Sunburned planet turns hot face away from star

By Phil Plait | October 21, 2010 7:00 am

More exoplanet news, and yet another instance where the more we look at them, the weirder they get.

spitzer_upsandbSpitzer Space Telescope sees in the infrared, so in a way it can measure the heat from an object. The orbiting observatory was pointed at the star Upsilon Andromedae, one of the very first stars known to have exoplanets. One of those planets, Ups And b, orbits so close to the star that it makes a complete circle around it every 4.6 days.

That close to the star, the planet should be tidally locked: one side always facing the star, and very hot; the other facing away, and much cooler. As the planet goes around the star, from our vantage point we see first the hot face, then the cooler one, and back again, a cycle repeated every 4.6 Earth days.

Now, we can’t separate the planet from the star; it’s way too close for that. But the planet gives away its presence by the hot and cool faces. We expect to see the hot face when the planet is on the far side of the star from us: it’s then we see the lit, hot face of the planet pointed toward us (the orbit is tilted enough that the star doesn’t get in the way). That hot face gives off infrared light, which adds a tiny bit of infrared light to the total we see from the system. 2.3 days later, the back side of the planet is presented to us. It’s cooler, gives off less infrared, and we see a dip in that light.

By measuring the amount of light, and when we see it, we can infer quite a bit about the planet, like how hot it really is. But astronomers got a surprise…


This graph shows that surprise. It plots the brightness of the system over time, where time is measured in units of the orbital period of the planet. 0 is when the planet’s back side is to us, 0.5 is when the lit side is toward us half an orbit later, and 1 is back to the beginning again.

The dark line is what we expect: it should be brightest in the IR when the lit side faces us, and darkest when the unlit side faces us. But that’s not what is seen. The dots represent measurements of the amount of IR light coming from the planet, and there is a nice sine wave drawn through them. But look: the wave is out of synch with what we expect. It looks like there is a hot spot on the planet all right, but it’s about 80° out of phase with the orbit. It’s as if the hot part of the planet is facing to the side of the star instead of right at it!

The diagram on the top showing the planet might make this easier to see. The hot spot on the planet faces us not when the lit side is facing us, but quite a bit before then. When the lit side faces us, the light actually dims, meaning the hot spot is brighter — hotter — than the part of the planet where its star is shining right down on it.

What the heck is going on? Why would the hot spot on the planet not be directly under the star, blazing overhead just under 9 million km (5.5 million miles) away? Mind you, the star is hotter than the Sun, too, so that planet is getting cooked.

So, after all that buildup, here’s the answer: no one knows. The planet shouldn’t be doing what it’s doing. It’s a mystery. It’s almost certainly a gas giant at that mass, so maybe it’s rotating so rapidly that it sweeps the hottest air around to a spot 80° away from the star. Small offsets from what’s expected have been seen with other planets like this, but 80°? To say that’s difficult to understand is an understatement. Maybe there is something happening beneath the clouds, and there’s a vast convective cell of upwelling hot gas from deep inside the planet. That’s not easy to support either.

Again, the bottom line is, this is a perplexing event. It’ll probably keep theorists and atmospheric modelers at their computers for some time to come. Scientists love a challenge, especially one from out of the blue, so to speak.

You may think this is all well and good, observations of a distant object that’s not even a dot in the sky, but just dots in graph. But think again. Live in the northern hemisphere? Go outside on the next clear night. This time of year, face east a little after sunset, when the sky is dark. You’ll see a giant square in the sky, sitting on one of its vertices like a baseball diamond. That’s Pegasus. Streaming off the star on the left side of that square (think of it as third base) is a parallel curved set of lines made up of three stars each. That’s Andromeda.

Ups_and_mapIf you look carefully — binoculars would help — you’ll see a star between the two curves. That’s Upsilon Andromedae. It’s a star, 44 light years away, but a real star, and orbiting it tightly is a massive Saturn-sized planet, whipping around just above the fires of the star below. That planet too is real, it’s a world, and it’s doing something weird, something unexpected, something at the moment we can’t explain.

We’ll figure it out eventually — we humans are pretty good at that when we set our minds to it. But never forget, that world is real. We can’t go there, we can’t see it, but it betrays its existence to those who know how to detect it. And the more we learn about this hidden world and the odd things it does, the more we can understand about our own world laid out before us.

That’s what science does. And it does it very, very well.


Comments (55)

Links to this Post

  1. Have scientists found an alien city? | October 21, 2010
  1. Grant Gordon

    I’m a little confused, the post seems to indicate that they can detect the planet because of the infrared light it adds to the system. But then goes on to say that the amount of infrared light is out of sync with where the planet is. How do they determine where the planet is in relation to earth to get the expected brightness?

  2. Georg

    That close to the star, the planet should be tidally locked: one side always facing the star, and very hot; the other facing away, and much cooler.

    is that true for a gas giant?
    What is “rotation” in a gas giant, where “wheather” is
    presumably so wild, that Jupiter looks peaceful in comparision?
    Hmm thinking a second time about that, maybe the
    wheather isn’t that stormy. Presumably that planet has
    a lower temperature gradient than Jupiter from core to surface.
    Wheather is driven by such gradients especially if some
    temperature of condensation (gas/liquid) are included in that range.
    The latter maybe is something the planed lacks.

  3. Juergen

    Could it be that the hot side is literally melting? If the hot side of the planet has a lower density than the cold side, that would shift the center of gravity, wouldn’t it? Hence, the planet would turn so that the center of gravity is closer to the star, which in turn would mean that the hot and less dense side turns away from the star – and cools down, while other areas heat up…

    Mind you, just an amateur idea…

  4. Jason

    I skimmed a couple of articles about this and wasn’t quite sure what was going on. Thanks for the clear explanation of the phenomena, even if the why is still unknown.
    It is definitely weird. Is it possible the planet is NOT tidally locked after all? Could the right rotational rate account for this?

  5. Lavocat

    You’ve made me want to go out and buy the best set of binoculars I can find.

    Thanks, Phil. You are always a great way to start out the day.

    The world truly is an awesome place.

  6. Mark

    I might be asking the same thing as Grant Gordon, but how do the astronomers know that they’ve put the zero on the x-axis in the right spot? How did they rule out being mistaken about where in its orbit this planet is?

  7. Adding a third vote to the “how do they know they got the position right” crowd. Could you clear up how they know the planet’s position if they are going off of the increase/decrease in IR?


    Grant Gordan (#1):

    How do they determine where the planet is in relation to earth to get the expected brightness?

    Presumably via the doppler spectroscopy method.

  9. Jens Ayton

    In the denser sampling region, it looks like there’s a second harmonic component. Of course, that could easily be an illusion.

  10. psuedonymous

    I would presume that doppler shift is used to determine the orbital period and location. We wouldn’t be able to tell whether the planet is orbiting clockwise or anticlockwise (i.e. when it is on the ‘left’, and when it is on the ‘right’), but we can tell when the star is approaching (blue shift), ‘stationary’ (no shift), and receding (red shift), and thus when it’s planet is closest and furthest from us. A receding star means the planet is approaching, etc.

  11. Johan Stuyts

    Why does our solar system have to become more boring with every discovery we make? 😉

  12. Is it possible the planet has a large amount of libration that is bringing its hot side to almost 90 degrees of the line of site to the star? i.e., it’s not completely tidally locked but whipping back and forth wildly on its axis?

  13. Messier Tidy Upper

    @ ^ drksky : That sounds like a reasonable plausible explanation to me. :-)

    Great discovery and one that’s got me scratching my head. These exoplanets are strange alright .. and wonderful. Love it! :-)

    @ 11. Johan Stuyts Says:

    Why does our solar system have to become more boring with every discovery we make?

    But ..wait we’re making plenty of fascinating discoveries about our own solar system too! 😉

    I don’t think our system is boring because we don’t have a Pegasid (Hot Jove) type planet or a SuperEarth or Eccentric Orbiter type planet for that matter.

    However, if it is then at least its safe. Remember a Hot Jupiter existing in our solar system would probably mean it migrated inwards from where Jupiter formed so Earth wouldn’t have formed to begin with! 😉

  14. Messier Tidy Upper

    For more on Upsilon Andromedae & its exoplanetary system see :

    for Kaler’s “Planet Project” page on it complete with photographic finderchart & for its wikipedia page :

    is the link to click. :-)

    Plus we already know this system is seriously strange as the BadAstronomer has noted on this blog before :

    Hmm .. I wonder if the weird orbit of Upsilon Andromedae d could have any bearing on this newest mystery or not?

  15. rob

    you are all wrong! wrong i say! it’s a hollowed out moon. no, wait, that’s no moon, it’s a space station! the et’s that run it have to keep it roatating for cooling. if it were tidally locked the space station would be destroyed!11!!!!!11! they are probably gonna steal our tv signals and repost them on the galactic youtube!!11!!!1!! we gotta send some network coporate lawyers out there to put a stop to them stealing earth’s intellectual property!1!!!!!!11!

  16. Joel

    @ 11. Johan Stuyts: “Why does our solar system have to become more boring with every discovery we make?”

    Does it, though? Just suppose there’s a terrestrial planet out there with life on it, and they’re all looking at Sol saying “Wow, there seems to be a big gas giant orbiting Sol, but it seems like it’s orbiting at a distance of more than 800 million kilometers from the star! That’s nearly a hundred times further than Zog is from our own star! Why isn’t our system that interesting?”

  17. Isn’t this roughly what you’d see if the ‘planet’ were actually a very large orbiting power collector? The near side of the collector would be dark, so as to absorb as much light as possible; and the far side would be bright, radiating away the heat, so as to achieve a large as possible temperature gradient to do work with?

  18. Gary Ansorge

    14. rob Says:

    “we gotta send some network coporate lawyers out there to put a stop to them stealing earth’s intellectual property”

    No problem here. Under Galactic Law 41,416,172,subsection A126, all received Em signals from a non space based civilization, herein after known, in Human parlance as a Dyson Type II Civilization, shall be considered in the public domain and may be looted,,,er, utilized for the common good and entertainment. Should the species generating such Em succeed in becoming a Dyson Type II and thus capable of initiating suit for restitution of proceeds from such pillage, form A 177 must be filed by a living representative of said species at Galactic Headquarters in the Perseus cluster.

    We’re awaiting your response.

    Caution: If you succeed in such communication with our representatives, you may be held liable for polluting the interstellar medium with pornographic imagery.

    Gary 7

  19. XPT

    I’m not sold on assuming that any planet close to its star should be tidally locked. Can’t it develop a different orbital resonance? Maybe if it has moons?
    Also, its axis could be very tilted, atmosphere bands could have different rotational periods, rings could screen radiation…
    Our home solar system is weird enough that we should not go out there and assume stuff about our neighbors!

  20. Messier Tidy Upper

    With apologies to Dorothea Mackellar riffing off her ‘My Country’ poem* :

    I Love a Sunburnt Planet

    I love a sun burned planet
    A world of searing gas
    Of mighty spots a-blazing!
    This Pegasids a blast.

    I love this world of fire
    So stormy, hot and wrong
    I love its scalding atmosphere
    That’s driven off head-long

    Core of my system this planet
    Upsilon Andromedae b
    O’ what a world of wonders!
    If we could only see

    For b its spot is shifted from where it oughta b
    Whilst d is far delivered
    In an orbit hard tilted see.
    And c itself like Venus
    the distance that it roves
    And yet tis so much bigger
    Than even Jupiter grows!

    Core of my heart this system
    What marvels there might be
    A system full of mysteries
    Off Pegasus it be.

    I love this sun burned planet,
    A world of F8 star
    So strange it is, so puzzling
    Alas on it there’s no bar!

    No pub, nor likely people
    Its empty, sad to say
    Until we can investigate
    in person, it will stay that way.

    – Stevo Raine aka Messier Tidy Upper

    * A fave Aussie poem sometimes set to music even – and much better than my over tired doggrel natch! 😉

  21. owlbear1

    The magnetic fields between the planet and the star are focusing the star’s corona to that point.

    That would be my guess.

  22. BJN

    How about the planet having a hot layer of atmosphere that’s distorted by the solar wind so that most of that layer is offset away from the star?

  23. ozprof

    My goodness Messier, I have not heard that one since I was in school way back when Noah was building the ark!!!!!! Oh that triggered some memories……….

    Personally though I prefer Bruce Woodley and the Seekers song “I am Australian”.

  24. owlbear1

    Ya know, I’d spend two hundred years as a dis-embodied brain to see that star system.

  25. wurmfood

    Understanding that my knowledge on this comes almost entirely from reading your blog and listening to Astronomy Cast, along with a basic understanding of physics…

    Would the effect of the heat on the planet cause the exposed gas to expand, creating a tidal bulge in the planet? That bulge would slow down that part of the the planet at a consistent rate that should have the bulge behind the line connecting the center of the bodies. I think it would be testable with sensitive enough equipment (and assuming we can see it transit, and the cold side is leading the transit), as we would see the dimming caused by the first half of the planet be less than the dimming caused by the second half of the planet due to the increased size.

    Or I could have my physics all messed up. Neat problem to think about, though.

  26. Alexander Benjamin

    @ drksky I don’t know, it seems to me that a planet with that kind of libration wouldn’t have such a strongly defined hot spot. It would probably be more diffused and smeared out, because any point on the surface facing the star won’t have had enough time to accumulate significant amount of heat than the surrounding region since it’s being rotated through such a large number of degrees by the libration.

    Do the astronomers have any idea on how well defined the spot is? It seems to be that ideas involving atmospheric movements and/or rotations of the planet won’t work fast enough to continuously displace the hot spot without smearing it all out.

  27. andy

    Well given the substantial orbital misalignment between the outer two planets of Upsilon Andromedae A, I guess the inner planet needs to have misaligned weather to fit in…

  28. Peachy

    It’s just a glitch in the Matrix. Now that we’ve noticed, some “scientist” will have to retcon an explanation.

  29. amphiox

    re #11;

    Our solar system has a habitable planet in it! How can that be considered boring?

    I mean, just imagine what it would be like to be observing Sol from Upsilon Andromedae, and finding earth falling out of the doppler signal. . . .

  30. andy

    One possibility that comes to mind here is that we are not necessarily seeing at a constant depth into the planet. It might be that we are seeing higher-altitude layers of the planet on the dayside (perhaps due to cloud formation), but deeper into the planet near the terminator, or something like that. That might help shift the phase curve without needing to significantly displace the actual hotspot.

    Having more data at more wavelengths would be very useful!

  31. flibbertigibbet

    Wow. Thanks, Phil. This is so fascinating! It makes my infrared studies of our own planet pale in comparison. I can’t wait to find out the answer to this mystery.

  32. Anchor

    Super-rotation of the atmosphere: a phenomenon known to be at work on Venus for many decades. Simulations for hot-jupiter-type exoplanet atmospheres typically yield hotspot offsets of 30 degrees. An 80-degree shift is remarkable but an unusually swift super-rotation is still a plausible explanation.

    Diversity of worlds. Variety is the spice of nature.

  33. Jim Baerg

    In our solar system venus has odd high speed winds moving the upper atmosphere around the planet every 4 earth days while the solid surface rotates with a period of over 100 earth days. Perhaps something similar could shift the hot spot on Upsilon Andromedae b.

  34. Peachy

    Excuse my ignorance (I’m just a lowly biologist), but what does “tidally locked” mean on a gas giant, anyway. Upper atmosphere, lower atmosphere, a solid surface (assuming there is a solid one)?

  35. I may have missed the small detail somewhere, but is the hot spot ahead of the sun-side by eighty degrees, or is it trailing behind by eighty?

    I’m presuming it’s racing ahead, since trailing behind seems like it’d be less confusing (with the idea of the heat starting some sort of boil that doesn’t surface until the planet has turned).

  36. Johan Stuyts @ 11.

    Hey. In this context “boring” is good. Old Chinese curse, “May you live in interesting times.” Living in a solar system with one of these would be interesting.

  37. Brian Too

    I marvel that we can detect the the addition of the planet’s reflected energy, in the glare of it’s star. Of course I guess this is a natural consequence of some of the current planet finding methods.

  38. been wondering .. I’m sure this has nothing to do with this phenomenon but what effects does relativity have on all these massive exoplanets orbiting very close to the planet. I know that relativity impacted Mercury causing “the perihelion of Mercury’s orbit precesses around the Sun”. Since they are a lot closer to their suns you would think it would a greater impact.

  39. The hottest time of day on Earth is just before sunset…that is, the Earth is about 80 degrees out of sync, just like this exoplanet! So maybe the same explanation holds: the planet is rotating and has day and night, and the hottest time of day is just before sunset.

  40. JB of Brisbane

    @Messier #20 – Aww, man, you beat me to it!

  41. Corey

    Think of the way that the earth’s ocean tides “follow” the moon; perhaps the atmosphere of the star follows a similar pattern.

  42. Gebraden Kip

    @jared.daniel #40 – The weirdness of this planet is that it is not hottest just before sunset, but just after sunrise!

    By the way, I’m pretty sure the hottest time of day on Earth is not just before sunset…

  43. Kurt

    Almost 90 degrees out of phase? IR reflection by some part of the atmosphere blasted away from the sunward side and swept preferentially to one ‘side’ of the planet by a persistent circulation? For a while there I thought it fit a half period sine wave with larger amplitude, which wouldn’t require circulation. Wild speculation is fun!

  44. Gebraden Kip wrote:
    >The weirdness of this planet is that it is not hottest just before sunset, but just after sunrise!

    I agree that would be weird… but if the planet is spinning the other way, it would be just before sunset after all. Big if, but maybe.

    >By the way, I’m pretty sure the hottest time of day on Earth is not just before sunset…

    Check Not sure if that corresponds to 80 degrees out of synch, but probably not too far off.

  45. Chris Winter

    Fascinating. Quite a bit of wild speculation went through my own mind, but I decided none of it was defensible. I look forward to seeing what the experts come up with.

    JBS Haldane scores again!

  46. AJ in CA

    @19: That’s what I was wondering, particularly if the planet is very uniform throughout.
    I know Mercury is orders of magnitude further out from the sun then this planet, but doesn’t Mercury have a stable 2:3 spin-orbit resonance? Is it possible that there could be resonant islands of stability closer in?

    I remember reading an old sci-fi story about the “coldest place in the solar system,” with the implication that it’s Pluto, before at the end they reveal that it’s on the dark side of Mercury. Apparently it was once thought that Mercury is also tidally locked?

    PS: I also missed the part about it being a gas giant. Can gas giants said to be tidally locked? With that kind of heat, wind speeds are going to be insane.

  47. AJ in CA

    @Messier: You should sing that and put it on Youtube. It could be a new weekly event – Friday Astrokaraoke! 😀

  48. Leon

    Hey, maybe they finally found a planet that acts the way most people think the Moon does–that is, it doesn’t rotate on its axis!

    Sorry Phil, couldn’t resist a jab at that popular misconception.

  49. andy

    Thing is with Mercury is that the 3:2 rotation state it is in is only stable for planets on eccentric orbits with a permanent asymmetry (i.e. solid planets). In fact, Mercury’s current orbital eccentricity is too low for capture into the resonance, but fortunately for the theoretical models it turns out that when you integrate Mercury’s orbit over long periods of time, it turns out that the planet experiences periods with substantially higher eccentricity than the present value, allowing for the resonant rotation to be set up.

    This does not work for gas giants: for these kind of planets the predicted configuration is pseudosynchronisation, where the rotation is a continuous function of the eccentricity (at zero eccentricity you get a 1:1 relation between the orbital eccentricity and the rotation period). Upsilon Andromedae Ab is in a highly circular orbit, so the predicted rotation state is pretty much a synchronised state. Essentially you have to explain the offset via weather and other such phenomena. Various simulations suggest intriguing possibilities for how these planets behave, including hemisphere-scale polar vortices and high-velocity equatorial jets (the simulations seem to disagree on whether we should expect such jets to be prograde or retrograde however).

    Hot Jupiters might seem to be the boring typical result of exoplanetary surveys (primarily because they are easy to find, not because they are particularly common – in fact current results suggest that hot Jupiters are much rarer than long-period gas giants like the ones in our solar system) but there is an awful lot we don’t know about them.

  50. AJ in CA

    Andy: Thanks, very informative post!! You answered several of my questions before I even asked :)

    And I’d been wondering about that multitude of hot Jupiters as well – are they really so numerous, or is it just that an enormous mass orbiting very close to its parent star is easiest to detect by pretty much any method (doppler shift, transit, visual/infrared, etc)?

  51. Messier Tidy Upper

    @48. AJ in CA Says:

    @Messier: You should sing that and put it on Youtube. It could be a new weekly event – Friday Astrokaraoke!

    Ah, if only *I* could sing which I can’t! It would be great if someone with a good voice could try it! Thanks. 😉

    @ 34. Peachy Says:

    Excuse my ignorance (I’m just a lowly biologist), but what does “tidally locked” mean on a gas giant, anyway. Upper atmosphere, lower atmosphere, a solid surface (assuming there is a solid one)?

    Good point. I’m not sure – & I’m kicking myself for not asking that very thing myself now.

    Any solid core here would almost certainly be deeply covered in a Jovian type atmosphere and then a transition to an immense liquid gas ocean .. and under such pressures and temperatures that it would be molten form those alone. Or, perhaps crystalised like diamond a la Clarke’s Space Odyssey vision? Or corroded away by the force of a scouring vastly deep layer of strange liquid metallic hydrogen. Jovian and Superjovian gas giant cores are one big mystery!

    @23. ozprof :

    Personally though I prefer Bruce Woodley and the Seekers song “I am Australian”.

    Cheers. I’ve heard that song before – but not like that by them. :-)

  52. Bryan

    So close to its parent star, would a gas giant be spherical, or might it be tidally stretched due to the pull of the star? That may have some effect on the way the atmosphere heats and cools- I would expect in such a situation that you’d be looking at a much denser atmosphere when you look at the side of the planet. But I’m just guessing, really.

  53. AJ in CA

    @Messier: You’d be surprised. I didn’t think I could sing at ALL until I took beginning (and then intermediate) Voice at the local college. It really helped me enjoy and appreciate other peoples’ singing a lot more, too.
    Unfortunately I’m still pretty shy, and I don’t know of that song, anyway 😛

    I’d love to see some kind of really tough atmospheric probe sent out to study the deep Jovian atmosphere. Like one of those bathyspheres they used to go to the bottom of the Marianas Trench, but smaller and stronger and with modern materials. I wonder how deep it’d be possible to go, or if it’d even be possible to build something that’d survive several million atmospheres of pressure? I believe that’s where you start running into metallic hydrogen and other neat stuff like that.
    The Galileo probe sent us all kinds of interesting data, and that one only functioned down about 150 km past the 1 earth atmosphere equivalent point.

  54. Andy

    Is it possible that the “hot spot” is on the leading edge of the planet, and that because of it’s proximity to the star, is plowing superheated gas before it?


Discover's Newsletter

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


See More

Collapse bottom bar