I am just reading the review article Advances in exoplanet science from Kepler (arxiv preprint: http://arxiv.org/abs/1409.1595), and I found a remarkable paragraph (last paragraph in section "Properties of planetary systems", page 341):

Just as important as the discoveries made by Kepler are its non-discoveries. So far, Kepler has found no co-orbital planets, which share the same average semi-major axis — like the Trojan asteroids found accompanying Jupiter and the Saturnian satellites Janus and Epimetheus. It has also found neither exomoons nor 'binary' planets orbiting one another.

There is no further explanation or interpretation on that.

What does the authors want to say with this remark? Do they infere that due to the non-discovery, the probability of such objects is very low?

I'm interested especially why the authors think this is important? What does it say?

Edit (October 2018):

I would like to update this question, to mention that that now - four years later - an exo-moon has very likely been discovered by Alex Teachey and David Kipping, using Kepler data (with subsequent observations with Hubble).

  • 3
    $\begingroup$ The "follow up question" probably should be its own question. $\endgroup$
    – Kyle Kanos
    Commented Sep 21, 2014 at 20:30
  • 2
    $\begingroup$ Kepler is a primarily a transit-based instrument. That means if finds exo-planets because they pass (repeatedly) between their star and the telescope. So ask yourself ... how would co-orbiting objects or binary planets or large moons show up? This is the kind of question that you should be able to puzzle out for yourself and it will make you a better scientist. $\endgroup$ Commented Sep 21, 2014 at 20:58
  • $\begingroup$ @dmckee I thought about it and have some picture in my mind. However, my main question is not HOW it does, but why the authors of the review (which are Kepler scientists) say it's important that they didnt detect any of those. $\endgroup$ Commented Sep 21, 2014 at 21:04
  • $\begingroup$ @KyleKanos True, thanks. Removed the question, and added the link to the wikipedia article (which pretty much explains it). $\endgroup$ Commented Sep 21, 2014 at 21:11
  • $\begingroup$ It tells you something about how planets form. No binary planets means they probably don't form by fragmentation of larger objects, but more likely the coalescence of smaller things. Of course the significance of the non-discovery depends on the sensitivity of the experiment. Exomoons are hard to find, unless they are a significant fraction of the (giant) planet size. $\endgroup$
    – ProfRob
    Commented Sep 22, 2014 at 11:31

1 Answer 1


The Hunt for Exomoons with Kepler Project has so far failed to find any exomoons. This is a negative finding (so far). Negative findings are always a bit trickier to explain than are positive findings. This negative finding might mean something very significant, or it might have very little significance:

  • Maybe exoplanets are much less likely to have exomoons that the abundance of moons in our solar system would suggest; or

  • Maybe the close-in exoplanets that Kepler was predisposed to find are much less likely to have exomoons; or

  • Maybe the group hasn't studied enough exoplanets and so far have just been a bit unlucky. From reading through their papers, their approach is exceeding CPU intensive, taking decades of CPU time per planet investigated; or

  • Maybe their technique isn't as good at detecting exomoons as they think it is; or

  • Maybe it's something else that explains the negative results.

As for why Lissauer et al. put that little teaser of a paragraph in their Nature article, that's a good question. It's a bit of a parenthetical remark (i.e., non-essential). If it's not essential, what's that remark doing there? The article in question was about not just what has already been found in the Kepler dataset but also about what is still waiting to be found. The uses of transit time and transit duration variations by the Hunt for Exomoons with Kepler Project could add significant value to the Kepler dataset. As for Lissauer et al. didn't say anything more, it's a bit premature right now to make anything of the so-far negative results from the project. There are too many maybes right now.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.