Therefore most of the heat is carried away via the oceanic crust; that is, it is conducted through the crust, then convected to the ocean surface and finally radiated to space. The loss is most rapid on the seafloor where very hot, newly formed crust comes into direct contact with cold seawater. This happens along segments of the midocean ridges where the tectonic plates are separating most actively. Over time—geological time—different segments of a ridge become active; from time to time the activity dies down at one place and starts up at another. Submarine volcanoes erupt wherever the activity happens to be most vigorous; lava is extruded through volcanic conduits from kilometers deep in the crust, and cold water floods in; it is quickly heated by the hot rock to temperatures as high as 300 or 400°C (the high pressure at depth raises the boiling point of water even higher than this). The heated water escapes upward through separate vents, pipes, and cracks. In this way local hydrothermal circulation systems are set up within areas of the crust.9 Huge volumes of water are circulated, so much that it takes only about 10 million years—a short period in geological terms—for the whole world ocean to pass through one or another hydrothermal system; this means that the whole ocean has circulated through hot new crust hundreds of times during the earth's lifetime so far.10 Indeed, "the hydrothermal circulation is the cooling radiator ... of the earth's engine."11

An individual hydrothermal system has a limited lifetime; as the seafloor spreads, it is carried away from the midocean ridge where it originated into a cooler region where it dies away. At the same time, a new hydrothermal system establishes itself in a new patch of hot seafloor, close to the ridge.

At the heart of an active hydrothermal system, the hottest water emerges through submarine geysers known as hydrothermal vents, which spew out superheated water with the force of a fire hose. The hot water rises into the surrounding cold seawater in opaque, billowing plumes known as black smokers, which look like the smoke from an oil fire. The color comes from suspended chemicals, mainly sulfides. Cooler vents creates white smokers. Warm water, cool enough not to scald you if you could put a hand in it, seeps out more gently through innumerable fissures.

The seafloor surrounding hydrothermal vents is the home of the hydrothermal vent fauna, a group of invertebrates that live in the earth's most recently discovered (1977) natural ecosystems.12 There are no plants. The animals, many of them belonging to species new to science, live and thrive near the vents, at a safe distance from the scalding water but close enough to benefit from the warmth. The rich supply of sulfides in the water is the energy source for big populations of chemosynthetic sulfur bacteria (see chapter 12). The bacteria are at the bottom of the ecosystem's food chains; at the top are weird creatures such as giant tube worms, giant bivalves (clams and mussels), sea anemones, and eyeless shrimps. Spectacular photos and videos, taken from manned submersibles, have made the hydrothermal vent fauna familiar to a wide public.

Photos have also been taken, without artificial light, using an exceptionally sensitive camera. The photos show the underwater darkness to be relieved by a very faint glow, too dim to be perceived by humans but presumably bright enough to register with the seemingly blind shrimps, which have now been found to have light-sensitive organs on their backs.13 How the energy for this faint light is generated is at present unknown; its source may be the chemical and physical reactions happening where superheated, mineral-rich water under high pressure emerges from the vents and makes contact with the cold water of the surrounding deep ocean.

Hydrothermal vents are a topic where the interests of geophysicists, oceanographers, biologists, and physical chemists converge. They are places where the earth's internal heat is being vigorously dissipated, where unique ecosystems flourish with no help at all from solar energy, and where light is produced in unusual ways.

Turmoil at the Surface: Earthquakes

The energy trapped in the earth's interior, underground and out of sight, attracts no public attention until it bursts out of bounds. People who never go near volcanoes, or into earthquake zones, can live a lifetime without becoming aware of the vast stores of energy trapped below the ground; it has to be experienced to be appreciated. An earthquake is one of the ways the earth rids itself of a portion of the energy it must dissipate; other ways are volcanic eruptions and mountain building.

Let's consider earthquakes first. They occur when the edges on either side of a break in the lithosphere scrape past each other. The break may be along the contact between adjacent tectonic plates; this is true of most large earthquakes and explains why they happen at plate boundaries. Or it may be a fracture—a fault—anywhere within a plate where the rock has been deformed by pressure from a distant plate collision until it snapped.

Then, whatever the origin of the break, the masses of rock on the two sides of it slide past each other. The sliding isn't smooth. At first friction holds everything in place while stress builds up; eventually the force pushing the masses overcomes the friction, and they suddenly jerk past each other. The jerking movement—often a sequence of several jerks—is an earthquake shock and its aftershocks. The relative movement of the two masses may be up and down, sideways, or a mixture of the two. A vertical movement at the surface leaves a cliff—a fault scarp—as evidence; a lateral, sideswiping motion, such as happens along California's San Andreas fault, leaves unmistakable discontinuities in roads, fences, and streams.

In terms of energy changes, what happens is this: while they are held fast by friction, the rock masses accumulate elastic potential energy in the internal deformations caused by the tremendous pressure. Then, when the stress becomes too great, the frictional force locking them together fails and the masses slip—the potential energy becomes kinetic energy. The abrupt movement of massive bodies of rock is a seismic event—an earthquake. What becomes of the energy released?

It is dissipated in several ways: some is converted to gravitational PE, stored in rock masses that are lifted to a higher elevation than they occupied before the quake. A fraction of the energy is used in smashing things—rocks and all manner of human constructions—which entails breaking chemical bonds. Another fraction is converted to thermal energy by friction at the place of the rock displacement. And what remains is transported away from the scene as seismic waves, to be dissipated, eventually, some distance from the site of their origin (see chapter 17).

Some earthquake energy, surprisingly, is added to the earth's spin energy.14 The displacement of rock accompanying an earthquake usually shifts heavy rock downward, closer to the earth's center and thus closer to its axis of rotation than it was before the quake. In this way the earth's spin energy is being increased, at present, by about 21 x 1013 kJ per year (this looks like a huge number, but it is only one-trillionth of the currently existing spin energy; see chapter 14). The increase causes the earth's spin to speed up, just as a spinning skater speeds up when she pulls in her outstretched arms. The faster spin would shorten the length of the day by an exceedingly small amount if the effect were not masked by the much more pronounced lengthening of the day caused by the drag of ocean tides.

The downward movement of a subducting tectonic plate under the pull of gravity also entails a loss of gravitational PE, in the same way that a landslide at the earth's surface causes surface rocks to lose gravitational PE (see chapter 9).15 The principle is the same even though the details are spectacularly different. At the surface, a mass of rock loses its gravitational energy by toppling off a precipice, hurtling through the air at high speed, and crashing at the bottom. Inside the earth, a falling mass of solid rock sinks through ductile rock at a speed of only a few millimeters a year; as it sinks, it slowly warms to the temperature of the rock surrounding it until it becomes indistinguishable from it. In both cases—landslide and subducting slab—a small fraction of the lost PE becomes added to the earth's spin energy. And in both cases most of the gravitational PE is converted to KE, which in turn is converted to thermal energy because of friction and drag. In a word, it is dissipated.

Turmoil at the Surface: Volcanoes

Next, consider volcanoes. Obviously the earth loses heat when volumes of red-hot lava gush out through the crust and cool off in the open air. Not so obvious is the reason the lava became molten in the first place. Volcanic lava (known as magma before it emerges into the open) usually comes from no

Figure 15.2. Ascending magma chambers full of magma (black) In the section shown in figure 15.1. Different kinds of magma chamber, labeled 1, 2, and 3, are formed by (1) the melting of ascending currents of mantle material; (2) frictional melting on the upper surface of a subducting slab; (3) the heat of a hot spot at the core-mantle boundary (far below the border of the figure).

Figure 15.2. Ascending magma chambers full of magma (black) In the section shown in figure 15.1. Different kinds of magma chamber, labeled 1, 2, and 3, are formed by (1) the melting of ascending currents of mantle material; (2) frictional melting on the upper surface of a subducting slab; (3) the heat of a hot spot at the core-mantle boundary (far below the border of the figure).

great depth: it is molten mantle rock. It does not come from the core; although the outer core is liquid iron capable of flowing, it is far too dense to rise to the earth's surface. The problem becomes, Why should mantle rock melt?

Most volcanoes are near the boundaries of tectonic plates. Mantle rock melts, and the magma collects to fill magma chambers, in two very different environments: where two plates separate, and where two plates meet with one subducting below the other. The rock melts in these two environments for two quite different reasons. Figure 15.2 shows what happens (note the sites labeled 1 and 2).

Where plates separate, a current of ductile mantle rock ascends from below; as the rock creeps up, the pressure weighing it down decreases until the combination of pressure and temperature is such that the rock liquefies; this is the process feeding volcanoes on midocean ridges.

Where two plates meet, the subducting slab sinks down into the mantle on which it had been floating. Frictional drag between the cool, dense slab and the warmer upper mantle rock through which it sinks generates sufficient heat to melt the overlying warm rock but not the cool, sinking slab.16 The reason the upper rock melts and the lower rock doesn't is that the rock above contains considerable moisture, absorbed from wet ocean sediment, and wet rock melts at a lower temperature than dry rock does.

Unlike most volcanoes, which are found at plate boundaries, some erupt nowhere near the boundaries. These are hot-spot volcanoes, fed by plumes of hot rock rising from much greater depths (site 3 in fig. 15.2).17 The plumes originate at localized hot spots at the bottom of the mantle where it is heated to tremendously high temperatures by currents in the liquid outer core. The temperature difference between a plume and the rock around it may be as great as 1,500°C.

A hot-spot volcano is unaffected by the drift of the tectonic plates across the earth's surface nearly 3,000 km above the hot spot itself. In this respect hot-spot volcanoes behave quite differently than plate-boundary volcanoes do. The latter erupt wherever a boundary happens to be, because the magma sources travel with the plates; a hot-spot volcano, in contrast, is left behind by the plate moving over it; each time it erupts it punches a new hole through the ever-drifting crust and builds a new volcano some distance behind the site of the preceding eruption. As a result, the successive eruptions from a hot-spot plume create a row of volcanoes. The youngest volcano of the row is at the back of the line, with progressively older volcanoes ahead of it.

Rows of hot-spot volcanoes are seldom straight lines; more often they are gently curved arcs. Sometimes the arcs trail across a continent, sometimes across an ocean. If the summits of an arc of submarine hot-spot volcanoes are high enough to emerge above sea level, the result is a volcanic island chain. Several of them occur in the Pacific—for example, the Hawaiian Islands, the Aleutian Islands, the Kuril Islands, and the Tuamoto Archipelago.

Note that each of the three kinds of volcano we've considered receives its lava (or magma, while it's underground) from a different source. Midocean-ridge volcanoes get it from rising currents in the mantle, which melt when the pressure is low enough; the volcanoes above subducting slabs get it from moist, upper mantle rock that has been heated to its comparatively low melting point by friction; and hot-spot volcanoes get it from hot plumes rising from the bottom of the mantle. Note particularly that in no case does the melting require "new" thermal energy or energy from an outside source. The energy that volcanoes let loose has been inside the earth all along, waiting to be dissipated. There is plenty more down there still.

Generalizations about the energy released in volcanic eruptions would be meaningless because they are so variable. But it can safely be said that some past eruptions were larger by far than any experienced by human beings since our species evolved 3 or 4 million years ago. Eruptions two or three thousand times as powerful as the Mount St. Helens eruption of 1980 have left their mark on the earth.

On several occasions in the past, a newly developed hot spot has sometimes fed a long sequence of eruptions lasting for a few million years and leaving behind overlapping layers of volcanic rock (basalt) that now cover whole landscapes to considerable depths.18 Each eruption obliterated every living thing in its path. Examples of the present-day remains of such events are the sheets of lava forming the Columbia Plateau of Washington and Oregon, which flowed out of the ground in a series of eruptions about 17 million years ago and buried more than 100,000 km2 of land. A similar series of eruptions 65 million years ago produced the enormous terraced sheets of lava known as the Deccan that occupy most of peninsular India; their area is more than 500,000 km2.

These sequences of tremendous lava floods have happened at long intervals, on the order of tens of millions of years. Another such sequence could begin anywhere at any time. The earth still has plenty of energy to dissipate.

Mountain Building

Volcanic eruptions and earthquakes are local events in which the earth's energy is dissipated in short, spectacular bursts. At the same time, the earth is continuously shedding huge amounts of energy in the long, slow process of mountain building. It happens when the continents borne by tectonic plates collide and deform each other. The upper layers of rock are compressed, and the way they respond depends on their structure.

Folding And Faulting DiagramsFolding And Faulting Diagrams

Figure 15.3. (a) Mountains built by normal faulting; the dotted line shows the surface before the faulting. (b) Mountains built by the folding and thrust faulting of sedimentary strata. In both diagrams, half arrows show the direction of relative movement at each fault; recently eroded material in the valleys is shown scribbled.

Figure 15.3. (a) Mountains built by normal faulting; the dotted line shows the surface before the faulting. (b) Mountains built by the folding and thrust faulting of sedimentary strata. In both diagrams, half arrows show the direction of relative movement at each fault; recently eroded material in the valleys is shown scribbled.

Two possible outcomes are beautifully illustrated in the Rocky Mountains, which were forced up, and continue to be forced up, by the push of the Pacific plate against the western margin of the North American plate.19 In the southern Rockies (Colorado and southward) the pressure has forced the crust to arch upward. Tension over the top of the arch has caused vertical or near-vertical cracks (normal faults) to open up, separating the crust into a number of steep-sided blocks, some of which have been driven farther upward by compression; many of these blocks, consisting of ancient granite from which overlying sediments have mostly been eroded, now form precipitous mountain peaks (fig. 15.3a).

In the northern Rockies, by contrast, the compression forced thick, horizontal layers of sediments to slide forward over the underlying rock. Friction resisted the slide and forced the sheets of rock to buckle into a series of folds. Under continued pressure, many of the folded sheets fractured and were thrust forward and upward over the sheets ahead of them along gently sloping thrust faults; the overlapping thrust sheets stacked up to form mountains several thousand meters high (fig. 15.3b). This is how the folded mountains of Wyoming, Montana, and Alberta came into existence.20

Both processes obviously entail the dissipation of enormous amounts of energy. The rock faces on either side of a fault grind past each other against unimaginably strong frictional resistance. Most of the movement probably happens in sudden spurts, separated by long intervals; frictional heat is generated whenever there is movement.21 The folding of rock consumes much energy too. It takes place deep underground, where the high temperature makes the rock plastic enough to bend rather than break under intense, slow-acting pressure. Energy is consumed in stretching and rupturing the intermolecular bonds that hold the rock together, leaving it permanently deformed.

As rock is raised to a higher elevation in the process of mountain building, it automatically gains gravitational potential energy (see fig. 15.4); in this way some of the KE of drifting tectonic plates becomes converted to PE and "stored" in the mountains as they rise. It doesn't all remain stored, though. Some of it is dissipated as it accumulates, because a growing mountain sinks to some extent at the same time as it grows. In the act of sinking, the crust dissipates the surplus energy (gravitational PE) in two ways. In the first place, the added weight of the growing mountain causes the lithosphere supporting it to sag; the lithosphere is an elastic solid, able to bend without breaking under pressure that does not exceed its elastic limit. The sagging lithosphere stores

Iceberg Below Surface
Figure 15.4. Lower panel: The weight of mountains at the surface stretches the elastic lithosphere supporting their weight. The lithosphere sags into the viscous asthenosphere it floats on. Upper panel: Sketch of one of the folded mountains in the diagram below (Mount Kidd, Alberta, 2,972 m high).

energy as elastic PE; then, as the load is removed by erosion, the elastic PE energizes an exceedingly slow "rebound."

In the second place, the sag in the floating lithosphere displaces some of the athenosphere it floats on (see fig. 15.4). In the same way that an iceberg floating in the sea displaces its own weight of water, the sagging lithosphere displaces its own weight of viscous asthenosphere material: it gives the material the energy to ooze to one side.22

We have now completed the energy budget for mountain building. To summarize: the source is the kinetic energy of drifting tectonic plates. It is dissipated in several ways: as frictional heat, when faults slip; in the breaking of chemical bonds, when rocks are forced into permanent folds;23 and as gravitational potential energy temporarily parked in the uplifted masses of rock. This gravitational PE is quickly (in geological terms) transformed, some to elastic PE in the sagging lithosphere and some to the KE of the oozing asthenosphere material displaced by the sagging lithosphere. Ultimately, erosion removes the mountain, the lithosphere rebounds, releasing its stored elastic PE, and the as-thenosphere oozes back to where it was before; drag causes a large fraction of the asthenosphere's KE to be dissipated as heat.

The energy that went into building the mountain is now all accounted for. The most difficult part—putting in the numbers—will not be attempted. Possibly the biggest difficulty is to assess how the PE stored for a while in the uplifted mass of the mountain becomes apportioned between elastic PE in the lithosphere and KE in the asthenosphere.

0 0


  • brian lee
    How do asthenosphere behave?
    5 years ago

Post a comment