15 May 2015

Fun with Killer Asteroids

I wrote the post below in January of this year for my WIRED Science Blog Beyond Apollo during a senseless media frenzy centered on a not-very-threatening near-Earth asteroid called 2004 BL86. Over the past several days, some media outlets have sought to do the same with asteroid 1999 FN53, a roughly 700-meter-wide space rock that passed about 10 million kilometers from Earth yesterday (Thursday, 14 May 2015). Probably it's a good time to repost.

Itokawa, sampled in 2005 by Japan's Hayabusa spacecraft, is in the same size class as 2004 BL86. Image credit: JAXA
To date, human beings have spotted more than 500,000 asteroids. These range in scale from 950-kilometer Ceres, the first asteroid discovered, way back on the first day of the 19th century, to unnamed boulders. Little asteroids (say, the size of a bus or a house) far outnumber the big ones.

Ceres resides in the Main Belt between Mars and Jupiter, as do the vast majority of asteroids. Only about 15,000 follow paths around the Sun that bring them near Earth's orbit. That's a crucial detail, by the way; they approach Earth's orbit regularly, but not necessarily Earth itself. As with the Main Belt asteroids, little near-Earth asteroids greatly outnumber large ones.

The largest of the near-Earth asteroids is 1036 Ganymed, which measures about 33 kilometers across. It has a stony composition much like that of the second-largest near-Earth asteroid, banana-shaped 433 Eros, which measures 34 kilometers by 11 kilometers. Eros never draws nearer than about 27 million kilometers from Earth, or about 70 times the Earth-moon distance; Ganymed never passes closer than about 56 million kilometers. Both of these small bodies were discovered before 1925.

Eros is unique because a derelict American spacecraft called NEAR Shoemaker rests on its surface; though designed as an orbiter, it landed on Eros on 12 February 2001, at the end of its mission, and continued to transmit for about two weeks. Eros has peculiar "ponds" made of fine dust; it is thought that NEAR Shoemaker happened to fall on one, softening the force of its impact.

A day or so ago, a 325-meter asteroid designated 2004 BL86 passed Earth. To get a sense of perspective, 325 meters, or roughly as wide as the Tour Eiffel is tall, is kind of big for a near-Earth asteroid. As asteroid flybys go, it was a close shave; 2004 BL86 passed about 1.2 million kilometers from Earth. That distance is a bit more than three times the distance between the Earth and the moon.

Any time an asteroid is due to pass Earth – even if it will pass more than a million kilometers away – the popular-audience space media kicks into inaccuracy overdrive. Adjectives I heard used to describe 2004 BL86 included "giant," "huge," "mountain-sized," and "dangerous." Phrases used to describe its minimum-approach distance included "so close you’ll be able to see it," "very close," and "a close encounter." None of this language was accurate. One media source even called it the biggest asteroid to approach Earth in 200 years; in fact, this was the closest approach of this asteroid for 200 years.

The news media are not the only ones that commit such errors. Space educators who should know better also play up the "threat" from "killer" asteroids when a body like 2004 BL86 passes the Earth-moon system. They place objects like 2004 BL86 in the same category as the "dinokiller" that struck Earth 65 million years ago.

Doing that falls short in the reality department in at least a couple of ways: for one thing, the impactor that ended the Cretaceous was an extraordinary object for a near-Earth asteroid, of roughly the same magnitude as Eros or Ganymed, and such bodies hit Earth on time scales of tens of millions of years; for another thing, an object nearly as large as the dinokiller impactor struck Earth 35 million years ago where now Chesapeake Bay is located, and it caused no mass extinction. It wasn't even known until the mid-1990s, and then only because one of its byproducts is brackish well water in eastern Virginia. So, even for truly big impactors, mass extinction is not inevitable.

Because of the poor quality of information they receive, many people with only a casual interest in space have developed the mistaken notion that asteroids are frightening things. In fact, they are data-packed fossils of the formation of our Solar System. The appropriate emotion to feel when one of these objects passes by Earth is not fear; it is fascination.

As proof of the sheer nifty-ness of asteroids, I offer this: as 2004 BL86 passed Earth on 26-27 January, observers aimed telescopes at it. By carefully gauging changes in the amount of sunlight reflected off the asteroid, they found that it might not travel through space alone. Earth-based radars then confirmed that a moon about 70 meters across circles 2004 BL86 at a distance of about 500 meters. How cool is that?

I think by now you realize that I do not endorse exploiting asteroids to scare people, no matter how slow a news day it might be. Just for grins, though, how about we imagine that 2004 BL86 tried to live up to the fearsome adjectives used to describe it and had actually struck the Earth?

The nice people at University College London (UCL) and Purdue University have conspired to create a handy online impact modeling tool called "Impact: Earth!" I prefer the less graphics-intensive 2010 version – to be found here – which is called, more prosaically, "Earth Impact Effects Program." The latter operates faster and allows me to use my imagination more.

The minds behind this modeling tool are careful to warn us that it might not be perfect. In fact, they warn that, if one enters "peculiar impact parameters" they refuse to be responsible for what happens. The model does, however, provide results in line with those arrived at in scientific studies of impact effects, and the explanatory document that accompanies it is convincing.

We know from spectral analysis that 2004 BL86 is another stony asteroid like Eros and Ganymed; they are quite common. To be more precise, it is a V-type asteroid, meaning that it might well be a chip knocked off Vesta, the third-largest and second-most-massive asteroid in the Main Belt. We know that, given the shape and tilt of its orbit about the Sun, 2004 BL86 might be a bit more likely to intersect Earth near the equator than near the poles. Now we know that it has a moon, which should be taken into account when modeling impact effects.

So, first we choose an impact site. I spin my 16-inch Earth globe – around and around it goes, and where it stops, nobody knows – and halt it with my finger. I look at the place that I have picked: it is in the Pacific just east of the Japanese island of Honshu. I do not like that choice; after all, the poor folks there are still picking up the pieces after the giant earthquake-tsunami-reactor meltdown disaster of 11 March 2011, and a nearby impact would be piling on.

I spin the globe again; this time my finger falls on the Atlantic Ocean about 300 kilometers east of The Bahamas. People there have to deal with killer hurricanes, but if this experiment is to have meaning I have to be dispassionate. So, it's east of The Bahamas for our impact site (sorry, Bahamians and their neighbors).

The modeling software allows me to select my distance from the impact point. Of course, I am tempted to put myself far enough away that I could conceivably be in a cafe in Paris, but I will instead suck it up and put myself in harm's way. I'll imagine that I am in Puerto Rico, about 300 kilometers south of the impact point. After all, I've long wanted to visit Puerto Rico to see Arecibo Observatory and old San Juan.

Next, I will enter the impactor's size, starting with 2004 BL86 by itself (I will add the newly found moon later). Now I need to decide on its density. I select "dense rock" with a mass of 3000 kilograms per cubic meter. The average asteroid impact velocity is 17 kilometers per second, but I will ramp it up a bit to 23 kilometers per second because of the shape of 2004 BL86's orbit about the Sun. The most probable impact angle is 45°, so I’ll go with that. I want to avoid "peculiar impact parameters," after all.

Almost done. The last step is to define the target density. Three hundred kilometers east of The Bahamas is deep ocean. In fact, the deepest part of the Atlantic, the Puerto Rico Trench, is close by. I enter a target density for "water" of 1000 kilograms per cubic meter and a compromise depth of 3000 meters.

OK. All set. Here comes our asteroid. I click on the "calculate effects" button. According to the model, an impact on the scale of a 2004 BL86 impact occurs – can this be right? – about every 84,000 years. That seems pretty often, but it is 10 times longer than recorded human history.

2004 BL86 begins to disintegrate 59 kilometers above the ocean. It shatters into many small pieces by the time it hits the water. The pieces splash down in an ellipse measuring about 0.9 kilometers long by 0.6 kilometers wide. This produces a "crater" – a splash, really – about 7.9 kilometers wide. Fragments reach the sea floor, forming a submerged crater field. The largest crater in the field measures 194 meters across by 69 meters deep.

The impact fireball is below the northern horizon as viewed from Puerto Rico, so I feel no wave of heat from the impact. If the impact took place at night, I would see a brilliant flash on the horizon. The seismic effects at the impact site resemble a magnitude 3.6 earthquake. Three hundred kilometers away in Puerto Rico, I feel nothing.

For people used to hurricanes, the atmospheric effects of the hypothetical 2004 BL86 impact are a walk in the park. The roar of the impact is about as noisy as loud traffic. Air blown outward from the impact site reaches Puerto Rico traveling at a speed of 7.61 meters per second, or about 17 miles per hour.

The tsunami the impact generates reaches Puerto Rico's north coast 35 minutes after the impact. The wave is 14.4 meters high or less. Some coastal towns are inundated.

It's important to point out that we knew of the 2004 BL68 flyby well in advance. We can thus assume that we would know of a 2004 BL86 impact well in advance. It would not have been too difficult to calculate where it would hit. Because of this, it is reasonable to assume that the coastal towns could have been evacuated before the impact occurred. We can also assume that ships and aircraft would be kept out of the impact area in the hours before 2004 BL86 struck; these steps, not much different from those taken ahead of a hurricane or during a volcanic eruption, would dramatically reduce loss of life and damage to property.

What about 2004 BL86's 70-meter-diameter moon? I leave all the impact model parameters the same except the impactor diameter and click the button. The moon barely reaches the ocean surface, creating no splash and barely any wind. Its effects are lost among those of 2004 BL86. According to the UCL/Purdue model, lone impactors the size of 2004 BL86’s moon hit Earth every 2200 years; given that our recorded history is not pocked with accounts of such impacts, it would seem that when objects that size strike Earth, they are not much noticed.

These results are suggestive, not definitive. I will repeat that the modeling software is by its own designers' admission not perfect. Though I would defend my inputs as plausible, GI/GO applies. The point is, however, that it seems that a body the size of 2004 BL86 does not have much effect on the Earth when it strikes. No mass extinction occurs, the climate does not shift to some new harsh state, and the effects on human lives even close to the impact site are similar to those that people have long endured because of volcanoes, hurricanes, tornadoes, earthquakes, and warfare.

Do I argue here that we should treat near-Earth asteroids as a non-threat? Of course not. We should find all of them. We have the technology to do that (and by some estimates, we're almost done). We should also test techniques for deflecting them away from Earth. As we do these things, we can study them to learn more about our Solar System.

Perhaps along the way we can teach ourselves how to make mining them profitable. As far back as the 1960s, some people – notably Dandridge Cole – suggested that we convert asteroids into habitats or interplanetary transports. Isaac Asimov once called them "stepping stones to the stars." That is, he suggested that asteroids might enable a slow human migration outward that might never end.

Many of us have become convinced that every asteroid is a killer. That belief is a downer and the evidence contradicts it. However, that people are willing to believe something about anything as esoteric as remote rocks in space is interesting. It causes me to wonder whether people might be ready to sign up for a more hopeful vision of asteroids. Could we become as excited about the certainty that asteroids are a new frontier for exploration and the possibility that people might live among them as we are now about the remote chance that one might destroy us?


  1. I think you discount the 14.4m tsunami... The 2011 Japan tsunami was 'only' 3m. That would produce a wave almost five times the size! More than 'a few' cities will be inundated. That's far larger than the swell - and more energetic - than a hurricane surge. o-o

    1. Crissa:

      I did some checking on the NOAA website regarding the Japan tsunami, and the results are confusing. Some charts show a max height of 9.5 meters but the text mentions a max height of 38.9 meters. There's reference to fishing equipment found 30 meters up a hillside and water surging inland up to 10 kilometers. Three-meter heights occurred, too, but farther from the earthquake epicenter.

      My assumption is that the smaller number is the actual tsunami height at the coastline and the higher numbers record surges caused by water being funneled inland by harbors and rivers.

      14.4 meters is the maximum ("14.4 meters or less") the impact would produce, and the word I used concerning number of coastal towns affected was not "few," it was "some." It is not possible to give a precise number of towns that would be affected, so I did not try. The word I selected to describe the number of towns affected does not minimize the potential number, it expresses uncertainty about the potential number.

      I checked some storm surge numbers for hurricanes. Katrina's surge reached 8.5 meters, but maximum water level reached 10 meters. Go figure. Hurricane Camille in 1969 set the previous records, and before that the record-setter was an unnamed 1947 hurricane.

      So 14.4 meters is within the range of human experience in the past 50-60 yrs. That is, as I say, a maximum. It could be much less. Given the imprecise nature of the model, it could be more.

      I think the important thing here is that, if we can predict in advance a flyby of a given asteroid, we can predict an impact in advance. Potentially we could predict the impact years or decades in advance. That means we could evacuate affected areas. It also means that we could potentially deflect the asteroid so that it wouldn't hit the Earth at all.

      Given the choice between a hurricane/storm surge that might give me a couple of days of warning, an earthquake/tsunami that gives me perhaps an hour of warning, and an asteroid/tsunami that gives me years or decades of warning, I think I'd choose the asteroid. It has the potential to cause more damage (though that's not certain), this is true; but it also gives us much more time to act.



I like hearing from my readers. No rules except the obvious ones - please keep it civil and on topic.