A History Of Planet Detection in 60 Seconds

Last week I gave my first proper talk to a conference of PhD students from nearby universities. Being an easily distracted man, rather than actually write my talk, I decided to spend the day before putting together an animation of the entire history of Planet Detection, from 1750 to 2015. It shows the orbital period (x-axis), planet mass (y-axis), radius (circle size)* and detection method (colour) of the 1800+ planets now known.

Made by Hugh Osborn
Made by Hugh Osborn

More Details

The idea of this plot is to compare our own Solar System (with planets plotted in dark blue) against the newly-discovered extrasolar worlds. Think of this plot as a projection of all 1873 worlds onto our own solar system, with the Sun (and all other stars) at the far left. As you move out to the right, the orbital period of the planets increases, and correspondingly (thanks to Kepler’s Third Law), so does the distance from the star. Moving upwards means the mass of the worlds increase, from Moon-sized at the base to 10,000 times that of Earth at the top (30 Jupiter Masses).

The colours are also important – dark blue shows the solar system planets (which include Ceres and Pluto for a few deacades each); In light blue are RV planets, which began the gold rush in 1995 with the discovery of 51 Peg; In maroon are Direct Imaging planets; in orange the microlensing discoveries; and in green those planets found by the transit method.

You might see a few patterns beginning to emerge:

The top left has a dense cluster of large worlds. These are the Hot Jupiters. We know of loads of these, even though they’re pretty rare, simply because they are easiest to find. Being so close to their star they produce the biggest radial velocity signals (light blue) and are most likely to transit (green). Ground-based transit surveys like WASP cant find anything beyond ~15 days, causing the sparse region to the right of this group.

The top right cluster is a population of Jupiter-like worlds that Radial Velocity is best at finding – anything beyond 10 years is too long at the moment to have a full signal.

The bottom group is from the Kepler space telescope. This clustering is the only one that’s actually real and not just a systematic effect. This is because Kepler was capable of finding every type of planet down to ~1 Earth radius. So this clustering shows that there are more Earth and super-Earth sized planets than any other. Hopefully we can begin to probe below it’s limit and into the Earth-like regime, where thousands more worlds should await!

Hope you enjoy it, and feel free to borrow it for your own use!

*Where Mass or Radius were unavailable I used the Mass-radius relations of Weiss & Marcy. Information from exoplanet.eu, so it might be a bit wrong. Thanks to Matt Kenworthy for suggestions. Pulsar planets are not plotted.

14 thoughts on “A History Of Planet Detection in 60 Seconds

  1. I like the sweep of time, but between 1800 and 1850 there were four extra planets – not just Ceres, but Vesta, Pallas, and Juno as well.

    1. You’re right – turns out the Wikipedia article I used (which just showed Ceres) wasn’t accurate. But, even if Vesta, Pallas and Juno were plotted, they’d be way off the bottom of the graph.

  2. Hey, awesome graph! I liked it immediately and linked it to my exoplanets professor, he had a nice suggestion that would remove the observational bias.
    If you plot period on the abcissa instead of mass, he suggested,

    This removes that fact that a short period planet is discovered in a
    short time because one usually needs at least one period (more like
    three for transits and two for RV) to make sure it is a planet.
    — Günther Wuchterl

    (hopefully I did the tags right)

  3. Why are pulsar planets not plotted? It’s a bizarre decision to leave out the first-ever confirmed extrasolar planets on a graph that purports to be “an animation of the entire history of Planet Detection”. Until it is in fact “entire”, you shouldn’t label and promote it as such.

    1. I may be somewhat bias based on my particular field of exoplanets, but to me Pulsar planets aren’t extrasolar planets. “Planet”, in my book, first and foremost means orbiting a star. Its the same reason I don’t plot the handful of “free-floating” worlds. No star: no planet. And given that 90% of papers in my field cite the ‘first exoplanet’ as 51 Peg in 1995, I don’t think I’m alone in this.

      1. > “Planet”, in my book, first and foremost means orbiting a star.

        Pulsars are neutron *stars*, as I’m sure you know. Show me an official description that says they don’t qualify as such. Making up arbitrary definitions for personal preference doesn’t change anything.

        Your particular field of exoplanets is well-known for wilfully ignoring the pulsar planets. (Perhaps more because they’re not considered to be interesting candidates for habitability?) But that doesn’t erase their existence. By ignoring them, you ignore a reality that has an impact on theories of planetary formation and evolution, as well as influencing the direction that exoplanetary research takes. So yes, I would agree with you that this is a bias, but not a good one. Remember that the pulsar planets were a surprise. There are probably other surprises in store for your field, which are less likely to be discovered if everyone involved is wearing blinkers and theorists are working with incomplete data. Something to think about. I hope.

  4. Hugh, I really like this article and the animation. It certainly highlights how active exoplanet science is now and how far it has come in a relatively short time.

    However, I must agree with NMGyrl that it seems, in my opinion, ignorant to leave out pulsar planets. It is even more confusing that your personal definition of a star does not encompass pulsars. Since when is a neutron star not a star? Maybe you can explain further. As far as I understand PSR B1257+12 was the first discovered/confirmed exoplanet – just because it doesn’t orbit a main sequence type does not change that, right? Neither does the sheep effect of exoplanet literature quoting on mass otherwise 😉

    I am happy to change my mind if there is suitable evidence otherwise, based on IAU description of planets/stars.

    As a formation scientist I don’t have any particular allegiance to a detection method, but pulsar planets are particularly interesting anyway since they may be evidence of a secondary formation mechanism.


    1. Well I guess I was taught that to tick the box that says ‘star’, there must be fusion going on. That’s certainly what draws the ~15Mjup mass limit, and why my astro lecturers told me White Dwarfs and Neutron “stars” aren’t officially stars either.

      But this is all semantics, isn’t it? When I update the plot for any new planets discovered, I’ll probably just add them in.

Leave a Reply

Your email address will not be published. Required fields are marked *