When a star eats its own

By Phil Plait | September 14, 2010 12:00 pm

What happens when a star with a planetary system (or perhaps a close stellar companion) gets old, expands into a red giant, and engulfs its neighborhood?

This:

chandra_bppsc

"This", in this case, is the star BP Piscium (or just BP Psc), a star a bit less than twice the mass of the Sun located about 1000 light years away. The picture is actually a composite of both an optical image from the Lick Observatory (in white and green) and X-rays using the orbiting Chandra Observatory (purple).

The jets of matter streaming away are usually seen around young stars. When a star forms, there is a thick disk of material surrounding it. Due to processes not fully understood yet (though we know it has to do with the disk, the star’s spin, and the star’s magnetic field), matter and energy can be focused into those two beams, and they can blast away from the star’s poles at high velocity, stretching for several light years.

But there’s a monkey in the wrench here: BP Psc isn’t a young star.

We know this for several reasons. One is that it’s nowhere near a star-forming region, and young stars with beams like that are never found outside of stellar nurseries. Second, spectra indicate it’s a red giant, a star like the Sun at the end of its life. And third, while it is emitting some X-rays, as young stars do, it’s not nearly bright enough or emitting them in the way young stars do. In fact, the X-ray spectrum matches that of an old star that is spinning rapidly.

All in all, it really looks like BP Psc is an old star masquerading as a young one. How is it doing that?

The key is in those beams. We think you need a disk surrounding the star to help focus them. How does an old star get a disk? Ah, well, when a star like the Sun gets old and becomes a red giant, it expands. A lot. The Sun can go from a diameter of a little over a million kilometers to a monster size of a hundred million kilometers or more! When the Sun does this (in about 6 billion years or so), it’ll swallow Mercury and Venus. We’re not sure the Sun can get big enough to consume the Earth, but we do know some stars get that big.

chandra_bppsc_artWhen a star expands, its rotation slows too. The usual example is of a figure skater who can increase their rate of spin by drawing in their hands — you’ve seen it many times, no doubt. The opposite is true as well: if you spin rapidly, you can open up your arms to slow down. Stars do the same thing: a star like the Sun that takes once per month to spin now might take years to rotate once around when it becomes a red giant.

Now imagine a big planet like Jupiter (or a very small red dwarf star) orbiting close in to a star like that. It may orbit the star in a matter of days or weeks, and remember the star itself rotates far more slowly than that. When it expands and swallows up that planet, the planet won’t just vaporize; it can take years for it to totally be destroyed. And during that time it’s orbiting inside the star! A star is just gas, and when it becomes a red giant it gets big and its density drops even more. As bizarre as it sounds, something much denser like a planet can orbit inside a red giant for some time.

So the planet is inside the star, mixing it up like a whisk in a bowl of batter. This can cause all sorts of havoc, including making the star to expel material along its equator to form a disk. A lot of that matter might be from the companion itself, broken up in the chaotic process.

What we’re left with is a weird star: it’s old and big and used to have a companion (or several), but now is spinning rapidly, wearing a thick disk around its middle, and blasting out those beams of material.

In other words, BP Piscium!

Interestingly, there’s some thinking that this disk might be thick enough to actually form planets once again. We can’t be sure about that with this example, but we’re getting very good at finding planets around stars. It would be amazing if, one day, we found what look like very young planets orbiting a very old star. If that were the case then we might have more evidence of what’s going on inside of BP Psc.

planetarynebulaeI’ll also note that we have lots more evidence of similar events: some stars blow off shells of material when they die. These inaptly-named planetary nebulae (called that because in a small telescope they can look like the disks of planets) come in all manners of bizarre shapes, including some indicating their parent stars were spinning rapidly. Since a solo red giant spins slowly, it’s been thought for years that these nebulae indicate the stars are swallowing either a binary star companion or perhaps inner giant planets. In fact, my old research advisor Noam Soker postulated this to explain planetary nebulae long before the first super-Jupiter planetary companion was ever found.

One of the things I love about science is how everything fits together. Since we’re describing reality, things have to fit together! So when we find an object like BP Psc, one that causes us to scratch our heads, we can turn to other observations, other knowledge, and try to figure out what’s going on. That’s why science is far more than just a compendium of knowledge; it’s a tapestry, a complex weave, where everything relies on everything else. And just like an exceptional tapestry, it’s a thing of beauty and wonder.

Image credits: BP PSC: X-ray (NASA/CXC/RIT/J.Kastner et al), Optical (UCO/Lick/STScI/M.Perrin et al); Illustration: NASA/CXC/M.Weiss; Planetary nebulae: NASA/HST/Bruce Balick

CATEGORIZED UNDER: Astronomy, Cool stuff, Pretty pictures
MORE ABOUT: BP Psc, Chandra

Comments (22)

Links to this Post

  1. Science cum Metaphysics « Speculative Heresy | September 27, 2010
  1. Great pictures and explanations as usual Phil.

  2. OtherRob

    Why would a star that “swallowed” its companion spin more rapidly than a solo red giant? Is the companion star or planet dragging the red giant around with it?

  3. Kyle

    Cool stellar cannibalism. Do we have much clue as to the temperature gradient as you go deeper into a red giant’s atmosphere? How much “warmer” would it be for an object at Mercury’s orbit than Venus’s an would something Jupiter’s size cause higher temps because it is churning thing up so much?

  4. Dr BA: These last couple of years, you always finish your science posts with a paragraph praising science and reality. Is it hard to always come up with that paragraph? Always having to find yet another way of saying how the universe is cool? I find it very poetic sometimes. Maybe you should collect these last paragraphs and make a book out of them.

  5. DrFlimmer

    @ OtherRob

    The planet is hold by the star (due to gravity) normally with a “long” arm. When the planet crashes into the stars then the arm becomes shorter and shorter. And, thus, the spin rate of the star must increase (just as the skater analogy….)

  6. Dave

    @OtherRob, @DrFlimmer: The planet came to exist so far from the star in the first place because it was moving too fast to fall. (That’s what it means to orbit.) When the star expands, upper layers of its atmosphere puff up on top of the lower layers. The atmosphere isn’t orbiting. (That is the definition of an atmosphere.)

    So, yes, the planet is moving much faster than the atmosphere. The forward-facing side of the planet would experience ground-vaporizing wind erosion until nothing is left. The energy released by this blasting would be comparable to (somewhat less than) its kinetic energy.

    The arm doesn’t get shorter, the skater gets fatter.

  7. James

    “Science is far more than just a compendium of knowledge; it’s a tapestry, a complex weave, where everything relies on everything else.”

    Fantastically said. This is exactly what I love about science and it’s a message that’s too often neglected.

    PS. Out of interest, what sort of timescale are we talking about for the conversion of ordinary yellow star into red giant? I’d always wondered and vaguely assumed that it was on sort of typical stellar timescale, but your post makes it sound like it is extremely rapid.

  8. Paul

    @OtherRob: thanks for asking my question – how do you get a rapidly spinning red giant?

    @Dave: if the skater is the planet, how is it that the rink (star) is spinning rapidly. The Space Shuttle undergoes a similar effect, yet doesn’t trigger a measurable change in the Earth’s spin,

    (I know, I know, you do enough tight spins and to the skater at least, the rink will _seem_ to be spinning)

  9. If you can have a planet orbiting inside a star presumably you could have a planet orbiting just on the “surface”? It would look like a grain of sand rolling around on a beach ball I suppose but it would look cool.

  10. Messier Tidy Upper

    Awesome post BA thanks. :-)

    With a name like BP Piscium I’m surprised no one has yet suggested those are jets of oil gushing out from the star! ;-)

    (Yeah, I know its a typical variable star designation with nothing to do with petrol or mining corporations at all, moving right along.)

    As far as planets or even stars within stars go, Betelguese may be one example of this :

    “A star inside a star. A 1986 study by three Harvard-Smithsonian astronomers gave evidence for a companion star orbiting the famous red giant [sic – its actually a supergiant star.] Betelgeuse … Their data suggest that this close companion is in an elliptical orbit that dips inside the red-glowing, outermost atmosphere of Betelgeuse. In one interpretation … the companion star could be a blue-white companion of a few solar masses. Such a system could form when the red giant [sic] expands and engulfs part of the orbit of the second star: The system would last only a short time, astronomically speaking, since drag forces will cause the second star eventually to plunge far into the giant [sic] and to merge with it.

    Source : Page 36, ‘Cycles of Fire: Stars, Galaxies and the Wonder of Deep Space’, William K. Hartmann & Ron Miller, Workman publishing, 1987.

    There hasn’t been any more discovered on this & I don’t know if its been confirmed or refuted but I find it an amazing prospect to consider.

    There are also theories suggesting this explains the odd eruption of red supergiant V838 Monoceros.

    .. young stars with beams like that are never found outside of stellar nurseries.

    Never? Never *ever*? Hmm .. the cosmos is mighty big and I do think I recall reading somewhere of an exception to this. Of sun-like stars forming outside of main star-forming regions and then there’s also the phenomena of “runaway” stars (Mu Columbae, AE Aurigae, 53 Arietis, etc ..) ejected from such regions – could they be ejected so early on that they are still forming protostars?

    Odd to think of a vast spinning red giant star – with such tenous outer layers I wonder if its rate of mass loss is, well, massive? ;-)

  11. Messier Tidy Upper

    @9. shane Says:

    If you can have a planet orbiting inside a star presumably you could have a planet orbiting just on the “surface”? It would look like a grain of sand rolling around on a beach ball I suppose but it would look cool.

    I doubt it would last long *on* the surface given the drag factor -the stars atmospher would draw it in pretty quickly – and also blow a lot of its atmosphere away – Hot Jupiter’s are already known to be evapourating away leaving huge comet-like tails of their outer atmosphere’s flowing away.

    One example is HD 209458 b or “Osiris” which was the first Hot Jupiter found by the transiting method was also the first to have its atmosphere detected – “boiling” off this exoplanets surface.

    One study around 2006-2007 suggests that all HotJoves closer than 0.15 AU (24 million km) will eventually be destroyed by this catastrophic evaporation. Of course this doesn’t mean they can’t fall into their stars as well! ;-)

    (Or be engulfed when their parent sun expands to gianthood swallowing them. Think the BA has posted at least one example of that here too.)

    In the novel ‘Sunstorm’ by Clarke & Baxter a superovian exoplanet – from a fictional Altair system – was used as a weapon by advanced aliens who plunged it into our daytime star where it is, according to the novel, already orbiting so it can trigger theeponymous Earth-sterilisingevent.

  12. Messier Tidy Upper

    Typo fixed paragraph :

    In the novel ’Sunstorm’ by Clarke & Baxter (great read btw.) a superjovian exoplanet – from a fictional Altair system – was used as a weapon by advanced aliens who plunged it into our daytime star where it is, according to the novel, already orbiting deep inside our Sun so it can trigger the eponymous Earth-sterilising event.

    See link : http://en.wikipedia.org/wiki/Sunstorm_(novel)

    For more on the evapourating HotJove Osiris see :

    http://blogs.discovermagazine.com/80beats/2010/07/16/osiris-the-scorched-exoplanet-with-a-comet-like-tail/

    Or be engulfed when their parent sun expands to gianthood swallowing them. Think the BA has posted at least one example of that here too.

    For the example I was thinking of there see :

    http://blogs.discovermagazine.com/badastronomy/2010/05/20/star-om-nom-nom-planet-aieee/

    Plus I think there have been a few more similar “exoplanet consumed by star” stories posted here and elsewhere too.

    I hope folks here find these links useful / interesting / enjoyable. Do they? :-)

  13. mfumbesi

    I found beauty in your narration, maybe I’m a pervert who like seeing dyeing starts swallow their companions (they be planets or stars).

  14. Dave

    @Paul: The planet causes the star’s atmosphere to speed up both by contacting it, and adding additional high-velocity material. I’m not an astrophysicist, but I’d guess that atmosphere is extremely tenuous.

    Jupiter’s mass is 0.1% of the Sun’s mass. If the top independently-rotating layer of the red giant’s atmosphere accounts for even 1% of its total mass, and ends up absorbing a Jupiter, it could get appreciably spun up.

    Denser material falling in from the outermost layer would also speed up as it fell, perhaps having a greater effect on lower, denser layers of the star. The total energy released through *that* process would be equivalent to the planet ceasing its revolution and falling smack into the star at extreme speed — more than the kinetic energy of orbital motion which is transferred to the outermost layer.

  15. Messier Tidy Upper

    @13. mfumbesi :

    I shouldn’t pick on typos I know – & FSM knows, I make enough of them myself
    – but “dyeing” stars? Which colours would you want to dye them? ;-)

    @ 3. Kyle Says:

    Cool stellar cannibalism.

    Is it really “cannibalism” if a star eats a planet rather than another star? If you really like “cannibalistic stars” I suggest you type Black Widow pulsar into Wikipedia. ;-)

    Do we have much clue as to the temperature gradient as you go deeper into a red giant’s atmosphere? How much “warmer” would it be for an object at Mercury’s orbit than Venus’s an would something Jupiter’s size cause higher temps because it is churning thing up so much?

    Well that would depend on a number of things, not least the nature of the object.

    Venus is actually hotter than Mercury despite being further from our Sun because the dense Cytherean atmosphere traps more heat creating a runaway Greenhouse effect.

    The Jovian or gas giant planets are still incredibly hot at their cores and they shine in the infra-red emitting more energy than they recieve from sunlight. Temperatures on (or should that be *in*) Jupiter itself, for instance, range from a minimum of 125 degrees Kelvin at the cloud-tops to a maximum of 30,000 K at the core.

    As for “churning things” up, I’m not sure but I suspect so. Having a planet or worse star inside the outer “red-hot vacuum” atmosphere of the red giant seems bound to have some major effects and impacts to me.

    I have read somewhere that companions (stellar &/or planetary) are suspected of playing a significant role in shaping planetary nebula although that’s a later stage of the system’s evolution.

  16. Messier Tidy Upper

    From :

    http://blogs.discovermagazine.com/visualscience/2010/07/15/astronomer-mike-brown-on-arty-exoplanets/

    This artist’s concept illustrates the hottest planet yet observed in the universe. The scorching ball of gas, a “hot Jupiter” called HD 149026b, is a sweltering 3,700 degrees Fahrenheit (2,040 degrees Celsius) – about 3 times hotter than the rocky surface of Venus, the hottest planet in our solar system. The planet is so hot that astronomers believe it is absorbing almost all of the heat from its star, and reflecting very little to no light. Objects that reflect no sunlight are black. Consequently, HD 149026b might be the blackest known planet in the universe, in addition to the hottest.

    Then there’s also this interesting study :

    http://blogs.discovermagazine.com/80beats/2010/04/13/astronomer-earth-like-planets-are-common-but-stars-have-eaten-many/

    indicating that BP Piscium is far from alone in this class.

    And even in an “earth-like” 360 day orbit aroudn aerd giant a planet can get pretty hot see :

    http://blogs.discovermagazine.com/badastronomy/2007/08/03/new-planet-with-earthlike-orbit-nah/

  17. MadScientist

    Don’t eat that little blue planet; it will give you a very bad case of gas.

  18. Doug M.

    James @7: a star like the Sun would take several hundred million years to go from late-main sequence to red giant. During this time it will steadily grow bigger, brighter, cooler and redder. Stars in this stage of life are called “subgiants”.

    So it’s not a very rapid process at all.

    Doug M.

  19. Howard

    How hot would it be inside the red giant? Would it be enough to melt or even vaporize a planet that got caught inside?

  20. Messier Tidy Upper

    @17. MadScientist Says:

    Don’t eat that little blue planet; it will give you a very bad case of gas.

    Actually I would ‘ve thought the big gassy banded (& ringed and many multi-mooned one’s) would be the one’s giving the red giant the gas! ;-)

    @19. Howard asks:

    How hot would it be inside the red giant? Would it be enough to melt or even vaporize a planet that got caught inside?

    Depends what the planet is made out of! ;-)

    A world like Pluto made largely of ice would fall apart, meltand subliminate long before. A rocky earth-like world or a planet mostly composed of iron or rock or a carbon planet that is made of asphelt, silicon carbide and diamond would last a lot, lot longer! ;-)

    Red giant surface temperatures are 3,700 degrees Kelvin and below & I expect much of their outer layers are too. This would vary upon the star and its exact spectral class with class M0 being at the warmer edge of the range and M9 at the cooler side – and some variable stars may vary their spectral class too from memory.

    Whether a given star’s luminosity class is I (supergiant), II (bright giant) or III (giant) might also have an effect worth considering here as might its metallicity. (Abundance of elements beyond helium.) However I am unsure of this. Anyone know and care to enlighten us further please?

    Correcting for typos for myself #16 :

    Even in an “earth-like” 360 day orbit around a red giant a planet can get pretty hot see :

    [Dr Phil Plait, 2007 Aug. 3rd :]“The star is a giant. That means it’s bigger than the Sun, and in this case HD 17092 is about 10 times the diameter of the Sun. That means a planet orbiting it will receive more heat from it at a given distance. I have the equations written out if you care to look, but the bottom line is that the temperature of a planet compared to the Earth’s depends on the temperature of the star and the square root of the star’s size. When you do the math (and I hope I did it correctly), you get that the temperature of the planet is about 2.3 times that of the Earth (assuming lots of things like: it has the same reflectivity as Earth, it rotates rapidly, and other stuff that will affect this calculation, but probably not by much).

    So actually, the planet’s surface will be pretty hot: roughly 500 Celsius, or 900 Fahrenheit. That would melt lead and zinc! Remember, though, the planet has more than 4 times Jupiter’s mass, or well over 1000 times Earth’s mass, so it’s probably a gas giant with no real surface at all.”

    From that last link.

    Come to think of it the temperature figure quoted there for HD 17092 is 4600 K which is hotter than the figure I gave. This may mean I’ve got the figure wrong or perhaps the BA was including orange giants in with the “red giant” class there & HD 17092 is in fact a K type star rather than a M type “proper” red giant.

  21. Tribeca Mike

    Absolutely fabulous. Thanks.

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

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

ADVERTISEMENT

See More

ADVERTISEMENT
Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »