The return of Sauron’s planet

By Phil Plait | October 27, 2012 7:00 am

First it was there, then it wasn’t, and now it just may be back again: the first exoplanet directly observed orbiting a normal star, Fomalhaut b, has had quite a ride.

[This post has a bit of detail to it, so here’s the tl;dr version: new analysis shows an object orbiting the star Fomalhaut may actually be a planet, enveloped in a cloud of dust. We can’t for sure it exists, but we can’t say it doesn’t, either! Earlier claims of it not existing may have been premature. Also, at the bottom of this post is a gallery of direct images of exoplanets.]

First a brief history. In 2008, astronomers revealed huge news: they had successfully taken images of planets orbiting other stars. Up until then, the only evidence we had of exoplanets was indirect, either by their tugging on their stars which affects the starlight, or by having them pass between their stars and us, dimming the starlight.

But, along with Gemini telescope pictures of a family of planets orbiting HR 8799, Fomalhaut b was the first planet ever seen directly, as a spark of light in a picture. Here is that historic shot:

It’s Sauron’s eye! [Click to embiggen.]

The object is labeled. It doesn’t look like much, but the important thing to note is that it moved between 2004 and 2006 (see picture below), and it was definitely in both images taken two years apart. That means it wasn’t some bit of noise or detector error. Moreover, the movement was consistent with what you’d expect from a planet. Not only that but the star Fomalhaut is surrounded by a vast ring of dust – Sauron’s eye – and the inner edge of the ring is sharp. That’s what you would expect if a planet was orbiting inside the ring; its gravity sweeps up the dust on the inside of the ring. Given the brightness, we were looking at an object with a few times Jupiter’s mass, much smaller than a star, so definitely a planet.

All in all, it looked good, and it looked real.

Then, in early 2012, some astronomers threw a Pluto-esque wet blanket on the news. A planet that big should be bright in the infrared. Fomalhaut is a youngish star, only a few hundred million years old. Any planet more massive than Jupiter should still be hot, radiating away the heat of its formation. They looked for it in the infrared, and it wasn’t there.

Uh oh.

To make things worse, they found that if you extrapolate the orbit of the supposed planet using its movement, it should cross the ring. That’s bad, because its gravity would disrupt the ring after a few million years tops. The ring is there, so that planet means the planet must not be.

Their conclusion: this object is a clump of dust, a cloud, orbiting the star. That fits the data, and a planet doesn’t. Cue the sad trombone.

But wait! We’re not done!

A team of astronomers took another look at the original Hubble data. They reprocessed it and re-analyzed it, and what they found is that the object is definitely real; there’s definitely something there. OK, that’s a good start. They also see it in Hubble images taken with a bluer filter (it wasn’t spotted before), which is great: more confirmation it exists, and gives a broader color coverage. That’s important, as I’ll get to in a sec.

They also took images in the infrared using the gigantic 8.2 meter Subaru telescope in Hawaii. Interestingly, they did not detect it, which corroborates the finding that whatever this thing is, it’s dark in the IR.

So what is it?

This is where things get interesting. First, they can rule out a dust cloud! A clump of stuff floating out there around the star would have to be pretty big to be so bright. So big, in fact that shear forces would rip it apart in a pretty short time – like a few tens of thousands of years, much less than the age of the star. That makes it pretty unlikely it’s a dust cloud.

So if it’s a planet, that can be checked, too. Because they saw it in a blue filter, the astronomers in the new study were able to look at computer models of what planets should look like through the telescope. They assume an age, a mass, and so on, then crunch the numbers and see what matches the observations. What they found is intriguing: a lower mass planet, something with half the mass of Jupiter or so, fits most of the data. It’s not bright in the infrared because it’s smaller than expected, but still big enough to reflect the light from Fomalhaut and show up in the blue. And it turns out it fits the data even better if you assume it’s surrounded by a smallish dust cloud. A big dust cloud would dissipate, but if there’s a planet in its middle the gravity will hold it together. It’s like Pig Pen from the Peanuts comic.

And there’s still one more thing. They reanalyzed the motion of the object, and using more accurate techniques, determined it does not cross the ring! In fact, the orbit looks like it lines up with the ring pretty well, just smaller. That’s right where you want a planet to truncate the inner edge of the ring.

So hey, it looks like the planet could be there after all. Here’s a fun and explanatory video from NASA on all this, themed for Halloween:

Let me be clear here: this is not proof the object is a planet! It is, however, good enough evidence to say that we can’t kill the planet status just yet. We still need more observations, in different colors, to help nail down what it is. More Hubble observations would be nice as well, to see how it’s moved over the past few years.

So it’s not so much Sauron’s planet as it is Schrödinger’s planet.

And I have to add a wry note: if it does turn out to be enshrouded in a small dust cloud, then it’s still not the first exoplanet directly imaged! We’re seeing the dust cloud, not the planet itself. So that honor goes to the HR 8799 planets seen by the Gemini telescope.

Hey! Clearly, details matter here. Still, no matter what it is, it’s interesting, and it shows something very important: science is a process. New observations, new analysis can change what we think. That is the greatest strength of science! It adapts, as must we, when new information comes in.



[Below is a gallery of exoplanets that have been directly imaged using telescopes on ground and in space. Click the thumbnail picture to get a bigger picture and more information, and scroll through the gallery using the left and right arrows.]

exoplanet_betapic3
exoplanet_1rxs1609b
exoplanet_2m1207b
exoplanet_betapicb1
exoplanet_betapicb2
exoplanet_fomalhautb
exoplanet_fomalhautb2
exoplanet_gliese581c_art
exoplanet_hr8799
exoplanet_hr8799b_art
exoplanet_hr8799cspec
hr8799e
hst_hr8799_1998
lkca-15-b_andstar
vlt_betapicb_2011


Comments (25)

  1. Chris

    It really just has a cloaking field around it.

  2. Dave

    Hi Phil,

    Nice post!

    Whatever we’re seeing, it’s definitely not a planet. Since there’s no (detectable) IR, the visible light we’re seeing must be reflected, not thermally emitted. The object has a reflecting area of something like 400 times Jupiter’s. So, as you allude to in your post, we’re seeing dust, not a single object with 20x the radius of Jupiter (i.e., twice the radius of the Sun). To create this much dust, you just need to turn a ~50-km asteroid into dust, and it’s conceivable that this has happened in the last few hundred to few thousand years (and we’re seeing a transient dust cloud now).

    Now, it’s certainly conceivable that this dust is around a smallish planet instead of just floating alone, and it might be that the planet hypothesis should be preferred (and I have no reason to disfavor it!), but the new study in no way contradicts the study from early 2012 — all that can be confirmed now (as could be confirmed back at the beginning of this year too) is that there’s a large cloud of dust around Fomalhaut and no associated (detectably) self-luminous planet.

  3. There’s actually two very similar papers on this on the Arxiv (I won’t add urls but Google is your friend): 1210.6620 (which is the one reported here) and 1210.6745, which is purely a re-analysis of the Hubble data that comes to essentially the same conclusions. I guess there’s a story there, since they came out on consecutive days.

    The second paper has a neat little diagram illustrating the constraints on any possible planet. It also discusses (and doesn’t rule out) the collision theory mentioned by Dave above.

    it’s still not the first exoplanet directly imaged

    I don’t see this as wry so much as saying “it’s not a planet, it’s something even cooler”.

  4. nobody

    “It’s Sauron’s eye! [Click to embiggen.]”

    What? Just “Click to embiggen.”? How about “Click to Ensauronate”? lol

  5. DanM

    I love the ‘zombie planet’ video – thank you to the Goddard folks. Although it does make me vaguely jealous that I don’t have a video geek willing to help me turn my latest papers into fun-‘n-interesting animated videos for public consumption. Or, more accurately, that I don’t have the budget to pay for that… I’m pretty sure NSF wouldn’t approve of that particular proposed budget reallocation. Hmph.

  6. James evans

    Click to Annatarenate

    Click to Gorthaurenate

    Click to Necromancenate

    How’s that for sailing away into the geeky blue yonder? For going above and beyond the call of geek duty? For reaching back for a little something extra on my geek fastball?

    Uh?

    Then, in early 2012, some astronomers threw a Pluto-esque wet blanket on the news…Their conclusion: this object is a clump of dust, a cloud, orbiting the star.

    Project Saruman?

    Oh, snap! Just outdid my incredibly dorky self there! That one might take even the most devoted Tolkien fan a while to decrypt.

  7. Jess Tauber

    Come see me- I’m Fomalhautan, and all alo(a)ne…

  8. Shouldn’t the clump or planet be more properly called “The Mote in Sauron’s Eye?”

  9. ctj

    why is it that all of these astro artists get the light source wrong? case in point: the gliese 581 system shows three planets, each illuminated by a different source, none of which is the star depicted (never mind how close they’re depicted in proximity).

  10. andy

    In fact the day after the paper associated with this press release hit the arXiv (arXiv:1210.6620 [astro-ph.EP]), another paper about this object was published (arXiv:1210.6745 [astro-ph.EP]) which came to the conclusion that there is some evidence that the source may be extended, and the hypothesis that the object is a dust cloud from a collision of two exo-Kuiper belt objects is still viable.

  11. Herb King

    Notice the fuzzy shadow of the Enteprise class starship covering the center section from middle to left.

  12. anon

    @andy:

    Some clarifications:

    – neither paper has been ‘published’ but both have been made public.

    arXiv:1210.6620 (the paper in question) has been ‘accepted for publication’ in ApJL and apparently will be published soon.

    The other, which appeared the day after, has its status listed as ‘submitted’.

    Apparently, there will be yet another paper on Fomalhaut b (from the original discoverers?) and probably many more as people argue (again) about this system.

    – the evidence for Fom b being extended is, IMO, very weak. The F814W measurement is the one really driving the ‘extended object’ hypothesis, but the signal-to-noise ratio is so low . Plus, extended emission =/ not-a-planet since in either case the source of the emission is dust, not a planet atmosphere. If the dust cloud is bound to a planet it just has to be smaller than the region defining that planet’s gravitational influence. And at the separation of Fomalhaut b, this region is quite large compared to the object size, whether partially resolved or unresolved. So no problem here.

    – while the first paper’s discussion of models explaining Fomalhaut b’s emission was a bit terse and probably could have used more meat (maybe they are doing a follow-up paper?), the second one completely hand-waves over glaring problems with exo-KBO collision model.

    As explained in the first paper (and apparently ‘declared to be true’ but not detailed in the discovery paper), the dust from any such collision would shear out extremely fast compared to the lifetime of the star (~450 Myr according to the most recent estimate from Mamajek (2012)) and other mechanisms would completely remove the dust in a likewise rapid timeframe. In other words, we would have to be catching this collision ‘just now’.

    Of course, we can’t logically rule this out: nature is weird. But this still strikes me as a rather ‘deus ex machina’ explanation, no less odd than the weird planet accretion + giant ring system idea from the discovery paper but not any more plausible.

    Furthermore, the second authors mistakenly derive the collision rate appropriate for planetesimals *within* the dust belt while Fomalhaut b is located well outside of it. There’s no dust at the location of Fomalhaut b, at least none that we can measure, and the ALMA image (probing larger grains –> better proxy for the parent bodies, i.e. the things that are colliding) is, if anything, even sharper than the HST image. If there’s any ‘unbound dust cloud’ floating around Fomalhaut it should be *within* the ring, not interior to it.

  13. Joseph G

    I wonder if it ever really gets dark on the night side of that planet? The Zodiacal light must be overwhelming. Like a night with low overcast over a brightly lit city…

  14. anon

    “There’s no dust at the location of Fomalhaut b, ”

    Meant to say that there’s no dust *belt* at the location of Fomalhaut b. That is, in the images with good spatial resolution (HST, ALMA) the dust belt lies entirely exterior to Fomalhaut b.

    If Fomalhaut b were within the belt (or, rather, mixed in with the belt’s parent bodies: objects best, though imperfectly, probed by the ALMA data), that would be trouble for the planet hypothesis: it would disrupt the belt unless it were something more like an asteroid in mass, not a gas giant or Neptune-mass planet as various planet-sculpting papers have proposed (http://arxiv.org/abs/0811.1985 and http://arxiv.org/abs/astro-ph/0605372). But it isn’t.

  15. andy

    - neither paper has been ‘published’ but both have been made public.

    Actually being made public is exactly what “published” means – I never claimed it was published in a peer-reviewed scientific journal. And in any case in these current times of “science by press release” it certainly seems that an arXiv publication is more than enough to go on to justify calling in the press.

    As explained in the first paper (and apparently ‘declared to be true’ but not detailed in the discovery paper), the dust from any such collision would shear out extremely fast compared to the lifetime of the star (~450 Myr according to the most recent estimate from Mamajek (2012)) and other mechanisms would completely remove the dust in a likewise rapid timeframe. In other words, we would have to be catching this collision ‘just now’.

    Which is exactly what the second paper is saying, so to claim that it “completely hand-waves” this away is somewhat disingenuous.

    Furthermore, the second authors mistakenly derive the collision rate appropriate for planetesimals *within* the dust belt while Fomalhaut b is located well outside of it.

    They actually do state that the location at Fomalhaut b should be lower – your characterisation of this as that they “mistakenly derive” the wrong rate appears to be a misrepresentation. Furthermore a ring-crossing orbit is not ruled out with the current data (even the 1210.6620 paper notes this), so we could perhaps consider that the collision occurred within the belt.

    To be clear here: I am certainly in favour of the planet hypothesis, however I do not consider this conclusion to be certain at the current time, nor am I a particular fan of somewhat hyperbolic press releases giving the impression that these ongoing debates have been firmly resolved.

  16. anon

    - “Actually being made public is exactly what “published” means – I never claimed it was published in a peer-reviewed scientific journal. And in any case in these current times of “science by press release” it certainly seems that an arXiv publication is more than enough to go on to justify calling in the press.”

    I think the disagreement here is just semantics. My point is simply that neither paper is published in refereed literature, the sense that we use when we say something is ‘published’. Certainly I never think of an article a colleague or I post on the arxiv or on a webpage or on FB, etc to be “published” until it is actually published: as in ‘published in a journal’.

    -“Which is exactly what the second paper is saying, so to claim that it “completely hand-waves” this away is somewhat disingenuous.”

    The reason I call it hand-waving is that the paper estimates how often these collisions occur that may (or may not) produce a detectable dust cloud but doesn’t squarely face how short-lived such an unbound dust cloud is and thus process the fact that such a fleeting event is unlikely to be witnessed/an unlikely explanation.

    – “They actually do state that the location at Fomalhaut b should be lower – your characterisation of this as that they “mistakenly derive” the wrong rate appears to be a misrepresentation. ”

    No. The rate they quote is roughly “one collision every century”. That rate is explicitly derived from their estimate for the mass of the disk ring for collisions *in the ring* and thus describing it as such is not a misrepresentation. It does not tell you the collision rate at the location of Fomalhaut b, where there is no evidence for a belt of debris material. The alternative is that the collision may have happened elsewhere but the debris cloud just now finds its way to Fomalhaut b’s location: that is but a small subset of possible collision outcomes.

    Also, the following statement “Even if the rate at the Fomalhaut b position may be lower, given the approximations of our model the rate is high enough to make such a ∼ 50year-old event plausible.” doesn’t follow: in their derivation, the collision rate scales as the mass of the disk squared. So a factor of 10 reduction in the local mass of planetesimals (a highly conservative inference if the ALMA map is a good tracer of the parent bodies’ locations) yields a factor of ~ 100 reduction in the collision rate at that location. Sure, it is plausible (in fact, almost guaranteed) that at some point in Fomalhaut’s history there were 50 km objects colliding in its debris ring. But that does not mean it is plausible that it just happens to be happening now and where there is no detectable debris ring, instead of where there is such material.

    – “however I do not consider this conclusion to be certain at the current time, nor am I a particular fan of somewhat hyperbolic press releases giving the impression that these ongoing debates have been firmly resolved.”

    Certainly agree. 1210.6620 is far more measured than the media reports imply, and the press/Twitter is just really getting carried away. Fomalhaut b’s status is far from certain and probably will be for a very, very long time.

  17. Messier Tidy Upper

    Great news, great NASA early Halloween clip, great write up here BA. Just wish we could fly there FTL and check out the reality right now. Cheers. :-)

    Guess many (most?) folks reading this will already know but if you wish to see Fomalhaut in the night sky – latitude, weather and time permitting -here’s a hopefully handy tip :

    You can use the first two stars in the front of the Great Square of Pegasus, Markab (Alpha) and Scheat (Beta) as pointers and follow the line upwards (or downwards) till you get to the first reasonably bright star which will be it. Unless one of the planets has moved across the line in Aquarius or Pisces! (Not happening now or for ages I think.)

    The two “back leg” stars of the Square of Pegasus btw (Algenib and Alpheratz /Sirrah) lead similarly to Beta Ceti – Diphda or Deneb Kaitos.

  18. You can use the first two stars in the front of the Great Square of Pegasus, Markab (Alpha) and Scheat (Beta) as pointers and follow the line upwards (or downwards*) till you get to the first reasonably bright star which will be it.

    * For clarity – that’s upwards in the southern hemisphere but downwards in that upside down northern hemisphere! ;-)

    Fomalhaut (& the rest of Piscis Austrinus, the Southern Fish constellation plus Aquarius, Pisces and more) is on the side towards the ecliptic and away from Andromeda.

    October is a good month for viewing Pegasus and Messier 31 and the Andromeda galaxy’s companions.

  19. Nigel Depledge

    James Evans (6) said:

    Click to Annatarenate

    Click to Gorthaurenate

    Click to Necromancenate

    How’s that for sailing away into the geeky blue yonder? For going above and beyond the call of geek duty? For reaching back for a little something extra on my geek fastball?

    Not bad, although that first one might be better as Celebrimborenate, as Celebrimbor was one of the elven smiths who learned the art of ring-making from Annatar (and also, coincidentally, drew the signs in Ithildin on Narvi’s west door to Moria).

    And was your reference to Saruman an allusion to his status as a kind of satellite of Sauron?

  20. Matt B.

    So in what sense is the gas cloud not a part of the planet? It surrounds the planet and is gravitationally bound to it. Wouldn’t you have to say that the planet simply has, due to high temperature, an extended atmosphere?

    It would be interesting if it turned out the cloud didn’t surround the planet, but orbited it like a moon. I doubt that’s possible, because of tidal effects, but it could explain misreadings of velocity.

    I love the modification of the Goddard credit at the end of the video.

  21. Joe

    @ DanM: Follow the link in my name and shoot me an email. I’d be interested in talking about what you have in mind.

  22. flip

    I loved the video! Sadly, I wonder if it adds fuel to the fire of conspiracy-minded people.

    On the other hand, it’s wonderful that a government organisation can have such a great sense of humour!

  23. andy

    So in what sense is the gas cloud not a part of the planet? It surrounds the planet and is gravitationally bound to it. Wouldn’t you have to say that the planet simply has, due to high temperature, an extended atmosphere?

    The modelling indicates a dust cloud, not a gas cloud. If so it probably originates from collisions between objects orbiting the planet, rather than being an extension of the planet’s atmosphere. Such a satellite swarm would be a more massive analogue to the irregular satellite systems around the giant planets of our solar system. See Kennedy & Wyatt (2011) “Collisional evolution of irregular satellite swarms: detectable dust around Solar system and extrasolar planets”

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 »