Margot J.-L., Pravec P., Taylor P., Carry B., and Jacobson S. (2015) Asteroid systems: Binaries, triples, and pairs. In Asteroids IV (P. Michel et al., eds.), pp. 355–374. Univ. of Arizona, Tucson, DOI: 10.2458/azu_uapress_9780816532131-ch019. Asteroid Systems: Binaries, Triples, and Pairs Jean-Luc Margot University of California, Los Angeles Petr Pravec Astronomical Institute of the Czech Republic Academy of Sciences Patrick Taylor Arecibo Observatory Benoît Carry Institut de Mécanique Céleste et de Calcul des Éphémérides Seth Jacobson Côte d’Azur Observatory In the past decade, the number of known binary near-Earth asteroids has more than qua- drupled and the number of known large main-belt asteroids with satellites has doubled. Half a dozen triple asteroids have been discovered, and the previously unrecognized populations of asteroid pairs and small main-belt binaries have been identified. The current observational evidence confirms that small (<20 km) binaries form by rotational fission and establishes that the Yarkovsky-O’Keefe-Radzievskii-Paddack (YORP) effect powers the spin-up process. A unifying paradigm based on rotational fission and post-fission dynamics can explain the forma- tion of small binaries, triples, and pairs. Large (>20 km) binaries with small satellites are most likely created during large collisions. 1. INTRODUCTION (Merline et al., 2002c; Noll et al., 2008). As such, they pro- vide an invaluable window on accretional, collisional, tidal, 1.1. Motivation and radiative processes that are critical in planet formation. The distribution and configurations of the multiple-asteroid Multiple-asteroid systems are important because they systems also provide a rich array of constraints on their en- represent a sizable fraction of the asteroid population and vironment, their formation, and their evolutionary pathways. because they enable investigations of a number of properties Observations rely primarily on groundbased telescopes and processes that are often difficult to probe by other means. and the Hubble Space Telescope (HST). For an up-to-date The binaries, triples, and pairs inform us about a great variety list of binaries and triples in the solar system, see Johnston of asteroid attributes, including physical, mechanical, and (2014). We describe observational techniques only briefly thermal properties, composition, interior structure, formation because this material is available elsewhere (e.g., Merline et processes, and evolutionary processes. al., 2002c). A few emerging techniques will be described in Observations of binaries and triples provide the most more detail. Likewise, we refer the reader to other texts for powerful way of deriving reliable masses and densities for an extensive history of the field (e.g., Merline et al., 2002c) a large number of objects. The density measurements help and highlight only a few of the developments here. us understand the composition and internal structure of mi- nor planets. Binary systems offer opportunities to measure 1.2. History thermal and mechanical properties, which are generally poorly known. Early search programs for asteroid satellites were unsuc- The binary and triple systems within near-Earth asteroids cessful, returning negative or dubious results, such that the (NEAs), main-belt asteroids (MBAs), and transneptunian authors of the Asteroids II review chapter chose the prudent objects (TNOs) exhibit a variety of formation mechanisms title “Do asteroids have satellites?” (Weidenschilling et al., 355 356 Asteroids IV 1989). The chapter provides an excellent discussion of the objects are substantially affected by the Yarkovsky-O’Keefe- physics of several formation mechanisms that were postu- Radzievskii-Paddack (YORP) effect during their lifetime. lated at the time. The perspective changed with the flyby of For typical NEAs and MBAs, this dividing line corresponds (243) Ida by the Galileo spacecraft in 1993 and the discovery to a diameter of about 20 km (Jacobson et al., 2014a). of its small satellite Dactyl (Chapman et al., 1995; Belton We define very small asteroids as those with diameters of et al., 1995). Groundbased efforts intensified and resulted in <200 m. This is the approximate size below which many the discovery of a satellite around (45) Eugenia by Merline asteroids are observed to spin faster than the disruption rate et al. (1999). Several other discoveries followed in rapid of a body with no shear or tensile strength wd = 43πρG , succession. The relatively small sizes of the MBA satellites where r is the density and G is the gravitational constant. suggested formation in subcatastrophic or catastrophic colli- We use two additional acronyms. The YORP effect is a ra- sions (Durda, 1996; Doressoundiram et al., 1997). diation-powered rotational acceleration mechanism for small The discovery of MBA satellites, coupled with analysis asteroids (Rubincam, 2000). The binary YORP (BYORP) ef- of terrestrial doublet craters (Bottke and Melosh, 1996a,b) fect is a radiation-powered acceleration mechanism that may and anomalous lightcurve observations (Pravec and Hahn, expand or contract the orbits of some synchronous asteroids 1997), suggested the existence of binary asteroids in the near- (Ćuk and Burns, 2005). Earth population as well. The unambiguous detection of five NEA binaries by radar cemented this finding and indicated 2. OBSERVATIONS that NEA satellites form by a spin-up and rotational fission process (Margot et al., 2002). Lightcurve observers reached Several observational techniques are available for discover- the same conclusion independently (Pravec and Harris, ing, detecting, and studying binaries, triples, and pairs, each 2007). Both radar and lightcurve observations revealed that, with its strengths and weaknesses. This section describes far from being rare, binary asteroids are relatively common recent results and illustrates the complementarity of the ob- (Pravec et al., 1999, 2006; Margot et al., 2002). By the time servational techniques that characterize individual asteroid the Asteroids III review chapter was written, a more decisive systems and entire populations. title (“Asteroids do have satellites”) had become appropriate (Merline et al., 2002c). This review focuses on the develop- 2.1. Radar Observations of Near-Earth ments that followed the publication of Asteroids III. Asteroid Systems 1.3. Terminology Radar has proven to be a powerful method of detect- ing secondaries to NEAs, enabling the discovery (as of Two- and three-component asteroids that are gravitation- September 2014) of the satellites in 71% of the 49 known ally bound will be referred to as binary asteroids (or bina- multiple-component NEA systems, including 33 of 47 ries) and triple asteroids (or triples), respectively. (Triple is binaries and both undisputed triple systems. Of the 14 favored over the more directly analogous terms trinary and binary NEAs discovered via optical lightcurve techniques, ternary because of long-established usage in astronomy.) 6 have been confirmed with follow-up radar observations Asteroid pairs denote asteroid components that are geneti- during later apparitions. Overall, radar detections suggest cally related but not gravitationally bound. Paired binaries that about one in six NEAs larger than 200 m in diameter or paired triples are asteroid pairs where the larger asteroid are multiple-asteroid systems (Margot et al., 2002; Taylor is itself a binary or triple asteroid. The larger component et al., 2012a), although 200 m is not a sharp cutoff. Three in binaries, triples, and pairs is referred to as the primary binary NEA systems identified by radar have primary component or primary. The smaller component in binaries components with suggested diameters of 120 m to 180 m: is referred to as the secondary component or secondary. 2003 SS84 (Nolan et al., 2003), (363599) 2004 FG11 (Tay- There has been some confusion in the literature about lor et al., 2012c), and 1994 CJ1 (Taylor et al., 2014). For the meaning of the word “asynchronous.” Here, we adopt comparison, the largest primaries of binary NEAs imaged the terminology proposed by Margot (2010) and later with radar — (5143) Heracles (Taylor et al., 2012b), the implemented by Jacobson and Scheeres (2011b) and Fang possible triple (276049) 2002 CE26 (Shepard et al., 2006), and Margot (2012c). Binaries with an absence of spin-orbit and (285263) 1998 QE2 (Springmann et al., 2014) — are synchronism are called asynchronous binaries. Binaries with more than an order of magnitude larger at >3 km in diam- a secondary spin period synchronized to the mutual orbit eter. It is likely that ~8-km-diameter (1866) Sisyphus has a period are called synchronous binaries. Binaries with both secondary based on analysis of frequency-only observations primary and secondary spin periods synchronized to the obtained on four separate dates in 1985 (S. Ostro, personal mutual orbit period are called doubly synchronous binaries. communication, 2001). If generalization to systems with more than one satellite is Radar observations can be used to detect asteroid satel- needed, we affix the terms synchronous and asynchronous lites because of the ability to resolve the components of the to the satellites being considered. system both spatially (along the observer’s line of sight) It is useful to present results for small and large asteroids. and in terms of frequency (due to Doppler shifts from the We place an approximate dividing line at the size at which rotational and orbital line-of-sight velocities), resulting