A&A 406, 751–758 (2003) Astronomy DOI: 10.1051/0004-6361:20030641 & c ESO 2003 Astrophysics Close encounters of asteroids before and during the ESA GAIA mission A. Fienga, J.-F. Bange, A. Bec-Borsenberger, and W. Thuillot Institut de m´ecanique c´eleste et de calcul deseph´ ´ em´erides (IMCCE) – Observatoire de Paris, UMR 8028 du CNRS, 77 avenue Denfert-Rochereau, 75014 Paris, France Received 15 October 2002 / Accepted 15 April 2003 Abstract. Observation of close encounters of asteroids is a powerful method to determine their masses. A systematic search of such close encounters of asteroids with diameters larger than 40 km has been made thanks to a procedure to select the most efficient phenomena by means of the observable gravitational deflection. This study allows us to give lists of such single (one encounter) and multiple (several encounters between two pairs of asteroids) phenomena that will be observable from ground based astrometric telescopes from 2003 to 2022. We also give lists of single and multiple phenomena spanning 2010-2022 and implying less sensitive deflections only accessible by space astrometry. These last encounters may be observed during the ESA GAIA space mission. Key words. minor planets, asteroids – astrometry – celestial mechanics – methods: observational 1. Introduction mass of Ceres (Schubart 1971; Landgraf 1988; Goffin 1991; Williams 1983; Viateau & Rapaport 1998; Michałak 2001). The determination of the masses of asteroids remains a primor- With similar methods, several papers recently led also to the dial problem to be solved in order to improve our knowledge determination of the mass of smaller asteroids from ground- of the dynamics of the asteroid families. Such progress could based observations (Viateau & Rapaport 2001, 1997; Viateau provide information on their origin and on the formation pro- 2000; Michałak 2001). Today, the ground-based observations cess of the solar system. Furthermore it could also give op- seem to be of limited use for the search for new opportunities portunities to improve the accuracy of the theory of motion of of mass determination. Favourable close encounters are rather Mars which is now limited by the uncertainties on the asteroid rare. A systematic search was carried out by Kuzmanoski & masses (Standish & Fienga 2001). Kne˘zevi´c (1993), who have computed opportunities of very Unfortunately, direct determination of asteroid masses is close encounters spanning 1993–2043. Hilton et al. (1996) pub- ffi di cult. Less than twenty of them are currently available lished their prospects for determining masses of large aster- (Michałak 2001). Furthermore the uncertainties remain high, oids based on their own computations of close encounters from even for the largest asteroids Ceres, Pallas and Vesta (Hilton 1950 to 2017. More recently Gal´ad (2001), and Gal´ad & Gray ff 1999). The analysis of the perturbing e ects of the mass of an (2002) studied much longer lists of asteroids and gave dates of asteroid on the orbit of a target asteroid is the usual method to close approaches from 1967 to 2023. perform direct mass determination. For this purpose close or- bits of asteroids involving one or several encounters have to be The accuracy of recent ground-based observations by detected and computed. These resonant orbits are thus able to CCD astrometry, reduced using astrometric catalogues, can accumulate weak perturbations all along large periods of time reach 50 milliarcsec or mas (see for example Fienga 1998). of the order of ten or hundred years (Schubart 1974). Unfortunately, the need to include older and less accurate ob- servations, performed with other techniques, imposes a strong After the first determination of the mass of Vesta by Hertz restriction on the detection of perturbing effects of a given or- (1968), the number of direct mass determinations using reso- bit by another asteroid. Furthermore, the next important step nant orbits increased, leading to several determinations of the for asteroid mass determination will be achieved thanks to the Send offprint requests to: A. Fienga, e-mail: [email protected] new observational methods: mainly adaptive optics and space Tables A.1–A.8 are only available in electronic form at the CDS astrometry. The recent advent of adaptive optics has lead to the via anonymous ftp to discovery of new satellites of asteroids and subsequently will cdsarc.u-strasbg.fr (130.79.128.5) or via us to get new determination of masses (Merline et al. 2000; http://cdsweb.u-strasbg.fr/cgi-bin/qcat?J/A+A/406/751 Margot et al. 2000). But, beyond a deflection, the mass de- or http://www.imcce.fr termination requires one to observe large enough arcs of the Article published by EDP Sciences and available at http://www.aanda.org or http://dx.doi.org/10.1051/0004-6361:20030641 752 A. Fienga et al.: Close encounters before and during the GAIA mission Table 1. Main characteristics of the GAIA mission. than 40 km (among the first 8000 objects) taken from the ASTORB database (Bowell 2001). We used the orbital el- Magnitude limit 21 ements from this database as integration constants with the = Number of objects: stars 1.3 billions (to V 20) Bulirsch and Stoer algorithm (1966) of numerical integration. 5 6 asteroids 10 to 10 The planetary perturbations of all the planets were taken from Accuracy: stars 10 µmas the VSOP ephemerides (Bretagnon 1982) asteroids 0.1 mas Mission duration ≥5 years It is an essential requirement for asteroid mass determina- tions to evaluate the characteristics of the gravitational event efficient for mass determination and to provide an estimate of satellite orbit in order to deduce the mass through the mea- their efficiency. surements of its orbital period. Considering these new tech- From 1971 till now, different criteria were considered. The niques of observations and the new accuracy of the astrometric minimal distance between the two asteroids, i.e. the impact pa- reductions, we think that close encounter observations remain rameter, is the usual criterion. The necessity of finding close very valuable and we give, in this paper, several dates of enough orbits can explain this choice. But actually, the mini- favourable close encounters from 2003 to 2022 which could mal distance is not the most determining parameter. As pointed allow us to determine asteroid masses thanks to such 50 mas out by Hoffman (1989) and Kuzmanoski & Kne˘zevi´c (1993), ground-based observations. the relative velocity at close encounter (or “encounter veloc- From space, two kinds of mass measurements are possible. ity”) is also an important characteristic to be considered. The On the one hand, spacecraft can furnish opportunities for in situ masses of the bodies involved in a close approach is another asteroid mass determinations. This was the case with the fly-by parameter of importance. Thus, a criterion of maximal deflec- of the small asteroid 243 Ida by Galileo (Belton et al. 1995), tion, a number directly related to the deviation occurring in a 253 Mathilde and 433 Eros by NEAR (Yeomans et al. 1997, close approach between two asteroids, can also be used as a 2000). On the other hand, data obtained by astrometric satel- selection criterion. lites have been proved to be helpful for mass determinations, Bange (1998) defined a criterion of deflection proportional / v2 v due to the very good accuracy they provide in a short time in to mp rm 0, where 0 is the relative velocity between the two comparison with ground-based observations. The ESA satellite bodies, mp is the mass of the perturbing asteroid and rm is the Hipparcos (ESA 1997) was the first instrument devoted to as- minimal distance at close encounter. This criterion is more sen- trometric measurements of positions of stellar and solar system sitive to the encounter velocity than the other criteria described objects. Among them, the astrometric data of minor planets led in literature. This method was already used to study and se- an estimation of the mass of asteroid 20 Massalia (Bange & lect close encounters of minor planets observed by Hipparcos Bec-Borsenberger 1997; Bange 1998) which was the first deter- (Bange & Bec-Borsenberger 1997). mination of an asteroid mass using the short time span of close Related to these characteristics, we define the concept of encounter observations. The GAIA mission is now planned by a mutual gravitational event (MGE). A mutual gravitational ESA to follow the Hipparcos mission in this way. Providing event occurs when the relative velocity becomes minimum or many more observations of greater accuracy (see Table 1), its the impact parameter becomes minimum or the criterion of de- data, planned to be taken between 2010 and 2022, will be of flection becomes maximum. These events (minimum of rela- the greatest interest to study close encounters among asteroids. tive velocity or impact parameter, or maximum of criterion) This project will allow a limiting accuracy two orders of mag- could occur at different epochs or within a delay of several nitude better than its predecessor, Hipparcos. It is presently es- days. In such cases, we have chosen to consider separately each timated to lead to positions, proper motions and parallaxes of event. This type of gravitational event is then seen as multiple about one billion objects, with an expected accuracy of 10 mi- events. In the case where the minimum of relative velocity, the croarcseconds, along with multicolour multi-epoch photometry minimum of the impact parameter and the maximum of the cri- of each object. terion of deflection occur once at the same time, then one can In the first section of this paper, we define criteria and pa- consider the event as a single event. rameters used to identify close encounters efficient for mass Furthermore, as described by Michałak (2001), the cri- determinations.