Hubble spots a chunk of ice 6.7 billion km away!

By Phil Plait | December 16, 2009 1:00 pm

I love stories like this: Hubble spotted a small iceball — only about a kilometer (0.6 miles) across — orbiting the Sun 6.7 billion kilometers (4.2 billion miles) away. And the best part? It was an accident. But it was found on purpose.

hst_kboartThe object wasn’t seen directly; it passed in front of a star, momentarily blocking the star’s light. The star is one that Hubble uses to point the telescope itself; it’s one of several thousand guide stars used to keep the telescope aimed at astronomical objects, a bit like stellar benchmarks.

Hubble’s Fine Guidance Sensors are telescopes used to lock onto these guide stars. They don’t take pictures, but instead use a technique called interferometry to (extremely) precisely measure the starlight. If the telescope starts to drift, even a teeny amount, the FGS will sense this and the motion can be corrected.

If a small object passes in front of a guide star it will momentarily block (what we call occult) the starlight, causing a dip in brightness as well as a diffraction pattern, a fluctuation in the starlight caused as the light bends around the object. Knowing this, astronomers dug through 4.5 years of archived data from the FGSs and found the tell-tale sign of such an occultation.


Using this data, the astronomers were able to determine the distance and size of the object. The distance puts it in the Kuiper Belt, the torus of icy comet nuclei orbiting the Sun out past Neptune. The size is remarkable: the smallest known Kuiper Belt Object (or KBO to those in the know) before this was about 20 kilometers across. This is therefore the smallest KBO ever found! Assuming a typical color for the object, it would be about 35th magnitude, or one-trillionth as bright as what you can see with your unaided eye!

Yowza. That’s also far dimmer than anything directly detected by Hubble, by a factor of about 50.

Again, this object was found by going through 4.5 years of archived Hubble FGS data; however, all the FGS data since launch (in 1990) is available, so it’s possible more events like this are waiting to be discovered. There’s real science in this, too: KBOs this small probably come from the collisions and grinding together of bigger objects; getting a handle on the members of this population will let us know more about the history of the Kuiper Belt and the solar system itself.

Also, it’s good to remember that not everything Hubble sees is on purpose! The fact that all the observations are archived means that people will be able to go treasure hunting for years to come. Hubble is the gift that keeps on giving.

Artwork credit: NASA, ESA, and G. Bacon (STScI)

CATEGORIZED UNDER: Astronomy, Cool stuff

Comments (41)

Links to this Post

  1. Hubble zag kleine Kuiper-ijsbol… per toeval | Astroblogs | December 19, 2009
  1. TG

    Also probably a good illustration of how so much light gets filtered by dust, ice chunks and other matter in the cosmos on its way to our eyeballs!

  2. We humans have created tools for almost as long as we’ve existed. Yet now and then we discover a tool so awesome that even we ourselves never realized fully just what it could do.

  3. Erik

    Very interesting. You might want to clarify that the first image is an artist’s impression of what the object looks like, I went the Hubble site link looking to find out if that was an actual image or not.

  4. Donald

    Dr. Hilke Schlichting did a fascinating public lecture on her work and this discovery last month at the University of Toronto. Great evening.

  5. Gary Ansorge

    WoW! Such a great close up of the comet,,,er,”artists rendition?” Oh well. Never mind.

    Someday, we may actually have ‘scopes capable of such close up shots, hopefully, sooner rather than later.

    Ah, old data. Always something new to discover. I wonder if analyses of the collected SETI data will, in future times, reveal we had already discovered ET but didn’t realize it, because our analyses techniques were just so darned primitive. (future PhD candidate, are you paying attention?).

    Gary 7

  6. Richard

    Doing a quick calculation with Kepler’s Third Law gives it a period of ~300 years. Pretty consistent with a Kuiper Belt Object.

  7. Paul M

    Kudos to the Hubble team for not discarding all this data. It is hard to guess what future analysis may find it invaluable.

  8. What was wrong with the perfectly good word “occlude”, which was already a verb cognate to the adjective “occult”?

  9. Cain

    I assume since we only have the single point we wont be able to discover the orbit?

  10. Scott

    What are the chances of this happening? Also, did was the chances of this happening within expectations given our knowledge of the Kuiper belt? I’m rather curios to know.

  11. Just curious… How do they know it’s 1km in size, and not just the 1km tip of a much larger object?

  12. Nekura

    What is the time scale of the blip? Seconds, minutes, hours?

  13. !AstralProjectile

    I suppose the Terrestrial Planet Finder will find a lot more of these. Fortunately their signature is unique.

  14. Mike

    Can I have a hubble telescope for xmas?

  15. Gary Ansorge

    14. Mike:

    Coitently. Now all you have to do is put it in orbit,,,

    Actually, in a sense, you DO have a Hubble. As do I. It’s the one in which we’re all part owners.

    Gary 7

  16. So, is that about 45 AU?
    miles and km get confusing in the outer solar system.

  17. Rachel

    I had the same question as Ken: how do they know the size AND the distance? What prevents this from being a jupiter-sized object much, much farther away?

  18. Jamey

    I have to third the question:

    We have one observation of an dip in the light available to one of the guidance sensors. From this, they somehow managed to derive the speed *and* the distance, to get the size. Or they managed to get the speed and size, to derive the distance… I’m missing something.

    How do they know it’s all the way out in the Kuiper Belt, instead of a small piece moving through the Asteroid Belt, or was kicked out by Saturn, or any one of a large number of other possibilities? What are the error bars on this thing? For that matter, how do they know it wasn’t a nut in the next orbit out from Hubble? Astronomers love to throw around their standard candles, but a lot of those standard candles seem to be based on measurements on other standard candles with similar large error bars – and don’t error bars multiply?

  19. I am truly amazed and in awe of the discoveries of this incredible telescope! I am so happy that it will continue to make new discoveries for several years to come. However, it saddens me that, eventually, this technological wonder will suffer the fate of many of it’s space-faring ‘brothers’, a fiery death plunge back to its home. I wish there was a way to boost it into a higher orbit, that would keep it safe for 50 or a hundred years. Hopefully by that time we will have developed another way of retrieving it, and bringing it safely home. I think the public could support an effort to at least keep it safely in orbit until we can save it, and bring it back to place in a place like the Smithsonian. How hard would it be to use robotics to attach a booster rocket to Hubble?

    (I suppose it’s going to far to ask if they could use Hubble to find my TV remote! :) Yes, I do know that Hubble cannot be pointed at Earth.)

  20. The HST FGS’s remain not only wonderful interferometers, but
    photometers as well. But, one small point. Phil’s post correctly says:
    “all the FGS data since launch (in 1990) is available, so it’s possible more events like this are waiting to be discovered,” but it wasn’t until HST Cycle 4 that the downlink telemetry format was changed to provide 40 Hz temporal sampling of the FGS guiding data to the ground to allow such serendipitous discoveries. We actually had jumped on that and used those data to search for previously unknown binary stars of very close angular separations of the components (and found lots of them). An overview is here:

    The published paper here:

    And for anyone who wants to dig into the catalog it’s available through Vizer
    as CDS catalog: J/PASP/110/1012

    and all ancillary data here:

    Of course that was 10 years ago, and there is a LOT more data in the HST archive
    that hasn’t been looked at in this way yet since. Hmmm… sounds like a Cycle 18
    archival proposal… There’s gold in them there bits.

    Glenn Schneider (still, and always a “Hubble Hugger”)

  21. Plutonium being from Pluto

    Awesome news & an amazing find. :-)

    I second MichaelL’s (19) idea and also wonder how we can know its distance based on a single observation as Rachel & Jamey (17 & 18) asked & that what we saw was the full object and not just part of something much larger like Ken B (11) said?

    Impressive. Very 8)

  22. Ah, very cool Glenn. So the occultations may not be resolved temporally in the older data.

  23. AnN Marie

    I’m kinda secretly imagining that it is an Imperial Star Destroyer which is about 1km in length. Unmanned though, because it was a long, long time ago….

  24. Bruce

    A chunk of ice floating around in space?!? Now global warming is destroying our whole solar system! Help us, Al Gore! You’re our only hope!!!

  25. Rachel

    OK, I took one for the team….and actually read the published paper. I am not an astronomer, so I may be making up everything you read below.

    1. They were looking for KBO objects specifically, and picked an area of sky where KBOs were likely. They also looked for events lasting about the predicted time length of a KBO occluding a star.

    2. They used both the time of occlusion and the diffraction pattern to estimate size and distance. They also assumed some parameters, namely a circular orbit and an inclination of 14 degrees.

    3. They don’t know for sure this is a KBO object, but judge it’s likely, based on where they were looking and the general event parameters:

    We note that for objects on circular orbits two solutions can fit the du-
    ration of the event. However, the other solution is at a distance of 0.07 from
    the Earth, and is therefore unlikely. It is also unlikely that the occulting
    object was located in the Asteroid belt, since the expected occultation rate
    from Asteroids is about two orders of magnitude less than our implied rate.
    Furthermore, an Asteroid would have to have an eccentricity of order unity
    to be able to explain the duration of the observed occultation event.

    So it’s not for sure, it’s just an educated guess. They extrapolated the rest from there.

  26. CR

    I’m with MichaelL… why must Hubble (like so many other artifacts) be de-orbited instead of pushed to a higher orbit for later retrieval? (I know we can’t safely retrieve it now with our current spacecraft, and I know we can’t just leave it ‘as is’ to come crashing down some random place on its own.) What’s the harm of a higher, longer lasting orbit, other than it posing a hazard to future space navigation? That’s not even such a concern, since it’s not like we wouldn’t know exactly where it was.

  27. Plutonium being from Pluto

    What do we call it? Does this object have a name or even a designation yet? Either official minor planet number or unofficial provisional listing or Xena-like nickname?

    A very minor nitpick & probably a losing fight here but I think the proper name is actually the Edgeworth-Kuiper Belt as Kenneth Edgeworth beat Gerard Kuiper in proposing its existence in via papers in 1943 and 1949 versus Gerard Kuiper’s in 1951. Both men’s ideas of the Edgeworth-Kuiper belt were apparently fairly vague proposals ( I haven’t read them myself but that’s what I’ve read) and have since had to be revised and new ideas – and considerable controversy – considered.

    Just a thought for what its worth :

    Maybe we should term this outer region the ice dwarf belt or trans-Neptunean belt or, (after the asteroid belt), simply the Cometary belt or disk or shell instead? 😉

    For more info on this – and an even more obscure precursor who apparently mentioned it first of all (named Fred Leonard) see :

    The Kuiper Belt website of the guy who co-discovered with Jane Luu the very first Leonard-Edgeworth-Kuiper Cometary belt object 1992 QB1 or Cubewan the original o! 😉

  28. T.E.L.


    The reason is cost. It would be a considerable engineering project to design, build & launch a gadget to couple with Hubble and boost it. I conservatively guess at least one billion dollars. And all for what?

  29. Plutonium being from Pluto

    @ 28 T.E.L. : “For what?”

    Preserving a vital and historic telescope that may even be repaired and used again even if not as the ultimate space observatory anymore?

    Rewarding and showing proper gratitude and respect to perhaps the worlds greatest ever telescope and its team for all the fantastic knowledge and images and joy they’ve brought us?

    Giving our grandchildren something beyond the merely practical but also a sign that we care about history enough to keep it around for them to enjoy too?

    Vintage cars are worth a lot and are treasured honoured antiques that still bring pride and joy to many human hearts, we’ve saved and preserved the old wooden frigate USS Constitituion the old British ship-of-the line HMS Victory (Lord Nelson’s Flagship), until the fire (& I think still after too) the last of the great clipper ships the Cutty Sark and so forth. People invest huge amounts of time and effort on saving, maintaining and rstoring to their original state historic planes and ships and cars and make them a big part of their lives. Emotionally and historically they count for an awful lot.

    Is Hubble different?

    Is the HST worth less effort and care and attempt at preservation for future generations?

    I support CR (25) & Michael L’s (19) call to boost Hubble higher and keep it flying. :-)

    Let’s keep the Hubble Space Telescope around – like good wine and aged whisky and cheese it will rise in value and prestige over time! 😉

  30. Plutonium being from Pluto

    NB. Drat! The link posted in my comment #26 doesn’t seem to work in terms of getting you to the exact page there – seems the top page address thingummy stays the same while you click on other things. (Different form of website set-up or something?)

    So you need to look at the left hand side menu bit, click ‘introduction’ under the capitalised ‘Kuiper belt’ then go :



    Why is it called that? The pronunciation and the history of the name.

    Click on underlined semi-url-y hypertexty thing. That’ll get you to the Why “Kuiper” Belt? page which I was meaning.

    To get to the page / section thingy on the scientists and Leonard Vs Edgeworth vs Kuiper vs Trans Neptunean vs Cometary belt~wise naming & priority issue I was referring to there.

    (Phew! Takes a breath.)

  31. Plutonium being from Pluto

    While astronomers just discovered this tiny world of *cold* ice, we’ve also co-incidentally just at the same time made a really huge discovery of a hot ice & water world twice Earth’s mass and three times its radius – but less dense! This latest low mass Super-Pluto / mini-Neptune style exoplanet is located 42 light years away hugging in closely to the warmth its dim red dwarf sun GJ 1214.

    For more info on this see :

    (& apparently the news is _also_ on CNN complete with an artists illustration too.)

    I’ve already emailed & facebook messaged the BA with this news and hopefully he’ll post about it soon but just thought I’d mention it here too. :-)

    Celebratory over-tired doggrel :

    Big worlds, little worlds, cold worlds hot worlds, worlds for everyone!
    Dense worlds, fluffy worlds, gas worlds, rock worlds, worlds under every sun!

    We’ve found them here, we’ve found them there, round Fomalhaut and Gliese
    We’ve measured transits even photo’ed them – albeit a little hazy!

    A tiny blip, a spectral shift, the merest, faintest trace
    Is all we need to spy a world among the teams that race!

    Artists paint as theorists guess, they follow every finding,
    These fresh found worlds inspire minds to brilliance that’s just blinding! 😉

    (Yeah, I know I’m going to bed now.) 😉

  32. T.E.L.


    Vintage cars are collectible because there are still plenty of them to be had. They’re cheap compared to this sort of thing. It’s even cheaper to preserve the Cutty Sark than to bring down Hubble. You want to spend a billion (or more) just to put Hubble in storage? How about all the other EQUALLY historic pieces of space hardware? Shall we spend huge sums to put all of them in storage? Where does this end? Do you want to at least DOUBLE the cost of EVERY spaceborne instrument for the sake of sentimentality? When do you expect to be bringing Hubble back to the ground? Do you have a good intuitive idea the magnitude of the problem? This would NOT be a piece of cake. We’re talking about putting together a dedicated system just for bringing down one instrument.

    And here’s a newsflash: Hubble isn’t the world’s greatest ever telescope. It’s one instrument out of thousands. It’s heyday is over. The future is coming. Get used to it.

  33. Jamey

    Saving Hubble – this kind of thing is why we *SERIOUSLY* need a space-based operations center, where we could send up the boost module to an orbit near it, have an astronaut maneuver over to it, attach it, and boost it.

    We *NEED* to quit de-orbitting stuff. It cost way too much to get it up there to start with. We really need to collect it and store it so that it can be recycled once we’ve established a real presence in space, and not this minimal, ephemeral presence we call the ISS.

  34. Ken

    @29 Plutonium:

    Well, personally I think it could be done for a lot less than $1B. Maybe a couple hundred mil. Maybe less if it could be allowed to be a “high-risk” mission, with a relatively high chance of not working.

    So which *new* science mission would you propose cancelling in order to free up the funding to do it?

    Or would you like to start passing around the tin cup now? Note you’ll need to raise 4-5 orders of magnitude more funding than to preserve an old sailing ship …

    (Though if someone *did* set up a nonprofit to collect funds for this, I’d happily toss a few bucks into the hat).

  35. Gary Ansorge

    Hubble is at 559 km, which is nearly twice the altitude of the ISS. The ISS loses about 2 km/month in altitude (due to drag from the tenuous atmosphere at that orbit) but atmospheric density drops off rapidly with altitude, so the Hubble loses a bare fraction of that. I expect if left alone, it would take several decades for the Hubble orbit to decay enough to re-enter and burn up. I don’t have exact figures, just gestimates. Still, our cost to orbit is dropping and I expect if they just left Hubble where it is, long before it would absolutely need to be brought down, we’d have a very cost effective means to access it and move it to a “historical” orbit. The Vasimir 200kw thruster is a very real possibility for boosting objects to higher orbits and maintaining them for very long time frames.

    GAry 7

  36. T.E.L.


    For the foreseeable future it would be irresponsible to leave Hubble where it is on faith that in a few years launches will be dirt cheap. Do you really think that before it comes down on its own, it’ll become downright affordable to just go up and just do what we want with it? And what is a “historical” orbit? Is it the kind where tourists will be going up to see it on weekends?

  37. Edward Carney

    The OED has interesting info on occultation vs. occlusion.

    The term occlusion originally came into the written language in the mid-16th C. to denote the blocking or obstruction of an orifice, tube, stream, etc. So navigation on the Mississippi was subject to occlusion (1786) as were the bronchi (1881).

    The term “occultation” for the disappearance of a star or planet has been in use since 1556. Originally it referred to the object’s disappearance in the light of the setting or rising sun. Later, it became a more general term for the concealment of a celestial object by another (1601). This word was also used in the 14th C. in Old French to mean the same thing.

  38. Asimov Fan

    As I understand this – & I could be wrong – they need to send a mission up to add a “de-orbiting” pack to Hubble anyway.

    If so then wouldn’t it be equally cheap /expensive to send a “boost it into a higher orbit where it’ll stay for a few hundred years so they could preserve it for the future” pack instead? I wonder …

  39. T.E.L.

    Azimov Fan,

    Not necessarily. Deorbiting only requires one deta-v of just enough to bring its perigee down into the atmosphere. Boosting it higher means two delta-v’s, one to raise its apogee and another to raise its perigee (to circularise the orbit). Two are more expensive than one.


Discover's Newsletter

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


See More

Collapse bottom bar