A typical final product of cometary orbit determination is a set of osculating orbital elements, derived at some fixed epoch, usually close to the moment of the perihelion passage. These elements are influenced by all perturbations a comet could suffer during its passage through the planetary system towards its perihelion, namely gravitational perturbations from major and minor bodies, relativistic effects and non-gravitational forces, including those resulting from physical processes on the cometary nucleus. If we want to study the dynamical history of such a comet, i.e. its past motion before it entered the planetary region, it is necessary to trace backward its motion among the planets, taking into account all the perturbations mentioned above (assuming they are also included in the orbit determination process, which is a standard procedure nowadays) up to the heliocentric distance at which all those perturbations can be neglected. In this paper we calculate barycentric original orbital elements at a heliocentric distance of 250 AU, according to the suggestions by Todorovic-Juchniewicz (1981) and adding an extra "security margin''. We call this distance the "Planetary Perturbation Limit'' (PPL).
As the initial values for such a calculation we used the osculating orbital elements, published in the Catalogue of Cometary Orbits (Marsden & Williams 1997), using only orbits with the quality class 1 or 2, according to the classification proposed by (Marsden et al. 1978). This automatically excludes all comets having determined non-gravitational parameters. Additionally, we supplemented this dataset with those for several long-period comets whose orbits were published up to the middle of 1999; more precisely, we modified the orbits of 9 comets, added 20 new orbits and removed one orbit because of non-gravitational parameters published recently. All modifications to the Catalogue of Cometary Orbits are summarized in Table 1.
For some statistical analysis presented later in this paper we additionally restrict the sample of the observed long-period comets to those with observed perihelion distances q>1.5 AU, which, according to Oort (1990), should exclude all comets with potentially significant non-gravitational effects.
| Comet | Source | Class | Comment |
| C/1996 J1-B | MPC33450 | 1A | modified |
| C/1997 A1 | MPC30957 | 1A | modified |
| C/1997 BA6 | MPC34420 | 1A | modified |
| C/1997 D1 | MPC30957 | 1A | modified |
| C/1997 J1 | MPC31482 | 1A | modified |
| C/1997 J2 | MPC33451 | 1A | modified |
| C/1997 L1 | MPC30738 | 2A | modified |
| C/1997 O1 | MPC30957 | 2A | modified |
| C/1997 P2 | MPC30738 | 2A | modified |
| C/1997 T1 | MPC32168 | 1A | added |
| C/1998 H1 | MPC32168 | 2A | added |
| C/1998 K1 | MPC33451 | 1B | added |
| C/1998 K2 | MPC32594 | 1B | added |
| C/1998 K5 | MPC33857 | 1A | added |
| C/1998 M1 | MPC34733 | 1A | added |
| C/1998 M2 | MPC32865 | 1B | added |
| C/1998 M3 | MPC34733 | 1A | added |
| C/1998 M5 | MPC34733 | 1A | added |
| C/1998 M6 | MPC33188 | 1B | added |
| C/1998 Q1 | MPC32595 | 2B | added |
| C/1998 T1 | MPC33451 | 2A | added |
| C/1998 U1 | MPC33650 | 2A | added |
| C/1998 U5 | MPC34421 | 1B | added |
| C/1998 W3 | MPC33857 | 2A | added |
| C/1999 A1 | MPC33858 | 2B | added |
| C/1999 F1 | MPC34734 | 1B | added |
| C/1999 F2 | MPC34734 | 1A | added |
| C/1999 K2 | MPC34735 | 2A | added |
| C/1999 K6 | MPC34735 | 2B | added |
| C/1995 O1 | MPC32410 | - | removed |
Finally, we performed our investigation of the past motion for a sample of 327 long-period comets, 195 of which have class 1A or 1B orbit, while 132 comets are of class 2A or 2B.
For such a population we calculated barycentric, original orbits, some details of which are found in Sect. 2, and then, for those comets, whose original aphelion distances were greater than the PPL, we performed backward numerical integrations of their motion, described in detail in Sect. 4. In that section we also discuss the importance of stellar perturbations, giving some examples, and propose a revision of the criterion for calling a particular comet dynamically "new''. In Sect. 3 we present some statistical characteristics of the results obtained in the present investigation as well as a comparison with similar results previously published, e.g. Yabushita (1989), Oort (1950, 1990). Section 5 shows several attempts to find a correlation between the past motion parameters and various physical characteristics of comets. At the end we discuss the possibility of distinguishing new (first time near the Sun) and old comets on the basis of their dynamical history and give some final remarks and conclusions.
Copyright ESO 2001