In the 1998 film Armageddon a group of oil rig workers lead by Bruce Willis are tasked with preventing a large asteroid from colliding with the Earth and wiping out civilization. They succeed by drilling into the centre of the asteroid and detonating a nuclear bomb, shattering it and causing it to divert from its collision course.

Ignoring the problem of landing on the asteroid and the lack of any formal NASA training received by the oil rig workers, would the approach of detonating the bomb actually work? Do we have a nuclear device powerful enough, and how far away would it have to be detonated to avoid a collision with the Earth?

  • 1
    Of course not - the sorts of tax increases that would be necessary to fund a mission to save humankind would impose unbearingly high costs of business. Decisions about how to deal with massive asteroids are, like everything else, best left to the judgement of individual consumers (yes, this is satire). That said, the second episode of Phil Plait's Bad Universe looked at this and concluded that no, it wouldn't, although I'm too tired to look up a better cite right now :p
    – Jivlain
    Apr 15, 2011 at 13:55
  • @Jivlain - there needs to be no significant tax increase if you get the dumb government bureaucrats and politicians out of the business of building rockets overpriced by orders of magnitude and let private companies do it.
    – user5341
    Apr 15, 2011 at 14:15

2 Answers 2


It certainly would not work as shown in the movie. The asteroid was too close by the time the bomb went off. According to Phil Plait:

Imagine the asteroid four hours before impact. Each half must move away fast enough to cover 6400 kilometers (the Earth's radius) to miss the Earth. Anything less than this means an impact. In turn, this means each half must be accelerated to a speed of 6400 km/ 4 hours=1600 kilometers an hour. That's about 1000 miles per hour, or about twice as fast as a passenger jet. But wait! This asteroid is 1000 kilometers across! It is extremely massive, and something with that much mass would take an enormous amount of energy to get moving that fast; about a hundred billion megatons, or very roughly the same amount of energy the Sun produces every second. Needless to say, one bomb ain't gonna do it. A billion or so might though. I don't think even Bruce Willis is up to that task.

In his book "Death from the Skies", he talks about real ways to prevent an asteroid colliding with the earth. The simple answer is that the further away it is, the less you need to deflect it. Given enough time, a simple 100kg satellite could tow the asteroid into a new orbit that would miss the earth - but you'd need a LOT more than four hours.

  • 2
    Good to know that Plait can analyze situations well enough to debunk ridiculous Hollywood movies. Maybe there's hope for him yet... lol... I crack myself up sometimes... /wipe_tear
    – Michael
    Apr 16, 2011 at 5:14
  • 1
    If you notice the asteroid soon enough, you can simply have a satellite paint one side of it white, the other black. The difference in radiation pressure due to reflecting/absorbing light from the sun is then enough to deflect it (heard this in a talk of an astrophysicist)
    – Lagerbaer
    Apr 16, 2011 at 17:30
  • 2
    Wouldn't that presume the asteroid isn't spinning?
    – fred
    Apr 17, 2011 at 12:35
  • Just as importantly, a one hundred billion megaton bomb going off that close to the earth... will instantly kill the entire population of the earth that is facing the bomb. The other half in the shadow if said bomb might survive the explosion, but the smoke from the resulting fires will probably black out the sun for 2 or 3 years at a bare minimum. Other effects, like the radiation stripping off the atmosphere for example, might have even worse effects. In other words, the cure the movie proposes is just as bad as the disease.
    – Ernie
    Sep 12, 2012 at 18:38
  • ... and zombies.
    – Adamski
    Oct 18, 2013 at 8:01

The answer (like many other scientific answers) depends on the dose - or in this case, the power of the nuke.

For example, according to one set of simulations, a small nuke wouldn't be enough since the asteroid would "reassemble" from blown-apart chunks:

Don Korycansky of the University of California, Santa Cruz, and Catherine Plesko of the Los Alamos National Laboratory in New Mexico simulated blowing up asteroids 1 kilometre across. When the speed of dispersal was relatively low, it took only hours for the fragments to coalesce into a new rock.

However, a big enough explosion may do the job:

"It (the asteroid) would be blown to smithereens," said study lead author Bob Weaver of Los Alamos National Laboratory in New Mexico, who presented the findings here Dec. 13 at the fall meeting of the American Geophysical Union.

"There was a lot of back-of-the-envelope stuff," he told SPACE.com. By contrast, Weaver said, he and his colleagues developed sophisticated codes to model exactly what would happen in such an explosion.

In previous studies, the researchers had modeled how a nuke would affect a solid asteroid about 1,650 feet (500 meters) long and with a shape based on a known oblong space rock called Itokawa.

That work found that a nuke could blast the asteroid into pieces too small to pose much danger. And the explosion would send these pieces flying apart at tremendous speeds - fast enough that they would not re-aggregrate to threaten the Earth again.

In the new study, the researchers modified their codes. The size of the asteroid remained the same, at 1,650 feet long, but now the researchers stipulated that it was 25 percent porous. It was attacked with a 500-kiloton nuclear blast - about 20 to 30 times more powerful than the U.S. blast inflicted on Nagasaki, Japan, to help end World War II.

The researchers exploded the nuke on the space rock's surface, as in previous models. And once again, the asteroid was blown to tiny, harmless bits that spread outward too quickly to be drawn back together. The best results were achieved by positioning the explosive on the asteroid's short side, Weaver said, but any orientation they tried did the job.

  • 1
    Wait a minute, where would the "harmless bits" go (outside of a computer simulation - where tracking billions of particles is hard, so I'd assume deletion under a certain size)? Unless you can blow those off the collision course, you have a cloud of small particles, with approximately the mass of the original rock, on a collision course with Earth. If they would all burn up in the atmosphere, wouldn't that bring a nuclear winter? (plus the planet would absorb the impact in billions of tiny collisions instead of a single huge one, but the total energy transferred would be similar, no?) May 18, 2011 at 15:13
  • 1
    @Piskvor - the problem is not the energy added to the atmosphere, it's energy applied to the surface. So if all the crap burns up in the atmosphere, it's presumably significantly better. Earth gets hit but tons of small scale cosmic garbage continuously.
    – user5341
    May 18, 2011 at 16:07
  • What about the "nuclear winter" scenario though? Small garbage burns up (and becomes dust or soot) - this is not significant as an isolated incident; but if you have a huge cloud of dust entering the atmosphere all at once, this would affect life (and climate) significantly - you may remember the effects of the (relatively minor) Eyjafjallajökull eruption in 2010; a dispersed asteroid could yield an order or two of magnitude higher volume of airborne dust. May 18, 2011 at 16:21
  • (pedantry @self: the term "nuclear winter" is a misnomer, of course, as the effects would be closer to a volcanic winter, with dust entering the atmosphere from outer space instead of a volcano eruption) May 18, 2011 at 16:39
  • @Eyjafjallajökull eruption is totally unrelated - the problem was solid particles that did NOT (unlike fast moving meteorite) burn up in the atmosphere. "Nuclear Winter" is also supposed to be (if it even exists) all smoke and soot
    – user5341
    May 18, 2011 at 16:43

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .