What if all the Kepler exoplanet candidates orbited one star?

By Phil Plait | August 14, 2012 7:00 am

This is pretty cool: astronomer Alex Parker took all the planet candidates found by the Kepler telescope – nearly 2300 planets in all – and made an animation showing what they would look like if they all orbited one star.

Dr. Parker had to do some scaling to make this work. For example, the actual size of each planet is known relative to its parent star, which he then scaled to fit the star shown in the animation. He scaled the distance from the star in a similar way. He describes it all on the page for the video.

I have to admit, it’s hard to know if there’s anything scientific we can learn from this. It’s fun to play with data, and it does often happen that by doing so you can see hidden relationships, things that aren’t obvious when displayed in normal ways (I’ve had that happen to me as well just playing with data – and Parker is very good indeed at playing with data; see Related Posts below for more cool stuff he’s done). That may very well turn out to be true in this case, or it may simply turn out to be an interesting demonstration. But as he points out, since this is animation was done to scale using all the Kepler planet candidates, one thing you see immediately is that there is always at least one transit going on! In other words, looking at all the Kepler host stars, no matter when you look, there are probably a dozen transits or more occurring at that moment.

That to me is actually shocking. I mean, it makes sense, and given some thought I would’ve realized it on my own. However, the context I always put this in is that just a few years ago we didn’t know of any transiting planets, and in fact for a while after the first were found a lot of astronomers scoffed at the idea. Now, though, the evidence is so overwhelming there is no doubt these transiting exoplanets exist.

And yet in all that time we didn’t know these planets existed, in all that time astronomers were looking for them and didn’t see them, in all that time some were found and other astronomers scoffed, there was never a time when there wasn’t a planet transiting a star.

And that’s just in the tiny patch of sky Kepler looks at; the entire sky is over 300 times bigger. So if there are a dozen or so transits going on in just the Kepler field, as Parker states, that means there are thousands of them going on in the sky even as you read this. Every day, all day, for millions and billions of years.

All those planets, perhaps millions of them, hidden in plain sight. All we needed to do was actually look for them.

So the usefulness of Parker’s animation is clear to me; its impact on me was profound. It reminded me once again that science evolves, and that my own biases – all our biases – must evolve with it. Otherwise, who knows what we’ll miss?

Tip o’ the dew shield to the Scientific American blog.

Related Posts:

Piano sonata in the key of Kepler-11 (music by Alex Parker based on planet orbits)
Music of the spheres (more music by Alex Parker, this time based on supernovae)
New study: 1/3 of Sun-like stars might have terrestrial planets in their habitable zones
Motherlode of potential planets found: more than 1200 alien worlds!


Comments (36)

  1. Anyone know if Trent Reznor is an astronomy buff? The FIRST time I heard “02 Ghosts I” was during that Cassini montage.

  2. steve

    Fascinating video, and completely hypnotizing. I could see the large, orange, fast moving one closest to the star even after it zoomed all the way out! What a monster that planet must be!

    Are we able to see planets that don’t transit yet? Like the part of the video showing the system from top-down: if they are not on the same plane of observation can we still detect the star’s wobble? There must be so many more….

  3. bryan elliott

    That’s awesome. Now do one that has gravity; I wanna see carnage.

  4. Doug Little

    Are the circles 1AU apart?

  5. Chris

    @3 bryan
    I was thinking the same thing. That would be computationally costly, even on fast computers. 2300 bodies orbiting in close quarters, all perturbing each other. Would be fun to see though.

  6. Georg

    All we needed to do was actually look for them.

    Since when is such a measurement of tiny (some % at best)
    variation of stars brightness possible on such timescale?

  7. dessy

    Gravity sim would be awesome but we still know very little about the orbits of these planets. This sim has assumed a circular orbit (because it is the best information we have at the moment) but it would be fair to assume that at least some of these bodies would have highly elliptical orbits. Nevertheless – its a wonderful sim. Highly hypnotic!

    @ Doug Little. The circles represent the orbits of Mercury, Venus and Earth (source: from the description on the Vimeo page.)

  8. Keith Hearn

    #4 Doug Little: The circles are the orbital distances of Mercury, Venus, and Earth.

  9. Adam

    Truly amazing to think that mere years ago we had no idea there were planets around other stars. And now we have thousands of them. Our once small neighborhood has gotten a lot bigger and more complex feeling. It is things like this that keep me humble.

  10. The field as shown looks a bit crowded and makes me wonder whether such an arrangement could possibly stabilize like that, rather than hurling planets out (and occasionally into the star).
    Under “Related posts” I note astronomy-based music; I make some of that myself.

  11. Tris

    Novice space guy question here …

    What determines the speed at which a planet travels while orbiting? It makes sense to me that the planets closer to the star move faster, but farther out, there are some planets moving much faster than others at roughly the same distance from the star. Why?

    This is fascinating stuff.

  12. Jimmy

    Steve, unfortunately some of the planets in that video have been shown to be false positives, including that one. From the video’s description:

    “A fraction of these candidates will likely be ruled out as false positives as time goes on, while the remainder stand to be confirmed as real planets by follow-up analysis. For example, the large orange object in a very close-in orbit was shown to be a background eclipsing binary blend by arxiv.org/abs/1207.2481”

    The article on arxiv suggests that the false-positive rate for certain orbits is a lot higher than previously suggested.

  13. Jimbo

    What if all the Kepler exoplanet candidates orbited one star? Well they’d all influence each other gravitationally and the result would be nothing like the animation!

    Yes there’s always one and today it’s me.

  14. I think asking for the gravity sim is the funniest thing I’ve read today. I would pay real coin to see that.

    But on a more serious (and possibly heretical) note, I am wondering if what we are seeing here, with so many planets being found that orbit so close to their stars, many of these planets quite large in relation to their parent stars, is the result of a commonly repeating error of some sort. To explain, I would expect to see some sort of distribution in these planetary systems that would have planets of varying sizes both far and near from the stars. Seeing something unexpected like multiple-Jupiter-mass planets so close to a star might be interesting in some of these, but in so many, and in such a preponderant number? Could it be that what we are seeing is some sort of observational error that leads to that conclusion so frequently?

    I mean, these stars are quite far away, and the amount of dimming is so slight, that a measure of error would be expected. Could the preponderance of results that show planets orbiting so close to their parent stars be a result of fuzziness that results from observational error? I would not be at all surprised to find over time, as the techniques get more refined, and our tools get better and better that in at least some of these cases, a non-trivial “some,” that our knowledge of these planets’ mass, size and orbital period alters significantly.

    Is anyone else thinking this, or am I alone and wondering this out of ignorance?

  15. HvP

    HartinHajovsky, you are partly right.

    The fact that we more commonly see large planets orbiting very close to their stars is the result of an observational bias, but it’s not necessarily an error. It just so happens that those are the planets that give the strongest signal because they have the largest effect, and so those are the ones that are the easiest to find. With that in mind it’s easy to understand why they might be overrepresented.

    Imagine standing 100 yards away from a random assortment of various sized balls and you have to identify how many of each you can count. It’s likely that you are going to notice the basketballs and soccer balls first even though there may be ten times as many golf balls and tennis balls out there – and you might never see all the marbles.

    As our instruments have become more precise we have started seeing more and more smaller planets orbiting further out. We are even starting to see planets within the near Earth-sized range orbiting at comfortable distances from their stars. They just happen to be harder to detect because they take longer to compete their orbits and have less effect on their stars.

    Keep in mind that the further away a planet orbits the longer it takes for it to complete one orbit. A close in planet with a two month orbit can be confirmed in one-half-year (three full orbits) but a planet that takes two of our years to complete an orbit would take six years to be confirmed.

  16. blueskitten

    #12: In my understanding, the larger mass planets that are closest in are the easiest to spot, which is most likely why we see a disproportionate number of them among the transiting planets identified to date. They have smaller orbits so their transits are more frequent, and they also block more light so the dimming is more pronounced. So as Kepler watches for longer periods, we can find those more distant and smaller planets with longer orbits.

  17. JohnT

    I think we would see the big crash if you had many in orbit

  18. soul_biscuit

    And weren’t a lot of these trans-Jupiter planets orbiting close to their stars discovered before Kepler, by the wobble method? I would think that a near-orbiting gas giant would cause a bigger wobble than a far-orbiting one, leading to another source of observation bias.

  19. Cool to look at and it does convey the variety of planets that have been found pretty well. Adding realistic gravity similar would be fun, briefly.

  20. Valdis Kletnieks

    Am I the only one who’s not at all surprised that out of a collection of 2300 bodies that were detected by their transits, there’s always a transit going on? Kind of like going through a large city, finding 2300 policemen based on their uniforms – and then being surprised that everybody in the sample is wearing a police uniform. 😉

  21. Ken Howard

    I’m slightly confused about your wordage, Phil. I understand that this video uses data from Kepler concerning all of the planets that it has found so far. And it truly is amazing! But what I do not understand is your statements about transits (when the planet moves in front of it’s host star).

    Why would exoplanets NOT transit their star? If it revolves around the star, of course it transits in front of it, right? It might not be a transit to US, because we might not be viewing from the correct plane/POV, but it happens…

    Or are you referring that, after combining all of the data, the fact that some of these planets would transit each other is amazing? In that case, isn’t the “amazement” sort of “wrong”, because the only reason for the transits in the video is because someone took all the data and put it in there…but that’s not “real”, because all of those planets aren’t really orbiting the same star.

    Or am I just completely missing the point? It’s been a very long day!

  22. HvP

    Ken Howard,

    What he’s saying is this shows us that if you imagine all of these planets once again separated out into their native star systems scattered across the sky – at any one moment at least one of them will be undergoing a transit event at any time you happen to be observing.

    Transit events are fairly fleeting if you think about it. It takes maybe a couple of hours for a planet to transit the star (from our perspective) and then several months or years to complete the rest of its orbit. So imagine laying out under a field of stars knowing that all around you there is the brief intermittent winking of transiting planets flitting across those myriad of stars. Like the reverse of lightening bugs on a summer evening, but in slow motion – brief intermittent dimmings randomly scattered all over the sky.

  23. Torbjörn Larsson, OM

    Now if they could do that with the Earth Similar planets (ESI >= 0.7) of The Habitable Exoplanets Catalog!

    in all that time astronomers were looking for them and didn’t see them, in all that time some were found and other astronomers scoffed, there was never a time when there wasn’t a planet transiting a star.

    That is what some people in the biz calls a deep-of-field observation, I would assume.

    My own take was that we haven’t really got started on the equivalent Mercury orbital distance yet. Hopefully the next data release will remedy that!

    @ 2:

    Yes, wobble methods works at least as well for ordinary surveys what I know, check the exoplanetary databases for method used with their statistical tools.

    But for a stationary field-of-view survey at a medium distance (so many stars surveyed) such as Kepler, transits is much more economical.

    @ 14:

    I would not be at all surprised to find over time, as the techniques get more refined, and our tools get better and better that in at least some of these cases, a non-trivial “some,” that our knowledge of these planets’ mass, size and orbital period alters significantly.

    As already noted, there is an observation bias here from many sources.

    Indeed, the early observations of many giants and suspicions that you needed rather large stars and high star metallicities (i.e. elements other than hydrogen and helium) have been modified.

    Now they claim, unless I am mistaken, more planets of smaller sizes (Kepler, Corot), more smaller planets the further out you go (Kepler, IIRC), proportionally more planets around M stars, and as many terrestrial planets regardless of star metallicity within wide limits. A veritable planet bonanza!

  24. James Evans

    @#3, bryan elliot:

    That’s awesome. Now do one that has gravity; I wanna see carnage.

    Park the Death Star in orbit around that swarm, and let’s have a turkey shoot.

    Though I suspect, with so many targets, overheating will become an issue, and the commander will require the men to double their efforts before the station is once again fully-operational for the Emperor’s troubling, unexpected visit.

  25. David Gormley

    Shouldn’t some of these planets be moving in the opposite direction?

  26. Andrew W

    Anyone know when we can expect the next batch of confirmed planets from the Kepler team?

  27. Jess Tauber

    I’m surprised none of you suggested that by combining the data from all the exoplanets we could create a kind of computer program that would send us a message from the original galactic planetary system about it’s properties (like that STNG episode with the DNA from the various humanoid races that so disgusted the Romulans that they were related to Klingons…)

  28. Satan Claws

    @Tris (#11):
    The star’s mass and the distance to the star. Look up: Kepler’s third law.

  29. Satan Claws

    Anyway, what surprised me the most was how many planets were discovered (relatively) close to their host star. Granted, confirming the presence of planetary companions relatively far from the star is harder, but this animation puts it in quite a different perspective!

  30. Messier Tidy Upper

    Superluminous animation & write up. Cheers! 😀

    @25. David Gormley asked : “Shouldn’t some of these planets be moving in the opposite direction?”

    Pretty certain I recall reading of several backwards (retrograde) orbiting exoplanets being discovered but I’m not sure whether or not it was the Kepler space telescope that found them. So, quite likely, yeah. Will have to check.

  31. Ken Howard


    Thank you! That makes a load of sense now! I appreciate it 😉

  32. Messier Tidy Upper

    On doing that checking for tonight :

    Looks like it was WASP (Wide Angle Search for Planets – La Silla observatory Chile, ESO) that found retrograde exoplanets :


    But then again :


    HAT -P-7b may be retrograde and is also known as Kepler-2b … (Or not 2b! A.k.a. the Hamlet Planet?)

    .. although that may just have been a confirmation test rather than a discovery.

    OTOH, this via Universe today :


    Implies that Kepler itself *has* found retrograde Hot Joves but does NOT explictly state so.

    Which means .. ? [Shrug.] Still no conclusive answer here from me I’m afraid. :-(

  33. Andreas H

    I think the most interesting thing we can learn from this video is how heavily biased the results from Kepler are. Don’t get me wrong, they are great results but the method used by Kepler heaviliy favors big planets in close orbits.

    Of course we can’t just extrapolate Keplers findings and assume similar density and amount of planets for wider orbits, but it is a pretty conservative assumption to at least expect a fair amount of additional planets in wider orbits.

    If anything, Kepler has shown that Planets are a common thing in this universe, more common than stars! While the last bit is not yet solidified 100% it’s by no means a stretch to say these days.

    We are all very understandably excited about Curiosity, but thinking about the fact that there are billions of planets waiting for exploration is an almost overwhelming feeling.

    At the same time I wonder, what if FTL technology is indeed impossible? What if in thousands of years we manage to perfectly cartograph our surrounding universe but will forever be captives of our own star system, unable to bridge the distances of space. What if the Fermi Paradox is no paradox at all, but just the confirmation of the physical universe as we understand it now? What if we will never be able to explore space?

  34. beer case

    @Andreas H:

    No FTL does not mean we are trapped in our own solarsystem. It just mean it’s gonna take a bit more time to get out there.

    That the fermi paradox seems seems real, could simply be a matter of not listening to the right frequencies:


  35. Andreas H

    Of course it doesn’t mean we are technically restricted.

    The real problem will not be propulsion, but communication. Look at the developement of various civilizations in our own history, they were not limited by how far they could travel but by how far they could still hold a steady stream of communication.

    Even if we can expand our own lifespawn to virtual immortality and conquer various technologies that would allow us to live forever on space ships, to actually explore space and to colonize it we need the ability to communicate efficiently enough to hold a “normal” attention-span. Ideally this means real time communications, we might work with a delay of a couple days at most, but anything longer will make it impossible to stay connected…

  36. Julian

    Given the observational bias, (ie, it’s much easier to detect large planets in close orbit than it is to detect small planets in large orbits), what’s even more humbling is that this graphical representation is just the tip of the iceberg within the systems that have been currently surveyed.

    It should be pretty straight-forward for astronomers to use existing data to factor for observation bias and model how many planets exist in larger orbits with their mass distribution. Probably already done 100 times over. Not my field so I don’t know…

    Anyway, great stuff! Science + data visualization FTW.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!


See More

Collapse bottom bar