Tracking Near-Earth Asteroids

Bloggers, our modern-day town criers, seemed to follow the celestial soap opera of asteroid 2004 MN4 with a level of urgency topped only by astronomers themselves. Slashdot, an Internet site summarizing science and tech news, posted the first whisperings about the object on Friday, December 24, 2004, at 1:30 PM:

“Introducing Asteroid 2004 MN4: from the hopefully-you'll-hear-nothing-more-about-it dept…Numerous readers wrote in with bits about a potential asteroid collision: ‘…asteroid 2004 MN4 is currently listed as having a 1/233 chance of hitting the Earth…If it strikes the Earth it will release an energy of 1,900 Megatons of TNT….’ So, in summary, there's a 1-in-233 chance of the worst disaster in recorded history happening on April 13, 2029, and a 232-in-233 chance of nothing happening. Have a nice day!”

As for the worst disaster in history part--too early to tell. If it struck land, the 320 m diameter rock could pummel a crater “only” 3 to 5 km wide (more of a regional than a global catastrophe). If it struck ocean, it could kick-start a tsunami to rival the Indonesian one that occurred just two days after this post appeared. Still, in the astronomy business, odds as low as 1 in 233 are, well, odd. And the ensuing plot twists only got more nail-biting. As posted on Slashdot:

December 24, 22:14 PM: Update: A 1 in 62 chance.

December 25, 6:31 PM: “2004 MN4, Even Higher Probability” 1 in 45.

December 27, 3:56 PM: “2004 MN4 Asteroid Odds Inching Up Again.” Now 1 in 37: the highest risk ever reported in space rock history.

December 27, 8:44 PM: “2004 MN4 Probably Won't Kill Us.” The odds had sunk like a stone, to 1 in 56,000.

Like any good drama, this one is still unfolding. While the rock will miss us in 2029, in February scientists had a new prediction: it may have a second chance at Earth collision, in 2034.

Asteroid orbit paths don’t veer wildly in a matter of days or months. So why did 2004 MN4’s chances of impact do just that? A closer look at each episode of the saga reveals how astronomers calculate orbits, how impacts are predicted, and just how much you should bite your nails.

Episode 1: Discovered!

The asteroid’s debut was unremarkable. It was first noticed in June 2004 the way all near-Earth objects (NEOs) are: through a telescope. To qualify as an NEO, an object must be an asteroid (a chunk of rock or iron) or a comet (an icy chunk of rock) that passes within 200 million km of the Sun at some point on its loop around it. (Earth orbits the Sun at 150 million km away.)  Three astronomers searching for a different asteroid with the University of Arizona’s Kitt Peak telescope found this one instead. They took photos of it low in the western sky six times in two days.

In each black-and-white frame, stars glow brightly, yet remain still from shot to shot. Any starlike point moving among them is likely an asteroid or comet. For each image, the team identified the object's coordinates in space, then extrapolated a range of possible orbit paths called solutions.

Episode 2: Lost … or Hidden?

For six months, 2004 MN4 was too faint and too hard to find to show up on any observer’s telescope.

Episode 3: Reemergence

On December 18, 2004, an Australian astronomer spotted what he thought was a new NEO. He emailed his findings to the Minor Planet Center, a group at the Harvard-Smithsonian Center for Astrophysics which collects, catalogs, and disseminates data on tens of thousands of asteroid observations per day by astronomers around the globe.

Astronomers at the Minor Planet Center identified the Australia object as 2004 MN4. In an instant, its observation arc went from short (two days) to substantial (six months). The farther apart an asteroid’s observations are spaced, the more accurate the estimates of its complete orbit. The Minor Planet Center was able to predict that the margin of error for 2004 MN4’s orbit would cross Earth’s orbit. The object’s impact probability--the likelihood that it would actually hit us--was also calculated: a 1 in 2,500 chance of collision on Friday, April 13(!), 25 years hence. Less than half a percent of the 3,000 known NEOs have ever achieved such a high impact probability. 2004 MN4 became an overnight celebrity.

Episode 4: Conflict

“Semipro” astronomer Roy Tucker spent the holidays tracking 2004 MN4. Tucker had a personal stake in the outcome: he was one of the three observers who had discovered the object at Kitt Peak. Tucker is an instrumentation engineer for the University of Arizona by day, and takes 30,000 image sets of starfields a year by night, mostly via $13,000 worth of observational equipment sheltered in a white wooden pop-top shack in his backyard.

For three nights, Tucker contributed his data to the Minor Planet Center’s accruing pile from other worldwide observers. Usually, further observations of an object quickly eliminate its impact probability. But the odds, surprisingly, kept rising daily. “It was kind of creepy,” laughs Tucker. “I wondered if I should quit sending in observations.” By December 26, the asteroid reached the highest-ever Earth-impact risk: a likelihood of 1 in 37. Tucker says he wasn’t worried. “But now that I think about it, if someone told me there was a 1 in 37 chance that my plane would crash, I probably wouldn’t get on it.”

Episode 5: Pre-covery

On December 27, two astronomers from the Tucson observatory Spacewatch hunted for 2004 MN4 not in the sky, but in their file cabinets. They found a set of photos from March 2004 with a dot so dim it had been overlooked. It was 2004 MN4. The asteroid’s observation arc was now nine months long. Since 2004 MN4’s orbital period is just over 10 months, the nine months of data was more than enough to recalculate its orbit solution more accurately. As a result, the object’s probability of Earth impact plummeted from 1 in 37 to 1 in 56,000. The tension over, most fans of 2004 MN4 moved on to the next big rock.

Episode 6: A Close Shave

Certain asteroids are interesting enough, large enough, and likely to pass close enough to Earth to warrant a few hours of coveted observation time at one of the world’s two radar telescope observatories. So far, about 300 asteroids, both NEOs and those in the main asteroid belt between Jupiter and Mars, have been analyzed with radar. Radar can yield a range of details far more precise than can optical equipment, including the object’s speed, distance, size, shape, spin, and surface properties. In January 2005, astronomers imaged the barely-radar-detectable 2004 MN4 at Puerto Rico’s Arecibo Observatory. The data revealed that the March optical observationsthe ones that inspired so much hubbubwere off-target.

Jon Giorgini, an analyst in the Solar System Dynamics Group at NASA’s Jet Propulsion Laboratory, calculated the new radar-based orbit solution. “The solution actually moved much closer to Earth than it had been before,” he says. The data revealed that we’re in for a much closer shave in 2029 than the optical data from had predicted2004 MN4 will pass a mere 36,400 km away from us, one-tenth the distance to the Moon. (In fact, the radar solution shifted the orbit to a path that the 10-month optical observations were 99.99995 percent certain it wouldn’t go!)

But the good news is twofold: the radar solution is so precise that astronomers are indubitably clear that it will miss Earth in 2029. (The impact probability went from 1 in 56,000 to an infinitesimal numberostensibly zero.) The other bonus is that the object will be visible to naked eye that year, a bright, fast point of light careening through the constellation Cancer. No other known asteroid has been visible that way.

Episode 7: Fear the Future?

Asteroid 2004 MN4’s close brush with Earth in 2029 means that our planet’s gravity will perturb its path in a way astronomers cannot predict perfectly at the moment. Unlike the stable, utterly predictable orbits of planets, asteroid orbits are chaotic. This means that a small influence on an asteroid’s path at one moment could hugely offset its trajectory many years in the future. “The uncertainty region, the region of space where it could statistically be, gets spread out as a result of its near-Earth encounter in 2029,” says Giorgini.

Current predictions are that 2004 MN4 could come close by Earth in 2034, or perhaps in 2035 through 2038, or in later years this century. Determining exactly how close, however, will have to wait until 2012, the next time 2004 MN4 is in radar range. “If we get radar data in 2012, we should be able to predict its position out to 2070 reliably,” says Giorgini. “It might be possible to exclude encounters in the 2030s sooner than that, but not the later encounters. Not until we get those measurements.” Seven years to find out this story’s next chapter? Now that’s some suspense.

The whole idea,” says Brian Marsden, director of the Minor Planet Center at the Harvard-Smithsonian Center for Astrophysics in Cambridge, Massachusetts, “is to try and find these things before they find us.” Given the potential for asteroids to literally and figuratively impact life on Earth in a profound way, asteroids have been quite sought after since the first and largest one, Ceres, was discovered in 1801. After 1898, when astronomers discovered that the 433rd, Eros, frequently passed within only 19 million km of Earth, the pursuit got even hotter.

Cambridge has been an asteroid-and-comet-hunting hub for most of that time. In 1839, Harvard set up an observatory there. In 1850, the first daguerreotype (an early photograph) of a star (Vega) was taken with the Harvard College Observatory’s world-renowned 15-inch refractor telescope. This paved the way for photography and other imaging techniques to revolutionize astronomy as the decades passed. In fact, everything we know about a near-Earth objectan asteroid or comet that passes through Earth’s neighborhoodbegins with a set of black-and-white images of it against a starry background.

Astronomical Paparazzi

In 1882, an astronomer in South Africa borrowed a reporter’s camera, tied it to a telescope, and took the first modern-process photograph of a celestial object. It was of the comet 1882II, which is visible to the naked eye. Unlike daguerreotyping, this technique used negatives. It also allowed for long exposures to light streaming from stars, asteroids, and other space objects too faint to see even through a telescope eyepiece. As a result, the photographs revealed 1882II flanked by a background of never-before-seen stars. The pictures became a public sensation. Harvard College Observatory began aggressively photo-documenting the sky that year. By the time it stopped in 1989, the observatory had amassed half a million glass photographic plates of the entire available heavens, imaged many times over.

“Technically, each Harvard plate could contain at least one asteroid, if one were to look for it,” says Marsden. Since asteroids reflect sunlight, they look like starry points among the real stars. (The word “asteroid” actually means “starlike.”) “When astronomers look for asteroids, they point the telescope at a patch of sky,” says Tim Spahr, technical near-Earth asteroid specialist for the Minor Planet Center. “They take a long-exposure image, and repeat. Then they’ll ‘flip’ through the pictures in the set. The stars appear stationary, but anything close by, like an asteroid or comet, will move.”

As rocky relics from the formation of our Solar System, most asteroids reside in the main asteroid belt. This ring of dust, asteroids, and meteoroids (smaller asteroids) circles the Sun at a distance of around 400 million km, sandwiched between the orbits of Mars and Jupiter. (Earth orbits the Sun at 150 million km away.) Sometimes, Jupiter’s gravity or other factors will “kick” an asteroid out of the main belt into a new orbit. Some of these ejected objects have a chance of intersecting Earth’s orbitor Earth itself someday.

Stars are orders of magnitude more distant than asteroids. The nearest, after our Sun, is about 40 million million kilometers away. Stars aren’t actually stationary, but because of their great distance from us, they appear at a standstill when photographed over time. The motion of asteroids, comparatively, is obvious.

A handful of photographs from one or two nights of observation reveals only a few details about a newly discovered asteroid. Astronomers can calculate its velocity (speed and direction) and its two-dimensional position in space relative to our Earth vantage point. But its true position in 3-D spaceits distance from Earthis trickier. “Imagine looking at a plane overhead. Is it 30,000 feet away? 50,000 feet? At this point, in all honesty, we make an educated guess,” says Spahr. Only by making more observations of an asteroid’s motion, then applying the physical laws of how objects orbit the Sun, can astronomers calculate its exact position and path through space.

Astronomers can get a preliminary idea of an asteroid’s composition, size, and shape at this stage by analyzing its reflected light. For example, measuring the light’s spectra, or its component wavelengths, can determine if the object is made of dull stone or brighter iron, nickel, and other metals, or some combination. Huge numbers of early star and asteroid spectra were taken at the Harvard Observatory, by placing a prism over the telescope lens.

All Automated, All The Time

Such asteroid analysis, which was hand-worked a century ago, is now fully computerized. “These days, astronomical observation is almost never done by a person looking through a telescope,” says astrophysicist Grant Stokes. Still, it wasn’t until the late 80’sthe 19 80’swhen photography’s successor came about: the CCD, or charge-coupled device.

Stokes is the head of the NASA-funded Lincoln Near Earth Asteroid Research (LINEAR) program. LINEAR’s two telescopes, located near Socorro, New Mexico, hunt asteroids in fulfillment of a mandate Congress issued to NASA in 1998 to find 90 percent of near-Earth objects a kilometer in diameter and larger by 2008. “A kilometer is near the threshold for global environmental damage if the object indeed hits Earth.” says Stokes.

Every clear night, automated digital cameras on LINEAR’s telescopes image large sections of the sky accessible from the New Mexico site. Five frames of each section are collected over a two-hour period. A 6 cm x 5 cm CCD chip in each camera acts as an electronic photographic plate, but is a hundred times more sensitive, so long exposures are no longer needed. When photons of light hit a grid of sensors on the CCD, the asteroid’s brightness can be captured, quantified, and analyzed much more precisely than with a photograph. Also, because the CCD output is digital, LINEAR’s computers can find asteroids in it immediately.

LINEAR takes upwards of 6,000 frames per night and can cover all the available sky in a month. Objects moving with unusual speeds or directions are flagged as “interesting”: potential near-Earth objects. The nightly lot of positional data is then handed over to the MinorPlanetCenter. So far, LINEAR and the several other large sky surveys have found 700 km-and-up asteroidsan estimated 60 percent of the objects in the Congressional mandate.

Orbit Hunting

Each day, the Minor Planet Center computes orbits for every single object reported from observatories worldwide. “It takes probably 6 to 10 images over several nights to determine an orbit,” says Spahr. “And to make a prediction of how close that could come to Earth in the future, we need more observationsover several weeks or months.”

Minor Planet Center posts the initial orbit solutions of a few dozen of the new “interesting” objects online per month. That way, astronomers at universities and amateur observatories--often one person sitting in their den, checking their backyard telescope via computer--can follow up with more observations. If LINEAR is the bulldozer in a gold mine, these smaller outfits are the pan-sifters, pointing their smaller-field-of-view telescopes at the right patches of sky to track intriguing objects.

The longer the baseline of observations, the more accurate the orbit calculations get. “We can, of course, observe a new object in the future,” says Marsden. But for near-Earth objects, “astronomers are often impatient to find out whether there really is a danger to Earth. So it helps a lot if we can find images of them on old photographic plates.”

“Over geologic time there's a hundred percent chance of a very large asteroid impact on Earth,” says Stokes. “But will it happen in the next hundred years, in the next couple of hundred years?  It's highly unlikely.” Still, the risk is clearly worth the government funding projects like LINEAR and the Minor Planet Center. If an object is found to be Earth-bound for certain, the save-the-world idea is that we’d somehow nudge it off course before it has a chance to arrive.

“Suppose that $3 million does ensure us against a really large object that would put an end to the human race,” suggests Marsden. “Well, the world’s gross product is about $30 trillion, so here we're spending $3 million to save $30 trillion. That's a pretty good insurance rate.”

Sixty-five million years ago, an asteroid half as long as Manhattan slammed into the tip of Mexico’s Yucatan Peninsula, forming a crater near the present-day coastal village of Chicxulub. Also about 65 million years ago, half of all the plant and animal species alive at the time--including dinosaurs--died out. Coincidence? AMNH geologist Dr. Denton Ebel doesn't think so. His latest research, conducted with the University of Chicago's Lawrence Grossman, also adds never-before-proven details about the impact.

Can you see the Chicxulub crater?

It’s mostly offshore, but a tiny bit is on the peninsula. You can actually see this edge from space, but it’s very subtle. The crater is huge. It’s the second-largest on the planet. The asteroid itself would have been about 10 kilometers in diameter to make that crater. Compare that to Mount Everest, which is 7 kilometers high. When the tip of this asteroid was hitting Earth, its top end was up past where you can breathe.

Clearly there’s no asteroid there now. So besides the crater, what evidence do scientists have that one struck?

In the early 1980’s, geologist Walter Alvarez was looking at the fluctuations of meteorites hitting Earth over geologic time. His father, a Nobel-Prize-winning physicist, helped. They went to Italy to study layers of earth sediment, looking for trace elements that are rare on Earth’s surface but common in meteorites. They were surprised to find a significant spike of iridium [an element abundant in some asteroids] in one sediment layer. Since layers of sediment can be dated, the iridium layer turned out to have been deposited 65 million years ago. The Alvarez team figured a giant asteroid impact at that time was responsible.

It turns out that a mass extinction 65 million years ago happens to coincide with this iridium layer. So scientists began looking more closely at fossil data on different speciesa lot from this museumto build up a more complete picture of the cataclysm. Only after this work was the actual crater found, in 1991.

How did scientists find the craterby satellite?

No. The crater is now mostly buried deep under sediment. When an asteroid impacts Earth, the rocks are reworked in a certain way. Geologists drilling for oil had taken rock cores from the Yucatan. Scientists looked at those cores later on and found these impact signatures. People also found that a layer of spherules [small beads of glassy material] that were thrown up by the impact got thicker and thicker toward the Yucatan area. These spherules are like the splash droplets of melted rock from the impact.

Melted rock from the asteroid itself or from the ground at the impact site?

A combination, but mostly material from the actual rocks on the ground that the asteroid hit. But what you also get with large impacts is a rising column of very hot vaporized rock, including most of the meteorite itself, that forms right where the energy is concentrated.

Kind of like a mushroom cloud from an atomic bomb?

Right — the stem of the mushroom is the material coming up. Then the mushroom cap part spreads out laterally, and the vaporized rock in it condenses into small, molten spherules. When the spherules re-enter the atmosphere, they cool and solidify like little marbles. The impact was so big that the material went into the stratosphere [the atmospheric layer 10-50 km above Earth]. The stratosphere is where the jet stream is. If stuff gets up there, such as all the vaporized iridium from the asteroid and rock from the crater, this air current can spread the material around the world very rapidlylike within a few hours.

Wow. What effect did these glass raindrops have on life on Earth?

When the spherules came back down, friction with the atmosphere created a pulse of hot infrared radiation, like a heating coil in an oven. The heat pulse traveled at the speed of lightmuch faster than the particles. So all at once, all over the planet simultaneously, the heat pulse basically cooked anything that was out there, cooked it alive. But small creatures that burrow underground, or live underwater, or hibernate in caves would have had a much better chance of surviving. These species are the ancestors of all the species alive today.

But the bigger dinosaurs were toast.

Nothing that big survived as far as we know. Some people say nothing bigger than a killer whale survived. Others say nothing bigger than a breadbox! We see evidence of the heat pulse in the sediment layerssoot particles and burned pollen mixed with the spherules.

Literally dinosaur toast. So where does your research come in?

What we did was calculate, for the first time, what the identity and composition of the material condensing from the plume theoretically should have been. And the evidence we have on Earth [the spherules] happens to match our calculations. Also, other scientists thought since spherules were dispersed globally 65 million years ago, their origin must have been many meteorites entering Earth’s atmosphere around the same time. We said, no, you don’t have to invent these hypothetical meteorites, you can get all the global spherules from just one impact instead, one plume which was carried around the world through the stratosphere. And we have the crater to prove it.

What about competing dinosaur extinction theories?

Well, some people think that intense volcanism in India 66 million years ago may have caused global climate change that contributed to the dinosaurs’ extinction. Those eruptions were very slowly oozing out of fissures in the ground and spreading out over large areas. But nothing was shooting high into the air, let alone into the stratosphere. Such oozing volcanoes can cause greenhouse warming in the lower atmosphere, which can raise global temperature over a long period of time. But there's no evidence that that would cause a mass extinction. My research only adds details to the impact that were not there before.

What’s the likelihood of another asteroid as big as Chicxulub happening again?

Every hundred million years is tossed around. The problem with big impacts is that we don’t really know, because these are unique events of which we have very little record. Most asteroid-impact craters get erased over time by erosion, deposition, and/or plate tectonics. Still, the probability is extremely low. And scientists have identified most of the dangerous ones. Put it this way: the probability of someone throwing a meteorite at you is greater than the possibility of you getting hit by one from space.