≡ Menu

Exoplanet Discoveries via PC

Because the financing for missions like Kepler is supported by tax dollars, it’s gratifying to see the public getting actively involved in working with actual data from the Kepler team. That’s what has been going on with the Planet Hunters site, where 40,000 users from a wide variety of countries and backgrounds have been analyzing what Kepler has found. Planet hunter Debra Fischer (Yale University), a key player in the launch of Planet Hunters, has this to say:

“It’s only right that this data has been pushed back into the public domain, not just as scientifically digested results but in a form where the public can actively participate in the hunt. The space program is a national treasure — a monument to America’s curiosity about the Universe. It is such an exciting time to be alive and to see these incredible discoveries being made.”

So far, so good on the citizen science front. Using publicly available Kepler data, Planet Hunters has found two new planets, both of them discarded initially by the Kepler team for a variety of technical reasons. Fischer believes the odds on the detections being actual planets are 95 percent or higher. The candidate planets have periods of 10 and 50 days, and radii from two and a half to eight times that of the Earth. One of them is conceivably a rocky world, though not in the habitable zone. Several dozen Planet Hunters users had spotted the planet candidates.

Image: One of the tutorial figures explaining how to use Planet Hunters at the site. The time it takes a planet to complete one orbit is called the orbital period. For transiting planets, this can be determined by counting the number of days from one transit to the next. Planets in longer period orbits will be more challenging to detect, both for humans and for computers because a transit will not appear in every 30-day set of light curve data. Large planets with short orbital periods are the easiest ones to detect. The most challenging detections will be small planets with long orbital periods. These will require patience and care, but are the real treasures in the Kepler data. Credit: Planet Hunters.

Following up the detection, astronomers used the Keck Observatory to study the host stars. A new study is to be published on the discoveries in the Monthly Notices of the Royal Astronomical Society, marking the first time the public has used NASA space mission data to find planets around other stars. So while the heavy lifting continues to be done by the Kepler team itself, public science has proven to be a helpful supplement, bringing more eyes to the data at hand. And, of course, the next round of Kepler data provides just as intriguing a hunting ground.

When it began, Planet Hunters was described as a bet on the ability of humans to beat computers, at least occasionally, because of the way people can use pattern recognition. The Kepler team uses computer algorithms fine-tuned to analyze light curve data because of the sheer number of stars the mission is working with. But while computers excel at finding what they are trained to find, the potential for surprise is always there as tens of thousands of users put pattern recognition to work to examine light curves, track down anomalies, and pay close attention to transit signals. For this kind of analysis, using powerful computers with a widely distributed human backup component is proving ideal for the task at hand.


Comments on this entry are closed.

  • Mike September 22, 2011, 10:09

    Great article about the Planet Hunters citizen science project! I’d like to point out why only 2 exoplanet candidates are listed in the paper submitted by Debra Fischer and others. They culled the best 2 candidates from a list of 10 observed only during Kepler’s first month of observation. Planet Hunters has found many more planet candidates over the 9 months they’ve been working. According to their website at planethunters.org they have detected 69 planet candidates so far. Of which I’m proud to have helped detect 5. These are exoplanet detections seperate from the Kepler science teams list of planet candidates, which stands at 1235, though that will likely be changed after Fridays news and public data release from the Kepler team.

  • stephen September 22, 2011, 10:47

    You had an article on 2010 TK7 ‘s horseshoe orbit. I wonder how many exoplanets might be in that kind of situation. They might be in binary star systems, too. Brown dwarfs and super-Earths might be in horseshoe orbits; will that make the planet finders’ jobs even more difficult?

  • Istvan September 22, 2011, 12:27

    It’s reading about initiatives like Planet Hunters that makes me wish I had more time on my hands. That would be really a fun thing to get involved with.

  • JohnQ September 22, 2011, 13:07

    Along these lines I wanted to share the following fascinating (to me anyway) story: http://www.komonews.com/news/local/130128788.html
    (there are probably other, better sources, but this is where I first heard of it.)
    Now, this has nothing to do with planet hunting as such, but it occurs to me that the use of game-like techniques, having proved so successful in the analysis of protein folding, might also have application for these human mind intensive planet searches.
    It’s an idea — no comment on the practicality, of course — and I stand proudly by it.

  • jkittle September 22, 2011, 16:15

    ALERT! spokesman from CERN announces that some Neutrinos are found to travel FASTER than Light.

    let me see:
    Possibly No Higgs particle
    Dark energy
    Dark matter
    particles traveling faster than light..
    maybe some new physics coming our way

    maybe the journey stars are closer than we had dared hope!

  • Tom Baty September 22, 2011, 16:49

    I know that this is off topic, but I am sure there would be some interest in this, I would post a link but I am having a PC problem. The people over at CERN did an experiment with Muson Nuetrinos and it appears that they found one that travels faster then light. They have NOT announced a discovery but are asking for confirmation from other sources. I found the idem on MSN news site. If can resolve my computer problem I will give you a link.

    Regards, Tom

  • Tom Baty September 22, 2011, 16:56

    PS……I just googled CERN and there are articles there from Washington Post and San Fran News. So it is out there, don’t get you hopes up could be a false reading….but!!!


  • Paul Gilster September 22, 2011, 19:37

    I think we shouldn’t jump to conclusions this early in the process — the likelihood of this being a systematic error in the measurement of the neutrinos seems most likely here. More as the story emerges, of course.

  • Rob Henry September 23, 2011, 0:13

    As a bewildered outsider, I would love to know why, despite repeated results suggesting a negative squared rest mass, virtually all physicists insist that the electron neutrino is not a tachyon. Martin Gardner explained pretty eloquently why it is not possible to have a charged tachyon, but I don’t know of any equivalent reason for uncharged particles.

    As for the statement from jkittle’s link that this breaks Einstein’s special relativity that is only true if other electron neutrinos had been shown to travel slower than light. The best I can make sense of that statement is, because it would be possible that information could be sent faster than light the idea of freewill might have to be abandoned – but why not just say that then.

    If anyone can help resolve my state of confusion I would be grateful.

  • henk September 23, 2011, 1:40


    those Neutrinos go 0.0025% faster than light. How do you come to the conclusion that a journey to the stars are closer. It is faster than light,But not thar much faster. Still it is very interesting.

  • Ronald September 23, 2011, 4:51

    ref. CERN and the FTL neutrino’s: the effect is only 0.0025%, not insignificant, but not exactly ‘the way to the stars’ either. Seems so much more likely that some much more mundane effect is playing a role here, such as maybe a more precise calibration of the real (maximum) speed of light.

  • Ronald September 23, 2011, 4:53

    Back to topic, a question, maybe to Mike: when participating in this Planet Hunter’s quest, is one assigned a star, or can one also choose a star oneself?
    (Of course I could go to the site and find out, but I am just being a bit busy and lazy here).

  • Paul Gilster September 23, 2011, 8:04

    I’ll be posting about the CERN results later today, so let’s move further discussion of those questions to that thread, as soon as the post appears. That way we can keep on topic here re Planet Hunters and exoplanetology.

  • Mike September 23, 2011, 11:29

    To Ronald, Planet Hunters will assign you a star. You will get a roughly 30 day length of stellar light curve that you will look at for any sign of a transit. You will mark any possible transit you see or indicate no transit. After your done with the first light curve you will get another roughly 30 day segment of another stars’ light curve. You don’t pick your star, there’s no point in choosing anyway. Keplers target stars are all anonymous dim ordinary main-sequence dwarfs. Or they should be after the Kepler science team culls the initial target list.
    But anyway the best thing is to avoid me confusing you with my comments is for you to follow the link Paul Gilster put in his article that leads directly to the planethunters.org website. They explain everything very well there and also provide a short tutorial about examining their data for exoplanets.
    One thing, if you’re running Internet Explorer as your browser you may have to use a differnet one to access Planet Hunters. I recomment Google Chrome.
    You can download it free from Google, it installs easily, is easy to use and it’s not resource intensive. You can still keep your Internet Explorer for other sites, Google Chrome won’t replace it, you can choose which browser to run. I hope this helps answer your questions.

  • Ronald September 23, 2011, 18:16

    @Mike: thanks!

  • coolstar September 24, 2011, 2:39

    While I congratulate the Planet Hunters on their finds, I’m somewhat pessimistic (and was from the start) that this is actually a good “crowd sourced” project.
    “good” here in the sense that it’s an efficient use of resources; it’s certainly good public relations and a great way to get people interested in science. First and probably most importantly, both these finds WERE flagged by the Kepler pipeline software and at very high significance (not surprising as the 50 day orbit stands out like a sore thumb and even the 10 day orbit transit is better than quite a few ground based transit observations. “Better” in the sense that it’s easy to see something interesting is going on. The precision is roughly 100x higher.). THEN the pipeline software dropped these transits from further analysis (why, doesn’t appear to be completely understood, at least not by the authors). This strongly suggests that a really good idea would be to put human eyeballs on EVERY light curve where this has happened (high s/n transit, later dropped by the software). The WASP team already does something very much like this in analyzing their ground based transit search data in that every software flagged transit is checked by eye.
    Secondly, the Planet Hunters false alarm rate is, not surprisingly, really high (60% of the 10 “best” candidates flagged). LOTS of time and effort (NOT completely wasted effort, as a lot was learned about the false positives) into resolving those false positives. It’s probably reasonable to assume that the Kepler team has already a) implemented something like what I’ve suggested and/or b) tweaked their pipeline software (almost certainly a continuing process). I suspect that these may be both the first and last planets found by the Planet Hunters. When this project was announced, I said publicly that I doubted any planets would be found that weren’t already flagged by the software, and that’s still true (what I didn’t predict was that the software would find high s/n transits, then later ignore them!). I think perhaps one should look to spectral classification as an astronomical example where the “mark-1 human eyeball” MAY still be better than the best computer program.

  • coolstar September 24, 2011, 22:16

    Off topic, but since Paul moderates this himself, this is slightly easier than sending him an email: nice new paper by Wes Traub on estimating eta-earth around FGK stars has appeared at
    Very readable paper, I found, with some nicely derivied relationships I’ve been too lazy to do myself.
    He also does a direct comparison between his result and that of Catanzarite and Shao. They differ by about a factor of 30x(!) which seems to be well accounted for by the different assumptions. My personal prejudice, for what little it might be worth, is that Catanzarite and Shao will end up closer to the mark, though Straub’s analysis is strictly more correct.
    The CS paper did have a particular bias that motivated it: that a space-based astrometric mission is needed before an imaging mission to study earth analogs, and that might be a good thing to keep in mind.
    I wish all papers were as well written and organized as this one by Traub!

  • ljk September 24, 2011, 23:01
  • Paul Gilster September 25, 2011, 7:48

    coolstar writes:

    Off topic, but since Paul moderates this himself, this is slightly easier than sending him an email: nice new paper by Wes Traub on estimating eta-earth around FGK stars has appeared at
    Very readable paper, I found, with some nicely derivied relationships I’ve been too lazy to do myself.

    Interesting paper indeed, and one I want to look at after the Orlando sessions. I’ve got a bit of a backlog of papers that are finding their way, slowly but surely, onto the site. We’ll get to this one soon.

  • ljk November 9, 2011, 14:59

    Speaking of this exoplanet finder’s namesake:

    Around Kepler’s “Dream”

    Authors: Jean-Pierre Luminet

    (Submitted on 18 Jun 2011)

    Abstract: Johann Kepler (1571-1630) is sometimes considered as a precursor of science-fiction novels with the writing of “Somnium, sive opus posthumum of astronomia lunaris”.

    In this work published posthumously in 1634 by his son Ludwig, Kepler intends to defend the Copernican doctrine by detailing the perception of the world for an observer located on the Moon. Although Kepler was not the first to write the fantastic account of a voyage from the Earth to the Moon and “pass around a message”, we show here that his “Dream” is made conspicuous under many aspects.

    Firstly, his author is one of the most remarkable minds of the history of sciences. Secondly, the “Dream” constitutes the missing link between the texts of pure imagination by Lucian of Samosata (IInd c. AD), and the fantasy novels based on scientific discoveries by Jules Verne, at the end of XIXth century. In the third place, the complete text presents an extraordinary structure in encased accounts, built on several series of explanatory notes drafted with the passing of years, as new astronomy progressed.

    Also it should be retained the dramatic role that its diffusion, although confidential during Kepler’s lifetime, had on his own personal and family life, as well as the major influence this book exerted on a whole current of speculative literature devoted to space travel.

    Comments: 7 pages. Conference given in march 2010 at Paris Observatory – University of Versailles St Quentin. To be published in Alliages, 2011 (in French)

    Subjects: History and Philosophy of Physics (physics.hist-ph); Popular Physics (physics.pop-ph)

    Cite as: arXiv:1106.3639v1 [physics.hist-ph]

    Submission history

    From: Jean-Pierre Luminet [view email]

    [v1] Sat, 18 Jun 2011 11:06:25 GMT (239kb)