Stellar tantrum blasts away part of its planet's atmosphere!

By Phil Plait | June 28, 2012 9:45 am

Life must suck for HD 189733b.

It’s a planet orbiting the star HD 189733, about 63 light years from Earth. It’s similar to Jupiter, being slightly more massive and slightly bigger. Unlike our own big brother, though, HD 189733b is far closer to its parent star, orbiting just about 4 million kilometers (about 2.5 million miles) above its surface! That means the cloudtops of the planet are at a scorching 840°C (1500°F), so hot that the atmosphere of the planet is getting blown away by the star itself, creating a comet-like tail of gas escaping from the planet.

And now, adding insult to injury, astronomers have discovered that the star is prone to cosmic hissy fits — and this may actually blow torch even more air from the planet.

[Illustration by NASA, ESA, and L. Calçada]

This pretty nifty: the researchers targeted the star and planet using Hubble in 2010 and didn’t see anything amiss. In 2011, though, they observed it with Hubble again, but also used NASA’s Swift satellite, which is sensitive to high energy emission like extreme ultraviolet and X-rays. They happened to be looking when the star erupted in a massive flare, quickly quadrupling its brightness in X-rays alone. And because the planet is so close to the star, it took the full brunt of this event, its atmosphere puffing up and actually getting blasted away into space by the fierce light from the star!

Just a few hours later, as seen from Earth, the planet passed directly in front of its star (like Venus did during its transit earlier this month). The atmosphere by this point was really streaming away from the planet in the aftermath of the flare, and was also blocking a bit of the star’s light. In the Hubble data the astronomers not only detected that absorption, but they could measure it to see how much hydrogen the planet was losing. It turns out something like 1000 tons of hydrogen was screaming away from the planet every second! And that’s a lower limit; they could only detect neutral hydrogen — that is, atoms that still held on to their electrons. There was probably plenty of ionized hydrogen they couldn’t detect.

NASA put out a short video that explains this as well:

HD 189733b is actually the second planet discovered to be losing its atmosphere due to its close proximity to its star — HD 209458b is the other — but this is the first time changes in the atmosphere have been detected. We’re really getting good at examining these planets! HD 189733b has been seen to have hydrogen, water vapor, carbon dioxide and carbon monoxide in its air (see Related Posts below for more on that), and we’ve been able to literally map the distribution of heat in its atmosphere.

Incredible. It’s amazing to think that we’re watching planets boil away as they get blasted by their stars! A third planet like this has been found that’s not a gas giant, so it really is boiling, its surface a seething cauldron of molten materials.

The Universe is a crazy place. And the more we learn about it, the crazier stuff we find. I’ll note that the Hubble observations were made using the Space Telescope Imagine Spectrograph, a camera I worked on back in the day. We used it to observe stars, supernovae, black holes, asteroids, and even HD 209458b in fact. But I never thought it would one day be used to watch as a planet slowly evaporated into space.

Crazy. And so, so amazing.

Related Posts:

Sunset on an alien world
Hubble spies carbon dioxide 600 trillion kilometers away!
Hubble smelt who dealt it
Astronomer make first map of extrasolar planet!

CATEGORIZED UNDER: Astronomy, Cool stuff, Pretty pictures

Comments (15)

Links to this Post

  1. Guerrilla Monkey – Today’s bookmarks | June 29, 2012
  1. SkyGazer

    Awesome observation!
    And indeed another example that you never can look up enough.

  2. Chris A.

    Given how many exoplanets have been discovered to be suffering a similar fate, one has to wonder how many have come and gone long before we gained the tools to catch them in their death throes. My hunch is that their deaths don’t span a terribly long time relative to the age of the Milky Way. (One more factor for the Drake equation: Average lifespan of the planet itself.)

  3. Chris

    Just to put in perspective if the planet was outgassing 1000 tons per second all the time, it would take 200 billion years to lose one Earth mass. Of course the rate is not steady and other mass is being lost but conservatively that planet will be around for quite awhile.

  4. Wzrd1

    Now, THAT planet is a SERIOUS wet sauna!
    Water was also detected from this planet.
    Considering the relative youth of the star, one ponders what its mass was before the star began to seriously erode the atmosphere!
    Perhaps it was a close contact binary that lost mass and migrated out enough to be where it is today.

    @Chris A, I seriously doubt that the Drake equation can be considered, due to the proximity to the star making the planet or any satellites FAR inside of the habitable zone.

  5. ChazInMT

    Oh Cripes!!! I hope this does not get out to the general public because they most certainly will start asking if this could happen to Earth if the Sun has a big solar flare.

    The relationship numbers here are Mind Boggling though. If this planet was in our solar system and our Sun was an 8.5Inch diameter bowling ball, Earth is 78 Feet away from the “Sun”, Mercury is 30 feet away. The subject planet here being blasted is 18 INCHES AWAY! at this scale. Put a 1inch marble 18 inches away from the bowling ball Sun and ponder how is it able to remain there for even a day? And, the planets Star is getting 4 times brighter….When has our Sun even gotten 5% brighter? They are freakishly close distances combined with freakish instabilities to make this happen. But all people will see out of this is that a Sun can blast a planet when it wants to.

  6. Chris

    @5 ChazInMT

    Our Sun is always eroding its planets’ atmospheres. It’s just nowhere near as dramatic as that.

    And, the planets Star is getting 4 times brighter

    Important point, it was 4 times in the x-rays. The difference between a B-class and X-class flare is 1000 times. Although as I look at the spaceweather site they say this is equivalent to an X three million flare. Our puny Sun had only a X-20 something.

  7. scgvlmike

    I’m curious to know about how much mass the planet absorbs from its parent star as a result of normal outgassing from the star over the same time period, when not being blasted with flares (i.e., the particles streaming out as solar winds in our system, but ” HD 189733ar” is a bit of a mouthful)

  8. @Chris (#3):
    I found a paper on arXiv(click on my name for the link) suggesting that some exoplanets being blasted by their stars might be completely gone on a time scale of ~1 Gyr.

    @Wzrd1 (#4):
    While I agree that a planet so close to its star is not considered a good candidate for habitability, I’m always cautious when it comes to making assumptions about what constitutes “habitable,” given that we are attempting to extrapolate from a rather small sample set of n=1 (i.e. all the planets which are known to harbor life). I mean, who would have guessed, 30 years ago, some of the ridiculously inhospitable places on Earth that are now known to harbor life? Also, note that the Drake equation in its original form makes no assumptions about what constitutes habitability of a planet beyond its association with a star, and there are some who have speculated that even “rogue” planets floating freely through the galaxy far from any stars might possess mechanisms for maintaining temperatures conducive to life.

  9. Kevin I

    Three words. Cosmic hissy fits. Band name!

  10. According to Wikipedia, the HD 189733 system has an official Variable Star designation: V452 Vulpeculae. (It’s a BY Draconis variable, meaning the changes in its brightness are due to the presence of sunspots [starspots] as the star rotates.)

    What’s odd, though, is that the spectrum for HD 189733 A (it’s a binary star; the planet is orbiting HD 189733 A, and HD 189733 B is suspected to be a red dwarf about 200+ A.U. away from the primary) is plain old spectral class K1 V, and is only absolute magnitude +6.2 (about a quarter to a third as bright as our Sun).

    Is it some kind of oddbal orange flare star? Why are its UV and X-ray emissions on the order of 20 times more powerful than the Sun’s?

  11. How can planets form so close to their sun? I would think that the immense gravity of the sun would have prevented the disk of dust and gas from coalescing. This is how the asteroid belt in our solar system formed right? The gravity of the outer gas giants disrupted all the asteroids from joining together.

  12. Tribeca Mike

    “This pretty nifty”? Try telling that to the residents of HD 189733b.

  13. MaDeR

    @Eric: simple – it did not form that close. It formed far away and migrated closer.

  14. Gary Ansorge

    11. Eric

    “How can planets form so close to their sun?” They probably can’t.

    The latest model I’ve seen for planetary formation suggests that planets tend to form far from the parent star and then migrate inward, either because of interactions with other planets in the system or due to external perturbations (such as an encounter with another star system).



Discover's Newsletter

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


See More

Collapse bottom bar