The upper limit to a planet

By Phil Plait | September 7, 2006 12:29 pm

See that little dot next to the bright one? Here’s an interesting puzzle: is it a planet, or a star? If you’re thinking "Here we go again!" then you’re not too far off the mark.

With all the Pluto nonsense going on, the meaning of the word "planet" is under fire. Mostly, the definition has been causing grief due to how it applies to objects at the small end of the scale. But what about the upper end?

Take Jupiter. It’s the biggest planet in the solar system. What happens if you dump mass into it? Let’s throw Saturn, Uranus, Neptune into it. Of course, Jupiter will get more massive. Oddly, it won’t get much bigger: more mass means more gravity, and since Jupiter is mostly gas, it compresses. That extra compression will pretty much balance out the extra volume being added, and the size stays the same. Weird, huh? Weirder yet is that Jupiter is right at the lower end of the size to do that. Saturn will get bigger as you add mass, but once you get to something about Jupiter’s mass, the size stays pretty much the same even if you dump ten times as much stuff onto it!

But as you compress a gas, it heats up (do a web search on Boyle’s Law). So at the center, it gets hot. As the mass increases, it gets very hot. At some point, the mass will be enough that the core temperature will reach the point where hydrogen atoms will fuse together, forming helium (actually, it’s a lot more complicated than that, but this is the main process). This releases a lot of energy, heating things up more. The core expands, and this means the star expands too. As it does, it cools (Boyle’s Law again, but in reverse this time). Eventually, a balance is reached such that the heat generated expands the gas, while the gravity compresses it, and the two balance. Literally, a star is born.

The magic mass for this is around 0.08 times the Sun’s mass. Anything with this mass or more is a full fledged star. If it has a smaller mass, it isn’t a star. It’s called a brown dwarf.

So brown dwarfs have an upper mass limit. But what is the lower limit? Where is the borderline between what we call a planet and a brown dwarf?

Jupiter has a mass about 1/1000th of the Sun. You might sometimes hear that Jupiter is a failed star, but that’s Bad Astronomy: it would need to be more than 80 times more massive to become a star! So if it’s a failed star, it’s a really failed one. But if we dump mass into it, when does it become a brown dwarf?

The answer is most likely never. Why? Well, we get into dumb, arbitrary planet definitions again! Most astronomers think of BDs as forming like stars: collapsing under their own gravity from a cloud of gas and dust. Planets, however, form from the disk of material circling a star. That’s not a gravitational collapse process, but more of fragmentation and accretion. However, this distinction is a bit silly: you could have two objects, both of which look exactly the same, but one formed from gravitational collapse, and one formed from disk fragmentation. Yet one would be a brown dwarf, and the other a planet. That’s dumb.

But what would you use as a mass limit? There really isn’t anything that distinguishes between, say, something that has 10 times Jupiter’s mass and something that has 20. One’s more massive, but other than that they’re pretty much the same. So a mass cutoff between planet and BD is arbitrary, and you know how I feel about that. Arbitrary rules do not a definition make.

So astronomers fudge, and just say that something more than about 13 times Jupiter’s mass is a BD. Why? Well, theoretically, this is about the upper limit to the mass of an object that will form from fragmentation. Not terribly satisfying, is it? It isn’t to me, but there you go. Nature doesn’t always provide us with a line in the sand, labeled "That be planets, here be brown dwarfs". It’s around this mass that deuterium, an isotope of hydrogen, will fuse. The problem here is that this is not a sustainable source of energy; a BD fusing deuterium can only keep it up for a few million years or so before running out. What is it then? And the mass limit is not rock solid: other factors come into play (elemental abundances, convection, etc.) that can change things such that a lower mass object can fuse deuterium. So while there is a physical change around 13 Jupiter masses — fusing deuterium versus not fusing it — it’s still a fuzzy border.

Which brings us, finally, to the object in the picture above. The bright object is a star, named CHXR 73. It’s actually really faint, a low-mass red M-type dwarf. But the fainter object is CXHR 73b, its companion. This is a newly-released Hubble image of the pair, and astronomers say that the mass of 73b is about 12 times that of Jupiter. So what is it? A planet, or a brown dwarf?

The mass indicates it’s a planet. But, and this is a big but, it’s really far from its parent star: about 31 billion kilometers (19.5 billion miles). That’s about 200 times the distance of Earth from the Sun, and according to all theories that’s too far out for a planet to form. The original disk of junk circling the star would never reach that far out, especially around a dinky star like CXHR 73. So most likely 73b formed from gravitational collapse out of the original gas cloud that the primary star formed from. That makes it a brown dwarf.

I think this is all very silly. We may never know for sure how 73b formed, and I absolutely guarantee we’ll find objects that’ll be even closer to the line than this, causing even more confusion. Worse, if we label it one way, it may cloud our judgement. If I think of it as a planet, I might overlook some star-like qualities of it, and vice-verse. That’s why labelling things can not only be silly, but outright damaging to science. You have to tread very carefully.

CATEGORIZED UNDER: Astronomy, Cool stuff, NASA, Science

Comments (37)

Links to this Post

  1. StrayPackets · Links: Wednesday, 20 September 2006 | September 20, 2006
  1. andy

    I thought the 13 Jupiter-mass limit was to do with the lower limit for deuterium fusion rather than the formation process, which makes it sound a bit more reasonable.

  2. Oh, nuts! You are absolutely right, and I screwed up. I’ll fix it.

  3. The Bad Astronomer Says: “That’s why labelling things can not only be silly, but outright damaging to science. You have to tread very carefully.”

    But isn’t labelling things fundamental to categorizing them? And isn’t that the basis for most sciences?

    – Jack

  4. Kirk: “Bones, there’s a… thing… out there.”
    Bones: “Why is any object we don’t understand called a ‘thing’?”

  5. Rumour Mongerer

    For me, one of the fundamental properties of a star(*) is its ability to generate various parts of the electromagnetic spectrum, e.g. visible light, x-rays, etc., etc.(+), which planets don’t (yes, there are radioactive decaying materials, but this is a secondary process to being a “planet”), whereas planets reflect those waves from the stars beside them.

    As for BDs and potential BDs…do they emit radiation? If not, then they are not a star. From then, differentiating between BD, planet and asteroid, etc., is merely a matter of scale and lines in the sand.

    (BTW, 73b doesn’t orbit our sun, so can’t be a “planet”…just to be the first (as I write this) to say that…)

    (*)I’m not an astronomer nor do I play one on TV, so if someone says “aha, what about this then!?” I probably won’t be able to argue about it…

  6. Rumour Mongerer

    Just quickly add a few more points.
    1) I should have probably more simply said something like “natural nuclear fusion reactor” to describe a star.
    2) I’m also willing to extend the ‘star’ definition to something that was a natural fusion reactor in some part of its natural life cycle. If a cold lump of rock was a star at some point, we coudl still call it a star now. The earth never will be, so it isn’t a star. (We could stellar manipulate it into being a nuclear fusion reactor, but that’s why I used the world ‘natural’).

  7. Jamie

    I think making a distinction between objects based on the way they formed makes sense. We do it with rocks. Granted, rocks that form differently have different properties, but the general premise still holds, at least to a lay-person like me.

  8. HvP

    I understand where you’re coming from on that Rumour Mongerer, but of course this isn’t hard and fast either. Practically everything emits low frequency radiation like infra-red – even space dust. The human body emits IR but also radio waves. Jupiter emits a greater quantity of radiation than it receives from the Sun. This is because its gravity is causing the gasses to collapse into high pressures near the core.

    The Earth also generates both IR and radio itself. If something is hot enough it emits visible light radiation as well – think of lava or a light bulb, this is incandescence. When the Earth and other planets were very young they most likely glowed in visible light for a while before the surface cooled. This is basic thermal radiation and it’s just a side effect of being made up of atoms.

    What we’re left with then is ionizing radiation, including UV and higher wavelenths. Well, some brown dwarfs emit X-rays which pushes the scale even higher. It’s complicated, no doubt, and this may still be a better approach than simply size and mass classification, but it’s not a magic bullet.

  9. HvP

    Of course, I was working on that when you posted your augmentation. Feel free to ignore 😉

  10. buczas

    Distinction based on origin makes perfect sense, no matter how similar the “final products” are. Take chemical aromas and pigments in food – they are “identical to natural” ones, as the label informs us. Nonetheless, they are artificial, man-made, and fall into distint category: artificial aromas/pigments.

  11. We should just tag the object:

    Earth: wet, disk formed, habitable
    Jupiter: gasbag, disk formed, unhabbitable

    CXHR 73b: ?gravitational colaps, prosulmbly gasbag, …

    so this way one object can be have exactly the same tags as an other but only differ i
    n the formation tag :)

  12. andy

    I had a nice post with a point to make, but a computer thinks it’s spam. Ah well. Your loss.

  13. Tim G

    Some folks use the term “sub-brown dwarf”.


  14. Rumour Mongerer

    Thanks for the explanation anyway, HvP. I knew there were problems with that. 😉

    As Jorge points out, we need better tags anyway. If someone says “Hey, look, it’s a planet!”, that doesn’t really tell you anything about it, other than (maybe) it’s a large mass orbitting a star. Is it earth size? Gas giant? Asteroid with attitude? “Planet” by itself is relatively useless.

  15. Mike

    I believe what you’re trying to come to grips with here is a Sorites Paradox. See

    And no I don’t have a solution and I still don’t see what was wrong with saying Pluto is a planet just because we want to call it a planet.

  16. BA Brother

    Hear, hear, Mike!

    Why do humans find it necessary to fiddle with stuff like this? It’s been a planet pretty much as long as I’ve been alive. Nothing about Pluto has changed in that time that I’m aware of to alter its status.

    Don’t these scientists have better things to do, like fighting the all-encompassing republican ignorance in this country?

  17. Space Cadet

    Jorge and RM must have read the whole article. I think our host’s point is the danger in our tendency to catagorize things, not what are the best criteria. “If I think of it as a planet, I might overlook some star-like qualities of it, and vice-verse.”

    A wonderful essay, Mr. BA. Eloquently stated.

  18. Actually, BA Brother (and I know who you are!), Pluto’s size has “shrunk” over the years. As measurements have gotten better, the measured diameter of Pluto has gotten smaller. At first it was though to be as big or bigger than Earth, but that’s because it’s very reflective. Once the albedo was determined better, it was understood that Pluto wasn’t as big (a shiny small object can be as bright as a dark big one). But size is secondary to the planet definitions; it only comes into play since mass determines the ability to make the object round.

  19. How about using the Gas Giant definition as a differentiator,

    So we can have

    1) Stars – objects with currently ongoing thermonuclear activity
    2) Brown dwarfs – gas giants with some history of thermonuclear activity
    3) Gas giants -no history of thermonuclear activity
    4) Planets

  20. If an astronomical body that burns D (or 3He, or Li, or Be, or any of those other unstable lights) is a brown dwarf, then why don’t Tokomak proponents talk about “harnessing the power of the brown dwarf?” Why don’t nuclear annihilation opponents warn against “extinguishing civilization with the flash of a thousand miniature brown dwarves?”

    All anthropogenic fusion is D-D or D-T fusion, and yet this is described as the process that lights the sun and stars. So why not call any actively fusing body a star?

  21. gopher65

    Lab Lemming says:

    “All anthropogenic fusion is D-D or D-T fusion, and yet this is described as the process that lights the sun and stars. So why not call any actively fusing body a star?”

    I think they do Lab. Brown Dwarves are a type of star. My memory hauls up this as the definition of a star (though it is faulty at best. My memory that is heh):

    1) It formed from a gas cloud collapsing, not from the accretion disk of another star (what, praytel, happens when an M-class star forms in the wake of massive 150 solar masses star? Surely it forms from the accretion disk of the star. In our solar system Jupiter took most of the surviving material from the accretion disc, and it is about 1000 times smaller than the sun. If you have a 150 solar mass star, and a 0.08 solar mass star forming from its accretion disc, that smaller star 2000 times smaller (I think). So I don’t see why this couldn’t happen. Is that smaller star not a star just because it formed from the accretion disc of another star? What happens if it is a brown dwarf that forms this way instead of an M-class star? Is it automatically a planet?)

    2) It has to be able to sustain some sort of fusion over a relatively long period of time. How long this is I don’t know. I suspect that it probably has to be a self sustaining reaction. But a brown dwarf can sustain (limited) fusion as long as a Blue Supergiant can. *shrugs*

    These definitions are really wacky.

  22. gopher65

    Ug. I forgot to spellcheck and proof-read that. I’m sorry.

  23. JustinK

    I have no scientific training, so this is something of a shot in the dark, but why not import a distinction used by philosophers, which distinguishes between definitions and descriptions? It is hard to define some things in life, such as love, where you can speak of certain facts (e.g., biological changes that take place when experiencing “love”), but where it is almost impossible to define the real essence. In such cases, when you are trying to talk about the topic, descriptions must suffice. Couldn’t the same thing apply to discussions of planets? Couldn’t scientists agree to call their decisions descriptions (implying some elasticity), rather than definitions (implying a rigid, immutable list of criteria)?

  24. I may be a bear of little brain, but the gas law linking pressure and temperature is not Boyles Law, its Charles Law

    Bad “Bad Astronomer” here?

  25. Irishman

    Jamie said:
    >I think making a distinction between objects based on the way they formed makes sense. We do it with rocks. Granted, rocks that form differently have different properties, but the general premise still holds, at least to a lay-person like me.

    But igneous, metamorphic, and sedimentary rocks are all still rocks. We’re talking about the most basic of words, very general categories, not the subcategories within.

    That’s the real problem – astronomers are trying to take a very generic word (planet) and make it convey some level of detailed definition. “Planet” is a vague category between “asteroid” (a chunk of rock) and “star” (a fusing object). So is a brown dwarf a star or a planet? Do we make it a new category wedged between the other two, and how do we draw the lines?

    Mike, I understand the issue with the Sorites paradox. However, there’s a problem with one of the specific examples:

    1 grain of wheat does not make a heap.
    If 1 grain of wheat does not make a heap then 2 grains of wheat do not.
    If 2 grains of wheat do not make a heap then 3 grains do not.
    If 9,999 grains of wheat do not make a heap then 10,000 do not.
    10,000 grains of wheat do not make a heap.

    There does not seem to be any logical justification to conclude step 2 from step 1.

    Counter Example:
    1 grain of wheat does not make a pair of grains.
    If 1 grain of wheat does not make a pair, then 2 grains of wheat do not make a pair.

    But 2 of something is the definition of “pair”. That is a contradiction. Therefore, the logical structure is invalid.

    There may be a way to exemplify the Sorites paradox, but that logical structure is not valid.

  26. What about not drawing boundaries/borders/limits here? What if there is no finite boundary present? What if the difference between a planet and a brown dwarf is a bit of a grey area, something like a boiling-point versus a boiling-range? Do we need to revise insights here then?
    Historically seen, drawing borders is as much beneficial for tax(onomy) as inciting wars.

    As for Sorites. The answers is > 3. The minimum amount of grains (or particles) to form a heap in a 3D-space. Or am I crossing some filosophic boundary here?

  27. Charles’ law is temperature and volume at constant pressure.

    Boyle’s law is volume and pressure at constant temperature.

    I think you want Gay-Lussac’s second law, which states that at constant volume, pressure / temperature is constant.

    Or just use the ideal gas law, which describes all possibilites.

  28. As a wordsmith, I would like to speak in defence of taxonomy.
    Taxonomy is always helpful, but not always immutable, as knowledge changes. I don’t think the Bad Astronomer or anyone else should dismiss a taxonomy of physics just because it is not foolproof and may change in the future.
    Making a definition of a planet/Kuiper belt object/brown dwarf is helpful for the purposes of communication because it allows “things” to be labelled so that their main properties are able to be communicated quickly and succinctly between people. It is only when the label becomes more important than the object it describes that problems may arise. This has been the case with those who wish to adhere to calling Pluto a planet “because it has always been thought of as a planet”.
    When evidence arises to challenge the assumptions that the labels are based on, we ought to change the labels. (This happens in biological taxonomy all the time: plant and animal genus and species names change frequently with new knowledge, and at the species level new species are being added all the time — I read about a new species of spider in my newspaper just this morning.)
    All that needs to happen is to have species, subspecies (and more, if necessary) of planets, brown dwarfs, etc. That way, if you have an example of an anomaly like CXHR 73b, you put it in the genus of brown dwarfs, because you think that’s how it formed based on your knowledge at the moment, and call it a new species because of its size/fusion status. Then, when your knowledge about it causes you to rethink its classification (either because of how it formed or because of its size or because of its relation to other similar objects yet to be discovered) you simply give it a new species or genus name, or put it into another family, or whatever is necessary to define it properly.

  29. OK if it is Gay-Lussac’s second law, then Phil still has an error by refering to Boyles Law

  30. SF Reader

    OK, so let’s collect some more data! Such as, what’s CXHR 73b’s spectrum? Is it glowing on its own, or does it just reflect its primary?

    That doesn’t speak to origins, but it should clarify things, or at least be interesting!

  31. exarch

    Obviously, you’re dealing with a spectrum here. Just like you can call a certain wavelength of visible light “red”, when moving along the spectrum, there comes a point where the colour you’re seeing is no longer red. There is no exact cut-off, but that’s not important either, unless you want to insist on classifying things as “red, blue or green”. But why is that important, except to make communication faster?

    As far as I can tell, the definitions of “star” and “planet” aren’t mutually exclusive, so it shouldn’t come as a surprise that there’s overlap between the two categories. So perhaps an object can be both?

  32. Elwood Herring : “Why is an object we don’t understand called a thing?”

    Because its shorter than saying “thingummeywhatsit!” Or for that matter
    ‘object we don’t understand’

    Why what else would youcall it?

    An Unidentified Flying (very-far out) object?

    Well ok it ain’t reallyflying more orbiting but anyway …

    Unidentified Celestial Object mayhaps?

  33. Brown dwarfs (dwarves?) are intermediate obbjects falling between stellar and planetary natures.

    They come in two variants spectral types recently (late 90’s early 00’s?) added to the end of the familar OBAFGKM spectral classification : L & T.

    I see no reason why there can’t be a certain amount of overlap – and why we can’t recognise a given object as having characteristics of both star and planet – eg. boasting flares and weather alike.

    Seems useful and worth adopting a ‘sub-species’ type of descriptive label tothe term planet as is effectively done now describing an object as a superjovian or hot jupiter or exo-Neptune or gas giant or terrestrial or ‘ice dwarf’ type planet.

    I also see no reason why we can’t be upfront about saying its a borderline object right onthe brown dwarf /superjovian planet boundary .. science unlike religion can after all admit and even take advantage of uncertainty.

    Ditto with dwarf planets – “ice dwarves” these are borderline planets – and in my view we should describe them as such. In all these things there’s acontiuum of objects along a range of mass, size, central temperatures, ability to fuse specific elements etc ..

    Jupiter radiates considerable radio and infra-red energy far more than it receives (reflects) from the Sun. It doesn’t make it a star and not quite a planet as we usually think of them – what it is, is Jupiter or a Jovian or gas giant planet which amounts to saying the same thing.

    CHXR73b (great name -not!) – I presume – radiates even more IR, radio and possible other wavelengths, can perhaps fuse deuterium and yet isn’t quite massive enough to be a brown dwarf. So what is it? A Superjovian planet? A very low mass L-type brown dwarf star? Or just CHRX73-b?

    Pluto is round, reflects sunlight and has a differentiated core, an atmosphere, a major satellite system with one comparatively huge moon and two other small ones and perhaps rings too. Yet it falls right on the borderline of what we’d call a planet – below the line if you follow the IAU stricture, above if you think they can “shove Pluto where the Sun don’t shine”! So what is it then? A planet? An ice dwarf? A dwarf planet? A giant cometary nucleus or minor planet? Well, perhaps its just Pluto …

    An interesting world in its own way as Jupiter and CHRX-73b also all are.

    Lets stop geting hung up on names – and possibly consider adding to the word ‘planet’ so we can tell usefully what sort of ‘planet’ a planet (or sub-brown dwarf or ice dwarf) is rather than disqualifying objects based on questionable rationalisations and (mis?) understandings.

  34. Der

    “Hear, hear, Mike!

    Why do humans find it necessary to fiddle with stuff like this? It’s been a planet pretty much as long as I’ve been alive. Nothing about Pluto has changed in that time that I’m aware of to alter its status.?”

    Because science need to accept its mistakes and correct them. Get over it, nostalgia retards.

  35. UserGoogol

    Irishman @ 25: That’s misrepresenting the argument. The claim “for all n, if n grains of wheat is not a heap, then n + 1 grains of wheat is not a heap” is meant to be a claim about heaps in particular, not a generic logical statement about the nature of collections in general. And it does seem mighty peculiar to say that removing a single grain of wheat from a heap of wheat makes it into a not-heap. There are indeed sorts of collections where removing a single item does change the class of that collection, but heaps seem to be a much looser kind of category. After all, if you have a pile of wheat which satisfies the typical functional properties of “heaps,” (for instance, you can roll around in it if you want) taking a single grain isn’t going to change that.

  36. Hah! You fooled me for a bit here linking this one on facebook, BA! 😉

    Thought I was reading something new and shocked to see a comment by ‘mungascr’ – my very early username here after a Turkish Angora cat that once owned me plus my initials. Quite flummoxed until I noticed the date!

    (FWIW. Munga is the Swahili word for cat & also a type of horsefeed.)

    [Old Ben Kenobi] Ah, there’s a name I’ve not heard in many years. Many years! [/Old Ben Kenobi] 😉 )

    Guess it shows how little things change on this issue – & that my memory’s not what it could be. :-(

    Any update on the status of CXHR 73b btw?

    NB. @ 34. my former self from six years past :

    Brown dwarfs (dwarves?) are intermediate objects falling between stellar and planetary natures. They come in two variants spectral types recently (late 90′s early 00′s?) added to the end of the familar OBAFGKM spectral classification : L & T.

    We’ve now added a third brown dwarf spectral class – Y – as well for the very coolest and most planet-like variety of them which I guess CHXR-73b may well be as long as it isn’t still too young and hot. Wiki- page for brown dwarfs with classification illustrations of types L, T & Y linked to my name here. CHXR-73b doesn’t have a wikipage of its own at least as of now.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!


See More

Collapse bottom bar