On Binary Asteroids

By David Polishook – Tel-Aviv University

The mere existence of binary asteroids was in doubt until 1993 when the spaceship Galileo obtained a close-up image of (243) Ida and Dactyl. Since then, binary asteroids have been found using a range of techniques: i) direct imaging using adaptive optics and large-aperture telescopes or the HST, can discover big Main Belt (MB) binaries with a large separation between the components, such as (45) Eugenia and (90) Antiope; ii) radar observations can detect much smaller binaries and satellites but this can be done only for near-Earth asteroids such as 1999 KW₄ and 2000 DP₁₀₇; iii) time-series photometry can exhibit two additive periods in the asteroid lightcurve that reflect the rotational period of the primary body and the orbital period of the secondary fragment. (See review paper by Merline et al. 2002) This technique is best used for binaries with asynchronous periods where the different frequencies are easily noticed. Eclipses and occultations can also sometimes be seen in such lightcurves and can reveal binarity for synchronous asteroids even when the orbital and rotational periods are equal, such as for Pluto and Charon or for the NEA (69230) Hermes. In these cases, very steep V-shaped minima represent eclipse events while U-shaped maxima are produced by both components.

Up to June 2006, 96 binaries are known among the small bodies in the Solar System and one-third of them are NEAs – this is a very high fraction compared to the ~1% ratio of all known NEAs to all known MBAs. This anomaly can be solved by either of two explanations: a selection effect, with present-day techniques unable to detect binary MBAs with the characteristics of binary NEAs (e.g., small size, small separation, etc.), or a specific formation mechanism with the creation and evolution of binary asteroids depending on their Solar System location. Some ideas suggested for this dependence are collisions in the Main Belt, which can disrupt an asteroid into two or more fragments. The problem with such a scenario is that the disrupted fragment must not gain too much energy in order to remain in orbit around the main body. Another option, more likely to happen in the denser Main Belt, is the gravitational capture of a fragment from an asteroid by another. This scenario requires the presence of a third body in order to conserve angular momentum and energy. Bottke and Melosh (1996) suggested that when an NEA passes near a terrestrial planet it can be disrupted by the planet’s tidal force. A fourth mechanism depends on the efficiency of re-emitting Sunlight. This effect, called YORP, is a torque on a rotating irregular body due to the absorption and subsequent re-emission of sunlight. When YORP spins up the asteroid (though spin down is another possibility) the body can pass the critical rotation speed threshold of a "rubble pile" structure and be disrupted into a binary (Bottke et al., 2002).

Pravec et al. (2006) recently summarized some important properties of asynchronous binary NEAs. The diameter ratios of secondaries to primaries (Ds/Dp) are almost always smaller than 0.5 when the primary diameter is less than two km. The secondaries are generally more elongated than the primaries, and orbit them in an almost circular orbit (e<0.1) with a semi-major axis that is 1.5-3 times the primary diameter. The secondaries’ rotation periods are usually synchronized with the orbital period, both with a median value of ~16 hours with a distribution tail reaching ~11 hours on one side and at ~40 hours on the other. The distribution of the primaries’ rotation period is more localized and ranges between 2.2 to 2.8 hours. This high spin rate is very close to the critical rate at which rubble pile asteroids disrupt, suggesting that the origin of such systems involved a fast rotating progenitor.

Known MBA and TNO binaries with asynchronous rotation have different properties, due to selection effects. Ds/Dp ranges from one (equal-sized components) to small values (tiny satellites). The distances between the primary and the secondary are larger than those of NEAs (otherwise they would not have been detected). The rotation periods of primaries are longer (several to a few tens of hours) and so are the orbital periods. The synchronous binaries’ properties are similar to the asynchronous Main Belt binaries, including one NEA binary with synchronous motion, (69230) Hermes, that has a rotation period of 13.89 hours.