Since there are many more small asteroids than large ones, run-of-the-mill collisions with the Earth will be made by the little guys. But the longer you’re prepared to wait, the more devastating the impact you can expect. On average, once every few hundred years the Earth is hit by an object about 70 meters in diameter; the resulting energy released is equivalent to the largest nuclear weapons explosion ever detonated. Every 10,000 years, we’re hit by a 200-meter object that might induce serious regional climatic effects. Every million years, an impact by a body over 2 kilometers in diameter occurs, equivalent to nearly a million megatons of TNT—an explosion that would work a global catastrophe, killing (unless unprecedented precautions were taken) a significant fraction of the human species. A million megatons of TNT is 100 times the explosive yield of all the nuclear weapons on the planet, if simultaneously blown up. Dwarfing even this, in a hundred million years or so, you can bet on something like the Cretaceous-Tertiary event, the impact of a world 10 kilometers across or bigger. The destructive energy latent in a large near-Earth asteroid dwarfs anything else the human species can get its hands on.

As first shown by the American planetary scientist Christopher Chyba and his colleagues, little asteroids or comets, a few tens of meters across, break and burn up on entering our atmosphere. They arrive comparatively often but do no significant harm. Some idea of how frequently they enter the Earth’s atmosphere has been revealed by declassified Department of Defense data obtained from special satellites monitoring the Earth for clandestine nuclear explosions. There seem to have been hundreds of small worldlets (and at least one larger body) impacting in the last 20 years. They did no harm. But, we need to be very sure we can distinguish a small colliding comet or asteroid from an atmospheric nuclear explosion.

Civilization-threatening impacts require bodies several hundred meters across, or more. (A meter is about a yard; 100 meters is roughly the length of a football field.) They arrive something like once every 200,000 years. Our civilization is only about 10,000 years old, so we should have no institutional memory of the last such impact. Nor do we.

Comet Shoemaker-Levy 9, in its succession of fiery explosions on Jupiter in July 1994, reminds us that such impacts really do occur in our time—and that the impact of a body a few kilometers across can spread debris over an area as big as the Earth. It was a kind of portent.

In the very week of the Shoemaker-Levy impact, the Science and Space Committee of the U.S. House of Representatives drafted legislation that requires NASA “in coordination with the Department of Defense and the space agencies of other countries” to identify and determine the orbital characteristics of all Earth-approaching “comets and asteroids that are greater than 1 kilometer in diameter.” The work is to be completed by the year 2005. Such a search program had been advocated by many planetary scientists. But it took the death throes of a comet to move it toward practical implementation.

Spread out over the waiting time, the dangers of asteroid collision do not seem very worrisome. But if a big impact happens, it would be an unprecedented human catastrophe. There’s something like one chance in two thousand that such a collision will happen in the lifetime of a newborn baby. Most of us would not fly in an airplane if the chance of crashing were one in two thousand. (In fact for commercial flights the chance is one in two million. Even so, many people consider this large enough to worry about, or even to take out insurance for.) When our lives are at stake, we often change our behavior to arrange more favorable odds. Those who don’t tend to be no longer with us.

Perhaps we should practice getting to these worldlets and diverting their orbits, should the hour of need ever arise. Melville notwithstanding, some of the chips of creation are still left, and improvements evidently need to be made. Along parallel and only weakly interacting tracks, the planetary science community and the U.S. and Russian nuclear weapons laboratories, aware of the foregoing scenarios, have been pursuing these questions: how to monitor all sizable near-Earth interplanetary objects, how to characterize their physical and chemical nature, how to predict which ones may be on a future collision trajectory with Earth, and, finally, how to prevent a collision from happening.

The Russian spaceflight pioneer Konstantin Tsiolkovsky argued a century ago that there must be bodies intermediate ill size between the observed large asteroids and those asteroidal fragments, the meteorites, that occasionally fall to Earth. He wrote about living on small asteroids in interplanetary space. He did not have military applications in mind. In the early 1980s, though, some in the U.S. weapons establishment argued that the Soviets might use near-Earth asteroids as first-strike weapons; the alleged plan was called “Ivan’s Hammer.” Countermeasures were needed. But, at the same time, it was suggested, maybe it wasn’t a bad idea for the United States to learn how to use small worlds as weapons of its own. The Defense Department’s Ballistic Missile Defense Organization, the successor to the Star Wars office of the 1980s, launched an innovative spacecraft called Clementine to orbit the Moon and fly by the near-Earth asteroid Geographos. (After completing a remarkable reconnaissance of the Moon in May 1994, the spacecraft failed before it could reach Geographos.)

In principle, you could use big rocket engines, or projectile impact, or equip tile asteroid with giant reflective panels and shove it with sunlight or powerful Earth-based lasers. But with technology that exists right now, there are only two ways. First, one or more high-yield nuclear weapons might blast the asteroid or comet into fragments that would disintegrate and atomize on entering the Earth’s atmosphere. If the offending worldlet is only weakly held together, perhaps only hundreds of megatons would suffice. Since there is no theoretical upper limit to the explosive yield of a thermonuclear weapon, there seem to be those in the weapons laboratories who consider making bigger bombs not only as a stirring challenge, but also as a way to mute pesky environmentalists by securing a seat for nuclear weapons on the save-the-Earth bandwagon.

Another approach under more serious discussion is less dramatic but still an effective way of maintaining the weapons establishment—a plan to alter the orbit of any errant worldlet by exploding nuclear weapons nearby. The explosions (generally near the asteroid’s closest point to the Sun) are arranged to deflect it away from the Earth.[32] A flurry of low-yield nuclear weapons, each giving a little push in the desired direction, is enough to deflect a medium-sized asteroid with only a few weeks’ warning. The method also offers, it is hoped, a way to deal with a suddenly detected long-period comet on imminent collision trajectory with the Earth: The comet would be intercepted with a small asteroid. (Needless to say, this game of celestial billiards is even more difficult and uncertain—and therefore even less practical in the near future—than the herding of an asteroid on a known, well-behaved orbit with months or years at our disposal.)

We don’t know what a standoff nuclear explosion would do to an asteroid. The answer may vary from asteroid to asteroid. Some small worlds might be strongly held together; others might be little more than self-gravitating gravel heaps. If an explosion breaks, let’s say, a 10kilometer asteroid up into hundreds of 1-kilometer fragments, the likelihood that at least one of them impacts the Earth is probably increased, and the apocalyptic character of the consequences may not be much reduced. On the other hand, if the explosion disrupts the asteroid into a swarm of objects a hundred meters in diameter or smaller, all of them might ablate away like giant meteors on entering the Earth’s atmosphere. In this case little impact damage would be caused. Even if the asteroid were wholly pulverized into fine powder, though, the resulting high-altitude dust layer might be so opaque as to block the sunlight and change the climate. We do not yet know.