On August 24, 2003, the LINEAR search system in the Spaceguard Survey discovered near-Earth asteroid 2003 QQ47. As is the case with many newly discovered NEAs, the initial orbit was highly uncertain and included several low-probability cases of possible future impacts. The orbit information was posted on the Internet by the JPL Sentry and Pisa NEODys systems. At one point, with only 6 days of observations reported, the formal odds of an impact in 2014 briefly rose slightly above one-in-a-million, and then went virtually to zero as more data were reported. This is standard operating procedure for dealing with newly discovered NEAs.
Unfortunately, the UK NEO Information Center decided that this asteroid deserved special attention, and on September 2 they issued a press release calling attention to the danger of collision in 2014. Since the NEO Information Center is supported by the UK government, this quasi-official prediction was widely reported in the British press. While most stories correctly noted the very small odds of hitting, they still treated this as a serious warning of a threat to Earth. The story was also reported in Europe, the USA, and Australia, but more moderately than in the UK.
The result is another round of criticisms of astronomers, triggered by the NEO Information Center release (which they withdrew on September 3). This is not the first time, of course. For your information, two background discussions follow: A story by Robert Britt from Space.com that deals with the media reactions to 2003 QQ47, and a section from a scientific paper in press that describes five previous cases where the supposed danger of impact was widely reported in the press.
ASTEROID DOOMSDAY 'RISK' EVAPORATES AFTER MEDIA FANS FLAMES
Space.com, 3 September 2003 by Robert Roy Britt Senior Science Writer posted: 09:45 am ET 03 September 2003
A newly discovered asteroid that generated doomsday headlines around the world yesterday morning was, by the end of the day, reduced to innocuous status as additional observations showed it would not hit Earth.
Meanwhile, a whirlwind of media hype has astronomers and asteroid analysts arguing among themselves -- again -- about how they should disseminate information to the public. By one expert account, it was business as usual in the Near Earth Object (NEO) community, a loose-knit group of global researchers who find, catalogue, analyze and frequently spout off about asteroids that might one day slam into our planet.
Asteroid 2003 QQ47 was discovered Aug. 24 by the Lincoln Near Earth Asteroid Research Program (LINEAR). Based on limited data collected during just a few days in late August, astronomers at first could not rule out the possibility that the giant rock would hit Earth. They gave it 1-in-909,000 odds of impact in 2014 and catalogued it as a 1 on the Torino hazard scale, a designation that merits "careful monitoring."
Its size -- three-quarters of a mile wide (1.2 kilometers) -- explains some of the attention 2003 QQ47 received. Were a rock that big to hit Earth, the climatic consequences would be global and it would cause, at the least, widespread regional devastation.
Most experts do not believe the mainstream press should waste time reporting on such an object. Several other newfound asteroids receiving similar designation in recent years have fallen off the list within days, as more observations allowed for refined orbital projections.
Nonetheless, a press release issued early Tuesday by the British government's Near Earth Object Information Center fueled widespread media coverage, including a wire story by Reuters that many asteroid experts saw as inflammatory. Headlines were over-the-top, most researchers felt. They included "Armageddon set for March 21, 2014" and "Earth is Doomed."
By late yesterday, however, more observations allowed astronomers to conclude there was no chance for impact in 2014.
The incident was just one in a long series miscues involving astronomers, their public relations efforts, and a media eager to report potential doom.
"It would appear that all the lessons learned from five years of our PR blunders, media gaffes and errors of judgement have been forgotten," said Benny Peiser, a social anthropologist and asteroid analyst at Liverpool John Moores University in the UK.
A handful of similar scares -- about one per year -- have evaporated in similar fashion as professional astronomers go about their business of finding and tracking potentially dangerous asteroids.
The NEO information Center, whose press release ignited the latest fiery press coverage, issued a follow-up statement early this morning. The NEO Information Center aims to keep the public and media informed of these kinds of issues, as they unfold rather than after the fact," the statement said. "This approach ensures we can promote understanding of the process of asteroid detection, tracking and risk assessment."
Kevin Yates, project manager for the center, had said in the original press release that additional observations would likely eliminate reveal a reduced risk. Today, Yates said, "Openly sharing this sort of information, in a non-sensationalist way, should help to dispel the popular myth that governments and astronomers would keep the discovery of a dangerous asteroid secret. I hope the coverage of this story will give the general public more of a feel for how the assessment of risk evolves over time as more observations are made."
The NEO Information Center's statement today concluded with a bizarre note of praise for the media that sounded defensive to others in the NEO community. "The NEO Information Center would like to thank the media for what, on the whole, has been responsible coverage of this story. Almost all of the press and broadcast coverage has included reference to our original statements that the probability of impact was very low at just 1-in-909 000, and that the Torino rating was likely to drop following further observations."
"Undermining our integrity"
Peiser did not share the center's rosy view for how the whole thing unfolded. He runs an electronic newsletter called CCNet, a forum for discussing the research and risks associated with NEOs, as well as the impact of media coverage on the public view of asteroid research and the credibility of the researchers.
"I'm afraid that any attempt to justify an ill-timed and unnecessary media campaign doesn't bode well for the NEO community's efforts to avoid false asteroid alarms that only risk undermining our integrity," Peiser wrote in the latest edition of CCNet today. Peiser leveled this accusation at the center: "Crying wolf becomes official policy."
The first and most notorious false asteroid alarm dates back to 1998. Then an astronomer went public with data showing that asteroid 1997 XF11 had a chance of hitting Earth in the year 2028. Once the asteroid was rendered harmless by more observations, a debate began as to if, when and how to release preliminary asteroid data to the media and the public.
Though new agencies, institutions and programs have since been set up to better manage the situation, little has changed. A similar scare developed last summer, when British media hyped the potential danger of 2002 NT7. In that situation, astronomers were candid and vocal in their criticism of the British press.
Like the Return of Elvis
One thing has changed of late: There is an increasing sense of sarcasm in the media with each new asteroid scare. Some reporters and editors are getting wise to the long odds -- or perhaps tired of having to report on them -- and doing more than just sensationalizing the data.
One story yesterday made light of the initial chances of 2003 QQ47 hitting Earth. Sky News, a British publisher, said a bookmaker was taking bets on the prospect. A spokesman for William Hill bookmakers likened the 1-in-909,000 odds of doom to the chance that a manned expedition to Mars would arrive and discover the Loch Ness Monster there, or the equally probable scenario that Elvis Presley would reappear and marry Madonna. We now know that the latter two scenarios are far more likely than the world ending in 2014 due to an impact by asteroid 2003 QQ47.
Copyright 2003, Space.com
Excerpt from a chapter in a forthcoming book on asteroid impact mitigation
IMPACTS AND THE PUBLIC: COMMUNICATING THE NATURE OF THE IMPACT HAZARD
Clark R. Chapman
Richard P. Binzel
LEARNING FROM EXPERIENCE: FIVE NEAS THAT MADE THE EVENING NEWS
Much of the press coverage of the NEO impact hazard has been excellent. Some journalists have worked hard to master the intricacies of this low-probability risk and to translate the information into a form that is accessible to a wider public. However, reporting of individual cases has revealed problems in the communications between astronomers and the non-specialist media.
In this section we briefly review five cases in the past decade where newly discovered NEAs were widely reported (or misreported) in the press as posing a risk of impact with the Earth. Each case involves communications failures of a different kind. These examples illustrate the frustrating breadth of ways that things can go wrong.
What constitutes an asteroid being newsworthy, and hence capable of spawning a scare, has changed gradually over the past decade. The selection criteria are such things as the probability and proximity in time of the potential impact. As new records have been set for each parameter, the media apply new standards to rumors and reports of potentially hazardous NEAs. Perhaps astronomers as well learn how to communicate better. Despite this, with the anticipated increasing sky coverage and deeper surveys, it seems inevitable that impact alarms will continue to make frequent appearances in the media.
(1) 1997 XF11 The first modern impact scare was associated with asteroid 1997 XF11 (now asteroid # 35396), an approximately 1-km diameter NEA discovered on December 6, 1997. By early in 1998, preliminary orbital calculations showed a possible close pass by Earth in October 2028. Brian Marsden, Director of the Minor Planet Center at the Harvard-Smithsonian Center for Astrophysics, was following the orbit in early March when the most recent observations suggested that XF11 might pass well inside the orbit of the Moon. On March 11 he made this information available and quickly distributed a Press Information Sheet in which he stated that "the chance of an actual collision is small, but it is not entirely out of the question. From the calculated miss distance and its estimated error, Marsden and other astronomers suggested that the probability of an impact could be as high as one-in-a-thousand, and news media all over the world put the story on the front page.
Unlike Marsden, Don Yeomans and Paul Chodas at JPL had the software to estimate the actual odds of hitting, and within two hours of Marsden's announcement they had calculated that an impact was impossible -- although the position of the asteroid in is orbit had a substantial uncertainty, there was zero possibility of a collision. (The error ellipse was effectively a line that passed well off the Earth, rather than a fat ellipse). Marsden did not withdraw his comments, however, until prediscovery observations from 1990 were found a day later that improved the orbit and demonstrated than the asteroid would not only miss, it would not even come close to the Earth in 2028. Only then did the media run retraction stories, mostly along the lines that new data had corrected the original prediction.
The XF11 episode demonstrated the need for rapid calculation of impact odds as well as nominal asteroid orbits and miss distances. In addition to the JPL team, Andrea Milani at Pisa and Karri Muinonen in Helsinki quickly demonstrated this capability. The second lesson was that astronomers should check their results with colleagues before going public if they wished to avoid the embarrassment of a scientific debate in the public eye. Third was a more general concern that astronomers looked rather foolish when a prediction of a possible impact was corrected or withdrawn within less than a day. It was suggested that perhaps the IAU should vet such predictions, and in this case the IAU issued a statement that "contrary to preliminary reports, there is no danger of its colliding with the Earth in 2028. Like all Earth-crossing asteroids, XF11 may someday hit our planet, but this seems to be an event for the distant future, and at present we are more at risk from some unknown asteroid colliding with the Earth than from XF11 or any other object already discovered."
In summary, the problem with XF11 was premature announcement without calculating a formal impact probability or consulting with colleagues. The solution seemed to be better software and more consultation before making announcements.
(2) 1999 AN10 On January 13, 1999, the LINEAR search program discovered a mile-wide NEA designated 1999 AN10. Milani and his colleagues at Pisa showed that there was a less than one-in-a-million chance of AN10 passing through a keyhole (see below) during a 2027 close approach so that it would impact the Earth several decades later. In the year since XF11, great strides had been made in orbit prediction, including the concept of keyholes defining a very restricted set of orbits at one pass that could lead to a resonant return at a subsequent pass. Although the result was unremarkable from the perspective of the impact probability, it was of interest to many dynamicists, and Milani prepared a manuscript and asked colleagues for an informal review within two weeks. Receiving no complaints, he posted his paper with no fanfare on his website.
Benny Peiser of John Moores University, the moderator of the CCNet Digest Internet forum, discovered the unheralded paper on Milani's website and charged cover-up. The widely-reported media story was not about the impact risk, but about the idea that astronomers would conspire to withhold from the public information on a possible (even though exceedingly unlikely) impact for weeks. As a result, Milani challenged his colleagues and the IAU to come up with a better way to review and (if appropriate) to publicize such cases. Shortly thereafter the IAU Working Group on NEOs set up its technical review committee, committed to carrying out a review within 72 hours. The Torino scale was also adopted in June 1999 to aid in public understanding as described above.
In summary, the problem with AN10 was that the media could read scientific websites and extract information on low-probability future impacts. As a result, the scientists lost control of the discussion and were subject to potential criticism for being either too open or for suppressing information, according to the taste of the critic. The solution seemed to be to formalize the review and let the IAU, through its technical review, provide an international, professional context for any released information.
(3) 2000 SG344 The IAU review process was soon exercised for the very small object 2000 SG344, which was interesting primarily because of the possibility that it may be a spent rocket upper stage in an orbit very similar to that of the Earth. For a few days, however, it appeared that SG344 might have an impact probability as high as 1 in 500, triggering the IAU 72-hour technical review. The review was completed on a Friday afternoon, confirming the relatively high impact probability, and accordingly JPL issued a press statement and the IAU posted the results on their webpage. Media interest was immediate, as this was the highest impact probability ever predicted and confirmed.
Within a few hours, new observations were available that showed no impact was possible. Unfortunately, on the weekend no-one was correcting the statements of alarm posted by JPL and the IAU. Two mistakes thus compounded the problem: a statement was issued that was technically correct but did not anticipate the availability of new data, and no-one was ready to issue a formal correction over the weekend. Once again the astronomers looked foolish.
In summary, the problem with SG344 was a literal application of the IAU guidelines, which called for release of information as soon as the technical review was complete, without recognizing that the data upon which the orbital calculations had been based were improving daily. The solution was, as a minimum, to relax the strict IAU guidelines. It was also suggested by some that no information should be released until all possible data were collected, even if this meant a delay of weeks or possibly months, since in nearly all cases the refinement of the orbit would eliminate the possibility of an impact.
(4) 2002 MN Asteroid 2002 MN was a small asteroid that flew past the Earth at a distance of 0.0008 AU (one-third of the lunar distance) three days prior to its discovery. The media judged this to be a failure on the part of astronomers because there was no prior warning, whereas actually it was a success: the asteroid was found and catalogued. Orbit calculation showed that it did not pose any danger of future impact (for the 50-100 year window that was used to search for future close passages).
In summary, the problem with 2002 MN was that it was discovered, as often happens, after closest approach rather than before. The community had difficulty explaining that the Spaceguard Survey was not designed to detect NEAs on their final plunge to Earth, and that finding an object after closest approach is just as useful as finding it before, and is in fact nearly as likely, since many asteroids move into the night sky from the sunward side of the Earth.
(5) 2002 NT7 Asteroid 2002 NT7, with the relatively large diameter of 2 km, was discovered in July 2002. By this time the calculation of impact probabilities was fully automated, and on July 18, NT7 was posted on the risk page of both the Pisa NEODys and JPL Sentry systems, showing a possible but very unlikely (of order one in 100,000) impact just 17 years in the future. Because new data were coming in and NT7 remained a zero on the Torino Scale (although very close to Torino Scale = 1), it was decided not to call for a formal IAU technical review or to make any public statements, pending improvements in the orbit.
On July 24 this remote chance of impact became an international media story when the BBC picked the information up from the Internet and reported that the asteroid was "on a collision course with Earth". As expected, however, additional observations quickly eliminated the possibility of an impact. The all clear for any impact in 2019 was released on July 26, and by August 1 continuing orbital improvements also eliminated a lower-probability impact in 2060 -- a progression of events that reflects the normal working of the Spaceguard system. So why all the media fuss about NT7, especially in the UK press? This is unclear. Especially provocative was the BBC story that called NT7 "the most threatening object yet detected in space." As the asteroid itself receded from interest, the media story focused on the sensationalist reporting.
Science journalist Robert Britt concluded in a story in Space.com that "The whole affair, over an asteroid that is almost certainly harmless, illustrates the stylistic ocean that separates American and British media and scientists' tactics in dealing with them". The following quotes are from his report. Duncan Steel suggested that asteroid stories have become so common that in his country they either make headlines or they're not used at all. Unless a reporter "makes it sensational, the editor will nix it. Ditto (especially) for the printed media." Don Yeomans said that he was unprepared when "the media blitz struck." "Most of the six interviews I did with BBC reporters Tuesday night began with their assumption that there would be a collision," Yeomans said "One is then forced to back up and try to explain the real situation and the fact that there is not really a story here. They didn't wish to hear that." Yeomans later concluded that journalists and scientists both need to strengthen efforts to help the public understand how asteroid risks are determined. "There is plenty of blame to go around," he said.
In summary, NT7 became a media story in spite of the astronomers. The press obtained information directly from technical webpages and gave it a sensationalist spin. Astronomers then spent several days trying to correct the misimpressions. While there is no clear solution, most of us concluded that scientists should have responsible statements ready to provide to the press or post on websites as an antidote to this sort of outbreak. But no one suggested that we should try to keep impact predictions secret. That is a clear lesson from all of the encounters with the press from XF11 to NT7.