I had expected to describe the Far Ultraviolet Explorer, FUSE, which was successfully launched on 1999 June 24. However, so much other news has come up that it will again have to wait for a future issue.
The biggest news by far this month is the successful launch of Chandra, third of the NASA Great Observatory series, which I described a year ago when it was still known as AXAF, the "Advanced X-ray Astrophysics Facility". The Einstein Observatory, launched in 1978, was the first large focusing X-ray telescope in space, and resulted in a sensitivity improvement of about 1000 over its predecessors. Ever since Einstein's enormous success, X-ray astronomers have been yearning to follow up along the many alluring paths it first suggested. Now, after 21 years, the doors are finally open.
A year or so ago, based on the experience with 1997 XF11, I predicted that many more asteroid impact false alarms would be forthcoming. This expectation was based partly on the obvious fact that the chances of an object passing "very close" (say 65,000 km, ~10 earth radii) to us, is much larger ( ~100 times) than the chance of our being actually hit. However it was also partly due to the details of the way in which poorly-known orbits are improved, inevitably giving us a little suspense in the midst of the process. This month we will study this process in more detail in a typical case, which also happens to be the current "threat champion" in the hazardous asteroid business.
In the good old days we never noticed near passages, quietly happening all the time. Now that we can actually take action to affect the result if need be, we have a program to detect all the dangerous asteroids >1 km in size within the next decade, so that they can be monitored. So it is inevitable that lots of close calls are going to be detected, evaluated, and almost all eventually dismissed as more accurate measurements become available showing no impact is possible.
Sure enough, two new potential impactors have turned up in the past few months. With the collection of more observations, and yet again (as for 1997 XF11) with the help of previously unrecognized historical observations, neither object now appears to pose any threat in the near term. Nevertheless, the situation for 1999 AN10 (size, about 1 km) as it appeared based on the data available in early June, was interesting. Since it is also likely to prove typical of many future alarms, I think it is worth study in some detail. The circumstances are illustrated in the figure below.
The "Impact Plane", so-called, is actually the plane through the center of the Earth, perpendicular to the velocity of 1999 AN10 in a co-ordinate frame in which the Earth is at rest, on 2027 August 7. The small circle at the origin (0,0) in the lower right is the "target": namely, us. The point at which 1999 AN10 will pass through the 2027 August 7 impact plane cannot (of course) be exactly known; yet because its orbit had been fairly well determined, we did know that it will pass somewhere within a small region in this plane. This region, even based on the observations available on June 1, does not include the Earth. So far so good, it seems.
It is characteristic of such possible regions that they are nearly elliptical in form, with the highest danger being at the center of the ellipse, and tapering away in all directions from that most hazardous point. The probability of 1999 AN10 passing through any given point, if plotted in proportion vertically above that point, generates a shape that looks like a smooth hill. It may not surprise you greatly that a cut through the center of the hill, perpendicular to either of the two ellipse axes, has the familiar bell-curve shape: the "normal curve" of probability.
The line labeled "uncertainty ellipse" is that region, based on the orbital solution of early June. It appeared to be a line rather than an ellipse simply because it was extremely long in one dimension, and very narrow in the perpendicular direction, so the width could not be ascertained by eye at the scale of the figure. Notice also that, since the normal curve does not ever return down completely to zero, we must be satisfied to plot a contour level >0 on the smooth hill, one chosen so low that the danger the asteroid will pass outside it is judged negligible. Any very small value can be chosen; 1% is a common choice. Fortunately, since the normal curve declines extremely rapidly to minuscule values, any reasonable choice does not affect the plotted size of the elliptical contour region too markedly.
Yet because the ellipse was so narrow, there was essentially no chance that it would strike us in 2027, for the possible region as it passed nearest the Earth was basically a long narrow strip, about 40,000 km away, but only a few hundred km wide. Or, the asteroid might easily have passed much further than 40,000 km, for the tip of the ellipse as drawn is approximately at (-430,000; 150,000) km. This is about 450,000 km away, well beyond the Moon. It might seem odd at first that one could say only that the asteroid will pass at a nearly unknown distance, between 450,000 km and 40,000 km from the Earth, and still assert with great confidence that it will not impact. The geometry of the enormously elongated ellipse in the figure, which is the typical situation, should help to make this more clear in the many encounters of hazardous asteroids that are inevitable in the future.
But 2027 itself is only part of the story. Even in the event of a miss, it is clear that if 1999 AN10 were to pass near the Earth, its current orbit would be strongly perturbed, conceivably into a new path which might impact on some future date. Because an elliptical Keplerian orbit is, after all, periodic, retracing itself year after year, any new orbit would necessarily pass near the Earth's orbit in the future, barring perturbation by some third body in the Solar System. The three points marked "keyholes" indicate such possible hazards, in that passage near those particular points might have changed the orbit to give a dangerous encounter in the three future years given. The details of an encounter following any of the "keyhole" passages would depend extremely sensitively on the precise position in the encounter plane in 2027. (The point for 2034 seems to be far outside the target ellipse, but because the ellipse falls off fairly slowly in the long dimension, given the inevitable arbitrariness in drawing the probability contours noted above, its probability still merited attention.)
The figure is based on the one set of data. Most sets of observations will result in some such ellipse; as more observations are added, the resulting new ellipses shrink, becoming successively smaller and smaller. But because any new data ought to be consistent with the old, addition of new observations should only reduce the size within the confines of previous ellipses, without moving it elsewhere in the impact plane. The result of continuing observations is therefore normally a set of nested ellipses of decreasing size, converging on the true point of passage through the impact plane. Early in the process, the error ellipse is so large that, even though it may contain the Earth, the likelihood of impact is negligible. Much later the ellipse shrinks until most likely, as here, the Earth is excluded. In between there is necessarily a period when the danger rises to a maximum. If this maximum is in the range above say, 0.1%, we have a false alarm likely to attract media attention.
For 1999 AN10, most recently, a pre-discovery observation on Palomar Sky Survey plates, found by A. Gnadig and A. Doppler and dating back to 1955, has shown that the 2027 encounter will in fact miss all of the keyholes in the diagram, passing Earth at 389,000 km. This removes the danger until an approach in 2076 February. Nevertheless, presuming a miss in 2076, it still seems that 1999 AN10 will remain a threat that needs monitoring for about the next 600 years, when the current series of close approaches should end. For now, it is the most hazardous earth-approaching asteroid known in the 1 km-or-above size class.
An especially good site for the most recent analyses of currently developing hazards is at the Space Mechanics Group of the University of Pisa: http://newton.dm.unipi.it/neodys/; the figure came from there. The standard repository for asteroid information is the IAU Minor Planet Center, at SAO; its URL is http://cfa-www.harvard.edu/cfa/ps/mpc.html .
Another remarkable object, 1998 KY26, was discovered last year passing at a distance of 800,000 km, or just twice that of the Moon. In the 1999 July 24 issue of "Science", Steve Ostro of JPL and an international team report it is unique in several regards:
Based on radar and optical observations, Ostro et al. suggest it is likely to be a carbonaceous chondrite, composed of complex organic compounds and, probably, several thousand tons of water bound chemically in the minerals.
A large number of other events this month deserve mention, if only in passing. First, Deep Space 1, its technology demonstration objectives accomplished with near perfect success, will fly by the barely-known 1-2 km asteroid 1992 KD, on July 29 at approximately 04:46 UT, at a speed of 15.5 km/s and a distance of only 15 km. At the very least this should be exciting! Then, on July 31, Lunar Prospector -- its primary mission all complete -- is to be deliberately crashed into a likely ice-bearing polar crater, in the hope of raising a cloud of water vapor that will be illuminated by the Sun and observed by waiting telescopes on Earth. Finally, Cassini, having flown by Venus again in June, will fly by Earth on August 18, and then be on its way to Jupiter and the outer Solar System.