IRREGULAR SATELLITES






What are irregular satellites?


Planetary satellites in the solar system are divided into two types: Regular satellites orbit in circular, equatorial orbits close to the planet. One example of a regular satellite system is the galilean satellites of Jupiter. Irregular satellites orbit the planet much further out. Their orbits are eccentric and inclined and are subject to strong solar perturbations which primarily causes them (the orbits) to precess.


                                                      Phoebe, an irregular satellite 
                                                      of Saturn, as imaged from the 
                                                      Cassini spacecraft. 
                                                      Credit: NASA/ESA/ASI




An example of a regular satellite system

Examples of irregular satellite orbits
Where do they come from?


Unlike the regular satellites which formed with the planet, irregulars originally formed in heliocentric space (eg the asteroid belt) and were later captured in a closed orbit around the planet. Since any such capture will involve energy dissipation in one form or another, such capture must have taken place while the planets were still forming and dissipative mechanisms were still in existence (Kuiper 1956).

Therefore, their physical and orbital characteristics reflect, and can provide information on, conditions that prevailed during planet formation, especially the latter stages. In the 1970s, three main theories were put forward to explain the existence and orbital distribution of these objects:

  • Pollack, Burns & Tauber (1979) showed how objects in heliocentric orbit can become captured by gas drag within the circumplanetary gaseous envelope of the giant planets or CGE (a bloated proto-atmosphere) with subsequent fragmentation when aerodynamic stresses overcame their structural binding strength. However the same process that captures these objects will also force the fragments to spiral down to the planet very quickly (a few decades compared to a CGE lifetime of 10^5 yr). The population we see today according to this model must have been bodies ``lucky'' enough to be captured and fragmented just before the CGE dissipated.
  • Colombo & Franklin (1971) postulated a collision between an asteroid and a temporary outer satellite to explain the two irregular groups at Jupiter. However, such a collision will not create two distinct groups, one prograde and one retrograde such as the ones we see at Jupiter. Thus, at least two collisions are called for under this scenario.
  • Heppenheimer & Porco (1977) suggested a ``pull-down'' capture mechanism. In this model, objects with heliocentric orbits similar to the planet's can enter retrograde planetocentric orbits under the action of a third body (in this instance the Sun) but are nevertheless energetically unbound to the planet in a two-body sense. The mass growth of the planet decreases the two-body energy towards a bound state and permanent capture. The main historical difficulty here is that such a process will not produce prograde satellites. A variant of this scenario (libration-point capture as proposed in Heppenheimer 1975) will, but a dissipation mechanism to render the capture permanent is still required.




Why study them now?


A revolution in our understanding of these objects took place in the last 7 years. Searches conducted with large aperture telescopes equipped with large format CCDs discovered 74 new objects and revealed for the first time their general characteristics as a population (Sheppard and Jewitt, Nature, 2003; Gladman et al., Nature, 1998; 2001; Holman et al., Nature, 2004).

This can be likened to the discovery of the first 100 asteroids which enabled Hirayama to establish the existence of asteroid families (Hirayama, AJ, 1918) and Daniel Kirkwood to identify gaps in the semimajor axis distibution of asteroids (Kirkwood, Obs, 1882) which are now known to be due to the clearing effect of mean motion resonances with Jupiter.






The orbital distribution of known irregular satellites as of December 2004. The sizes of the symbols 
are log-proportional to the sizes of the satellites. The x-axis denotes the distance from the planet 
in Hill radii. One Hill radius is the theoretical limit at which a satellite ceases to orbit the planet 
and instead orbits the Sun. Y-axis denotes orbital inclination. Retrograde orbits have inclinations 
greater than 90 degrees.

In fact, one can use general features in this distribution to test the three main models of the satellites' formation:

  1. The absence of size-distance sorting, in other words the fact that small fragments occupy the same orbital distances as large fragments. If fragmentation of parent bodies took place before the CGE dissipated as Pollack et al postulate, then small, less massive fragments would have migrated inwards faster under the action of gas drag and ended up closer to the planet than larger ones. The absence of such a trend implies that fragmentation took place after the gas dissipated (Gladman et al., Nature, 2001) and the numerous faint satellites we now see in orbits similar to those of larger objects are the fragments of those collisions (Sheppard and Jewitt, Nature, 2003; Nesvorny et al, Icarus, 2003).
  2. There are less prograde than retrograde satellites as well as inclination ``gaps'' . Astakhov et al (Nature, 2003) showed that initial temporary capture within the Hill sphere would produce such a preferential distribution of objects since objects in the inclinations we see today would remain temporarily trapped for longer (a few times 10^4 yr), giving gas drag time to dissipate their orbital energy and enable permanent capture. The timescale for significant orbital evolution without fragmentation under gas drag is of the same order (Cuk and Burns, Icarus, 2004) meaning that capture of those objects is more likely to become permanent under the Pollack et al scenario. This also implies that such objects were of low initial energy with respect to the planets, ie in nearby or resonant orbits. Astakhov et al also attributed the difference in the prograde/retrograde ratio at Jupiter and Saturn to the fact that the jovian regular satellites (Callisto in particular) are more effective than its Saturnian counterpart (Titan) in clearing objects in the ``forbidden'' inclination zone around 90 deg. Orbits in that region undergo so-called Kozai librations the result being that their pericentres would enter the regular satellite region and be lost to the system, either by direct collision or by gravitational scattering.
  3. The irregular populations at all four planets, if corrected for observational bias, appear to be similar. This presents major problems for current formation scenarios for the ice giants. If Uranian and Neptunian irregulars were captured according to the Pollack et al. model, the inferred CGEs are significantly more extensive than the most current outer giant formation scenarios indicate (Boss, ApJ, 2003). It is a major question then how Uranus and Neptune ended up with these satellite populations, especially if, in addition, significant orbital migration of these planets took place (Thommes et al., Nature, 1999). A similar difficulty exists for ``pull-down capture'' as planetary mass growth needs to be both significant and fast for this mechanism to work, something that is not supported by current modelling.
Last Revised: 2005 January 10th
WWW contact: aac@arm.ac.uk