This month asteroids are on everyone's minds after the excitement about 1997 XF11. The story began on 11 March, when the Central Bureau for Astronomical Telegrams (CBAT) in Cambridge, Massachusetts, issued an IAU Circular (IAUC 6837) with an item concerning a newly-discovered object, dubbed 1997 XF11. Based observations spanning just 88 days, the orbital solution indicated the object would pass a mere 46,400 km from Earth on 26 October, 2028 AD. The IAUC was not exactly sensational in tone; the possibility of an actual collision with the Earth was not explicitly mentioned. However, Brian Marsden, the author, did say that "Error estimates suggest that passage within 0.002 AU [299,000 km, less than the distance of the Moon] is virtually certain, this figure being decidedly smaller than has been reliably predicted for generally fainter PHAs ['Potentially Hazardous Asteroids"] in the foreseeable future." The implication was that the Earth was near the middle of the "error ellipse" and thus could be struck.
The error ellipse in this case is a 2-dimensional elliptical region, plotted on the plane through the center of the Earth, perpendicular to the path of the asteroid on 26 October, 2028. It is associated with the probability distribution for 1997 XF11 to pass through a given point in that plane. In three dimensions, that probability distribution would look something like a smooth elliptical hump, or mountain, shaped like the usual bell-shaped normal distribution when cut through either the long or short axis. It was immediately clear that "virtually certain" meant that the ellipse was smaller in both dimensions than the Moon's orbital radius. How much smaller was very interesting, but not immediately clear from the IAUC.
We often say "3 sigma" in science as a kind of jargon for "pretty certain", meaning within 3 times the standard measure of the width of the normal curve. This measure, sigma, is defined so that there is about a 68% probability that something described by the normal curve will lie less than sigma away from the mean; the "3 sigma" translates to about 1 chance in 750. So if Marsden's "virtually certain" meant that sigma = 100,000 km (so 3 sigma was 300,000 km), and the center of the ellipse was just 46,000 km from Earth, then the implication was clear that the odds that we would be hit were most likely on the order of 1%, probably a little less. As the size was roughly estimated to be of order 1 km, even a 1% chance was quite interesting, to put it mildly.
Sigma is large for poorly determined orbits, and gets smaller and smaller as more and better observations are made. For some newly-discovered asteroids, sigma would be millions of km. For others with well-determined orbits, it might be 100 km or even less. If sigma were huge, enormously larger than the Earth, then even if the ellipse were perfectly centered on us, we would have little to fear in practice, because the chance of actually being hit would be negligible. At the other extreme, we would similarly be quite safe if the center of the ellipse were only, say, 500 km above Los Angeles, if sigma were only 20 km, because we would know the orbit so precisely that we could say with great confidence that it would miss, in spite of an extremely close call. (In such a situation we would obviously be strongly motivated to measure the orbit very carefully!)
In the case of 1997 XF11, the implied size of the ellipse and its nearness to the Earth created a fair sensation. Fortunately for all our nerves, the orbit was already well determined enough that it could be accurately extrapolated backwards in time, as well as forwards to 2028. Thus in only a few hours, among the thousands of archival photographic plates that have been taken, it was discovered that 1997 XF11 appeared on two pre-discovery 0.46 m Schmidt exposures obtained in 1990 March by K.A. Lawrence and E. F. Helin at JPL, as part of the Planet-Crossing Asteroid Survey. By the end of the next day, improved orbits had been calculated both at SAO and JPL, and it quickly became apparent that 1997 XF11 would pass well beyond the Moon. So we have had a not-very-near escape.
Nevertheless, we should all bear in mind that regardless of the verdict on 1997 XF11, this sort of false alarm is going to happen again, probably many times, because there are literally millions of objects in the 10 to 100 m size range, still floating around out there and big enough to be bad news if they hit London or Johannesburg. Naturally most miss, and we just have not noticed them before. Naturally, smaller ones are much more common than big ones. But a few do hit every century that are really good sized: (like Tunguska 1908 -- probably 50 Megatons or so); a famous near-miss in 1972 August over the western USA (a fireball actually grazed the atmosphere, close and large enough to be spectacular in the daytime over a track many hundreds of km long, but then escaped into space); and one that seems to have struck southern Greenland just this winter (see "Aviation Week and Space Technology", late 1998 February).
Only recently have there been systematic programs to detect these little (and not so little) guys before they surprise us. With sensitive survey programs now under way - which we will describe in more detail next time - it seems inevitable that lots and lots of such things will be turning up in the future, long before we need to take action to prevent an actual disaster. So the process just played out by 1997 XF11 will almost surely be repeated: a potentially dangerous object will be discovered, new or old observations will accumulate, and the error ellipse will get smaller. For most, we will soon discover that we are safe for now. But for a few objects, the ellipse will get not only smaller, but also closer; so that we will have a more or less extended period of suspense before we finally learn the truth. And eventually the truth we learn will be that an impact is certain. Even then, since small asteroids and defunct comets are more numerous than larger ones, we will probably not need to take action. Indeed, in the past few years it has become clear that impacts the size of small nuclear explosions are fairly common, but almost always happen in conveniently uninhabited places, like Tunguska, Greenland, or the South Pacific.
In this situation there are at least two important ways that amateur astronomers can contribute. The first is simply to be aware and help educate those around them and the public in general about the prospect that is in store, and what it means: that the threat is quite real but generally very long-term; that many false alarms are inevitable; that numerous potential threats will need to be followed, but should not worry anyone unduly; that we may expect some genuine near misses; that there will be some sizable yet harmless impacts; and eventually (probably not for many decades or even centuries) we must expect some impact threats which we will need to take action to prevent. Most likely the best means for doing so will be reasonably clear in view of the technology and capabilities available to us at that time. Certainly (in view of the public's suspicion and tendency to see conspiracy everywhere) we will need learn to avoid premature scenarios of doom.
The other contribution is that astrometric observations have become practical with telescopes of modest size, due to the metric stability and sensitivity of CCD's, and the new availability of large numbers of very accurately known Hipparchos astrometric reference stars. Especially in the southern hemisphere, where well-equipped amateurs are relatively few, there are great opportunities for time-critical follow-up observations which could easily make all the difference between getting a good orbit for an object or losing it altogether. Even were the object to be recovered at a later date, since it is usually much easier to deflect a body if we act sooner than later, early and accurate information could very possibly have enormous economic and human consequences.
In any event, along with the ordeal we have in store of watching and worrying about yet another kind of global disaster as these little worlds whiz by, the good news is that we will probably have lots of practice getting our act together, before a real disaster comes into focus -- unless we are very unlucky, of course!
Details of some asteroid survey programs, NEAR at Mathilda, Mars Global Surveyor science, and Water on the Moon: all will have to wait for next time.