Exomoons? Kepler‘s On The Hunt


Recently, I posted an article on the feasibility of detecting moons around extrasolar planets. It was determined that exceptionally large moons (roughly Earth mass moons or more), may well be detectable with current technology. Taking up that challenge, a team of astronomers led by David Kipping from the Harvard-Smithsonian Center for Astrophysics has announced they will search publicly available Kepler data to determine if the planet-finding mission may have detected such objects.

The team has titled the project “The Hunt of Exomoons with Kepler” or HEK for short. This project searches for moons through two main methods: the transits such moons may cause and the subtle tugs they may have on previously detected planets.

Of course, the possibility of finding such a large moon requires that one be present in the first place. Within our own solar system, there are no examples of moons of the necessary size for detection with present equipment. The only objects we could detect of that size exist independently as planets. But should such objects exist as moons?

Astronomers best simulations of how solar systems form and develop don’t rule it out. Earth sized objects may migrate within forming solar systems only to be captured by a gas giant. If that happens, some of the new “moons” would not survive; their orbits would be unstable, crashing them into the planet or would be ejected again after a short time. But estimates suggest that around 50% of captured moons would survive, and their orbits circularized due to tidal forces. Thus, the potential for such large moons does exist.

The transit method is the most direct for detecting the exomoons. Just as Kepler detects planets passing in front of the disc of the parent star, causing a temporary drop in brightness, so too could it spot a transit of a sufficiently large moon.

The trickier method is finding the more subtle effect of the moon tugging the planet, changing when the transit begins and ends. This method is often known as Timing Transit Variation (TTV) and has also been used to infer the presence of other planets in the system creating similar tugs. Additionally, the same tugs exerted while the planet is crossing the disk of the star will change the duration of the transit. This effect is known as Timing Duration Variations (TDV). The combination of these two variations has the potential to give a great deal of information about potential moons including the moon’s mass, the distance from the planet, and potentially the direction the moon orbits.

Currently, the team is working on coming up with a list of planet systems that Kepler has discovered that they wish to search first. Their criteria are that the systems have sufficient data taken, that it be of high quality, and that the planets be sufficiently large to capture such large moons.

As the team notes

As the HEK project progresses, we hope to answer the question as to whether large moons, possibly even Earth-like habitable moons, are common in the Galaxy or not. Enabled by the equisite photometry of Kepler, exomoons may soon move from theoretical musings to objects of empirical investigation.

12 Replies to “Exomoons? Kepler‘s On The Hunt”

  1. I can imagine a planet being noticeably tugged by a moon if the moon’s orbital period and the duration of the transit were of the same magnitude… but how likely is that? Venus takes 7 hours to transit the Sun – most of the gas giants we’re spotting have much faster orbits and so (I assume) transit in minutes – by which time I expect a moon wouldn’t have moved enough to tug its parent planet in a different direction.

    If such an effect really can be teased out of the Kepler data, then I’ve clearly underestimated our capabilities!

    1. Don’t they do both? The TTV looks during the integrated orbit, the TDV during the transit.

    2. There are two different ways the presence of a moon can be detected here that i can think of:

      1. The entry and exit timings are displaced from transit to transit caused by the dislocation of the planet as the moon are orbiting. The values are increased by the distance between the two, and the larger the mass of the moon the more signal. Large separation means a slow orbit.

      2. The difference between the entry and exit timings, the duration of the transit. This requires, as you say, relatively quick moon-orbit in order to show any signal. Fast orbit means small separation.

      I suspect 1. is significantly easier (yet still difficult) to detect than 2.

  2. That we see captured moons locally is encouraging though. Say, Triton.

    And they didn’t always had a nice planetary disk to migrate and capture comfortably against, it seems they had to rely on being originally binaries in most cases. (Again, Triton, see the link.)

    Or it simply happens once for a blue moon.

    1. a case could be made that all moons are captured.
      large moons could have moons of their own.
      how about primordial ‘wandering moons’ that have resided in multiple star systems?

      1. That all moons are “captured” could possibly be true now when the brown dwarf type of giant planet coalescence from the disk seems moot. (The alternative core collapse scenario, where a core slowly aggregates gas until it can rapidly grow by draining it from the disk, is supported by most exoplanet observations.)

        As Earth-Moon, Mars-Deimos-Phobos and Pluto-Charon demonstrates, that capture can go through impacts and subsequent capture and coalescence of ejected mass.

        Wandering moons would have the larger problem of capture happening without an auxiliary break like those that disk effects sets up in the early protoplanetary system.

  3. Just made me think of Pluto here, orbiting a point outside its own body. So clearly, if the moon is on the opposite side as when we first saw a transit, the transit will be later (or earlier) and if the moon is orbiting in the same direction as the planet, duration of transit might be longer as well.

  4. There was a related article here at UT on detecting rings around extrasolar planets. I think the detection involves the light signature. When a planet crosses another star there is a characteristic optical signature it presents. The transit is not just a step function that turns the luminosity down a bit and then back up. There are complicated boundary spikes and so forth. This has a bit to do with the Kirchoff rule for EM radiation passing through a hole or around “mask.” It is a bit complicated so I just mention it, where for those interested you might look in Jackson’s “Classical Electrodynamics” to get the full scoop. Now if the planet is dragging an entourage of moons there will be slight variations on the optical timing signature for the transit. I think those are what the method might possibly pick up.


  5. So our present equipment couldn’t detect an exomoon the size of Ganymede or Titan? These worlds could have just as much potential for hosting life as their parent planets.

    1. Yes, they have habitability potential, which is one reason why it is exciting there is against expectations likely observability of some sort.

      If large moons can be seen, there may be potential for extrapolating distributions (i.e. how many of some mass range) to smaller moons, however bad such a guesstimate will be.

Comments are closed.