The Size of Planets, or What is a Planet?

I started thinking about this when I saw a TV show on how people have demoted Pluto from a planet to a "dwarf planet", or even a "Kuiper Belt Object". I have to say, it made me a bit mad to see Pluto treated so poorly. The more I looked into it, the less sense it made. It comes down to the question of how to define a "planet", and it turns out that there was no good definition. That is, the definition was largely traditional, meaning "those things which have historically been designated planets". Which certainly included Pluto.

With the discovery of the Kuiper Belt, Pluto was designated a Kuiper Belt Object (KBO). Later, the International Astronomical Union declared it a "dwarf planet". There is still great controversy among scientists over this simple question, showing that it is less a matter of science than definitions. So why the controversy? What is a planet? Can something be in the Kuiper Belt and still be a planet?

To consider these questions, let's first review the structure of the solar system. The solar system consists of a number of large bodies orbiting around the Sun. Copernicus showed, and later Galileo and others proved that the Earth is one of these bodies. The others that were discovered fairly early were Mercury, Venus, Mars, Jupiter and Saturn. Uranus and Neptune were added in the late 1700's, and Pluto in the early 20th century. The key features of these designated planets were that they were big enough to see, and they all orbited in more or less circular orbits directly around the Sun. Smaller objects were also known, such as as comets and asteroids. Comets were much smaller (that is, the cometary nucleus was tiny though the tails could be huge), and they followed highly elliptical orbits. Some of them had periods in the hundreds or thousands of years. With their long tails, strange orbits, these things were obviously not planets. Asteroids were similarly not considered planets, mainly because of their small size. It was discovered that most asteroids existed in or originated from a belt between the orbits of Mars and Jupiter, and theories were raised that they were either the remnants of some planet that had disintegrated long ago, or else material from the early solar nebula that had failed to condense into a planet. In either case, they were something less than planets. The largest of the known asteroids is Ceres, with a radius of a bit under 500km, so that was implicitly below the lower limit of what could be considered a planet.

Another class of bodies were obviously the moons, including our own Moon and the moons of other planets. Some of these were quite large. Some of them are not only larger than Pluto; they are larger than Mercury (e.g., Ganymede, Titan). No one disputes that Mercury is a planet. So obviously size is not the distinguishing characteristic alone. The key feature of the moons is that their motion is more complicated - in addition to orbiting the Sun, they also orbit around a larger body (the planet). For example, the Moon orbits the Earth, Europa and Ganymede orbit Jupiter, Titan orbits Saturn. But since the planets themselves orbit the Sun, so do the moons.

Moons follow a somewhat zigzag orbit around the sun.

As we can see in the figure to the right, the moon's actual path around the Sun isn't a circle or ellipse, there is also a motion towards and away from the Sun, as the moon circles the planet. This figure is greatly exaggerated to show the effect. In actual fact, with the Earth's Moon, for example, the radius of its orbit around the Earth is only around 0.00255 A.U., while the radius of its orbit around the Sun is the same as the Earth's, that is, 1 A.U. So the zigzag motion is only of an amplitude of about 0.25% of its orbit around the Sun.

The situation is further complicated by the fact that moons don't really orbit planets, but rather that both planet and moon orbit around a common center of gravity. If there is a great difference between the moon and planet's masses, then the center of gravity is close enough to the planet that we can fudge and say that the moon is orbiting the planet. This is more true for say Jupiter and Europa, than it is for Pluto and Charon, whose masses are much more similar.

We can see from this description that if you get into the nitty gritty of the actual orbit around the Sun, there is a great range of variation between orbits of moons close to the Sun versus moons of planets far away from the Sun; moons of small planets versus moons of large planets; moons that revolve very fast around the planet versus moons that have much slower orbits, etc.

And yet no one seems to have a problem calling them all moons. We have Deimos, a Martian moon which is a little chunk of rock about 10 km in size, which would barely deserve a name if it were not orbiting Mars. And we have Ganymede, a Jovian moon, which is bigger than Mercury. There is very little in common between these two. One barely has any gravity, the other has a fully differentiated iron core with a working magnetosphere. If Ganymede or Titan or Europa were circling the Sun rather than Jupiter, I doubt anyone would have a problem calling them planets. But they are moons for the same reason that Deimos is a moon, because they circle a larger body in addition to orbiting the Sun.

Definitionally, then, we may consider any object to be a moon rather than a planet if it orbits around a larger body in addition to orbiting around the Sun. Perhaps some day people might get around to saying that Deimos and Phobos are too small to be real moons, but right now the hate is directed against Pluto.

This may seem somewhat arbitrary, but definitions often are. In the sky, there are only chunks of rock or gas or ice which follow various complicated paths. Whether we call them planets or moons or asteroids or comets depends on how we define these terms. Unfortunately, "planets" have historically not been well defined. Their definition has been more of tradition, being regarded as planets historically.

What about Pluto then? Pluto has historically been a planet, but it suffers from a few problems. One is that it is the smallest planet. Another is that it happens to lie in the Kuiper Belt, which is again a somewhat arbitrarily defined region of space. The third is that there is a lot of ice on it (not water ice). Let's consider these one at a time.


Pluto has a radius of around 1200 km, compared to the next smallest planet Mercury, which has a radius of 2440 km. Obviously, it is a small planet. But size is relative. Even the Earth is tiny compared to Jupiter. How big does it have to be to be considered a planet? One criteria to differentiate planets from just lumps of rock floating around in space is that planets are big enough to have sufficient gravity to overcome the mechanical strength of the rocks of which they are made. In other words, the gravity is powerful enough to crush its mass into the shape of a sphere. Not only is Pluto at that limit, it is far beyond it. Obviously the minimum size at which an object will be spherical rather than some odd shape depends on the materials from which it is made. But objects as small as 200 km in radius, such as Saturn's moon Mimas are spherical. Pluto, with a radius of 1200 km, is far beyond that.

Unfortunately, Pluto has been largely overlooked by NASA, and no probes have reached it yet. For this reason our knowledge of Pluto is poor, compared to the other planets. However, we know that Pluto probably has a differentiated rocky core, with a thick layer of ice over it. Pluto has an atmosphere for part of its year when it approaches the Sun. As it moves further away, the atmosphere freezes and falls to the surface as snow. NASA finally launched the New Horizons probe in 2006, which is expected to reach Pluto in 2015, at which point we will know a lot more about Pluto than we do now.


Pluto is located in the Kuiper Belt. The Kuiper Belt is part of a large range of objects, collectively termed "trans-Neptune" objects. Transneptune objects are typically divided into 3 categories:

Kuiper Belt Objects (green dots) in the Solar System. The Sun is the red dot in the middle, planets are blue. The brownish-orange dots between the Sun and the Kuiper Belt are scattered objects known as Centaurs. The pink dots are Jupiter's swarm of asteroids, the Trojans and the Greeks. Image from Wikipedia.

Kuiper Belt

Roughly disk/donut shape about 30 A.U. to 55 A.U. from the Sun (bulk of the material is in the 40-48 A.U. range). Contains at least 70,000 objects that are 100 km or more in size. Pluto is the largest of these. They share many characteristics, such as composition (lots of methane and carbon dioxide ice), long and sometimes eccentric orbits, etc. Pluto, unlike the other planets has an orbit significantly inclined to the ecliptic, which is also not uncommon in KBOs.

Scattered Disk

This is a Kuiper Belt like region, but more dynamic, and scattered. Objects in this region can have much larger orbits (up to 100 A.U.) which are often highly inclined to the plane of the ecliptic. These objects were possibly originally Kuiper Belt objects that were displaced by Neptune. The largest of these is Eris, a planet-like object similar to Pluto, and even larger (1300 km radius). At 97 A.U., it is currently also the farthest known object in this belt.

Oort Cloud

This is a huge cloud of material reaching the very edge of the solar system. It is often subdivided into an inner donut shaped cloud at 2,000 A.U. - 20,000 A.U. (also known as the Hills Cloud) and an outer spherical cloud from 20,000 A.U. to 50,000 A.U. The Oort cloud is the source of the long period comets. It's been estimated that there are probably over a trillion cometary nuclei of 1.3 km or larger size in the Oort cloud. Much of the material is icy, like comets, but rocky material has also been found.

There is no question that Pluto is located in the Kuiper Belt and is therefore a Kuiper Belt object. The question is, why does that preclude it from being a planet as well? There is no reason, unless you make location part of a planet's definition.

The IAU tried to bring a quantitative aspect to this question. They went ahead and defined a planet as:

Obviously, Pluto qualifies on the first two counts. The third is the problem. The logic was, look at Ceres in the asteroid belt. It's big (not as big as Pluto though), and its orbit includes zillions of other asteroids. If it were a planet, it would have collected them all and grown bigger and then it would be a planet. But because it hasn't, it's still just an asteroid. Since Pluto also exists in the middle of a bunch of similar debris which it hasn't picked up, it can't be a real planet either.

Two graphs showing Pluto's orbit in relation to that of Neptune. The orbit is inclined to the ecliptic. From Wikipedia.

But there are problems with this approach. First, the asteroid belt and the Kuiper belt aren't really comparable. The asteroid belt is very small by comparison (see figure above) . It's located between the orbits of Mars (1.66 A.U.) and Jupiter (5.2 A.U.). Even in this little 3.5 A.U. range, most asteroids are clumped into an even thinner band. By contrast, the Kuiper Belt is 25 A.U. wide, and much of it isn't neatly within the ecliptic plane like the asteroid belt. To expect Pluto to clear this area is asking way too much. It would come down to an argument of size - if Pluto were bigger, it would have cleared more, but even if it were Jupiter sized, I seriously doubt it could have cleared the whole Kuiper Belt.

While "clearing the area" around it sounds like an objective criterion, there is no mention of how much exactly does it need to clear its area. Even Jupiter, the biggest planet hasn't fully "cleared" its orbit. Jupiter runs around with its own swarm of asteroids sharing its orbit. The "Greeks" and "Trojans" are clusters of asteroids that share Jupiter's orbit, in the L4 and L5 points 60 degrees ahead of and behind Jupiter. Most of the inner planets also have orbits that come very close to orbits of asteroids, as the dinosaurs discovered 65 million years ago, right here on Earth.

For these reasons, I find the 3rd part of the definition unsatisfactory. First, it has the feel of appeasement to it: tacking on the "clear the orbit" part to please the people who want to see Pluto disqualified. What scientific reason is offered for this criteria? None, that I can see. I could understand it if we were talking about unstable orbits, if Pluto or objects near it were flying all over the place, constantly crashing into each other. It wouldn't be a planet then; it wouldn't be much of anything - it would just be an object in the process of formation. But this is not the case. Pluto is in a stable orbit. Stuff around it is also in stable orbits. They don't crash into each other, at least, not any more than things crash into the Earth. They've been like this pretty much since the beginning of the solar system, so far as we know.

Second, there is no measure of "clearing the orbit" provided. How much orbit does it need to clear? Is it right to expect it to clear as much orbit as Jupiter does? After all, it is far smaller. What if a couple big asteroids happen to bump into each other near Jupiter? Does it cease to be a planet until it clears their debris?


The third argument is that Pluto, unlike the other planets, has an awful lot of ice covering it. This is about as stupid as things get. Why the heck should it matter what the planet is made of? We have rocky planets like Earth, we have gas giants like Jupiter. No one has a problem with one planet being made of rock and another of mainly hydrogen/helium. We have planets that are dense and heavy like Earth (density 5.5x that of water), and we have planets like Saturn that would float on water. No problem there. But Pluto, which is more than twice as dense as Saturn is somehow not a planet, because it's got too much ice.

This really burns me. I heard some guy on TV, who runs a planetarium or some such in New York, say "if we brought Pluto to the orbit of the Earth, all that ice would vaporize and it would have a tail like a comet - that's embarrassing for a planet!". Yeah, and if we stuck a tail to his rear, he'd look like a jackass. What does that prove? Pluto is very far away. It's not surprising that there is a lot of ice on it. Besides, what does the composition have to do with whether something is a planet or not? Since when did we start defining planets by composition? What are we going to do when we start discovering more extrasolar planets and find that they don't fit our notions? Will we reject them simply because they're not made to the same specs as those in our solar system?

Pluto's composition simply shows that like other Kuiper Belt objects which are very far away, it has a lot of ice on it. We already know Pluto is a Kuiper Belt object. The question is, why does that disqualify it from being a planet? The Kuiper Belt contains objects ranging in size from specks of dust to as big as Pluto. What is wrong with distinguishing between them, calling some "dust specks", others "asteroid sized" and still others "planets"? There is no reason, unless you arbitrarily place the limit for a planet at the orbit of Neptune.

Other Counterarguments

Some people say, well, if you only choose the first two rules of IAU's definition, we'll have 30+ planets, because you have to count moons too. I don't see why. We already have a good definition for a moon - something that orbits a larger body, in addition to orbiting the Sun. This is a perfectly workable definition. In fact, it has worked for centuries, and not many people are up in arms to change it.

Obviously, if we include Pluto, we should also include Eris in the Scattered Disk, Sedna (possibly in the Oort Cloud) and a few more of the larger Kuiper Belt objects. The number of planets will grow. There is nothing wrong with this. It's just progress - as we explore more of the solar system with better instruments, we find more things. Why dumb this new knowledge down by saying "no more planets".

Towards a Working Definition of a Planet

As I have argued above, the definition of "planet" has been largely traditional. There is nothing wrong with this, and there is no reason to change it on a whim. That doesn't mean that we can't further refine the definition as our knowledge grows. Typically, we have thought of planets as objects like Earth. We now know that many planets aren't like Earth, some are gas giants, some are very hot, some are very cold. Some have a surface, some don't. Through all of this, a couple of things have stayed constant - they do orbit the Sun, and they are fairly big.

Size is definitely both a traditional and necessary part of the definition. It's traditional because historically, we could only see the bigger stuff. It's necessary because obviously there is a difference between a chunk of rock you can hold in your hand, and a place you can visit, stand on, explore. The cutoff for size has to be a matter of choice. There is no law writ on the heavens that says "they must be this size or larger". But the choice is not purely arbitrary - there is some history, some knowledge behind it. Historically, Pluto was of sufficient size to be called a planet.

I agree with the IAU that it should be large enough to assume a spherical shape under its own gravity. But beyond that, I would go a bit further. The precise cutoff you choose for the size has an impact on how many planets we have in the solar system. If you cut the size limit down to 200 km radius (the smallest known size at which a body will assume a spherical shape under its own gravity, such as Saturn's moon Mimas), you're going to end up with a lot of planets. Several asteroids are bigger than that. Not to mention dozens, and potentially hundreds or thousands of transneptunian objects. This is not a good cutoff, because it would group very small rocks such as Mimas with the likes of Jupiter. There is little cognitive benefit to such a sweeping category.

If you raise the limit to 1000 km radius, the number of planets is much smaller. This brings a planet much closer to the popular concept of a relatively big place, which (if it were suitable) could be colonized, have cities, nations, etc. Our concept of measurement is very much based on our own experience, of things we see, touch and feel, first hand. With some mathematical knowledge, we can conceptually understand very large numbers, such as those used in astronomy. But this is quite abstract. For anything more intuitive, we rely on much smaller units, that we can see, grasp, pick up, etc. This is why we measure in feet or meters, weigh in kilograms or pounds, etc. Not only today, in all of human history, whether the measure was a cubit or a yard, it was something we could relate to on our own human scale. I mention this because I am trying to explain that although ANY cutoff limit for planet size would seem arbitrary, there can still be something reasonable behind it. In the case of land or surface areas, we think in terms of blocks, or cities, or nations. By those standards, an object 1000 km in radius would be large enough to have an earth-like sense of distances - neighborhoods, towns, metropolises, states and countries.

This has nothing to do with whether we actually visit a planet, far less colonize it. It's simply a consequence of our sense of size, by which most people would refuse to call a ball of rock you can hold in your hand a planet, but might well consider the term suitable for a body with a 1000 km radius. I am not arguing for the 1000 km number, I am just saying that a human choice of some number does not have to be completely arbitrary. There can easily be a reasonable explanation for it, in the same way there is an explanation for why we measure in pounds and feet - because measurements are a human tool to relate the magnitude of various things to whatever humans can inspect, by touch, feel, weight, etc.

It doesn't really matter whether the cutoff is 1000 km or 1100 km. But I can suggest two numbers for convenience: either 1000 km because it's a nice round number which is easy to remember, or the size of Pluto (1195 km radius) because it is something already well recognized as a planet. In either case, the object fits the necessary definitions of planet, having an orbit around the Sun, being large enough to be spherical, being large enough to intuitively feel like a planet to most people, being large enough that the number of planets in our solar system by this definition won't ever be too large, beyond the memorization capacity of average people.

Finally, I'd like to leave you with this cheerful thought.

I think this is largely a political issue, not a scientific one. The facts are these:

Since this is still a raging controversy, it is obvious that no good answers to these questions have been proposed. I have yet to hear why "clearing the orbit" is a NECESSARY part of being a planet. It's worth remembering that the IAU represents less than 5% of all astronomers. This issue is definitely not over.

Meanwhile, the only people who benefit from this controversy are a few high profile museum/planetarium administrators, who get trotted out for interviews with the media frequently, to make stupid jokes and assert that Pluto isn't a planet, as if their own opinions will change the minds of millions of people who are not convinced.

Update: From MSNBC - The IAU will not debate the question of Pluto's status in their 2009 meeting. Some comments from the article:

"The IAU is not Holy Mother Church, speaking ex cathedra," Mark Sykes, director of the Arizona-based Planetary Science Institute and an advocate for Pluto's planethood, said in an e-mail ... []. "The issue continues to be debated," Sykes observed. "Scientists continue to write papers where Pluto and other such objects are referred to and treated as planets, because the science being discussed (e.g., atmospheric processes, mantle convection, differentiation) are shared with objects like the Earth."

Alan Stern, a planetary scientist at the Colorado-based Southwest Research Institute and principal investigator for NASA's New Horizons mission to Pluto, turned down an invitation to speak at the IAU's Rio meeting. "I'm not there because the IAU seems to have become irrelevant," he told me today via e-mail.

I guess we have exhausted the official means at this point. Perhaps the question of Pluto's classification will be revisited when the New Horizons probe reaches the planet, and collects more information about the outer solar system. But it seems just as likely that the demotion of Pluto will have become entrenched in people's minds by that time, and new information will be interpreted in line with it.