Saturn
Saturated with satellites
Douglas P. Hamilton

Advances in detector technology have led to a rash of newly discovered moons
around the giant planets. Saturn currently has the most known satellites,
but for how long?

Parents of small children are expected to know the answers to questions,
like "Which planets in the Solar System have the most moons?" Due to a
surge of discoveries of distant planetary companions over the past five
years, beleaguered parents everywhere would be hard pressed to answer this
question correctly. Today, the planets with the most known natural
satellites are Saturn with 30 moons, Jupiter with 27, Uranus with 21 and
Neptune with 8. But this top-four list was ordered differently last year,
and still differently the year before that.  Only during the Voyager flybys
of the 1980s was the discovery of new moons as rapid. On page XXX of
this issue1 an international group of astronomers led by Brett Gladman
reports the discovery of a dozen new saturnian satellites using
highly-successful ground-based observations. 

The new discoveries have not come from telescopes in space, or from
the largest telescopes on Earth, but from medium-sized instruments
(3-5 m diameter), which can efficiently scan large regions of
space. Ground-based surveys complement spacecraft imaging because they
are sensitive to small moons far from a planet, but are insensitive to
nearby satellites owing to light scattered from the planet.
Conversely, imaging by a passing spacecraft can easily find small
moonlets close to their parent planet, but cannot be used to
efficiently search for distant satellites.

Rapid improvements in detector technology make systematic ground-based
searches much more tractable than even a decade ago; today digital CCD
arrays with 10,000 x 10,000 pixels can cover a quarter-square-degree patch
of sky (roughly equivalent to the area of the full Moon) in a single
exposure. Equally important are improvements in computers, which allow
data to be analysed in real time, and the development of sophisticated
algorithms that can automatically detect and flag any faint, slowly moving
objects.

Despite these improvements, the observational task is still daunting. The
region of space that needs to be covered in a thorough search is a planet's
'Hill sphere', within which satellites can orbit stably. The projected area
on the sky covered by the Hill sphere is related to the planet's mass: for
Jupiter, Saturn, Uranus and Neptune, these areas are a respectable 48, 22,
6, and 7 square degrees, respectively. Even with a
quarter-square-degree field of view there is a lot of space to be covered
and, partially for this reason, Gladman's group focused first on more
distant Uranus and Neptune2,3, before proceeding to Saturn1. 
  To cover Saturn's entire Hill sphere Gladman et al. planned a
careful systematic campaign involving multiple telescopes and
coordinated follow-up observations. 

Remarkably, this latest Saturn survey is sensitive to objects with a
brightness of 23rd magnitude1. On the magnitude scale, brighter
objects have lower numbers. So this means that, with reasonable
assumptions for the reflectivities of the new moons, they have a complete
inventory of objects with radii larger than about 4 km circling Saturn (Fig.
1). The Gladman group estimates that their previous studies of Uranus and
Neptune are nearly complete, covering about 90% of the Hill sphere down to a
similar magnitude limit3. While a complete survey of Jupiter's environs has
not yet been completed, an impressive step in this direction was made
earlier this year by Sheppard et al.4 who added ten additional jovian moons
to the single one detected a year ago6.

Because they shine by reflecting sunlight, 23rd magnitude satellites
around Jupiter, Saturn, Uranus and Neptune have very different radii:
approximately 1, 4, 16, and 36 km, respectively. This makes it
difficult to compare different satellite populations because the
observations discriminate against more distant planets.  For example,
all of the new saturnian satellites are probably too small to have
been detected had they orbited Uranus or Neptune instead. Similarly, a
group of three very faint objects that Gladman et al.1 spotted moving
near Saturn but then lost (when angling for moons, unlike when angling
for fish, it is the small ones that get away) would have been more
easily tracked had they circled Jupiter instead.

Nonetheless, complete surveys of satellite populations around
individual planets can reveal clues to the origin and early history of
the planets.  Indeed, Gladman et al. observe that the saturnian
satellites are grouped into distinct families (Fig. 1) with similar
orbital properties, just as the main belt asteroids and distant jovian
satellites are.  Families are thought to be formed either during
cratering collisions that break fragments off parent bodies, or during
more violent catastrophic collisions that completely destroy the
parent.

The current orbits of the satellites (Fig. 1) gives some idea of these
relationships, but similarities in their long-term orbital properties
(particularly the average tilt of the orbital plane) give a clearer
indication.  The existence of satellite families implies there are
substantial unseen populations of smaller bodies around all of the
giant planets, because collisions that chip 1-km fragments off 10-km
moons occur far more frequently than ones that produce 10-km fragments
from 100-km moons. The three little ones that got away at Saturn are
just the tip of the iceberg.

If each satellite family was derived from a single parent moon, where
did the parents originate? It has long been believed that these
distant satellites were captured early in the history of the Solar
system when a large disk of gas and dust, the Solar nebula, still
surrounded the Sun.  There are several possible capture mechanisms:
involving interactions with the planet's own gaseous nebula7, other
objects within the planet's Hill sphere8, and the expansion of the
Hill sphere as the planet grows in size9.  Gas drag from the Solar
nebula probably facilitated all of these processes by slowing down the
velocities of isolated objects.  But the satellite capture process is
not well understood and remains an outstanding problem in planetary
science.

The next challenge is a survey of Jupiter's entire Hill sphere down to
23rd magnitude. Such a survey is complicated by the large search area
(more than for the other three giant planets combined), the
potentially huge number of detectable 1-km-sized moonlets, the more
rapid orbital motion of jovian satellites, and the greater likelihood
for confusion with main-belt asteroids. Nonetheless, a multitude of
jovian satellites probably lurk undetected in the vast region of space
controlled by this giant planet. It will probably not be long before
massive Jupiter regains the title of the Solar System's most-mooned
planet.

Douglas P. Hamilton is an Associate Professor in the Department of
Astronomy, University of Maryland, College Park, Maryland 20742, USA.
e-mail: hamilton@astro.umd.edu

1. Gladman, B. J. et al. Nature (this issue).
2. Gladman, B. J. et al. Nature 392, 897-899 (1998). 
3. Gladman, B. J. et al. Icarus 147, 320-324 (2000). 
4. Sheppard, S. S. et al. IAU Circular 7555 (2001). 
5. Scotti, J. V. et al. IAU Circular 7460 (2000). 
6. Heppenheimer, T. A. & Porco, C. C. Icarus 30, 385-401 (1977). 
7. Pollack, J. B., Burns, J. A. & Tauber, M. E. Icarus 37, 587-611 (1979). 
8. Colombo, G. & Franklin, F. A. Icarus 30, 385-401 (1971).

Figure 1 The current orbits of the outer saturnian satellites. The radius of
the Hill sphere is about 65 million km, which is equal to 1,100 Saturn
radii, or 0.43 AU (1AU is the mean Earth-Sun distance). Saturn is at the
center and the white objects on inner nearly-circular prograde orbits are
the classical satellites, Titan, Hyperion and Iapetus. Prograde objects
circle Saturn in the same way that its classical satellites do, whereas
retrograde objects orbit in the opposite direction. The 12 new satellites
discovered by Gladman et al.1 can be grouped into families according to
their orbital properties: cyan and green objects are the new prograde groups
whereas magenta and red dots (including previously-known Phoebe) are all
retrograde satellites. Animations of these and all other Solar System
satellites can be found at http://janus.astro.umd.edu/solarsystem.html.
(Figure provided by M. Asbury.)