Tuesday, April 12, 2011

They’re all planets - part 1

As I wrote last in my last post, I take exception to the 2006 International Astronomical Union (IAU) definition of “planet.” It’s not because I harbor some kind of irrational stalker obsession with Pluto, which got demoted from “planet” to “dwarf planet” as a result of the new definition. I just think the IAU got it wrong.

The IAU defines “planet” as any object in our Solar System that: (1) orbits the Sun, (2) has “cleared its orbital neighborhood” of smaller objects, and (3) is massive enough to have achieved hydrostatic equilibrium (a roughly spherical shape). The IAU calls an object which meet only the first and third criteria a “dwarf planet,” which it insists is a distinct category, not a sub-category of “planet.”

There is a good reason that the definition specifically refers only to objects in our own Solar System. Make no mistake: the IAU’s definition is a political one, not a scientific one. It was written to ensure that we would have only eight planets and not hundreds as we continued to discover additional objects approximately the size of Pluto in the outer reaches of the Solar System. It’s true there is a difference between the larger planets and objects like Pluto, but there are better ways to classify those differences. If they wanted to draw the line for purely aesthetic reasons and to simplify elementary school textbooks, they should have just done so. Instead, we were given an official definition that amounts to sloppy pseudo-science. When examined closely, it generates hypothetical exceptions and contradictions at every turn.

Is it only a gun when it’s smoking?

I especially disagree with the first two points listed, which define “planet” based on circumstantial criteria rather than intrinsic ones. Limiting “planet” to only objects that orbit our own Sun is ludicrous. (And, I must add on behalf of aliens everywhere, racist.) But even if you make the first criterion broad enough to include exoplanets (that is, planets orbiting stars other than our own), you are still leaving out a lot. Consider that many computer simulations of how our Solar System took its present shape propose a hypothetical fifth giant planet (the size of Neptune and Uranus), which eventually moved into a hyperbolic orbit and was ejected from the system by gravitational interactions with the other giant planets. If such a “rogue” object exists, it would no longer be orbiting a star but moving independently in interstellar space. What exactly is this object – 14 times the mass of Earth – if not a planet?

Also, consider that there are two moons in our own Solar System (Ganymede, which orbits Jupiter, and Titan, which orbits Saturn) that are larger in diameter than our smallest “planet” Mercury. It is convenient for the IAU that there are none even larger. And more convenient still that we do not live on one. The largest gas giant planets could easily sustain Earth-sized “moons” in orbit around them. Habitable moons like Endor (Star Wars) and Pandora (Avatar) are not far-fetched: they surely exist in the real world just as they do in science fiction. (Maybe even some with adorable teddy bears or giant blue aliens.) Moons were often referred to in the past as “secondary planets,” and even today astronomers like former NASA Associate Administrator Alan Stern call them “satellite planets.” If they would be considered planets orbiting the Sun, are they really that much different because they orbit a planet? Titan even has a substantial atmosphere, something Mercury lacks. Who is to say which of the two is more planet-like?

No planet is an island

The goofiest part of the IAU definition of “planet” is this business about ”clearing one’s neighborhood.” In one sense, there is little doubt that Mercury, Venus Earth, Mars, Jupiter, Saturn, Uranus and Neptune are the dominant objects in their orbital zones, and that this is true to a degree far beyond that of other Solar System objects. But none of them has “cleared” its orbital path completely: this is simply sloppy language. The orbits of the inner planets, for example, are all constantly crossed by asteroids thrown their way from the Main Asteroid Belt by the gravitational perturbations of Jupiter. At least four planets – Earth, Mars, Jupiter and Neptune – share their orbits with “trojan asteroids,” which maintain stable orbits by staying in those planets’ Lagrangian points – gravitational “blind spots” 60° ahead of or behind the dominant planet in its orbit. Mercury is locked in a 3:2 spin-orbit resonance with the Sun (it rotates on its axis exactly three times for every two revolutions around the Sun), so in a sense one can say that it is the Sun, not Mercury, that dominates there. Likewise, Saturn appears to have stabilized in its current orbit by achieving a near orbital resonance with Jupiter (it revolves around the Sun once for every two revolutions Jupiter makes). Neptune’s orbit is crossed by Pluto and many of its Kuiper Belt kin, although they are only able to do so because they are in orbital resonance with the giant planet (revolving exactly three times for every two times Neptune does).

So you see that the issue is not a simple one. And again, even if we use a very relaxed definition of “clearing one’s neighborhood,” it remains partly the luck of the draw. Pluto doesn’t dominate its orbit, but an object Pluto’s size would probably be massive enough to dominate Venus’ orbit. Likewise, move Mercury out past the Kuiper Belt to about 60 AU (that is, 60 times the average distance from the Earth to the Sun) and it would no longer get the job done. And that’s only about a tenth of the average distance of Sedna, the most distant dwarf planet we’ve discovered so far.


But the Solar System extends much further than that. Move the largest terrestrial planet in the Solar System, Earth, out some 3,000 times its current distance and it too loses dominance. (Really, at that distance, it would take so long to revolve around the sun that “orbital dominance” begins to lose all meaning anyway.) Granted, this is an extreme case, but considering that some astronomers have theorized that Earth-sized or larger bodies may exist at even greater distances, it is worth noting. If we were to find an object larger than the Earth in such an orbit, would we really call it a “dwarf” planet?

The size of the shoe does not determine the size of the foot

Here is a hypothetical scenario. (By the way, considering that there are upwards of a septillion star systems in the observable universe – that’s a one with 24 zeroes after it: 1,000,000,000,000,000,000,000,000 – there’s a good chance that any scenario that could happen has happened somewhere, so this isn’t just speculation for speculation’s sake.) In our quest to discover exoplanets, we have discovered many “hot Jupiters” – large gas giants that formed in the outer regions of the system and then migrated inward to a near-star orbit. Indeed, it appears that in our own system Jupiter and Saturn may have formed further out before stabilizing at their current orbits. Suppose for a moment that Jupiter had not stopped at 5.2 AU, and had instead migrated all the way to the Earth’s orbit at 1 AU. In the process, let us suppose that four things happened: 1) the influence of Jupiter’s gravity caused Mercury and Mars to move together and enter a co-orbital relationship where they trade orbits periodically (like Saturn’s moons Epimetheus and Janus); 2) Venus was flung into the outer solar system and ended up in an eccentric orbit that crossed and was resonant with Saturn’s (similar to Pluto’s relationship with Neptune); 3) Earth maintained its current orbit by staying in one of Jupiter’s stable Lagrangian points; and 4) Jupiter’s four largest moons were spun away and established stable, dominant orbits of their own. According to the IAU, in this scenario Venus and Earth would absolutely not be considered planets, Mercury and Mars would almost certainly not be considered planets, and Jupiter and Saturn would be very questionable at best because each would share its orbit with an object of significant size (Earth and Venus, respectively). Meanwhile, Ganymede, Callisto, Io and Europa would all be planets, despite all being a fraction of the mass of the other objects mentioned.


I hope that this last example shows just how problematic it is to define “planetary status” by circumstantial rather than inherent criteria. This part of the IAU’s definition of “planet” reminds me of the barycenter definition of “double planet” that I argued against last time. Although celestial objects can and should be categorized according to their orbital properties, this is not what makes them planets. To say it does flies in the face of common sense. What makes a planet a planet? Size and size alone. Any sufficiently large (“too big to be a space station”) substellar object is a planet, no matter where it is or what is around it.

Are you a dentist or a human being?

At the risk of dwelling too much on this point, I want to make it clear that an object’s orbital characteristics (what it orbits and how) are indeed important, they are just fundamentally different than the object’s inherent properties. It’s like comparing a creature’s behavioral characteristics with its physiological ones. Both are important, and both can be used to classify the creature, but only the latter describes what the creature is at the most basic level. For example, whether an organism is warm or cold blooded (physiological) is important in determining whether or not it is a mammal. Whether or not it is carnivorous (behavioral) is not: carnivorous creatures can be mammals, reptiles, fish, insects, even plants. That is quite a range.

To put it another way, your carnivorous dog can subsist on a diet of corn meal and vegetable protein. It cannot survive breathing water.

There are lots of ways to describe a creature’s behavior, some very basic (aggressive, nocturnal, predatory) and some quite complex (dentist, shopaholic, overlord). Likewise, the orbital behavior of planetary objects can be described in different ways. Some of these are easy to quantify with simple numbers: eccentricity (how elongated the orbit is), inclination (how tilted the orbit is in relation to the equatorial plane of the system), period (how long it takes for the object to complete one revolution), etc. Other behaviors are more complex, such as object’s relationship to other objects in its orbital path. While I do not agree with the IAU’s decision to use these relationships to define “planet,” I do agree that it is worth classifying objects according to this behavior. Here is a system of terminology for doing just that:

Orphan” describes any substellar object that does not orbit another body: for example, “rogue” objects that have been ejected from their star systems, or “free” objects that developed in dust clouds that were not massive enough to become a star. We have witnessed comets being ejected from our own Solar System, and there are undoubtedly other free-floating objects of every conceivable size zipping through the vast emptiness of interstellar space.

A “dominant” object is one that controls the orbital characteristics of virtually all of the other objects in its orbital path around a star. This means that the objects that share or cross its path are much smaller objects governed by its gravitational influence. There are boring mathematical ways of trying to determine this (for example, the Stern-Levison parameter Λ or Steven Soter’s planetary discriminant μ), but the point is that there is a huge gap between dominant and non-dominant objects, so at least in our Solar System it is pretty obvious.

Classical” refers to objects with reasonably stable, low-eccentricity (<0.24) orbits around a star, which are neither dominant nor directly under the gravitational control of a dominant planet. In our own Solar System, this includes most Main Belt asteroids and classical Kuiper Belt objects (known as “cubewanos”) that orbit completely beyond the influence of Neptune.

Unstable” objects are those whose eccentric orbits cross areas within the control of a dominant planet, and whose orbital paths are thus more likely to be altered over time. This includes comets, centaurs and most planet-crossing asteroids.

Detached” objects are similar to unstable ones in that they have high eccentricity orbits (>0.24), but their orbits are either greatly inclined or distant enough that their closest approach still leaves them free from the gravitational influence of the dominant planets, making them relatively stable like classical objects. Dwarf planet Sedna is a good example of a distant detached object: its highly eccentric orbit (0.8527) takes it as far away from the Sun as 961 AU and only as close as 76 AU, well out of the influence of the nearest dominant planet (Neptune at 30 AU).

Resonant” objects are those in a synchronized orbital relationship with a larger object. For example, plutinos (a class of Kuiper Belt Objects that includes Pluto) are in 2:3 resonance with Neptune: for every three revolutions Neptune makes around the Sun, plutinos make exactly two. This allows plutinos to have relatively stable orbits even though they cross Neptune’s path. Other resonant groups exist within the Kuiper Belt, as well as in the more distant, more dynamic Scattered Disc region. Resonant objects are less stable further in due to perturbations from multiple major planets, but they do exist. Moons may also be resonant with larger moons, as in the case of Saturn’s moon Hyperion, which is in 3:4 resonance with the larger moon Titan.

Co-orbital” describes a special kind of 1:1 resonant relationship, where multiple objects share very similar orbits and complete revolutions in the same amount of time. The orbits need not be exactly the same for a co-orbital relationship to exist, and in fact the dynamics of their gravitational interaction can get quite complicated. While “trojan” and “satellite” relationships are technically co-orbital (see below), the term is most often used to refer to relationships where two objects exchange gravitational energy each time they approach in such a way that they never collide. The most common is what is also referred to as a “horseshoe orbit,” because of the path each object appears to travel from the perspective of the other. Several asteroids are in horseshoe orbits around the Earth (including 3753 Cruithne, 2002AA29, 2003YN107 and 2010SO16), and two of Saturn’s moons, Janus and Epimetheus, are co-orbital with each other in this way. In these cases, whenever the two objects approach each other, the object with a closer (and thus faster) orbit around the Sun is shifted further out while the slower object is shifted to a closer orbit. (This shifting is negligible for an object much more massive than its co-orbital companion, such as the Earth.) The best analogy is two cars going the same speed on a circular racetrack: the inner lane is slightly shorter, and so the car in that lane should complete the circuit slightly faster than a car in an outer lane. However, as it approaches the other car, it moves to an outer lane, while the other car moves further in where the track is shorter and thus starts to pull away again. When it has almost caught up to the other car, they again switch lanes and thus neither car ever passes the other. Occasionally, objects in horseshoe orbits will move into what is known as a “quasi-satellite” orbit for a limited time, where instead of oscillating back and forth, they appear to remain in very slow orbit around the other object. In fact, the revolution of one around the other is an illusion, as both are actually in orbit around a third body. For example, beginning in 1996, it appeared that Earth had a second moon, as 2003YN107 slow spiraled around it, completing one revolution of the Earth per year at a distance of about 15 million km. But the tiny asteroid (about 10-30 meters in diameter) was actually orbiting the Sun the entire time, and in 2006 it returned to its normal horseshoe orbit.

Trojan” describes a special kind of co-orbital relationship where a smaller object orbits entirely within a larger object’s L4 or L5 Lagrange points, areas approximately 60° ahead of or behind the larger body in its orbit. This is the most stable kind of co-orbital relationship, becoming unstable only when the trojan exceeds about a tenth of the mass of the larger object. At least four planets (Jupiter, Neptune, Mars and now Earth) have confirmed trojans, and two of Saturn’s moons are accompanied in their orbits by trojan moons (Tethys is preceded by Telesto and followed by Calypso, while Dione is bookended by Helene and Polydeuces). Other planets and moons may also have trojans or dust clouds at these points. Because of the gravitational dynamics of this relationship, trojans are not always exactly the same distance from the larger planet, but seem to trace a “tadpole” shaped orbit around the Lagrange point.

A “satellite” or “moon” is an object which does not orbit a star directly, but is in a stable orbit around another object which does. “Regular satellites” are close-orbiting moons with low eccentricities. These generally have low-inclination, prograde orbits and are tidally locked (that is, one side always faces the object they orbit). Most regular satellites were formed in situ as the planet they orbit was forming, although there are exceptions such as Neptune’s largest moon Triton (which was an independent object captured by Neptune’s gravity when it passed too close) and the Earth’s only moon Luna (which scientists believe was formed when a larger object collided with Earth). “Irregular satellites,” by contrast, are more distant moons that are always captured objects. Irregular satellites tend to have more eccentric and inclined orbits, and are predominantly retrograde (that is, they revolve backwards compared to the rotation of the object they are orbiting). Satellites may also be called “resonant,” “co-orbital” or “trojan,” per the above definitions. Moons of objects too small to be considered planets themselves are often called “moonlets,” as are moons that have not cleared their orbits of debris (i.e., moons that are imbedded in a planet's rings).


A “sub-satellite” is a moon of a moon. These are theoretically possible, although one has never been observed.

It doesn’t have to be shaped like a hammer to pound a nail

Getting back to the IAU definition, the third criterion comes closest: insisting that a planet be massive enough to achieve hydrostatic equilibrium rightly puts the emphasis on an intrinsic property. Namely: “This is a freaking huge hunk of stuff in space!” Big enough that its own gravity crushes it in all directions until it is almost completely spherical. However, while this definition is much better, I still have a few nitpicky objections. First, at the present time it is hard to know which of the new, far-flung (dwarf) planet candidates are actually round, because even our most advanced telescopes can’t make out that level of detail. We know from observation of closer objects like asteroids and icy moons that rigid, rocky bodies achieve hydrostatic equilibrium somewhere between 550-950 km in diameter (the gap between the largest rocky body not in hydrostatic equilibrium, the asteroid known as Pallas, and the smallest one in hydrostatic equilibrium, “dwarf planet” asteroid Ceres). We also know that bodies made primarily out of ices are less rigid and achieve hydrostatic equilibrium at smaller sizes – somewhere in the neighborhood of 400 km in diameter and maybe smaller. (In fact, Methone, a tiny, egg-shaped moon of Saturn made of ice “fluff,” appears to be in hydrostatic equilibrium despite having a mean diameter of only 3.2 km.) So we can make some assumptions about hydrostatic equilibrium based on our assumptions about their sizes, but that leaves an awful lot of guesswork, which explains why the IAU has only officially granted “dwarf planet” status to five of the fifty or so objects that appear to meet the criteria.

The “roundness” criterion gives me pause for another reason as well. There are only four objects in the Solar System over 400 km in diameter that are known to not be in hydrostatic equilibrium: asteroids Pallas (544 km), Vesta (529 km) and Hygeia (430 km), and Neptunian moon Proteus (416 km). All four are a little too lumpy to be in hydrostatic equilibrium, but are pretty close to spherical nonetheless. At least two of them appear to be irregular at least in part due to past traumatic collisions: Vesta has a big chunk missing, while Proteus appears to have accreted from the violently shattered pieces of Neptune’s original moons in the wake of the chaos when Neptune captured Triton. Should these four objects be excluded from a category that includes smaller objects that just happened to form from different materials or under different circumstances?

Before you answer, consider that even objects that are demonstrably in hydrostatic equilibrium are not perfectly spherical. Uranus’ moon Miranda, for example, experienced extreme geological activity in its past that has caused it to become misshapen. As a result, its axes differ by as much as 3%. While this is still less than the 8% difference between Proteus’ longest and shortest axes, it is still a significant difference. Also consider that extremely fast rotation has caused dwarf planet Haumea to become highly elongated; it is still basically ellipsoidal, but its polar and equatorial diameters vary by a whopping 49%. The Earth’s rotation, by comparison, only causes its polar and equatorial diameters to vary by 0.1%.

So far there has been little debate, but I think it is a matter for one. And I, for one, side in favor of Proteus. It may be a rubble pile, but it’s a freaking huge rubble pile. Under the right circumstances – some tidal heating from a closer orbit to Neptune, for example – it would have been able to finish becoming as spherical as Miranda.

The words coming out of my mouth

What I propose is a classification system based purely on an object’s size. There are two ways to measure size in this case: mass or spatial dimensions (volume, diameter, etc.). Both are valid and each has certain advantages and disadvantages. Because it is easier to determine for smaller objects, I propose using spatial dimensions for solid objects: specifically mean volumetric diameter (the diameter the object would have if it were perfectly spherical), which is directly related to volume but does not vary exponentially and is thus easier to work with. Using spatial dimensions ignores some of the differences between rocky, higher density objects and icy, lower density ones, and so there will be some differences in terms of mass, gravity, atmospheric origin and retention, and so forth. But it also means that the objects of the same size class will be of similar shape and surface area.

The fluid, low-density nature of gas giant planets, however, creates a different dynamic. Because gasses expand and contract easily, diameter is a poor indicator of the nature of these planets. For example, the effect of increasing gravitational forces may cause their diameters to stay the same or even shrink as mass increases. Meanwhile, their diameters will expand if the gasses in their atmosphere are warmer (such as in an orbit closer to a star) or shrink if those gasses are colder. For these reasons, pure size is an impractical measuring stick for gas giants and I propose that mass be used instead.

I further propose that the division between “planet” and “not quite a planet” is a blurry one, and that several broad categories are necessary to explain the gradual change. The basic dividing line will be placed roughly at the boundary for rocky bodies to achieve hydrostatic equilibrium: mean volumetric diameter of about 600 km. Objects above this boundary are “planets,” regardless of all other criteria; however, it is important to be cognizant that objects just above it are barely planets and objects just below it are almost planets, and so the categorization should reflect this.

The broad categories for “planets” are “giant planets” (those massive enough to retain light gasses like hydrogen and helium in their atmospheres), “terrestrial planets” (roughly Earth-like in size and potentially suitable for human colonization) and “dwarf planets” (like terrestrial planets but without any of the amenities like atmospheres or substantial gravity).

Below these categories, objects are divided into “demi-planets” (which may or may not be in hydrostatic equilibrium, but will most likely be somewhat round or potato-shaped, and to share other traits with planets), “planetoids” (which are not likely to be rounded by their own gravity), “meteoroids” (objects small enough to pose no substantial threat to mankind should they crash into the Earth) and other categories too boring to mention unless you’re a scientist.


(Note that in the first draft of this article, I had originally called demi-planets “greater planetoids,” but I was never comfortable with it. The term “demi-planet” does a much better job of describing their status of being almost-but-not-quite-planets. The term also creates a nice parallel in that planets are named after deities, and demi-gods are almost-but-not-quite-deities.)

The term “asteroid” is used to denote non-dominant objects that orbit the Sun inside the orbit of Jupiter. This is a term that indicates location, not physical properties. While most asteroids are planetoids, they do not have to be: at least one asteroid, Ceres, is a dwarf planet. Likewise, “centuar” is used to denote non-dominant objects which orbit the Sun beyond Jupiter but within the orbit of Neptune, and “comet” describes objects with extremely elliptical orbits that cross multiple regions of the Solar System. Finally, “plutoid” is used to denote any trans-Neptunian object. (This is an expansion of the current IAU definition of the term, which only refers to trans-Neptunian dwarf planets.)

If one sub-stellar object orbits another, it is called  a “moon.” Any moon big enough to be a planet, however, is still understood to be a planet, just one that happens to orbit another planet. Such a moon may be called a “major moon” or “major satellite” or “satellite planet” or “secondary planet.”

Likewise, “intermediate moons” (or “intermediate satellites”) will be understood to be demi-planets, and “minor moons” (or “minor satellites”) will be understood to be planetoids. (As an aside, certain satellites now listed as “major moons” in many references become “intermediate moons” under this definition: Enceladus, Mimas and Miranda. This puts them in the same category as similar-sized Proteus, whose previous exclusion as a major moon was strictly a case of hydrostatic equilibrium bias.)

The term “major planet” will be used to indicate the eight largest planets in our Solar System (Jupiter, Saturn, Neptune, Uranus, Earth, Venus, Mars and Mercury), as well as any other planets (real or imagined) that dominate their orbits. This provides a term that syncs with the current IAU definition, without making the definition of “planet” dependant on it: an object does not lose “planet” status if it is not dominant; it only loses the adjective “major.”

The term “minor planet” has traditionally been used to encompass everything else. That’s bound to be confusing, since most “minor planets” are planetoids (the two terms have traditionally been synonymous), but some will now be classified as small planets. I doubt I could convince the Minor Planet Center to change its name, which may have been why the IAU insisted that dwarf planets were not technically “planets” in the first place. For clarity, the term “minor planet” should probably be avoided, but I do like that it shows that objects of different sizes are part of a steady continuum. (In a way, they’re all planets!)

The term “SSSB” (small Solar System body) adopted by the IAU in 2006 should be discontinued in favor of “planetoid.” It was adopted because of an increasing realization (thanks to main belt comets, damocloids and centaurs) that the division between planetoids and comets was artificial, and they wanted to create a new term that encompassed both. Why not just say that comets are now recognized as a type of planetoid, along with asteroids, centaurs, trans-Neptunian objects (TNOs) and the rest? “Planetoid” is sufficiently generic sounding; why inflict us with a pointless, unpronounceable acronym? (And also one that is location-specific, as it only applies to objects in our own system, not other star systems.) That was dumb and unnecessary, so I hereby kill it with fire.

Finally, there are a number of planetoids that share a name with an object that would now be considered a planet under this definition: for example, asteroid (52) Europa shares a name with one of Jupiter’s major moons. I propose that all planets and demi-planets in our Solar System be afforded unique names, and in cases of overlap, the smaller body should have the word “Minor” appended to its name. The asteroid Europa thus becomes Europa Minor.

But wait, there’s more

In “part 2,” I will go into more detail about my recommendations for classifying planets and planetoids by size. Stay tuned...

5 comments:

  1. WoW! Do you ever luv classifications, or what?! lol Robert Jenkin I am Upper Canadian! Toronto aka The City of Notes

    ReplyDelete
  2. This comment has been removed by a blog administrator.

    ReplyDelete
  3. This was amazing to come across in a google image search to find a picture of Nessus. I know pretty much nothing about astronomy, but have been becoming fascinated after learning some astrology (don't judge!). In terms of how I think of the planetary bodies in our solar system and the universe at large, I will be adopting your system. And am even more inspired to learn more about our amazing universe.

    ReplyDelete