The Minor Planet Center today issued an MPEC that listed eight more outer moons for Saturn. They are all very small, in retrograd orbits, and typically have orbital periods of two to three years. They were discovered with the eight meter Subaru telescope in Hawaii, by Jewitt and company.

Edit: As Jyril points out below, there is a ninth satellite that the MPC, for some reason, listed in a seperate posting.

Nine new moons, actually. S/2004 S 19 in the previous MPEC is also a new satellite.

That makes 56 Saturnian moons versus 63 Jovian moons. At this rate, Saturn will soon regain the planet with most satellites title. In addition, new moons of Saturn tend to be a couple of times bigger than Jupiter's (4-6 km vs. 1-2 km). Based on that, Saturn may have many more irregular satellites than Jupiter not yet detected. It is obvious that Uranus and Neptune are understudied in this respect, so there's probably many more satellites yet to be discovered.

Can anyone help me? I've forgotten how to read these orbital elements from the MPECs to find out basic information (in particular, the semimajor axis). Here's one example set of elements:


--Emily

You can find an explanation of the orbital elements here.

The 3 elements you're probably most interested are:

a -- semimajor axis in AU,
e -- orbital eccentricity and
P -- orbital period in years.

Thanks ugordan! I thought that's what "a" meant but it seemed very surprising on the face of it for the semimajor axes to be 10% of an AU or bigger. These outer planets have really huge systems! No wonder Cassini can't spot these moons.

--Emily

Well Neptune has a couple of moons that have semi-major axes at around 0.3 AU

Neptune's outer irregular satellites have very large eccentricities (around 0.5) which allow them to reach distances of almost 0.5 AU from Neptune! Furthest of them spends over 25 years making one orbit around Neptune.

Given how poorly Neptune's irregular satellites are known, it wouldn't be surprising if even more distant satellites are found. Because of the great distance from the Sun, Neptune's Hill sphere extends over 0.75 AU so there's plenty of room left.

The Hill spheres of outer solar system bodies are pretty impressive -- I worked out Pluto's once and it came to something like 700,000 kilometres, not bad for such a small body.

So what are the Hill spheres of Jupiter, Saturn, and Uranus?

--Emily

Jupiter: 53.2 million km = 0.35 AU
Saturn: 65.5 million km = 0.44 AU
Uranus: 70.1 million km = 0.47 AU
Neptune: 116 million km = 0.77 AU

In comparison,

Earth: 1.50 million km
Mercury: 221,000 km
Pluto: 7.58 million km

So it is really the distance from the Sun that matters most.

The actual equation is

r ~ a * ( m/3M ) ^ 1/3

where r is the approximate radius of the Hill sphere, a orbital distance, m the mass of the orbiting body and M the mass of the parent body (in SI units). The mass of the satellite within the Hill sphere must be negligible compared to the other masses.

Titan probably has larger Hill sphere than any other satellite. Still, the radius of its Hill sphere is mere 52,400 km. Callisto, whose orbital radius is larger, has slightly smaller Hill sphere (50,100 km). Jupiter's greater mass and smaller orbital distance makes Ganymede's Hill sphere much smaller, only 31,700 km in radius. Add various disturbances and it becomes clear that no large satellite can have own moons in our Solar System.

I am not smart enough to set up a simulation to see if this is possible, but there may be a congenital (if that is the right word) reason why we don't see satellites of moons.

Upon formation, let's say a moon like Dione, could be expected to be rotating non-synchronously with the host planet. Seems I have seen figures around 8 to 10 hours. Over time, the moon will tide lock to the planet, but during the period when it is not, any satellites of the moon will be subject to tidal effects from the moon, and will be accelerated in their orbit around the satellite, just as our moon is being accelerated today by tidal effects.

The orbital evolution of such satellites should be interesting. As they 'loft' away from their host moon, the effect weakens and the satellite may be expelled ever so gently from the 'Hill sphere' of the moon.

What then?

Well the little satellite is still in orbit around the host panet, and not moving away very quickly from the original 'host' moon.

I feel such objects are quite likely to wind up in a Trojan relationship with the moon.

Objects such as Pandora and Calypso may be old satellites of moons, conveniently left for us to explore, time capsules of the materials swept up by the original host moon, and never processed thermally or tectonically.


In the case of a larger moon such as Titan, we might expect the 'spin off' process to be rather more energetic and the cast off satellite might wind up in a higher orbit around the host planet, rather than in a Trojan relationship.

Hyperion might be such a body. It's crater devestated surface recording the influx of debris coalescing into Titan, back in the time when Saturn's moon's (and their satellites) were forming.

Uh, SI units are good, but are they required in this equation (as long as consistent units are used)?

Umm... no, if you don't want an SI result.

Provided you measure the masses using the same units and are consistent for those two the result will be in whatever standard you choose for the orbital distance. So we could measure the masses in MegaElephants and use the recently discovered GigaCellphone for orbital distance and it would still work fine yielding results that are quite neat since a GigaCellphone is around 10 million km.

It would be interesting, though, if some of these outer-irregular satellites of the giant planets were of the double-asteroid type, i.e. two similarly sized bodies in close orbit.

Cassini might be able to resolve this sort of thing (albeit just barely).

If the irregular satellites originate from captured and disrupted asteroids, it wouldn't be surprising if some of them are close (or contact) binaries.

Aren't they way too far to be detectable by Cassini?

Seems like Iapetus would have a uniquely large Hill sphere for a moon . . . .

Hi,

using the Hill sphere formula given in this chat, it becomes indeed obvious that the larger satellites have Hill spheres about 1-2 orders of magnitudes smaller than the planets:
Ganymede 0.00021 AU
Callisto: 0.00033 AU
Titan: 0.00035 AU
Iapetus: 0.00025 AU

Mercury, in comparison, has 0.0015 AU.

However, tiny asteroid (243) Ida has only 0.00005 AU, or 8200 km, but owns a satellite. How does this work?

T.

Simple... the satellite has to stay closer (in practice, substantially closer) than 8200 km from Ida. Otherwise the sun will snatch it away.

Staying within the Hill sphere, though, is no guarantee of a stable orbit when we're talking about satellites of satellites. The Galilean satellites all have Hill spheres, but that doesn't mean they can have moons of their own. For example, anything going into orbit around Callisto (within Callisto's Hill sphere) would probably be able to stay in that orbit for a year or two. Eventually, though, the huge gravitational perturbations from Jupiter would destabilize it. It would most likely end up crashing into Callisto.

This is one of the problems that has to be considered for the Europa Orbiter; it will need to constantly stabilize itself, and that's going to eat up fuel.

Dactyl's orbit is poorly constrained, but it is roughly 100 km.