Some background: yes indeed, the asteroid 9942 Apophis (uh-PO-fis) is going to pass by the Earth on April 13, 2029. Yes, it will pass about 18,000 miles above the surface, and should be visible through binoculars from Asia, Africa, and Europe; at its nearest point it should be moving across the sky about 80 times faster than the Moon! Apophis absolutely will not strike the Earth in 2029.
But the flyby in 2029 will be so close that the Earth's gravitational pull will definitely change the asteroid's orbit. And that's the problem: we don't yet know precisely how much the orbit will be changed. As things stand right now, there is a tiny, 1-in-250,000 chance that the flyby in 2029 will change the asteroid's orbit in such a way that leads to Apophis impacting the Earth in 2036.
The 1-in-37 chance that Apophis would hit the Earth was based on early calculations, when we didn't know exactly where the asteroid was. As more images came in, we were able to better calculate Apophis's orbit, and the odds of impact got less and less.
But how could we not know where an asteroid is? Can't we see it with telescopes?
Yes we can -- but not clearly. When you take a picture of your face, you can do it in such a way that each dot of the photograph is "life size" -- one inch on the photo corresponds to one inch of your face. Things change when you take pictures of distant objects. When photographing a field, one inch on the photo might correspond to 100 feet in real life. For a far off mountain range, one inch might correspond to 10 miles. And when photographing space, one dot might represent 10000 miles.
When photographing an asteroid from Earth, you almost never actually see the asteroid's surface. What you see is a dot of light, meaning that somewhere in a 10000 mile region of space, there is an asteroid. By examining multiple photos of space, you can narrow down where an asteroid is, but there will almost always be some uncertainty.
Former Apollo astronaut Rusty Schweickart started the B612 foundation to come up with plans for what to do about Apophis. The first thing he suggests (which the European Space Agency wants to try) is to fly a lander to the asteroid the next time it comes into the Earth's neighborhood in 2013. The lander will have a radio beacon on it, allowing radiotelescopes on Earth to triangulate the asteroid's position precisely down to a few _feet_. With that information, we should be able to calculate precisely where Apophis will be from now to the year 2070. More than likely, we're going to find that Apophis poses no threat to the Earth, even taking the 2029 gravitational shift into account.
But suppose we get unlucky with our 1-in-250,000 chance, and we find that the 2029 flyby puts Apophis smack into the Earth in 2036. What then?
The smartest thing to do is probably the simplest. NASA astronaut Ed Lu came up with the idea of launching a massive 20-ton spacecraft to fly next to the asteroid without touching it. The gravitational field of the spacecraft will gently tug at the asteroid for years, eventually pulling the asteroid out of the danger zone for striking the Earth. To get faster results, you can add an ion engine to the spacecraft so that it can more effectively tug the asteroid-- this can change the asteroid's speed by 0.02 mph, which is all that you need to save the Earth.
Other ways to move Apophis might entail attaching a solar sail to the rock, so that the pressure of sunlight can gently shift the asteroid -- remember all we need is to change its velocity by 0.02 mph. Other ideas might be spraying ice on one side of the asteroid and/or crushed dust on the other side; the white ice might act like a solar sail, or the dark dust might absorb (and then give off) so much heat that the small thrust should move the asteroid.
The gravitational tug is probably the safest and simplest way to move an asteroid. Unfortunately, the concept wasn't developed until 2005, after the movie Armageddon, so in the popular imagination the only way to destroy an asteroid is to BLOW THE SHIT OUT OF IT WITH NUCLEAR WEAPONS, YEAH!!!
This is probably the stupidest thing we could do. This was mentioned in the movie "Independence Day", when General Gray said "the only thing that would accomplish is turn one dangerous falling object into many"; in other words you'll be just as dead if you're hit with a shotgun blast as if you're hit with a howitzer.
Remember, we're discussing "what if" scenarios: Apophis has a 1 in 250,000 chance of hitting the Earth, and I think the odds are going to go more in our favor as we learn more.
But since you asked, suppose everything fails and Apophis is headed directly for Earth. What happens then?
If the asteroid hits a rocky place, the resulting crater will be 2.2 miles across. That's more than twice as large as Ardsley itself, and about 1/10th the size of Yonkers. But that's just the crater! If you were 28 miles away from the impact, the heat would be about 20 times as intense as the Sun; and will set trees and buildings on fire. Nine seconds later, the seismic shaking will arrive; it will be equivalent to a 6.6 Richter earthquake, and will cause consuderable damage to most structures. A minute and a half later, the ejecta will come raining down; molten chunks of lava the size of soccer balls. But that's OK, because the blast wave, which will arrive 40 seconds later, will feel like a wind 140 miles per hour, and will knock down what remains of the buildings and burning trees.
If the asteroid were to impact Ardsley, all of Westchester/Putnam/ Rockland from Harriman to West Point to Brewster would be destroyed. Fairfield County Connecticut, from Ridgefield to Long Island Sound, would be on fire. On Long Island, pretty much all of Nassau County and a chunk of Suffolk County out to Cold Spring, Huntington, and Bethpage would be wiped off the map. The structurally sound buildngs of lower Manhattan might still be standing, but Queens, the Bronx, Harlem, and northern Brooklyn would be rubble, and Central Park would be one giant forest fire. Most of northeastern New Jersey, from Bayonne to West Milford, would be destroyed.
We're lucky this is NOT GOING TO HAPPEN!