SCIENCE'S DOOMSDAY TEAM VS. THE ASTEROIDS
By Guy Gugliotta
April 9, 2005
Astronomer David Tholen spotted it last year in the early evening of June 19, using the University of Arizona's Bok telescope. It was a new "near-Earth object," a fugitive asteroid wandering through space to pass close to Earth.
Tholen's team took three pictures that night and three the next night, but storm clouds and the moon blocked further observations. They reported their fixes to the Minor Planet Center in Cambridge, Mass., and moved on.
Six months later, Tholen's object was spotted again in Australia as asteroid "2004 MN4." In the space of five days straddling Christmas, startled astronomers refined their calculations as the probability of the 1,000-foot-wide stone missile hitting Earth rose from one chance in 170 to one in 38. They had never measured anything as potentially dangerous to Earth. Impact would come on Friday the 13th in April 2029.
The holidays and the tsunami in South Asia pushed 2004 MN4 out of the news, and in the meantime additional observations showed that the asteroid would miss, but only by 15,000 to 25,000 miles -- about one-tenth the distance to the moon. Asteroid 2004 MN4 was no false alarm. Instead, it has provided the world with the best evidence yet that a catastrophic encounter with a rogue visitor from space is not only possible but probably inevitable.
It also demonstrated the tenacity of the small band of professionals and amateurs who track potential impact asteroids, and highlighted the shortcomings of an international system that pays scant attention to their work.
"I used to say the total number of people interested in this was no more than one shift at a McDonald's restaurant," said David Morrison , an astronomer at NASA's Ames Research Center and a student of near-Earth objects for nearly three decades. "Now it's maybe two shifts." Awareness of the apocalyptic potential of near-Earth objects has been slow to develop. It took years for Nobel laureate Luis Alvarez and his son Walter to win acceptance for their 1980 research showing that a near-Earth object impact quite likely caused the extinction of the dinosaurs 65 million years ago.
"The human brain wouldn't grasp reality until it had somewhat more direct evidence," said Colorado-based planetary scientist Clark R. Chapman of the Southwest Research Institute, another longtime expert on near-Earth objects. "Alvarez provided that."
The vast majority of near-Earth objects are asteroids -- huge rocks or chunks of iron that travel around the sun in eccentric orbits that cross Earth's path periodically. The rest are comets -- ancient piles of dust, stones and ice that come in from the edges of the solar system.
"The good news is that comets represent 1 percent of the danger," said Donald K. Yeomans, who manages NASA's Near-Earth Object Program at the Jet Propulsion Laboratory. "The bad news is that should we find one, there's not a lot we can do about it. . . . We detect them only nine months from impact."
Asteroids, by contrast, generally offer decades or even centuries of warning -- unless they are too small to detect, in which case there is no warning at all. But today's technology enables astronomers to get a fix on any asteroid capable of causing a global "extinction event" -- six miles in diameter or bigger.
Asteroid 2004 MN4 is a "regional" hazard -- big enough to flatten Texas or a couple of European countries with an impact equivalent to 10,000 megatons of dynamite -- more than all the nuclear weapons in the world. Even though it will be a near miss in 2029, that will not be the last word.
"You don't know what the gravitational effect of the Earth will be," said Brian G. Marsden, who oversees the hunt for near-Earth objects as director of the Minor Planet Center at the Harvard-Smithsonian Center for Astrophysics.
"In 2029, the [close encounter with] Earth will increase the size of the orbit, and the object could get into a resonance with the Earth," he added. "You could get orbit matchups every five years or nine years, or something in between." In fact, 2004 MN4 could come close again in 2034, 2035, 2036, 2037, 2038 or later.
So, what can be done? The first thought, dramatically depicted in the 1998 movies "Deep Impact" and "Armageddon," is to nuke the intruder into small pieces so it will burn up in Earth's atmosphere. Many scientists say, however, that this is unacceptably sloppy -- instead of obliterating the target, the bomb could break the asteroid into large radioactive chunks capable of transforming huge stretches of Earth into wasteland. Or the explosion could deflect but not destroy the asteroid, putting it on a future collision course. A nuclear strategy would also forever require a stockpile of doomsday weapons.
"The cure's worse than the disease," said former Apollo astronaut Russell L. "Rusty" Schweickart. He is a board member of the B612 Foundation, a group of experts promoting a space mission by 2015 to send a "tugboat" spacecraft to a near-Earth object, dock with it and gently alter its speed enough to change its orbit -- to show that it can be done. (B612 is the name of the asteroid home of "The Little Prince," in the Antoine de Saint-Exupery story.) "You want to delay or speed up the asteroid a little," said Berlin-based Alan Harris, chairman of the European Space Agency's Near-Earth Object Mission Advisory Panel. "What kind of surface do you have: Is it rocky? Dusty? Rubbly? How much force can I apply? I don't want to break it up -- unless I really break it up."
B612 has a design but little money, while ESA has spent only a nominal amount to study the feasibility of a reconnaissance mission to an asteroid. NASA, at $4 million a year, is currently the big spender for near-Earth object research. With this, NASA maintains a database at JPL to plot and record orbits for all known near-Earth objects, and contributes money to the Minor Planet Center and to sky surveys underway at telescopes in Arizona, California, Hawaii, New Mexico and Australia.
The money was authorized after a push from Congress led by Rep. Dana Rohrabacher (R-Calif.), a conservative, and former House Science Committee chairman George E. Brown Jr. (D-Calif.), known as one of Congress's most liberal members before his death in 1999. "I have a vision of something terrible happening, and I feel compelled to see that it doesn't happen," Rohrabacher said.
NASA's task -- which Congress imposed in 1998 -- is to find 90 percent of the estimated 1,100 near-Earth objects that are one kilometer (0.6 miles) or greater in diameter by 2008. As of mid-March, JPL's database included 762 of these.
On March 1, Rohrabacher introduced the George E. Brown Jr. Near-Earth Object Survey Act, mandating $40 million for a two-year start-up to survey every object 100 meters (328 feet) across or larger, of which there may be 300,000. To date, Marsden has registered 3,265 near-Earth objects of all sizes.
Tholen, of the University of Hawaii, is a frequent contributor in the search for threatening objects. He specializes in "Atens," a subspecies that orbit mostly between the Earth and the sun and are difficult to see in the glare of the sun. To spot Atens, astronomers must work at dawn or dusk.
Tholen's team, on a field trip to the University of Arizona's Steward Observatory, had booked an hour on the evenings of June 19, 20, 23 and 24, 2004. They found a new Aten on the first evening and saw it again on the second evening. It was about 106 million miles away.
The team recorded the sightings and sent them electronically to Marsden, who published the object's position, which he named 2004 MN4 in accordance with a complicated coding system based on the date of discovery.
Tholen waited for another opportunity, but rain clouds cloaked the sky. When the storm passed, the moon was squatting right where the team wanted to look. For the next six months, nobody looked for it.
Then, on Dec. 18, astronomer Gordon Garradd, working at the Siding Springs telescope in Coonabarabran, Australia, 240 miles northwest of Sydney, spotted what he thought was a new near-Earth object, "brightly lit and traveling fast," he recalled. He took four images in his first set, then followed up with two more sets.
Marsden's team put Garradd's data on the center's Web page, a signal for astronomers to get more fixes. On Dec. 20, JPL produced its solution. Chance of impact was one in 2,500 -- nothing to get excited about. "Usually the probability goes down with more observations," Marsden said.
Not this time. On Dec. 23, the risk rose to one in 270, and rose steadily over Christmas and beyond. "We'd never had anything this big come this close, and we'd never predicted anything like it," Marsden said. "It was quite fantastic." The asteroid was 9 million miles away -- about as close as it would get this trip.
By Dec. 26, the impact probability had risen to one chance in 38. What the plotters needed was a "precovery," an overlooked observation from before Tholen's initial June fixes to yield a more precise orbital solution. In Tucson, astronomers at the Spacewatch Project, at the University of Arizona's Lunar and Planetary Laboratory, started searching their archive. Spacewatch has been surveying the solar system for 20 years, and precovery is a specialty.
"We store [our images] on DVDs," Spacewatch leader Robert S. McMillan said. "If there's something that wasn't automatically sorted by our software, we can usually find it -- if we were looking in the right place at the right time."
They were. On Dec. 27, Spacewatch astronomers Jeffrey Larsen and Anne Descour found 2004 MN4 in a series of images taken March 15, more than three months before Tholen's sighting. They passed the word to JPL, which issued a news bulletin: "An Earth impact on 13 April 2029 can now be ruled out." Since then, astronomers have continued to observe 2004 MN4 whenever possible, but most of the time it is obscured.
"It would be awfully nice to have information so we don't get surprised," said Schweickart, who advocates flying a small interceptor mission to plant a transponder on 2004 MN4 that would constantly radio its location, tagging it like a grizzly bear. "Our favorite little asteroid might provide enough reality here to provoke people. Maybe we should get serious."
WHEN TO ACT ON IMPACT PREDICTIONS
by Rusty Schweickart
It would be helpful would be if people understood the typical time history of the impact probability of any NEO that has a chance of an impact. As additional telescopic observations are made over time, the uncertainty in the asteroid's position gets smaller and smaller. So long as the Earth remains within the "uncertainty ellipse" the probability of impact will generally increase. At some point, as the error ellipse shrinks, it no longer includes the Earth within it. At that moment the probability of impact drops to zero regardless of how high it was. In most instances this pattern is very short lived and the asteroid drops from a TS=0 (TS is Torino Scale) to a zero probability of impact (TS=0 does not mean zero probability.. it means very low probability.. another confusing factor). In a few instances an asteroid has reached TS=1 and then suddenly dropped to zero. In the instance of 2004MN4 it went from TS=2 to TS=4 in a period of 4 days, and then suddenly dropped to zero. Once in a thousand years (approximately) the impact probability of a NEO will continue to rise until it hits us. That's the one we want to deflect 10 or more years before that happens.
The NORMAL pattern is for the probability of impact to increase as more observations are made and then for it to drop suddenly to zero. This is NOT due to errors made by astronomers, as some have said. It is due to additional observations shrinking the uncertainty over time with more observational data.
I believe that the NEO community needs to begin thinking seriously about why we're doing this and act accordingly. We're not discovering these objects simply to know (finally) when one will hit before it hits. We're looking for them in order to know one will hit in time to deflect it before it hits. This means that we need to have adequately accurate information adequately in advance of a potential impact to permit a deflection mission.
Adequately accurate translates (approximately) to having an error ellipse about the size of the target to be missed. Generally this means an error ellipse (3 sigma) about the size of the Earth (accounting for gravitational focusing). However, if there is a gravitational encounter with the Earth or another body within 10 years or so of the putative impact, then the time to deflect shifts to prior to that early encounter and the "object" to be missed is the resonance keyhole at the pre-impact encounter. If this pre-impact encounter is significant the size of the keyhole can be quite small. For example, in the case of 2004MN4, as the asteroid passes close by the Earth in April 2029, the keyhole associated with the 7/6 resonance (potential Earth impact in 2036) is approximately 600 meters across or about 5 orders of magnitude less than the size of the Earth itself.
In essence then "adequate accuracy" in this case means an error ellipse about 600 meters long. And, since a deflection would have to take place prior to 2029, "adequately in advance" would back up to about 2014 for a commitment to launch a deflection mission. (decision, planning, manufacturing, launch, rendezvous, deflection together requiring approximately 15 years)
The good news is that the delta-V (change in velicity) required to miss the keyhole is very small, approximately 10^-6 meters/sec. The bad news is that in order to obtain an error ellipse of 600 meters by 2014, we will need to place a radio transponder on the asteroid before that date. Optical measurements alone will still yield an error ellipse at least 2 orders of magnitude too large to make a rational deflection decision by that date.
Needless to say, the probability that such a deflection mission will be needed is very small, about one in 10,000 based on current knowledge. However, by 2014 (or so), if the error ellipse is 60 km long because we've relied only on optical (and even radar) data, the most we'll be able to say is that the probability of impact is something like 1 in 140 (or, more likely 139 chances of a miss out of 140), even if it is headed for an impact! With a radio transponder on it we will be able to say either "it will miss.. absolutely" (the most probable result by far) or... the chance of an impact is better than 50%. In the former instance we don't launch a deflection mission... in the latter we will.