≡ Menu

Large Exomoons Shown to Be Detectable

The search for sub-planetary scale features in other solar systems continues, with encouraging news from the Hunt for Exomoons with Kepler project. A moon around a distant exoplanet is a prize catch, but as we’ve also seen recently, scientists are weighing the possibilities in detecting exoplanetary ring systems (see Searching for Exoplanet Rings). Confirming either would be a major observational step, but exomoons carry the cachet of astrobiology. After all, a large moon around a gas giant in the habitable zone might well be a living world.

David Kipping (Harvard University) and colleagues at HEK have released a new study that tackles the question of how detectable exomoons really are. Published online today by the Astrophysical Journal, the paper examines 41 Kepler Objects of Interest, bringing the total number of KOIs surveyed by HEK thus far up to 57. The paper demonstrates that the process is beginning to move out of the realm of computer simulations and assumption-laden theory to actual data from Kepler. The paper’s goal is to determine how small a moon could be detected in each case given the kind of signatures that flag an exomoon’s presence.


Image: After surveying nearly 60 exoplanets for moons, the HEK team have derived empirical limits for each world, demonstrating an ability to detect even the smallest moons capable of sustaining an Earth-like atmosphere (“Mini-Earths”) for 1 in 4 cases studied. Whilst a confirmed discovery remains elusive, the painstaking survey of 60 planets spanning several years reveals what is possible with current technology. Credit: The Hunt for Exomoons with Kepler (HEK) Project.

An examination of an exoplanet that does not turn up an exomoon thus leads to a statement of how massive a moon has been excluded by the current data, which means the team is learning much about the sensitivity of its methods. From the paper:

… based on empirical sensitivity limits, we show for the first time that the HEK project is sensitive to even the smallest moons capable of being Earthlike for 1 in 4 cases (after accounting for false-positives). In terms of planet-mass ratios, we find even that the Earth-Moon mass-ratio is detectable for 1 in 8 of cases, posing a challenge but not an insurmountable barrier. Mass ratios of ∼ 10−4, such as that of the Galilean satellites, have never been achieved. However, if Galilean-like satellites reside around lower-mass planets than Jupiter, of order ∼ 20 M, then we do find sensitivity, as demonstrated by the limit of 1.7 Ganymede masses achieved for Kepler-10c.

This is encouraging news, for the team can now make statements about the actual mass of a detectable exomoon. In 1 of 3 planets surveyed, an exomoon with Earth’s mass is detectable. Kipping believes that we can move down to the smallest moon thought capable of supporting an Earth-like atmosphere and still detect it in 1 of 4 of the cases studied. No exomoons have yet been detected but we are learning just what our capabilities are. Says Kipping:

“Here we report on our null results and the first estimate of empirical sensitivities. Ultimately, we would like to actually discover a clear signal and are pursuing some interesting candidates we hope will pan out. Either way though, I like to recall what the Nobel Prize winning American physicist Richard Feynman said about science: ‘Nature is there and she’s going to come out the way she is, and therefore when we go to investigate it we shouldn’t pre-decide what it is we’re trying to do except to find out more about it’.”


Image: The Moon has about 1% the mass of the Earth posing a challenge for the HEK team, since such configurations are detectable for 1 in 8 planets surveyed. The much larger Pluto-Charon mass-ratio of 11.6% is much more detectable. Credit: Hunt for Exomoons with Kepler Project.

No exomoons turn up in the 41 KOIs surveyed in the study, with four, KOI-0092.01, KOI-0458.01, KOI-0722.01 and KOI-1808.01, showing up as false positives for an exomoon. Stellar activity is a likely cause, as the paper comments:

When dealing with a handful of transits, quasi-periodic distortions to the transit profile, such as those due to spots… can be well fitted by the flexible exomoon model. However, since an exomoon is not the underlying cause, this model lacks any predictive power and thus should fail F2a [a follow-up test described in the paper]. We therefore suggest that stellar activity is likely responsible for these four instances.

KOI-1808.01, in fact, passes the basic criteria for an exomoon detection, but the paper explains that the observed transit signal is distorted by the effects of star spots. Transit timing variations observed at KOI-0072.01 (Kepler-10c) seem to point to an additional planet in the system rather than an exomoon.

Thirteen of the KOIs produce some kind of spurious detection, assigned by the paper to effects like perturbations from unseen bodies, stellar activity or instrumental artifacts. Through the range of KOIs the project has studied thus far (57), 46 null detections are found from which upper limits on an exomoon’s mass can be derived. The paper reminds us that “…exomoons live in the regime where correlated noise is present and one must employ methods to guard against it when seeking such signals.”

The declared purpose of the Hunt for Exomoons with Kepler project is to ‘determine the occurrence rate of large moons around viable planet hosts,’ a task with implications for the abundance of life in the universe, for if habitable moons are common, there could be more of them than habitable planets, and conceivably more than one orbiting a single planet. An additional benefit of studying exomoons is that they can teach us about solar systems formation by showing us planet/moon systems in a variety of configurations.

The paper is Kipping et al., “The Hunt for Exomoons with Kepler (HEK): V. A Survey of 41 Planetary Candidates for Exomoons,” submitted to the Astrophysical Journal (preprint).


Comments on this entry are closed.

  • Andrew LePage March 20, 2015, 10:43

    I have been a big proponent of habitable moons for decades (see https://www.centauri-dreams.org/?p=31557) and look forward to the first “exomoon” being found, habitable or otherwise. But in addition to the KOIs mentioned in this paper which proved to be “false positives”, there was also the case of potential moon orbiting Kepler 90g which turned out to be a subtle instrument artifact.


    Just goes to show how difficult this task of finding exomoons is going to be.

  • Ross Turner March 20, 2015, 13:52

    I’m interested in this as well, and it will be exciting when the first detection is made.

    There could be a lot of issues regarding habitability of such moons, but first things first. Hopefully they will find one (some) soon.

    There is also apparently a size limit ratio of the moon to the gas giant. I can’t remember the article or the link, but I recall the result was something like a maximum of 1 mars mass to 1 Jupiter mass. Would like to find that link again. I’ll post it if I can find it. If true, this mass ratio will effect which gas giants moons can be detected around, given the current detection limits.

  • Daniel Högberg March 20, 2015, 13:53

    I find the tought of habitable exomoons really exciting. It all stems from my young years when I read 2001 by Clarke and later the movie Avatar of course. Im sorry to say that I failed catastrophically in school so that I can never furfill any dreams of working in the fields of astronomy or physics. But I can plow through all the sci-fi books that has ever been written and I can see all the sci-fi movies that has ever been filmed, and I can dream..

  • Ross Turner March 20, 2015, 14:00

    Here is one article from 2012, mentioning exomoon mass requirements for surface habitability, no mention of the moon to planet size ratio, but there is a mention of the gas giant possibly capturing a larger object as it migrates in towards its star, which eliminates the size ratio issue, if it exists:


  • galacsi March 20, 2015, 16:30

    There are a lot of super-sized Jupiters among the new found exoplanets and there should be a lot of big exomoons around them. but if “Mass ratios of ∼ 10−4, such as that of the Galilean satellites, have never been achieved. ” , we will have to wait for some new progresses before finding them. I am quite impatient !

  • Andrew Palfreyman March 20, 2015, 18:24

    The idea of habitable exomoons makes the somewhat disappointing abundance of super-earths more palatable.

  • Andrew LePage March 20, 2015, 21:16

    A more detailed review of the mass and other requirements for habitable exomoons can be found in the on-line version of my December 1998 article in Sky & Telescope:


    A more technical discussion of the issues can be found in a fully referenced piece I wrote in 1997 for SETIQuest Magazine:


    While these two pieces are about 17 years old, they still accurately reflect the issues involved with habitable exomoons.

  • Adam March 20, 2015, 21:40

    Assuming a 2:1 ratio of radii and a 1.5 Earth radius primary, then the usual mass relations for Earth composition planets produce a 2 gee surface gravity on the Primary and 0.68 g on the habitable moon. That’s a similar contrast to the planets in Larry Niven’s Known Space yarns – Jinx, a super-Silicate planet (1.68 g), and We Made It (0.6 g). Whether humans would adapt to 2 g is unknown.

  • Ashley Baldwin March 21, 2015, 6:55

    Complex things habitable moons. Rene Heller in Toronto has published extensively on this subject based on countless sophisticated simulations. There are two types of moon, ‘regular’ and ‘irregular’. Regular moons form from the same aggregation of material as their parent planet . Heller’s simulations have found that the largest such a moon could be is about 0.2Earth mass or the same mass as Mars. The lower mass limit to hold onto an Earth like Atmosphere lies somewhere between 0.2 and 0.3 Earth masses so touch and go. Not just for the gravity necessary but also for the radioactive isotopes and internal heat of formation to retain a par liquid iron core that can create a long term protective magnetic Dynamo. This of course is complicated by moons being locked into synchronous rotation with their planet which could potential stop them rotating quick enough, but studies have been able to create decent magnetic dynamos of different types. To be close enough to the parent planet to be protected by its magnetic field leads to excessive tidal heating ( think Io) which in turn leads to he dreaded runaway greenhouse effect.( think Venus!)
    Another problem Heller found is that the biggest moons ( think Ganymede) form round the biggest planets, with Mars size moons forming around monster gas giants with ten times Jupiter’s mass. Such moons would be made from a lot of ice and as with their gas giant planet will form by definition beyond the system’s ice line. To be in the habitable zone the planet would have to migrate inwards, something we know is possible , but once in the warm habitable zone it’s ice moon would melt and likely end up as a ocean world, probably without tectonics to help create a secondary atmosphere and maintain a carbon cycle.
    The other type of moon is an ‘irregular moon’. This is a free body that is captured by a migrating planet ( think Triton) possible a large planet migrating inwards. The smaller the planet the less the closing velocities and the greater the chance of capture rather than destruction or even ejection from the system. So a Neptune seized planet would do nicely. Irregular moons are not size limited and could be Earth mass or bigger and also be made of rock and iron if formed inside the ice line. Good news as long as it doesn’t too close to the patent planet to get tidally heated and rotates quicker enough to create a strong enough magnetic field to avoid having its atmosphere ripped away by the stellar wind which will be much stronger early in a system’s life. Far enough out and it will almost act like an independent planet.
    The only final warning is that we all remember Shoemaker-Levy slamming into
    Jupiter . Large planets have big gravity which attracts more impactors . Not good for a potentially inhabited moon if it gets in the way. Once again, as with Jupiter, this effect might be mitigated by the presence of a large gas giant further out to “draw fire” but there is no doubt an Earth sized exomoon is more likely to be subject to bombardment than a free Earth. Not insurmountable though.
    All based on simulation of course and as has happened in the past reality can throw up surprises. Let’s find those moons !
    Heller has published extensively on this and his work is very readable and available on arxiv.

  • Erik Landahl March 21, 2015, 13:58

    Hunt for Exomoons with Kepler a.k.a. HEK . . .
    Astrophysicists, astroengineers, astrogeologists, et al. continue to strain their brains to find words to describe their project that will result in a catchy acronym. They will soon have to hire public image firms to create catchy acronyms for their latest satellites and probes. :)

  • Adam March 21, 2015, 21:50

    Let’s get one thing cleared up re planets, magnetic fields and stellar winds. Atmospheres aren’t stripped away at a meaningful rate in the present-day solar-wind environment on Venus, Earth or Mars – it takes tens of aeons to remove a fraction of an atmosphere, at present. Magnetic protection is only meaningful when the star is in its early magnetically active phase, losing rotational energy to an enhanced stellar wind. This phase lasts 0.1 – 1.0 Gyr, or so, depending on the star. Atmospheric erosion, by the stellar wind, is an early process in *any* habitable planet’s history. If a planet’s magnetic field lasts long enough to get through that phase, then it can die off and that’d be of no real consequence to atmospheric evolution.

    I get a bit tired of hearing dumb-ass arguments over the need to give Mars a magnetic field to terraform it, so its atmosphere doesn’t go bye-bye from the Solar Wind. Ain’t going to happen – in this system or any other that’s 4-5 aeons old. Thus I felt the need to make this point.

  • John March 23, 2015, 6:23

    A large moon of a gas giant might be a super Europa, but it might equally be a super Io. We could make guesses based on distance from the primary, but is there any hope of extracting a surface spectrum from the Moons of a super gas giant, and distinguishing different sprectra from each other?

  • spaceman March 23, 2015, 19:16

    Would the HEK project be able to detect a Jupiter analogue with moons similar in size to our Jupiter’s moons inwards of 0.5 A.U.?

  • Eniac March 23, 2015, 21:07

    In our solar system alone we have at least three examples of irregular moons: Luna, Triton, and Charon. Stands to reason there’d be a few in every system. Also, the “fact” that regular moons around gas giants cannot be bigger than the ones around Jupiter and Saturn is a bit shaky if based on those two examples.

    Theoretical models are notorious for following experimental evidence, so we may well find some large regular exo-moons despite the current theory, and new theoretical models will quickly be created to explain them.

    As Adam says, the importance of magnetic fields has been greatly exaggerated, but gravity is an obvious must-have to retain an atmosphere. Size and insolation are overwhelmingly what matters for habitability.

  • Lionel March 25, 2015, 5:24

    Agree with the comments by Eniac above

    I have not yet seen Titan raised here – why is this satellite so different to the rest in our solar system and how does it fit into these models which we are giving so much power to? Why did it retain an atmosphere 4.5 times more dense than Earth’s while other moons have no atmospheres to speak of? If our solar system didn’t contain Titan then wouldn’t we have just made models that rule out satellites with such thick atmospheres?

  • Holger March 30, 2015, 5:36

    “In our solar system alone we have at least three examples of irregular moons: Luna, Triton, and Charon. ”

    The Moon is not irregular by Ashley Baldwin’s definition, but a third type – formed from the impact, not capture, of a planet-sized body, and much smaller than the impacting body.
    I can’t imagine Charon to be irregular – the chance of tiny Pluto capturing it in the vastness of the Kuiper belt must be negligibly small, and the existence of Pluto’s smaller moons in regular orbits strongly implies a common accretion history of the system.

  • Lionel Ward March 30, 2015, 14:02

    @Holger very true, and this third type – impact induced moons – may be more popular than captured moons.

    Here’s a relevant paper that just came up on google, a key thing their model implies is that there could be an upper limit to the size of impacting photo-planets that can form exomoons because bigger planets will have higher vapour fractions after impact, and that’s apparently bad for moon formation:


    ” Wada et al. (2006) suggest that a vapor-rich disk is dynamically unstable and that it may not be suitable for moon formation. If this is the case, the mass and composition of a planet may affect the satellite formation process. Here, we show results from giant impact simulations of planets with various masses and compositions. We use the model suggested by Nakajima & Stevenson (2014) to estimate the vapor mass fractions of the disks. We find that the more massive and the more ice-rich the planet is, the higher the vapor mass fraction of the disk becomes. This indicates there is an upper limit of the planetary mass to form an impact-induced moon and the limit depends on the planetary composition. This upper limit is a few Earth masses for a rocky planet, and about an Earth mass for an icy planet. “

  • Lionel Ward March 30, 2015, 14:03

    Sorry the hyperlink hasn’t come out right above. If interested please google:

    Constraints on Exomoon Formation – Nakajima, Miki; Genda, Hidenori; Asphaug, Erik; Ida, Shigeru

  • Eniac March 30, 2015, 23:23

    @Holger: I suppose I used “irregular” wrong. I meant any moon not subject to the 0.1% mass ratio “rule”. It is not much of a rule if there are so many violations just in our own system….

  • Ashley Baldwin April 5, 2015, 5:44

    Magnetic fields undoubtedly become less important with time though even this varies from star to star. The sun is unusually “quiet” in terms of excessive chromospheric activity but even many similar stars have been shown to exhibit high XUV outbursts long term. Smaller stars and especially M dwarfs can remain active for extended periods. Proxima Centauri, which is older than the sun, continues to flare. Whilst late stellar activity and wind in particular is significantly less as stars “spin down” ,no longer posing the risk of stripping atmospheres , magnetic fields still play and important long term role in helping protect planetary surfaces from prolonged exposure to dangerous XUV and also even more dangerous Cosmic rays originating from outside the solar system . This is a big if not insurmountable obstacle to extended manned exploratory periods on Mars ,which of course has a negligible magnetic field . Although not on Mars ( further increasing radiation exposure ) longer term a thick atmospheric blanket will help protect too , if it’s survived the early years of planetary life or been produced subsequently via tectonics. Given the imminent arrival of TESS and a detailed look at M dwarf planets we need to remember that they have the added problem of an extended pre main sequence life during which the proto M dwarf produces high levels of XUV. The habitable evaporated cores work looks at one way of bypassing this by planetary formation further out, avoiding excessive early erosion by “stellar aggression” via the dear old “inverse square law” before migrating into the the HBZ when things have calmed down. Ironic that the protective magnetic Dynamo produced by a rotating planet is required to protect against radiation that has been captured ,stored and released by a similarly created stellar magnetic field .

  • Ashley Baldwin April 5, 2015, 5:52

    Irregular and regular moons aren’t my definition by the way ! Far to clever. Rene Heller coined the term , but youre right , there is a “third way”( sounds like a political term ) . I will ask Prof Heller this question. Collisions are obviously common in the early solar system and in general “irregular” moons are produced by low velocity collisions between smaller planets given the lower gravity and closing speeds. A Neptune sized planet is ,ore likely to have a substantial irregular moon than say Jupiter , as evidenced by Triton. Some of the smaller Jupiter or Saturn moons particularly ( nearer the Kuiper belt) may well be the shattered left overs of such collisions . The issue though is being big enough to hold on to a substantial atmosphere should your parent gas/ice giant subsequently migrate to the HBZ and these smaller post collision moons are unlikely to do this as the minimum mass Heller’s team has calculated to do this is between 0.2-0.3 Mearth. All simulation though so the sooner we find some exomoons the better.

  • Ashley Baldwin April 5, 2015, 6:08

    In terms of terraforming Mars , Paul and I discussed atmospheric thickening as the best way of terraforming Mars. As Adam correctly points out , atmospheric loss , especially the thin atmospheres of habitable terrestrial planets would only be lost very slowly , predominantly by the “Jeans” escape mechanism , on planets around mature stars with low activity rates. Although Mars would struggle longer term to hold on to an atmosphere it would do so for tens of millions of years or more -so no risk to any civilisation existing on a far shorter timescale . Bombarding the Martian poles at low velocity with either high volatile content asteroids or comets would melt and release their CO2 and water which supplemented with the impactors’ own supply would thicken the atmosphere . An issue would be to avoid releasing sun blocking dust ( hence the requirement for low velocity impact) . The melting energy would be kinetic-thermal only with no nasty nuclear explosions and radiation involved. No breathable atmosphere created initially , but would elevate surface temperature and pressure , hopefully allowing liquid water and also protect from radiation without the need of a magnetic field . A lot more pleasant environment in which to live. Sadly although feasible in the not too distant future ,we both felt that such a technique could be exploited as a weapon should the pin point accuracy required be turned to sinister purposes.