≡ Menu

Interstellar Wanderers

Because of my fascination with exotic venues for astrobiology, I’ve always enjoyed Karl Schroeder’s novels. The Canadian writer explored brown dwarf planets as future venues for human settlement in Permanence (2002), and in his new book Lockstep (soon to be published by Tor, currently being serialized in Analog), Schroeder looks at ‘rogue’ planets, worlds that move through the galaxy without a central star. Imagine crimson worlds baked by cosmic radiation, their surfaces building up, over the aeons, the rust red complex organic molecules called tholins. Or consider gas giants long ago ejected from the system that gave them birth by close encounters with other worlds.

Objects like these and more are surely out there given what we know about gravitational interactions within planetary systems, and they’re probably out there in huge numbers. I’m not going to review how Lockstep uses them just yet — in any case, I haven’t finished the book — but we’ll return to its ingenious solution to time and distance problems in a future post. Right now I just want to mention that one of Schroeder’s characters muses upon ‘a hundred thousand nomad planets for every star in the galaxy.’ Now that’s some serious real estate.

If the number sounds like a novelistic exaggeration, it’s nonetheless drawn from recent work. Schroeder is invoking the work of Louis Strigari (Stanford University), who has studied the possibilities not only of planets ejected from their own systems but those that may form directly from a molecular cloud. The figure of 105 free-floating planetary objects for every main sequence star is from a 2012 paper in Monthly Notices of the Royal Astronomical Society (you can read more about Strigari’s ideas in ‘Island-Hopping’ to the Stars).

Rogue planets would be tricky to find but gravitational microlensing should help us set constraints on their actual numbers, and as we’ll see below, direct imaging has its uses. If rogue worlds are available in such quantities, we can imagine a starfaring culture capable of exploiting their resources. We can even speculate that a thick atmosphere that can trap infrared heat coupled with tectonic or radioactive heat sources from within could sustain elemental forms of life even in the absence of a star. Tens of thousands of objects in nearby interstellar space would obviously be a spur for exploration.

A Newly Found Orphan World

Eighty light years from Earth floats a solitary planet that has been discovered through its heat signature in data collected by the Pan-STARRS 1 wide-field survey telescope on Maui. In mass, color, and energy output, the world is similar to directly imaged planets. As you might expect, PSO J318.5-22, a gas giant about six times the mass of Jupiter, turned up during a search for brown dwarfs, delving into the datasets of a survey that has already produced about 4000 terabytes of information. The discovery was then followed up through multiple observations by equipment on nearby Mauna Kea, with spectra from the NASA Infrared Telescope Facility and the Gemini North Telescope indicating the young, low-mass object was not a brown dwarf.


Image: Multicolor image from the Pan-STARRS1 telescope of the free-floating planet 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.

“We have never before seen an object free-floating in space that that looks like this. It has all the characteristics of young planets found around other stars, but it is drifting out there all alone,” explained team leader Dr. Michael Liu of the Institute for Astronomy at the University of Hawaii at Manoa. “I had often wondered if such solitary objects exist, and now we know they do.”

The find is interesting on a number of levels, not least of which is that observations of gas giant planets around young stars have shown that their spectra differ from those of L- and T-class brown dwarfs. Young planets like these, according to the paper on this work, show redder colors in the near-infrared, fainter absolute magnitudes at the same wavelength and other spectral peculiarities that suggest the line of development between brown dwarfs and gas giant planets may not be as clear cut as once assumed. The paper makes clear how complex the issue is:

PSO J318.5−22 shares a strong physical similarity to the young dusty planets HR 8799bcd and 2MASS J1207−39b, as seen in its colors, absolute magnitudes, spectrum, luminosity, and mass. Most notably, it is the first field L dwarf with near-IR absolute magnitudes as faint as the HR 8799 and 2MASS J1207−39 planets, demonstrating that the very red, faint region of the near-IR color-magnitude diagram is not exclusive to young exoplanets. Its probable membership in the β Pic moving group makes it a new substellar benchmark at young ages and planetary masses.

A landmark indeed, and here the Beta Pictoris moving group, a collection of young stars formed about twelve million years ago, is worth noting. Beta Pictoris itself is known to have a young gas giant planet in orbit around it. The newly detected PSO J318.5−22 is lower still in mass than the Beta Pictoris planet and it is thought to have formed in a different way. The paper goes on:

We find very red, low-gravity L dwarfs have ≈400 K cooler temperatures relative to field objects of comparable spectral type, yet have similar luminosities. Comparing very red L dwarf spectra to each other and to directly imaged planets highlights the challenges of diagnosing physical properties from near-IR spectra.

The beauty of objects like these from an astronomical point of view is that we don’t have to worry about filtering out the overwhelming light of a parent star as we study them. Co-author Niall Deacon (Max Planck Institute for Astronomy) thinks PSO J318.5−22 will “provide a wonderful view into the inner workings of gas-giant planets like Jupiter shortly after their birth.” The discovery also gives us much to think about in terms of future explorations as we contemplate a cosmos in which perhaps vast numbers of planets move in solitary trajectories through the galaxy.

The paper is Liu et al., “The Extremely Red, Young L Dwarf PSO J318-22: A Free-Floating Planetary-Mass Analog to Directly Imaged Young Gas-Giant Planets,” in press at Astrophysical Journal Letters (preprint). Also intriguing is Abbot and Switzer, “The Steppenwolf: A proposal for a habitable planet in interstellar space” (preprint).


Comments on this entry are closed.

  • David Cummings October 10, 2013, 9:00

    Maybe once we get out to the Oort Cloud we’ll discover that we can just keep going, step-by-step, across the interstellar void, without requiring any excessively long hops.

    In any case, Oort Cloud humans are going to have very strong ideas on how to get to the stars, and those ideas very well might include rogue planets.

  • Rob Flores October 10, 2013, 11:45

    So, can we tell if this object was created as a stand alone substellar object OR
    did it maybe acrete in the outer fringes (400+ AU) of a forming solar system and got ejected due to gravitational nteractions with other newly formed suns. Is there a way to figure it out?

    Astrobiological potential?
    Assuming this planet has a rapid spin like jupiter, it must have a
    gargantuan Magnetic field.
    If it has other moons of comparable size their interaction will induce tidal heating, which is one source of energy, if the heating is milder than Io maybe
    life can arise there.
    What if there is only One major moon on regular orbit (no heating) which passes through the strongest part of the magnetic field. Could a strong magnetic field interaction with such large moon be a ‘source’ of energy to enable the posibility of simple living orgnisms arising?

  • stevie October 10, 2013, 14:33


    The only version I can find on amazon is in French. I don`t read (speak) that.

  • Alex Tolley October 10, 2013, 15:51

    @Stevie – which Amazon are you looking at?

    The US (.com) has the English versions. Here is the Kindle version link

    I checked the content of the Kindle, hard and ppk versions, and the language is English.

  • Alex Tolley October 10, 2013, 16:03

    Assuming a sphere 5 ly in radius, 100k planetary objects implies that there is on average 1 every ~0.1 ly away. Even at 10K planets, that is 1 every ~0.25 ly away.

    Depending on the composition, they might make the star hopping via STL quite interesting as these bodies show offer a lot better resources than Oort cloud snowballs.

    Presumably they are moving relative to our sun, so what might be the frequency of near approaches to the sun, even entering the inner solar system?

  • Marshall Eubanks October 11, 2013, 0:17

    As it happens, this was the subject of one of my presentations to this year’s 100 Year Starship Symposium (in the Saturday “Destinations and Hidden Objects” session). I addressed the connection between recent work on nomadic planets (planets not orbiting any star) and starship destinations; my conclusion is that such planets are likely to be considerably closer to us than any star and thus will be the first destinations for interstellar travel. From that viewpoint, the first goal of interstellar travel is to find the nomadic planets close to us; fortunately they are (just) within the sensitivity of the ALMA array.

    The presentation is at

  • Michael October 11, 2013, 0:58

    Maybe these wanderers are the cause of comet storms

  • David Cummings October 11, 2013, 7:12

    Oort cloud snowballs are good. Water is good. Methane is good.

  • Marshall Eubanks October 11, 2013, 13:12

    A new type of habitat for life is the “Hydrogen planet” – a planet formed far from any Sun that kept its original complement of Hydrogen. If the hydrogen atmosphere is sufficiently thick, the Hydrogen IR bands get pressure broadened enough to keep in the heat from radionuclides, and there could be liquid oceans of water on the surface for billions of years. This raises the interesting question of whether there could be a hydrogen biosystem, where the biology breathes in hydrogen and eats oxidizing foods produced by bacteria at volcanic vents. See

    (“Life-sustaining planets in interstellar space?”) and
    (“Hydrogen Greenhouse Planets Beyond the Habitable Zone”) for more.

  • Alex Tolley October 11, 2013, 14:04

    Life needs CHON + P + S + many other elements in various amounts. Technology will need a host of elements, including metals. Do Oort snowballs have all the requirements?

  • Alex Tolley October 11, 2013, 14:36

    so what might be the frequency of near approaches to the sun

    Assuming stars in our region are 10 ly apart, a cube 10 ly on a side would contain 10^5 planets, or 100/ly^3.

    Using Marshall’s drift of 20 arc seconds for average velocity = ~ 10^-4 ly/yr.
    So we can assume that 1 planet will drift by within a ly every 100 years. Assuming the Oort is 0.5 ly in extent, a planet should drift through the Oort every 400 years. That seems rather too frequent to me.

    If we use just 1 planet/ly, we get a planet disrupting the Oort every 40k years. My BoE calculations may be off, but do these numbers fit at all with known lunar impacts, given expected comet frequencies with each disturbance?

  • Wojciech J October 11, 2013, 16:14

    With such abundance of worlds and resources it seems dubious that we would see the never-ending expansion of any civilization that some theorist assume invokes Fermi Paradox. Especially combined with the ability to go smaller on nano-scale or to digitize consciousness.
    Alex Tolley-indeed this is a good question and one I would love to know answer to? For example-how widespread are elements like uranium or other isotopes in asteorids and Oort objects?

  • Sean M. Brooks October 12, 2013, 13:13

    Very interesting, this article about “rogue” planets. It reminded me of the first novel I recall reading which uses the concept, SATAN’S WORLD, by Poul Anderson. A rogue planet also plays a role in another Anderson novel, ENSIGN FLANDRY.

    In SATAN’S WORLD, a rogue planet smaller than a gas giant swung near enough to another star that it’s heat warmed the rogue, enabling it to thaw and again have an atmosphere. And entrepreneurs in SATAN’S WORLD quickly realized such a planet would be very useful as a site for industrial processes too costly or dangerous to use on inhabited worlds. The radioactive waste heat given off by these industries would act to prevent Satan from freezing again as it moved back again into interstellar space.

    The next step, I assume, would be to look for smaller rogue planets.

    Sean M. Brooks

  • David Cummings October 12, 2013, 18:27

    Alex, I agree with you that it’s unlikely to find anything lifelike on an Oort snowball. I was thinking of them more in terms of convenient resourse opportunities for explorers from Earth.

  • Alex Tolley October 13, 2013, 12:09

    @David Cummings – my point was that the life on the world ship needs more elements than may be available on snowballs. Not to mention the non-living technology. I seriously doubt there is life in the Oort. But to use it as a resource beyond fusion fuel, it must offer all the elements lost to imperfect recycling in the worldship. A simple example, if there was no (or insufficient) magnesium, then the green plants relying on it for chlorophyll would be unable to grow. End of living worldship.

  • Marshall Eubanks October 13, 2013, 14:02

    I would assume that small bodies in deep space (Oort cloud or nomads) are most likely mostly volatiles (ice of some sort), as new comets seem to be mostly volatiles. Larger worlds (Mars sized and above) I would assume to be rocky planets, with substantial amounts of heavier elements (true metals and even radionuclides). Now, how easy it is to get at those resources is another matter – after all, we don’t typically regard all of the nickel-iron in the Earth’s core as an available resource. I suspect somewhere in between may be the “sweet spot” in terms of resource extraction.

  • coolstar October 13, 2013, 18:18

    I’m not aware of any serious person who takes Louis Strigari’s estimate of 10^5 rogue planets per bound planet seriously. As others have stated, the limits on micro-lensing experiments, impacts in the solar system, the lack of sufficient material in proto-stellar disks, and the fast observed turnover of numbers of objects per their mass as one gets in the brown dwarf range makes this extremely unlikely. I suspect there are pretty strong dynamical limits also. Lots of folks do solar system formation modelling and I’m not aware of any of them seeing ejections of any where close to this number (by several orders of magnitude).

  • Andrew Palfreyman October 14, 2013, 15:33

    Any serious considerations about using these bodies as way stations has to take cognizance of energy budgets. It seems that this particular discovery is warm and slowly cooling, a process that presumably occurs on a geological timescale, so will – as far as we’re concerned – remain warm for a good while. Energy could presumably be extracted for useful purposes via a heat engine operating between the surface or the interior, and cold vacuum. Terraforming might even be possible, powered by this method. I don’t know off-hand how injurious to human health is interstellar radiation, which might demand shielding. But it doesn’t seem impossible to imagine creating a viable habitat on such worlds.

  • Eniac October 14, 2013, 22:19


    I’m not aware of any serious person who takes Louis Strigari’s estimate of 10^5 rogue planets per bound planet seriously.

    I was also wondering why the first one we see is 80 ly away, when there are supposed to be 100,000 floating around within 5 ly distance. Something sure seems fishy about this 10^5 number…

  • Holger October 16, 2013, 15:17

    “I’m not aware of any serious person who takes Louis Strigari’s estimate of 10^5 rogue planets per bound planet seriously.”

    Strigari estimates “up to” 10^5 rogue planets per main-sequence STAR, not per bound planet, and he includes all objects larger than Pluto in the tally – the vast majority of the 10^5 objects would be smaller than the Moon, only ~700 would be bigger than Mars, and probably less than one would be Jupiter-sized. This doesn’t seem so unreasonable to me, given that we can only discover the very largest of them over interstellar distances. (There are probably several undiscovered “super-Plutos” even in solar orbit, according to Mike Brown: https://en.wikipedia.org/wiki/Planet_x#Probability )

  • Eniac October 16, 2013, 20:49

    Moon sized bodies in interstellar space would be completely frozen and dark. It would be impossible to detect any of them even if there were 10^9 per cubic light year. The best we can do is an upper limit by calculating how often we would see occultations of stars. Detection by occultation would allow us to tell they are there, and estimate how many, but leave us no chance to actually observe one for longer than that single blip. Nor could we locate them to use as a travel destination, either.