≡ Menu

Crowd-Funding the Exomoon Hunt

I’ve been trying to figure out why exomoons — moons around planets that orbit stars other than our own — have such a fascination for me. On the purely scientific level, the sheer amazement of discovery probably carries the day, meaning that I grew up in a time long before we had confirmation of any exoplanets, and now we’re talking about getting data on their moons. But there’s also that sense of the exotic, for we can wonder whether gas giants in the habitable zone, which may be more plentiful than we realize, might have life on their own rocky moons.

David Kipping (Harvard-Smithsonian Center for Astrophysics) has been a key player in the exomoon hunt for some time now (search under his name in the archives here and you’ll retrieve articles going back for years). David is now working with a ‘crowd-funding’ source called Petridish.org to fund a new mini-supercomputer that will go to work on the Hunt for Exomoons with Kepler (HEK) project. The idea behind HEK is to use Kepler data to look for transit timing variations (TTV) and transit duration variations (TDV), perturbations in the motion of the host planet that should flag the presence of a large exomoon. The detection of exomoons down to 0.2 Earth masses seems feasible with these methods, as Kipping has determined in earlier work.

Help us find the first exomoon is getting plenty of attention. The beauty of Petridish.org is that it lets individuals become a part of science one project at a time, playing an important role in the kind of things that get funded. Have a look at the site and you’ll see a wide range of projects ranging from a study of wolf populations on Isle Royale National Park (Lake Superior) to the collection of rock samples in Antarctica. Each project has a short video explaining the work at hand and the funding goal, along with the rewards for donors, which could be souvenirs of some kind or, for large donations, having the project named after the donor. Needless to say, backers are also on the fast track for updates on the research.

With over 1000 new planetary candidates just released by Kepler, the exomoon possibilities are getting more and more interesting, but Kipping points out that hunting for a single moon takes about 6 years of computer time:

Searching for moons requires the most sophisticated statistical techniques, many of which we have borrowed from cosmologists studying the Big Bang and dark energy. The systems we model have complex dynamical interactions and produce strange, asymmetric light curves requiring a lot of computer power. But we are *almost* there. A mini-supercomputer would have a huge impact on our search, so please do consider supporting us!

The fund-raising project still has eleven days to run and is making excellent progress. But faster computer processors would bump up the speed for HEK’s work, and with almost two weeks to go, Kipping is hoping the project can not only acquire the needed machine but upgrade it to state-of-the-art standards. Have a look at what HEK is doing with crowd-funding, and be aware, too, of Kipping’s paper “The Hunt for Exomoons with Kepler (HEK): I. Description of a New Observational Project,” available on the arXiv site and now accepted for publication in The Astrophysical Journal. For more on HEK, see the Centauri Dreams post New Exomoon Project Will Use Kepler Data.


Comments on this entry are closed.

  • djlactin March 22, 2012, 10:11

    This is astonishing! In my lifetime, we have gone from looking at stars to determine whether they show Doppler shifts that indicate the presence of a planet, to looking for wobbles in the Doppler shift that indicate moons around them. Who knows what people of my childrens’ generation may find. Would that I could live for a millennium!

  • James D. Stilwell March 22, 2012, 12:12

    A gas giant in a habitable zone with a rocky earth like moon?
    Now that’s very interesting. More than interesting. JDS

  • Rob Flores March 22, 2012, 12:50

    Looking at extra solar Moons with biocapable enviroment we are
    confronted with some restrictions, some obvious, others not so much.

    Habitability Zone

    Orbital Period around the host planet, would seem like a simple issue, but in the details it’s not. For a large moon with more land than seas, only a close -in orbit will work otherwise temperature differential will create massive storms/dust storms. This moon would really have to be close in to avoid this effect. And because of tidal effects the moon – planet mass ration could not be too high, (or you get a milder form of Io) I think something like
    1/2 Neptune mass & Earth mass might work.
    However if you have Moon with more than 75% water then you can slide
    the orbit of the moon around the planet further out if tidal forces are too
    much, and use the oceans as buffer a against the temperature differential.
    I still see storms but maybe only with Gale force winds, as opposed to near supersonic hurricanes.

    Are we capable of finding high reflectivity Moons(follow the water ) that are near Earth size with the instrumentation available today? Probably not at the longer Kepler distrances of >700LY. But What about closer stars.?

  • Mike March 22, 2012, 12:53

    I’m also fascinated with the HEK project. As Paul mentions we’ve seen the list of exoplanets grow and now were on the edge of seeing the detection of exomoons. What’s next? If we could only live for a millennium indeed!
    Please take a moment to check out David Kipping’s HEK project. I’m sure most people will find it an interesting project and it offers a chance for people to be involved in the historical first detection of an exomoon.
    The HEK project has made great progress with it’s funding but some more donations would enable the purchase of the fastest processor option for the mini supercomputer.
    Considering the amount of data that it will have to crunch the faster the better. I made my contribution by just cutting back on the beer budget for a little while. Every little bit helps.

  • FrankH March 22, 2012, 15:34

    Unless that gas giant captured an already existing terrestrial planet in the star’s HZ, its moons will be very small, rocky and airless.

    In the very, very unlikely event that it managed to accrete an Earth-sized moon BEFORE it spirals in, that Moon will be mostly ices… many of those ices are gases at HZ temperatures, so you still end up with a much less massive, airless moon.
    I’m not saying that it’s a waste to look for exoplanet moons, because it’ll produce new techniques for mining the data, but the chances of finding a habitable moon are pretty close to zero.

  • Daniel March 22, 2012, 18:03

    FrankH @

    I disagree with you FrankH

    When the scientist found the First exoplanet around 51 Pegasi in1995, no one could imagine that hot jupiters was possible to exist.now we know that many out there.it’s not the majority but they exist.

    Scientist thought,that multiply star system it’s unlikely to have planets ,and today we know many multiply star system has planets, probably most of them

    By analogy, I think that the same can happen with earth-size,earth mass exomoons around gas giants , Perhaps can be plenty of earth-size exomoons out there.

    And if are plenty of earth-size,earth-mass exomoons out there, I want see,what good theoretic explanation they will be give for it

    I think that, before jump to conclusions, let’s wait for the observational data.

    The universe it’s full of surprise.

  • coolstar March 22, 2012, 20:58

    FrankH makes some good points above, I think. It’s not even clear that gas giants systems can form moons with enough mass to hold atmospheres in the HZ (some recent theory leads one to think that maybe they can’t). I think this is a cool project, but the chances of finding a Pandora are probably a small fraction of the seemingly already small chance of finding a terrestrial sized planet in the HZ. Who knows, but I’d guess in the range of 0.1 to 0.0001 as likely as a terrestrial sized planet, with 0.1 being VERY optimistic…..

  • Martin J Sallberg March 23, 2012, 4:46

    Why only be able to contribute by donating money? Why not do like SETI, let people contribute by downloading raw data and letting their computers do it? Many small computers can be as good as one big.

    Habitable moons around extrasolar gas giants would certainly increase the number of potential habitats out there. It may be more common than current models predict, just keep observing.

  • Mike March 23, 2012, 10:07

    To Martin, some one asks the same question at the Petridish website. David Kipping’s answer is “the tasks can’t be efficiently broken up into small chunks. They require multiple CPUs with very fast communication between them.”
    And there are other reasons this job doesn’t lend itself very well to distributed computing. There is more information available at Petridish.org.
    about that and other project details.
    Exomoons are a fascinating subject and certainly deserve more observing and modeling.

  • FrankH March 23, 2012, 15:09

    There’s a good article describing the limits to a moon’s size:
    Robin M. Canup and William R. Ward “A common mass scaling for satellite systems of gaseous planets” Nature 441, 834-839 (15 June 2006)


    Simulations match real world examples (our gas giants) very well.

    The chances of an Earth sized moon forming with a gas giant and surviving as an Earth-sized moon after its migration to the HZ are vanishingly small; you’ll have better luck finding a unicorn.

    The other option, capturing an existing Earth sized planet in the star’s HZ is not much better; from the Kepler data it appears that Earth-sized (0.75 – 1.25 the size of Earth) planets in ** their star’s HZ ** are very rare, so now you have two unicorns to hunt down.

  • Ole Burde March 23, 2012, 15:58

    Our own solar system is probably a rare case , for reasons we still dont understand . Hot jupiters belongs in a more “normal” system where things might be radically different , but if life somehow got started in the otherwise TEMPORARY stage of a waterrich moon in the habitable zone , it could have changed the rules of this moons geophysical development as it did on earth . An active ozone layer produced by fotosytesis and other biorelated chemical feedback mekanisms is maybee the reason earth has been able to hang on to its water .
    If we considered the probabbiliy an earthlike ecology system being stable for billions of years , without knowing that it actually existed on earth , it would appear close to zero . On an exomoon even less.
    it could be , that life exists on a great number of these exomoons , but that it would never last long enough to produce more than bacteria . For human traveloors , an expected stabilityperiod of 50 millyears more , would be more than enough .

  • Rob Henry March 23, 2012, 16:45

    FrankH is right about theory seeming to make large captured moons extraordinarily rare. So rare in fact that the chances of our system having one can be discounted… yet Triton exists.

  • Daniel March 23, 2012, 18:38

    By the way

    what about earth-size/Earth mass planet formation around Brown Dwarf? what is the prospect for it ? Is It the same for earth-size/Earth-mass exomoons around Gas Giants? (of course if the theoretical model hold true about exomoon formation around Gas Giants by the futures observations, not matter if this article is from nature still a theoretical model that need to be prove by observations )

  • Rob Henry March 23, 2012, 19:12

    Theories that model the mass of the moon systems that form in place around giant planets have long placed their likely mass as 0.01% that of their planet. Given our state of actual exomoon data, I put it to you all that we should place more emphasis on the upper limits of these models, which has typically been about twice this. Thus, given the number of planets detected of two or more Jovian masses, it would not destroy our current theories to find a large number of exomoons heavier than Mars.

    That would still be very exciting, but I have a horrible feeling that exomoons might have to be bigger than that to be detected by this method.

  • Daniel March 23, 2012, 22:32

    “So let’s go finds this unicorns lol any way if they don’t look for observational data, we will never know

    well about Earth-size planet in their star’s HZ ,I still suspect that hidden out there,the Stars in the Kepler field it’s too noise,we will need more transit to confirm a earth-size around a G dwarf type of star,and any way if Kepler was look for Venus,Earth analog,Kepler wouldn’t detect none, and as for cool stars, there are few cool stars in the Kepler field,for example M dwarf stars in the Kepler field are about 3000 stars,and maybe parameters for the cooler stars is incorrect, like this publish Article:

    Near-Infrared Spectroscopy of Low-Mass Kepler Planet-Candidate Host Stars: Effective Temperatures, Metallicities, Masses and Radii


    if this really hold true,there are a least 7 Earth-size planets around M dwarf stars

    KOI 448.02 (M0-V Primary) — Radius 1.85 Earth — 240 K — Year 43.62 days
    KOI 463.01 (M3-V Primary) — Radius 0.93 Earth — 232 K — Year 18.48 days
    KOI 812.03 (M0-V Primary) — Radius 1.16 Earth — 228 K — Year 46.19 days
    KOI 947.01 (M1-V Primary) — Radius 1.24 Earth — 254 K — Year 28.60 days
    KOI 1361.01 (M0-V Primary) — Radius 1.58 Earth — 232 K — Year 59.88 days
    KOI 1422.02 (M2-V Primary) — Radius 0.85 Earth — 249 K — Year 19.85 days
    KOI 494.01 (M1-V Primary) — Radius 1.05 Earth — 268 K — Year 25.70 days

    And this isn’t bad number of earth-size HZ planet candidates for a small simple of 3000 M dwarf stars

    Spitzer telescope right now try to validate a least 2 of this candidates,as you can see on this Kepler Conference presentation.

    some of them (like the KOI 947.01 and KOI 1361.01) are under way to be validate by the Spitzer

    http://connect.arc.nasa.gov/p1jngti8c8g/?launcher=false&fcsContent=true&pbMode=normal (Validation of Habitable-Zone Super Earth Kepler Candidates with Warm Spitzer.)

    I don’t see why many people so pessimist,still many Kepler data to be process.

    I’m still optimist “

  • Eric Goldstein March 24, 2012, 1:59

    An Earth-mass exomoon may have already been found.

    Here’s an abstract to a paper recently presented:

    MOA-2011-BLG-262Lb: Free-Floating Planet with an Exomoon or Planetary System with Halo Kinematics
    Prof. David Bennett
    University of Notre Dame

    Abstract: MOA-2011-BLG-262 is the shortest duration microlensing event to date with a planetary mass companion, and the lens-source relative proper motion is unusually high. The high relative proper motion suggests a nearby lens, but if this is the case, then the primary must have a mass of a few Jupiter masses. This would imply that the secondary is an exomoon of about an Earth mass. The other alternative is that the velocity of the planet host star is quite large, implying that it resides in the Milky Way’s halo. I will describe what the current and future data might say to resolve this issue.

    The abstract is here:

    I read about this work at the Extrasolar Visions II forum:

  • Ole Burde March 24, 2012, 8:15

    If no exomoons can be found with the present data , it ‘s just another reason to get more and better data .To put things in perspektive , the whole Kepler program is a cheap thing as compared to the James Webb telecope , and if this one cant do the job that REALLY interrests the general pulic ( as opposed to looking-for-god-in-the-sky astronomers) then something else must be done .
    A bigger , better and much longer lasting Kepler 2 could be designed to the minimum specifications where it could detect exacly the kind of exomoons we would expect acording to the 0.01% relationship .
    A less than Mars sized Exomoon would probably not be biologicly stable for billions of years like Earth , but if we found hundreds of them a few of them could be in their life bearing period .

  • coolstar March 24, 2012, 14:32

    The Canup and Ward paper that FrankH linked to above (and I was too lazy to find) gives ratios of moon to giant planet masses of 10^-4 NOT 10^-2, which would limit exomoons forming with their hosts to being Mars sized, at best. A Mars sized object in the habitable zone MIGHT be able to hold on to an atmosphere sufficiently long, and might not….The Kipping project is still worth the effort, though it’s quite unlikely to find an unicorn. It will get great results on transit timing variations though, and is worth the modest funding required for that alone.

  • Ole Burde March 24, 2012, 18:11

    “A Mars sized object in the habitable zone MIGHT be able to hold on to an atmosphere sufficiently long , and might not ”
    Long enough for what ? If an ocean of liquid water can exist for a hundred million years , life might get started no matter the atmosphere looks like . And if life takes hold , it will probably produce an oxygen atmosphere , including an ozone layer , that will be capable of protecting itself AND the remaining water for a much longer period than would otherwise be expected . Mars might have been a habitable planet for almost a billion years…thats perhabs 15 % of the solar systems life time , so if we looked at a random moment , there would be a good chance .

  • coolstar March 24, 2012, 22:59

    @Ole Burde I don’t think the ozone is as important as a strong magnetic field, which the moon itself probably won’t have as it’s likely to be tidally locked. It’s parent may have a VERY strong field however, which could help a lot (and possibly hurt). A billion years isn’t very long given our one data point on the subject: the earth only had (apparently) single-celled life at that stage….

  • Ole Burde March 25, 2012, 18:22

    @ coolstar
    “Researchers once believed that small bodies like Jupiter’s major moons could not possess such strong fields at all. But these moons orbit deep inside Jupiter’s own powerful magnetosphere. According to models developed by Graeme Sarson (University of Exeter) and his colleagues, a strong ambient field helps initiate the circulation needed to produce a vigorous dynamo effect in the core of an even slightly active moon, leading to a strong field for the moon itself. Taken together, these observations and models hint that planet-size moons can maintain protective magnetic fields that they would not have in isolation.” (from Andrew J Lepage ., Sky&Telescope)
    The tidal locking will give such a moon a very long day , from 2 to 30 times onger than ours . This would not prevent life from starting in a bathtup-temperature ocean .
    If Mars had been a moon , it might have kept both its magnetic field and its tectonic activity runing for a longer period .

  • coolstar March 26, 2012, 19:10

    @Ole Burde I suspect the idea that the planet’s magnetic field would stimulate a dynamo effect in the moon itself is highly speculative; especially given that we don’t understand how such dynamos actually work! Protection by the primary’s magnetic field we DO understand, at least somewhat better, though it’s possible that interactions with it could actually be destructive to a moon’s atmosphere (by generating super auroras and huge currents, etc). Looking for a habitable exomoon is certainly doable, as Kipping’s excellent work has proven, FINDING one may be an entirely different story……I’ll stick with my original guess of 0.1 to 0.0001 as likely as finding a terrestrial sized planet in the HZ (though trojan planets might also be found). There might be a better chance of actually finding a binary planet, though none of these options are likely to increase the total number of habitable planets appreciably. Note also that the earth most likely didn’t even have an O2 atmosphere after even a billion years…..

  • FrankH March 27, 2012, 14:03

    @coolstar – if Mars were in a HZ, it would just barely be able to hold on to nitrogen and oxygen over geologic time and it would loose its water vapor. If its density were closer to Earth’s (or Mercury’s) it would be at the cusp of loosing its water vapor, but could retain an atmosphere.
    A Mars sized planet with the typical density of a gas giant moon would quickly become a much smaller, airless, rocky world.

  • Ole Burde March 27, 2012, 15:58

    We dont really have an idea about how many earthlike planets there are yet , it could be 0.000001% or 5% , both would fit the data well enough .
    To compare an unknown to another equally unknown is not the best way to reach conclutions . A more logic way is to compare the extra costs of including moons in the search , to the procect as a whole . So far it has been only a small addition in cost ,to cover an aditional possiblity . There are no reason to rule out moons before the frequency of earthlike planets is solidly established to be bigger than our worst fears…

  • Rob Henry March 27, 2012, 19:11

    FrankH, makes a good point with his low density objection to holding an atmosphere, but let‘s not overplay it. A body half Martian density would need 4x its mass to have the same surface gravity, but note that it only needs 1.4x its mass to have the same escape velocity.

  • coolstar March 27, 2012, 20:07

    @FrankH Yes, I agree totally.
    @ Ole Burde If the eta earth is 5% then there’s likely less than one habitable exomoon in the Kepler data base…..I’ve always said the search is worthwhile, but liking hunting for an unicorn (thanks, FrankH) it’s likely to be just as successful. Eta earth is NOT completely unknown, by the way, it’s almost certainly less than 0.10.

  • ljk April 4, 2012, 23:51


    Outer-Planet Moons Found — and Lost

    The orbital positions of more than a dozen small moonlets around Jupiter and Saturn are so uncertain that they are effectively lost.

    Prior to the year 2000, Jupiter had 17 known moons, about half of which moved in “irregular” orbits that were highly inclined and eccentric and often traveling in reverse (retrograde), with respect to Jupiter’s spin. Saturn had 18 moons in all, though only Phoebe occupied an unusual orbit. Jupiter’s irregular moons had paths that seemed to be clustered, suggesting that they were related — likely chunks set loose during long-ago collisions.

    Discovery images, taken in 2001, of a 22.1-magnitude moonlet around Jupiter that’s just 2½ miles (4 kn) across. Spotted as part of a survey for new Jovian moons, this one — later named Hermippe — didn’t get away. But some of its siblings are now considered lost.

    S. Sheppard / D. Jewitt
    To track down more members of these oddball families, University of Hawaii graduate student Scott Sheppard and his adviser, David Jewitt, started an ambitious project to find new irregular satellites around Jupiter and Saturn. Using the big guns on Mauna Kea — first the university’s 2.2-m reflector and the 3.6-m Canada-France-Hawaii Telescope, and later the 8.2-m Subaru telescope, Sheppard and Jewitt were incredibly successful. In a single 2003 article, they announced the discovery of 23 irregular Jovian satellites, clustered into five distinct dynamical groups.

    Meanwhile, another observing group, led by Canadian astronomers Brett Gladman and JJ Kavelaars, had likewise been quite successful tracking down new outer-planet satellites, particularly around Saturn. NASA’s Cassini orbiter also chipped in a few Saturnian moonlets after its arrival in late 2004.

    The great majority of Jupiter’s 67 known moons travel in retrograde orbits, meaning they travel in the direction opposite the planet’s spin. Click on the image for a larger view.

    Scott Sheppard
    The upshot of this finding frenzy is that the count of planetary satellites now stands at an amazing 170: Earth has the Moon, Mars has 2, Jupiter 65, Saturn 62, Uranus 27, and Neptune 13. Among the dwarf planets, Pluto has 4, Eris 1, and Haumea 2.

    However, just because all these moons have been found doesn’t mean astronomers actually know where they are right now. Dynamicists Robert Jacobson and Marina Brozović (Jet Propulsion Laboratory), along with four collaborators, have analyzed the uncertainties of some 100 small outer-planet satellites whose orbits are distant, highly inclined, and eccentric.

    Their surprising conclusion, summarized here and detailed in an article submitted last month to Astronomical Journal, is that the locations of 10 Jovian and seven Saturnian satellites are known so poorly that they are effectively lost. “We have very little idea of where to point a telescope to observe them,” the authors conclude. “A new survey of the planetary environs will be required to find them.”

    “Lost” Satellites with Highly Uncertain Orbits


    Mean distance (106 km)

    Eccen- tricity

    Inclina- tion

    Period (years)



    2003 J 12






    2003 J 3






    2003 J 15






    2003 J 10






    2003 J 9






    2003 J 5






    2003 J 19






    2003 J 23






    2003 J 4






    2003 J 2







    1007 S 2






    2004 S 13






    2006 S 1






    2007 S 3






    2004 S 17






    2004 S 12






    2004 S 7






    Source: R. Jacobson & others

    Moreover, the team finds that 11 additional outer-planet moonlets could become lost over the next decade unless someone steps up to the eyepiece, so to speak, to pin down their orbits better.

    The count of AWOL objects would be higher had the team not undertaken an intensive effort to recover faint outer-planet satellites during 10 observing runs in 2009-11 with the CFHT in Hawaii and Hale 5-m reflector on Palomar Mountain, California. Observers attempted to track down 40 objects with dicey orbits and found 38 of them. It helped that, in September 2010, Jupiter had one its closest-ever oppositions and Uranus just over 1° from it in the sky.

    New-found satellites are given temporary identification codes: for example, two new ones swept up by graduate student Michael Alexandersen (University of British Columbia) and others during the recovery work in 2010 are designated S/2010 J 1 and J 2. Eventually, once their orbits are known precisely, they’ll get mythologically relevant names.

    A “satellite-recovery team” used the Hale 5-m telescope to snare the 22.8-magnitude moonlet Arche in October 2011. The glare from nearby Jupiter is evident at upper left.

    B. Gladman & others
    So how do “found” satellites become “unfound”? The simple answer is that their orbits were too poorly determined initially for long-term certainty. “You can’t just find them,” Gladman grumbles. “You have to find them and track them.” Many of these irregular moonlets are very faint (23 to 25 in magnitude) and lie far from their planets, he explains, positional follow-ups require the use of big telescopes for long observing runs. Unfortunately, it’s not the kind of scientific yield that’s going to gain much support when observatory time gets allocated.

    All of the lost and near-lost moons were discovered by Sheppard, Jewitt, and their collaborators, the most recent ones in 2007. “We wanted to complete all the known satellites with modern CCD technology,” Sheppard told me, keying on dynamically related irregular clusters. “We definitely wanted to recover everything,” he explained, but observations to follow up the Jovian finds were largely ruined by a run of bad weather. His team got more time last year on one of the twin Magellan telescopes, and although those observations haven’t yet been analyzed, Sheppard expects that some of the missing moons will turn up.

    And if not? It wouldn’t be the first time an outer-planet satellite became lost. In 1975, Charles Kowal and Elizabeth Roemer announced their discovery of a tiny moonlet in an inclined, elliptical orbit averaging 4.6 million miles (7.4 million km) from Jupiter. Then its whereabouts became unknown — until 2000, when Sheppard, Jewitt, and two others spotted it in their survey. Named Themisto in 2002, this one is not in any danger of being lost again.

    Posted by Kelly Beatty, April 4, 2012