What is a planet?

Eris and its moon Dysnomia (Artist's impression)
Eris and its moon Dysnomia (Artist’s impression). Image Credit: NASA, ESA, and A. Schaller (for STScI)

Recently it was revealed that Eris – that pesky worldlet that precipitated the removal of Pluto’s status as a planet – may actually be smaller than Pluto after all, albeit only by a couple of kilometres in radius. Naturally, this has got some folks thinking about the definition of a planet again.

The way that Eris’ diameter was refined was pretty clever – astronomers realised that the distant dwarf planet would be passing in front of a very faint star when viewed from certain places on Earth. They weren’t sure exactly where the occultation path would be except that it’d be visible from South America at least. As it happens, it was observed by astronomers in Chile, and at least three groups saw the light from the star blink out as Eris passed in front of it. By timing how long the occultation took, they could figure out an upper limit to Eris’ diameter, and it turns out that Eris can’t be larger than in 2340 km diameter. Pluto, by comparison, is 2344 km in diameter (give or take 20 km).

So it seems that at best Eris is actually similar in size to Pluto, and is more likely to actually be smaller. So that’s good news for Pluto, which now appears to have regained its crown as the largest object in the Kuiper Belt… but is it enough to make it a planet again?

Well… no.

When Pluto was “demoted” I have to admit I was a little put out by it, but I rapidly realised that it had to be done. What people may not recall is that Pluto has always been considered “the odd one out” since its discovery in 1930 – of the (then) nine planets, we had four terrestrial planets in the inner solar system, four gas giants in the outer solar system… and then Pluto, which was smaller even than any of the terrestrial planets. This was made worse when we started to discover other objects out there beyond Neptune – the first was 1992 QB1, a small body about 100-200 km in diameter orbiting between 40 and 45 AU (further than Pluto) that was discovered in 1992 by David Jewitt and Jane Luu. Since then, hundreds more objects have been discovered in orbits beyond Neptune – as a result, Pluto was no longer alone in the outermost reaches of the solar system, raising further questions about its planetary status, and people had been wondering what would happen if something was discovered in the outer system that was larger than Pluto anyway. Would that new object be a 10th planet? What if we found others, would they be Planets #11, 12, 13, or more? When Mike Brown and his team discovered Eris in 2005 (back then it was nicknamed “Xena”) and that seemed to be larger than Pluto, that lit the fuse on the powderkeg. Something clearly had to be done.

The IAU weighs in

In 2006, the International Astronomical Union passed a resolution defining a planet as:

The IAU…resolves that planets and other bodies in our Solar System be defined into three distinct categories in the following way:
(1) A planet [1] is a celestial body that (a) is in orbit around the Sun, (b) has sufficient mass for its self-gravity to overcome rigid body forces so that it assumes a hydrostatic equilibrium (nearly round) shape, and (c) has cleared the neighbourhood around its orbit.
(2) A “dwarf planet” is a celestial body that (a) is in orbit around the Sun, (b) has sufficient mass for its self-gravity to overcome rigid body forces so that it assumes a hydrostatic equilibrium (nearly round) shape [2], (c) has not cleared the neighbourhood around its orbit, and (d) is not a satellite.
(3) All other objects [3] orbiting the Sun shall be referred to collectively as “Small Solar System Bodies”.

[1] The eight planets are: Mercury, Venus, Earth, Mars, Jupiter, Saturn, Uranus, and Neptune.
[2] An IAU process will be established to assign borderline objects into either dwarf planet and other categories.
[3] These currently include most of the Solar System asteroids, most Trans-Neptunian Objects (TNOs), comets, and other small bodies.

It also resolved that planets and dwarf planets are actually “two distinct classes of object” so dwarf planets aren’t actually a sub-type of planets.

However, this didn’t make everyone happy – in fact it made some scientists very unhappy for a number of reasons (for those who think that scientists are all objective and unemotional about their work, they’re not – they can get very intense about things 🙂 ). Personally, I don’t like it much either. For one, it defines planets only in our own solar system (orbiting the Sun), and as we now know – and knew at the time – planets patently are found around other stars. Another problem is that it doesn’t clearly define “clearing the neighbourhood”. Another thing is that “dwarf planet” is misleading – what if we find an object that is as big as Earth, but still hasn’t “cleared its orbit” – is that a dwarf planet, even though it’s the size of Earth? If they’re called “dwarf planets” then why go on to state that they’re not actually planets at all? If that’s the case, then they should be called something else!

It seems like a very unsatisfying, inelegant, rushed, and poorly considered definition to me and in my opinion it hasn’t really been improved over the years. And obviously, it defines Pluto as a “dwarf planet” too, which got a lot of people up in arms who didn’t like the idea of losing Pluto as a planet (there was an addendum in 2008 that defined further “plutoids” as being dwarf planets that orbit beyond Neptune’s orbit, so Pluto is now a Plutoid, but I think that’s a pretty unnecessary and rather pointless subdivision).

A better way?

There are IMO much better definitions out there – this one from 2000 by Alan Stern & Howard Levison (hereafter referred to as S&L), and this one by Steven Soter from 2006. Stern, incidentally, is one of the most vocal opponents to the IAU definition – if I recall correctly one of his main criticisms of the IAU resolution is that it combines physical definitions (the shape) with orbital dynamics (clearing the neighbourhood), which is pretty inelegant and is a valid concern. He also happens to be the Principal Investigator on the New Horizons mission currently on its way to Pluto (which was a planet when the probe left Earth, and now isn’t), so it’s probably not surprising that he has strong feelings about this! Anyway, if you’re interested in planet definitions then definitely read those papers – they’re both fairly readable and don’t contain too much horrible maths.

Both of those in my opinion come up with more logical ways to determine what is a “proper” planet and what is not. The nice thing about the S&L paper is that (unsurprisingly, given the identity of one of the authors) it does separate out the physical parts of the definition from the dynamical parts. They first define a “planetary body” as an object that cannot (and could never) generate energy through nuclear fusion (otherwise it’d be a brown dwarf or a star) and that also is large enough that its shape is determined largely by gravity rather than other force (essentially, that its mass is high enough that its gravity forces it into a spherical shape). Then they go on from there to define what “planet” is. At this point then, Pluto is a planetary body because it’s in a bound orbit around a star (or stars) – and the classical nine planets are included as such along with the largest asteroids and KBOs.

Having done that, they then continue to subdivide planetary bodies down into different types based on whether or not they’ve cleared their neighbourhoods, and define a numerical property (Λ, the upper case of the greek letter Lambda) that can be calculated that shows how well they have ‘cleared their neighbourhood’. This is where it gets really interesting – the eight largest planets have large values of Λ that are much greater than 1, and objects like Pluto and the asteroids have very small values of there of Λ that are much less than 1. So S&L set a boundary between “uberplanets” (Λ > 1) and “unterplanets” (Λ < 1) In the S&L system, Pluto is therefore actually an “unterplanet” along with Ceres… but the key point is that it’s still a type of planetary body (remember that Pluto isn’t a planet the IAU definition, it’s a “dwarf planet” instead, and they’re explicitly not a subtype of planet according to the IAU). However, it’s interesting to note that if Pluto was moved to Mercury’s distance from the sun, it would jump above the Λ = 1 line and become an “uberplanet”.

What I like about the S&L definition is that it’s methodical and consistent. It’s also measurable too, although they do get into further subdivisions that I don’t think are really that necessary.

The Soter paper comes up with another value called the planetary discriminant, with the symbol μ (the greek letter mu). Essentially, the planetary discriminant is the mass of the object divided by the total mass of the other stuff (not including itself) in its “orbital zone”. Soter defines two bodies to be in the same orbital zone if their orbits cross, their orbital periods are non-resonant, and their semi-major axes differ by less than one order of magnitude. Again, he finds that μ is very large for the big eight classical planets, and less than 1 for objects like Pluto and Ceres. He suggests that μ=100 should be the critical value that separates a planet from a non-planet – in other words, if an object is at least 100 times more massive than the other stuff that crosses its orbit, then it’s a planet.

While that seems pretty elegant, Soter then continues to define objects based on where and how they accreted and formed, which isn’t actually something that can be unambiguously gathered by observation. As a result, I think Soter’s classification system isn’t actually as useful and clear as the S&L system. It’s also not clear what happens if one has a situation where several large objects of a similar mass are in the same orbital zone – if we found a system with say five similar earth-sized objects that were somehow in the same orbital zone, then would they not be planets (μ would be ~0.25 in that case)?


In all of these classification systems, Pluto is still reduced from being a “proper” planet in the sense that Earth or Jupiter or Neptune is. I think that’s just inevitable, because Pluto’s nature is very clearly different to that of the major planets. We know now that it’s one of the larger members of a belt of bodies that are within its “orbital zone”, much like the asteroids closer to the sun. It’s interesting to note that when the first asteroid (Ceres) was discovered in 1801, three other large asteroids were discovered at similar distances from the sun in the next three years, eventually leading astronomers to believe that there was a belt of bodies in the region between Mars and Jupiter (though they were actually classified as planets at first!). I think that had Eris and the other TNOs been discovered around 1930 along with Pluto then it would have been clear to astronomers back then that they were looking at another asteroid belt that was further from the sun, but the 60+ year gap between the discovery of Pluto and 1992 QB1 was long enough to get Pluto entrenched in our minds as “the ninth planet”.

I think it’s also clear that it’s actually really hard to come up with a consistent, universal definition for a planet, which is surprising since it’s something that a lot of people took for granted before. I think the S&L paper does the best job out of the system considered here, but even then that still has some issues, though it’s vastly superior to the IAU definition IMO (in a later post here, I may discuss my own ideas for planet definitions to further confuse matters).

Finally, I think the most important thing to realise is that Pluto is still out there merrily orbiting the sun as it always has, and I’m quite sure that it really doesn’t care what we call it ;). When New Horizons gets there in 2015 it’s going to reveal the last frontier of the solar system to us, and I’m quite sure that Pluto and Charon (and Nix and Hydra, Pluto’s recently discovered smaller moons) will have a lot of interesting surprises for us!

5 Responses to “What is a planet?”

  • Your speculation as to what astronomers in the 1930s might have made of Pluto’s standing as a planet if they had discovered a few more KBOs within a few years is an interesting one. I’m inclined to agree that it would have declined gracefully into “minor planet” status as Ceres did.

    I think it is worth observing that 1930s estimates of Pluto’s mass encouraged its misidentification as a planet. Nicholson and Mayall (December 1930. “The Probable Value of the Mass of Pluto”. Publications of the Astronomical Society of the Pacific 42 (250): 350.) estimated its mass as 1.0 M♁ on the assumption that it was Le Verrier’s Planet X. That estimate stood until 1948 and lent grew weight to Pluto’s being considered in a class with the terrestrial planets. Kuiper produced a lower estimate (0.1 M♁) in 1948, but wasn’t until 1978 (the discovery of Charon) that we realised that Pluto’s mass was as truly insignificant as it is (one fifth of one percent of the mass of Earth).

    I suggest that if astronomers in the 1930s had realised that Pluto was 1/500 of the mass of Earth (much smaller even than Earth’s moon) they would have been more willing to recognise it as in a class with Ceres rather than in a class with Earth.

  • Interesting – I found that article here. They pegged it at 1.08 earth masses (you used the symbol for Mercury there, not Earth 😉 ), apparently. I’m not sure that it was down to any assumptions about it being “Planet X” though, it looks more likely to me from that paper that the equations they used (or the numbers they were putting into them) weren’t that accurate.

    It was Percival Lowell that came up with “Planet X”, BTW – Le Verrier figured there was an unknown planet beyond the orbit of Uranus and predicted its position, and thus discovered Neptune, but he never referred to it as “Planet X”.

  • The designer of this font may have screwed up. “♁” ought to come out as an ‘Earth’ symbol. “☿” ought to be the ‘Mercury’ symbol.

    Some typographers rather irritatingly substitute a circle with a cross on top of it for the proper ‘Earth’ symbol (a quartered circle), but I am loth to use the circled-plus symbol ⊕ as a substitute for the Earth symbol because (1) in a proper font the circle doesn’t touch the ends of the cross, and (2) it is a different symbol, goddammit!

  • Oops, I guess I got confused with the symbols there (astrological symbols are not my strong point, unsurprisingly!).

  • The only ones I ever use are ☉ and ♁, the ‘Earth’ and ‘Sun’ symbols. I find them convenient as subscripts to denote the mass and diameter of Earth, the mass and luminosity of the Sun, which are unfortunately unavoidable as units.

    ☿, ♃, and ♄ I can’t keep straight or even tell reliably from the symbols for the astrological ‘signs’.

Comments are currently closed.