≡ Menu

‘Rogue’ Planet Population in the Galaxy

We’ve recently looked at gas giant planet formation, and specifically the stages in which Jupiter seems to have formed — this is the work of Thomas Kruijer (University of Münster) and colleagues as summarized in A Three Part Model for Jupiter’s Formation. Whether or not the details of Kruijer and team’s model are correct, it seems evident that gas giants must form quickly, based on current theories. These involve the formation of a large solid core, with gas accretion building up a thick atmosphere at a time when the disk around the parent star is still rich in materials.

In this thinking, planets like the Earth come along much later than the gas giants that are the first to form. Get a few million years into the evolution of a stellar system and there should be evidence of a gas giant, if one is going to form, but terrestrial worlds can take up to 100 million years to emerge. This has captured the interest of Nader Haghighipour (University of Hawaii), whose work was presented at the General Assembly of the IAU meeting in Vienna.

Haghighipour is interested in so-called ‘rogue’ planets, and his focus is on early system formation as a way of exploring their population in the galaxy. Bruce Dorminey has a good entry on this in Forbes, recounting the basics of Haghighipour’s statistical study, one that involved 500 simulations of terrestrial planet formation, including multiple models.

What gets intriguing here is that Haghighipour believes most stellar systems wouldn’t be very good at ejecting planets out of the parent system. This has bearing, of course, on the question of how many rogue planets, free-wandering worlds without a star, might be out there.

It is an issue that is a long, long way from being resolved. Without the reflected light of a primary, such planets are extremely hard to detect except through gravitational lensing, where the light of a background star may be affected by the curvature of spacetime in the presence of a gravitational well, the distortion giving us information about the intermediate object. Used in the exoplanet hunt, gravitational microlensing allows us to spot planets down to rocky, terrestrial-sized worlds around stars thousands of light years away, with the significant limitation that such observations are one-off affairs. You can’t realign the stars to do a second run.

The Microlensing Observations in Astrophysics (MOA) survey, based in New Zealand, as well as the Optical Gravitational Lensing Experiment (OGLE), using a 1.3 meter telescope in Chile, have found a small number of possible ‘rogue’ planets of roughly Jupiter’s mass. A 2011 paper made the case that given microlensing probabilities — and taking into account the efficiency of the equipment at these installations and the rate of such lensing — there could be as many rogue planets in the Milky Way as there are stars. Louis Strigari (Stanford University) has likewise estimated high numbers of rogue planets ranging from Ceres-size on up to gas giants.

Image: One apparent free-floating planet turned up in a search for brown dwarfs. This multicolor image is from the Pan-STARRS1 telescope, showing PSO J318.5-22, in the constellation of Capricornus. The planet is extremely cold and faint, about 100 billion times fainter in optical light than the planet Venus. Most of its energy is emitted at infrared wavelengths. The image is 125 arcseconds on a side. Credit: N. Metcalfe & Pan-STARRS 1 Science Consortium.

But at the same IAU meeting, according to Dorminey’s report, Yutong Shan (Harvard University) presented evidence from the K2 mission, using data from both the revived Kepler and ground-based surveys. Shan’s team was hunting for free-floating planets, but while they found candidates, none were conclusive — in other words, the detections may be brown dwarfs.

For his part, Nader Haghighipour believes that there aren’t as many rogue planets to find as some believe. The reason: It’s when the protoplanetary disk is seething with youthful energies that planets are likely to be ejected — few terrestrial worlds available then — and only 1-2 percent of these would leave the system. Shan’s Kepler study is ongoing, but as Dorminey notes, the early results seem to support Haghighipour in his doubts over rogue planet ejections.

But we are early in this work. Let me also draw your attention to work at the University of Oklahoma from Xinyu Dai and Eduardo Guerras. Working with data from the Chandra X-Ray Observatory and microlensing models of their own design, the duo have been studying the black hole at the center of quasar RX J1131–1231. Here we have a background quasar being lensed by the foreground black hole, and Dai and Guerras show that these emissions near the Schwarzschild radius of the black hole can be affected by planets nearby in its galaxy.

What we get here is another ‘rough cut’ at rogue planets, this time in another galaxy, and in it, the authors calculate about 2000 objects per main sequence star, in sizes ranging from the Moon to Jupiter. Dai and Guerras, in other words, come up with numbers that come closer to supporting Strigari than Haghighipour. All of which makes a strong case that on the subject of rogue planets in this or any galaxy, what we still don’t know vastly outweighs what we do.

The microlensing paper mentioned above is Sumi et al., “Unbound or distant planetary mass population detected by gravitational microlensing,” Nature 473 (19 May 2011), 349-352 (abstract). The Strigari paper is “Nomads of the Galaxy” Monthly Notices of the Royal Astronomical Society Vol. 423, Issue 2 (21 June 2012 – abstract). The paper on RX J1131–1231 is Xinyu Dai & Eduardo Guerras, “Probing Planets in Extragalactic Galaxies Using Quasar Microlensing,” Astrophysical Journal Letters Vol. 853, No. 2 (2 February 2018). Abstract.


Comments on this entry are closed.

  • Gideon Marcus September 4, 2018, 11:14

    Excellent and fascinating article. In 1963, we don’t even have brown dwarfs (the concept was advanced this year as “black dwarfs”, but the name is confusing.)

    • Ljk September 4, 2018, 17:55

      I think they should be called collapsars, to match with pulsars and quasars.

  • H. Floyd September 4, 2018, 15:35

    Tantalizing — and spooky. Just as we’re getting serious about near-Earth objects, here come stealth rogue planets.

    Imagine if our interstellar visitor ‘Oumuamua — first spotted only a year ago next month — had been a dark Neptune. It’s enough to provoke Melancholia.

    • Robin Datta September 4, 2018, 17:56

      A Neptune-sized rogue planet on a Oumuamua trajectory? How much disruption would it cause in the Solar System?

      • H. Floyd September 4, 2018, 20:40

        Little, if any? I also imagine there’d be little, if anything, we could do but watch.

        We find astronomy thrilling because it shocks and matures the ego without make-believe. Science gives us tangible Copernican moments for the Self.

        A rogue gas giant visibly passing into our Solar System would be a world-shatteringly Copernican moment for the whole species. I wonder what that would do to us?

        • Paul Gilster September 5, 2018, 8:47

          And just last night I re-watched The Man from Planet X, which involves a rogue planet making a fast pass through our system. Great fun and eerily directed by the gifted Edgar G. Ulmer.

  • Sanford Z Meschkow September 4, 2018, 16:39

    Yes, H. Floyd, it’s enough to make one wonder if it’s time to remake George Pal’s “When Worlds Collide” featuring a rogue brown dwarf with a close-in Earth-sized moon that is easier to terraform than Mars.

    • H. Floyd September 4, 2018, 22:15

      I’m ashamed to say I missed that classic film, but I’ll fix that!

    • ljk September 5, 2018, 8:57

      No remakes. The 1951 does not need it.

    • ljk September 5, 2018, 9:06

      This is from the Wikipedia entry on When Worlds Collide:


      The 1998 film Deep Impact originated as a joint remake of When Worlds Collide and an adaptation of the 1993 Arthur C. Clarke novel The Hammer of God, and the project was originally acknowledged as such, although the finished film did not acknowledge any of its sources since it was judged as being different enough to not require it.[17]

      Paramount Pictures began preproduction on a remake of When Worlds Collide circa 2013. As of August 25, 2015, no release date had been announced.[18]

      If they do to WWC what they did to The Andromeda Strain in 2008, may it remain in the can on a shelf – and no I don’t care if that is an outdated technical film term.

  • Antonio September 4, 2018, 17:02

    I always favor experimental data over simulations, so I’m not with Haghighipour here, I think there are plenty of rogue planets.

  • Brett Bellmore September 4, 2018, 18:08

    ” with the significant limitation that such observations are one-off affairs. You can’t realign the stars to do a second run.”

    With telescopes in space you could. Not realign the stars, but realign the telescopes. With multiple scopes to get some idea of the direction the image is moving, you should be able to reacquire it.

    • Andrei September 6, 2018, 5:58

      We know the microlensing have happened.
      Yes with only one telescope we do not know in which direction the object have moved, if it have moved from ‘north’ or at 90 degrees – from east to west or opposite.
      The lightcurve look the same.

      But it would take a wide network with many highly movable telescopes in space to catch a second event – unless we were very lucky.

      • Andrei September 6, 2018, 6:02

        Wrong by me, microlensing only work for objects in other galaxies. For our galaxy a search would be for “occultaions”.

        • Paul Gilster September 6, 2018, 9:18

          Actually, microlensing is often used to detect exoplanets here in our galaxy. It’s a useful tool because it can detect planets at much greater distances than radial velocity or transit techniques.

          • Andrei September 6, 2018, 20:06

            Thank you for the correction Paul. :)

  • wdk September 4, 2018, 19:28

    While the current discussion is somewhat skeptical about the prospects for significant numbers of rogue planets, I’d like to submit some notes taken from a lecture given by William Bottke of the Southwest Research Institute about 7 years ago.

    Dr. Bottke, while in Houston, gave an evening’s lecture on the Nice ( as in France) Model of solar system formation and development at the Lunar and Planetary Science Institute, and I mused on the matter for our local AIAA newsletter, much in the same spirit as Centauri Dreams.

    To get to the heart of things:
    When the Nice Model was first presented in 2005 in sev-
    eral papers by the team of Gomes, Levison, Morbidelli and Tsiganis, it was proposed that the four large gas giant planets (Jupiter, Saturn, Uranus and Neptune) resided between ~5.5 and ~17 AUs,
    more tightly packed than they are now, though not with Jupiter any closer to the sun

    Beyond this a large, dense disk of small rock and ice
    planetesimals remained (~35 Earth masses), extending to
    35 AUs. Over several hundred million years, as a result
    of encounters such as described, Jupiter edged slightly
    further toward the sun to its than its present position and the remaining planets moved farther out, scattering small bodies to their current general positions, as well as ejecting the majority deep into space.

    But when Jupiter and Saturn cross into a mutual 1:2 mean
    motion resonance, the entire system of outer planets becomes unstable, with in-creased eccentricities and eventual angular momentum
    exchanges, scattering most of the remaining primordial disk.
    In the inner solar system this sudden influx of matter was
    registered as the Late Heavy Bombardment.

    It is noted that in ~50% of the Nice simulation cases, within
    the first billion years of solar system history Neptune and
    Uranus change places with respect to distance from the

    Simulations related to the Nice Model attempt to identify configurations of the early solar system which can
    best explain the stable configuration of planets and distribution of small bodies we perceive now. Thousands of initial configurations were tested for the early reports, as described above, and the pro-
    cess has continued to the pre-sent day. The result: we see a
    planetary migration mostly OUTWARD.

    The more recent findings reported by Bottke and by other
    investigators indicate that even better extrapolations
    forward to present day conditions can be obtained if a
    fifth giant planet is included in the early solar system set. As
    reported by David Nesvorny of SwRI, based on 6,000 simulations
    of the early solar sytsem, with 5 giant planets,
    the end state was ten times more likely to lead to today’s configuration of planets rather than in cases involving only four such bodies. The fifth
    gas giant is “ejected” completely from the solar system
    by a close encounter with Jupiter. “This possibility appears to be conceivable in view of the recent discovery
    of a large number free-floating planets in interstellar
    space, which indicates that planet ejection should be
    common.”, concludes the study.
    So, if you take this as anecdotal evidence…

  • kzb September 5, 2018, 8:35

    The quasar microlensing puzzle goes back to at least the 1990’s.
    These studies apparently show a large number of planetary mass objects in the lens galaxy. Notice in this paper, the 2000 planets per star is actually a lower limit. So it is “at least” 2000 planets per star. The problem is microlensing studies in our own galaxy don’t give us anything like this number. IMHO, this is because they are not planets. They are AU-sized gas clouds of planetary mass. At large distances these act as gravitational lenses, but closer to home they do not, so they are not detected by this method. I just wish more work could be done on this subject area; it seems our closest interstellar neighbour could be one of these clouds and most importantly they could be the galaxy dark matter. This site is a good starting point to see what I am on about:


    • Alex Tolley September 5, 2018, 11:28

      That is where I was tip-toeing. in my question below. The distribution is important to account for the rotation rate of stars in galaxies. The combined mass to account for the macro-effects on gravitational lensing and obviate the need for some special, but so far undetected, dark matter..

      • kzb September 10, 2018, 13:44

        My point was, these objects are unlikely to be planets. They are compact gas clouds and they have escaped detection in our own galaxy, except as so-called Extreme Scattering Events in the radio.

  • Alex Tolley September 5, 2018, 9:51

    Is there enough combined mass, and any spatial distribution, in these rogue planets to have any measurable impact on the visible stars in a galaxy?

  • Frank Smith September 5, 2018, 11:49

    If there are millions or billions of rogue planets in the galaxy, would not a fair number of stars eventually gravitationally capture one of them?

    I would think the orbit of a captured rogue planet would be really far out and eccentric, but even so, a Jupiter-sized distant planet should be detectable.

  • andy September 5, 2018, 13:42

    The Sumi et al. (2011) results on the apparent discovery of a large population of free floating/wide orbit giant planets by microlensing have been subsequently challenged, see Mróz et al. (2017). This brings the inferred population of planets from microlensing more in line with what would be expected from observations of young clusters.

    • kzb September 6, 2018, 7:19

      But the paper under discussion finds thousands of planets per star.
      This is by microlensing. It is not the first study to find this by any means. What it adds to the debate is they are more certain the effects are truly due to objects in the lensing galaxy rather than intrinsic to the quasar itself.
      So, you have this massive discrepancy between microlensing in a distant galaxy finding thousands of “planets” per star, and local studies which find 3 or 4 orders of magnitude fewer. I think I have given the possible answer to this problem in my earlier post.

  • Tom Baty September 5, 2018, 17:58

    One more add about “When World Collide” The book which the movie was based on was written in the early 1930’s. Like many movies I think the book was better as the movie left out half of the story and was more descriptive
    The squeal book “After Worlds Collide” came out in the mid-30’s . Although interesting it not as good as its predecessor. That one was not made into a movie. Don’t get me wrong I thought the movie was one of the best SF from the 1950’s along with “Destination Moon” Both are classics.

    • ljk September 6, 2018, 9:26

      George Pal wanted to make After Worlds Collide, but his 1955 SF film, The Conquest of Space, bombed at the box office, so there was no support for the sequel.

      • Paul Gilster September 6, 2018, 9:33

        I hadn’t realized that. Did Pal get After Worlds Collide up to having a working screenplay?

        • Tom Baty September 6, 2018, 14:06

          IJK has it right…I saw the Conquest of Space too. It was ok but nothing great. From what I know Paul they never got as far as a screenplay. for After Worlds Collide.

          • Paul Gilster September 6, 2018, 14:34

            Yes, Conquest of Space really seemed wooden to me. What a shame if it shot down what might have been a much better film.

            • ljk September 6, 2018, 15:49

              Conquest was also not much fun. It had some silliness in the beginning with that old standby, Da Guy from Brooklyn, then it got downright serious and depressing. Not exactly a cheerleader for space exploration – the exact opposite of the book of the same title which the film was supposed to be influenced by.

              The animated Walt Disney series on Man in Space was much better in every way. Sorry, George, and I am big fan of Destination Moon, War of the Worlds, and WWC.

              • Alex Tolley September 7, 2018, 15:58

                The religious overtones at the end were also rather jarring, but still being echoed today whenever Mars colonization is discussed, but also secularized with a “Prime Directive”.

                • ljk September 10, 2018, 9:29

                  It probably comes from the ancient belief that Earth and humans were not at the center of the everything as has been wrongly interpreted for centuries, but actually at the bottom of the Universe similar to a big cosmic pit, with only Hell being lower at the very center/bottom. As one moved higher into the heavens, things were considered more pure, with God and Heaven with His Angels at the very top.

                  This is one reason why Galileo and his contemporaries were so excited (or angry and frightened) about the heliocentric theory, because it elevated us to being among the stars, assuming every star was a sun with its own retinue of planets and moons.

                  So those who still think of the heavens as being Heaven would consider it outrageous and blasphemous for mere mortal humans to travel there via rockets rather than the Judeo-Christian religion’s rules and regulations. With the way things are going education-wise these days, I would not be surprised if such attitudes are still held in certain places.

                  I am not sure if this is anecdotal or not, but Nikita Khrushchev once said that Soviet cosmonauts had penetrated the heavens and did not see either God or angels floating around up there – a holdover from this very old view.

                  • ljk September 10, 2018, 9:44

                    This is the book that really opened my eyes to how ancient Western thinkers really viewed Earth and humanity’s literal place in the cosmic scheme:

                    The Book Of The Cosmos: Imagining The Universe From Heraclitus To Hawking, by Dennis Danielson (Editor)


  • Dmitry Novoseltsev September 6, 2018, 2:45

    Here you can pay attention to possible erroneous assumptions in the Drake equation (and derived from it the Seager equation with biomarkers): that civilizations can occur only near stars and on each inhabited planet there can be only one civilization.
    Imagine that on a “rogue” exoplanet such as ice earth or super-earth, covered with ice crust and frozen gases, due to tectonic and geothermal activity occur ice seas and oceans of liquid water involved in the exchange of matter with solid rock crust, but separated from each other by a thickness of ice.
    At successful coincidence of circumstances each may occur own isolated subbiospheres. That is, the level of biodiversity on such a wandering “rogue” exoplanet can be much higher than on Earth with a single biosphere, where the atmosphere and hydrosphere are in a state of constant circulation, moving biological species.
    In this case, external observation will not detect any typical biomarkers, because the seas and oceans are separated from the surface by a thick layer of ice, impervious to gas exchange.
    Then you can present a good case when in each subbiosphere comes up with its own intelligent life. Over time, they can create their SETI analogue based on the connection through the ice thickness on acoustic or seismic waves.
    Available to monitor their technosignature will be the only advanced civilization of the ice wandering extrasolar planets, has found a way to come to the surface and organize a large-scale operation.
    If the number of wandering exoplanets in the galaxy by the new estimates will be significant, the Drake equation and the Seager equation may require revision.

    • Alex Tolley September 7, 2018, 15:54

      The problem with such a scenario is that such worlds, like icy moons, have a very low energy flow. Terrestrial life is primarily driven by sunlight, a small part of which, is used to create local order (life). This provides for a rich biosphere that dwarfs a biosphere driven by chemistry alone seeping up from the mantle.

      Hydrothermal vents provide local energy-rich environments, but even here, organisms have migrated to the hotspot from elsewhere. Once the smoker “dies”, so does the local ecosystem. Without a larger, sun-driven biosphere to supply evolved forms to draw on and populate vents, those vents would likely be just bacterial.

      Clarke suggested that Europa had had many “civilizations”, each living around its hydrothermal vent until it died (2061: Odyssey Three. While it makes for a good tale, I don’t see the Europan sub-surface ocean being able to support such rich ecosystems and teh energy flow is so low.

      • ljk September 10, 2018, 9:38

        I cannot find the post at the moment, but Arthur C. Clarke once wrote a piece about the cracks on Europa being some kind of road network.

        Clarke wrote it in such a way that it could be interpreted either as humor or serious, perhaps so that if they were actual roads made by native Europans he could claim credit for the idea, whereas if they were not, Clarke could just say he was just being funny – didn’t you get the joke?!

        Reminds me of the Soviet astronomer Igor Shklovskii, who co-authored the 1966 book Intelligent Life in the Universe with Carl Sagan, claiming circa 1960 that the Martian moon Phobos was a giant hollow satellite containing all the knowledge of the ancient Martian race. In later years he said he was only joking, but if you look at the earlier accounts, the man was keeping a pretty straight face literary-wise.

  • Dmitry Novoseltsev September 10, 2018, 10:10

    Well, Europe and other ice moons are quite small, and their “geothermal” activity is due only to the tidal forces of the giant planets. But the ice super-earth, even wandering without its own star, can cool down long enough, and its thermal activity can also be very high, at the level of supervolcans. In this case, the released energy is not distributed in the entire volume of the single atmosphere and hydrosphere and is not spent on their mechanical mixing, but is localized in liquid reservoirs near cracks in the planetary crust. Of course, additional modeling is required, but the probability is non-zero.