≡ Menu

Speculations on Starless Worlds

Yesterday’s paper from Matt Clement and team reminded us of the enormous transformation that can take place in a planetary system as it lurches toward eventual stability. Gas giants have so much to say about how this process occurs, with their gravitational interactions sometimes ejecting other worlds from the system. Ejected planets are often called ‘rogue’ planets because they wander the galaxy without orbiting a star. Their numbers may be vast.

Clement and team think we may have ejected an ice giant from our early system, as we discussed yesterday. Whatever the case, I’ve been talking about rogue planets for about ten years, and as I look back, I run into intriguing finds like PSO J318.5-22, which is described in a 2013 paper from Michael Liu and colleagues (citation below). Says Liu (University of Hawaii):

“We have never before seen an object free-floating in space that looks like this. It has all the characteristics of young planets found around other stars, but it is drifting out there all alone.”

We learn from the paper that this object shows similarities to young, dusty planets in terms of luminosity, mass, spectrum, etc. It’s also useful because it’s likely a member of the β Pic moving group, meaning we can glean something about its age, about 12 million years. Have a look.

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 Capricorn. The planet is extremely cold and faint, with most of its energy emitted at infrared wavelengths. The image is 125 arcseconds on a side. Credit: N. Metcalfe & Pan-STARRS 1 Science Consortium.

Estimates of the number of rogue planets are all over the map, with one recent one being 2000 objects ranging in size between the Moon and Jupiter per main sequence star. That’s from a 2018 study by Xinyu Dai & Eduardo Guerras (University of Oklahoma), but if we want to jump to the high end, we can go with Louis Strigari (Stanford University) and colleagues: 105 compact objects per main sequence star. See Island Hopping to the Stars for Strigari, and Detection of Extragalactic Planets? for Dai and Guerras, where I give citations.

Now we have word of a small rogue world probably about the size of the Earth. Przemek Mroz (California Institute of Technology) is lead author of the study. Detected through gravitational microlensing, which is about the only way it could have been found with our current technologies, the world is labelled OGLE-2016-BLG-1928. This is the smallest rogue candidate yet identified, with the microlensing event having a timescale of a scant 42 minutes.

Image: An artist’s impression of a gravitational microlensing event by a free-floating planet. Credit: Jan Skowron / Astronomical Observatory, University of Warsaw.

Gravitational microlensing occurs when a foreground star (or planet, in this case) moves in front of a background stellar object, causing the light from the more distant star to be magnified. A brief burst in magnification becomes the signal identifying the star and any associated exoplanet, but in the case of rogue planets, we have no central star. Here’s Mroz on the matter:

“If a massive object (a star or a planet) passes between an Earth-based observer and a distant source star, its gravity may deflect and focus light from the source. The observer will measure a short brightening of the source star. Chances of observing microlensing are extremely slim because three objects – source, lens, and observer – must be nearly perfectly aligned. If we observed only one source star, we would have to wait almost a million year to see the source being microlensed.”

Image: Changes of brightness of the observed star during the gravitational microlensing event by a free-floating planet. Credit: Jan Skowron / Astronomical Observatory, University of Warsaw.

Given how few of these objects we have detected, and the wide range of estimates in the population of rogue planets, it’s hard to make too many statements about them. Are they all the likely result of gravitational interactions in an infant or maturing stellar system? Is there a risk of mistaking ultracool brown dwarfs for planets in this regime (I assume the answer is yes)? In the case of OGLE-2016-BLG-1928, are we absolutely sure there is no host star? Consider this passage from the paper:

The discovery of OGLE-2016-BLG-1928 demonstrates that current microlensing surveys are capable of finding extremely-short-timescale events. Although the mass of the lens cannot be unambiguously measured, properties of the event are consistent with the lens being a sub-Earth-mass object with no stellar companion up to the projected distance of ∼ 8 au from the planet. Thus, the lens is one of the best candidates for a terrestrial-mass rogue planet detected to date. This population of low-mass free-floating (or wide-orbit) planets may be further explored by the upcoming microlensing experiments.

Would OGLE have detected a star here if the planet in question was at Saturn’s 10 AU distance from it? This was one tricky detection, “at the edge of current limits of detecting short-timescale microlensing events,” according to the authors. It’s suggestive of a rogue planet. Looking ahead, we’re also moving toward a space-based microlensing capability in the Nancy Grace Roman Space Telescope (WFIRST). Future surveys will doubtless turn up rogue planets that will add to the ground-breaking work accomplished by dedicated ground-based surveys like OGLE.

And I wonder: What happened to that third ice giant that was ejected from our early Solar System? Is it out there wandering the galaxy, yet another rogue far from the star of its birth?

The paper is Mroz et al., “A terrestrial-mass rogue planet candidate detected in the shortest-timescale microlensing event,” Astrophysical Journal Letters Vol. 903, No. 1 (29 October 2020). Abstract / Preprint. The Liu paper is “The Extremely Red, Young L Dwarf PSO J318-22: A Free-Floating Planetary-Mass Analog to Directly Imaged Young Gas-Giant Planets,” Astrophysical Journal Letters Vol. 77, No. 2 (22 October 2013). Abstract.

{ 17 comments… add one }
  • Michael Fidler November 4, 2020, 9:35

    Thanks for the article, made me think, have been looking into the low end in brown dwarfs, the L, T’s and Y’s. What we are seeing is a mass but have no idea as to its actual shape, think of what an advance civilizations could make in these objects. Condos for 10 to the power of 50 alien beings, with all the comfort of home and national parks the size of California at you front door! They could be completely invisible to outside visitors except for the lensing. This would be a lot easier to do then Dyson type objects around stars, the center of gravity is relatively small and no out side gravity forces to tear it apart. Energy would be no problem since hydrogen could be extracted from space or what is left from the original object. Reference book; The Next Ten Thousand Years, by Adrian Berry, 1974.

  • Michael L A David November 4, 2020, 11:26

    During formation a Star can have many ‘Proto-Planets’ forming around it, all interacting with each other. And many of them (if they are very lucky), make it into adulthood (like so many other things in life), and I am sure, that in many cases, there are a multiplicity of planets around various Stars which have, (against all odds) survived into adulthood. Depending of course the type of Parent Star they orbit. But as you know, many things can occur within a developing Solar System, which can cause the ejection of a Planet, the which I will not go into here, but sufficient to say, that as a result of those many destabilizing factors, several of those Planets can be and are ejected from that System. That is so very true, but what about at the other end of that Stars life cycle, when it either, (as in the case of ‘Sun-like Stars’) just fades into obscurity, thus allowing its family of Planets to slowly ‘drift away’, no longer able to hold onto them. And in many cases that might mean several Planets, an x number of ‘gas Giants’, and even a few ‘Rocky Planets’ as well. Stripping away that Stars legacy, thus allowing those Planets of various descriptions, to then ‘Free Float’ among the Stars. And if that were the case, then the numbers for such ‘free spirits’ might be in their Trillions, because, despite early thinking, ‘Planet Formation’ is as common as ‘Star Formation’, one being the result of the other. And so, you are left with an X number of such Planets, but having said all that, isn’t it believed that ‘Titan’ Neptune’s ‘adopted’ child is just that, a ‘Captured’ Moon, but captured FROM WHERE, was it from within this Solar System itself, OR, is a fugitive from an Alien System? Thus raising to possibility that man such objects get incorporated into other Systems this way. and so that leads to the question I intended to ask you, can any of those ‘Rogue Planets’ ever eventually find their way into another System by accident, thus boosting its Planetary total by however many?!

    • Ron S. November 4, 2020, 14:31

      “can any of those ‘Rogue Planets’ ever eventually find their way into another System by accident, thus boosting its Planetary total by however many?”

      Of course the answer is yes, but that is uninformative. The better question is the likelihood, and that is almost certainly tiny. The problem is one of trajectory (vector: direction and relative speed). For a massive body such as a planet there is a very narrow range of trajectories that will permit a 3-body interaction of the type required for capture.

      Bodies formed from the same nebula as the rest of the stellar system have trajectories far more likely to cause interesting interactions. Mutual trajectory influence can occur when a rogue passing through, and although more likely it is still of low probability. Capture is far more probable with low mass bodies such as comets.

  • Jens Fridthjof Knudsen November 4, 2020, 17:35

    Is there any chance the Webb telescope could find a few of those that are reasonably close to us?

  • Henry Cordova November 4, 2020, 23:41

    Why is ejection and capture always considered as the origin and fate of starless planets? Maybe they were never associated with a star. This is just the low end of the stellar mass function–below the red and brown dwarfs come the gas giant planets, and so on, smaller and smaller in ever-increasing numbers, These may be the most common discrete objects in the universe, although because of their small size they only account for a very small fraction of its mass.

    My guess is Jupiter-like objects are formed routinely, in huge numbers, in all star-forming regions: failed stars. They are all over the place, we just can’t see them because they are dark.

    • Alex Tolley November 5, 2020, 12:10

      Has anyone modeled the possibility of stellar material of small amounts condensing only to planetary sized objects, rather than to even small stars with or without a retinue of planets? Is your argument valid or is there a cutoff where the cloud is too small to gravitationally coalesce at all?

      • Bruce D Mayfield November 5, 2020, 18:21

        This lower limit probably has been modeled, but just thinking about it should tell a person that there must be some lower mass limit beyond which no body can form. Forces working against the gravitational contraction of a small dust and gas cloud would include temperature and external tides, and the high vacuum of nearly empty space will always be working against the weakest force of all. Only the combined masses of huge numbers of particles in a given volume of space will permit collapse to commence.

        We know that objects as low in mass as brown dwarfs can form by themselves. Just guessing, but objects several times the mass of Jupiter might be the extreme low end of what can form on its own.

      • wdk November 5, 2020, 20:35

        A.T, and H.C.,
        The answer to that question or issue might lie in the so-called Jeans Instability or Jeans mass limit. Back in the days when I was in school and stellar theory centered on stars, and exoplanets meant a few astrometry efforts, when I ran into pre-main sequence discussions of
        stellar evolution, it appeared that coalescence of clouds needed at least dozens of solar masses, if not thousands. Granted, not all the mass of the collapsing cloud would be applied to a star, but the early
        theorists were not to worried about what would happen with the excess if the trigger mass was so much more than a solar mass.
        But if I could just find my notes for that particular quarter – it seemed like the lowest mass anticipated coming out of the Jeans Mass limited process was about 0.08 solar mass. I was not very good at following this argument, but what a coincidence? It would be a decade or two later before there were any real data points to consider: doppler detections of exo-planets that had to form somehow. And the argument there was based on the stability or instability of UV and IR detected circumstellar disks. This was a different ball game on mass limits since the star was doing much of the work shepherding clouds and dusts. But a modern review of
        the Jeans’ mass limit – or cloud instability – might be worth examination. My guess though is that circumstellar disk formation and subsequent planetesimal formation is simply much more efficient for generating planets – including kicking them out when small planets interact with brown dwarf or jovian objects.

  • ljk November 5, 2020, 11:25

    Makes me wonder if we could find out if any of these objects are actually artificial WorldShips? Or if a species found a way to leave on their home world for a new star system? Some of these rogue worlds may not be so rogue after all.

    • Alex Tolley November 5, 2020, 16:52

      Interesting idea. However, it seems rather drastic to me, dragging around a massive world to migrate with. These worlds would need a huge amount of energy to escape their home system, more to navigate it, and then to rendezvous with a target star.
      I prefer Michael Fidler’s idea that they have become permanent homes. They are very hard to find in deep space, so reducing the risk of a predator civilization destroying it or the inhabitants. The worlds wander, but the inhabitants have no destination to reach, just the intent to remain hidden. A cold world would allow very deep occupation with very little heat flow at depth. They could house many independent biospheres too.
      OTOH, they could be large versions of Rama, intended to be used as zoos to contain inhabitants of many worlds in biospheres that reproduce their home world’s.

      • Michael C Fidler November 6, 2020, 10:05

        They could also explore and that may be the only signals that may be visible. I would suspect there may be normal trade routes. You see the problems we encounter living on a solar world, just think how many aliens that may exist. A Jupiter mass object would make a huge home and there is the possibility of making groups of objects from the large mass. Earthling’s love the adventure and novelty of what is found by exploring, just imagine huge worlds in deep space that could watch the exploration of billion of planets and all the unusual creatures that live on them.

    • Robin Datta November 5, 2020, 22:36

      Perhaps we should assign a high priority to “rogue” bodies for SETI-type monitoring: “Tractor crew to base: we have docked with the rogue comet, and will shortly inform you of our trajectory to orbit”.

  • Geoffrey Hillend November 5, 2020, 16:00

    There is a mass limit of the protostar gas and dust cloud or diffuse nebula for it to make a body in space like a brown dwarf. Once it gets too much smaller than that there is not enough mass to start a collapse according to current astrophysics. Consequently, gas giants are probably rogue.

  • Michael C Fidler November 7, 2020, 10:16

    Have been trying to calculate the number of objects that could be within 1/2 light year of the Sun. Use the rough figure of 200 objects per star and the nearest star system being a little over 4 light years it should be a radius of 2 light years around the Sun. Volume of a sphere formula:
    volume = 4/3 * π * r3 or use the calculator here; https://www.gigacalculator.com/calculators/volume-of-sphere-calculator.php

    This gives the amount of 33.51 with 200 planets, divide that by 4.1888 for the volume of a sphere of 1 light year radius gives volume of just 8 which would have 25 free floating planets. The 1/2 radius Light year divide by 8 again and end up with 3 free floating planets within 1/2 light years of the Sun. Anyone see a flaw in the amounts please advise!

    The discovery of these objects have 3 possible ways, of course the gravitational microlensing events. Something this close may also cause an occultation of a star like some of the more distant Kuiper belt objects. This would give us the ability to see the actual shape of these planets, rings, satellites or artificial shapes. Any atmosphere could also be probed. The third method would be looking for extremely cold objects in data archives that are moving at a high rate. The Large Synoptic Survey Telescope / Rubin Observatory should turn up some of these by all three methods.

    If something unusual is found a interstellar probe could be sent at .25 the speed of light, like the Breakthrough Starshot, that could reach the objects in 2 years with a 6 month data return time. So now is the time to start looking for these almost invisible/ cloaked planets and their technosignatures.

    • Alex Tolley November 7, 2020, 17:44

      The easiest calculation starts with your 200 objects within a sphere of radius 2 ly (halfway to Alpha Centauri).

      To get the numbers for each volume, 1/8 for r=1 ly, 1/64 for 1/2 ly.
      Multiply 200 by these ratios to get the number in each sphere radius.

      However, I think your basic methodology is wrong due to the volumes not packing compactly, although this is not important given your initial estimate of 200 objects/star. I would start out with the number of stars in a volume of radius r, multiply that number by 200/stars, then repeat your calculation.


      12,500 star SYSTEMS of all types of stars within 100 ly.
      (src: https://www.strudel.org.uk/blog/astro/000480.shtml)
      objects = 200 x 12500 = 2.5 million.
      Therefore object within 1 ly = 2.5E6/(100^3) = 2.5
      Therefore object within 1/2 ly = 2.5E6/(200^3) = 0.3

      So about an order of magnitude smaller than your estimate using the distance to the nearest star.

      Using stars rather than star systems for multiple star systems increase the numbers by about 30%.

      Therefore the key variable is the initial guess of objects per star[system].

      As for organizing a search, and a flyby, this seems reasonable, especially if Breakthrough Starshot has been sufficiently developed to allow for probes to be sent out on demand and report back their findings. If that can be done before the mission to Proxima, it would make a good precursor mission after reporting on Kuiper belt and inner Oort cloud objects.

      • Michael C Fidler November 8, 2020, 10:09

        Yes, but Alpha Centauri stole ours!!! Looks like Gaia, Nancy Grace Roman Space Telescope and Rubin Observatory – LSST will have to be online for a while before we will have those numbers even close. Your method looks like it is more accurate going from the outside in instead of the inside out? But there is 4 times as many Red Dwarfs then the stars we see visually in the sky.

      • Michael Fidler November 10, 2020, 2:32

        Looks like the 4th method is already getting results!

        NOVEMBER 9, 2020
        Faint super-planet discovered by radio telescope.

        BDR J1750+3809 (dubbed “Elegast” by the discovery team) was first identified using data from the Low-Frequency Array (LOFAR) telescope in Europe, and then confirmed using telescopes on the summit of Maunakea, namely the International Gemini Observatory and the NASA InfraRed Telescope Facility (which is operated by the University of Hawai’i). Directly discovering these objects with sensitive radio telescopes such as LOFAR is a significant breakthrough, because it demonstrates that astronomers can detect objects that are too cold and faint to be found in infrared surveys, and perhaps even detect free-floating gas-giant exoplanets.


        Direct radio discovery of a cold brown dwarf.


Leave a Comment