CHAPTER 2: THE COSMIC SHOOTING GALLERY

Uniformitarian philosophy makes much sense. Most natural processes are cyclical. Rivers empty into oceans, yet the oceans don't overflow. In the words of the Old Testament, there is "nothing new under the Sun". Clearly nature must be stable for the familiar Earth to be so ancient. That life has been sustained for billions of years seems to mean that ecological changes have been minimal. We judge the past by our own experiences: the present is the key to the past. We can hardly imagine that processes or events far beyond our experience might have been important for the history of our planet. A catastrophic impact from a cosmic projectile seems as remote and unfathomable as a thunderbolt hurled by Zeus or an invasion by space aliens.

Yet evidence of cosmic impacts is there for those who look. In this chapter we describe the Earth's cosmic neighborhood, and why we know that we have indeed been bombarded by some of our smaller neighbors in space.

We begin our quest for a cosmic perspective on the Moon. Because the Earth's surface is constantly reworked by erosion and geological change, it preserves a poor record of its past. Even "ancient" mountain ranges are scarcely older than 5% of our planet's age. The Moon, in contrast, is a geologically dead world, whose mountains and craters were formed aeons ago. No continents press into one another on the Moon, and volcanoes no longer erupt. There is no air, no water, and no wind erosion. The surface of the Moon is a window onto its past, preserving four billion years of history.

When Galileo and his successors looked at the Moon through their telescopes, they saw a surface covered by the shallow circular depressions they called craters, after the Greek word for cup. Craters are the quintessential lunar landform. With a modest telescope we can count thousands of them, ranging from numerous cavities 1 km across up to Clavius, fully 250 km in diameter. Still larger impact features are called basins rather than craters. Some of the huge ringed basins are filled with dark lava plains, forming the familiar spots we call the "Man in the Moon." From lunar orbit, spacecraft have photographed millions of craters, down to mere hollows the size of a swimming pool. Returned moon rocks bear pits down to microscopic dimensions. How curious it is that these ubiquitous features, and the lessons they hold for our own planet, were misunderstood and ignored for centuries.

We now know to use the Moon to infer the history of the Earth -- the perspective taken in this chapter. But the opposite approach was applied during the 19th century and much of the 20th, when our world served as a model for the Moon. On Earth, craters are rare (in the United States, they are often in national parks), and most of them are volcanic. Some are the familiar cinder cones like Sunset Crater in Arizona, or on the summits of high mountains like Mauna Loa in Hawaii. Others are pit craters, collapse craters, and even volcanic explosion craters. By analogy, lunar craters were assumed to be volcanic, although few terrestrial volcanoes have the shallow, bowl-shaped appearance of most features on the Moon. By circular reasoning, scientists assumed that the Moon must be like the Earth, so we were unlikely to learn anything very revolutionary about the Earth from looking at the Moon.

A few geologists suggested a century ago that the ubiquitous lunar craters might have an impact origin, but the concept made real sense only after scientists calculated the energy of impacts by cosmic debris. Interplanetary velocities of asteroids and comets are so great -- tens of kilometers per second -- that each ton of projectile typically carries 100 times more kinetic energy of motion than the chemical energy of a ton of high explosives. So an impacting ton of asteroid explodes with the energy of 100 tons of TNT, and impact craters are, in effect, explosion craters. More geologists were converted to the impact hypothesis after seeing aerial photos of the Moon-like, crater-scarred battlefields of World War I. In 1924, British astronomer Charles Gifford wrote "The fact which has not been taken into account hitherto in considering the meteoric hypotheses [for formation of the lunar craters] is that a meteor, on striking the Moon, is converted, in a very small fraction of the second, into an explosive compared with which dynamite and T.N.T. are mild and harmless." Final proof that the lunar craters were formed by impacts, however, awaited the Apollo explorations that revealed the lunar surface to be an extremely ancient, battered landscape.

Knowing that each lunar crater was made by an explosive impact, our perspective is fundamentally changed. The Earth and the Moon have moved together through the solar system since their birth. If the Moon has been battered by cosmic debris, the Earth must have been as well. Only the smaller components of this cosmic rain burn out as they pass through Earth's atmosphere: any projectile large enough to make one of the telescopically visible lunar craters would have had a similar effect on the Earth. Were it not for the continual erosion and reworking of the Earth's surface, which fills in and erases craters, our planet would be as densely cratered as the Moon.

The United States space program of the 1960s and 70s played a critical role in stimulating scientific interest in cratering. As exploratory spacecraft like the Mariners and Voyagers ventured to the other planets, we found that craters are the common coin of planetary geology. Mercury, Venus, Mars, the moons of Jupiter and Saturn -- all have cratered surfaces. Impacts are a part of planetary history, competing with internal forces such as volcanism in sculpting geologic landforms. Only on the Earth and Venus (and a couple of volcanically active moons) do the internal forces dominate. Elsewhere, as on the Moon, impacts have generally shaped most of the surface features we see.

Today we cannot look at the Moon without recognizing the lesson it teaches us about our own planet's history of bombardment from space. But this is a modern viewpoint, a product of the planetary exploration programs of the past three decades.

To read the records of lunar and terrestrial history, we need a clock. The crater-scarred Moon shows us that the Earth must have been struck repeatedly by cosmic debris, but it does not immediately tell us when these impacts happened. Perhaps, we might suggest, the era of bombardment was confined to the distant past, when the planets were young and the Earth's crust was still molten and could not retain craters. If so, the lunar record has little relevance to the recent history of the Earth. Can we test this hypothesis?

Before the Apollo expeditions to the Moon, we could not determine the age of the lunar surface just from looking through a telescope. Maybe it is extremely old, and craters were formed in the early era of planetary accretion. Maybe we are safe today from such impacts. However, as soon as the first moon-rocks were returned to Earth by the Apollo astronauts, the age of the lunar surface could be measured. Most of these rocks are igneous, formed in volcanic eruptions that formed the dark "Man in the Moon" lava plains. The ages when the molten lava solidified and became rocks were measured from the decay of radioisotopes to be 3.3 to 3.9 billion years old -- very old, but up to a billion years younger than the age of the Moon itself (and the Earth), which is 4.5 billion years. The lava plains of the Man in the Moon had erased all previous topography, including craters, but they are no longer smooth today. Even a small telescope reveals that craters have formed in the plains due to continuing bombardment that has persisted beyond the accretionary era. Indeed, it is still continuing.

How do we read the cosmic scorecard represented by these fresh lava plains? We can see just five craters on the plains that exceed 50 km diameter -- that's one or two such craters formed every billion years (on average) since the plains were formed. Since the Earth is struck by the same cosmic debris as the Moon, we can calculate that it was probably hit about 150 times during the same period of 3½ billion years, since our planet has approximately 30 times more land area than the lunar maria. That's one new 50-km crater formed on land every 20 million years [check]. Indeed, as we will discuss later, this is just about the rate of crater formation rate inferred from geological studies of ancient terrestrial impact craters.

These cratering rates assume that the cosmic rain has been steady during the past 3½ billion years. We don't know how correct this assumption is. Probably there have been a few heavy showers interspersed with long dry spans, during which fewer impacts took place. Yet measured ages for both lunar and terrestrial craters tell us that at least the larger craters have formed throughout geological history. Some astronomers think that the smaller, more frequent impacts come in storms, which would be particularly scary if we were unlucky enough to be in a rare storm right now. These interesting ideas are speculative, however, and there is no compelling evidence for such variations over time.

Each impact crater on the Earth or Moon is the result of a huge explosion. Something smashed into the surface, releasing a burst of energy and excavating a large cavity in the ground. From the size of the crater we can estimate the energy of the projectile, but little else. Just what are these cosmic bullets?

One reason 19th century astronomers and geologists failed to realize that lunar craters were made by impacts was the lack of a known source for the projectiles. At that time they did not know that there were asteroids loose in the solar system that could run into the Earth or Moon. Asteroids are small rocky or metallic bodies ranging up to about 1000 kilometers in size. They were all thought to be safely stored in the asteroid belt, out beyond the orbit of Mars. Not until 1932 was the first small asteroid found on a path that crossed the Earth's orbit. This little rock, only a few kilometers across, was named Apollo (for the Greek Sun god) because it approached the Sun so closely. Another of the early asteroids that came near the Earth and Sun was called Icarus for the same reason. Since then, about 200 more Earth-crossing asteroids have been discovered.

The Earth-crossing asteroids are not so numerous as their larger main-belt siblings, but their irregular, elongated orbital paths bring them close to the Earth and other planets. Occasional close encounters with the gravity fields of planets send them veering off into unpredictable paths that continue to pose a threat. It has been estimated that one in four Earth-crossing asteroids will eventually score a direct hit on Earth.

The Earth-crossing asteroids are all too small to be photographed from Earth in any detail, even with the Hubble Space Telescope. However, the new techniques of planetary radar have allowed us to probe the shape and physical nature of several of these mysterious objects. The use of radar to obtain images has been pioneered by Steven Ostro of the Jet Propulsion Laboratory (JPL) in Pasadena, California. Very powerful radar pulses are transmitted from big radio antennas on Earth. A few minutes later, the pulses intercepted by the little asteroid are reflected back toward Earth and, some minutes after that, faint echoes are received -- either by the radio telescope that transmitted the pulses, or another dish aimed at the object and precisely tuned to the correct frequency. Detailed computer analysis of the returned signals enables Ostro to infer the shape and rotation of the object. He has successfully used radar instruments operated by NASA as part of its Deep Space Net of tracking antennae, as well as the giant 1000-foot (300-m) radar telescope, the largest in the world, operated by Cornell University at Arecibo in Puerto Rico.

If the asteroid is unusually close so that the returned radar signals are strong, they can be processed into actual pictures of the asteroid. By the mid 1990s, Ostro had reconstructed excellent images of three asteroids: Castalia (500 m across), Toutatis (5 km across), and Geographos (2 km across). Each is an elongated, highly irregular object, as might be expected for a fragment spawned in some ancient collision between two asteroids. Unexpectedly, however, 2 of these asteroids appear to be double objects, perhaps created when two roughly spherical objects collided at low speed to create a binary. Castalia, in particular, is unmistakably double. Today there is speculation that many of the small Earth-approaching asteroids may be binaries or multiple-component "boulder piles".

To learn more about these mysterious objects, two spacecraft have been launched on exploration missions. The first attempt was made by the U.S. Department of Defense as a part of its growing interest in defending the Earth from impacts, a subject we discuss in detail in a later chapter. The Air Force Space Command and the Naval Research Laboratory collaborated in 1994 to send a spacecraft called Clementine, built using some of the technology developed for the "star wars" Strategic Defense Initiative, toward the Earth-crossing asteroid Geographos via the Moon. Unfortunately the spacecraft failed soon after leaving lunar orbit, and no asteroid images were obtained.

A larger spacecraft called NEAR (for Near Earth Asteroid Rendezvous) was built by NASA as part of its planetary exploration program and launched from Cape Canaveral on February 17, 1996. It passed through the main asteroid belt in June 1997, taking the first-ever close-up images of one of the common class of black asteroids -- this one is called Mathilde. NEAR's chief target is the large Earth-approaching asteroid Eros, which will be reached by the spacecraft in 1999. If successful, NEAR will orbit Eros for more than a year, returning a great deal of data to the Earth, before finally being sent to crash-land on the asteroid's surface.

As the first space mission ever dedicated to studying asteroids, NEAR carries a payload of instruments specifically designed to learn about the nature, including the chemical composition, of Eros. Eros will be mapped in exquisite detail, as only an orbiter can do it. (When a spacecraft only flies past an object, en route to something else -- as Galileo flew past the main-belt asteroids Gaspra and Ida in the early 1990s -- we are always left with the nagging question: What might have been on the other side?)

NEAR, which is about the size of a Volkswagen Beetle, will be watching in case Eros holds some surprises for us, like some moons or unexpected geological activity. Even if Eros turns out to look just like a giant example of a meteorite we already have samples of here on Earth, we can reflect on the fact that this object -- about 20 km across -- could well be the next major extinctor of life on this planet. Computer calculations show that Eros has about a 50-50 chance of ending its existence by colliding with Earth. Although we know such a collision cannot possibly occur during our lifetimes, it could well do so in the next few millions of years, and it will make the K/T boundary dinosaur extinctor look paltry by comparison.

So as NEAR scrutinizes Eros for clues about its origin, or for practical information that might facilitate eventual mining, this enormous "flying mountain" with the name of "love" might instead be seen as the Sword of Damocles hanging over the heads of Earth's species incapable of establishing themselves elsewhere when the potential destruction happens.

There are indirect, and cheaper ways, of studying the physical and chemical nature of asteroids. For instance, we can learn about them by studying meteorites, which are asteroidal fragments that have survived passage through our atmosphere. From these rocks we learn that some asteroids are "primitive", meaning that they represent the original unmodified material that formed the building blocks for the planets. The primitive meteorites and their asteroidal parents are composed of rocky minerals mixed with fine metallic grains. Other asteroids were somehow heated to their melting points, permitting their rocky and metallic parts to separate. The meteorites that originate from the once-molten cores are nearly pure iron-nickel alloy, while the samples of their mantles and crusts resemble terrestrial rocks. Although metallic meteorites are common in museums, their presence results from selection effects: they make it through the Earth's atmosphere almost undamaged and, once fallen, iron-nickel meteorites are much easier to identify than stony meteorites. In fact only about 2% of the meteorites that fall are metallic. This and other evidence indicates that metallic asteroids must likewise be quite rare, so the great majority of asteroidal projectiles are rocky objects.

Comets, like asteroids, are small, solid bodies that orbit the Sun. They differ from asteroids by having ices (including water ice and dry ice) near their surfaces. When a comet approaches the Sun, the heat vaporizes the ices to drive out an enormous dusty atmosphere (the comet's "head"), which is often swept away from the Sun into a tenuous "tail", typically tens of millions of kilometers long. Most of the light we see from a comet is reflected from the thin gas and dust of the head and tail. In addition there is most certainly a solid nucleus of rock and ice that is the source of the visible atmosphere. The existence of solid nuclei was generally recognized in the 1940's but not really proven until 1986 when the European Giotto spacecraft took pictures of a very substantial object about 10 km across in the middle of the head of Halley's Comet. Unfortunately, we do not have cometary samples in the form of meteorites, since their icy remnants burn up in the atmosphere, but we are pretty sure that if we were struck by a cometary nucleus, it would be very nearly like being hit by a rocky or metallic asteroid of the same mass and incoming velocity.

We need a simple term for these cosmic projectiles, including both Earth-crossing asteroids and comets. They are now usually called NEOs, for Near Earth Objects. By coincidence, this terminology emphasizes the newness of the perceived threat of impact from NEOs. . . the idea of impacts by NEOs could even be called neo-catastrophism (pun intended).

Although many of the NEOs will eventually strike the Earth, it is still a long time between such cosmic collisions. The target (Earth) must be in exactly the same place at the same time as the NEO. Since interplanetary space is very big, the chances of such co-location are very small. Typically, an NEO will cross the Earth's orbit many millions of times, with the Earth elsewhere, before there is an actual hit. It may be many tens of millions of years before an NEO hits the Earth, Venus, or the Moon, or is gravitationally ejected by near-misses. Thus there is plenty of opportunity to spot these objects and track them before they strike our planet.

Comets and asteroids can be seen with ordinary telescopes (otherwise we would not know that NEOs exist), but they are not easy to find. With rare exceptions, astronomers on their mountain-tops do not spend their time "scanning the skies." Instead, they focus on particular problems involving single stars or galaxies. Besides, the great telescopes of the world (and the Hubble Space Telescope in orbit) are designed to peer intently at patches of the heavens so small that thousands of such fields-of-view would be required to fill the bowl of the Big Dipper. Thus astronomers don't necessarily know when something changes in the sky. When one of the countless stars of the firmament explodes, it does so unwatched. Maybe it becomes brilliant enough eventually to be noticed, but astronomers always come on the scene in the aftermath, like police after the burglar has fled.

There are a few telescopes of different design. Instead of magnifying a small piece of the heavens, they capture a wide part of the sky. During the 1980s and early 1990s just two telescopes on Earth were routinely used to watch for new objects in the solar system. One was the 18-inch wide-field telescope on Palomar Mountain, in southern California -- the mountain made famous for the 200-inch Hale telescope, once the world's largest. The other is a 36-inch telescope operated by the University of Arizona in a search called Project Spacewatch. The few users of these telescopes, many of them unpaid volunteers, have been the only astronomers about whom it could properly be said that they "scan the skies."

In 1995 the 18-inch wide-field telescope on Palomar discontinued its survey program, although two new automated search telescope are taking its place in 1997 and 1998. Even with this modest increase in telescopic power, however, only a small portion of the sky will be scanned each month, and there is no assurance that an incoming projectile will be found before it hits. It is entirely possible that when the next big one hits the Earth, the first we will know about it is when we feel the ground shake and watch the fireball rising above the horizon.

** * * * *

Let's conclude this chapter by introducing one of the pioneers in asteroid hunting -- and one of the leading characters in our impact saga. He is Gene Shoemaker, now best known as co-discoverer of Comet Shoemaker-Levy 9, which crashed into Jupiter in July 1994. Long before the great comet crash, Shoemaker had established himself as the world's leading expert on impacts and impact craters, and his work had already earned him the Presidential Medal of Science and membership in the U.S. National Academy of Sciences. Shoemaker, a charming and energetic scientist who recently retired from the U.S. Geological Survey, was trained as a field geologist. He founded the field of planetary geology (which he originally called astrogeology) and helped train the astronauts for the Apollo Moon landings. Later he participated in the Voyager robotic mission to the outer planets and helped interpret pictures of the cratered moons of these distant planets. In the midst of these activities, Shoemaker also took up a new career as an observing astronomer.

Shoemaker was one of the first scientists to calculate the average rate of impacts on the Earth from the lunar data, at the same time going "into the field" to discover and explore the impact scars on our planet. But he was not content with such historical research on past impacts; he also wanted to know the current cratering rate, as well as the nature of the projectiles that produced these craters. Very few astronomers cared about the NEOs, and no one in the 1970s was carrying out a systematic effort to discover and catalogue these interplanetary wanderers. Characteristically, Shoemaker decided that the best way to remedy this situation was to undertake the search himself.

Through his association with the California Institute of Technology (Caltech) as a part-time Professor of Geological and Planetary Science, Shoemaker got access to the small 18-inch telescope on Palomar Mountain. In collaboration with JPL scientist Eleanor Helin and later with his wife Carolyn Shoemaker, he made the personal commitment to spend one week out of each month photographing the sky with this small telescope. Each region would be photographed twice and the two images compared to look for any objects that were moving against the background of stars. This is a formidable task, because each individual photographic plate contained XX,000 [check] stellar images. The simplest way to pick out the moving asteroids was to view the pairs of plates through a stereo microscope. When aligned properly, all the star images appear to be in the same plane, while any moving object appears to "float" above the background stars. Carolyn Shoemaker proved to be particularly adept at recognizing these floating objects, and she is the person who actually discovered most of the team's new asteroids and comets.

The Palomar NEO search began in 1973. In 1982 it split into two efforts, each using the 18-inch telescope for one week per month. One team was headed by Eleanor Helin and the other by the Shoemakers. In the early years about one NEO per year was found by each group, but in the 1990s the pace accelerated. When the program began only xx of these objects were known, but by the middle of 1995 xx were known, xx of them discovered at Palomar. These discoveries form the basis for estimating the hazard of impacts and developing the strategy for dealing with them, as we will discuss in later chapters of this book. Gene Shoemaker himself has played a leading role in all of these activities, and he is truly the father of modern NEO and impact research, as well as something of a father figure to many younger scientists. The "Mom and Pop" team of Gene and Carolyn Shoemaker have probably done more to save the Earth from impact catastrophe than anyone else alive, and they will reappear frequently throughout this narrative.

* * * * * * *

One of the most important products of the searches for NEOs is an estimate of the number of comets and asteroids with orbits that cross the orbit of the Earth, and hence an estimate of the frequency of impacts by objects of various sizes. For those interested in the impact issue, whether one is an astronomer, geologist, insurance actuary, or civil-defense planner, it is essential to know this impact rate. Thus we conclude this chapter with a quantitative estimate of this cosmic rain that pelts down continuously on our planet.

From the NEO surveys carried out by Gene Shoemaker and his astronomer colleagues, we can estimate the number of near-Earth asteroids in various size ranges. The largest such asteroids, named Ivar and Betulia, have diameters of about 8 km, roughly half the size of the object that killed off the dinosaurs when it struck the Earth 65 million years ago. We are confident that there are no undiscovered near-Earth asteroids larger than these two, but once we move down to smaller sizes, our catalogues are less complete. The numbers of these objects must therefore be estimated statistically. Our telescopes are watching only a part of the sky for a part of the time, so many objects are missed. To calculate the true numbers, we correct our incomplete observations, in much the same way you could calculate the total number of trucks that pass along a highway during the course of a day even if you only actually counted the trucks for 5 minutes out of each hour.

Combining the data from astronomical surveys with the record of lunar craters (which represent a long-term average), Shoemaker has estimated that there are 400 Earth-crossing asteroids larger than 2 km in diameter, 2000 larger than 1 km, 10,000 larger than 500 m, and roughly 300,000 larger than 100 m across. These numbers are good estimates for the larger objects, but become increasingly uncertain as we move to smaller sizes, where only a very small fraction of the NEOs have actually been discovered.

Probably the most meaningful way to present these statistics on the NEO population is to make a plot of the average frequency of impacts of a given size or larger over the whole Earth. Shoemaker first derived such a plot in 1981, and an updated version is illustrated in the Figure. In addition to asteroid size, this figure shows the energy (in megatons of TNT equivalent) for each impact. Thus, for example, we can see on the plot that Earth is struck by an NEO with energy of 100 megatons or more (diameter 100 m or more; the size of a large city office building) approximately once per millennium. (It is purely coincidence that this particular set of values comes out in such round numbers, but it makes them easy to remember). Larger impacts are much rarer: the interval for one million megatons (2 km diameter) is approximately once every one million years (another number that is easy to remember).

These estimated average impact rates will come up repeatedly in the chapters that follow. To evaluate the implications of these impacts, we must know many additional factors, especially concerning the physical and environmental consequences of impacts of various sizes, but this impact rate curve remains essential.