By counting the asteroids flying past our planet, astronomers estimate that Earth gets hit by a big one once every 100,000 years. By tallying craters here and on the moon, geologists come to similar conclusions. The energy released in such collisions equals 10m atom bombs, sufficient to cause global environmental mayhem and to kill up to half of humanity.
Just how big a piece of space rock is needed to do this is not clear. Half a mile in diameter might be enough, depending on its composition and arrival speed. All scientists agree that the "asteroid the size of Texas" depicted in the film Armageddon is preposterous. No such lumps exist in our neighbourhood; and a much smaller piece of debris would do the job.
Most people are familiar with the theory that the dinosaurs were wiped out by an asteroid or comet impact 65m years ago. The dinosaur-killing impact was caused by an object five to ten miles in diameter-the crater in Mexico has now been mapped. Such projectiles arrive rarely, with tens of millions of years between events. They cause the extinction of species on a global scale. An impact capable of killing just 25 or 50 per cent of humanity is a lesser thing, yet a terrifying prospect-and much more frequent. Should we be worried? Should we do something about it?
An insurance policy?
The government has guidelines for use in assessing significant potential accidents. A matrix is used, relating probabilities of occurrence to likely outcomes (such as numbers of deaths). A once-per-100,000-years event, such as the one discussed here, means a one-in-100,000 chance per annum, if such events are distributed randomly in time. The guidelines say that we simply have to live with such a risk, if it is likely to cause fewer than 100 deaths. It is a "tolerable" hazard. At between 100 and 10,000 deaths, substantial expenditure to reduce the probability of an event is justified. It was this kind of calculation which drove the recent investment of several billion pounds to enhance safety in various nuclear reactors.
An event of this probability which would kill more than 10,000 people is defined as "intolerable" and preventative measures are mandated. On this scale, asteroid impacts are super-intolerable, and the government must act, or else abandon its own guidelines.
But how much should we spend? Consider it rationally, as you would an ordinary insurance policy. If an insurance company wants to survive, it must charge a premium larger than the annual expectation of loss (AEL) for drivers of your age, sex, car model, postcode and so on. Just three things are needed to calculate the AEL caused by a big asteroid impact. We have the annual probability: one in 100,000. We have the deaths resulting: at least 10m in Britain. The last component is an economic valuation of a human life; in this case a British one. The government has set such a value to calculate the amount it will spend to avoid deaths through accidental causes, such as road crashes (by improving lighting, barriers and so on) or in health and safety at work campaigns. A decade ago, the figure was ?800,000. Now it might be closer to ?1m. (By comparison, the US government is willing to spend $4m to avoid each accidental death.)
Taking the three components, we arrive at an AEL of ?100m for Britain. Rationally-although of course we are not rational in dealing with risks-that is the minimum we should be spending in order to tackle the asteroid hazard. But it is hard to imagine how to spend such a sum. We do not have the people with the necessary skills to employ. In any case, we can find a much cheaper solution. The cost/benefit ratio is so extreme that it makes no sense to avoid the small expenditure required. It is as if fully comprehensive car insurance were available for ?1 a year.
Your chance of dying because of a cosmic impact is about one in 10,000-higher than the one-in-30,000 chance of meeting your end in a plane crash (for an average flyer in a western country). This may seem ridiculous. The great infrequency of asteroid impact (compared even to rare natural disasters), combined with their extreme potential destructiveness, make for some strange calculations of probability. But we have seen what happens to a planet in the firing line. In 1994, 20 comet fragments slammed into Jupiter, causing damage over an area much greater than that of Earth. We don't have to get a bloody nose ourselves before we realise that it will hurt.
Some asteroid history
Many ancient civilisations had a superstitious dread of comets. In Shakespeare's Julius Caesar, Calpurnia urges her husband not to go to the Senate on the Ides of March in 44BC, telling him: "When beggars die, there are no comets seen;/ The heavens themselves blaze forth the death of princes."
The knowledge that comets can wreak real havoc on our planet came not long after Shakespeare's time. When Edmond Halley calculated the orbit of the comet that bears his name, he realised that its path crosses that of Earth, making a collision feasible. He raised this at a meeting of the Royal Society in 1694. His particular concern was the origin of the biblical deluge: "the great agitation that such a shock must necessarily occasion in the sea." He suggested that the Caspian sea might have been produced by a cometary collision.
One hundred years later, as Paris was being rebuilt after the French revolution, Baron Georges Cuvier, one of the founders of palaeontology, studied the strata laid bare by construction work. Cuvier regarded the biblical flood as simply the most recent in a series of catastrophes, and he found evidence for earlier extinctions in the rock layers. Byron also took up the theme. In 1822 he spoke of future cometary impacts spelling the doom of mankind, but he suggested that we might be able to "hurl mountains into their paths by means of steam," and thus save ourselves. We might perhaps consider Byron the inventor of the idea of anti-comet (and asteroid) defence.
Cuvier and Byron were of the early catastrophist school, which saw terrestrial history as a series of little-changing eras punctuated by massive upheavals. During the mid-19th century, the rival uniformitarian school gained ascendancy. Its leaders in Britain were Charles Lyell and Charles Darwin. Darwin thought of biological evolution occurring through the slow but steady accumulation of slight changes over very long periods.
Although the possibility of comet impact has resurfaced from time to time over the past couple of centuries, the small number of comets, together with the small area of Earth, has given rise to the comforting notion that there must be many tens of millions of years between impacts. And until the early 1900s, when radioactive dating established the antiquity of Earth (which formed, with the rest of the solar system, about 4.54 billion years ago), few scientists in any discipline considered our planet to be more than 100m years old. This meant that it was feasible to believe that there had been no cosmic impacts on Earth, ever.
All this changed when asteroids began to be found in large numbers throughout the solar system. Comets are brilliant objects, easy to find because of the large cloud of vapour surrounding their solid nuclei. Largely composed of ice, when they approach the sun some of that ice forms a gaseous cloud that may be larger than a full-sized planet. This reflects a great deal of sunlight, as do their familiar tails, making many comets visible to the naked eye.
Not so asteroids. They are bare lumps of rock, most of them quite small, and they reflect only a small fraction of the sunlight striking them. We have learnt that asteroids are mostly stone from studying meteorites, which seem to be simply chips off asteroids. Meteorites are mostly stony, although there are also metallic meteorites.
Asteroids seem to be debris left over from when the planets formed. Each planet was built up from countless trillions of smaller lumps which agglomerated in the early days of the solar system. Close to the sun, the temperature was high enough to make ice and other volatile materials evaporate, and the resultant gases were swept outwards, leaving the inner planets Mercury, Venus, Earth and Mars as rocky bodies with metallic cores. But in the outer solar system, where the temperature was much lower, the icy material remained stable and so the planets Jupiter, Saturn, Uranus and Neptune formed as huge balls of fluid and gas.
Where did asteroids form? Most observed asteroids orbit the sun in the "main belt," lying between Mars and Jupiter. It seems that they would have formed another planet except that the pull of Jupiter's gravity stopped this from happening. Occasionally, Jupiter's gravity gives an asteroid a strong enough tug to knock it out of the main belt, and into a more elongated egg-shaped orbit crossing the paths of the inner planets, Earth included.
The first asteroid was discovered on the first day of the 19th century; it is about 600 miles across. Other large asteroids followed, all of them far away, in near-circular orbits between Mars and Jupiter. In 1898 the first earth-approaching asteroid, called Eros, was discovered. It comes close, but cannot strike Earth. (At least not soon: computer modelling of its path indicates that it could collide with us within the next million years.) We know a lot about Eros, because Nasa has had a satellite orbiting around it for a year.
In the 1930s three asteroids which could collide with the earth were spotted, one which comes as close to us as the moon. (That asteroid, Hermes, was soon lost, and may return.) As the decades passed, a few more such bodies were added to the data banks.
With improved technology, discovery of earth-approaching asteroids has greatly accelerated. The first dedicated search projects began in the 1970s, in California; before then all discoveries had been the accidental side-effect of other astronomical work. By the mid-1980s several dozen such objects were known. Late in that decade the first project began using an electronic camera to scan the sky automatically, at the University of Arizona. At about the same time I began a new search programme, using a wide-field telescope owned by Britain and located at the Siding Spring Observatory in Australia. We carried on looking for seven years, until the money ran out.
The discovery surge
The real blossoming in asteroid discovery rates has come in the past three years. This is largely because of two separate projects, using similar telescope systems, operated by the US Air Force. One is located at an observatory on the island of Maui in Hawaii, the other in New Mexico. Both are good sites, but their productivity stems from having wide-field telescopes able to skip quickly from one part of the sky to another, together with the latest in electronic detectors, developed as part of the Star Wars programme. Several other asteroid search projects are underway in the US-three in Arizona alone-and Nasa is now spending several tens of millions of dollars a year on a posse of satellites that will reconnoitre various asteroids and comets over the next decade or so. The net discovery rate is now about one earth-approaching asteroid per night.
The growing number of "possible asteroid impact" stories in the newspapers in recent years comes in part from the rate at which new objects are being found, but also from the theoretical work being done on their likely future movements. Whereas science fiction depicts just a few days or weeks notice before an impact, it is now possible to foresee a collision some decades ahead. Achieving this requires very precise observations and software which can follow how the asteroid trajectory will alter bit by bit under the gravitational tugs of the planets. Any asteroid due to strike the earth would most probably make several near-misses before actually colliding, each time suffering small but significant deviations of its path. It is only in the past two years that the software necessary to trace this celestial dance has become available.
In essence, the programmers set up millions, or even billions, of hypothetical asteroids in their computers, based on the best available data from astronomers' observations. They then extrapolate the movement of these objects over the next century or so, and look for possible collisions. The accuracy of the original data is crucial, because slight errors in the starting positions could mean the difference between a projected hit and a miss.
Each night's new data leads to a re-run of the software. In most cases, the results show that an impact is not possible. Of some concern are those for which the probability grows. Most worrying are those that have significant collision probabilities but become "lost"-insufficient information was obtained to enable a telescope to be pointed at a particular part of the sky and pick them out. We have a few of those on file. We must hope that we'll find them again before they find us.
How many earth-approaching asteroids remain to be discovered? It all depends on the size that interests you. There might be up to 2,000 objects larger than half a mile in size. Something like 40 per cent of those are in the data banks, but it would take over 25 years to find the rest at the current discovery rate. So if there is one with our number on it, due to strike in, say, 2005, then it's unlikely that we'll see it first. We must raise our game.
These are only the large ones, though. There are many more small asteroids. If we decided to go down to a quarter of a mile in size, then we'd need to find around 10,000 objects, and so far we have less than 10 per cent of those. It might be judicious to look for still smaller objects. In 1908, a lump of rock 60 or 70 yards across entered the atmosphere above Siberia. It blew up at an altitude of about five miles because of the hypersonic shock experienced entering Earth's atmosphere. (Asteroids typically hit the atmosphere at a speed of 15 miles per second.) Even though it did not reach the ground intact, the ten megaton blast this natural missile produced ignited and flattened almost 1,000 square miles of forest. If the next one arrived above London, everything within the M25 would receive the same treatment. Such events may be expected somewhere around the world about once per century.
Given that frequency, one might expect several such calamities to have occurred over populated regions in recorded history. And it seems they have, but were not recognised as asteroid events. A meteorite explosion in China in the 15th century is thought to have killed 10,000. Another possible impact, in the 12th century in New Zealand, a couple of centuries after the Maoris arrived, has been mooted. France may have suffered a mid-9th century event. And the conditions leading to the Justinian plague of 540AD may also have been caused by an encounter with a comet.
The avoidance strategy
What could we do if we discovered an asteroid was due to hit us soon? For a small asteroid, evacuate the target area. For a large one (if nothing else) we could lay in food supplies. The main cause of death after an impact might be starvation, because the amount of dust thrown up into the stratosphere would obscure the sun for several years, leading to global cooling and the loss of entire growing seasons. (Big impacts produce vast craters, injecting many times their own weight of pulverised rock into the upper atmosphere.)
It does appear, though, that we have the technology to divert such a threatening missile, given enough warning. These projectiles move through space at hypervelocity, but giving one a nudge so as to change its speed by just a few inches per second would be enough to get it to miss Earth. However, a substantial lead time would be essential.
The only feasible technology for knocking an asteroid off its course would be a nuclear weapon. But we would have to nuke our cosmic enemy gently. A nuclear explosion on its surface would break the asteroid into hundreds of fragments, most of which would still be heading for Earth. We would merely have turned a cannonball into a shotgun blast.
Instead, we would use a stand-off detonation, maybe a mile above the asteroid's surface. In space there is no air, and so no blast wave of the type one sees in films of nuclear tests. However, the neutrons and X-rays generated in the explosion would irradiate the asteroid, causing the nearer surface to be heated very substantially. A layer, perhaps a foot to a yard deep, would evaporate, but on only one side of the target. And that would give it the little sideways shove required.
The dawn of the space age, which has made it possible to spot asteroids and understand their threat, has also equipped us to tackle any object which happens to have a date with Earth. In 50 years time, our descendants will laugh about "the asteroid threat," regularly deflecting them to preserve the planet. But we must ensure that we reach that stage.
The international surveillance programme, as recommended by the government's recent task force, is critical. The task force, which reported last September, included Harry Atkinson, a long-time science administrator, Crispin Tickell, former British ambassador to the UN, and David Williams, former president of the Royal Astronomical Society.
Their report confirmed the nature and level of the impact hazard: a small probability of occurrence, but with such phenomenal consequences that we must take it seriously. The recommendations made covered a suite of actions that Britain could take: alone, in collaboration with the US (the world leader in this work by far), or jointly with European and other nations. Already several countries have started well-funded research on specific aspects of the problem. Japan has built a new observatory dedicated to looking for Earth-approaching asteroids, while Italian mathematicians have written sophisticated software to improve estimates of the chances of particular asteroids running into us soon. If the government followed the task force's recommendations, it would cost about ?10m each year. Specifically, Britain would lead a pan-European effort, including the use of a special search telescope in the southern hemisphere. (Currently no one is searching the southern skies, so we are partly exposed.)
Life on earth has existed for at least 3.8 billion years. Until 570m years ago it was little more than slime. The top predators have varied from era to era. Geology and palaeontology indicate that evolution has been punctuated by the occasional arrival of rocks from space. Astronomy tells us that such impacts must have occurred, and will happen again unless we intervene. The dinosaurs couldn't see their nemesis coming. We can.
