≡ Menu

Pondering Recent Exomoon News

‘Exomoons’ — moons around planets around other stars — are another of those new frontiers of modern astronomy. It’s astonishing to reflect that 51 Pegasi b, the first exoplanet orbiting a main-sequence star, was discovered as recently as 1995, a time when we could only suspect that planets might be common and were only then working out the best ways to find them. Now we have thousands of planet candidates, the search is on for true Earth analogues, and the idea that we might make as fine-grained a discovery as an exomoon is an exhilarating prospect.

So is the recent paper from David Bennett (University of Notre Dame) and colleagues the breakthrough we’ve been waiting for? The answer is no because we have no way of knowing whether this suggestive find is a true moon around a planet or perhaps two larger objects in gravitational synch at a much further distance. But either case is intriguing. Here are the possibilities for the 2011 event called MOA-2011-BLG-262, detected by gravitational microlensing:

  • A gas giant of three to four times Jupiter’s mass at a distance of 1800 light years, the moon being, at about half the mass of Earth, absolutely gigantic in comparison to moons in our Solar System and orbiting its host world at a distance of 0.13 AU, quite a large separation. This is what Bennett and team call the ‘fast model,’ ; or
  • A much more distant brown dwarf orbited by a Neptune-mass planet. This is the ‘slow model’ (see the paper for details on the use of these terms).

Now brown dwarfs with planets are interesting enough, so I would find either of these scenarios fascinating. But what makes the exomoon picture still more riveting is that these objects are evidently out there on their own, with no associated star. If we really are looking at an exomoon and a planet, then the question of how they came to be where they are emerges. New Scientist quotes exomoon hunter David Kipping (Harvard University) as saying “It almost begs the question as to whether we can really call these objects ‘moons’ or whether some other name is more apt.” What that other name might be is left to the imagination of the reader.

So unusual would a free-floating planet with a moon of its own be that the paper on this work leans preferentially to the brown dwarf solution. From the preprint:

…an apparently free-floating planet with a half Earth-mass moon would be a new class of system that was not previously known to exist. Such a new discovery would require strong evidence, so our favored model for this event is that it is a low-mass star or brown dwarf orbited by a planet of about Neptune’s mass.

Or as the abstract puts it: “The data are well fit by this exomoon model, but an alternate star+planet model fits the data almost as well.”

If that seems to take the sizzle out of this story, please reconsider. This interesting duo was identified by gravitational microlensing, in which the passage of an object in front of a more distant star bends the light from that star in ways that can be quantified. In the case of MOA-2011-BLG-262, the light of the distant star was magnified about seventy times beyond the norm. About an hour later the second, smaller increase in brightness occurred, suggesting the passage of two objects in front of the star. Because microlensing events don’t repeat, we may never be able to untangle the true story of MOA-2011-BLG-262. But the work of Bennett and colleagues points to future microlensing breakthroughs:

…it should be possible to definitely distinguish similar models for future events if they are observed with high cadence from multiple sites. Very high cadence observations on 1-2m class telescopes are able to measure the light curves precisely enough to distinguish similar models (Gould et al. 2006), such as the fast and slow models for MOA-2011-BLG-262. If these high cadence observations are taken from observatories separated by thousands of kilometers, then the terrestrial parallax effect can be used to measure the lens masses (Gould & Yee 2013).

Screenshot from 2013-12-27 09:30:38

Image: The left panel shows a K-band (infrared) image from the VISTA 4m telescope from the VVV (Vista Variables in the Via Lactea) survey. The field observed by the Keck-2 telescope in K and a zoom of this field are shown to the right. The arrow indicates the microlensing source star, separated by 0.51 arcsec from its nearest neighbor. Credit: David Bennett.

‘Cadence’ as used in the passage above refers to how long it takes to re-observe the same target. High cadence can be combined with other techniques, and the paper notes that observatories separated by thousands of kilometers can take advantage of terrestrial parallax to increase and refine the microlensing detection rate. Ultimately, we’ll need new instruments:

…the development of large robotic networks of 1m class telescopes, such as the Las Cumbres Robotic Telescope Network (Brown et al. 2013), will substantially improve the rate of terrestrial microlensing parallax mass measurements. Thus, if systems resembling the planetary-mass host models for MOA-2011-BLG-262 are common, the combination of high cadence microlensing surveys, rapid realtime event detection by these surveys, and high cadence follow-up observations should enable the definitive discovery of rogue exoplanets with moons of nearly an Earth mass within a few years.

Recall the ground-breaking work of David Kipping and colleagues on using transit timing (TTV) and transit duration variations (TDV) to spot an exomoon signal in transit data, and the ongoing Hunt for Exomoons with Kepler project that Kipping runs. Add in this recent microlensing story and the possibility of future microlensing discoveries of the kind described above. It’s clear that we have exmoon detections in our future, and who knows what we may learn about the characteristics of rogue objects that, if MOA-2011-BLG-262 is a true indication, may not be traveling alone.

The paper is Bennett et al., “A Sub-Earth-Mass Moon Orbiting a Gas Giant Primary or a High Velocity Planetary System in the Galactic Bulge” (preprint).


Comments on this entry are closed.

  • andy December 27, 2013, 13:27

    The fast model still leaves the question of whether the primary is best regarded as a gas giant planet or a (sub-)brown dwarf. Is such a system likely to have survived ejection from a planetary system – I would guess that planet scattering would not be good for the satellite system.

    On the other hand, it might be thought of as a planet around a very low mass brown dwarf (i.e. stellar-like formation mechanism). Does anyone know of any studies that tried to model the evolution of the discs that have been observed around low-mass brown dwarfs (e.g. OTS 44, Cha 110913-773444) — is a half-Earth mass planet at 0.13 AU a likely outcome?

  • jeff December 27, 2013, 16:39

    at such a large AU separation is it at all plausible that they both were at one point two distinct planets of a shared stellar system, were both kicked out by some gravitational disruption (by which I just mean another star or cluster of stars drifting too close), and ended up a-drift on trajectories similar enough that the one ended up in orbit around the other?

    Obviously not a likely scenario, but still curious if that could be a possible outcome of such a near collision between stars.

  • Dr Ashley Baldwin December 28, 2013, 8:48

    Even assuming this half Earth mass body is a moon and that it is orbiting a big gas giant rather than a brown dwarf , a semi major axis of 0.13 AU gives an orbital diameter ( if not too eccentric ) of about of nearly 24 million miles. If that sort of variation was applied to an orbiting planet it would give it one hell of an eccentric orbit. Even if the gas giant is in the middle of the HBZ for its star , the moon’s orbit is going to be pushing the limit of that zone and possibly going outside ( or inside ) it for atleast part of its year. The variation in the moon’s distance from its parent star is going to be enormous . I recognise this may be an extreme example, and that what is most exciting is an Earth dimension exo moon , but it illustrates an important point. To be in their planetary HBZ, exo moons are going to have to be a significant distance away from that planet to avoid radiation belts, gravitational tides, eclipses etc . This in turn will impact on their distance to the parent star (as they orbit the planet) , that distance increasing as a fraction of the orbital distance . The smaller the star and related HBZ, the greater that fraction. This is a factor that will impact on the overall stellar flux the moon receives ( partly mitigated by short resultant eclipse periods ). The bigger the moon’s planetary orbit the greater this variation. In the case of a Red Dwarf star for instance , with an average HBZ semi major axis of .25 AU even for an M O, such variation would be critical .

  • Adam December 29, 2013, 7:50

    To be stable a Moon would need to be within the Hill Sphere of its Primary. The Hill Sphere radius is computed as the mass-ratio of the Moon/Primary to the 2/5 power times their mutual orbit’s semi-major axis around the star. This assumes low orbital eccentricity around the star. For example, our Moon orbits Earth at a distance of 384,400 km, which is ~1/3 the Hill Sphere radius. Jupiter, which masses ~1/1048 the Sun and orbits at 5.2 AU has a Hill Sphere out to 0.322 AU, or ~48 million km. It’s outermost moons orbit at about 3/5 the Hill Sphere distance. The Hill Sphere scales with the orbital radius, by definition, so placing a Gas Giant at 1 AU around a solar mass star means a smaller Hill Sphere and thus smaller stable orbits. If the stellar system in question was a low mass Red dwarf, then the half-mass “moon” could really be a Trojan planet, captured in the L4 or L5 position of the Gas Giant.

  • Dr Ashley Baldwin December 29, 2013, 19:59

    Thanks Adam. Very thought provoking. I’m not an astronomer or physicist, but a keen follower of exoplanet science and this is new to me. The concept you describe is clearly extremely relevant to the whole science of large earth sized bodies interactions within the extended Hill Spheres of large gas giant sized planets. Especially those within the HBZ of their respective stars . The more so the smaller the star with correspondingly smaller HBZ . It also draws attention to two totally new but important concepts to all but experts , namely Trojan planets and what exactly constitutes a moon. Are our current RV and TTV methods able to identify such bodies , of such potential for life and maybe the only way in those systems where the HBZ lies within the star’s synchronous orbit zone? Is a Trojan planet tidally locked and if so, to what? Its parent planet or its parent star? Sounds like a potential custody battle . How stable can the orbits of such bodies be too?

  • Michael December 30, 2013, 4:19

    Dr Ashley Baldwin said

    ‘Is a Trojan planet tidally locked and if so, to what? Its parent planet or its parent star? Sounds like a potential custody battle . How stable can the orbits of such bodies be too?’

    I believe the moon will be tidally locked to both because the Lagrange points are where the forces are in balance and so both the parent planet and Star will hang in the same positions of the sky respectively. Masses in the L4 and L5 points are quite stable for 100’s of millions if not billions of years, however they can easily be disrupted by nearby masses. If this moon is in one of these Lagrange points it could explain why it is so massive as material will want to collect there.

    @Adam said

    ‘If the stellar system in question was a low mass Red dwarf, then the half-mass “moon” could really be a Trojan planet, captured in the L4 or L5 position of the Gas Giant.’

    Nice theory by the way and if so we should be able to work out the orbital dynamics of the system better as there are tight constraints on the locations of the Lagrange points.

  • Adam December 30, 2013, 5:59

    Hi Ashley
    There have been several papers on Trojan planets – well worth looking for them on the arXiv or NASA’s ADS.
    Stability is definitely an issue for crowded systems, but fortunately the L4/L5 points are attractors and planets caught there require a lot of jostling to leave. It’s not inconceivable for a Trojan to be semi-captured in a loose orbit around the primary planet, but mostly they will trail or lead the planet in its orbit.

  • Harry R Ray December 30, 2013, 10:52

    THE MYSTERY DEEPENS! Just today, a NEW paper on this subject was added to the Extrasolar Planets Encyclopedia, titled:”New Method to Measure Proper Motions of Microlensed Sources:Application to Candidate Free-Floating-Planet Event MOA-2011-BLG-262″ The “NEW” measurement CLEARLY (as opposed to ONLY a HIGH PROBABILITY in the Bennet paper) a high velocity lens as opposed to a high velocity source! What is most intreaguing is in the FINAL SENTENCE of the preprint “…These values are then important imput into a baysian analysis of the event”. Correct me if I am wrong,but, in the original paper,Bennet et al based their PREFERENCE of thye Brown Dwarf-Planet model OVER the Free Floating Planet-Exomoon model on a Baysian analysis that didNOT FAVOR the exomoon interpretation! Could the NEW DATA in the Skowran et al paper make a REVISED Baysian analysis MORE favorable to the exomoon interpretation? Someone help me out,here,please!

  • Harry R Ray December 30, 2013, 11:27

    ADDENDUM TO THE ABOVE COMMENT; THERE APPEARS TO BE another EXOMOON ISSUE SURFACING, although DEFINITELY NOT an exomoon DETECTION! In the paper by Daniel Tamayeo:” Consequences of an Eccentric Orbit for Fomalhaut b”, he FAVORS a scenario where Fomalhaut b is a SUPER EARTH surrounded by a dust cloud maintained by several “IRREGULAR SATILITES”! Does “IRREGULAR imply stable ECCENTRIC 0orbits or just UNSTABLE orbits! If it is the former, exomoons may be the BIGGEST news story of 2013!

  • Dr Ashley Baldwin December 30, 2013, 17:25

    Thanks Adam. I’ve found the articles and intend to plough through them. I often read arXiv even as a non astronomer .
    In relation to tidal locking , would it be possible for the body at the L4 or L5 Lagrange to not be tidally locked to either star or planet Michael ? Just a thought. Interesting concept for side stepping the synchronous orbits impediment to HBZs of low mass stars if it was, especially if it is a collecting point for orbital detritus.

  • andy December 30, 2013, 18:59

    Harry R Ray: what’s with the ALL CAPS all over the place? It certainly doesn’t help when reading your posts.

    If you read the Skowron et al. paper, you would have come across this part:

    The stellar lens with the Jupiter-mass planetary companion located in the Galactic bulge is, a priori, much more likely explanation for the MOA-2011-BLG-262 microlensing event, than the close-by Jupiter-mass planet with the Earth-mass moon in the Galactic disk. By reducing the likelihood of the lens being in the bulge, our measurement brings those two explanations on par.

    So it is still far from conclusive, even though the likelihood of it being an exomoon is increased.

    As regards the term “irregular satellites”, it is the same term as used to describe the features of satellite systems of the solar system gas giants, which are divided into inner regular satellites on orbits roughly coplanar with the planetary equator, and outer irregular satellites which are often on eccentric and inclined (sometimes retrograde) orbits. E.g. Callisto is an example of a regular satellite of Jupiter, while Himalia is an irregular one. You may want to try a Google search for the term to find out more information.

  • Eniac December 30, 2013, 20:06

    I am confused. If this were a true planet and its moon, there must be a star there, too. If it is a brown dwarf with a planet, then there would be no third object. As I understand microlensing, there would be no way to miss the main star, planets are normally just tiny minor perturbations overlaying the main lensing signal. So, are the alternate models just a free floating planet vs red dwarf and planet, or is a full star-planet-moon system also on the table?

    If it is just free-floating planet with moon vs. brown dwarf with planet, one is just a smaller version of the other, and my excitement would be much contained.

  • Paul Gilster December 30, 2013, 20:12

    Eniac writes:

    If it is just free-floating planet with moon vs. brown dwarf with planet, one is just a smaller version of the other, and my excitement would be much contained.

    I think these are exactly the alternatives, a free-floating planet and moon vs. brown dwarf with planet. The paper leans toward the latter explanation.

  • c howard December 31, 2013, 6:23


    To be stable a Moon would need to be within the Hill Sphere of its Primary.

    why is the Hill Sphere radius computed as the mass-ratio of the Moon/Primary to the 2/5 power times their mutual orbit’s semi-major axis ?
    what is the reasoning behind 2/5 power ?

  • Michael December 31, 2013, 15:15

    @Dr Ashley Baldwin

    ‘In relation to tidal locking , would it be possible for the body at the L4 or L5 Lagrange to not be tidally locked to either star or planet Michael ?’

    I am of the opinion that any material forming in the L4 and L5 points would be under the influence of both the Planet and Star and most likely be damped of rotation early on in the formation process. If it formed first and then was captured it could still have rotation depending on whether it is close to the Star to have been tidally locked in the first place.
    Off topic I wonder if the interaction between Earth and Venus caused the hypothetical planet Theia to become unstable at one of the L points to eventually collide with the young Earth, after all it appears that Venus is locked to our Planet.

  • Dr Ashley Baldwin December 31, 2013, 21:51

    Thanks Michael.
    Ive seen several articles suggesting the increasingly suspicious Earth/Venus connection. Funny how it should be reinforced by exo moons. Those L4/5 Lagrange points are beginning to look very important in young star systems ,alone and on top of orbital resonance. I have to agree using “moon” for objects originating there is a misnomer. Any one up for some IAU lobbying? As the distances between orbiting bodies shrinks with decreasing star-body distance ,(such as the HBZs around small stars ) this is going to complicate habitable matters yet further , never mind XUV ,CME, flares and the like. No wonder they are finding more “orphan” planets. It must be like a terrestrial planet bowling alley ( or marbles) in close in M class systems.

  • Michael January 1, 2014, 16:23

    I see that it gets crowded around smaller stars so there is likely to be some pinball going on. I also see your point about whether it would be classed as ‘moon’ if it is at L4 or L5, maybe we could call it a ‘Lamoon’ in honour of Lagrange. Adam any ideas for naming them as you came up with the idea maybe ‘Adamoons’?

    As for the Planetary systems we have normal orbits such as most of our solar system, resonance orbits like Jupiter’s moons, elliptical orbit such as with Neptune and Pluto and possibly now Lagrange point ‘moons’. Interesting stuff and interesting times ahead.

  • Eniac January 1, 2014, 17:59

    Thanks, Paul, for the clarification. It seems, then, that the discussion started by Adam around Hill sphere and Lagrange points, while interesting, would not really apply to either of these models, seeing that they are both two-body systems.

  • Sean Sakamoto January 2, 2014, 10:37

    I had the pleasure of hearing Ray Pierrehumbert speak at the Hayden Planetarium about exoplanets and he brought up an interesting fact about exo-moons that I had never considered.

    Just like exoplanets need to orbit in goldilocks zone around a star to have liquid water and temperatures that are conducive to supporting life, exo-moons also have a goldilocks zone of orbit from their main planet. The need to be far enough away so that the gravity of the planet they orbit doesn’t cause constant volcanic activity which would also make life impossible. So an exomoon must orbit an exoplanet that’s in the goldilocks zone from its star, and then the moon must be in a goldilocks orbit to keep it geologically stable enough.

    I’m sure a lot of your readers already know this, but I wanted to mention it for anyone like me who found it interesting.

    Happy New Year, everyone!

  • andy January 2, 2014, 14:08

    Regarding objects in the Trojan points, for the case where the orbits are circular and the object is located exactly at the Trojan point, being locked to the star gives the same rotation as being locked to the planet, since the objects are fixed in an equilateral triangle. Things are presumably more interesting for objects on tadpole or horseshoe orbits around the Trojan point(s). Not sure whether the rotations of any of the Trojan satellites in the Saturnian system have been determined?

    I would not consider an object in a Trojan point, or any of the other 1:1 resonant configurations to be a moon – the dynamical situation is very different. I guess the term “Trojan planet” would do in analogy to the Trojan asteroids, and this has already been used in the scientific literature. “Quasi-satellite” has already been used for objects in another class of 1:1 resonance.

  • ljk April 10, 2014, 13:28