Several times a year, there is a story in the press about an asteroid that narrowly misses the earth. One of the nearest--and largest--misses was by one that whizzed by at little more than the distance of the moon in May 1996 and was clearly several hundred feet across. A closer encounter in December 1994 was with an asteroid perhaps only 20 feet in diameter. Comets can also be a hazard, as was illustrated earlier that year when a score of cometary fragments made a dramatic score on Jupiter. It is only a matter of time before the earth sustains a direct hit. Clearly, some objects are potentially more dangerous than others, and astronomers have often worried about how to distinguish the serious problems from the wolf-crying cases that are publicized all too often. Some would rigorously examine the probability that the range of uncertainty in the predicted position of each object over some interval of time includes that of the earth and then compute the kinetic energy of a possible resulting collision. Others are content to identify objects that may come close enough and perhaps be large enough to bear watching over a less definite timespan. The latter approach inspired the rather simple definition of a "potentially hazardous asteroid", the hundredth member of which was recognized at the beginning of September.
The longest-known asteroid in the broad category of what are nowadays called "NEOs", or "Near-Earth Objects" was discovered 100 years ago next August. This is Eros, and a calculation published last year indicated that there is actually a very good chance that it will collide with the earth during the next million years. Extend that to 100 million years, and much the same will be true for all the other 419 asteroids now classified among the NEOs--unless they collide with the sun first or are gravitationally ejected from the solar system. Sometimes the NEOs are separated into "Atens", "Apollos" and "Amors", according as to whether they have orbits smaller than that of the earth, whether they go closer to the sun than the earth does, or whether they pass somewhat beyond to the earth's orbit on the outside. Given that the closest distances of the Amors to the sun are currently as much as 30 percent larger than the earth's distance from the sun, it is also appropriate to include among the NEOs the majority of the known comets, thereby yielding a total of more than 1000 known NEOs.
But almost all of these 1000 NEOs will "not be a problem" over even tens of thousands of years! Even though these objects do pass from outside to inside the earth's distance from the sun and back, many of their orbits really come nowhere near the orbit of the earth. Recognizing this, the concept of an ECA, or "Earth-Crossing Asteroid" was introduced some years ago to refer only to those asteroids that might conceivably approach us, even when one considers the manner in which their orbits will gravitationally evolve over tens of thousands of years. But that calculation of what are termed "secular perturbations" is quite time-consuming, is meaningless if the actual observations cover only a few days, and can be significantly in error if there are close approaches to or low-order resonances with the major planets.
From an entirely practical point of view, we are really interested only if there is to be a collision with the earth over the next couple of centuries. Taking the probabilistic approach thus has merit in this connection, but it does not say anything about the situation beyond the time interval for which the computation--which is again quite time-consuming--is actually done. In any case, there is a danger in taking the "computation of the error ellipsoid" too seriously. This computation assumes one really has some measure of the accuracy of the observations, a point that is particularly troublesome for observations substantially isolated in time from the bulk of the data. It can also be in error if the assumed force model is wrong, as could happen if one ignores a close encounter with some other planet or asteroid; and if the object of concern is a comet, there is the problem of the very imperfect knowledge of the nongravitational force arising from the reaction of the comet to vaporization of its icy constituents. Furthermore, the calculation of the collisional kinetic energy must also be uncertain. This depends on the object's mass. The mass in turn depends on the density and the radius. Any estimate of the density is a pure guess. The radius depends on the albedo and the brightness. Except in those rather rare instances where detailed observations of the brightness have been made in different colors, the albedo, or efficiency with which the object's surface reflects sunlight, is again a guess. Finally, the observed brightness itself, which is in fact used to compute the absolute magnitude--a number showing how bright the object would be if it were located such that it, the sun and the earth were at the vertices of an equilateral triangle--is often itself quite a crude estimate. In short, one can go to a fair bit of effort to compute an impact energy that is significantly in error.
So we come to the definition of a potentially hazardous asteroid, or PHA. To be a PHA, an object should have an orbit that brings it quite close to that of the earth, at present and for a couple of centuries into the future. The object should also have an absolute magnitude bright enough that the object is likely to be large enough that a collision would result in widescale, perhaps even global, damage. The numerical limits can be debated, of course, but it is not unreasonable to choose an orbital distance of 5 percent of the distance to the sun (say, 5 million miles) and an absolute magnitude of 22 (which is likely to include all objects down to a diameter of 200 meters (650 feet), which is in turn at a level where an ocean landing is likely to yield devastating tsunamis on even distant shores). Of course, if the orbit has been computed from observations covering several years, we can examine the possibility of encounters with the earth at specific times. But even when an object has been observed for only a few days, the minimum distance between the orbits is generally well established, even though the actual location in the orbit is completely unknown. It is even sufficient to simplify the computation on the assumption that the earth's orbit is a circle. Furthermore, in many instances the minimum orbital distance is unchanged for a couple of centuries. Severe changes in this quantity occur mainly for the asteroids with revolution periods close to one-third that of Jupiter, a point that needs to be checked as the orbit computations improve. Close approaches to Jupiter are of course also a problem, but that applies mainly to short-period comets, for which it is unwise to use the PHA definition. Known long-period comets are also not a threat because, by definition, they can make no more than one approach to the sun over the course of two centuries.
The first PHA was Apollo, discovered in 1932 and eventually rediscovered many revolutions and 41 years later. The same was true of Adonis, discovered in 1936. The third PHA was Hermes, discovered on its particularly close approach to the earth in 1937 and lost ever since--a halfmile-sized object that is therefore one of the most dangerous NEOs. Half of the PHAs have been found since 1991, one quarter since 1994. The year 1997 has so far seen seven new PHAs. The first of these was under observation for only six days and is therefore lost. The others, including the latest, 1997 QK1, have been tracked for at least 20 days and can probably be reobserved in future years, if the effort is made to search for them. That effort would probably be worthwhile.
Brian G. Marsden
1997 September 25
This page may not be copied onto other Web sites, but other sites may place "clickable" pointers to this page. Photo-copies of this information sheet may be distributed for educational purposes, provided that no charge is made in doing so; such photocopies must include full credits. Journalists and writers may quote from this "information sheet", provided that proper credit is given. Further information is available by writing to Mail Stop 18; Smithsonian Observatory; 60 Garden St.; Cambridge, MA 02138; U.S.A.
Back to index of Press Information Sheets.
Index to the CBAT/MPC/ICQ pages.