≡ Menu

Enormous Ring System Hints of Exomoons

Might there be gas giant planets somewhere with moons as large as the Earth, or at least Mars? Projects like the Hunt for Exomoons with Kepler (HEK) are on the prowl for exomoons, and the possibility of large moons leads to astrobiological speculation when a gas giant is in its star’s habitable zone. Interestingly, we may be looking at evidence of an extremely young — and very large — moon in formation around a planet that circles the young star J1407.

That would be intriguing in itself, but what researchers at Leiden Observatory (The Netherlands) and the University of Rochester have found is an enormous ring structure that eclipses the young star in an epic way. The diameter of the ring system, based on the lightcurve the astronomers are getting, is nearly 120 million kilometers, which makes it more than two hundred times larger than the rings of Saturn. This is a ring system that contains about an Earth’s mass of dust particles, with a marked gap that signals the possibility of the large moon.


Image: Artist’s conception of the extrasolar ring system circling the young giant planet or brown dwarf J1407b. The rings are shown eclipsing the young sun-like star J1407, as they would have appeared in early 2007. Credit: Ron Miller.

The ring system itself was discovered in 2012 by Eric Mamajek (University of Rochester) and team, with Leiden’s Matthew Kenworthy and Mamajek now refining the observations and working out the details. What emerges is a ring system with over thirty separate rings. And you need to see the lightcurve, which is available below. Kenworthy’s enthusiasm about the find is evident:

“The details that we see in the light curve are incredible. The eclipse lasted for several weeks, but you see rapid changes on time scales of tens of minutes as a result of fine structures in the rings. The star is much too far away to observe the rings directly, but we could make a detailed model based on the rapid brightness variations in the star light passing through the ring system. If we could replace Saturn’s rings with the rings around J1407b, they would be easily visible at night and be many times larger than the full moon.”

Exoring model for J1407b from Matthew Kenworthy on Vimeo.

I love the many worlds presented to us in science fiction, but I’m hard pressed to come up with a depiction of anything quite like this. Says Mamajek:

“If you were to grind up the four large Galilean moons of Jupiter into dust and ice and spread out the material over their orbits in a ring around Jupiter, the ring would be so opaque to light that a distant observer that saw the ring pass in front of the sun would see a very deep, multi-day eclipse. In the case of J1407, we see the rings blocking as much as 95 percent of the light of this young Sun-like star for days, so there is a lot of material there that could then form satellites.”

The figure below, from the paper, gives a static view of the same data:


Image: From the paper. The caption reads: “Model ring fit to J1407 data. The image of the ring system around J1407b is shown as a series of nested red rings. The intensity of the colour corresponds to the transmission of the ring. The green line shows the path and diameter of the star J1407 behind the ring system. The grey rings denote where no photometric data constrain the model fit. The lower graph shows the model transmitted intensity I(t) as a function of HJD. The red points are the binned measured flux from J1407 normalised to unity outside the eclipse. Error bars in the photometry are shown as vertical red bars.” Credit: Matthew Kenworthy/Eric Mamajek.

As to J1407b, the planet these rings surround, the astronomers estimate that it has an orbital period of about a decade, with a mass most likely in the range of between ten and forty Jupiter masses. The gap in the ring structure points to a satellite in formation that has an orbital period of approximately two years around the gas giant. It becomes clear that if we can find more instances of early disks, we can begin to study comparative satellite formation around exoplanets. From the paper:

J1407 is currently being monitored both photometrically and spectroscopically for the start of the next transit. A second transit will enable a wide range of exo-ring science to be carried out, from transmission spectroscopy of the material, through to Doppler tomography that can resolve ring structure and stellar spot structure significantly smaller than that of the diameter of the star. The orbital period of J1407b is on the order of a decade or possibly longer. Searches for other occultation events are now being carried out (Quillen et al. 2014) and searches through archival photographic plates (e.g. DASCH; Grindlay et al. 2012), may well yield several more transiting ring system candidates.

The paper also points out possible ring structures around Fomalhaut b (anomalous bright flux in optical images) and Beta Pictoris b (anomalous photometry), though neither of these has been confirmed. The scientists involved are encouraging amateur astronomers to help monitor J1407 as the attempt to constrain the mass and period of the ringed planet J1407b continues. Observations can be reported to the American Association of Variable Star Observers (AAVSO).

The paper is “Modeling giant extrasolar ring systems in eclipse and the case of J1407b: sculpting by exomoons?” accepted for publication by the Astrophysical Journal (preprint).


Comments on this entry are closed.

  • Andrew LePage January 27, 2015, 10:54

    In an article I wrote 16 years ago for Sky & Telescope on habitable moons, I stated that it would probably be decades before we find the first exomoons (habitable or otherwise). While I usually avoid making such predictions in print, so far I have been proven correct. I certainly agree that these observations of J1407 are tantalizing, but the presence of an exomoon orbiting J1407b is just a scientific hypothesis to explain what was seen. Future observations will be required to verify that J1407b actually exists, that it possesses a ring system and that an exomoon is responsible for some of its structure. As for HEK, one of the better candidates for an exomoon detection, a subtle “blip” in the photometry associated with Kepler 90g, has been shown to be an instrument artifact.


    So the search for an exomoon continues. Hopefully the continued analysis of the Kepler data set will reveal some bona fide exomoon signatures in the near future.

  • Al Jackson January 27, 2015, 11:23

    The fraction of solar mass stars in binary systems is about 33%.
    I think the estimate now is that , depending on spectral type, most , if not all, single stars may have planets. (Remember most stars are less than a solar mass.)
    Considering that asteroid 2004 BL86 , which flew by the Earth-Moon system today has a moon, the existence of exo-moons , everywhere is not surprising.
    However this affinity of ‘pair up’ or ‘group up’ is not entirely explicitly clear. Stars , planets and asteroids just like to have company.

  • Alex Tolley January 27, 2015, 18:33

    That is a big ring system. Is it of recent origin, perhaps due to a major impact event? Coincidental? It would be good to confirm the planet it surrounds, otherwise it is possible this is a misinterpretation of the data.

  • Matthew Kenworthy January 28, 2015, 5:31

    Paul, many thanks for a detailed article on our paper!

    Andrew – exomoons are our best working hypothesis for explaining the detailed structure in the light curve. If you look at Saturn, nearly all the ring structure you see is either direct clearing by a moon, or indirectly from Lindblad resonances caused by moons. Of course, it could also be some kind of disk self-stirring or excitation of disk modes.

    We carried out a detailed search for J1407b itself, detailed in Kenworthy et al. 2015 in Monthly Notices:


    We did not detect J1407b, but we derived strong limits on its possible mass and orbital period. We are very confident that it exists, though, as we went through a relatively exhaustive list of alternative possibilities to explain the light curve in Mamajek et al. (2012) AJ:


    When the next eclipse occurs, 24 hour monitoring of J1407 will uniquely constrain the ring structure and then we can see if the structure is consistent with disk excitation modes or with resonant gaps and edges caused by exomoons. Either way around, it’s going to be great fun to find more of these systems in the next few years!

  • Andrew LePage January 28, 2015, 12:32

    @Matthew Kenworthy January 28, 2015 at 5:31

    Thanks for the links to the various papers! They are most appreciated. Naturally I agree that a ring system sculpted by an exomoon orbiting a giant planet or brown dwarf is certainly a plausible hypothesis to explain the observations. However, in my comment I am just pointing out that this hypothesis is based on a number of assumptions that do not rise to the point to make this a definitive detection of an exomoon (using the situation of Kepler 90g as an example where an exomoon hypothesis for Kepler’s observations unfortunately did not pan out recently). But like any good scientific hypothesis, this explanation does make a lot of predictions about what future observations should find in the years to come. I am certainly looking forward to see how this works out!

  • Wojciech J January 28, 2015, 13:51

    I wonder if colonization of such ring system is possible, especially if they are moonlets in it. The sight and diversity of possible cultures would be beyond spectacular. Or imagine indigenous life evolving on one of the moons(although I guess impact events could hinder this).
    On more scientific note, wasn’t there a suggestion that exo-moons could be detected around super-Jovians by analyzing radio wave emissions?

  • andy January 28, 2015, 16:16

    Discs in transit have been found for stellar binaries, most notably Epsilon Aurigae, which also exhibits some evidence for a complex disc structure around the secondary. So is this more an overgrown Saturn or a diminutive Eps Aur?

    As for Beta Pictoris, isn’t the transit window predicted to be sometime in the next couple of years? (Not sure what the latest orbit estimation is for this one…)

    On the subject of rings but on a rather smaller scale, there’s a recent suggestion that (2060) Chiron may support a ring system. This would make it the second centaur for which rings have been found, the first being (10199) Chariklo.

  • Matthew Kenworthy January 28, 2015, 16:47

    andy – yes, we were thinking of beta Pic too in this regard, its transit may well be soon. The structure in the light curve means that there’s no gas in the J1407b system to fluff up the disk.

    And yes, there are rings being found around rocks in the Solar System, I think it’s all very cool to see coming out.

    Wojciech – I was thinking of trying out to make a travel poster like the Kepler ones that cae out earlier in the month. If you were on an exomoon, it would be a truly spectacular sight. I’m not sure of the moon would have gravity enough for an atmosphere, but if it did and the moon orbited *slightly* out of the ring plane, you’d get trulay spectacular views. And you’d see gravity ripples along the trailing edge of the rings by the moon. Great stuff for the imagination!

  • FrankH January 28, 2015, 16:56

    With a mass between 8J and 40J, that “planet” pretty much brackets the masses of brown dwarfs. I don’t know if brown dwarf “moons” form the same way as moons of gas giants or if they form as regular planets.

    These rings remind me of Robin Canup’s work on the formation of moons around gas giants:


  • Harry R Ray January 29, 2015, 11:01

    I strongly disagree that the central body surrounded by the rings is a planet,REGARDLESS OF MASS! My reasoning is this:In our solar system, the degree of axial tilt is a function of mass for the four giant blanets, with Jupiter, the most massive, having by far the least tilt. This ring system could NOT have been seen in such great detail if it were edge on (or very nearly face on, like Uranus’s ring system), or somewhere in the middle, like Saturn’s. Therefore, the central object COULD NOT HAVE FORMRD LIKE JUPITER, via core accreation, with a slight asymitry of impact angles leaving a wery slight aial tilt, but; instead, like a star, which could have an equator far different in angle to the central star’s equator. To me, the only way this object could be a planet is if it formed via disk instability as opposed to cloud collapse, but even then, you would assume that conervation of angular momentum would align the two equators. But; since this issue had notbeen addressed in the literature (as far as I know of, correct me if I am wrong) a PURELY SCIENTIFIC PROOF of this logic is not currently available, and therefore, there MAY be exceptions, but I seriously doubt it.

  • Matthew Kenworthy January 29, 2015, 16:47

    Frank – The distinction between giant planet and low mass brown dwarf is quite fuzzy (somewhere around 10 to 20 Jupiter masses) , but the satellite formation should be very similar at all these masses.

    Harry – The most recent theories of core accretion versus gravitational instability don’t say anything about the equatorial alignment of the giant planets with respect to their ecliptic, so seeing this tilted ring system doesn’t inform on the formation mechanism.

    And as you point out with Uranus, equatorial tilts can be reset with planetary collisions, so the formation history can be erased that way too.

    As for the detail in the rings, that’s not due to the tilt – the level of detail is set by the size of the ring system compared to the size of the star.

    I hope that helps!

  • andy January 29, 2015, 19:10

    From the abstract of Canup & Ward (2006), they mention that the satellite loss process arises from gas-driven migration. So would the J1407b disc be more favourable for satellite survival due to its low gas content?

  • ljk March 12, 2015, 9:52


    UV Habitability of Possible Exomoons in Observed F-star Planetary Systems

    Satoko Sato, Manfred Cuntz

    (Submitted on 9 Mar 2015)

    In the present study we explore the astrobiological significance of F-type stars of spectral type between F5 V and F9.5 V, which possess Jupiter-type planets within or close to their climatological habitable zones. These planets, or at least a subset of them, may also possess rocky exomoons, which potentially offer habitable environments.

    Our work considers eight selected systems. The Jupiter-type planets in these systems are in notably different orbits with eccentricities ranging from 0.08 to 0.72. Particularly, we consider the stellar UV environments provided by the photospheric stellar radiation in regard to the circumstellar habitability of the system.

    According to previous studies, DNA is taken as a proxy for carbon-based macromolecules following the paradigm that extraterrestrial biology might be based on hydrocarbons. Thus, the DNA action spectrum is utilized to represent the impact of the stellar UV radiation. Atmospheric attenuation is taken into account based on parameterized attenuation functions. We found that the damage inflicted on DNA is notably different for the range of systems studied, and also varies according to the orbit of the Jupiter-type planet, especially in systems of high ellipticity.

    For some systems large values of damage are attained compared to an Earth-type planet at Earth-like positions in the solar system. A highly protective exomoon atmosphere would be required in most systems to foster habitable environments, notwithstanding extremophiles or systems based on nonstandard exobiology, which are beyond the scope of the present study.

    Comments: 30 pages, 7 figures, 3 tables; submitted to International Journal of Astrobiology

    Subjects: Solar and Stellar Astrophysics (astro-ph.SR); Earth and Planetary Astrophysics (astro-ph.EP)

    Cite as: arXiv:1503.02560 [astro-ph.SR]

    (or arXiv:1503.02560v1 [astro-ph.SR] for this version)

    Submission history

    From: Manfred Cuntz [view email]

    [v1] Mon, 9 Mar 2015 17:15:13 GMT (286kb)


  • ljk March 19, 2015, 9:02

    Race to find the first exomoon heats up

    22:00 17 March 2015

    by Jacob Aron

    Magazine issue 3013.

    THE galaxy could be stuffed with large moons that orbit alien worlds, according to an analysis of data from NASA’s Kepler space telescope. Such moons are thought to be the most likely places to find alien life, so groups are clamouring to find the first.

    Astronomers have raked in thousands of exoplanets, but their smaller moons have proved harder to come by. Kepler looks for exoplanets by watching how the light from a star dips as a planet passes in front, known as a transit. Moons should produce a smaller, secondary dip, but that dip’s timing varies because an orbiting moon can transit before, after or at the same time as its parent planet.

    A team led by David Kipping at the Harvard-Smithsonian Center for Astrophysics is searching for exomoons by modelling all the positions one could be in and looking for similar light signals in the Kepler data. But it’s such a computationally intensive strategy that they have had to crowdfund their own supercomputer and borrow NASA’s to crunch the numbers.

    Now Michael Hippke of the Institute for Data Analysis in Neukirchen-Vluyn, Germany, has road-tested a simpler method. While he didn’t find any clear exomoon signals, the results suggest an abundance of moons the size of Jupiter’s Ganymede, the largest in our solar system.

    Full article here:


  • ljk March 19, 2015, 9:04


    On the detection of Exomoons

    Michael Hippke

    (Submitted on 17 Feb 2015)

    Despite the discovery of thousands of exoplanets, no exomoons have been detected so far. We test a recently developed method for exomoon search, the “orbital sampling effect” (OSE), using the full exoplanet photometry from the Kepler Space Telescope.

    The OSE is applied to phase-folded transits, for which we present a framework to detect false positives, and discuss four candidates which pass several of our tests.

    Using numerical simulations, we inject exomoon signals into real Kepler data and retrieve them, showing that under favorable conditions, exomoons can be found with Kepler and the OSE method.

    In addition, we super-stack a large sample of Kepler planets to search for the average exomoon OSE and the accompanying increase in noise, the “scatter peak”. We find significant exomoon presence for planets with 35d<P<80d, with an average dip per planet of 6+-2ppm, about the radius of Ganymede.

    Comments: Draft. I invite the interested community to repeat the tests using your own workflow and detrending. Co-authors welcome. To be submitted

    Subjects: Earth and Planetary Astrophysics (astro-ph.EP)

    Cite as: arXiv:1502.05033 [astro-ph.EP]
    (or arXiv:1502.05033v1 [astro-ph.EP] for this version)

    Submission history

    From: Michael Hippke [view email]

    [v1] Tue, 17 Feb 2015 20:48:27 GMT (2394kb,D)