Fig. 8-0: (a) Hubble Space Telescope image of Comet Shoemaker-levy 9 in March 1994, four months before its collision with Jupiter. The comet had broken up into a “string of pearls” that ultimately impacted with Jupiter. (b) Mosaic of WFPC-2 images shows the evolution of the comet impact site on Jupiter. The comet broke into fragments, producing several impact sites. The change in one impact can be seen in the images from lower right to upper left, and the scar from a second impact appears in the third image. (Credit: (a) Courtesy of NASA; credit: H. Weaver (JHU), T. Smith (STScI). (b) R. Evans, J. Trauger, H. Hammel, and the HST Comet Science Team and NASA)
We are not alone in the Solar System. Since ancient times, it has been apparent that we have neighbors distinct from the stars, most obviously the moon and planets, with their bright presence and distinctive orbits. And occasionally one of the neighbors drops by for a permanent visit, seen in meteor showers or the occasional meteorite or comet that finds its way to Earth’s surface. Particularly in the early history of the solar system, interactions with the neighbors had a major role in Earth’s formation, and subsequent interactions have played a major role in the evolution of life and even today pose a threat of environmental catastrophe. From one point of view, Earth is simply made up of neighbors whose visits became permanent. If we could choose one planetary grain as the original Earth, then it ultimately grew to our current planet by the progressive accumulation of material through impacts, a process that still continues on a reduced level today. Therefore in a very real sense, we are our former neighbors.
Early Earth’s history is closely tied to that of our nearest neighbor, the moon. Dates from lunar rocks are largely between 3.0 and 4.4 billion years. The absence on the moon of modern volcanism, of an atmosphere to produce weathering, and of any tectonic movement make the moon a kind of “planetary fossil” that records events in our vicinity early in solar system history, a period from which no rocks on Earth survive. Study of the lunar rocks not only tells us about the history of our nearest neighbor, it also provides key insights into Earth’s early history that can be gained in no other way. The origin of the moon now appears to have been caused by the giant impact of a Mars-sized planet about 50 million years after the origin of the solar system. Owing to the heat generated from its rapid accretion, the moon likely underwent large-scale melting of its interior to produce an early “magma ocean.” Floating crystals of plagioclase rose to form the predominant rock type of the light-colored lunar highlands. Separation of other crystals led to layering of the lunar interior. Melting several hundred million years later produced the younger black lava plains of the lunar maria. Thermal considerations suggest that Earth likely also underwent very large scale melting with formation of a magma ocean early in its history. Both Earth and moon later underwent a “late heavy bombardment” of meteorites when reorganization of the orbits of the outer planets destabilized the asteroid belt, leading to many large impacts in the inner solar system around 3.8 Ga. It may be for this reason that this age corresponds closely with that of the oldest surviving rocks on Earth, and only after this time was life able to establish a permanent foothold on Earth’s surface.
Impacts that were the primary process of formation of planets and moons have continued through Earth’s history, with marked consequences for life, including the extinction of the dinosaurs 65 million years ago. Historic impacts are evident from comets crashing into Jupiter (see frontispiece), as well as surviving young craters on Earth. Future impacts are inevitable, either of near Earth objects from the asteroid belt, or comets released from the vast cometary storehouses of the Kuiper Belt and Oort Cloud, from the outermost solar system beyond Neptune.
Even a casual consideration of the world we observe around us shows that planetary evolution does not proceed in isolation. Instead, it involves close and diverse relationships with neighbors, large and small. Our size is negligible compared to the sun (Fig. 8-1); we depend on it for our orbit and our energy and light, and we are influenced by its varying magnetic field. The moon causes tides that influence all our shorelines and ecosystems. Jupiter is also vast in size, and gravitational effects from the outer planets influence our orbit sufficiently to be a major cause of climate variations, as we will see in Chapter 18. The outer planets also influence the orbits of all the objects that traverse the solar system, and some of these disturbed orbits impact Earth and other planets today as meteorites and comets. Larger meteorite impacts in the past led to mass extinctions of life that contributed to biological evolution, as we will learn in Chapter 17. And in the future, an asteroid or comet impact may be the most significant catastrophe that human civilization will face. We are influenced by and depend upon these interactions with our neighbors. In terms of energy, climate, life, and matter, we are in relationship with the solar system, and Earth’s habitability has been strongly influenced by these relationships.
Fig. 8-1: Extreme Ultraviolet Imaging Telescope (EIT) image of the sun, showing the relative size of sun and Earth. More than a million earth-sized objects could fit inside the sun’s volume. Earth is smaller than many sunspots. Prominences are relatively cool dense plasma suspended in the sun’s hot, thin corona that escape the sun’s atmosphere. Emission in this image shows the upper chromosphere at a temperature of about 60,000°K. Every feature in the image traces magnetic field structure. The hottest areas appear almost white, while the darker areas indicate cooler temperatures. (Information and image courtesy of NASA)
We also have much to learn about Earth’s history and habitability from the study of our solar system neighbors. Early Earth’s history cannot be studied directly because the oldest rocks on Earth are about 4 billion years old, and less than 1% of Earth’s surface contains rocks older than 3 billion years. Most of these rocks were not at the surface when they formed, and when they do become exposed at the surface, they are rapidly modified by erosion and biology. Earth’s surface tells us very little about early solar system conditions. To put this 4.55 Ga–3.8 Ga data gap in perspective, it is a longer time period than the entire fossil record when multicellular life developed and evolved, and it is the time when Earth formed its primary layers, plate tectonics may have begun, and primitive life may have first appeared. How can we fill this important hole in Earth’s history? Study of other solar system objects that preserve information from that time can tell us much about the earliest history of our habitable world.
In addition to the sun and the eight planets, the solar system contains a host of other objects. The most prominent of these are the more than 150 moons that circle the planets. The list does not end here. Circling the sun between the orbit of Mars and the orbit of Jupiter are about 100 billion asteroids. While most of them are 1 m or less in diameter, about 2,000 are greater than 10 km in diameter. Only one, Ceres, is more than 1,000 km in diameter. Jupiter and Saturn have rings made up of myriads of small objects. Finally, on the order of 1,000 billion comets are thought to ride in orbits more distant than Neptune’s. The recent studies of comets, which are considered the leftovers from the formation of our solar system around 4.6 billion years ago, show that they consist of ices wrapped around a rocky core. In this sense, comets are to the major planets (Jupiter, Saturn, Uranus, and Neptune) what asteroids are to the minor planets, namely, miniature versions. The famous “tails” of comets are the vapor created when in the course of their highly elliptical orbits they come close to the hot sun.
Fig. 8-2: Sizes of some of the moons of the four outer planets. The circles show the relative sizes of the moons, for which the scale is shown in the box on the lower right. Their distance from the center of the host planet is indicated on the horizontal axis. Note the scale is logarithmic. The moons appear to overlap because their size is greatly exaggerated relative to the distance scale. As can be seen, three of the moons are larger than Mercury, and five are larger than Earth’s moon. The numbers within the circles give the bulk densities of the moons (in gm/cm3). Io and Europa have densities close to that of our moon; the other moons have lower densities and must contain a substantial component of the ices.
Six of the sun’s eight planets have moons. As shown in Figure 8-2, all four outer planets have moons, and as our resolution of the outer solar system continues to increase and we can see smaller and smaller objects, the number of moons may increase still further. Jupiter has sixty-three, four of which are the large moons discovered by Galileo (Io, Europa, Ganymede, and Callisto) and two of which (Ganymede and Callisto) are larger than the planet Mercury! Saturn has fifty-three moons, one larger than Mercury. Uranus has twenty-seven, and Neptune has thirteen. Earth and Mars are the only terrestrial planets with moons. We have one large moon similar in size to the large moons of the outer planets. Mars has two very small moons.
The space probes sent from Earth to explore the solar system have photographed the moons of Mars, Jupiter, Saturn, Uranus, and Neptune, showing them to be solid objects of great diversity. Many of the moons as well as Mars and Mercury are dotted with craters made by the impacts of meteorites (Fig. 8-3). Others have no craters at all, suggesting an actively changing surface. Io has a smooth surface owing to its highly active volcanism (Fig. 8-4a). Europa (Fig. 8-4b) is entirely covered by ice that moves and deforms. The densest moons have densities similar to silicate rocks, but most of the moons of the outer planets have much lower densities, consistent with the cold environment in which the outer planets formed. Study of moons is a burgeoning field of planetary science, particularly as we try to understand the likely characteristics of solar systems other than our own and expand our views of the possible range of environments for life. For example, while the surface of Europa is ice, its density reveals a rocky interior, and there is evidence of a large liquid ocean beneath the snowball surface. It is natural to wonder whether liquid water and a rocky substrate led to the appearance of life at depths in the ocean there.
Some moons around the outer planets have characteristics that suggest they formed as mini solar systems around their parent planet; they have regular prograde orbits, circling the planet in the same direction as the planet is rotating, and they are aligned with one another around the planet’s equator. Moreover, Jupiter’s large moons show a regular decrease in density with distance from their planet, suggesting a possible gradient in the temperature of condensation caused by the formation and luminosity of the parent planet (Figs. 8-2 and 8-4). The moons with densities less than 3.0 must be considerably richer in volatile elements than are the terrestrial planets and contain substantial amounts of the ices. The miniature solar systems around Jupiter, Saturn, and Uranus have very small inner moons and large outer ones, similar to the organization of planets around the sun.
There is a second large class of moons that have higher inclinations (they do not rotate around the planet’s equator) and often have retrograde orbits. These moons are thought to have been captured from other regions of the solar system. The two small moons of Mars are readily explained by capture from the neighboring asteroid belt, but the asteroid belt is not a candidate for supplying the numerous moons of the outer planets. Instead, they are thought to come from a large region beyond the planets called the Kuiper Belt (Fig. 8-5a). Many hundreds of substantial objects have now been discovered beyond Neptune, and the Kuiper Belt is thought to include more than 70,000 objects larger than 100 km in size. Pluto is one of the largest of these objects, and it is now known to have a large companion named Charon. Neptune’s largest moon, Triton, has a retrograde motion and is actually 18% larger than Pluto. Pluto’s orbit actually crosses that of Neptune. Both Pluto and Triton are now thought be the largest representatives of Kuiper Belt objects. These considerations (among others such as its size and the inclination of its orbit) led to the removal of Pluto from the list of true planets.
Fig. 8-3: Craters are a common characteristic of almost all bodies in the solar system. Here are two examples. The top image is a photograph of the surface of Saturn’s moon Mimas. The main surface feature is the giant impact crater 130 km across in the upper right of the image; its walls are approximately 5 km high. A crater of similar relative dimensions on Earth would be as wide as Canada. Fractures on the opposite side of Mimas have been discovered and may have been created by shock waves from the impact traveling through the moon’s body. The bottom image is a portion of the surface of Mercury near its south pole. Other impacts are visible in many other figures of this chapter. (Images courtesy of NASA)
Fig. 8-4: The four “Galilean” moons of Jupiter, shown to scale. (a) Photograph of the surface of Io, the innermost of the four Galilean moons of Jupiter. Its bumpy surface is due to intense volcanic activity, which continually repaves the surface, destroying any craters that occur. (b) Photograph of the surface of Europa, another of the Galilean moons. Craters are mostly absent, because the ice-covered surface is in movement and is young in age. (c, d) The two outer moons of Ganymede and Callisto are made of rock and ice (see their densities in Fig. 8-2, and reveal partially cratered surfaces). Color images of Europa and Io can be seen in color plates 5 and 6. (Images courtesy of NASA)
The farthest reaches of the solar system are the Oort Cloud (Fig. 8-5), where billions of potential comets reside in vastly distant orbits extending much of the distance to the nearest star. Passing stars can perturb the orbits of Oort Cloud objects and send them rushing into the inner solar system as comets, many of which end up being accreted to one of the planets.
A striking aspect of the exciting new information coming from the moons and other objects of the outer solar system is their diversity. Io is the most volcanically active body in the solar system, and liquid sulfur appears to play an important role in the eruptive style. Ice-covered Europa is a satellite where the rocky interior underlies a planetary scale ocean whose surface is completely ice covered. Saturn’s Titan has an active surface climate driven by methane, which can be solid at the prevailing low temperatures and also forms rivers and lakes. Even greater diversity of environment and style must exist elsewhere in the galaxy, stretching our concepts of potential planetary environments and the diverse styles of life that might occupy them.
Fig. 8-5: Illustration of the the two major features of the outer solar system. The Kuiper Belt is a band of objects just beyond the orbit of Neptune. Note the tiny box representing the inner solar system. The former planet Pluto is the largest Kuiper Belt object. The Oort Cloud extends greatly beyond the Kuiper Belt and hosts many billions of objects, some of which are perturbed by the gravity from passing stars and enter into the inner solar system as visible comets. (Courtesy of NASA; http://www.nasaimages.org/luna/servlet/detail/NVA2~8~8 ~13317~113858:Hubble-Hunts-Down-Binary-Objects-at)
The moon (Fig. 8-6) is of particular interest to us as we explore the origin of our habitable planet. It is our nearest neighbor, more than one hundred times closer then either Venus or Mars, and potentially has much to reveal about solar system events in our immediate neighborhood that are no longer preserved on the constantly changing terrestrial surface.
First-order consideration of the Earth-moon system reveals a number of curious features. Earth is the only inner planet with a moon of significant size. The moon is unusual even in the broader context of the solar system moons discussed above because it is the largest moon relative to the size of its planet. And unlike moons around the outer planets, it has a lower density (3.1 gm/cm3) than its planet. Even more puzzling, the moon’s density is lower than the density of any of the inner planets. Recalling our discussion of density in Chapter 5, the moon must be made up almost entirely of rock, without a core of any significant size. The evidence from other planets and meteorites suggests that planetary differentiation is commonly associated with separation of rock and metal, so how would it be possible to form such a large object in the inner solar system with no metallic core? The circular lunar orbit is also unusual. It is within 1% of being a perfect circle, although the general solution for satellites is an ellipse. For example, Jupiter’s major moons have ellipticities of 4–15%.
Other puzzling features of the moon became apparent from the ages and chemical compositions of lunar rocks. The dates obtained on lunar rocks constrained the age of the moon to 4.43–4.52 billion years old. These numbers are 40 to 100 million years younger than the ages of chondrites discussed in Chapter 4. As we learned in Chapter 5, modeling of the solar nebular suggests that the major episodes of accretion of planetesimals would take less than 20 million years, so the age of the moon poses a bit of a puzzle. What happened to lead to its later formation?
Another puzzle is the low concentrations of siderophile elements in lunar rocks relative to chondrites. On Earth we explain this depletion by core formation. Since the low density of the moon precludes a significant core, how could the moon become depleted in siderophiles? The moon is also very depleted in volatile elements, not only the obvious lack of water and atmospheric gases but also moderately volatile elements such as K, Na, and Cl. For Earth the ratio of the moderately volatile K to the refractory U is 12,000. For the moon, K/U is only 2,000. This difference suggests that the material that formed the moon was subjected to temperatures considerably higher than was the case for the material that formed Earth. The moon is also highly depleted in all elements more volatile than potassium relative to Earth.
Fig. 8-6: Photograph of Earth’s moon. The dark-colored portions (called maria) are areas of the surface that were flooded by lunar basalt flows. The light-colored portions (called highlands) represent the original anorthositic crust. The difference in cratering intensity shows that the maria are younger than the highlands. (Courtesy of NASA)
There is one more piece of evidence bearing on the moon’s origin coming from careful measurements of the oxygen isotopes. Careful measurement of the relative abundances of the three isotopes of oxygen, 16O, 17O and 18O, showed slight differences between Earth and various classes of meteorites. The moon, however, is identical to Earth, suggesting a common parentage.
Hypotheses for the origin of the moon need to account for these diverse observations. The capture hypothesis has the moon accreted in an orbit similar to Earth, from which it was passively captured to enter orbit around Earth. This hypothesis does not explain the lack of a lunar core, and dynamically it is also very difficult to capture a large object like the moon and have it end up in a circular orbit.
The fission hypothesis is that the moon formed by fission from Earth after Earth’s core had formed. This hypothesis has appeal because it accounts for most of the lunar puzzles. There would be siderophile element depletion despite the lack of core because Earth’s core formation would occur prior to fission. Oxygen isotopes would be identical to Earth because moon and Earth were once combined. If the fission occurred at high temperatures, there might be a mechanism for loss of volatile elements. There are two difficulties with the fission hypothesis, however. The first is the young age of the moon. If early Earth were spinning fast enough to fission, why wouldn’t it happen immediately after Earth formation and separation of the core, which we learned in Chapter 7 happened less than 30 Ma after Earth formed? The more serious difficulty is that fission would require the early Earth to spin with a two-hour day, causing some of the mantle to be ejected into orbit, where it could coalesce to form the moon. This hypothesis requires the ad hoc assumption that Earth had a much faster spin than observed elsewhere in the solar system. Even more seriously, the total angular momentum of Earth and moon today are not consistent with such high spin rates, unless a great deal of material was lost to space.
The giant impact hypothesis proposes that another planet about the size of Mars had a grazing impact with Earth, ejecting large amounts of material into space around the joined planets, which condensed to form the moon. A giant impact resolves most of the lunar puzzles. It would explain why Earth had a large moon and other inner planets did not. It would also explain the late age of the moon, since its formation would have to postdate the major accretion of the planets. The impact would happen after the formation of cores on the two impacting bodies. Modeling of the giant impact (Fig. 8-7) showed that the high density of the metallic cores would cause the two cores to fuse in Earth, leading to a hot cloud of silicate debris surrounding Earth that would be siderophile-element depleted. Condensation and accretion of this debris at high temperatures would lead to the formation of the moon, depleted in both siderophile and volatile elements.
A criticism of the giant impact hypothesis is that it requires a unique event. Advances in modeling planetary accretion, however, show that giant impacts are likely in early solar system history, and that most of the planets probably ultimately accreted from large protoplanets impacting one another. Large impacts have now been invoked to explain why Mercury has an oversize core—a large direct impact could have caused the removal of most of the silicate mantle. Recent results also suggest that the vast differences between the two hemispheres of Mars may result from a giant impact. A giant impact has also been proposed to account for the reverse rotation of Venus and the fact that Uranus has a horizontal spin axis compared to the other planets. These various lines of evidence, coupled with increasingly detailed modeling that provides compelling images of the giant impact, have led most planetary scientists to currently favor the giant impact hypothesis for the formation of the moon, and have led more broadly to a recognition of the importance of giant impacts to planetary accretion and early solar system history.
The hypothesis is by no means proven, however. A difficulty with the giant impact hypothesis is that modeling results suggest the moon would be formed largely from material from the impactor rather than Earth itself. To be consistent with the oxygen isotope evidence, the impactor would have to have had the same oxygen isotope fingerprint as Earth—and there is no way to verify whether it did. Since it is possible that this was the case, or that future models may find a way to make more of the moon from Earth itself, the giant impact hypothesis does not have major problems based on current data, and the fission and capture hypotheses do have such problems. So the giant impact is currently the preferred model, but in view of the uncertainties, it ranks only a 5–6 on our theory scale.
Fig. 8-7: Numerical model of the giant impact hypothesis for the moon formation. This model has the proto-Earth hit by a Mars-size body (called Theia) at a velocity of 40,000 km/hour and at an angle of 45°. The impact causes moon-forming material to be ejected into the Earth orbit, producing a hot silicate vapor; as it cools down, a solid particle disk is created, and through accretion of these particles, the moon is formed. For color version see color plate 7. (Courtesy of Robin M. Canup, Southwest Research Institute)
Giant impact and fission share many features—both form the moon from Earthlike materials, after the formation of the core, from a hot cloud of debris surrounding early Earth. These common characteristics are likely to survive as our understanding of the moon’s formation continues to evolve.
After the giant impacts that marked the first 100 million years of the history of the solar system, impacts did not stop. When we look up at the moon, it is clear even to the naked eye that the lunar surface (Fig. 8-6) is pockmarked with craters of all sizes, some of them more than 1,000 km in diameter, and all of these must have happened after the moon had its primary differentiation into layers, since a rigid outer crust is a prerequisite for crater preservation. Impacts of significant size have been ongoing.
Today it is accepted that craters form as a result of impacting objects from space. This was not always the prevalent view. One of the difficulties in convincing early scientists that craters formed as a result of impacts is that most craters are circular. Impacts could come in at any angle, and if we carry out simple experiments firing projectiles in the laboratory, then low-angle impacts lead to elliptical craters, not circular ones. Furthermore, there was also often little evidence of the impacting object. Where was it? And there was often a lot of silicate melt around, suggesting that craters were caused by volcanic processes.
The breakthrough in understanding crater origin was the recognition that impacting objects travel at hypervelocities of 17–70 km/sec. At 70 km/sec, a meteoroid could travel from San Francisco to Paris in two minutes or travel from the moon to Earth in an hour and a half. The speed is about a hundred times faster than a bullet—Superman could not catch them. Energy increases as the square of the velocity, so a bullet at hypervelocities would be 10,000 times more damaging. At these high speeds, the pressures on impact are so great that a strong shock wave propagates from the point of impact. The shock caused by the impacting object rather than the object itself creates a circular crater some twenty times larger than the diameter of the impactor (Fig. 8-8). The impact also generates enough heat to largely vaporize the meteorite and melt the country rock, and creates high-pressure minerals never seen elsewhere on Earth’s surface. The presence of the high-pressure form of quartz, named stishovite, is considered definitive evidence of an impact origin for numerous nonvolcanic, circular craters on Earth, and convincing evidence that the same morphological features elsewhere in the solar system resulted from impacts.
Earth’s atmosphere plays an important role in impacts. Large meteoroids punch through the atmosphere and create a vacuum than can facilitate throwing some of the ejecta into space. Smaller meteoroids, however, are slowed down by the atmosphere. Many of them burn up in transit, making meteor showers, and others are drastically slowed by the friction of the atmosphere and have much less energetic impacts and better preservation of meteorite fragments. The moon is able to preserve a much more complete cratering record, because meteors of all sizes are able to arrive at the surface unchanged by interaction with an atmosphere.
Modern observations show that impacts remain important. On Earth today meteor showers are common, and these reflect a continuing inflow of solar system material. Larger recent impacts, since humans have inhabited the Earth, are also evident, such as the Barringer Crater in Arizona, where the Canyon Diablo meteorite, 50.0 m in diameter, impacted the Arizona desert to create a crater 1.2 km wide ~50,000 years ago (Fig. 8-9). In the early twentieth century, a likely comet exploded in the atmosphere over Tunguska, Siberia, leading to widespread devastation. And in 1994 the comet Shoemaker-Levy had a spectacular impact with Jupiter (see frontispiece). Another comet probably about 1 km in diameter impacted Jupiter in 2009.
Even larger terrestrial impacts are indicated by an unusual form of glassy rock called tektites (Fig. 8-10). Based on their aerodynamic shapes and surface textures, tektites are thought to form from liquids that crystallized to glass above Earth’s atmosphere and then were modified by ablation as they fell back through the atmosphere to Earth’s surface. Scientists are now convinced that these objects were formed during impacts that splashed material up from Earth’s surface well above the atmosphere. The most abundant of the tektites are those found throughout Southeast Asia and Australia in soils and streambeds and in marine sediments adjacent to these landmasses. These tektites all have ages of 0.7 million years and likely formed from a single large impact. Other groups of tektites are found in North America (with an age of 30.0 million years), in central Europe (age 13.0 million years), and in the African Ivory Coast (age 1.1 million years).
Fig. 8-8: Illustration of formation of an impact crater. Note that impacting objects travel at “hypervelocities” of 17–70 km/sec. At these high speeds, the pressures on impact are so great that a strong shock wave propagates from the point of impact. The shock caused by the impacting object rather than the object itself creates a circular crater some twenty times larger than the diameter of the impactor. The impact also generates enough heat to largely vaporize the meteorite and melt the country rock, and it creates high-pressure minerals never seen elsewhere on Earth’s surface. (Modified from B. French (1998). Traces of Catastrophe, Lunar and Planetary Institute Contribution No. 954, with permission)
Fig. 8-9: Aerial photo of Barringer crater. The crater is about 1.2 km (400 ft) in diameter and 170 m (570 ft) deep. It was created by the Canyon Diablo meteorite, a meteor 50 m in diameter that impacted the Arizona desert 50,000 years ago. (Courtesy of U.S. Geological Survey)
In the case of the European tektites, the actual impact crater is thought to have been located in Germany. Even though it has been partially erased by erosion over the last 30 million years, the presence of stishovite in the sedimentary rocks deformed by this impact proves its origin. These various lines of evidence show that impacts of a wide range of objects continue historically and in the geologically recent past. Planetary accretion is ongoing.
Since impacts have persisted over billions of years, the extent of cratering of a surface reveals information about the age of the surface. Earth has few impact craters because its surface is continually reworked by erosion, mountain building, and volcanism. In contrast, the oldest planetary surfaces have received so many craters that their surfaces are “saturated” with them—the entire surface is cratered, so that new craters simply destroy older ones. Surfaces intermediate in age are moderately cratered. This simple principle allows us to make estimates of the relative ages of the surfaces of moons and planets. Even for saturated surfaces, a relative age scale for craters can be constructed by looking carefully at their geological relationships. A crater that occurs within an existing basin, or destroys a preexisting crater rim, is younger, and “rays” of dust from impacts that overlie older craters also give relative ages. Craters with small numbers of impacts in their basins would also be younger than those with larger numbers of impacts (Fig. 8-11). By dating impacts and combining the dates with the relative timescale, a cratering history can be constructed.
Fig. 8-10: Photo of tektites, black glass objects whose shapes and textures bear witness of high-speed flight through the atmosphere. They are created from material melted and splashed high above the atmosphere by the impact of meteorites or comets. (Courtesy of Harvard Museum of Natural History)
Fig. 8-11: Image of impact craters on Callisto, moon of Jupiter. Estimates of the relative ages of the surfaces of moons and planets can be obtained by the extent of cratering. Even for saturated surfaces, a relative age scale for craters can be constructed by looking carefully at their geological relationships. A crater that occurs within an existing basin, or destroys a preexisting crater rim, is younger, and “rays” of dust from impacts that overlie older craters also give relative ages. Craters with small numbers of impacts in their basins would also be younger than those with larger numbers of impacts. By dating impacts and combining the dates with the relative timescale, a cratering history can be constructed. (Courtesy of NASA)
Fig. 8-12: Map of the global topography of Mars. The Mars Orbiter Laser Altimeter (MOLA), an instrument on the Mars Global Surveyor (MGS) spacecraft acquired the first globally distributed, high-resolution measurements of Mars topography. Topographic models have enabled quantitative characterization of global scale processes that have shaped the Martian surface, and as revealed on the image, Mars has two hemispheres with contrasting crater characteristics; the southern highlands have a large number of craters, suggesting an old surface, whereas the northern hemisphere plains are far less cratered, suggesting a much younger surface. See color plate 2. (From Smith et al., Science 284 (May 28, 1999):1495–1503; http://photojournal.jpl.NASA.gov/jpeg/PIA02031.jpg.
The lunar surface is pockmarked with craters of all sizes, ranging from large craters 1,000 km in diameter to microscopic craters from impacting planetary dust. (These small craters are possible on the moon because it has no atmosphere to slow down and burn up incoming planetary debris, and the craters have the chance to be preserved owing to the absence of weathering.) This suggests that the lunar surface is very old. Mercury has a completely pockmarked surface like the moon, suggesting an ancient age for its surface. Mars has two hemispheres with contrasting crater characteristics (Fig. 8-12). The southern highlands have a large number of craters, suggesting an old surface. The northern hemisphere plains are far less cratered, suggesting a much younger surface, though craters are still far more abundant than they are on Earth. Venus has very few craters on its surface, showing that its exterior has been resurfaced during its history. Cratering intensity thus reveals much about the relative histories and activities of planetary objects.
After the moon formed, it underwent interior modifications that led to the formation of diverse layers controlled largely by density differences, just as took place on Earth. The moon provides the only direct evidence we have to permit a comparison of the interior modifications of two planetary bodies. As it turns out, there are important similarities and important contrasts both in the processes and final results of lunar and terrestrial planetary differentiation.
A similarity is that the moon appears to have a metallic core, but the contrast is that it is very small. The inference from the low lunar density was confirmed by instruments left on the moon by astronauts that radioed back seismograms generated by moonquakes. These results suggest that the moon has a small core (about 2% of its mass). The core likely formed by immiscibility, as on Earth, but the small amount of metal in the moon allowed only a tiny core to form.
The moon has no atmosphere or ocean. The reason is that the moon’s gravity is so low that gaseous molecules can readily escape from the moon’s surface. Here we encounter one of the prerequisites for habitability. If a planet is too small, it can retain neither atmosphere nor ocean.
The lunar crust carries a complex story that reveals much about the moon’s evolution. Careful examination of the lunar surfaces shows that the density of impacts over the surface is not constant. The side of the moon that we see is divided into two distinct provinces of contrasting age. The whiter portions are more highly cratered than the plains that fill the giant craters. The smoothness of these surfaces led early observers to name them lunar maria (Latin for “seas”). The white regions of the moon are also at higher elevations, so they are called the lunar highlands (Fig. 8-6). The back side of the moon consists entirely of highlands terrain. The inference from the relative cratering age was confirmed by rocks returned from the highlands and maria by the Apollo and Luna missions in the 1960s. Study of these rocks then permitted a much more detailed assessment of how the two major terrains of the lunar crust were formed, and a remarkably detailed model of the early history of the moon.
Fig. 8-13: Histogram showing the temporal distribution of ages for various Mare basalts. A map of the moon is shown as a reference for sample locations. Note that the Mare basalts show a peak in activity about 1 billion years after the formation of the moon. Recent high resolution photography of the lunar surface suggests there may be very small amounts of younger lava flows, inferred even to be as young as 1.3 billion years based on their cratering density. (Modified after Hiesinger et al., J. Geophys. Res. 105 (2000), no. E12: 29,239–75, and 108 (2003):1–27)
Dates from lunar rocks showed that the dark lunar maria consisted of basalts formed mainly during the period 3.1–3.9 billion years ago (Fig. 8-13). Rocks from the light-colored highlands are older, having formed as much as 4.4 billion years ago. Little or no volcanism has occurred on the moon during the last 3 billion years. Since then the moon has been a “dead” planet. No convection cells churn in its mantle. No plates collide on its surface. No volcanoes erupt. Why this big difference? Again, it is the moon’s small size that is responsible. The small size of the moon leads to a large surface area/volume ratio, permitting heat to get out. And the low gravitational field of the moon means the pressure increases very slowly with depth, allowing melting to extend to great depths and extract the heat to the surface. Owing to its present cold and rigid internal state, great convection cells no longer carry heat from its interior to the top of its mantle.
The moon’s maria are made of a rock superficially akin to Earth’s basalt, and the highlands are rocks somewhat akin to Earth’s granite; the highlands anorthosites consist primarily of the mineral plagioclase feldspar, also the most abundant mineral of Earth’s crust, although the lunar feldspar is much more Ca rich because of depletion of the moon in the more volatile Na. Upon deeper inspection, however, the similarity between Earth’s crust and the moon’s crust breaks down.
As we saw in Chapter 7, continental crust on Earth is granite made up of multiple minerals that are the “minimum temperature melt” in the presence of water of a host of rocks—sediments, metamorphic rocks, basalts, or preexisting granites. These granitic melts crystallize to form quartz, two feldspars, and other minerals. Terrestrial granites have also concentrated the magmaphile elements to a hundred times higher levels than Earth’s mantle. Not so the lunar highlands. These rocks consist largely of a single mineral, the Ca-rich plagioclase end member anorthite, and their concentrations of magmaphile elements are often very low. Examination of the binary phase diagram from the previous chapter shows that melts of polymineralic substances do not lead to liquids that would have monomineralic compositions. The lunar Highland crust does not have the composition of a partial melt of any planetary interior, and clearly formed by very different processes than Earth’s continental crust.
The chemical compositions of the mare basalts also revealed bizarre compositions dissimilar to any terrestrial basalts. Whereas basalts on Earth generally have between 1 and 4 wt. % TiO2 and lavas at the high end are rare, many lunar basalts had TiO2 concentrations greater than 10.0 wt. %, while others had concentrations less than 0.5 wt.%. Clearly, processes that created both highlands and maria differed greatly from those that created crust on Earth. These new data challenged the creativity of terrestrial geoscientists. Could the well-understood principles of igneous petrology be used to understand the formation of the unusual compositions of the lunar crust?
Important additional clues came from trace element concentrations in the lunar rocks, particularly the rare earth elements (REE). The REE, also known as the lanthanide contraction series, occupy the lower interior of the periodic table (see Fig. 4-1) and have some very useful geochemical characteristics. Because they are all related by having additional electrons added to an inner rather than outer electron shell, all of them have a common outer electron shell structure. This leads to very similar geochemical behavior during igneous processes. As the number of protons in the nuclei increase, however, the ionic size of the REE decreases regularly over the entire series of seventeen elements. Because minerals discriminate for and against elements based on their charge and size, when only the size varies, the chemical differences from one REE to another tend to be gradual and smooth. This leads to “REE patterns” that are characteristic of different minerals and provide clues to what minerals have been involved in the genesis of different rocks.
All of the REE but one also have a common 3+ valence. The one element on the moon with a different valence is Europium (Eu) in the middle of the lanthanide contraction series. Europium can occupy two different valence states, 2+ and 3+, and this makes its behavior differ in important ways from the other REE. This difference in behavior is particularly marked for the mineral anorthite, because its mineral formula CaAl2Si2O8 has Ca in a 2+ valence site of appropriate size where Eu2+ can substitute very easily, while the other 3+ REE are too large to substitute for the Al3+ in the mineral. For this reason, plagioclase feldspars take up much more Eu than other REE, creating a REE pattern with a marked positive concentration anomaly for Eu. When the REE pattern in a rock has a Eu anomaly, it shows that plagioclase has played an important role in the generation of the rock. Rocks that are formed from the accumulation of plagioclase have positive Eu anomalies, and those that have seen plagioclase removal have negative Eu anomalies.
REE patterns of the lunar highlands rocks have marked positive Eu anomalies (Fig. 8-14), showing that they formed by accumulation of plagioclase minerals. This corresponds with their monomineralic character—somehow plagioclase minerals accumulated preferentially to form the lunar highlands. Mare basalts, on the other hand, have strong negative Eu anomalies (see Fig. 8-14)! This sentence deserves the exclamation point because there was no plagioclase in these rocks, and experiments showed that their chemical compositions would not have crystallized plagioclase at any pressure. If no plagioclase could have been removed from the rocks, how could they have a Eu anomaly? The answer is that the source region that melted to form the rocks had already undergone prior separation of plagioclase. Then the source region would be both depleted in plagioclase and have a negative Eu anomaly. Subsequent melts would inherit the Eu anomaly and have so little plagioclase in the melt that they would not crystallize plagioclase during cooling.
Fig. 8-14: Trace element patterns of Maria basalts from Apollo 17 and highlands anorthosites. Note that REE patterns of the Mare basalts have strong negative Eu anomalies, therefore the region that melted to form the rocks had already undergone prior separation of plagioclase; on the other hand, lunar highlands anorthosites have marked positive Eu-anomalies, showing that they formed by accumulation of plagioclase minerals. (Adapted from P. H. Warren, The Moon, in Andrew M. Davis, ed., Meteorites, Comets, and Planets, vol. 1 of Treatise on Geochemistry (Oxford: Elsevier Ltd., 2005))
To summarize the evidence, the lunar data showed old highlands crust up to 30km thick containing anorthositic rocks with strong positive Eu anomalies suggesting plagioclase accumulation. Mare basalts occurred hundreds of millions of years later, filling large impact basins, and had negative Eu anomalies despite the absence of evidence for plagioclase removal from the rocks. Moreover, Mare basalts included a wide diversity of compositions, from very high to very low TiO2 contents.
These lines of evidence were creatively explained by a model where the moon had a large magma ocean early in its history. Accretion of the moon following the giant impact could have generated enough heat to cause most of the early moon to melt, creating a magma ocean. One of the first minerals to crystallize would have been plagioclase. Measurements of density showed that plagioclase solids were lighter than the magma of the magma ocean, because the magma ocean had a high FeO content (with 56 protons per iron atom), and all the elements making up plagioclase were of lower atomic number. The plagioclase crystallizing over hundreds of kilometers of thickness of magma ocean would rise to the surface, creating thick anorthositic crust. The plagioclase crystals would preferentially incorporate Eu from the magma ocean, leading to REE with positive Eu anomalies in the lunar crust and a negative Eu anomaly in the residual magma ocean liquid from which the plagioclase had separated. All the later minerals that crystallized would inherit the negative Eu anomaly produced by early separation of the anorthositic crust. Mafic minerals such as olivine and pyroxene would accumulate in other layers. Since these minerals contain little TiO2, they would create Ti-poor source regions. Much later in the crystallization sequence, the dense Ti-rich mineral ilmenite (FeTiO3) would crystallize, and consequently accumulation of this mineral would generate Ti-rich source regions. The solidification of the magma ocean would thus lead to a spectrum of source regions from Ti-rich to Ti-poor, all characterized by a negative Eu anomaly. After the solidification of the magma ocean, the heat generated by radioactive decay or other processes ultimately heated the lunar interior, permitting it to ascend and melt, generating the spectrum of mare basalt compositions up to a billion years later in lunar history. After these melting episodes, the moon became so cool that no further melting was possible.
This simple scenario (Fig. 8-15) explains the major features of the lunar crust and corresponds with the compositions and ages of the lunar rocks. The story is based, however, on a very sparse sampling of the moon. Only ~390 kg of lunar rocks are available for study, and they are from limited portions of the lunar surface. The entire back side of the moon and the lunar poles are not sampled at all. It is a tribute to geochemistry and the creativity of lunar scientists that such a complete model of lunar crustal formation has been able to be generated, but we should also recognize that more complete sampling would inevitably lead to important modifications of this story. The lunar magma ocean merits only a 4–5 on our theory scale—there will be exciting scientific discoveries when more lunar rocks are recovered.
Fig. 8-15: Illustration of the lunar magma ocean hypothesis. Accretion of the moon following the giant impact could have generated enough heat to cause most of the early moon to melt, creating a “magma ocean.” The plagioclase crystallizing over hundreds of kilometers of thickness of magma ocean would rise to the surface, creating the thick anorthositic crust. Mafic minerals such as olivine, pyroxene and the Ti-rich mineral ilmenite would accumulate in other layers.
We know that impacts have occurred throughout the history of the solar system. Can we say anything about how the rate of impacts changes with time? One constraint comes from the current flux of impacts on Earth. About 40,000 tons per year of matter from space fall to Earth’s surface each year, equivalent in total to a rock about the size of a college science building. Most of this matter is dust, scattered all over the Earth, but there is about one meteorite per year weighing more than 20 gm for every 10,000 square km (about the equivalent of a major metropolitan area). Most of these small fragments are never seen or recovered (Fig. 8-16). A handful of larger newly fallen meteorites are seen and collected each year. Since meteorites are rapidly destroyed by weathering, all the meteorites collected by humans are relatively recent falls.
Fig. 8-16: Photo of the Carancas meteorite crater, Peru, which landed in 2007. The Carancas meteorite created an impact crater about 14 m wide in a rather remote area of that country. No casualties are known. If this insignificant impact had occurred in a populated area, there could have been hundreds of victims. Most such events occur in unpopulated regions (e.g., the ocean, high latitudes, etc.). (Image courtesy of Michael Farmer)
Could the present rate over billions of years lead to Earth’s current size? No—current accretion rates over 4.5 billion years would produce less than one ten-millionth of Earth’s mass. Impacts must have been vastly larger and more frequent in early solar system history. A natural preconception and first hypothesis for impact history would be a smooth exponential decline. Each time a planet or moon goes through its orbit it intersects a percentage of the objects with crossing orbits, gradually clearing out debris from the solar system like a giant gravitational vacuum cleaner. One could construct a “half-life,” for example, for the number of asteroids that cross Earth’s orbit, and over time this would lead to an exponential decline in numbers of asteroids, and associated impacts, similar to the decline of a radioactive isotope during its decay (Fig. 4-14c). The present impact rate would be one constraint, and we could use the cratering intensity on the moon to determine older points and specify the decay rate. This simple scenario, however, turns out to be only part of the story.
Constraints on a hypothesized exponential decline in cratering became possible after the return of lunar samples by the Apollo program. Early studies of the ages of impact breccias (broken-up rocks created by impacts) on the moon showed a clustering of ages between 3.9 and 3.8 billion years. Impact melts from lunar meteorites found on Earth gave the same age range. It appears that dozens of impact craters >300 km in size may have formed within this time interval. To account for these observations, a terminal cataclysm or late heavy bombardment (LHB) has been proposed, where cratering intensity increased markedly for a short period of time.
The LHB has been controversial because a mechanism for such late bombardment was hard to fathom. If planets gradually clean out their orbits, what would be the source of a large flux of impactors 700 Ma after the planets formed? What that would require would be some event that changed the numbers of objects in Earth- or moon-crossing orbits. If the solar system established its orbits very early on, how could that occur? As models of the early history of the solar system have become more sophisticated a mechanism has appeared, because the models suggest that the planets change their orbits in early solar system history. As the orbits change, different portions of the asteroid belt can become unstable, sending objects into the inner solar system. In particular, movement of the outer planets can greatly perturb the asteroid belt when Jupiter and Saturn pass through a mode where they have a 1:2 resonance, where Jupiter goes around the sun exactly twice as fast as Saturn. If this occurred at about 3.9 Ga, it would be a mechanism to generate a new flux of Earth- and moon-crossing objects. Recent studies of the distribution of asteroids in the asteroid belt provide strong evidence for outer planet migration early in solar system history. There is also evidence of a major heating event at the same time in the rare Martian meteorites. The confluence of evidence from the moon, lunar meteorites, Mars, the asteroid belt, and solar system modeling is leading to much greater acceptance of the LHB hypothesis.
Fig. 8-17: Illustration of how impacts may have varied in solar system history. An overall exponential decrease in cratering intensity is punctuated by a massive bombardment episode at about 3.8 billion years, termed the “late heavy bombardment” (LHB). These results show the great importance of impacts in the early solar system, which would have had far-reaching consequences for habitability of early planetary surfaces. (Adapted from Koeberl, Elements 2 (2006), no. 4: 211–16)
Figure 8-17 shows how impacts may have varied in solar system history. An overall exponential decrease in cratering intensity is punctuated by a massive bombardment episode at about 3.8 billion years. These results show the overwhelming importance of impacts in the early solar system, which would have had far-reaching consequences for habitability of early planetary surfaces.
Study of the moon has revealed much about major events in the early solar system that were not evident from study of Earth alone, and these events have significant implications for Earth’s early history.
If the giant impact hypothesis is correct, then shortly after its initial formation Earth underwent the immense catastrophe of giant, planetsize impact leading to ejection of material and formation of the moon. The energy of this impact was sufficient to lead to planet-scale melting, which upon solidification would have caused formation of an early crust and possible stratification of the mantle. Some have suggested that Earth got so hot that it formed a gaseous atmosphere of silicate gas. Depending on the atmospheric conditions, the magma ocean would have cooled rapidly, and one could construct models where it would become stratified. There is little evidence for a highly stratified mantle today, however, and it is possible that convection rehomogenized the mantle. That said, we do not have direct samples of the lowermost mantle, and some scientists argue for some layering of the mantle left over from the magma ocean event. Conclusions about the consequences for Earth of the giant impact rely almost exclusively on modeling, and models at that scale in the absence of known boundary conditions inevitably involve assumptions and creativity. The elegant calculations can be used to support plausible arguments for the consequences of the giant impact for Earth’s early history and subsequent evolution, but the actual facts remain shrouded in uncertainty.
If early Earth did have a magma ocean, why didn’t a large anorthositic crust form? If the lunar magma ocean could preserve 30 km of light anorthositic crust, wouldn’t a terrestrial magma ocean have led to a similar crust hundreds of kilometers thick? No, Earth would not have generated a significant anorthositic crust because of the limited pressure stability of anorthite. All minerals have a limited range of temperature and pressure over which they are stable. The key difference in this regard between Earth and the moon is that the pressure change with depth is much less on the moon because of the moon’s low gravity. That is why the astronauts could jump so high and so far when they landed on the lunar surface. Since rocks weigh much less on the moon than on Earth, pressure increases much more slowly with depth. Anorthite has a maximum pressure stability about 12 kilobars (1.2 GPa) (Fig. 8-18). The center of the moon, at a depth of 1,200 km, has a pressure of only 4.7 GPa (47 kbar), and anorthite is stable over 300 km of depth. Crystallization of 10% plagioclase over this interval would lead to a 30 km anorthositic crust. Earth, however, has an increase of 1 kbar (0.1 GPa) for every 3 km of depth, so anorthite is stable only to a depth of 36km! The equivalent anorthositic crust would be only a few kilometers thick, which would easily be destroyed by impacts and subsequent magmatism. The rapid increase of pressure with depth on Earth also makes it more difficult to generate a terrestrial magma ocean, because melting temperature also increases rapidly with pressure. For the same temperature that would melt the moon to create a 600 km magma ocean, Earth would melt only to some 60 km, trivial in comparison to the 3,600 km depth of the terrestrial mantle.
Fig. 8-18: Depth-temperature diagram with plagioclase stability for Earth and moon. Curves for Earth and moon differ because pressure increases more rapidly with depth on Earth. All minerals have a limited range of temperature and pressure over which they are stable. Anorthite is stable only up to about 1.2 GPa, or 12 kilobars. Because of its low total mass, the moon has a weak gravitational field and rocks weigh much less on the moon than on Earth. Anorthite is stable over 300 km of depth. Crystallization of 10% plagioclase over this interval would lead to a 30 km anorthositic crust. Earth, however, has an increase of 0.1 GPa (1 kbar) for every 3 km of depth, On Earth, the stability limit of anorthite of 1.2 GPa is equivalent only to 36 km, and plagioclase cannot crystallize below that depth. There is, therefore, no possibility of a thick anorthositic crust accumulating on the early Earth. It also takes much higher temperatures to melt Earth at depth.
The early Earth would also have experienced the extensive cratering for which there is evidence on the moon. If the moon suffered a terminal cataclysm, Earth would have received even more impacts owing to its larger radius and more extensive gravitational field. Based on radius alone, Earth would have received at least ten times more impacts than the moon. Adding in the Earth’s additional gravitational pull could make the number far larger than that. David Kring and colleagues have estimated that a 20 km crater might have been formed on Earth every thousand years, and a 1,000 km basin every million years, sufficient to sterilize Earth’s surface of any life that may have begun. The number of surviving rocks on Earth becomes vanishingly small earlier than 3.8 billion years, when the LHB would have ended. This could be a natural consequence of a terrestrial LHB. Only after that time was there sufficient stability of the surface that continental fragments had a chance of surviving. The LHB, then, reflects an important moment of passage for the history of the solar system, after which surface conditions would have become more stable, the rock record on Earth would have been able to be preserved, and the threat of surface sterilization by impacts would have eased. This overall framework provides us with a chronology for the early history of the Earth-moon system that we would not be able to obtain from Earth alone (Table 8-1).
Table 8-1
Hadean history: The first billion years of Earth’s history
| Year before present (Ma) | Time from the zero year | Event |
4,566 |
0.00 |
Condensation of the first solid matter in the solar system |
4,565 |
1 m.y.* |
Formation of planetesimals |
4,555 |
11 m.y. |
Igneous activity in planetesimals |
4,532 |
34 m.y. |
Core’s separation completed |
4,500 |
66 m.y. |
Moon forms from giant impact |
4,450 |
116 m.y. |
Atmospheric degassing largely completed |
4,404 |
162 m.y. |
Oldest zircon |
3,980 |
586 m.y. |
Oldest rock |
3,800–3,900 |
~800 m.y. |
Late heavy bombardment |
3,500 |
1,066 m.y. |
Evidence of life |
*m.y. = million years.
No matter how the moon actually formed, scientists have long inferred that through a process called tidal friction, energy associated with Earth’s spin is being gradually transferred to the moon. The extra energy gained by the moon speeds its movement around the Earth and lifts it to ever more distant orbits, as the tidal friction also slows down Earth’s spin. This inference from calculation turned to proof, thanks to the Apollo program. One of the tasks given to the astronauts who visited the moon was to put in place reflectors that would provide precise points from which a laser beam shot from Earth could be bounced back to Earth. By measuring precisely the time required for a pulse of laser light to travel to the moon and back to Earth, it is possible to establish the distance from a point on Earth to a point on the moon with an accuracy of about 1 cm. These measurements have been repeated regularly over a period of decades. They confirm that the moon is moving away from Earth at the rate of 38 mm per year.
Proof of the complementary change in Earth’s spin requires much longer timescales of measurement, which can be accessed only through the geological record. John Wells, a paleontologist at Cornell University, was aware that corals living in today’s reefs are banded. While the most prominent of these bands have been shown to be annual, the porosity of the calcium carbonate deposited by these organisms changes subtly with the seasons. These changes can be seen on medical x-rays taken of slabs cut from a coral head (Fig. 8-19). In addition to seasonal bands, Wells also saw weaker banding, which he attributed to the monthly tidal cycle and to day-night cycles.
If the moon has retreated at a rate of 4 or so cm per year, then several hundred million years ago there must have been both more days and more months in a year. Wells found that in fossil corals about 360 million years old there were about 400 daily bands associated with each annual layer. As there is no reason to suspect that the time required for Earth to orbit the sun has changed significantly, these results suggest that the day was shorter in the past than it is today—i.e., that Earth was spinning faster. The changes found by Wells in the number of days tell us that the moon was 1.2 × 104 km closer to Earth at 360 Ma than it is today. If the record in fossil corals is being properly read, then the moon’s retreat rate over the last 7% of Earth’s history has averaged 4 cm per year, just as it has over recent decades. Estimates of the length of day at 900 Ma are about nineteen hours. If we extrapolate back to early periods of Earth’s history, the day may have been as short as ten hours, leading to many more days in a year, and the moon would have been much closer and orbiting much faster, leading to a shorter month. This has important consequences for many aspects of the surface conditions of early Earth. Tides would have been much greater, which would have led to much more energetic shorelines and tidal environments, and the moon would have been twice as large in the night sky.
Fig. 8-19: X-ray photograph of a slab of coral. The annual growth rings are clearly visible. The dark bands represent growth during summer months. Proof that the most prominent growth bands in coral heads are caused by the seasonal changes in the character of the calcium carbonate they deposit was provided by studies of corals from the Eniwetok Atoll. The test of an early version of the hydrogen bomb conducted in this atoll in the 1954 produced an enormous amount of local radioactive fallout, so that the waters of the atoll became temporarily highly contaminated with the fission fragment 90Sr. Since the element strontium substitutes readily for the element calcium in the CaCO3 formed by coral, the 1954 growth band was ‘marked’ with radioactive strontium. When corals collected a decade or more after this event were analyzed, it was found that there was one growth band for each year that had elapsed since the 1954 test. Thus, the 90Sr marking of the 1954 band allowed the annual growth-band hypothesis to be verified. (Courtesy of Richard Cember, Lamont-Doherty Earth Observatory)
The recognition of the importance of impacts throughout solar system history, along with the mass extinction of the dinosaurs that is accepted to be the result of a meteorite impact 65 million years ago, naturally leads to the question of the likelihood and danger of impacts today and their consequences for Earth and human civilization.
We know that a major impact occurred in Siberia in the early twentieth century, and the energy released was equivalent to a large atomic bomb. Widespread destruction and loss of life could result from such an impact today, if it occurred either at sea or in populated regions, whether directly on land or by the creation of an enormous tsunami.
Impacts require Earth-crossing orbits, and such orbits can result from three different sources in the solar system. The first are the asteroids with Earth-crossing orbits, referred to as near-Earth objects (NEO). A systematic mapping program has identified the largest 1,000 of these, and it appears that none of them are likely to impact Earth in the near future. Note that many of the orbits are chaotic, and uncertainties increase considerably moving further out in time.
Two other sources of impacts come from the outer reaches of the solar system (see Fig. 8-5). Beyond the orbit of Neptune lies the Kuiper Belt, of which Pluto is the largest current representative. More than 1,000 Kuiper Belt objects have now been observed, and there are estimates of more than 70,000 of them larger than 100 km in size. Much further from the Sun is the Oort Cloud, where a massive amount of solar system debris shot into highly elliptical orbits rotates around the sun at distances up to one light year, or 50,000 times the distance of Earth from the sun. The Oort Cloud is the outermost portion of our solar system. It is so far away that objects there can have their orbits perturbed by neighboring stars or by the Milky Way galaxy itself. These perturbed orbits can then come rushing into the inner solar system. Halley’s comet, for example, spends most of its life in the Oort Cloud. The Shoemaker-Levy comet was not previously known and impacted Jupiter in 1994 with spectacular results (see frontispiece). A more recent impact of Jupiter by an unknown comet was accidentally observed by an amateur astronomer in July 2009.
It may appear surprising that impacts continue with such frequency despite the fact that materials in planet-crossing orbits have been largely removed from the solar system over its long history. This occurs because most current impacts arise from materials whose orbits have been recently perturbed. The billions of objects in the Kuiper Belt and Oort Cloud are in cold storage awaiting such perturbations. Comets are objects from these regions whose orbits were recently perturbed to send them zooming into the inner solar system. Some are quickly captured through impact. Others are ejected from the inner solar system. Still others, like Halley’s comet, make regular visits but lose a little bit of their mass with each orbit, so their total lifetime is only a few million years. That means that current comets are not remnants of earlier objects with Earth-crossing orbits. They are simply the most recent crop of perturbed orbits. Since there are billions of potential comets beyond the orbit of Neptune, and orbit perturbations happen periodically in the outer solar system, a steady supply of potential impactors is assured.
One might think from Hollywood that we could simply send up a rocket with a nuclear warhead to destroy the incoming object. Our missiles fly up to about 1,000 km from Earth’s surface. At 50 km/sec, an incoming comet passes that point 20 seconds before impact, and the comet is moving faster than anything our missiles are designed to intercept. Even if a missile were able to explode nearby, the impactor would simply break up into fragments and disperse damage over a wider area. While NEOs can be mapped and there would be many years of warning prior to impact, most Kuiper Belt and Oort Cloud objects cannot be mapped, and there would be little warning. Shoemaker-Levy was discovered only months before its impact with Jupiter. There is no effective defense against impacts of comets, and these are sure to happen at some future point in Earth’s history.
Not all of the materials from the early solar nebula were accreted by planets. More than a hundred moons, almost exclusively around the outer planets, reveal a great diversity of planetary accretion and style and also show the important role of capture through solar system history. Even after the major steps of formation of planets and planetesimals occurred, impacts continued to play a central role in early solar system history. Earth’s moon is an exceptional body in the solar system—the only significant moon of the inner planets, of very large size and low density relative to its parent planet, depleted in siderophile elements while lacking a significant core. A giant impact was the likely cause of the origin of the moon, and this event would have largely melted the early Earth some 50–100 million years after its formation. Study of the moon shows the importance and scale of early planetary differentiation, likely also to have influenced earliest Earth’s history, though the remaining evidence for a terrestrial magma ocean is far from clear. Impacts then progressively decreased until migration of the outer planets destabilized the asteroid belt and led to a “late heavy bombardment” during 3.9–3.8 Ga. This is also the age of the oldest rocks on Earth and suggests that from this time forward the terrestrial environment became more stabilized with a surface environment where an equable climate could be permanently established and life had the potential to flourish.
Vast amounts of debris remain in the solar system, ranging from the rocky materials of the asteroid belt to the billions of objects in cold storage in the Kuiper Belt and Oort Cloud of the outer solar system. Some of these objects become modified by the inevitable gravitational perturbations created by the outer planets and passing stars. Some of the perturbed orbits intersect the inner solar system, where ultimately they become captured by planets and moons to form modern impacts. Impacting objects from the solar system had an important influence on the evolution of life on Earth, and the process continues today, with potentially catastrophic consequences for human civilization at some unknown future time.
Neil McBride and Iain Gilmour, eds. 2004. An Introduction to the Solar System. Cambridge: Cambridge University Press,.
William K. Hartmann. 2005. Moons and Planets, 5th ed. Pacific Grove, CA: Thomson Brooks/Cole.
Robin M. Canup and Kevin Righter, eds. 2000. Origin of the Earth and Moon. Tucson: University of Arizona Press.
