HOW THE ASTEROID STORY HIT: AN ASTRONOMER REVEALS HOW A DISCOVERY SPUN OUT OF CONTROL

Brian G. Marsden


The following item may be copied and extracted, provided that credit is given, both to the author and to the Boston Globe newspaper.

This is an extended version of an article that appears on pages E1 and E2 in the FOCUS section of the Boston Sunday Globe on 1998 March 29. The piece was commissioned by the Globe.


There was a dramatic worldwide response to my announcement earlier this month that an asteroid known as 1997 XF11 would pass perhaps uncomfortably close to the earth on a date 30 years hence, with an exceedingly remote chance that it could strike, subject to further study of its path.

Less than 30 hours later, it was still true that the asteroid would pass close to us, though not excessively so, on that date. But the worry was off, and initial concerns had been replaced by jokes about astronomers making mistakes, as in, "No wonder math education in this country has its problems".

Along with such comments came accusations of "Chicken Little" behavior by "irresponsible" Harvard astronomers neglecting to check their "cockamamie calculations" with colleagues first (columnist William Safire), and complaints that some NASA scientists were muscling in on a matter that really didn't concern them, a case of "professional discourtesy" (Malcolm Browne, in The New York Times).

Of course, as many other scientists and journalists pointed out, the now greater distance from the earth of this asteroid's projected path did not alter the fact that there will come a next time, when a sizable comet or asteroid strikes with disastrous consequences, but it provided some reassurance that, for the moment at least, we can sleep easily.

So what went wrong with asteroid XF11? In my view from the eye of the storm, the problem was an overeager use of e-mail and public relations by a supervisory scientist at NASA's Jet Propulsion Laboratory (JPL). I also believe that for us not to make the announcement as we did would have led to condemnation that science was being stripped of its essential openness.

Asteroid passes are more frequent and closer than most people think. Just 30 years ago, on June 14, 1968, the asteroid Icarus, roughly the same size as 1997 XF11, passed 4 million miles from the earth. That seems like a tremendously safe distance (and it is), but much of the public and the press did not think so at the time. A younger and slimmer yours truly was interviewed about the Icarus encounter a day or two beforehand by a younger and slimmer Chet Curtis of Boston's Channel 5, a cult gathered on a peak in Colorado to escape from the anticipated slide of California into the Pacific Ocean, and I spent several minutes on the phone calming a woman whose mother was terrified about the impending end of the world.

It was all unnecessary. Icarus had been expertly tracked by astronomers from its discovery in 1949 frequently through 1967, and the distance of 4 million miles was precise. But because it was only the first time an asteroid had been predicted to come that close, many people went wild.

Icarus is the fifth entry on the list of "potentially hazardous asteroids", or PHAs, that have the possibility, over the next several centuries, of coming within 5 million miles of the earth and that are more than perhaps a tenth of a mile across. Icarus next comes to a distance of 5 million miles in 2015. And it will continue to induce scares, as it did in fact just a few weeks ago, when there appeared in many countries a claim in the press that it will definitely hit the earth just eight years from now! This was complete nonsense, for the miss distance in 2006 will be 30 million miles: even the planet Venus, almost as large as the earth, comes closer than that, and it does so every 19 months.

More of a worry in 1968 was the fact that there were perhaps half a dozen other known PHAs (though we didn't call them that then) that had been lost since within weeks of their discoveries. The first PHA, known as Apollo, had been discovered way back in 1932. It was a bit of long shot, but in 1973 I worked with two colleagues using the 61-inch telescope in Harvard, Mass., in a successful hunt for Apollo. As a result of this rescue, we now know where Apollo will be for centuries to come--and we know that the earth is safe from it.

On the other hand, Hermes, observed for only four days in 1937 (with the first known photograph also having been obtained at the Harvard station), is still very much lost. It could be anywhere around its orbit. Some October day (or, perhaps, an April day) in the years to come, Hermes could indeed be a threat to us. With luck, the programs that search for PHAs and other asteroids that come moderately near the earth will accidentally record Hermes well before that day. With appropriate measurements, we shall be able recognize that it really is Hermes. The rediscovery of Hermes is one of those prizes that quite a few astronomers wish for at the back of their minds. Furthermore, for each Hermes, there are 20 or more unknown possible threats of comparable size that we don't already know.

The first program set up specifically to search for "NEOs" (i.e., "Near Earth Objects", a more general term than PHAs) was initiated in 1973 by Caltech's Eleanor Helin in collaboration with the late Gene Shoemaker and using wide-field telescopes at the Palomar Observatory in California. The Helin and the Shoemaker programs went their separate ways in the early 1980s, when Gene began to collaborate with his wife, Carolyn. The other early entrant into the NEO search business was Tom Gehrels, who was photographing asteroids already in the 1950s and who in the 1980s set up the University of Arizona's "Spacewatch" project, the first program to make the asteroid detections efficiently by electronic means, rather than from scanning conventional photographs by eye. In 1995, Helin's program also "went electronic", and her team in California is now engaged in the analysis of images relayed to them six nights a month from a telescope operated by the Air Force in Hawaii.

The programs of the veteran researchers Gehrels and Helin have recently been joined by an excellent third search program, which also uses an Air Force telescope (in New Mexico) and is operated by a group at MIT's Lincoln Laboratory here in Lexington. My offices at the Harvard-Smithsonian Center for Astrophysics in Cambridge have worked, and continue to work, closely with all these search programs, as well as with other astronomers (professional and amateur) around the world who make follow-up observations of interesting objects that are found. We correlate and otherwise organize the data from the various groups and observers, compute orbits as appropriate and attend to the communication of the results under the auspices of the International Astronomical Union (IAU), the international association of professional astronomers.

It was an assistant in the Shoemaker program who, in 1989, discovered the asteroid Asclepius, which at a distance of only 430 thousand miles came closer than any known asteroid since the lost Hermes. There was the particular worry that Asclepius had actually made its pass a week before it was discovered, and that was because it came at us from the direction of the sun.

Beginning in 1991, there were several occasions when Spacewatch discovered tiny asteroids, only tens of feet across, at distances of 100 thousand miles and even less. Objects this small really are of little concern, because even if they hit, they are likely to burn up completely and harmlessly in the earth's atmosphere.

Then, in 1996, came the PHA with the closest known pass, 280 thousand miles, for an object perhaps one to two tenths of a mile across. It was discovered photographically by two enterprising students in Arizona, Tim Spahr and Carl Hergenrother, only a few days before its closest approach.

The subject of the latest concern, 1997 XF11, was discovered in December by Jim Scotti of Spacewatch and is the largest of the 11 PHAs added last year.

(A complete list of the 112 known PHAs appears in our Web site at http://cfa-www.harvard.edu/iau/lists/Dangerous.html. The site also has a list of all the predicted approaches of asteroids and comets within about 19 million miles of the earth during the next third of a century: see http://cfa-www.harvard.edu/iau/lists/CloseApp.html.)

On March 11, when my colleague Gareth Williams and I were making our regular monthly update of our Web files on March 11, we were startled to find 1997 XF11 singled out for a passage less than 30,000 miles from the center of the earth on Oct. 26, 2028. In more than 40 years of computing orbits, I had never seen anything like that before. Of course, we realized that the actual distance was uncertain and that the object could easily pass at or somewhat beyond the distance of the moon.

Nevertheless, to have the earth right at the center of the uncertainty range was startling. What really surprised us, however, was that nobody else --the folks at JPL, for example, or even a knowledgeable amateur astronomer with a PC--seemed to have done this same calculation yet, since all the information necessary had been available on the Web for five days.

In fact, nobody seemed to have taken much interest in the object at all, even though it was obvious already in December that XF11 was an asteroid that could come particularly close to the earth and that it was in the largest 20 percent of all PHAs. This reinforces the impression that we must hunt for PHAs because if we don't do so, we shall someday go the way of the dinosaurs. Yet some colleagues appear to assume that the PHAs we have already detected cannot possibly hurt us, and we don't need to study them any more.

With our attention drawn to XF11, the obvious next course to us was to encourage further observations, either in the future--with the object now getting rather faint and due to move into daylight in a couple of months--or in the past, with images perhaps to be found in photographic archives.

I therefore put out an "IAU Circular" stating the facts and telling observers where to look for XF11 between now and early July.

As we sometimes do with a particularly interesting development, we also put a popular version of it into our Web page. The popular rendition mentioned the possibility of finding images of XF11 on photographs taken in 1990 and earlier. By "clicking" on a particular year, one could see where in the sky the object was. If earlier images were found, they could provide more precise determination of the asteroid's course than we had available.

Eleanor Helin (who, as it happens, is also now located at JPL) and assistant Ken Lawrence checked the Web site, realized they had suitable Palomar telescope films in 1990, inspected them, quickly found the images of XF11 from March 22 and 23 of that year, measured them and reported the measurements to us. Great.

Williams worked the 1990 data into the orbit solution, revised the 2028 miss distance to 600,000 miles, and I prepared another IAU Circular with the new results. (This will still be the record predicted approach for an asteroid or comet during upcoming decades. Although we know of an asteroid that will come just a little closer in 2086, it seems to be smaller than 1997 XF11.)

So where did things go awry? The media, of course, make an easy scapegoat. But responsible science writers, such as Browne and David Chandler of the Globe, wrote quite commendable articles. These writers receive the IAU Circulars on a regular basis anyway. What's the alternative way of getting news out? Should we prohibit anyone with press credentials from even looking at our Circulars? Some of my colleagues are looking into ways of preventing information on wayward PHAs from getting to the press and the public. To me, that smacks of censorship.

No, I think the problem lay with some of those very colleagues, eager to make pronouncements. And the reason this is a problem now is the easy availability of e-mail. E-mail is a wonderful tool--often better than the telephone--for getting quick answers to urgent questions. But many e-mail devotees use it not just to send a message to someone but to copy that message simultaneously to a couple of dozen other people.

This prompts responses and remarks between other pairs of participants, with the messages again copied to all the members of the gang, plus, perhaps, a few more friends. Those friends also have their regular e-mail correspondents, of course, so both information and misinformation can proliferate rapidly, like a pyramid scheme.

Paul Chodas is a fine scientist at JPL, and on receiving the message about XF11, he performed some experiments on the uncertainty of the miss distance to see whether there was possibility the object might hit the earth in 2028. Chodas's computations were not, in fact, significantly different from mine. The difference was that an official there e-mailed his results to all and sundry, thereby also bringing them to the attention of the press.

Chodas and I agreed, as did two other scientists making similar calculations, that it would be very difficult to bring XF11 within 15,000 to 20,000 miles of the earth's center in 2028. But that is closer than the geosynchronous artificial satellites. Furthermore, the other scientists working on the problem agreed with me that it was rather rash to say that there was absolutely no chance of a strike, for it was impossible to claim there could not be other unconsidered effects (such as an abnormal distribution of errors, the ignored gravitational effect of a close approach to another asteroid, the possible reaction of the object to the explosive vaporization of ice on the surface).

A rip-roaring e-mail barrage on such small points quickly developed. As far as I was concerned, the argument was irrelevant. Matters would be solved once and for all if someone would do what I suggested and look for images of XF11 on past photographs. In any case, as Chandler later described in the Globe, there was an error in Chodas's computations--and Chodas was man enough to admit it.

And, of course, as soon as the 1990 observations became available, it was obvious that my group and the JPL group would obtain the same result with the 600,000-mile miss: revealing absolutely no danger. Again, what I thought inappropriate was the release of the new data by an official to colleagues and the press, with no mention that we and JPL were in agreement, given that the Circular with Williams's calculation (and a revision to our Web page) had been issued 90 minutes earlier. Contrary to reports, there was no "mistake" in either set of computations made by our group in Cambridge. What there was were new data, properly reported.

A few of my colleagues have questioned whether it was appropriate to issue an IAU Circular with a call for observations of XF11, when I could instead perhaps have contacted Helin and others directly with regard to checking past photographs. My response to this is that I did not know in advance whether she or others (which others?) would have appropriate old films. If there were no old films at all, it was obviously important to make future observations, and time for these was running short.

If there were no sense of urgency, there would be no incentive to search archives immediately, and they might not have been searched until future observations became impossible. In any case, it is precisely for the timely acquisition of follow-up observations like this that the IAU Circulars exist. If one asks observers individually, that is much more inefficient, and there are bound to be some who do not get asked--and who may as result feel a bit miffed. As it happened, we had a response from an astronomer in Germany concerning a possible image--not confirmed, it seems--of XF11 from way back in 1957. Would I have thought to solicit that astronomer directly?

The only other known positive past detections of XF11 to date are from the Shoemaker team, also made in 1990. The fact that there were such detections (but not the measurements themselves) was relayed to our offices some six hours after Helin informed us that her team had detections, and in the mean time she had provided the necessary measurements. Science may strive to be objective, but scientists are as competitive as anybody else. We made the information about where to look in 1990 available in the Web. That put all the film searchers on an equal footing. Nobody had an unfair advantage over anybody else.

At a meeting organized by NASA in Houston on March 17 some consideration was given to changing the procedures for announcing possible "end of the world" scenarios in the future. There was talk of establishing a special committee to "verify the orbit computations", as well as to alert those who might be able to make further observations and search for images in past archives. This is one of those proposals that may sound good, but as I have just indicated, there would clearly be problems as regards crediting priority for observations and archival searches--indeed, also, as regards crediting priority for ideas. "Science by consensus" is really not a palatable way to go, and decisions by committees are the food for cover-ups. In any case, the XF11 orbit computations announced were the correct conclusion from the data available at the time, and there was independent verification of this by both Williams and myself. Furthermore, it is surely naive to believe one can stop everyone everywhere from talking to his or her favorite reporter.

Certainly, I appreciate that there was a problem that the leaderships of both the IAU and NASA were embarrassed by this episode, in the sense that they had not in fact seen the initial Circular before they were overwhelmed by the press. I have indeed apologized to them for catching them unawares and hope that a way can be found to warn them and perhaps others of this kind of thing in the future.

I don't expect to be around in 2028 myself, but I can say that, weather and intervening disasters permitting, astronomically inclined residents of Massachusetts ought to be able to follow the asteroid on the morning and evening of the appointed day. A tiny starlike object will move from east to west (and in near full moonlight, if the night stays clear).

Few will remember it as 1997 XF11 then, because it will have received a real name, perhaps at the time of its 6-million-mile pass in 2002. But some will remember the brief fiasco in 1998 and perhaps reflect that this represented a genuine turning point. For one thing, perhaps 1998 would signal the beginning of a genuine breakthrough in significantly improving the quality of U.S. education in math and science all the way through grade 12. If so, the passage of XF11 in 2028 would be treated with the knowledge and the respect it deserves, with Heavensgate-type behavior and its associated ignorance a thing of the past. Furthermore, recalling the conclusions of the Spaceguard Survey Report of 1992 that a concerted survey both to discover and to track more than 90 percent of the most dangerous PHAs would take two to three decades, our successors watching the skies of 2028 might appreciate that XF11 provided the inspiration that led to funds to patrol the skies for PHAs much more thoroughly than before.


Brian G. Marsden is associate director for planetary sciences at the Harvard-Smithsonian Center for Astrophysics in Cambridge and director of both the Central Bureau for Astronomical Telegrams and the Minor Planet Center of the International Astronomical Union


Back to index of Press Information Sheets.

Index to the CBAT/MPC/ICQ pages.

MPC CBAT ICQ