The case of 1997 XF11


Guy Hurst has asked me to write some comments on the case of 1997 XF11, given the fact that there is still considerable controversy, in part due to the publication on IAUC 6879, of a partial rectification of the possible impact prediction for 2028. Today, June 10th, I have received a relayed e-mail from Brian Marsden (the original appeared in an Internet newsletter on CCNet), replying to further comments from his most persistent critic. This indicates that the regrettable situation of strong, and (frequently) unreasonable criticism of Brian is still continuing in some circles and deserves a public reply.

 In the last few years the danger of Near-Earth Objects (NEOs), asteroids and comets with the potential for impact, has been recognised. After years of being only respectable as a science-fiction plot, multiple studies have shown that the potential for impact is greater than had ever been imagined. One study even shows that the danger of dying in an asteroidal impact is actually greater than that of dying in a plane crash. At least one object of catastrophic potential (Hermes = 1937 UB) is completely lost, having been seen for just a 4-day interval in 1937. Even very small NEOs such as the current holder of the distance record, 1994 XM1, an 11-m diameter rock (for an albedo of 15%), have a kinetic energy around 1Mt at parabolic velocity (42 km/s), although the normal velocity of impact would have been more like 15-20 km/s; this reduces the impact energy by a factor of around 4-8, but still leaves a pretty big bang should the asteroid hit.

Even such small objects are not to be disdained should they fall near inhabited territory.

As of today, 517 close-approach asteroids are known, of which no less than 123 are classed as "potentially hazardous" (H < 22.0, equivalent to a diameter of approximately 200-m, and a minimum distance of approach of < 0.05AU). This is, without doubt, only a small fraction of the total population of NEOs. The potential for disagreeable surprises is ever-present. The very largest of the potentially hazardous objects, 1990 MU (almost 7km diameter) was only discovered at closest approach. In 1989, the object 1989 FC (4581 Asclepius) was discovered 8 days AFTER what was then the closest-ever approach by a large object (~350-m diameter).

As the Director of the Minor Planet Center, Brian Marsden knows that, any morning, he might walk into his office and discover that one of the previous night's batch of new asteroids might represent a serious danger to the Earth. This is a serious responsibility and one which he discharges with great dignity.

The specific case of 1997 XF11 is a particularly interesting one. It was discovered well away from closest approach, in fact, it was approaching aphelion. The Spacewatch Telescope which discovered it took just three frames, each separated by half an hour, confirming it as an asteroid, but a slow-mover. El Ni˝no prevented further Spacewatch observations. Had it not been for amateur observations this asteroid would have been rapidly lost. As it was, the observations by A.Sugie and T.Kojima (the latter with a 25-cm reflector!) recovered it twice when there was a real danger of the asteroid being lost had much more time elapsed between observations.

Even 2 weeks after discovery it was already evident that this was an object with the potential for a really close approach to the Earth. Unfortunately, despite being flagged as an interesting object in a MPC electronic circular, the observations still only trickled in to the extent that the addition of a single pair of observations could radically change the extrapolation of the ephemeris. This is what finally happened: two new observations were made at Mc Donald Observatory (Texas) in March and suddenly 1997 XF11 stopped being of only academic interest. With these observations a very close approach on October 26th 2028 became a distinct possibility.

When you see that an asteroid of nearly 2km diameter is scheduled to make an approach to less than 50 000km in the foreseeable future, you take it very seriously! Even if the real danger of an impact is very small indeed, it would be a fool who would ignore the possibility of a 20km crater appearing somewhere in Western Europe. An order of magnitude estimate of the explosive force would be 500 000Mt.

Brian reacted in what was, to me, a very measured way, with a carefully phrased IAUC, noting the near certainty of an approach to a distance less than that of the Moon and requesting further observations. Nothing too controversial there.

There was a rapid response in that Ken Lawrence and Eleanor Helin found images from the 1990 close approach within hours of the IAUC being announced, closely followed by images from Ted Bowell and Carolyn Shoemaker. Less than 30 hours after the initial circular, a new orbit was ready. This showed that the asteroid was slightly ahead of the predicted track and would actually pass closest approach in 2028 some 11 hours earlier than expected (at around 30km/s, this is a very large difference in physical position), but not unexpected for a 30-year extrapolation.

This converts into an increase from a distinctly uncomfortable <50 000km miss-distance, to a less than terrifying 960 000km.

Here, bad luck came into play. The initial calculations suggested it was virtually certain that the asteroid would approach closer than the Moon. In fact, the probability of an approach closer than the Moon varied from 90% in one estimate, to around 40% in another - still pretty good odds.

Had 1997 XF11 come closer than the Moon, I don't think that there would have been very much fuss. After all, "closer than the Moon" sounds pretty close to most people. In fact though, 1997 XF11 will pass around 2.5 times the distance of the Moon which looks much less impressive, although still a very small distance for a large object.

It turns out that we had been particularly unlucky statistically. Given the size of the errors in the calculations, even given the pessimistic error estimate, there was only a 1% chance that the asteroid would pass that far away. Statistics are particularly slippery and they can make a complete fool of you. In this case they behaved very badly... that's not Brian's fault!

New calculations have shown that the asteroid's node and the Earth would have coincided some time around 2040 and that an impact was a real possibility for some three years when the distance between the asteroid's orbit and the Earth is 0. The probability of an impact was small - perhaps 1 in 100 000 - but not zero. Before the 1990 pre-discovery observations came to light, there was a possibility that the unknown circumstances of the 2028 encounter could actually have resulted in an impact a decade or two later. Specifically, Brian showed that an impact in 2037 was a definite possibility (it would have required particularly bad luck, with the approach in 2028 changing the orbit by just the right amount, but it was not impossible).

What is important though is that we now know, with absolute certainty, that 1997 XF11 will not hit the Earth. It cannot because it will never be close to the Earth in its orbit during the three critical years. Beyond 2040 its orbit moves slowly away from the Earth's by about 4000Km per year and the danger of a close approach in the future, which might change the orbit back to a more dangerous one, will get steadily smaller. In fact, we already know that this possibility is practically zero.

Web navigators will have no trouble finding other sites with other scientific (and non-scientific) perspectives, including page with a modestly entitled text "JPL scientists save the Earth".

There has been some dreadful over-reaction to this incident. I wish people would look at the other side of the coin: suppose Brian had with-held the news, waiting for confirmation (or, worse, having contacted some groups in the hope of follow-up) and it had been leaked? The MPC would then have been accused, quite rightly, of a cover-up and hiding information of vital interest to the public.

I'm quite sure that the scandal would have been far worse this way.

What the incident has revealed is that proceedures for dealing with a potential impact hazzard need to be thought about. In this sense it has been a valuable dry run and, in the end, nobody has been hurt. The next time that Brian walks into his office, checks his computer and sees an asteroid like this flagged -- it could be tomorrow -- we may not be so lucky as to see the errors go on the safe side. You buy insurance in case of accidents that you are sure will never happen, but you do it anyway. When that happens, we may be grateful that we have have had some experience in how to react.