As I surmised long ago, the problem with fitting the comet's orbit is quite clearly that, if one tried a standard orbit solution, the sheer weight of the 1995 data would throw a large residual into the single 1993 position, basically on account of the systematic errors in the GSC reference-star system used for most of the measurements. After all, the many hundred 1995 observations covered only a small part of the sky. By substantially reducing (e.g., by a factor of ten) the relative weight of the 1995 data, it was in fact possible to fit the 1993 position completely satisfactorily. The alternative of invoking the effects of nongravitational forces on the comet seemed very unlikely at the comet's large heliocentric distance, even if one believed, as some did, that the comet's great activity involved enormous relative mass loss in terms of vaporizing carbon monoxide.
If the surmise were correct, it would presumably be found that the fit to the 1993 Apr. 27 observation would again improve as the comet's observed arc extended, with the observations during 1996, though continuing to be numerous, covering a greater area of sky. This did in fact happen, acceptable orbital solutions being possible quite early in the year with the relative weight of the current data reduced to only a factor of five. Since then there has been further steady improvement. Finally, when observations extended to mid-October, a satisfactory solution could be obtained that was fully able to incorporate the 1993 observation with unit weight.
A new orbit from observations (e.g., MPC 28557) extending to mid-December gives the (O-C) of the 1993 observation as -1.1 arcsec in R.A. and -0.5 arcsec in Decl. Actually, the point has been reached where it really does not matter whether the 1993 observation is included or not. The outcome is almost precisely the same whether considered in terms of the orbital elements and their formal errors or in terms of the near-perihelic sky position, which is now formally predictable to better than 1 arcsec.
Statements to the effect that occultations by the comet's nucleus can be predicted more accurately using orbital solutions that omit the 1993 observation are thus patently false. This is not to say that occultations near the time of perihelion can yet be precisely predicted. Sources of error are the small nuclear size, the star catalogues used and nongravitational forces in the cometary motion. The first of these will require last-minute relative measurements of comet nucleus and star to be occulted to better than 0.05 arcsec. In the absence of such measurements, star-catalogue errors are likely to result in near-perihelic unpredictability to some arcseconds. The possible effects of nongravitational forces are unknown, and they proved to be a severe problem in predicting the post-perihelic position of comet C/1996 B2 (Hyakutake), for example. While both comets are clearly very active, one might surmise that the larger relative size of C/1995 O1 (Hale-Bopp) will result in little nongravitational influence on this comet's motion.
The last perihelion passage of C/1995 O1 occurred close to 4210 years ago. In the absence of nongravitational forces near perihelion, the next return after 1997 will be some 2380 years hence.
Brian G. Marsden 1997 Feb. 5
Back to Astronomical Headlines.