≡ Menu

Mapping Black Holes in (and out of) the Milky Way

Some years back, I reminisced in these pages about reading Poul Anderson’s World Without Stars, an intriguing tale first published in 1966 about a starship in intergalactic space that was studying a civilization for whom the word ‘isolation’ must have taken on utterly new meaning. Imagine a star system tens of thousands of light years away from the Milky Way, a place where an entire galaxy is but a rather dim feature in the night sky. Poul Anderson discussed this with Analog editor John Campbell:

One point came up which may interest you. Though the galaxy would be a huge object in the sky, covering some 20⁰ of arc, it would not be bright. In fact, I make its luminosity, as far as this planet is concerned, somewhere between 1% and 0.1% of the total sky-glow (stars, zodiacal light, and permanent aurora) on a clear moonless Earth night. Sure, there are a lot of stars there — but they’re an awfully long ways off!

For more on galactic brightness, see The Milky Way from a Distance. The Anderson tale was originally serialized as The Ancient Gods in the June and July, 1966 issues of Campbell’s magazine. Long-time readers will remember its cover, which I ran back in 2012, along with a discussion of how artist Chesley Bonestell approached the cover art, which shows the distant galaxy as far brighter than it would actually appear. Bonestell brightened it even knowing this to make the cover interesting while still suggesting just how far away the vast ‘city of stars’ actually was in the story.

Where Black Holes Roam

Intergalactic space is, I would assume, about as empty a place as could be. Yet new work out of the University of Sydney delves into just what we might find if we could see what’s out there. And it turns out that it is quite a lot. The university’s David Sweeney is lead author on a paper in Monthly Notices of the Royal Astronomical Society. The researchers discuss what they describe as the ‘galactic underworld,’ which is comprised of the compact remnants of massive stars. In other words, stars that have collapsed onto themselves and produced neutron stars and black holes.

Remember that black holes and neutron stars form from stars more than eight times the size of our Sun. If less than about 25 times the mass of the Sun, the star forms a neutron star, its tiny sphere jammed with neutrons prevented from collapsing further by neutron degeneracy pressure. Sweeney and team say that thirty percent of the black holes and neutron stars out there have been completely ejected from the galaxy. Given the age of the galaxy, over 13 billion years, a vast number of such objects must have formed, the 30 percent ejected by the ‘kick’ induced by their creation in a supernova.

Image: A colour rendition of the visible Milky Way galaxy (top) compared with the range of the galactic underworld (bottom). Credit: Sydney University.

As you can see in the image, the galaxy’s underworld turns out to stretch well beyond the visible limits of the disk. Peter Tuthill (Sydney Institute for Astronomy) notes the challenges involved in creating this first chart of an unseen population:

“One of the problems for finding these ancient objects is that, until now, we had no idea where to look. The oldest neutron stars and black holes were created when the galaxy was younger and shaped differently, and then subjected to complex changes spanning billions of years. It has been a major task to model all of this to find them. Newly-formed neutron stars and black holes conform to today’s galaxy, so astronomers know where to look. It was like trying to find the mythical elephant’s graveyard”

The researchers used a stellar population synthesis computer code called GALAXIA, modifying it to include stars that have exhausted their nuclear fusion life cycle, leaving behind a remnant black hole or neutron star, and excluding stars below 8 solar masses. Additional custom code was then produced to capture velocity changes to the star caused by supernovae explosions (the so-called ‘natal kick’). The effects of the kick were added to each remnant’s velocity and transformed to galactocentric coordinates, with subsequent custom code showing evolution of the stars’ paths over time.

The distribution map that emerged depicts a galaxy, and thus its remnants, changing over time, so that the Milky Way’s present shape does not predict the distribution of neutron stars and black holes surrounding it. In fact, the relatively thin and flattened disk structure gives way to triple the scale height of the Milky Way we see.

Image: Point-cloud chart of the visible Milky Way galaxy (top) versus the galactic underworld. Credit: Sydney University.

As the paper notes:

The spatial distribution of compact remnants is different from that of visible stars. The remnants are more dispersed in the vertical direction with the scale height being about 3 times larger than that of the visible stars. This is mainly due to the significant velocity kicks received by the remnants at the time of their birth.

Also interesting are these two points:

The spatial distribution of BHs is more centrally concentrated as compared to the NSs due to the smaller velocity kick they receive.

For some remnants the kick is so large that their total velocity becomes greater than their escape velocity (40% of NS and 2% of BHs). We are able to estimate a Galactic mass loss in ejected compact remnants as 2.1×108M or ∼0.4% of the stellar mass of the Galaxy.

If 30 percent of the stellar remnants over the course of the galaxy’s evolution have been ejected into intergalactic space, that leaves 70 percent that still moves through the visible disk, so that neutron stars and black holes from the earliest days of the galaxy still move unattached to any nearby star through stellar neighborhoods like our own.

Black Holes and Their Neighbors in Space

In addition to these ‘free floating’ black holes, there are those in gravitational dance with nearby stars, leaving traces that are detectable. Making that point is the recent discovery of a black hole about 12 times the mass of the Sun at roughly 1650 light years from the Solar System, one that appears to be orbited by a visible star. This is “closer to the Sun than any black hole X-ray binaries with known distances…or any of the black holes identified through other techniques.”

The work, led by Sukanya Chakrabarti (University of Alabama, Huntsville), likewise highlights the role these remnants can play in the disk we see today. Says Chakrabarti:

“In some cases, like for supermassive black holes at the centers of galaxies, [black holes] can drive galaxy formation and evolution. It is not yet clear how these non-interacting black holes affect galactic dynamics in the Milky Way. If they are numerous, they may well affect the formation of our galaxy and its internal dynamics.”

Note the term ‘non-interacting,’ which the author uses to distinguish this kind of black hole from those that show an accretion disk of dust accumulating from another object. As you might imagine, interacting black holes – or the features they produce – are easier to detect at visible wavelengths.

Finding the black hole in this work involved analyzing data on almost 200,000 binary stars, as accumulated from the European Space Agency’s Gaia mission. The intent was to find objects that seemed to have a dark companion of large mass, looking for the gravitational effects of a black hole on a visible star. The most interesting sources were followed up by the Automated Planet Finder in California, Chile’s Giant Magellan Telescope and the W.M. Keck Observatory in Hawaii. Spectroscopic measurements confirmed that the binary system contains a visible star cataloged as Gaia DR3 4373465352415301632 orbiting a dark, massive object.

Image: The cross-hairs mark the location of the newly discovered black hole. Credit: Sloan Digital Sky Survey / S. Chakrabarti et al.

As to how this system of star and black hole originally formed, this interesting speculation:

Given the combination of the large mass of the dark companion and a semi-major axis of Gaia DR3 4373465352415301632 that is neither very large nor very small, the formation channel for this system is not immediately clear. However, the most natural scenario may be that the visible G star was originally the outer tertiary component orbiting a close inner binary with two massive stars.

So here we have a search for black holes bound to visible stars, with the authors estimating that perhaps a million such stars have black hole companions. That’s an early estimate for one population of black holes, but this object, in a 185-day orbit from the star, does not represent the class of black holes and neutron stars that may move through the galaxy untethered to any visible object, as found in the investigations of the Sydney team. Just how many black holes may be peppered through the several hundred billion stars of the Milky Way, and how widely spaced are they likely to be?

Finding untethered black holes, whether within or outside the galactic disk, is not work for the faint-hearted. Surely microlensing studies are our best way to proceed?

The paper is Sweeney et al., “The Galactic underworld: the spatial distribution of compact remnants,” Monthly Notices of the Royal Astronomical Society, Volume 516, Issue 4 (November 2022), pp. 4971–4979 (abstract / preprint). The black hole discovery paper is Chakrabarti et al., “A non-interacting Galactic black hole candidate in a binary system with a main-sequence star,” in process at the Astrophysical Journal (preprint).

{ 16 comments… add one }
  • Al Jackson October 21, 2022, 9:34

    Interesting, it had never registered on me the number of Analog Poul Anderson ‘astronomical’ themed stories that made ASFS covers. All illustrated by Chesley Bonestell. Could have been an Anderson suggestion to Campbell? There was also Supernova Jan. 1967, Starfog Aug. 1967 , Satan’s World May 1968…. any others? All Bonestell. (Tau Zero, To Outlive Eternity appeared in Galaxy June 1967 with no cover illustration.) (Starfog was sort of a flip side to Ancient Gods.)
    Anderson was a real grand master of SF story telling and even tho others Heinlein , Asimov and Blish had injected ‘hard science’ astronomy into their stories (they weren’t alone) , Anderson with his degree in physics was a super adept at mixing poetry with hard science.

    Not a lot of Bonestell ASF covers , H L Gold used a number on Galaxy but many times they did not illustrate a story.

  • Henry Cordova October 21, 2022, 10:02

    Could these be the missing MACHOS responsible for the mysterious ‘Dark Matter”?

    An extended cloud of MAssive Compact Halo ObjectS could be the invisible gravitational component that provides the cause of the unexplained non-Keplerian rotation observed in galactic disks.

    In the early galaxy, the formation of quickly evolving, hypermassive stars prone to supernova collapse was much more common than it is now, so a large number of dark objects would wind up populating the outer regions of the galaxy. Except for the occasional, brief gravitational lensing of faint distant galaxies, we would have no way of detecting their presence.

    • Ron S. October 21, 2022, 14:43

      I have no references immediately at hand, but it has been pretty well established in observational microlensing studies that there are far too few MACHOs to come anywhere near to equalling the mass requirement. These studies encompass brown dwarfs, Jovian class rogue planets, small & large BH, etc.

      • henry October 21, 2022, 21:46

        I’ve heard the same same thing, that the usual candidates for MACHOs, football-sized rocks, rogue planets, red dwarfs, etc are nowhere near plentiful enough, The MACHO hypothesis has been ruled out (as has the WIMP theory). But this new research suggests a very large number of very massive black holes would provide the mass, yet be few enough in number they would not leave any other evidence.

        We do know that the early universe experienced galaxy-wide bursts of star formation with many massive stars which rapidly evolved to neutron stars and black holes, now invisible. For some reason, the early universe seemed to favor this…

        Whether this is enough to simulate the effects of Dark Matter, I don’t know. But it is an interesting conjecture, and should be followed up. I’m familiar with the work you allude to, and it does sound convincing, but BH MACHOs provide an alternative to Dark Matter. I’ve always been suspicious of DM. It sounds too much like an ad hoc explanation for an inconvenient fact.

        Still, if this alleged Dark Matter substitute is indeed the bulk of matter in the universe, that would imply an extravagant number of hypermassive stars were created in the early universe to explain the anomalous observed rotation rates of galaxies, and the gravitational cohesion of clusters of galaxies. Why this should be so is not clear.

        One way to test this hypothesis is that these collapsed stars would be primarily located near their galaxies of origin, but would be relatively rare in the voids between galaxies.

      • Michael October 22, 2022, 3:14

        Although only a small percentage of compact objects form this huge cloud it could drag a lot of material with it. Also when these stars go supernova to form the compact object they eject much more material than forms in the object by a huge amount.

    • Mike Serfas October 23, 2022, 18:59

      The resemblance to a dark matter halo is striking, but – correct me if I’m wrong – I suppose it’s just coincidence? Like these objects, dark matter moves slowly enough that it doesn’t always escape the galaxy, but rapidly enough not to be bound to individual stars. I’m curious if someone could describe each of the proposals for dark matter in terms of the “natal kick” each would be expected to have. Wouldn’t it be neat if somebody could measure something for dark matter like that little sideways spike in the side-on distribution shown here, using gravitational lensing?

  • Antonio October 21, 2022, 12:43

    I wonder whether any planets would survive the violent creation of this pair. It would be quite an interesting place to live in.

  • Alex Tolley October 21, 2022, 12:55

    If our sun was part of a binary system with the companion a BH, how far away would the BH need to be to make it near impossible to detect?

    If such a BH did exist, would it not suck in rogue planets and other ejected [icy] bodies, producing a flash of energy from a short-lived accretion disk? If we knew where it was, would monitoring it provide information on the number of these free-floating bodies?

    • Ron S. October 21, 2022, 14:38

      “If our sun was part of a binary system with the companion a BH, how far away would the BH need to be to make it near impossible to detect?”

      There are several ways to detect a BH. One is by the accretion disk. Even with episodic in-falling matter it would be detectable at extreme distances. If it were less than, say, 1 ly distant, the noise from ISM material falling in would likely be detectable.

      Two is by orbital effects on the outer bodies, including KBO and Oort cloud. That appears to be discounted by observations, putting it at least 1 ly distant. At that distance it is far more likely to be a passing interloper than gravitational bound to the Sun.

      Three is by gravitational lensing. A solar mass BH within quite a large radius could not possibly escape detection because of the continuous observation of the sky by countless observatories. The apparent lensing radius would be large. So, no nearby BH of solar mass. Smaller BH are unobserved and, at least so far, have no credible path to formation.

      “would monitoring it provide information on the number of these free-floating bodies?”

      Probably about as much as observing Earth would tell us about the prevalence of life in the universe. Extrapolating from a data set of 1 is unreliable.

      • Michael October 22, 2022, 15:48

        Giai may be able to see the slight movement in stars as a BH or NS goes about its business.

  • Martin October 21, 2022, 18:16

    Minor correction:
    Where it says “If less than about 25 times the mass of the Sun, the star forms a neutron star” I believe you left out the decimal point. It should be 2.5.

    • Jed November 28, 2022, 1:14

      25 solar masses is correct. You’re thinking of the maximum mass of a neutron star, which is about 2.2 solar masses. The maximum mass of a progenitor star that will leave a neutron star remnant after supernova is 25 solar masses. Above that they tend to leave a black hole, unless metallicity is very high:


  • Martin October 21, 2022, 19:04

    Oops. You were talking about the progenitor star. My mistake.

  • ljk October 26, 2022, 15:32

    It is funny and sad that the media keeps calling these black hole and neutron star locations “graveyards” because they cannot imagine any native life forms being able to exist around them.

    Well, any natives they once had might be gone, or they and others realized the awesome energies and other benefits of building a society around a black hole…


    Maybe that’s where the advanced ETI are.

  • wdk October 31, 2022, 15:13

    Observation of dark matter based on galactic rotational rates out from their centers – it’s not uniform among galaxies. But considering the current controversy, whether DM can be attributed to black holes and neutron stars ejected into the depths, let’s consider spiral and elliptical galaxies. It is assumed that spirals continue star production and ellipticals have slowed down. Both have some type of variation from a visible mass profile for rotation. But are there characteristic profiles for each type?

    Just an idle thought: galaxies with active centers (Seyfert, quasars, etc.) have a tendency to emit highly charged material along magnetic polar axes. Can be observed to extend thousands or tens of thousands of parsecs in profile. But where does it go or settle out? Does it condense out all as baryons?

  • ljk November 14, 2022, 14:41

    Did a passing black hole help to dim the red giant star Betelgeuse recently (well, relatively recently from our perspective)…


    A very interesting related paper here:


    [Submitted on 27 May 2022]

    The Great Dimming of Betelgeuse seen by the Himawari-8 meteorological satellite

    Daisuke Taniguchi, Kazuya Yamazaki, Shinsuke Uno

    Betelgeuse, one of the most studied red supergiant stars, dimmed in the optical by ~1.2 mag between late 2019 and early 2020, reaching an historical minimum called “the Great Dimming.” Thanks to enormous observational effort to date, two hypotheses remain that can explain the Dimming: a decrease in the effective temperature and an enhancement of the extinction caused by newly produced circumstellar dust.

    However, the lack of multi-wavelength monitoring observations, especially in the mid infrared where emission from circumstellar dust can be detected, has prevented us from closely examining these hypotheses.

    Here we present 4.5-year, 16-band photometry of Betelgeuse between 2017-2021 in the 0.45-13.5 micron wavelength range making use of images taken by the Himawari-8 geostationary meteorological satellite.

    By examining the optical and near-infrared light curves, we show that both a decreased effective temperature and increased dust extinction may have contributed by almost the same amount to the Great Dimming.

    Moreover, using the mid-infrared light curves, we find that the enhanced circumstellar extinction actually contributed to the Dimming.

    Thus, the Dimming event of Betelgeuse provides us an opportunity to examine the mechanism responsible for the mass loss of red supergiants, which affects the fate of massive stars as supernovae.

Leave a Comment