Latest / SLEEPY HISTORY / The Impossible Size of the Smallest Galaxy We_ve Ever Found
Transcript
- 0:00Speaker 1: Somewhere in the constellation Ursa Major, roughly 30,000 light years
- 0:05Speaker 1: from where you are now, there is a gathering of
- 0:07Speaker 1: ancient stars so small and so faint that astronomers cannot
- 0:11Speaker 1: agree on what it is. It contains approximately 60 stars.
- 0:15Speaker 1: Its total mass is roughly 16 times the mass of
- 0:18Speaker 1: our Sun. Its light output is only 11 times the
- 0:22Speaker 1: brightness of a single star like ours. And yet, this
- 0:26Speaker 1: whisper of matter might be a galaxy. not a miniature
- 0:30Speaker 1: version of the Milky Way, something stranger, something that challenges
- 0:34Speaker 1: the very definition of what a galaxy is. What you
- 0:37Speaker 1: are about to hear is the story of the smallest,
- 0:40Speaker 1: faintest object ever found that might still deserve to be
- 0:43Speaker 1: called a galaxy, and why that word might be breaking.
- 0:47Speaker 1: If you enjoy deep journeys into the strangest corners of
- 0:50Speaker 1: the universe, I would really appreciate it if you could
- 0:53Speaker 1: hit that like button and subscribe. Now, settle in, get
- 0:57Speaker 1: yourself comfortable, Let's begin. When you hear the word galaxy,
- 1:08Speaker 1: a specific image forms in your mind. You see a
- 1:11Speaker 1: vast spiral of light, hundreds of billions of stars turning
- 1:15Speaker 1: slowly through space, arms trailing luminous dust, and newborn stellar
- 1:19Speaker 1: furnaces across distances so immense that light itself needs a
- 1:23Speaker 1: hundred thousand years to cross from one edge to the other.
- 1:26Speaker 1: You picture the Milky Way, Our home, with its roughly
- 1:30Speaker 1: 400 billion stars burning hydrogen into helium, in a disk
- 1:34Speaker 1: so wide that every civilization in human history could rise
- 1:38Speaker 1: and fall a thousand times before a single photon completes
- 1:41Speaker 1: the crossing. Or you picture Andromeda, our nearest large neighbor,
- 1:46Speaker 1: a trillion stars strong, its faint glow visible to the
- 1:50Speaker 1: naked eye, from two and a half million light years away.
- 1:54Speaker 1: These images are correct. They are also incomplete in a
- 1:57Speaker 1: way that matters enormously for what comes next. They are
- 2:00Speaker 1: incomplete because they establish an expectation. Galaxy means big. Galaxy
- 2:05Speaker 1: means luminous. Galaxy means an island universe. A self-contained metropolis
- 2:11Speaker 1: of stars massive enough to shape the curvature of space
- 2:14Speaker 1: and time around it. The ancient Greeks called the Milky
- 2:17Speaker 1: Way Galaxias, from their word for milk, because the band
- 2:21Speaker 1: of light stretching across the sky resembled spilled milk against
- 2:24Speaker 1: the darkness. When Edwin Hubble proved in the 1920s that
- 2:28Speaker 1: the so-called spiral nebulae were actually distant collections of stars
- 2:32Speaker 1: lying far beyond the Milky Way, the word galaxy expanded
- 2:36Speaker 1: to describe all of them. From that moment, galaxy became
- 2:40Speaker 1: synonymous with immensity. But the universe does not care about
- 2:43Speaker 1: human vocabulary. It does not adjust its structures to fit
- 2:47Speaker 1: neatly inside the categories we invent. And over the past
- 2:50Speaker 1: two decades, astronomers have discovered objects so small, so faint,
- 2:55Speaker 1: and so ghostly that the word galaxy begins to crack
- 2:58Speaker 1: under the pressure of what it is being asked to describe.
- 3:00Speaker 1: The trouble started in 2005. That year, astronomers mining data
- 3:06Speaker 1: from the Sloan Digital Sky Survey, an ambitious project that
- 3:09Speaker 1: mapped millions of stars across a large fraction of the sky,
- 3:13Speaker 1: found something strange. Embedded in the data were faint smudges
- 3:17Speaker 1: of stars hovering around the Milky Way. These were not
- 3:20Speaker 1: bright satellite galaxies like the Large Magellanic Cloud, Visible to
- 3:24Speaker 1: the naked eye from the southern hemisphere and home to
- 3:27Speaker 1: 30 billion stars, they were something far more subtle. Tiny
- 3:31Speaker 1: over densities of stars, barely detectable even with sophisticated matched
- 3:36Speaker 1: filter algorithms, designed to pick out faint stellar groupings against
- 3:40Speaker 1: the crowded foreground of the Milky Way's own disk. Without
- 3:44Speaker 1: the algorithms, a human eye scrolling through sky images would
- 3:48Speaker 1: never have noticed them. They were hiding in the noise.
- 3:51Speaker 1: Before these new discoveries, The known population of Milky Way
- 3:55Speaker 1: satellite galaxies consisted of classical dwarf spheroidals like Ursa Minor, Draco,
- 4:01Speaker 1: and Carina systems that had been catalogued for decades, each
- 4:05Speaker 1: containing a few million stars. But the objects emerging from
- 4:08Speaker 1: the Sloan survey data were something different entirely. They were
- 4:12Speaker 1: fainter by factors of a hundred or a thousand. Some
- 4:16Speaker 1: contained only a few thousand stars. Others appeared to hold
- 4:19Speaker 1: only a few hundred. Their total luminosities were less than
- 4:22Speaker 1: the output of a single bright red giant star. If
- 4:26Speaker 1: you placed one of these objects at the distance of Andromeda,
- 4:29Speaker 1: no telescope on Earth could detect it. Astronomers called them
- 4:33Speaker 1: ultra-faint dwarf galaxies, and they represented a class of object
- 4:38Speaker 1: that no one had predicted would exist in such extreme form.
- 4:42Speaker 1: The first question was immediate and obvious. Were these objects
- 4:46Speaker 1: actually galaxies at all? A few thousand stars clumped together
- 4:49Speaker 1: in space could easily be a star cluster. Globular clusters,
- 4:53Speaker 1: the dense, ancient balls of stars that orbit the Milky
- 4:56Speaker 1: Way by the hundreds, typically contain between 100,000 and a
- 5:00Speaker 1: million stars. But the smallest known globular clusters contain only
- 5:04Speaker 1: a few thousand. At first glance, these ultra-faint detections looked
- 5:09Speaker 1: as though they could be stripped-down remnants of old star
- 5:11Speaker 1: clusters rather than independent galaxies. They were the right size
- 5:15Speaker 1: for clusters. They were the right luminosity for clusters. They looked,
- 5:20Speaker 1: by every surface-level measure, like clusters. The answer came from
- 5:24Speaker 1: a measurement that would become the single most important diagnostic
- 5:27Speaker 1: in every classification dispute that followed. Velocity dispersion. Velocity dispersion
- 5:33Speaker 1: is the spread of speeds among the stars within a system.
- 5:36Speaker 1: How fast the individual members are moving relative to each other,
- 5:40Speaker 1: not how fast the system as a whole travels through space.
- 5:43Speaker 1: but how vigorously the internal members are jostling around within
- 5:46Speaker 1: their shared gravitational well. In a star cluster, the stars
- 5:51Speaker 1: are held together by their own collective gravity. The total
- 5:54Speaker 1: gravitational pull comes from the combined mass of all the
- 5:57Speaker 1: visible stars. Measure how fast those stars move, apply Newton's laws,
- 6:02Speaker 1: and you can calculate the total mass required to keep
- 6:05Speaker 1: them bound. If the calculated mass matches the visible stellar mass,
- 6:09Speaker 1: the system is self-gravitating. It holds itself together through the
- 6:13Speaker 1: gravity of the stars you can see. No mystery required.
- 6:18Speaker 1: No invisible scaffolding needed. But in the ultra-faint dwarfs, the
- 6:21Speaker 1: speeds did not match. Not even close. Stars inside these
- 6:26Speaker 1: tiny systems were moving far too fast for their visible mass.
- 6:30Speaker 1: The velocities implied total gravitational masses hundreds or even thousands
- 6:34Speaker 1: of times greater than what the visible stars could account for.
- 6:37Speaker 1: A system with a few thousand solar masses of visible
- 6:40Speaker 1: stars was moving as though it sat inside a gravitational
- 6:43Speaker 1: well belonging to something containing millions of solar masses. Something
- 6:47Speaker 1: invisible and enormously heavy was gripping those stars, preventing them
- 6:52Speaker 1: from escaping into the void. That something was dark matter.
- 6:55Speaker 1: Dark matter makes up roughly 85% of all matter in
- 6:59Speaker 1: the universe. It does not emit light. It does not
- 7:02Speaker 1: absorb light. It does not interact with electromagnetic radiation in
- 7:06Speaker 1: any way that any instrument has ever been able to detect.
- 7:09Speaker 1: It reveals itself only through gravity, through its pull on
- 7:12Speaker 1: visible matter and its bending of light passing nearby. Every
- 7:16Speaker 1: large galaxy in the universe sits inside a vast halo
- 7:19Speaker 1: of dark matter, extending far beyond the visible stars. In
- 7:24Speaker 1: the Milky Way, the dark matter halo is estimated to
- 7:27Speaker 1: contain roughly ten times more mass than all the stars,
- 7:31Speaker 1: gas and dust combined. It extends outward for hundreds of
- 7:35Speaker 1: thousands of light years beyond the edge of the visible disk.
- 7:38Speaker 1: Dark matter provides the gravitational skeleton on which the visible
- 7:42Speaker 1: galaxy is built. Without it, the Milky Way's outer stars
- 7:45Speaker 1: would fly off into intergalactic space. Their orbital speeds are
- 7:49Speaker 1: too high for the visible mass to hold them. Dark
- 7:52Speaker 1: matter keeps the architecture intact. In the ultra-faint dwarfs, dark
- 7:56Speaker 1: matter does not merely contribute to the gravitational budget, it
- 8:00Speaker 1: dominates it almost entirely. These systems are not galaxies in
- 8:04Speaker 1: the way most people imagine. They are not grand spirals.
- 8:08Speaker 1: They are not luminous ellipticals. They are tiny clusters of
- 8:11Speaker 1: ancient stars embedded inside dark matter halos, so massive relative
- 8:16Speaker 1: to their visible content that the stars themselves become almost
- 8:20Speaker 1: an afterthought. The visible matter is a rounding error. The
- 8:23Speaker 1: dark matter is the structure. The galaxy, in a meaningful
- 8:27Speaker 1: gravitational sense, is the halo. The stars are just traces,
- 8:32Speaker 1: luminous breadcrumbs marking the centre of something overwhelmingly invisible. This
- 8:36Speaker 1: realisation forced astronomers to confront a question that had never
- 8:40Speaker 1: seemed urgent before. What, exactly, is a galaxy? The answer
- 8:45Speaker 1: had always felt intuitive. A galaxy is a large, self-contained
- 8:50Speaker 1: collection of stars. But intuition collapses when you are staring
- 8:53Speaker 1: at a system containing fewer stars than a modest open
- 8:57Speaker 1: cluster in the Milky Way's disk, yet embedded inside a
- 9:01Speaker 1: dark matter halo that outweighs all its visible stars by
- 9:05Speaker 1: a factor of thousands. In 2012, astronomers Beth Wilman and
- 9:10Speaker 1: Jay Strader proposed a definition crafted to handle exactly this problem.
- 9:14Speaker 1: They wrote that a galaxy is a gravitationally bound collection
- 9:17Speaker 1: of stars whose properties cannot be explained by a combination
- 9:21Speaker 1: of baryons and Newton's laws of gravity. In plainer terms,
- 9:25Speaker 1: if you measure how the stars inside a system are moving,
- 9:28Speaker 1: and the motions require more mass than the visible stars
- 9:31Speaker 1: can provide, the system contains something beyond ordinary matter. It
- 9:35Speaker 1: is a galaxy. If the motions can be fully explained
- 9:39Speaker 1: by the stars alone, with no invisible excess, the system
- 9:43Speaker 1: is a star cluster. Within the standard cosmological framework lambda
- 9:47Speaker 1: cold dark matter, this definition reduces to a single diagnostic question.
- 9:52Speaker 1: Does the object sit inside its own dark matter halo?
- 9:55Speaker 1: If yes, it is a galaxy. If no, it is
- 9:59Speaker 1: a star cluster. Dark matter becomes the dividing line. Not
- 10:03Speaker 1: the number of stars, not the physical size, not the luminosity,
- 10:07Speaker 1: not the shape or the colour or the distance from
- 10:10Speaker 1: the Milky Way. The presence or absence of a dark
- 10:13Speaker 1: matter halo is what separates the smallest galaxy from the
- 10:17Speaker 1: largest star cluster. Two objects could contain identical numbers of stars,
- 10:22Speaker 1: occupy identical volumes of space, and look indistinguishable in any photograph.
- 10:27Speaker 1: But if one sits inside a dark matter halo and
- 10:29Speaker 1: the other does not, one is a galaxy and the
- 10:32Speaker 1: other is not. This definition is elegant. It is also
- 10:36Speaker 1: extraordinarily difficult to apply in practice. Measuring velocity dispersion requires
- 10:41Speaker 1: obtaining spectra of individual stars within the system. and determining
- 10:45Speaker 1: their radial velocities how fast each star moves toward or
- 10:49Speaker 1: away from us, to a precision of better than one
- 10:52Speaker 1: or two kilometers per second. For ultra-faint dwarfs containing only
- 10:56Speaker 1: a few dozen stars bright enough for spectroscopy, the entire
- 11:00Speaker 1: measurement depends on a small handful of data points. Binary stars,
- 11:04Speaker 1: pairs orbiting each other, can contaminate individual velocity measurements and
- 11:09Speaker 1: inflate the calculated dispersion of the entire system. Remove one
- 11:13Speaker 1: such contaminant from a sample of only a dozen stars
- 11:16Speaker 1: and the calculated total mass can drop by a factor
- 11:19Speaker 1: of 10. The difference between galaxy and star cluster can
- 11:23Speaker 1: hinge on whether a single anomalous velocity reflects a dark
- 11:26Speaker 1: matter halo or an undetected binary orbit that has nothing
- 11:30Speaker 1: to do with the system's gravitational potential. There is a
- 11:33Speaker 1: second diagnostic tool, less direct but sometimes equally revealing. Metallicity spread.
- 11:40Speaker 1: Metallicity is in astronomical usage, refers to the abundance of
- 11:44Speaker 1: elements heavier than helium in a star's atmosphere. The very
- 11:48Speaker 1: first stars in the universe formed from nearly pure hydrogen
- 11:51Speaker 1: and helium left over from the Big Bang. No heavier
- 11:54Speaker 1: elements existed yet. When those massive early stars died in
- 11:58Speaker 1: supernova explosions, they forged and scattered heavy elements, carbon, oxygen, silicon, iron,
- 12:06Speaker 1: and dozens of others into the surrounding gas. The next
- 12:09Speaker 1: generation of stars formed from that enriched gas, incorporating heavier
- 12:14Speaker 1: elements into their composition. Each successive generation became progressively richer
- 12:19Speaker 1: in metals. This process is called chemical enrichment, and it
- 12:23Speaker 1: operates as a one-way ratchet. Galaxies grow more metal-rich over
- 12:27Speaker 1: time as stellar generations build upon the chemical legacy of
- 12:30Speaker 1: their predecessors. A star cluster forms in a single burst.
- 12:34Speaker 1: All its stars are born from the same cloud of
- 12:36Speaker 1: gas at roughly the same time. and they share approximately
- 12:40Speaker 1: the same chemical composition. The metallicity spread within a typical
- 12:44Speaker 1: globular cluster is extremely small, often less than 0.1 in
- 12:49Speaker 1: the iron to hydrogen ratio. All the stars were made
- 12:52Speaker 1: from the same raw material, in the same place, at
- 12:55Speaker 1: the same moment. There was no second generation, no enrichment
- 12:59Speaker 1: from internal supernovae that stayed within the system. The debris
- 13:03Speaker 1: from dying stars simply escaped the cluster's shallow gravitational well
- 13:07Speaker 1: and dispersed into the surrounding galaxy. A galaxy, by contrast,
- 13:11Speaker 1: retains the debris from its own supernovae. Its gravitational potential
- 13:16Speaker 1: well is deep enough, thanks to dark matter, to hold
- 13:19Speaker 1: onto the gas enriched by exploding stars. New stars form
- 13:23Speaker 1: from that enriched gas, and their compositions differ measurably from
- 13:27Speaker 1: the first generation. Over time, the galaxy accumulates a spread
- 13:32Speaker 1: of metallicities across its stellar population. The oldest stars are
- 13:35Speaker 1: extremely metal-poor, Later stars are progressively more metal-rich. This range
- 13:41Speaker 1: of chemical compositions acts as a fingerprint, revealing that the
- 13:44Speaker 1: system captured its own supernova products and formed stars across
- 13:48Speaker 1: multiple episodes. The metallicity spread is indirect evidence for a
- 13:52Speaker 1: deep potential well, and therefore indirect evidence for dark matter.
- 13:57Speaker 1: So astronomers carry two primary tools for distinguishing a galaxy
- 14:00Speaker 1: from a star cluster at the faintest extremes– Velocity dispersion
- 14:04Speaker 1: reveals the total gravitational mass and whether dark matter is present.
- 14:09Speaker 1: Metallicity spread reveals whether the system retained enriched gas across
- 14:13Speaker 1: multiple generations of star formation, and therefore, whether its potential
- 14:17Speaker 1: well was deep enough to hold on to supernova ejecta.
- 14:20Speaker 1: When both measurements clearly indicate dark matter and chemical self-enrichment,
- 14:25Speaker 1: the classification is secure. The object is a galaxy. When
- 14:29Speaker 1: both measurements clearly show no excess mass and no chemical spread,
- 14:32Speaker 1: the object is a star cluster. The trouble arises when
- 14:35Speaker 1: the measurements are ambiguous, when the velocity dispersion is marginally
- 14:39Speaker 1: resolved but could be inflated by binaries, when the metallicity
- 14:43Speaker 1: spread is small but not quite zero, or when too
- 14:46Speaker 1: few stars have been measured to draw a confident conclusion.
- 14:49Speaker 1: When the system is so faint that only a handful
- 14:52Speaker 1: of stars can be observed spectroscopically, and every single data
- 14:56Speaker 1: point carries enormous uncertainty. At these extremes, The clean boundary
- 15:01Speaker 1: between galaxy and star cluster begins to dissolve into a
- 15:04Speaker 1: grey zone that unsettles everyone who works in this field.
- 15:07Speaker 1: Before Ursa Major 3 Unions 1 entered the picture, the
- 15:11Speaker 1: most extreme object pushing against this boundary was Segway 1. Segway.
- 15:17Speaker 1: One was discovered in 2007 in the Sloan Digital Sky
- 15:21Speaker 1: Survey data, lurking in the constellation Leo, roughly 75,000 light-years
- 15:26Speaker 1: from the Sun. It was initially considered a possible globular
- 15:30Speaker 1: cluster because of its small half-light radius, roughly 30 parsecs,
- 15:35Speaker 1: or about 100 light-years. Its total luminosity was staggeringly low.
- 15:40Speaker 1: With an absolute magnitude of roughly negative 1.5, it was
- 15:44Speaker 1: tens of millions of times less luminous than the Milky Way.
- 15:47Speaker 1: If you gathered all the light from every star in
- 15:49Speaker 1: Segway 1 and focused it into a single beam, it
- 15:53Speaker 1: would barely outshine a few hundred suns. but spectroscopy told
- 15:57Speaker 1: a very different story. Mala Gihar at Yale and Josh
- 16:01Speaker 1: Simon at the California Institute of Technology measured the radial
- 16:05Speaker 1: velocities of 24 member stars using the Keck 2 telescope
- 16:09Speaker 1: and found a velocity dispersion of approximately 4.3 km per second.
- 16:15Speaker 1: For a system containing roughly 1,000 solar masses of visible stars,
- 16:19Speaker 1: this was drastically too high. The implied mass-to-light ratio ranged
- 16:24Speaker 1: between approximately 1,300 and 3,400, meaning the system was overwhelmingly
- 16:30Speaker 1: dominated by dark matter, Segway. One appeared to contain nearly
- 16:35Speaker 1: a thousand times more invisible mass than visible mass. Its
- 16:39Speaker 1: stars were mere sparks drifting through the centre of an
- 16:42Speaker 1: enormous dark matter structure they could never reveal on their own.
- 16:45Speaker 1: Chemical measurements reinforced the galaxy classifications. Segway 1 stars displayed
- 16:51Speaker 1: a wide range of metallicities, with iron to hydrogen ratios
- 16:55Speaker 1: spanning more than two orders of magnitude. Some stars ranked
- 16:59Speaker 1: among the most metal-poor ever found anywhere in the universe,
- 17:02Speaker 1: with compositions barely enriched beyond the primordial hydrogen and helium
- 17:06Speaker 1: of the Big Bang. Others carried significantly more heavy elements.
- 17:11Speaker 1: This spread implied that Segway 1 had retained supernova ejecta
- 17:14Speaker 1: across multiple episodes of star formation, behavior characteristic of a
- 17:18Speaker 1: system with a deep dark matter potential well. A star cluster,
- 17:23Speaker 1: with its shallow gravity, could not have held onto that
- 17:26Speaker 1: enriched gas. It would have blown away with the first supernova.
- 17:30Speaker 1: On the strength of both kinematic and chemical evidence, Segway
- 17:34Speaker 1: 1 was classified as the least luminous galaxy known. It
- 17:38Speaker 1: became a cornerstone of dark matter research on small scales.
- 17:42Speaker 1: Its extreme mass-to-light ratio made it a prime target for
- 17:45Speaker 1: indirect dark matter detection experiments, which used gamma-ray telescopes to
- 17:50Speaker 1: search for signals from dark matter particle annihilation at the
- 17:53Speaker 1: centre of its dense halo. No signal was found, but
- 17:56Speaker 1: the constraints were valuable. Segway 1 was pushing the limits
- 18:01Speaker 1: of what astrophysics could learn from a handful of ancient stars,
- 18:04Speaker 1: but even Segway 1 prompted debate. Some researchers questioned whether
- 18:09Speaker 1: tidal interactions with the Milky Way had distorted its velocity structure.
- 18:13Speaker 1: Others raised concerns about foreground contamination. Milky Way stars at
- 18:17Speaker 1: similar velocities and similar positions on the sky could masquerade
- 18:22Speaker 1: as Segway-1 members, skewing the dispersion measurement. The classification held,
- 18:27Speaker 1: but not without persistent dissent. Segway. One taught the community
- 18:32Speaker 1: a difficult lesson. At these faint extremes, every measurement is fragile,
- 18:37Speaker 1: every data point is precious, and every classification is provisional
- 18:41Speaker 1: until confirmed by independent methods. Other ultra-faint dwarfs followed in
- 18:45Speaker 1: the years after. Boots 1, discovered in 2006, sits roughly 200,000
- 18:51Speaker 1: light-years away, with a total luminosity of roughly 30,000 suns.
- 18:56Speaker 1: Though far brighter than the most extreme ultra-faint dwarfs, its velocity, dispersion,
- 19:01Speaker 1: and metallicity spread both support a galaxy classification and its
- 19:05Speaker 1: elongated shape hints at ongoing tidal disruption. Wilman 1, also
- 19:10Speaker 1: found through the Sloan survey, remained stubbornly ambiguous for years.
- 19:14Speaker 1: Its velocity measurements were plagued by foreground contamination so severe
- 19:19Speaker 1: that different analyses studying the same data reached opposite conclusions
- 19:24Speaker 1: about whether it contained dark matter at all. The uncertainty
- 19:27Speaker 1: surrounding Wilman 1 was not an anomaly, but a warning sign.
- 19:31Speaker 1: At these faint luminosities, every diagnostic tool that works cleanly
- 19:35Speaker 1: for brighter systems begins to degrade, and the confident separation
- 19:39Speaker 1: of galaxy from star cluster starts to slip. Each ultra-faint
- 19:44Speaker 1: dwarf galaxy discovered forced astronomers to ask the same uncomfortable question.
- 19:49Speaker 1: How small can a galaxy get? Is there a floor,
- 19:54Speaker 1: a minimum number of stars, a minimum luminosity, a minimum
- 19:58Speaker 1: dark matter halo mass below which the concept of a
- 20:01Speaker 1: galaxy simply stops making physical sense? Theoretical models within the
- 20:05Speaker 1: Lambda Cold Dark Matter Framework predict that such a flaw
- 20:09Speaker 1: should exist. Below a certain halo mass, roughly between 100
- 20:13Speaker 1: million and 1 billion solar masses depending on the assumptions,
- 20:17Speaker 1: the gravitational potential well becomes too shallow to accumulate gas,
- 20:21Speaker 1: cool it, and compress it into stars. Conditions in the
- 20:25Speaker 1: early universe, particularly the intense radiation that accompanied the epoch
- 20:28Speaker 1: of reionization, permanently cut off the gas supply to the
- 20:32Speaker 1: smallest halos ensuring that only those halos massive enough to
- 20:36Speaker 1: have collected gas beforehand could ever form stars. Below this
- 20:40Speaker 1: critical threshold, dark matter halos exist in abundance, predicted by
- 20:44Speaker 1: the millions around every Milky Way-sized galaxy, but they remain
- 20:49Speaker 1: permanently dark. No stars, no light, no visible trace at all,
- 20:55Speaker 1: just invisible clumps of dark matter drifting silently through the halo.
- 20:59Speaker 1: How many of these dark halos exist remains one of
- 21:02Speaker 1: the most contested questions in modern astrophysics, and answering it
- 21:06Speaker 1: requires finding the luminous ones, the halos that did manage
- 21:10Speaker 1: to form stars, and working backward from their abundance to
- 21:14Speaker 1: infer the total population. This means that the faintest galaxies
- 21:18Speaker 1: we observe today are not simply the runts of the
- 21:21Speaker 1: galactic litter, they are fossils from the earliest epoch of
- 21:25Speaker 1: galaxy formation. Relics of the first dark matter structures to
- 21:28Speaker 1: cross the mass threshold and form stars, likely within the
- 21:32Speaker 1: first billion years after the Big Bang. Their stars are
- 21:35Speaker 1: among the oldest in the universe, forged from gas that
- 21:38Speaker 1: was barely enriched beyond its primordial composition. They carry chemical
- 21:43Speaker 1: fingerprints of the very first supernovae. They are direct windows
- 21:47Speaker 1: into the conditions that prevailed when the universe was still
- 21:50Speaker 1: assembling its first structures, and they are windows into dark
- 21:53Speaker 1: matter itself. Because ultra-faint dwarfs are almost entirely dark matter dominated,
- 21:59Speaker 1: their internal dynamics serve as direct probes of dark matter
- 22:02Speaker 1: behaviour on the smallest scales. The density profiles of their halos,
- 22:06Speaker 1: the minimum halo mass at which star formation becomes possible,
- 22:10Speaker 1: their survival rates against tidal disruption by host galaxies, all
- 22:15Speaker 1: of these observables place constraints on what dark matter particles
- 22:18Speaker 1: can and cannot do. Warm dark matter, which predicts fewer
- 22:22Speaker 1: small halos than cold dark matter, Fuzzy dark matter, which
- 22:26Speaker 1: creates halos with diffuse cores rather than dense cusps. Self-interacting
- 22:31Speaker 1: dark matter, which allows dark matter particles to scatter off
- 22:34Speaker 1: each other and reshape the inner structure of halos. Each
- 22:37Speaker 1: alternative model makes different predictions about how many ultra-faint dwarfs
- 22:41Speaker 1: should exist, how dense their central halos should be, and
- 22:45Speaker 1: how small the smallest galaxy can get before the concept
- 22:49Speaker 1: breaks down. The tiniest galaxies are not astronomical curiosities filed
- 22:53Speaker 1: away in catalogues and forgotten. They are among the most
- 22:56Speaker 1: powerful tests of fundamental particle physics available to observational astronomy.
- 23:01Speaker 1: By the early 2020s, the census of Milky Way satellites
- 23:04Speaker 1: had grown to roughly 60 known objects, a mix of
- 23:08Speaker 1: confirmed dwarf galaxies, confirmed star clusters, and a growing number
- 23:13Speaker 1: of ambiguous systems occupying the classification grey zone. Deep imaging
- 23:18Speaker 1: surveys kept probing fainter, pushing toward the detection limits of
- 23:22Speaker 1: current instrumentation. But even the most extreme known objects, Segway 1,
- 23:27Speaker 1: Wilman 1, Tucana 2, still contained at least a few
- 23:31Speaker 1: hundred visible stars and had luminosities of a few hundred
- 23:34Speaker 1: solar luminosities or more. Researchers wondered whether they were approaching
- 23:38Speaker 1: the bottom of the luminosity function, the faintest galaxies the
- 23:42Speaker 1: universe was capable of producing, or whether an entirely new
- 23:45Speaker 1: population of even smaller systems lurked below the detection threshold. Then,
- 23:50Speaker 1: in 2024, a survey operating from the summit of Mauna
- 23:55Speaker 1: Kea in Hawaii found something that rendered everything that came
- 23:58Speaker 1: before comparatively normal. It was a system so faint that
- 24:02Speaker 1: its total light output was only 11 times the luminosity
- 24:05Speaker 1: of our sun. It contained roughly 60 stars. Its total
- 24:09Speaker 1: stellar mass was approximately 16 solar masses. It was four
- 24:13Speaker 1: to five times less massive than the next faintest Milky
- 24:16Speaker 1: Way satellite ever recorded. Its physical size was typical for
- 24:20Speaker 1: a compact globular cluster, and the initial velocity measurements hinted
- 24:24Speaker 1: that it might be sitting inside a dark matter halo
- 24:27Speaker 1: weighing roughly one billion times more than all its visible
- 24:30Speaker 1: stars combined. Even its name reflects the confusion. The system
- 24:34Speaker 1: carries two designations simultaneously, Ursa Major 3 and Unions. One
- 24:41Speaker 1: because the scientific community cannot yet determine what it is.
- 24:45Speaker 1: One name assumes it is a galaxy. The other assumes
- 24:48Speaker 1: it is a star cluster. What happened when astronomers pointed
- 24:51Speaker 1: the most powerful spectroscopic instruments on Earth at this barely
- 24:55Speaker 1: visible smudge of ancient starlight is a story that would
- 24:59Speaker 1: shake the boundary between galaxy and star cluster harder than
- 25:03Speaker 1: anything that came before. The summit of Mauna Kea on
- 25:06Speaker 1: the island of Hawaii rises nearly 14,000 feet above the
- 25:10Speaker 1: Pacific Ocean. At that altitude, the air is thin and dry.
- 25:14Speaker 1: and the sky is so transparent that starlight arrives with
- 25:17Speaker 1: a clarity unmatched by almost any other location on Earth.
- 25:21Speaker 1: Perched near the summit sits the Canada-France-Hawaii telescope, a 3.6-metre
- 25:26Speaker 1: reflector that has been surveying the sky since 1979. It
- 25:29Speaker 1: is not the largest telescope on the mountain. The twin
- 25:33Speaker 1: Keck telescopes, each with 10-metre mirrors, tower nearby. The Subaru telescope,
- 25:40Speaker 1: the Gemini North Observatory, and several others share the summit.
- 25:44Speaker 1: But the Canada-France-Hawaii telescope possesses a particular advantage that makes
- 25:48Speaker 1: it indispensable for a certain kind of science. Its wide-field
- 25:52Speaker 1: imaging camera, called Megacam, can photograph an area of sky
- 25:56Speaker 1: roughly one degree across in a single exposure. This combination
- 26:00Speaker 1: of depth and breadth makes it an ideal instrument for
- 26:03Speaker 1: finding things that are faint, diffuse, and spread across large
- 26:07Speaker 1: patches of sky, exactly the kind of object that disappears
- 26:10Speaker 1: in the narrow gaze of a larger, more powerful telescope
- 26:14Speaker 1: focused on a pinpoint field of view. In the mid-2010s,
- 26:17Speaker 1: the Canada-France-Hawaii telescope joined a collaborative imaging campaign called the
- 26:22Speaker 1: Ultraviolet Near-Infrared Optical Northern Survey, known by its acronym UNIONS.
- 26:28Speaker 1: The survey pooled data from three telescopes across Hawaii, the
- 26:32Speaker 1: Canada-France-Hawaii telescope on Mauna Kea, the Pan-STARRS telescope on Haleakala
- 26:38Speaker 1: in Maui, and additional infrared observations from other facilities. The
- 26:42Speaker 1: goal was sweeping. Map the northern sky to a depth
- 26:45Speaker 1: sufficient to resolve individual faint stars at distances of tens
- 26:50Speaker 1: of thousands of light years, creating a census of the
- 26:53Speaker 1: Milky Way's stellar halo and its population of satellite companions.
- 26:57Speaker 1: UNIONS was designed not to find bright objects, but to
- 27:00Speaker 1: find the almost invisible ones. The faint smudges. The statistical
- 27:05Speaker 1: whispers hiding in fields crowded with foreground stars. Detecting an
- 27:09Speaker 1: ultra-faint satellite galaxy or an ultra-faint star cluster is not
- 27:13Speaker 1: like spotting a new star through a telescope. There is
- 27:16Speaker 1: no single bright point of light to notice. There is
- 27:19Speaker 1: no dramatic nebula or conspicuous concentration of stars that jumps
- 27:23Speaker 1: out of an image. What you are looking for is
- 27:25Speaker 1: a faint statistical excess, a region of sky where the
- 27:29Speaker 1: density of faint, old, metal-poor stars is slightly higher than
- 27:33Speaker 1: what the Milky Way's own stellar halo would produce by chance.
- 27:37Speaker 1: The signal is buried in noise. To extract it, astronomers
- 27:41Speaker 1: use a technique called matched filter analysis. They construct a
- 27:45Speaker 1: model of what a very old, very metal-poor stellar population
- 27:48Speaker 1: would look like at a given distance, including specific predictions
- 27:52Speaker 1: for the brightnesses and colours of stars along the main
- 27:55Speaker 1: sequence and the red giant branch. Then they scan the
- 27:58Speaker 1: survey data, pixel by pixel, looking for patches where real
- 28:03Speaker 1: stars cluster in colour magnitude space in a way that
- 28:05Speaker 1: matches the predicted model. Most patches show nothing. The filter
- 28:10Speaker 1: finds no excess. But occasionally, in one small area of sky,
- 28:15Speaker 1: a concentration appears, a handful of stars whose colors, magnitudes,
- 28:20Speaker 1: and spatial positions are all consistent with a single ancient
- 28:23Speaker 1: stellar population at a specific distance. That concentration is a candidate.
- 28:29Speaker 1: The difficulty of this search cannot be overstated. The Milky
- 28:33Speaker 1: Way's own stellar halo is not empty. It contains tens
- 28:36Speaker 1: of millions of stars scattered at various distances, spanning a
- 28:40Speaker 1: wide range of ages, metallicities and luminosities. Any patch of
- 28:45Speaker 1: sky observed at survey depth contains hundreds of foreground halo stars,
- 28:50Speaker 1: whose colours and magnitudes partially overlap with those of a hypothetical, ancient,
- 28:55Speaker 1: metal-poor satellite population at a similar distance. The signal you
- 28:59Speaker 1: are hunting for is not a cluster of stars that
- 29:02Speaker 1: stands out visually against a clean background. It is a
- 29:05Speaker 1: statistical excess of perhaps a few dozen stars sprinkled across
- 29:09Speaker 1: a region already populated by unrelated foreground objects whose photometric
- 29:15Speaker 1: properties are only subtly different from those of the target.
- 29:18Speaker 1: You are trying to find 60 particular grains of sand
- 29:21Speaker 1: on a beach by arguing that, in one small patch,
- 29:24Speaker 1: the grains are packed 3% more tightly than the background
- 29:27Speaker 1: average and that this excess is not a random fluctuation
- 29:31Speaker 1: but a real physical structure. The matched filter algorithm quantifies this.
- 29:36Speaker 1: It assigns a probability to each star based on how
- 29:39Speaker 1: closely it's color and magnitude match the predicted isochrone of
- 29:43Speaker 1: an old metal pore population at a given distance. Stars
- 29:47Speaker 1: that match well receive high weights. Stars that do not
- 29:50Speaker 1: match receive low weights. The algorithm then searches for spatial
- 29:54Speaker 1: concentrations of high weight stars and that exceed the expected
- 29:58Speaker 1: density of false positives from the foreground population. The threshold
- 30:02Speaker 1: for detection is typically set at a statistical significance of
- 30:06Speaker 1: 5 sigma or higher, meaning the observed concentration has less
- 30:11Speaker 1: than a 1 in a million probability of arising by
- 30:13Speaker 1: chance from the foreground alone. Even after a candidate passes
- 30:17Speaker 1: this threshold, confirmation is not automatic. A 5-sigma photometric detection
- 30:22Speaker 1: could still be a chance alignment of unrelated foreground stars
- 30:26Speaker 1: whose colours happen to mimic the target population. It could
- 30:29Speaker 1: be a tidal debris stream from a disrupted satellite that
- 30:32Speaker 1: is not gravitationally bound. It could be an artefact of
- 30:35Speaker 1: spatially varying survey depth, where a patch of sky observed
- 30:39Speaker 1: under slightly better conditions produces a spurious overdensity. Before investing
- 30:44Speaker 1: hours of precious spectroscopic time on a 10-metre telescope, the
- 30:48Speaker 1: Discovery team must cross-check the detection against independent photometric datasets,
- 30:53Speaker 1: verify that the candidate stars share consistent proper motions using
- 30:57Speaker 1: astrometric data from the Gaia space mission, and examine whether
- 31:01Speaker 1: the spatial distribution of candidate members is physically plausible for
- 31:04Speaker 1: a bound system rather than an elongated stream or a
- 31:08Speaker 1: patchy foreground fluctuation. URSA Major Three Unions One passed every check.
- 31:15Speaker 1: its candidate members clustered tightly in colour magnitude space along
- 31:19Speaker 1: a single old metal pore isochrone. Their spatial distribution was
- 31:23Speaker 1: compact and approximately round, consistent with a gravitationally bound system.
- 31:28Speaker 1: And critically, the Gaia proper motions of the brightest candidates
- 31:32Speaker 1: confirmed that they were co-moving through space as a coherent group,
- 31:36Speaker 1: ruling out a chance alignment of unrelated halo field stars
- 31:40Speaker 1: travelling in random directions. Simon Smith, a graduate student at
- 31:45Speaker 1: the University of Victoria, applied this technique to the Union's dataset.
- 31:49Speaker 1: The matched filter scan returned a detection in the constellation
- 31:52Speaker 1: Ursa Major, not far from the stars that form the
- 31:55Speaker 1: familiar shape of the Big Dipper. The concentration was extraordinarily faint.
- 32:01Speaker 1: It consisted of roughly 60 stars, tightly clustered in both
- 32:04Speaker 1: their positions on the sky and their locations in colour
- 32:08Speaker 1: magnitude space. The stars were consistent with a stellar population
- 32:12Speaker 1: older than 10 billion years, with a mean metallicity of
- 32:15Speaker 1: approximately negative 2.2 on the iron-to-hydrogen scale, meaning their iron
- 32:21Speaker 1: content was roughly 160 times lower than our sun's, a
- 32:26Speaker 1: composition characteristic of stars that formed in the earliest epochs
- 32:29Speaker 1: of the universe. The distance to the system was estimated
- 32:32Speaker 1: at approximately 10 kiloparsecs from the sun, or roughly 33,000 light-years.
- 32:38Speaker 1: In galactic terms, this is remarkably close. It is well
- 32:41Speaker 1: within the Milky Way's own halo. The object was not
- 32:44Speaker 1: a distant extragalactic discovery. It was practically next door. The
- 32:49Speaker 1: measured properties were staggering. The system's total luminosity was approximately
- 32:54Speaker 1: 11 solar luminosities. 11. Not 11,000. Not 11 million. 11
- 33:01Speaker 1: times the light of our sun. If you could somehow
- 33:03Speaker 1: replace our sun with this entire stellar system... Swapping a
- 33:07Speaker 1: single star for the combined output of roughly 60, the
- 33:10Speaker 1: change in brightness as seen from Earth would be modest.
- 33:13Speaker 1: Our Sun is an ordinary star of average luminosity. This
- 33:17Speaker 1: entire system was barely brighter. The estimated total stellar mass
- 33:22Speaker 1: was approximately 16 solar masses with an uncertainty of plus
- 33:26Speaker 1: six and minus five. 16 suns worth of material total
- 33:31Speaker 1: across all 60 or so stars. Most of those stars
- 33:34Speaker 1: are dim red dwarfs and ancient main sequence stars of
- 33:37Speaker 1: extremely low mass, contributing almost nothing to the total light output.
- 33:41Speaker 1: The system's physical size was compact, a half-light radius of
- 33:45Speaker 1: approximately three parsecs, or roughly 10 light years. This is small,
- 33:51Speaker 1: very small. For comparison, the nearest star to our sun,
- 33:55Speaker 1: Proxima Centauri, lies approximately 1.3 parsecs away. The half-light radius
- 34:01Speaker 1: of this system at roughly three parsecs, is barely more
- 34:04Speaker 1: than twice the distance between our Sun and its nearest
- 34:07Speaker 1: stellar neighbor. Within this compact volume, an entire candidate galaxy
- 34:12Speaker 1: concentrates roughly half of all its light. This combination of properties,
- 34:16Speaker 1: extreme faintness, tiny stellar mass, compact size, immediately placed the
- 34:22Speaker 1: system in uncharted territory. It was roughly four to five
- 34:25Speaker 1: times less massive than the next faintest known Milky Way satellites,
- 34:29Speaker 1: a pair of objects called Kim 3 and Delve 5.
- 34:33Speaker 1: It was approximately 10 times fainter than the faintest confirmed
- 34:37Speaker 1: ultra-faint dwarf galaxies. It was also approximately 10 times fainter
- 34:41Speaker 1: than the least luminous known globular cluster. Nothing previously discovered
- 34:46Speaker 1: occupied this corner of parameter space. Ha! The object sat
- 34:50Speaker 1: alone at the extreme end of every relevant axis. The faintest,
- 34:55Speaker 1: the least massive, the most extreme. Its compact size was
- 34:59Speaker 1: consistent with known globular clusters. Its luminosity was not consistent
- 35:04Speaker 1: with anything. The object needed a name, and here the
- 35:07Speaker 1: naming convention revealed the depth of the uncertainty. Newly discovered
- 35:12Speaker 1: dwarf galaxies orbiting the Milky Way are traditionally named after
- 35:16Speaker 1: the constellation in which they appear. Newly discovered star clusters
- 35:20Speaker 1: are traditionally named after the survey that found them. Because
- 35:23Speaker 1: the classification was genuinely unknown, the system received both names simultaneously.
- 35:29Speaker 1: Ursa Major 3, the dwarf galaxy designation, and Unions 1,
- 35:35Speaker 1: the star cluster designation. This dual naming was not a
- 35:39Speaker 1: bureaucratic formality. It was a public acknowledgement by the discovery
- 35:43Speaker 1: team that they could not tell what they had found.
- 35:45Speaker 1: The next step was spectroscopy. Imaging alone can tell you
- 35:49Speaker 1: where an object is, how bright it is, and roughly
- 35:52Speaker 1: what kind of stars it contains. but it cannot tell
- 35:55Speaker 1: you how those stars are moving. To measure internal motions,
- 35:58Speaker 1: astronomers need spectra, detailed measurements of starlight broken into its
- 36:02Speaker 1: component wavelengths, allowing precise determination of each star's radial velocity
- 36:08Speaker 1: through the Doppler shift. Blue shift means the star is approaching.
- 36:12Speaker 1: Red shift means the star is receding. The precision required
- 36:16Speaker 1: is extraordinary. Differences of a few kilometres per second tiny
- 36:20Speaker 1: fractions of a star's total orbital speed carry the entire
- 36:24Speaker 1: weight of the classification. Smith and colleagues obtained spectroscopic observations
- 36:28Speaker 1: with the 10-metre Keck 2 telescope on Mauna Kea using
- 36:32Speaker 1: the Demos Multi-Object Spectrograph, an instrument capable of observing dozens
- 36:38Speaker 1: of stellar spectra simultaneously. From these observations, they identified 11
- 36:42Speaker 1: stars whose radial velocities were consistent with membership in the system.
- 36:46Speaker 1: Eight of those 11 also had proper motion measurements from
- 36:49Speaker 1: the Gaia space mission, confirming that they were co-moving through
- 36:53Speaker 1: space as a coherent group. This was critical confirmation. The
- 36:57Speaker 1: system was not an optical illusion or a chance alignment
- 37:00Speaker 1: of unrelated foreground stars. It was a real, gravitationally bound structure.
- 37:06Speaker 1: From the 11 radial velocity members, Smith's team derived an
- 37:10Speaker 1: intrinsic velocity dispersion of approximately 3.7 km per second. with
- 37:16Speaker 1: an uncertainty range of plus 1.4 and minus 1.0. For
- 37:21Speaker 1: a system containing only 16 solar masses of visible stars,
- 37:25Speaker 1: this velocity dispersion was enormous. Apply the Virial Theorem, the
- 37:30Speaker 1: fundamental relationship between kinetic energy and gravitational potential energy, in
- 37:35Speaker 1: a system in equilibrium, and you get a mass estimate
- 37:39Speaker 1: in the hundreds of thousands of solar masses. The implied
- 37:42Speaker 1: mass-to-light ratio would be in the thousands. If the measurement
- 37:46Speaker 1: was correct and reflected the true gravitational potential, the system
- 37:50Speaker 1: would be overwhelmingly dominated by dark matter. The visible stars
- 37:54Speaker 1: would constitute a negligible fraction of the total mass. Ursa Major.
- 37:59Speaker 1: Three unions. One would not just be an ultra-faint dwarf galaxy.
- 38:05Speaker 1: It would be the darkest galaxy ever discovered. The system
- 38:08Speaker 1: with the highest ratio of dark matter to visible matter
- 38:11Speaker 1: ever measured. But Smith and colleagues were careful. they flagged
- 38:15Speaker 1: their own result with explicit caveats, cautioning against over-interpretation. The
- 38:20Speaker 1: velocity dispersion measurement was fragile. It depended sensitively on the
- 38:25Speaker 1: inclusion of two stars with velocities that deviated significantly from
- 38:29Speaker 1: the systemic mean. Remove the single largest velocity outlier, and
- 38:33Speaker 1: the dispersion dropped to approximately 1.9 km per second, still
- 38:38Speaker 1: potentially significant, but far less dramatic. Remove a second outlier
- 38:43Speaker 1: and and the dispersion became statistically unresolved, consistent with zero
- 38:48Speaker 1: internal spread beyond measurement errors. The difference between a dark
- 38:52Speaker 1: matter dominated galaxy and a mundane star cluster hinged on
- 38:56Speaker 1: whether one or two individual stars were genuine fast moving
- 38:59Speaker 1: members of the system, or binary stars whose orbital velocities were,
- 39:03Speaker 1: contaminating the measurement. Binary stars are the persistent saboteurs of
- 39:08Speaker 1: ultra faint dwarf galaxy research. When two stars orbit each
- 39:11Speaker 1: other in a close pair, the radial velocity of each
- 39:14Speaker 1: star oscillates as it swings toward and away from the
- 39:18Speaker 1: observer during each orbit. A single spectroscopic observation captures one
- 39:23Speaker 1: instantaneous velocity, but that velocity includes both the star's motion
- 39:27Speaker 1: within the larger system and its orbit motion around its
- 39:31Speaker 1: binary companion. If the binary period is short, the orbital
- 39:35Speaker 1: velocity can be several kilometers per second, easily large enough
- 39:38Speaker 1: to inflate the apparent velocity dispersion of the entire system.
- 39:42Speaker 1: The only way to disentangle binary motion from genuine systemic
- 39:45Speaker 1: motion is to observe the same star multiple times across
- 39:49Speaker 1: different epochs. If its velocity changes significantly between observations, it
- 39:54Speaker 1: is likely a binary. If it remains stable, it is
- 39:57Speaker 1: more likely a single star, whose velocity reflects the gravitational
- 40:01Speaker 1: potential of the host system. Smith's initial observations were single epoch,
- 40:06Speaker 1: there were no repeat measurements. The binary contamination risk was
- 40:09Speaker 1: acknowledged but unresolved. A companion study, coordinated with the discovery
- 40:14Speaker 1: team and published shortly afterward, took the investigation a step further.
- 40:18Speaker 1: Rafael Arani, then at Carnegie Mellon University, and collaborators used
- 40:24Speaker 1: N-body simulations to model the tidal evolution of the system
- 40:28Speaker 1: as it orbits the Milky Way. The question was straightforward.
- 40:31Speaker 1: Can a self-gravitating cluster of 16 solar masses survive on
- 40:35Speaker 1: this orbit? The orbit itself had been determined using the
- 40:38Speaker 1: system's position, radial velocity, and Gaia proper motions. It is
- 40:43Speaker 1: a halo-like trajectory that plunging as close as approximately 13
- 40:48Speaker 1: kiloparsecs from the Milky Way's centre at pericentre, and swinging
- 40:52Speaker 1: out to roughly 30 kiloparsecs at a pocentre. The radial
- 40:55Speaker 1: orbital period is approximately 0.4 billion years. At pericentre, the
- 41:00Speaker 1: system passes deep into the Milky Way's gravitational domain, where
- 41:04Speaker 1: tidal forces are intense. The Milky Way's gravity pulls harder
- 41:08Speaker 1: on the near side of the system than on the
- 41:10Speaker 1: far side, creating a stretching effect that strips stars from
- 41:14Speaker 1: the outer region's and gradually unbinds the structure. Irani's simulations
- 41:19Speaker 1: model the system as a self-gravitating cluster stars held together
- 41:24Speaker 1: only by their own mutual gravity. With no dark matter
- 41:28Speaker 1: and evolved it forward through the Milky Way's tidal field,
- 41:31Speaker 1: the results were unambiguous. Every self-gravitating model, regardless of initial mass,
- 41:37Speaker 1: fully disrupted within roughly 0.6 billion years from the present,
- 41:42Speaker 1: less than two orbital periods. The system would be torn
- 41:45Speaker 1: apart within a cosmically brief interval. This was not surprising.
- 41:49Speaker 1: The mean density of a 16 solar mass cluster compressed
- 41:53Speaker 1: into a 3 parsec half-light radius is comparable to the
- 41:56Speaker 1: mean density of the Milky Way itself, enclosed within the
- 42:00Speaker 1: pericentric radius. The system is not significantly denser than its
- 42:04Speaker 1: tidal environment. Tidal forces overwhelm it. To understand what tidal
- 42:08Speaker 1: disruption means physically, Consider what happens each time the system
- 42:12Speaker 1: reaches pericentre its closest approach to the Milky Way's centre,
- 42:16Speaker 1: roughly 13 kiloparsecs from the core. At this distance, the
- 42:21Speaker 1: Milky Way's gravitational field is intense. The circular velocity of
- 42:25Speaker 1: the galaxy at this radius is approximately 240 kilometres per second,
- 42:30Speaker 1: reflecting the enormous mass enclosed within that sphere. The tidal
- 42:34Speaker 1: field at pericentre does not simply tug gently on the
- 42:37Speaker 1: system from one direction, it creates a differential force. The
- 42:41Speaker 1: near side of the cluster is pulled more strongly toward
- 42:43Speaker 1: the galactic centre than the far side, and simultaneously the
- 42:47Speaker 1: far side is pushed outward by the centrifugal component of
- 42:50Speaker 1: the orbital motion. This stretching force acts like a pair
- 42:53Speaker 1: of invisible hands pulling the system apart along the axis,
- 42:57Speaker 1: connecting it to the Milky Way's centre. For a massive,
- 43:00Speaker 1: dense system like a typical globular cluster, 100,000 solar masses
- 43:05Speaker 1: compressed into a few parsecs, the internal gravity easily overpowers
- 43:10Speaker 1: the tidal stretching. The cluster passes through pericentre, loses a
- 43:14Speaker 1: handful of loosely bound outer stars, and continues intact. But
- 43:19Speaker 1: for a system containing only 16 solar masses spread across
- 43:22Speaker 1: three parsecs, the internal gravity is feeble. The tidal forces
- 43:26Speaker 1: at pericentre are comparable in strength to the self-gravity holding
- 43:30Speaker 1: the system together. Each passage strips away a significant fraction
- 43:34Speaker 1: of the remaining stars, The process is cumulative and irreversible.
- 43:39Speaker 1: Stars lost to tidal stripping do not return. They trail
- 43:42Speaker 1: behind and ahead of the system along its orbit, forming
- 43:46Speaker 1: tidal tails that gradually disperse into the Milky Way's halo.
- 43:50Speaker 1: Irani's simulations tracked this process in detail. They varied the
- 43:54Speaker 1: initial mass of the self-gravitating cluster across a wide range,
- 43:58Speaker 1: from just above the current observed mass up to several
- 44:01Speaker 1: hundred solar masses, testing whether a more massive progenitor could
- 44:05Speaker 1: survive longer by starting with a larger reservoir of stars
- 44:09Speaker 1: to lose. The result was unequivocal. Regardless of initial mass,
- 44:13Speaker 1: every self-gravitating model reached complete disruption within roughly 0.6 billion
- 44:18Speaker 1: years less than two full orbital cycles. The system would
- 44:21Speaker 1: dissolve entirely, its stars scattered along a tidal stream wrapping
- 44:25Speaker 1: around the Milky Way. No self-gravitating configuration survived. This created
- 44:30Speaker 1: a powerful argument. The stars in Ursa Major 3, Unions
- 44:34Speaker 1: 1 are older than 10 billion years. The system has
- 44:38Speaker 1: presumably been orbiting the Milky Way for many billions of years,
- 44:41Speaker 1: completing dozens of orbits through the tidal danger zone near Pericenter.
- 44:45Speaker 1: If it were self-gravitating, with nothing but its own stellar
- 44:48Speaker 1: mass holding it together, it should have been destroyed billions
- 44:52Speaker 1: of years ago. The fact that it still exists as
- 44:55Speaker 1: a coherent, bound structure strongly suggests that something additional is
- 45:00Speaker 1: stabilising it. something providing a gravitational well deep enough to
- 45:04Speaker 1: resist the Milky Way's tidal forces across many orbital periods.
- 45:08Speaker 1: The natural candidate is dark matter. Iranian colleagues explored what
- 45:12Speaker 1: kind of dark matter halo could plausibly host such a system.
- 45:16Speaker 1: Within the Lambda Cold Dark Matter framework, the first dark
- 45:20Speaker 1: matter halos to form stars are expected to have virial
- 45:24Speaker 1: masses of roughly 100 million to 1 billion solar masses.
- 45:28Speaker 1: Depending on the epoch of formation and the strength of
- 45:30Speaker 1: reionization feedback, a halo at the lower end of this range,
- 45:35Speaker 1: formed at a redshift of approximately two corresponding to roughly
- 45:39Speaker 1: 11 billion years ago, consistent with the inferred stellar age,
- 45:44Speaker 1: would develop a central density profile following the Navarro-Frank white form,
- 45:49Speaker 1: with a dense, cuspy core that resists tidal stripping far
- 45:54Speaker 1: more effectively than a uniform distribution of mass. The simulations
- 45:58Speaker 1: showed that if Ursa Major 3 formed at the centre
- 46:00Speaker 1: of a cuspy dark matter halo, with a virile mass
- 46:04Speaker 1: of roughly 1 billion solar masses, the halo's dense core
- 46:08Speaker 1: would survive the Milky Way's tidal field. Over many billions
- 46:11Speaker 1: of years and dozens of pericentric passages, the outer layers
- 46:15Speaker 1: of the halo would be gradually stripped away. But the
- 46:18Speaker 1: innermost region, where the stars reside, would remain intact. the
- 46:23Speaker 1: predicted velocity dispersion for stars embedded in such a surviving
- 46:27Speaker 1: remnant halo would be approximately one kilometre per second. This
- 46:31Speaker 1: prediction was broadly consistent with the observed velocity dispersion range
- 46:35Speaker 1: between roughly one and four kilometres per second, depending on
- 46:39Speaker 1: which stars were included, though it fell below the nominal
- 46:42Speaker 1: best fit value of 3.7. The implications, if correct, were extraordinary.
- 46:47Speaker 1: A one billion solar mass dark matter halo containing only
- 46:51Speaker 1: 16 solar masses of stars, would represent a mass ratio
- 46:55Speaker 1: of roughly 60 million to 1. For every unit of
- 46:59Speaker 1: visible matter, there would be 60 million units of dark matter.
- 47:03Speaker 1: This would eclipse the mass-to-light ratios of every other known
- 47:06Speaker 1: galaxy by an enormous margin. Segway 1, with its mass-to-light
- 47:11Speaker 1: ratio of roughly 1,000 to 3,000, would seem almost star-rich
- 47:16Speaker 1: by comparison. Ursa Major 3, be a galaxy in which
- 47:20Speaker 1: the visible component is so negligible that the system is,
- 47:24Speaker 1: for all practical gravitational purposes, a pure dark matter object
- 47:28Speaker 1: with a few ancient stars sprinkled through its centre as
- 47:31Speaker 1: witnesses to its existence. This prospect electrified the astrophysics community
- 47:36Speaker 1: for reasons that extended well beyond the record-breaking statistics. If
- 47:41Speaker 1: a dark matter halo of one billion solar masses could
- 47:44Speaker 1: host a galaxy with only 16 solar masses of stars,
- 47:48Speaker 1: It meant that the minimum luminosity for galaxy formation was
- 47:51Speaker 1: far lower than anyone had previously measured. It meant that
- 47:54Speaker 1: the universe could produce galaxies that are functionally invisible systems
- 47:58Speaker 1: containing dark matter, halos of enormous mass, but so few
- 48:03Speaker 1: stars that they are undetectable by any existing survey unless
- 48:06Speaker 1: they happen to sit extremely close to us. Ursa Major
- 48:10Speaker 1: Three Unions. One is only 10 kiloparsecs from the sun.
- 48:15Speaker 1: If it were located at the distance of most known
- 48:17Speaker 1: ultra-faint dwarfs 50 to 100 kiloparsecs, it would be completely
- 48:21Speaker 1: undetectable with current technology. This raised the unsettling possibility that
- 48:26Speaker 1: the Milky Way might be surrounded by hundreds or thousands
- 48:29Speaker 1: of similar objects. Dark matter halos containing only a handful
- 48:33Speaker 1: of ancient stars, lurking below the detection threshold in every direction.
- 48:37Speaker 1: It also placed severe pressure on alternative dark matter models.
- 48:41Speaker 1: Fuzzy dark matter. A model proposing that dark matter consists
- 48:45Speaker 1: of extremely light particles with de Broglie wavelengths comparable to
- 48:48Speaker 1: the size of dwarf galaxy cores predicts that halos below
- 48:52Speaker 1: a certain mass cannot form dense central concentrations. The extremely
- 48:56Speaker 1: high inferred density of the Ursa Major III dark matter halo,
- 49:00Speaker 1: if confirmed, would be very difficult to reconcile with standard
- 49:03Speaker 1: fuzzy dark matter models. Warm dark matter models, which predict
- 49:07Speaker 1: fewer small halos than cold dark matter, would also face constraints.
- 49:11Speaker 1: Self-interacting dark matter, which tends to create constant density cores
- 49:16Speaker 1: rather than cuspy centres, would struggle to produce the dense
- 49:19Speaker 1: inner profile needed to survive tidal stripping on such a
- 49:22Speaker 1: tight orbit. Each alternative dark matter model would need to
- 49:25Speaker 1: explain how this system exists, and for several models, explaining
- 49:29Speaker 1: it would be extremely difficult or impossible. The theoretical community
- 49:34Speaker 1: responded rapidly. Papers analysing the implications for dark matter physics
- 49:39Speaker 1: appeared within weeks of the discovery. Groups calculated predicted gamma-ray
- 49:43Speaker 1: signals from dark matter annihilation in the dense core. Others
- 49:46Speaker 1: modelled the expected tidal tails that should extend from the system,
- 49:51Speaker 1: if it was undergoing even mild stripping. The object's extreme
- 49:54Speaker 1: proximity to the Sun made it the closest known galaxy candidate,
- 49:59Speaker 1: if it was indeed a galaxy, offering a laboratory for
- 50:02Speaker 1: dark matter studies at a distance no other system could match.
- 50:06Speaker 1: Within months, multiple independent groups published analyses exploring the system's
- 50:10Speaker 1: implications across different branches of dark matter physics. Several teams
- 50:14Speaker 1: calculated the expected gamma-ray flux from dark matter particle annihilation
- 50:19Speaker 1: at the system's inferred central density. If the dark matter
- 50:23Speaker 1: halo was as dense as the velocity dispersion suggested, the
- 50:26Speaker 1: annihilation rate would be extraordinarily high per unit volume, and
- 50:30Speaker 1: the system's proximity only 10 kiloparsecs away would make the
- 50:34Speaker 1: signal potentially detectable by the Fermi Large Area Telescope. The
- 50:38Speaker 1: predicted flux depended sensitively on the halo's inner density slope,
- 50:43Speaker 1: with cuspy Navarro-Frank white profiles yielding signals orders of magnitude
- 50:47Speaker 1: stronger than chord profiles. A detection would not merely confirm
- 50:51Speaker 1: the presence of dark matter. It would constrain the annihilation
- 50:55Speaker 1: cross-section of dark matter particles, a fundamental parameter in particle
- 50:59Speaker 1: physics that no laboratory experiment has yet been able to measure.
- 51:03Speaker 1: Other groups examined the implications for the minimum halo mass
- 51:06Speaker 1: capable of forming a luminous galaxy. If a halo of
- 51:10Speaker 1: roughly 1 billion solar masses can produce a galaxy containing
- 51:13Speaker 1: only 16 solar masses of stars, then the overwhelming majority
- 51:18Speaker 1: of the halo's baryonic budget was never converted into visible
- 51:21Speaker 1: stars at all, a formation efficiency so low it challenges
- 51:26Speaker 1: the most pessimistic models of how star formation shuts down
- 51:29Speaker 1: in the smallest halos. This efficiency is far below what
- 51:33Speaker 1: even the most pessimistic models of reionization feedback had predicted.
- 51:38Speaker 1: It would imply that star formation in the smallest halos
- 51:40Speaker 1: is not merely suppressed but almost entirely extinguished, with only
- 51:44Speaker 1: a vanishing trace of baryonic matter managing to cool, collapse,
- 51:49Speaker 1: and ignite nuclear fusion before the gas is heated beyond recovery.
- 51:53Speaker 1: Still others examined what the system's density would mean for
- 51:56Speaker 1: fuzzy dark matter. Finding that standard fuzzy dark matter particle
- 52:00Speaker 1: masses in the range of 10 to the negative 22
- 52:02Speaker 1: electron volts would produce halo cores too diffused to match
- 52:06Speaker 1: the inferred central density, potentially ruling out a significant portion
- 52:11Speaker 1: of the fuzzy dark matter parameter space if the galaxy
- 52:13Speaker 1: interpretation held. But all of these implications rested on a
- 52:17Speaker 1: single unstable foundation, the velocity dispersion measurement. If the dispersion
- 52:22Speaker 1: of 3.7 kilometers per second was accurate, the dark matter
- 52:26Speaker 1: case was strong. If the true dispersion was closer to 1.9,
- 52:30Speaker 1: the value obtained when the single largest outlier was removed,
- 52:35Speaker 1: the case was still plausible but weaker, and the mass-to-light
- 52:38Speaker 1: ratio dropped significantly. If the dispersion was unresolved, consistent with zero,
- 52:44Speaker 1: then there was no kinematic evidence for dark matter at all.
- 52:47Speaker 1: The system could be a self-gravitating star cluster. and the
- 52:51Speaker 1: tidal survival problem would need to be explained through some
- 52:54Speaker 1: other mechanism, perhaps a population of unseen stellar remnants adding mass,
- 52:58Speaker 1: or perhaps a recent formation scenario in which the system
- 53:01Speaker 1: had not yet completed enough orbits to disrupt. Everything depended
- 53:05Speaker 1: on what happened when astronomers returned to the Keck 2
- 53:08Speaker 1: telescope with a new plan. They needed multi-epoch observations, the
- 53:13Speaker 1: same stars measured again, months or years later, to check
- 53:17Speaker 1: for binary velocity variations. They needed higher spectral resolution, sharper
- 53:22Speaker 1: measurements of each star's velocity to reduce the per-star uncertainty
- 53:26Speaker 1: that blurred the dispersion estimate. And they needed chemical abundances,
- 53:30Speaker 1: precise metallicities for individual stars to determine whether the system
- 53:34Speaker 1: showed the chemical enrichment spread expected of a galaxy or
- 53:38Speaker 1: the chemical uniformity expected of a cluster. The initial discovery
- 53:42Speaker 1: had identified the most extreme Milky Way satellite ever found.
- 53:46Speaker 1: It had shown that the object either contained dark matter
- 53:49Speaker 1: in quantities that would make it the darkest galaxy in
- 53:52Speaker 1: the known universe, or that it was the faintest self-gravitating
- 53:55Speaker 1: star cluster ever seen an object whose survival against tidal
- 53:59Speaker 1: destruction would require its own explanation. The two possibilities were
- 54:03Speaker 1: not merely different answers to the same question. They pointed
- 54:06Speaker 1: toward fundamentally different conclusions about how the universe builds structure
- 54:10Speaker 1: at the smallest scales, how dark matter behaves in the
- 54:14Speaker 1: lowest mass halos, And whether the word galaxy can stretch
- 54:17Speaker 1: to cover a system containing fewer stars than a typical
- 54:20Speaker 1: open cluster in the Milky Way's disk, the Discovery team
- 54:23Speaker 1: knew that only new data could resolve the tension. They
- 54:26Speaker 1: had planted a flag at the extreme boundary of observational astronomy,
- 54:30Speaker 1: at the threshold where 60 ancient stars either sit inside
- 54:34Speaker 1: an invisible dark matter mountain or drift alone through the
- 54:38Speaker 1: tidal violence of the Milky Way's gravitational field, bound only
- 54:42Speaker 1: to each other and to whatever unseen remnants might share
- 54:45Speaker 1: their tiny volume of space. What came back from the
- 54:48Speaker 1: follow-up campaign did not settle the question the way anyone expected.
- 54:52Speaker 1: It shifted the weight of evidence in a direction that
- 54:54Speaker 1: made the story more complicated, not less. William Cerny had
- 54:58Speaker 1: been studying the faintest satellites of the Milky Way for
- 55:01Speaker 1: years before Ursa Major 3 Unions 1 appeared in the data.
- 55:05Speaker 1: As a graduate student and then postdoctoral researcher at Yale University,
- 55:09Speaker 1: He had built a career around the objects that live
- 55:11Speaker 1: at the classification boundary, the systems too faint for easy answers,
- 55:16Speaker 1: too ambiguous for confident labels. He understood better than most
- 55:20Speaker 1: how fragile velocity dispersion measurements become when the entire calculation
- 55:24Speaker 1: rests on a handful of stars. And he understood that
- 55:27Speaker 1: the extraordinary claim attached to Ursa Major three unions, one
- 55:31Speaker 1: that it might be the darkest galaxy in the known universe,
- 55:34Speaker 1: demanded extraordinary scrutiny. The follow-up campaign began with a return
- 55:39Speaker 1: to the Keck Observatory on Mauna Kea. CERNI and collaborators
- 55:42Speaker 1: used two instruments across the twin Keck telescopes, Demos on
- 55:46Speaker 1: Keck 2, the same multi-object spectrograph that had provided the
- 55:51Speaker 1: original velocity measurements, and ELRIS on Keck 1, the low-resolution
- 55:56Speaker 1: imaging spectrometer, which offered complementary capabilities for measuring chemical abundances
- 56:01Speaker 1: through different spectral features. The goal was twofold. First, Obtain
- 56:06Speaker 1: new velocity measurements for the known member stars, ideally at
- 56:09Speaker 1: multiple epochs separated by months, to check for binary-induced velocity variations. Second,
- 56:16Speaker 1: measure the chemical compositions of individual stars with enough precision
- 56:20Speaker 1: to determine whether the system displayed a metallicity spread consistent
- 56:24Speaker 1: with a galaxy or the chemical uniformity expected of a
- 56:28Speaker 1: star cluster. The new observations expanded the confirmed member sample.
- 56:32Speaker 1: Cerny's team identified a 12th star as a radial velocity member,
- 56:36Speaker 1: bringing the total to 12 stars with secure membership. More importantly,
- 56:41Speaker 1: the new velocity measurements carried substantially improved precision. The original Smith,
- 56:46Speaker 1: et al., data had per-star velocity uncertainties that, while adequate
- 56:51Speaker 1: for identifying members, left room for significant ambiguity in the
- 56:55Speaker 1: dispersion calculation. The new data sharpened these measurements considerably, reducing
- 57:00Speaker 1: the individual uncertainties and providing a cleaner foundation for the
- 57:04Speaker 1: statistical analysis. The results overturned the original picture. When Cerny
- 57:09Speaker 1: and collaborators applied Bayesian analysis to the updated velocity dataset
- 57:13Speaker 1: incorporating the expanded member sample, the improved per-star precisions, and
- 57:18Speaker 1: a careful treatment of binary contamination, the velocity dispersion came
- 57:22Speaker 1: back dramatically lower than the initial estimate. The analysis favoured
- 57:26Speaker 1: a value of approximately 0.1 km per second. the 95%
- 57:31Speaker 1: credible upper limit was 2.3 km per second. The original
- 57:35Speaker 1: value of 3.7 km per second, the measurement that had
- 57:39Speaker 1: launched the darkest galaxy hypothesis, was now ruled out by
- 57:42Speaker 1: the new data. This was not a subtle revision. It
- 57:46Speaker 1: was a near-complete inversion of the kinematic evidence. The original
- 57:50Speaker 1: measurement had implied a mass-to-light ratio in the thousands, requiring
- 57:54Speaker 1: an enormous dark matter halo to explain the stellar motions.
- 57:57Speaker 1: The new measurement implied little to no excess mass beyond
- 58:01Speaker 1: the stars themselves. The gravitational signature of dark matter, which
- 58:05Speaker 1: had seemed so promising in the initial data, had largely disappeared.
- 58:09Speaker 1: What caused such a dramatic shift? The answer involved exactly
- 58:13Speaker 1: the culprit that the Discovery team had warned about. Binary stars.
- 58:18Speaker 1: One of the 11 original member stars had been flagged
- 58:21Speaker 1: by Smith and colleagues as a suspected spectroscopic binary. a
- 58:25Speaker 1: star whose velocity appeared inconsistent across different spectral lines or
- 58:29Speaker 1: observing conditions. The new multi-epoch observations confirmed this suspicion. The
- 58:35Speaker 1: star's velocity had changed substantially between the original measurement and
- 58:38Speaker 1: the follow-up, consistent with orbital motion around an unseen companion.
- 58:43Speaker 1: This single binary, when included in the dispersion calculation, had
- 58:48Speaker 1: inflated the velocity spread of the entire sample. Remove it,
- 58:52Speaker 1: and the dispersion drops. The new data also identified three
- 58:56Speaker 1: additional binary candidates among the member stars, further complicating the picture.
- 59:01Speaker 1: In a system containing only 12 confirmed members, even a
- 59:05Speaker 1: few binaries can dominate the statistics. Each contaminated velocity measurement
- 59:10Speaker 1: pulls the calculated dispersion upward, mimicking the gravitational signature of
- 59:14Speaker 1: dark matter where none may exist. Binary contamination is a
- 59:18Speaker 1: fundamental challenge that shapes every result at this luminosity frontier.
- 59:23Speaker 1: In a massive galaxy containing millions of observable stars, it
- 59:27Speaker 1: averages out For every binary moving toward the observer, another
- 59:31Speaker 1: moves away. The net effect on the overall dispersion is negligible.
- 59:35Speaker 1: But in a system with 12 members, the law of
- 59:37Speaker 1: large numbers does not apply. One or two binaries with
- 59:41Speaker 1: unfavorable orbital phases at the moment of observation can shift
- 59:44Speaker 1: the calculated dispersion by a factor of two or more.
- 59:47Speaker 1: The statistical machinery is simply not robust at such small
- 59:51Speaker 1: sample sizes. Every individual data point carries disproportionate weight. and
- 59:56Speaker 1: every individual binary carries disproportionate risk. Cerny's team went further.
- 1:00:02Speaker 1: Having addressed the kinematic evidence, they turned to the second
- 1:00:05Speaker 1: major diagnostic, metallicity. Using the LRIS spectrograph, they obtained measurements
- 1:00:11Speaker 1: of the calcium 2K absorption line for the 12 brightest
- 1:00:14Speaker 1: confirmed members. Calcium 2K is a strong spectral feature that
- 1:00:18Speaker 1: serves as a reliable proxy for overall iron abundance in
- 1:00:21Speaker 1: metal-poor stars. By measuring the strength of this line in
- 1:00:25Speaker 1: each star, Cernian collaborators could estimate individual stellar metallicities and
- 1:00:30Speaker 1: crucially determine whether those metallicities were uniform or spread across
- 1:00:34Speaker 1: a range. The results reinforced the kinematic finding. The 12
- 1:00:39Speaker 1: stars showed no evidence of significant internal metallicity variation. The
- 1:00:43Speaker 1: metallicities were uniformly low, centered around an iron-to-hydrogen ratio of
- 1:00:48Speaker 1: approximately negative 2.2, with a spread consistent with measurement uncertainty alone.
- 1:00:54Speaker 1: The upper limit on the intrinsic metallicity dispersion was small,
- 1:00:58Speaker 1: far smaller than the typical metallicity spreads of 0.3 to 0.7
- 1:01:04Speaker 1: dex observed in confirmed ultra-faint dwarf galaxies, where multiple generations
- 1:01:09Speaker 1: of star formation create chemically distinct populations. Instead, the chemical
- 1:01:15Speaker 1: uniformity of Ursa Major III unions, one more closely resembled
- 1:01:20Speaker 1: the nearly homogeneous compositions seen in ancient globular clusters, where
- 1:01:24Speaker 1: all stars formed in a single episode from the same
- 1:01:27Speaker 1: parental gas cloud, and no subsequent enrichment occurred within the system.
- 1:01:32Speaker 1: This was the second pillar of the galaxy hypothesis crumbling.
- 1:01:35Speaker 1: The first pillar elevated velocity dispersion, implying dark matter had
- 1:01:40Speaker 1: fallen to binary contamination and improved measurements. The second pillar
- 1:01:44Speaker 1: metallicity spread, implying chemical self-enrichment within a deep potential well,
- 1:01:49Speaker 1: had failed to appear in the data at all. Neither
- 1:01:52Speaker 1: of the two primary diagnostics used to distinguish galaxies from
- 1:01:55Speaker 1: star clusters at ultra-faint luminosities supported the galaxy interpretation. Cerny
- 1:02:00Speaker 1: and collaborators drew a careful conclusion. Based on the available evidence,
- 1:02:05Speaker 1: they stated that Ursa Major III Unions, one is more
- 1:02:09Speaker 1: consistent with a star cluster than with an ultra-faint dwarf galaxy.
- 1:02:14Speaker 1: They noted that there are no confirmed dwarf galaxies for
- 1:02:17Speaker 1: which both the velocity dispersion and the metallicity dispersion are
- 1:02:21Speaker 1: as low as the limits measured for this system. Every
- 1:02:24Speaker 1: confirmed ultra-faint dwarf has at least one clear indicator, either
- 1:02:28Speaker 1: a resolved velocity dispersion significantly exceeding what its visible stars
- 1:02:33Speaker 1: can account for, or a metallicity spread demonstrating chemical self-enrichment
- 1:02:37Speaker 1: across stellar generations, or both. Ursa Major 3 Unions 1
- 1:02:42Speaker 1: shows neither. But Cerny's team also exercised restraint. They did
- 1:02:46Speaker 1: not declare the galaxy hypothesis dead. They acknowledged explicitly that
- 1:02:51Speaker 1: the measurements could not fully rule out the presence of
- 1:02:53Speaker 1: dark matter. The velocity dispersion upper limit of 2.3 km
- 1:02:57Speaker 1: per second, while dramatically lower than the original estimate, is
- 1:03:01Speaker 1: not zero. A modest dark matter halo could still produce
- 1:03:05Speaker 1: a dispersion within this range without exceeding the upper bound.
- 1:03:08Speaker 1: And the metallicity dispersion limit, while small, rests on measurements
- 1:03:13Speaker 1: of only 12 stars, a sample that could miss a
- 1:03:15Speaker 1: subtle chemical spread if it exists among fainter, unmeasured members.
- 1:03:20Speaker 1: The dwarf galaxy scenario remains viable, pending higher-resolution chemical abundance
- 1:03:25Speaker 1: patterns and even tighter multi-epoch kinematic constraints. The paper's title
- 1:03:29Speaker 1: captured the measured tone of the conclusion. No observational evidence
- 1:03:33Speaker 1: for dark matter, nor a large metallicity, spread in the
- 1:03:37Speaker 1: extreme Milky Way satellite, Ursa Major 3 Unions 1. Not
- 1:03:42Speaker 1: a definitive rejection of the galaxy hypothesis. An absence of
- 1:03:46Speaker 1: positive evidence for it. A distinction that matters enormously in
- 1:03:50Speaker 1: a field where classification often depends on which uncertainty you
- 1:03:53Speaker 1: find more tolerable. The shift in evidence did not eliminate
- 1:03:57Speaker 1: the mysteries surrounding the system. It rearranged them. If Ursa
- 1:04:01Speaker 1: Major 3 Unions 1 is not a dark matter dominated galaxy,
- 1:04:05Speaker 1: a new question immediately takes the stage. How? Has it survived?
- 1:04:11Speaker 1: The tidal survival problem identified by Irani and collaborators in
- 1:04:15Speaker 1: 2024 remains. A self-gravitating cluster of 16 solar masses on
- 1:04:19Speaker 1: this orbit should be destroyed by the Milky Way's tidal
- 1:04:22Speaker 1: field within a cosmically brief interval. The stars are old.
- 1:04:26Speaker 1: The orbit has been repeating for billions of years. If
- 1:04:29Speaker 1: nothing but the mutual gravity of 16 solar masses of
- 1:04:32Speaker 1: stars holds the system together, it should have been torn
- 1:04:35Speaker 1: apart long ago. Removing dark matter from the explanation, does
- 1:04:39Speaker 1: not remove the need for an explanation. Something must still
- 1:04:43Speaker 1: be stabilising the system, or else its survival requires an
- 1:04:46Speaker 1: improbable coincidence of timing. The possibility that we happen to have...
- 1:04:50Speaker 1: Rostami Shirazi and collaborators, in a study published in 2025,
- 1:04:56Speaker 1: used N-body simulations to model this scenario explicitly. They found
- 1:05:01Speaker 1: that a dark matter deficient progenitor, a cluster that never
- 1:05:04Speaker 1: sat inside a dark matter halo, could reproduce the current
- 1:05:08Speaker 1: observed properties of Ursa Major, three unions. One, if it
- 1:05:13Speaker 1: retained a significant population of stellar mass black holes with
- 1:05:16Speaker 1: relatively weak natal, kicks. In this model, the black holes
- 1:05:21Speaker 1: provide the gravitational glue that dark matter was originally invoked
- 1:05:24Speaker 1: to explain. The system survives tidal disruption not because it
- 1:05:28Speaker 1: sits inside an invisible halo of exotic particles, but because
- 1:05:32Speaker 1: it harbors an invisible population of dead stars, massive enough
- 1:05:36Speaker 1: to deepen its gravitational well. This scenario is physically plausible.
- 1:05:41Speaker 1: Observations of other low-mass stellar systems have provided growing evidence
- 1:05:44Speaker 1: that black hole retention is more common than traditionally assumed.
- 1:05:49Speaker 1: Gravitational wave detections of merging black hole binaries by the
- 1:05:52Speaker 1: LIGO and Virgo observatories have revealed a population of stellar-mass
- 1:05:56Speaker 1: black holes spanning a broad range of masses, some of
- 1:06:00Speaker 1: which formed in dense stellar environments. If even a modest
- 1:06:03Speaker 1: number of such objects remain bound within Ursa Major 3
- 1:06:07Speaker 1: Unions 1, their combined mass could exceed the visible stellar
- 1:06:11Speaker 1: mass by a significant factor, stabilising the system without requiring
- 1:06:15Speaker 1: dark matter. The Dark Star Cluster model also offers a
- 1:06:18Speaker 1: natural explanation for the inflated velocity dispersion in the original data.
- 1:06:23Speaker 1: If unseen compact remnants contribute substantial mass to the system,
- 1:06:28Speaker 1: they deepen the gravitational potential well. Stars embedded in that
- 1:06:32Speaker 1: deeper well move faster than they would in a system
- 1:06:35Speaker 1: composed solely of visible stars. Even without dark matter, the
- 1:06:39Speaker 1: presence of black holes and neutron stars could produce a
- 1:06:42Speaker 1: velocity dispersion moderately above what 16 solar masses of luminous
- 1:06:46Speaker 1: stars would predict. This intermediate scenario, no dark matter, but
- 1:06:51Speaker 1: significant remnant mass occupies the space between the two extremes
- 1:06:55Speaker 1: that dominated the initial debate. But the dark star cluster
- 1:06:58Speaker 1: model has its own uncertainties. The number and mass distribution
- 1:07:02Speaker 1: of retained black holes depend on assumptions about the initial
- 1:07:05Speaker 1: mass function of the system's first stellar generation, the physics
- 1:07:09Speaker 1: of black hole natal kicks, and the dynamical history of
- 1:07:12Speaker 1: the cluster over 10 billion years. These parameters are poorly
- 1:07:16Speaker 1: constrained for an object this faint and this old. The
- 1:07:19Speaker 1: model is consistent with the data. It is not uniquely
- 1:07:22Speaker 1: required by it. Ursa Major Three Unions. One is not
- 1:07:26Speaker 1: the only system haunting this classification boundary. It is the
- 1:07:30Speaker 1: most extreme example, but it belongs to a growing population
- 1:07:34Speaker 1: of ambiguous satellites that resist clean categorization. Understanding why it
- 1:07:39Speaker 1: matters requires seeing it in the context of these other
- 1:07:41Speaker 1: contested objects. Wilman 1, discovered in 2005 in the Sloan
- 1:07:47Speaker 1: Digital Sky Survey data, was among the first systems to
- 1:07:50Speaker 1: expose how messy the classification boundary could become. Located approximately
- 1:07:55Speaker 1: 38 kiloparsecs from the Sun, it was initially reported to
- 1:07:58Speaker 1: have a high velocity dispersion consistent with dark matter domination.
- 1:08:02Speaker 1: But its systemic velocity was dangerously close to the velocities
- 1:08:06Speaker 1: of Milky Way foreground stars at similar positions in the sky,
- 1:08:10Speaker 1: meaning that contamination from unrelated halo stars could easily corrupt
- 1:08:14Speaker 1: the membership list. Different research groups analyzed the same data
- 1:08:19Speaker 1: and reached opposite conclusions. Some found a resolved velocity dispersion,
- 1:08:24Speaker 1: indicating dark matter. Others after applying stricter membership criteria that
- 1:08:29Speaker 1: removed likely interlopers, found a dispersion consistent with no dark
- 1:08:33Speaker 1: matter at all. The debate simmered for years without resolution.
- 1:08:37Speaker 1: Wilman 1 eventually received a tentative galaxy classification based partly
- 1:08:41Speaker 1: on an apparent metallicity spread among its members, but even
- 1:08:45Speaker 1: this evidence was questioned as potentially arising from foreground contamination.
- 1:08:49Speaker 1: The object remains one of the most contentious in the field.
- 1:08:53Speaker 1: Tucana 3 a satellite discovered in the Dark Energy Survey,
- 1:08:57Speaker 1: presented a different kind of ambiguity. Its velocity dispersion was
- 1:09:01Speaker 1: consistent with dark matter, but its tidal tails, streams of
- 1:09:05Speaker 1: stars being stripped by the Milky Way's gravity, suggested it
- 1:09:08Speaker 1: was undergoing active disruption. Tidal disruption can inflate a system's
- 1:09:12Speaker 1: apparent velocity dispersion by injecting kinetic energy into the stellar motions.
- 1:09:18Speaker 1: mimicking the signature of a dark matter halo, where the
- 1:09:21Speaker 1: true source is external gravitational torque. Disentangling the internal dispersion
- 1:09:26Speaker 1: from tidal effects requires sophisticated modeling that depends on assumptions
- 1:09:31Speaker 1: about the system's orbit, its initial structure, and the Milky
- 1:09:35Speaker 1: Way's gravitational potential. The classification of Tucana 3 remains dependent
- 1:09:39Speaker 1: on which model one trusts. Delve 1, Eridanus 3, and
- 1:09:44Speaker 1: several other recently discovered ultra-faint compact satellites occupy similar positions
- 1:09:48Speaker 1: in parameter space. Each sits in the region where the
- 1:09:51Speaker 1: physical sizes, luminosities and velocity dispersions of the smallest dwarf
- 1:09:56Speaker 1: galaxies overlap with those of the faintest star clusters. Each
- 1:10:00Speaker 1: requires case-by-case analysis that is limited by small sample sizes,
- 1:10:05Speaker 1: binary contamination risks, and the fundamental difficulty of measuring precise
- 1:10:10Speaker 1: velocities and chemical abundances for a few dozen extremely faint
- 1:10:15Speaker 1: stars at distances of tens of kiloparsecs. These are not
- 1:10:18Speaker 1: isolated puzzles. They form a pattern. The boundary between galaxy
- 1:10:23Speaker 1: and star cluster is not a wall. It is a
- 1:10:25Speaker 1: fog bank. Objects enter it from both sides. Star clusters
- 1:10:30Speaker 1: that appear unusually massive for their luminosity, dwarf galaxies that
- 1:10:34Speaker 1: appear unusually faint for their implied dark matter content. Somewhere
- 1:10:38Speaker 1: inside this fog, The clean distinction between dark matter-dominated and
- 1:10:42Speaker 1: baryon-only systems dissolves into measurement uncertainty, binary contamination, tidal effects,
- 1:10:50Speaker 1: and the irreducible limitations of trying to characterise a gravitational
- 1:10:54Speaker 1: system using the motions of 8 or 12 or 20 stars.
- 1:10:58Speaker 1: Why does any of this matter beyond the specialised world
- 1:11:00Speaker 1: of stellar dynamics and satellite taxonomy? The answer extends far
- 1:11:04Speaker 1: beyond classification for its own sake. If Ursa Major 3
- 1:11:08Speaker 1: Unions 1 is a galaxy, if future measurements reveal that
- 1:11:12Speaker 1: it does sit inside a dark matter halo, despite the
- 1:11:15Speaker 1: current absence of kinematic and chemical evidence, the implications for
- 1:11:19Speaker 1: dark matter physics are profound. It would mean that dark
- 1:11:23Speaker 1: matter halos of substantial mass can produce galaxies with absurdly
- 1:11:27Speaker 1: small stellar populations, pushing the minimum luminosity for galaxy formation
- 1:11:32Speaker 1: to limits that severely constrain the particle nature of dark matter.
- 1:11:36Speaker 1: Warm dark matter models which suppress the formation of small
- 1:11:39Speaker 1: halos would face significant tension. Fuzzy dark matter models which
- 1:11:44Speaker 1: create diffuse halo cores incapable of surviving strong tidal fields
- 1:11:49Speaker 1: would struggle to explain how such a dense system persists
- 1:11:52Speaker 1: on such a punishing orbit. The measurement would sharpen the
- 1:11:55Speaker 1: case for cold dark matter at the smallest scales, precisely
- 1:11:59Speaker 1: where alternatives have traditionally claimed the most leverage. If, instead,
- 1:12:04Speaker 1: the system is confirmed as a star cluster, a different
- 1:12:07Speaker 1: set of implications follows. It would mean that the faintest
- 1:12:10Speaker 1: end of the Milky Way satellite luminosity function includes objects
- 1:12:14Speaker 1: that are not dark matter halos at all, but ancient,
- 1:12:17Speaker 1: stripped-down remnants of once-larger stellar systems. This would complicate the
- 1:12:22Speaker 1: census of dark matter subhalos and the use of satellite
- 1:12:25Speaker 1: counts to constrain the minimum halo mass for galaxy formation.
- 1:12:29Speaker 1: It would mean that some fraction of the ultra-faint satellites
- 1:12:32Speaker 1: discovered in current and future surveys are not galaxies and
- 1:12:36Speaker 1: cannot be used as dark matter laboratories. The field would
- 1:12:39Speaker 1: need more sophisticated methods to separate genuine ultra-faint dwarfs from
- 1:12:43Speaker 1: imposter clusters masquerading at similar luminosities. And there is a
- 1:12:48Speaker 1: deeper practical consequence. Astronomers have used ultra-faint dwarf galaxies as
- 1:12:53Speaker 1: targets for indirect dark matter detection experiments. Gamma-ray telescopes like
- 1:12:58Speaker 1: the Fermi Large Area Telescope have pointed at systems including
- 1:13:01Speaker 1: Segway 1, Draco 2, and others, searching for the faint
- 1:13:06Speaker 1: glow of gamma rays that would be produced if dark
- 1:13:08Speaker 1: matter particles annihilate at the dense centers of dwarf galaxy halos.
- 1:13:13Speaker 1: The expected signal strength depends critically on the dark matter
- 1:13:16Speaker 1: density profile, which is inferred from the velocity dispersion. If
- 1:13:21Speaker 1: the velocity dispersion is contaminated by binaries, inflated by tidal effects,
- 1:13:26Speaker 1: or simply measured from too few stars, the inferred dark
- 1:13:30Speaker 1: matter density is wrong, and the predicted gamma ray signal
- 1:13:33Speaker 1: is wrong. Every dark matter search that uses an ultra-faint
- 1:13:37Speaker 1: dwarf as a target relies on the assumption that the
- 1:13:40Speaker 1: object is correctly classified as a galaxy, and that its
- 1:13:43Speaker 1: kinematic properties accurately reflect its dark matter content. Ursa Major 3. Unions.
- 1:13:50Speaker 1: One demonstrates how tenuous those assumptions can be, Cerny and
- 1:13:53Speaker 1: collaborators proposed a pragmatic methodological standard for the field going forward.
- 1:13:59Speaker 1: They argued that ultra-faint, compact Milky Way satellites, with strong
- 1:14:04Speaker 1: upper limits on both velocity dispersion and metallicity dispersion systems,
- 1:14:09Speaker 1: for which no positive evidence of dark matter or chemical
- 1:14:13Speaker 1: self-enrichment exists, should be regarded as star clusters until clear
- 1:14:17Speaker 1: positive evidence proves otherwise. The burden of proof, they suggested,
- 1:14:22Speaker 1: should fall on the galaxy hypothesis, not on the cluster hypothesis.
- 1:14:26Speaker 1: In a regime where measurement uncertainties are large, sample sizes
- 1:14:31Speaker 1: are small, and systematic effects like binary contamination can mimic
- 1:14:36Speaker 1: dark matter signatures, the conservative default should be the simpler explanation.
- 1:14:41Speaker 1: Stars held together by their own gravity, perhaps supplemented by
- 1:14:45Speaker 1: invisible stellar remnants, without invoking a dark matter halo, This
- 1:14:50Speaker 1: methodological proposal carries weight because it addresses a broader cultural
- 1:14:54Speaker 1: tendency in the field. Ultra-faint dwarf galaxies are scientifically exciting.
- 1:14:59Speaker 1: They are dark matter laboratories. They are fossil relics of
- 1:15:03Speaker 1: the earliest epoch of galaxy formation. They attract funding, telescope
- 1:15:07Speaker 1: time and theoretical attention. A new ultra-faint dwarf galaxy candidate
- 1:15:12Speaker 1: generates more interest and more follow-up than a new faint
- 1:15:15Speaker 1: star cluster. There is an understandable, if largely unconscious, incentive
- 1:15:19Speaker 1: to interpret ambiguous objects as galaxies rather than clusters. Cerny's
- 1:15:23Speaker 1: proposal pushes back against this tendency, by demanding that the
- 1:15:27Speaker 1: evidence meet a clear, affirmative threshold before the galaxy label
- 1:15:31Speaker 1: is applied. But not everyone in the community agrees. Some
- 1:15:34Speaker 1: researchers argue that the tidal survival problem is sufficiently compelling
- 1:15:38Speaker 1: on its own to favour the galaxy interpretation, even in
- 1:15:41Speaker 1: the absence of resolved velocity dispersion or metallicity spread. If
- 1:15:45Speaker 1: a self-gravitating cluster cannot survive the Milky Way's tidal field
- 1:15:49Speaker 1: for billions of years on this orbit, and if the
- 1:15:52Speaker 1: dark star cluster model requires specific assumptions about black hole
- 1:15:56Speaker 1: retention that remain observationally unverified, then perhaps the simplest explanation
- 1:16:01Speaker 1: is still dark matter, even if the kinematic signal is
- 1:16:04Speaker 1: too subtle to detect with current instruments. The absence of evidence,
- 1:16:09Speaker 1: in this view, is not evidence of absence. A dark
- 1:16:13Speaker 1: matter halo producing a velocity dispersion of 0.5 km per
- 1:16:17Speaker 1: second would be virtually impossible to detect with existing spectroscopic
- 1:16:21Speaker 1: precision in a system of 12 stars. The signal could
- 1:16:24Speaker 1: be real and simply below the noise floor. This is
- 1:16:27Speaker 1: the crux of the dispute as it stands today. Both
- 1:16:30Speaker 1: sides are scientifically legitimate. Both acknowledge the same data. They
- 1:16:35Speaker 1: differ on how to interpret silence. When the measurements say
- 1:16:38Speaker 1: neither clearly yes nor clearly no, does the default classification matter?
- 1:16:43Speaker 1: Is the object a star cluster until proven otherwise, or
- 1:16:46Speaker 1: a galaxy candidate whose dark matter content simply eludes current
- 1:16:50Speaker 1: measurement precision? The answer depends partly on philosophy of science,
- 1:16:54Speaker 1: and partly on what new data the next generation of
- 1:16:57Speaker 1: observations will deliver. And that next generation is coming soon.
- 1:17:01Speaker 1: What has become clear is that Ursa Major three unions,
- 1:17:05Speaker 1: one is not merely a classification problem to be solved
- 1:17:08Speaker 1: and filed away, It is a signpost marking the frontier
- 1:17:11Speaker 1: where astronomy's categories begin to fail. It sits in a
- 1:17:15Speaker 1: region of parameter space where no confirmed galaxy has ever
- 1:17:19Speaker 1: been found. Whether velocity dispersion is this low, the metallicity
- 1:17:23Speaker 1: dispersion is this small, and the stellar mass is this negligible,
- 1:17:27Speaker 1: whether it turns out to be the darkest galaxy in
- 1:17:30Speaker 1: the universe or the faintest star cluster ever seen, the
- 1:17:33Speaker 1: object has already accomplished something that neither answer can undo.
- 1:17:37Speaker 1: It has forced the question of what a galaxy is
- 1:17:39Speaker 1: into the sharpest focus the field has ever known, and
- 1:17:42Speaker 1: it has revealed that at the faintest extremes, the universe
- 1:17:46Speaker 1: offers evidence that is quieter than our instruments can confidently hear.
- 1:17:49Speaker 1: The deeper question is no longer just about this one system.
- 1:17:53Speaker 1: It is about the boundary itself, what it means, where
- 1:17:56Speaker 1: it sits, and whether it was ever as firm as
- 1:17:58Speaker 1: astronomers believed. The word galaxy was never designed for what
- 1:18:02Speaker 1: astronomers are now asking it to hold. The tension is
- 1:18:05Speaker 1: not new. Every generation of astronomers has encountered objects that
- 1:18:09Speaker 1: did not fit the categories inherited from the previous generation.
- 1:18:12Speaker 1: But the current tension has a special character, because it
- 1:18:16Speaker 1: touches not just what we observe, but the invisible scaffolding
- 1:18:19Speaker 1: we believe supports everything we observe. Dark matter is the
- 1:18:22Speaker 1: foundation of modern cosmological structure. If the line between galaxy
- 1:18:27Speaker 1: and star cluster is ultimately a line between dark matter
- 1:18:29Speaker 1: dominated and dark matter free, then failing to draw it cleanly,
- 1:18:34Speaker 1: means failing to map the dark matter itself. It was
- 1:18:37Speaker 1: born in ancient Greece to describe the pale river of
- 1:18:40Speaker 1: light arching across the night sky. It matured in the
- 1:18:43Speaker 1: 20th century to describe vast island universes containing billions of stars.
- 1:18:49Speaker 1: It expanded to include dwarfs, irregulars, ellipticals and every luminous
- 1:18:55Speaker 1: structure that carried its own gravitational identity across the cosmic landscape.
- 1:19:00Speaker 1: At each stage the words stretched to accommodate new discoveries.
- 1:19:04Speaker 1: At each stage the core intuition held A galaxy is big,
- 1:19:08Speaker 1: luminous and populated with stars on a scale that dwarfs
- 1:19:12Speaker 1: any other single astronomical object. Ursa Major. Three unions. One
- 1:19:18Speaker 1: sits at the point where that intuition either bends one
- 1:19:21Speaker 1: final time or finally snaps. Sixty stars. Sixteen solar masses.
- 1:19:27Speaker 1: Eleven solar luminosities. A volume of space ten light years across.
- 1:19:32Speaker 1: If this is a galaxy, then the word has travelled
- 1:19:35Speaker 1: so far from its origin that it now describes something
- 1:19:38Speaker 1: that could fit inside the distance between our Sun and
- 1:19:41Speaker 1: its nearest stellar neighbour, and still have room to spare.
- 1:19:45Speaker 1: If this is not a galaxy, then the word has
- 1:19:47Speaker 1: held its boundary, but the object that tested it has
- 1:19:50Speaker 1: revealed how arbitrary that boundary always was. To understand why
- 1:19:54Speaker 1: this matters, we need to step back from the classification
- 1:19:57Speaker 1: debate and look at what the smallest galaxies actually teach
- 1:20:00Speaker 1: us not about themselves, but about the universe's deepest architectural secrets.
- 1:20:06Speaker 1: because the strange truth is that the tiniest galaxies are
- 1:20:09Speaker 1: more scientifically valuable than the largest ones for answering the
- 1:20:13Speaker 1: questions that matter most. The Milky Way contains roughly 400
- 1:20:17Speaker 1: billion stars, vast quantities of gas and dust, a supermassive
- 1:20:22Speaker 1: black hole at its centre, and a dark matter halo
- 1:20:25Speaker 1: containing approximately 10 times the mass of all its visible
- 1:20:29Speaker 1: components combined. Separating dark matter's gravitational effects from those of
- 1:20:33Speaker 1: ordinary matter in such a complex system is extraordinarily difficult. Stars, gas, dust,
- 1:20:40Speaker 1: stellar remnants, molecular clouds, all of them contribute to the
- 1:20:43Speaker 1: gravitational budget. Disentangling the dark matter signal from the baryonic
- 1:20:48Speaker 1: noise requires sophisticated modeling that carries its own assumptions and uncertainties.
- 1:20:53Speaker 1: The dark matter is there, its effects are measurable, but
- 1:20:57Speaker 1: it shares the stage with a dense, complex, interacting population
- 1:21:01Speaker 1: of visible matter that obscures the details. Now consider an
- 1:21:05Speaker 1: ultra-faint dwarf galaxy, a system where the total stellar mass
- 1:21:09Speaker 1: is a few hundred or a few thousand solar masses,
- 1:21:12Speaker 1: where the gas has been stripped or exhausted billions of
- 1:21:14Speaker 1: years ago, where there is no dust, no active star formation,
- 1:21:19Speaker 1: no supermassive black hole, no molecular clouds, the visible matter
- 1:21:23Speaker 1: is negligible. The baryonic contribution to the gravitational potential is
- 1:21:27Speaker 1: a rounding error. What remains is almost pure dark matter.
- 1:21:31Speaker 1: The stars inside such a system are not important for
- 1:21:34Speaker 1: their own sake, they are important because they move, and
- 1:21:37Speaker 1: their motions trace the gravitational field of the dark matter
- 1:21:41Speaker 1: halo they inhabit with a directness that no measurement in
- 1:21:44Speaker 1: a massive galaxy can match. Ultra-faint dwarfs are the cleanest
- 1:21:48Speaker 1: dark matter laboratories in the observable universe. They are systems
- 1:21:52Speaker 1: where the dark matter signal is not buried under layers
- 1:21:55Speaker 1: of baryonic complexity, it is exposed. The velocity dispersion of
- 1:22:00Speaker 1: a few dozen ancient stars directly probes the depth and
- 1:22:03Speaker 1: shape of the dark matter potential well. The spatial distribution
- 1:22:07Speaker 1: of those stars reveals the density profile of the halo's
- 1:22:10Speaker 1: inner regions. The survival of the system against tidal disruption
- 1:22:14Speaker 1: constrains the halo's overall mass and concentration. Each measurement, however uncertain,
- 1:22:20Speaker 1: provides a constraint on the nature of dark matter itself,
- 1:22:24Speaker 1: on the mass of the dark matter particle, on whether
- 1:22:27Speaker 1: dark matter interacts with itself, on whether dark matter halos
- 1:22:30Speaker 1: develop dense cusps or flat cores at their centers. This
- 1:22:34Speaker 1: is why the classification of an object like Ursa Major
- 1:22:37Speaker 1: Three Unions One carries stakes far beyond astronomical bookkeeping. If
- 1:22:42Speaker 1: it is a galaxy, its dark matter halo is among
- 1:22:44Speaker 1: the densest and most extreme ever inferred. Its inferred central density,
- 1:22:50Speaker 1: if confirmed, would challenge models of fuzzy dark matter, which
- 1:22:54Speaker 1: predict that halos below a certain mass cannot form concentrated
- 1:22:58Speaker 1: central cusps, because the quantum wave nature of extremely light
- 1:23:02Speaker 1: dark matter particles smooths out structure on small scales. Reconciling
- 1:23:08Speaker 1: fuzzy dark matter with a halo dense enough to survive
- 1:23:10Speaker 1: the Milky Way's tidal field on the tight orbit of
- 1:23:13Speaker 1: Ursa Major 3 unions, one would require significant modifications to
- 1:23:18Speaker 1: standard fuzzy dark matter models, or perhaps rule them out
- 1:23:22Speaker 1: entirely at certain particle masses. Similarly, Warm dark matter, a
- 1:23:27Speaker 1: model in which dark matter particles have residual thermal velocities
- 1:23:30Speaker 1: from the early universe that suppress the formation of small
- 1:23:34Speaker 1: scale structure, predicts fewer low mass halos than cold dark. Matter,
- 1:23:39Speaker 1: if the galaxy interpretation holds, it strengthens the case that
- 1:23:43Speaker 1: dark matter is cold and that low mass halos form abundantly,
- 1:23:47Speaker 1: exactly as the standard lambda cold dark matter model predicts.
- 1:23:51Speaker 1: Self-interacting dark matter presents yet another set of constraints. In
- 1:23:56Speaker 1: models where dark matter particles scatter off each other at
- 1:23:58Speaker 1: rates significant enough to redistribute energy within halos, the central
- 1:24:03Speaker 1: regions of small halos are expected to develop constant density
- 1:24:06Speaker 1: cores rather than the steep density cusps predicted by cold
- 1:24:10Speaker 1: dark matter. A chord halo at the mass scale inferred
- 1:24:13Speaker 1: for Ursa Major 3 would be more vulnerable to tidal
- 1:24:16Speaker 1: stripping than a cuspy one. If the system has survived
- 1:24:19Speaker 1: for billions of years on its current orbit, its halo
- 1:24:22Speaker 1: must be dense enough at the centre to resist disruption,
- 1:24:26Speaker 1: a condition more naturally satisfied by cuspy cold dark matter
- 1:24:29Speaker 1: profiles than by the chord profiles of self-interacting models. The
- 1:24:34Speaker 1: survival of this system, if it is a galaxy, becomes
- 1:24:37Speaker 1: a test of dark matter microphysics. Each of these constraints
- 1:24:41Speaker 1: sharpens as the system in question becomes smaller and fainter.
- 1:24:45Speaker 1: In massive dwarf galaxies containing millions of stars, the constraints
- 1:24:49Speaker 1: are valuable but diluted by baryonic complexity and measurement averaging.
- 1:24:53Speaker 1: In ultra-faint dwarfs like Segway 1, The constraints are cleaner
- 1:24:57Speaker 1: but still rest on samples of a few dozen stars.
- 1:25:01Speaker 1: At the extreme represented by Ursa Major 3 Unions 1,
- 1:25:05Speaker 1: the constraint tightens to a razor's edge but so does
- 1:25:07Speaker 1: the risk that the entire measurement is wrong, contaminated or misinterpreted.
- 1:25:12Speaker 1: The power and the fragility are two sides of the
- 1:25:15Speaker 1: same coin. This tension connects directly to one of the
- 1:25:18Speaker 1: oldest unsolved problems in cosmology, the missing satellites problem. Lambda
- 1:25:24Speaker 1: cold dark matter simulations predict that a galaxy the size
- 1:25:28Speaker 1: of the Milky Way should be surrounded by thousands of
- 1:25:32Speaker 1: dark matter subhalos, small, gravitationally bound clumps of dark matter
- 1:25:37Speaker 1: orbiting within the larger host halo. Many of these subhalos
- 1:25:40Speaker 1: should be massive enough to host stars, forming satellite galaxies
- 1:25:44Speaker 1: detectable by current surveys, but the observed count of Milky
- 1:25:48Speaker 1: Way satellites falls far short of the theoretical prediction. As
- 1:25:52Speaker 1: of the most recent census, Approximately 60 satellites have been confirmed.
- 1:25:57Speaker 1: Simulations predict hundreds to thousands. Where are the rest? Several
- 1:26:01Speaker 1: explanations have been proposed, and each carries different implications for
- 1:26:05Speaker 1: fundamental physics. The simplest resolution invokes observational incompleteness. Current surveys
- 1:26:12Speaker 1: have not covered the entire sky to uniform depth. The
- 1:26:15Speaker 1: Sloan Digital Sky Survey mapped roughly one third of the sky.
- 1:26:19Speaker 1: The Dark Energy Survey covered a different region in the
- 1:26:21Speaker 1: southern hemisphere. Large portions of the sky, particularly those behind
- 1:26:26Speaker 1: the dense stellar fields of the Milky Way's disk, remain
- 1:26:30Speaker 1: poorly explored. Ultra-faint satellites at distances greater than roughly 50
- 1:26:35Speaker 1: kiloparsecs are beyond the detection limits of most existing surveys.
- 1:26:40Speaker 1: The missing satellites may simply be there, waiting for deeper
- 1:26:44Speaker 1: and wider surveys to find them. This explanation is strengthened
- 1:26:47Speaker 1: enormously by the existence of objects like Ursa Major 3
- 1:26:51Speaker 1: Unions 1, If a system with only 11 solar luminosities
- 1:26:55Speaker 1: can exist as a bound satellite of the Milky Way,
- 1:26:58Speaker 1: then the detection threshold for the faintest satellites is far
- 1:27:01Speaker 1: higher than previously appreciated. A system this faint would be
- 1:27:05Speaker 1: completely undetectable beyond roughly 20 kiloparsecs with current instrumentation. If
- 1:27:10Speaker 1: it were located at the typical distance of known ultra-faint
- 1:27:13Speaker 1: dwarfs 50 to 100 kiloparsecs, no survey on Earth could
- 1:27:18Speaker 1: find it. The Milky Way's halo could be teeming with
- 1:27:20Speaker 1: similar objects. each containing a handful of ancient stars inside
- 1:27:24Speaker 1: a dark matter halo of substantial mass, all hiding below
- 1:27:28Speaker 1: the detection floor. An alternative explanation for the missing satellites
- 1:27:32Speaker 1: invokes the physics of galaxy formation itself. Below a critical
- 1:27:36Speaker 1: dark matter halo mass, the gravitational potential well may be
- 1:27:40Speaker 1: too shallow to capture and cool enough gas to form stars.
- 1:27:44Speaker 1: A specific event in the early universe, the epoch of reionization,
- 1:27:48Speaker 1: established a mass threshold that permanently divided dark matter halos
- 1:27:52Speaker 1: into two populations. Those that had already collected enough gas
- 1:27:56Speaker 1: to form at least some stars, and those that would
- 1:27:59Speaker 1: remain starved and dark forever. The physics behind this threshold
- 1:28:03Speaker 1: deserves closer attention, because it connects the existence of an
- 1:28:07Speaker 1: object like Ursa Major 3 Unions 1 to conditions that
- 1:28:11Speaker 1: prevailed across the entire universe during its first billion years.
- 1:28:16Speaker 1: In the earliest epochs after the Big Bang, the universe
- 1:28:19Speaker 1: was filled with neutral hydrogen and helium gas. Dark matter
- 1:28:23Speaker 1: had already begun collapsing into halos gravitationally bound concentrations that
- 1:28:27Speaker 1: grew by accreting surrounding material. The gas followed the dark matter,
- 1:28:32Speaker 1: falling into the gravitational wells of these halos and accumulating
- 1:28:35Speaker 1: at their centres. In halos above a critical mass, the
- 1:28:39Speaker 1: gas could cool efficiently through atomic hydrogen radiation, losing thermal
- 1:28:43Speaker 1: energy and condensing into dense clouds that fragmented and formed stars.
- 1:28:48Speaker 1: The first generation of stars ignited inside these halos, perhaps
- 1:28:53Speaker 1: as early as 100 to 200 million years after the
- 1:28:56Speaker 1: Big Bang. But star formation came with consequences. the first
- 1:29:00Speaker 1: stars were massive, luminous and short-lived. They poured ultraviolet radiation
- 1:29:05Speaker 1: into the surrounding gas, ionising the neutral hydrogen atoms and
- 1:29:09Speaker 1: heating the intergalactic medium to temperatures of roughly 10,000 to 20,000 degrees.
- 1:29:15Speaker 1: This process, called reionisation, unfolded over several hundred million years
- 1:29:21Speaker 1: and was largely complete by the time the universe reached
- 1:29:24Speaker 1: an age of approximately one billion years. corresponding to a
- 1:29:27Speaker 1: cosmological redshift of roughly six. By the end of reionization,
- 1:29:32Speaker 1: the universe had transitioned from a neutral, opaque state to
- 1:29:35Speaker 1: an ionized, transparent one. The cosmic microwave background radiation traveled
- 1:29:41Speaker 1: freely through this newly transparent medium, eventually reaching our telescopes
- 1:29:45Speaker 1: as the faint afterglow of the Big Bang. Reionization transformed
- 1:29:49Speaker 1: the gas dynamics of the universe at every scale. Before reionization,
- 1:29:54Speaker 1: neutral gas could fall freely into dark matter halos of
- 1:29:58Speaker 1: any mass. After reionization, the heated gas had too much
- 1:30:02Speaker 1: thermal energy to be captured by the shallowest gravitational wells.
- 1:30:06Speaker 1: A dark matter halo needs a circular velocity of at
- 1:30:08Speaker 1: least roughly 15 to 20 kilometers per second, corresponding to
- 1:30:12Speaker 1: a virial mass of roughly 100 million to 1 billion
- 1:30:16Speaker 1: solar masses, depending on redshift and environment to gravitationally overcome
- 1:30:20Speaker 1: the thermal pressure of the heated intergalactic gas and pull
- 1:30:23Speaker 1: it inward. Below this threshold, gas accretion effectively ceases. The
- 1:30:28Speaker 1: halos still exist. They continue orbiting within the host galaxy's halo,
- 1:30:33Speaker 1: but they are starved of the raw material needed to
- 1:30:35Speaker 1: form stars. They become permanently dark. This is why the
- 1:30:39Speaker 1: stars in ultra-faint dwarf galaxies are almost universally ancient. The
- 1:30:44Speaker 1: galaxies formed their stars before reionization, during the brief window
- 1:30:48Speaker 1: when neutral gas was still available to small halos. Once
- 1:30:52Speaker 1: reionization swept through their region of the universe, their gas
- 1:30:56Speaker 1: supply was cut off and star formation stopped. The stars
- 1:30:59Speaker 1: we observe today in systems like Segway 1 and Boots
- 1:31:02Speaker 1: 1 were born in that narrow pre-reionization window more than
- 1:31:07Speaker 1: 12 billion years ago and have been passively aging ever since.
- 1:31:11Speaker 1: These ancient populations predate most of the stellar mass in
- 1:31:14Speaker 1: the modern universe, and their extreme chemical poverty marks them
- 1:31:18Speaker 1: as products of the earliest star-forming halos survivors from an
- 1:31:21Speaker 1: era that ended before the universe reached even a tenth
- 1:31:24Speaker 1: of its current age. The stars in Ursa Major 3,
- 1:31:27Speaker 1: Unions 1, are consistent with this picture. Their estimated age
- 1:31:32Speaker 1: exceeds 10 billion years, and their extremely low metallicity, an
- 1:31:36Speaker 1: iron to hydrogen ratio of approximately negative 2.2, indicates that
- 1:31:41Speaker 1: they formed from gas that had been enriched by, at most,
- 1:31:45Speaker 1: a tiny number of prior supernovae. they carry the chemical
- 1:31:48Speaker 1: fingerprints of a single brief burst of star formation in
- 1:31:52Speaker 1: an environment where heavy elements were exceedingly scarce. If the
- 1:31:56Speaker 1: system is a galaxy, its stars represent one of the
- 1:31:59Speaker 1: most marginal star formation events the universe has ever produced.
- 1:32:03Speaker 1: A dark matter halo that barely crossed the mass threshold,
- 1:32:06Speaker 1: captured just enough gas, formed just enough stars to leave
- 1:32:10Speaker 1: a visible trace, and then fell silent forever as reionization
- 1:32:15Speaker 1: heated the surrounding gas beyond its gravitational reach. This interpretation
- 1:32:20Speaker 1: places the system at the precise intersection of two of
- 1:32:23Speaker 1: the most important thresholds in cosmology, the minimum dark matter
- 1:32:27Speaker 1: halo mass for galaxy formation and the epoch of reionization
- 1:32:32Speaker 1: that established that minimum. Understanding where these thresholds lie is
- 1:32:36Speaker 1: essential for predicting how many luminous satellites the Milky Way
- 1:32:39Speaker 1: should possess and for constraining the properties of dark matter
- 1:32:42Speaker 1: particles through the abundance of the smallest galactic. Structures. If
- 1:32:47Speaker 1: this picture is correct, the missing satellites are not missing
- 1:32:50Speaker 1: at all. They are dark. They exist as predicted by
- 1:32:54Speaker 1: lambda-cold dark matter, but they never formed stars. The observable
- 1:32:59Speaker 1: satellite population represents only the tip of a vast substructure
- 1:33:03Speaker 1: of invisible dark matter clumps, each one a potential galaxy
- 1:33:07Speaker 1: that never was. Ursa Major Three Unions One, if it
- 1:33:11Speaker 1: is a galaxy. sits precisely at the boundary where this
- 1:33:15Speaker 1: dark to luminous transition occurs. A dark matter halo of
- 1:33:19Speaker 1: roughly 1 billion solar masses, containing only 16 solar masses
- 1:33:23Speaker 1: of stars, would represent the most extreme known example of
- 1:33:26Speaker 1: the galaxy formation process, operating just barely above the threshold
- 1:33:31Speaker 1: of viability. It would be a system that almost did
- 1:33:34Speaker 1: not form stars at all. A galaxy that exists only
- 1:33:37Speaker 1: because its halo was just massive enough, just dense enough,
- 1:33:41Speaker 1: at just the right moment in cosmic history, to collect
- 1:33:44Speaker 1: and cool a tiny wisp of primordial gas before reionization
- 1:33:49Speaker 1: shut the process down forever. Its handful of ancient, metal-poor
- 1:33:53Speaker 1: stars would be relics of that single, brief episode of
- 1:33:57Speaker 1: star formation, frozen in time for more than 10 billion years,
- 1:34:02Speaker 1: drifting through the Milky Way's halo as the oldest and
- 1:34:05Speaker 1: faintest witnesses to the moment when dark matter first learned
- 1:34:08Speaker 1: to make light. If, on the other hand, the system
- 1:34:11Speaker 1: is a star cluster, the implications are different but no
- 1:34:14Speaker 1: less important. It would mean that the luminosity function of
- 1:34:17Speaker 1: Milky Way satellites, the statistical distribution of satellites as a
- 1:34:21Speaker 1: function of brightness, reaches a bottom defined not by dark
- 1:34:25Speaker 1: matter halo physics, but by the survival limits of bare
- 1:34:29Speaker 1: stellar systems in the Milky Way's tidal environment. It would
- 1:34:32Speaker 1: mean that below a certain luminosity, the satellite population transitions
- 1:34:37Speaker 1: from dark matter-dominated galaxies to self-gravitating stellar remnants that can
- 1:34:42Speaker 1: mimic the appearance of galaxies without containing any dark matter
- 1:34:45Speaker 1: at all. Separating these two populations would become a critical
- 1:34:49Speaker 1: task for every future survey, and the tools for doing
- 1:34:53Speaker 1: so would need to be sharpened considerably beyond their current state.
- 1:34:56Speaker 1: Those future surveys are approaching rapidly. The VERA-C, Rubin Observatory,
- 1:35:02Speaker 1: currently being commissioned in Chile, will conduct the legacy survey
- 1:35:05Speaker 1: of space and time. a 10-year campaign to repeatedly photograph
- 1:35:09Speaker 1: the entire southern sky to unprecedented depth. The survey will
- 1:35:14Speaker 1: detect individual stars at distances far exceeding the reach of
- 1:35:17Speaker 1: any previous wide-field imaging program. Simulations predict that the Legacy
- 1:35:22Speaker 1: Survey of Space and Time will discover dozens to hundreds
- 1:35:25Speaker 1: of new ultra-faint satellite candidates around the Milky Way, pushing
- 1:35:30Speaker 1: into luminosity regimes where objects like Ursa Major 3 Unions
- 1:35:34Speaker 1: 1 may turn out to be common rather than exceptional.
- 1:35:38Speaker 1: Each discovery will require spectroscopic follow-up to determine velocities, chemical compositions,
- 1:35:44Speaker 1: and ultimately classifications. The community will face a flood of
- 1:35:48Speaker 1: new ambiguous objects, each one demanding the same painstaking case-by-case
- 1:35:53Speaker 1: analysis that has consumed years for the handful of systems
- 1:35:56Speaker 1: currently known. The Euclid Space Telescope, launched by the European
- 1:36:01Speaker 1: Space Agency in 2023, will contribute from a different angle,
- 1:36:06Speaker 1: Its wide-field infrared imaging will map the distribution of dark
- 1:36:09Speaker 1: matter across vast cosmic volumes through gravitational lensing, the bending
- 1:36:14Speaker 1: of light from distant galaxies by foreground mass concentrations. While
- 1:36:18Speaker 1: Euclid is not primarily designed to discover individual Milky Way satellites,
- 1:36:23Speaker 1: its data will constrain the overall abundance and mass distribution
- 1:36:27Speaker 1: of dark matter subhalos at scales relevant to the missing
- 1:36:30Speaker 1: satellites problem providing a statistical framework against which the observed
- 1:36:34Speaker 1: satellite population can be compared. The numbers from simulation-based predictions
- 1:36:38Speaker 1: give a sense of the transformation that is approaching. Current
- 1:36:42Speaker 1: estimates suggest that the legacy survey of space and time,
- 1:36:45Speaker 1: with its combination of depth and sky coverage, will be
- 1:36:49Speaker 1: sensitive to ultra-faint satellites down to absolute magnitudes of roughly
- 1:36:53Speaker 1: zero out to distances of approximately 30 to 40 kiloparsecs.
- 1:36:58Speaker 1: An object as faint as Ursa Major 3 Unions 1
- 1:37:01Speaker 1: with its absolute magnitude near positive 2, would need to
- 1:37:04Speaker 1: be closer than roughly 20 kiloparsecs to be detected even
- 1:37:08Speaker 1: by Rubin. Beyond that distance, systems at this extreme faintness
- 1:37:12Speaker 1: will remain below the detection threshold even for Rubin, but
- 1:37:16Speaker 1: within that volume the survey is expected to discover a
- 1:37:19Speaker 1: substantial new population. Simulations calibrated to the current satellite luminosity
- 1:37:24Speaker 1: function and corrected for the sky coverage and depth limitations
- 1:37:28Speaker 1: of existing surveys predict that the Milky Way may host
- 1:37:31Speaker 1: between 200 and 1,000 luminous satellites. In total, depending on
- 1:37:37Speaker 1: assumptions about the faint end of the luminosity function and
- 1:37:40Speaker 1: the efficiency of star formation in the smallest halos, if
- 1:37:44Speaker 1: the true number is toward the higher end, the Legacy
- 1:37:47Speaker 1: Survey of Space and Time could discover more new ultra-faint
- 1:37:50Speaker 1: satellites in its first few years of operation than all
- 1:37:54Speaker 1: previous surveys combined over the past two decades. But discovery
- 1:37:58Speaker 1: is only the first step. Each new candidate will require
- 1:38:01Speaker 1: spectroscopic follow-up to measure radial velocities, identify member stars, determine
- 1:38:06Speaker 1: velocity dispersions, and assess metallicities. This is where the bottleneck emerges.
- 1:38:13Speaker 1: Spectroscopy of extremely faint stars, magnitudes of 22, 23, 24,
- 1:38:20Speaker 1: and beyond, requires large aperture telescopes and hours of integration
- 1:38:24Speaker 1: time per target. The Keck telescopes, the Magellan telescopes, the
- 1:38:29Speaker 1: very large Telescope, and eventually the Extremely Large Telescope and
- 1:38:33Speaker 1: the 30-meter telescope will share this burden. But telescope time
- 1:38:37Speaker 1: is finite, and each ultra-faint satellite candidate requires not just
- 1:38:41Speaker 1: one epoch of spectroscopy, but multiple epochs to identify and
- 1:38:45Speaker 1: remove binary contaminations. A single system can consume tens of
- 1:38:50Speaker 1: hours of 10-metre-class telescope time before a confident classification emerges.
- 1:38:55Speaker 1: Multiply this by dozens or hundreds of new discoveries, and
- 1:38:58Speaker 1: the spectroscopic follow-up campaign becomes one of the largest observational
- 1:39:02Speaker 1: undertakings in stellar astronomy. The challenge is compounded by the
- 1:39:06Speaker 1: lesson of Ursa Major Three Unions, one itself. Initial single-epoch
- 1:39:11Speaker 1: spectroscopy produced a velocity dispersion that pointed strongly toward a
- 1:39:15Speaker 1: galaxy classification. multi-epoch follow-up reversed the picture. Any future discovery
- 1:39:21Speaker 1: that receives only a single spectroscopic observation will carry the
- 1:39:25Speaker 1: same risk of binary contamination and premature classification. The community
- 1:39:31Speaker 1: will need to develop standardized protocols for multi-epoch kinematic monitoring
- 1:39:35Speaker 1: of ultra-faint satellite candidates, a systematic, years-long campaign of repeated
- 1:39:41Speaker 1: spectroscopic observations designed to flag binaries, stabilize dispersion measurements, and
- 1:39:47Speaker 1: build the sample sizes needed for statistically robust conclusions. Chemical
- 1:39:52Speaker 1: abundance analysis adds another layer. Measuring individual stellar metallicities from
- 1:39:57Speaker 1: low resolution spectra provides useful but limited information. The strongest
- 1:40:01Speaker 1: classification evidence comes from high resolution spectroscopy that can resolve
- 1:40:06Speaker 1: individual elemental abundances. Not just the overall iron content, but
- 1:40:10Speaker 1: the ratios of alpha elements like magnesium, calcium, and titanium
- 1:40:15Speaker 1: to iron which trace the timescale and enrichment history of
- 1:40:19Speaker 1: star formation. These measurements require even more telescope time and
- 1:40:23Speaker 1: are feasible only for the brightest member stars in each system.
- 1:40:27Speaker 1: For a system like Ursa Major 3, Unions 1, where
- 1:40:30Speaker 1: the brightest members hover at faint magnitudes, high resolution abundance
- 1:40:35Speaker 1: analysis pushes current instrumentation to its limits. The next decade
- 1:40:40Speaker 1: of ultra-faint satellite research will be defined not only by
- 1:40:43Speaker 1: the pace of photometric discovery, but by the capacity of
- 1:40:47Speaker 1: the spectroscopic infrastructure to keep up. The flood of candidates
- 1:40:51Speaker 1: from the Legacy Survey of Space and Time will outstrip
- 1:40:54Speaker 1: the available telescope time for detailed follow-up, forcing the community
- 1:40:58Speaker 1: to prioritise targets, develop faster classification diagnostics, and accept that
- 1:41:04Speaker 1: some fraction of newly discovered systems will remain in the
- 1:41:07Speaker 1: ambiguous grey zone for years before their nature is resolved. Together,
- 1:41:12Speaker 1: These observatories will transform the landscape of ultra-faint satellite research
- 1:41:16Speaker 1: within the next decade. The question of how many objects
- 1:41:19Speaker 1: like Ursa Major, three unions, one, exist around the Milky Way,
- 1:41:24Speaker 1: and how many of them are galaxies versus star clusters,
- 1:41:28Speaker 1: will move from theoretical speculation to observational constraint. But the
- 1:41:32Speaker 1: transformation will also intensify the classification problem. More objects in
- 1:41:36Speaker 1: the grey zone means more disputed cases, more ambiguous measurements.
- 1:41:41Speaker 1: more systems where the evidence stubbornly refuses to point clearly
- 1:41:45Speaker 1: in either direction. And this is where the deeper question
- 1:41:48Speaker 1: finally surfaces. Not whether Ursa Major 3 Unions 1 is
- 1:41:52Speaker 1: a galaxy or a star cluster, but whether the distinction
- 1:41:55Speaker 1: itself remains coherent at these extremes. And this is played,
- 1:42:00Speaker 1: the Wilman and Strader definition, a galaxy is a system
- 1:42:04Speaker 1: whose properties require more than baryons and Newtonian gravity to explain,
- 1:42:09Speaker 1: is operationally clean, and scientifically useful for the vast majority
- 1:42:13Speaker 1: of astronomical objects. Apply it to the Milky Way, obviously
- 1:42:18Speaker 1: a galaxy. Apply it to Omega Centauri, the most massive
- 1:42:22Speaker 1: globular cluster in the Milky Way, widely regarded as a
- 1:42:26Speaker 1: probable stripped dwarf galaxy nucleus based on its multiple stellar
- 1:42:30Speaker 1: populations and chemical complexity. Apply it to Segway 1, a
- 1:42:35Speaker 1: galaxy supported by both kinematic and chemical evidence. Apply it
- 1:42:39Speaker 1: to a typical globular cluster like Messier 13, clearly a
- 1:42:43Speaker 1: star cluster with no dark matter required, but apply it
- 1:42:47Speaker 1: to Ursa Major 3 Unions 1, and the definition does
- 1:42:50Speaker 1: not fail, but it reaches a regime where applying it
- 1:42:54Speaker 1: becomes operationally impossible with current data. The measurements cannot distinguish
- 1:42:58Speaker 1: between a tiny dark matter halo producing a velocity dispersion
- 1:43:02Speaker 1: of half a kilometre per second, and no dark matter
- 1:43:05Speaker 1: at all. The chemical data cannot distinguish between a single
- 1:43:08Speaker 1: burst cluster population and an ultra-faint galaxy whose enrichment history
- 1:43:13Speaker 1: was so brief that its metallicity spread is indistinguishable from zero.
- 1:43:18Speaker 1: The diagnostic tools that make the definition work in brighter
- 1:43:21Speaker 1: systems lose their resolving power at this luminosity frontier. The
- 1:43:25Speaker 1: definition is correct in principle. It is unverifiable in practice.
- 1:43:29Speaker 1: This is not a failure of science. It is a
- 1:43:32Speaker 1: revelation about the universe. The universe does not draw sharp
- 1:43:36Speaker 1: boundaries between its structures, it builds gradients. Stars shade into
- 1:43:41Speaker 1: brown dwarfs, which shade into giant planets. Asteroids shade into
- 1:43:45Speaker 1: dwarf planets, which shade into planets. The question of where
- 1:43:49Speaker 1: to draw the line between these categories has fuelled debate
- 1:43:52Speaker 1: for centuries, a debate that famously stripped Pluto of its
- 1:43:56Speaker 1: planetary status in 2006. The galaxy versus star cluster boundary
- 1:44:01Speaker 1: is the same kind of problem. transplanted to the grandest
- 1:44:05Speaker 1: scales of cosmic structure, nature produces a continuous spectrum of
- 1:44:09Speaker 1: gravitationally bound stellar systems. At the high end, the systems
- 1:44:14Speaker 1: are unambiguously galaxies. At the low end, they are unambiguously clusters.
- 1:44:19Speaker 1: In the middle, in the fog, there are objects that
- 1:44:22Speaker 1: resist classification because the physical properties that distinguish galaxies from
- 1:44:27Speaker 1: clusters become unmeasurable. What makes the fog unsettling is what
- 1:44:31Speaker 1: it conceals. If the boundary between galaxy and star cluster
- 1:44:35Speaker 1: is a gradient rather than a wall, then the question
- 1:44:37Speaker 1: of how many galaxies orbit the Milky Way does not
- 1:44:40Speaker 1: have a clean answer. The number depends on where you
- 1:44:43Speaker 1: place the threshold, and the threshold depends on measurements that
- 1:44:46Speaker 1: are fundamentally limited by the faintness and sparsity of the
- 1:44:50Speaker 1: objects in question. The missing satellites problem, one of the
- 1:44:53Speaker 1: central puzzles of modern cosmology, may not have a single solution,
- 1:44:57Speaker 1: because the problem itself is partly definitional. Some of the
- 1:45:01Speaker 1: missing satellites may be dark matter halos with no stars
- 1:45:04Speaker 1: at all. Others may be systems like Ursa Major 3
- 1:45:07Speaker 1: Unions 1, hovering at the boundary, containing stars but defying
- 1:45:11Speaker 1: classification as either galaxy or cluster. The universe builds its
- 1:45:16Speaker 1: smallest structures in a regime where the conceptual machinery of
- 1:45:20Speaker 1: human science begins to grind against its own limits. We
- 1:45:24Speaker 1: want categories. Nature provides continua. We want clean distinctions. Nature
- 1:45:29Speaker 1: provides systems where the difference between galaxy and not-galaxy is
- 1:45:34Speaker 1: smaller than our measurement uncertainties. We want to count the
- 1:45:38Speaker 1: galaxies around the Milky Way. Nature builds objects that might
- 1:45:41Speaker 1: or might not qualify, depending on which spectrum we observe,
- 1:45:45Speaker 1: which binary we exclude, and which prior we impose on
- 1:45:49Speaker 1: our statistical model. Perhaps the most important contribution of Ursa
- 1:45:52Speaker 1: Major three unions, one to astronomy, is not the answer
- 1:45:56Speaker 1: it will eventually receive galaxy or star cluster, but the
- 1:45:59Speaker 1: question it has already forced the field to confront. What
- 1:46:03Speaker 1: does it mean for a category to break down? What
- 1:46:06Speaker 1: does it mean when the universe produces objects that fit
- 1:46:09Speaker 1: no existing label comfortably? And what does it mean that
- 1:46:12Speaker 1: the difference between the smallest galaxy and the largest star
- 1:46:15Speaker 1: cluster may hinge not on the nature of the objects themselves,
- 1:46:19Speaker 1: but on the precision of the instruments we point at them? Return,
- 1:46:23Speaker 1: one final time, to the object itself. Sixty ancient stars.
- 1:46:28Speaker 1: older than 10 billion years, drifting through the halo of
- 1:46:31Speaker 1: a galaxy containing 400 billion suns. Their combined light is
- 1:46:35Speaker 1: 11 times the luminosity of our own star. Their combined
- 1:46:39Speaker 1: mass is 16 times the mass of our sun. They
- 1:46:43Speaker 1: occupy a volume of space roughly 10 light years across,
- 1:46:46Speaker 1: a patch of the cosmos so small and so faint
- 1:46:49Speaker 1: that it eluded detection until a wide-field survey camera on
- 1:46:52Speaker 1: a mountaintop in Hawaii caught the faint statistical excess of
- 1:46:56Speaker 1: their collective presence against the crowded foreground of the Milky
- 1:47:00Speaker 1: Way's halo, they may be sitting inside an invisible mountain
- 1:47:04Speaker 1: of dark matter, a halo weighing roughly a billion solar masses,
- 1:47:08Speaker 1: a structure so much more massive than its visible contents
- 1:47:12Speaker 1: that the stars are barely more than decorative sparks at
- 1:47:15Speaker 1: the center of an enormous gravitational well. In that case,
- 1:47:19Speaker 1: they are the most extreme galaxy ever discovered, the faintest
- 1:47:22Speaker 1: whisper of star formation the universe has ever produced. A
- 1:47:26Speaker 1: fossil from the first billion years of cosmic history that
- 1:47:29Speaker 1: survived tidal destruction for 10 billion years because it was
- 1:47:32Speaker 1: cradled inside a dark matter cocoon, dense enough to resist
- 1:47:36Speaker 1: the Milky Way's relentless gravitational pull. Or they may be
- 1:47:39Speaker 1: clinging together through their own gravity, and the gravity of
- 1:47:43Speaker 1: dead companions, white dwarfs, neutron stars, stellar-mass black holes left
- 1:47:48Speaker 1: behind by more massive siblings that burned out long ago.
- 1:47:52Speaker 1: In that case, they are the faintest and most ancient
- 1:47:54Speaker 1: star cluster ever found. A system whose survival is owed
- 1:47:58Speaker 1: not to dark matter, but to the invisible mass of
- 1:48:01Speaker 1: stellar corpses that share their tiny volume of space. A
- 1:48:04Speaker 1: cluster that looks, from the outside, almost exactly like what
- 1:48:08Speaker 1: a galaxy would look like at this extreme. And that
- 1:48:11Speaker 1: can be distinguished from a galaxy only by measurements more
- 1:48:13Speaker 1: precise than current technology can reliably deliver. The unsettling truth
- 1:48:19Speaker 1: is not that astronomers found a small galaxy. It is
- 1:48:22Speaker 1: not that they found a faint star cluster. It is
- 1:48:25Speaker 1: that at the extreme lower boundary of structure formation, the
- 1:48:28Speaker 1: universe stops sorting its creations into neat compartments. A handful
- 1:48:33Speaker 1: of ancient stars may be the luminous fringe of a
- 1:48:35Speaker 1: vast dark matter architecture, or they may be the last
- 1:48:39Speaker 1: gravitationally coherent remnant of a stellar system slowly dissolving into
- 1:48:43Speaker 1: the Milky Way's halo. The answer matters enormously for dark
- 1:48:47Speaker 1: matter physics, for galaxy formation theory, for the census of
- 1:48:51Speaker 1: cosmic structure, but the answer also depends on measurements that
- 1:48:54Speaker 1: push against the fundamental noise floor of observational astronomy, where
- 1:48:58Speaker 1: the signal from dark matter, if it exists, is quieter
- 1:49:02Speaker 1: than the systematic uncertainties that surround it. Future telescopes will
- 1:49:06Speaker 1: find more objects like this one, Some will turn out
- 1:49:09Speaker 1: to be galaxies, others will turn out to be clusters.
- 1:49:12Speaker 1: A few perhaps will resist classification for decades, sitting in
- 1:49:16Speaker 1: the fog, challenging every assumption about what the universe builds
- 1:49:20Speaker 1: at the smallest scales it bothers to illuminate. And sitting
- 1:49:23Speaker 1: among them, roughly 33,000 light years from where you are now,
- 1:49:27Speaker 1: will be this one particular gathering of 60 ancient stars,
- 1:49:31Speaker 1: still orbiting, still bound, still refusing to say whether it
- 1:49:35Speaker 1: is the smallest galaxy the universe has ever made, or
- 1:49:38Speaker 1: merely the ghost of a star cluster that should have
- 1:49:40Speaker 1: dissolved billions of years ago. The boundary between galaxy and
- 1:49:44Speaker 1: star cluster, between visible matter and dark matter, between certainty
- 1:49:49Speaker 1: and classification, runs directly through it. And for now, the
- 1:49:53Speaker 1: universe is not giving us a clean answer. It is
- 1:49:56Speaker 1: giving us something more honest. It is giving us the
- 1:49:59Speaker 1: edge of what we know, the place where the map
- 1:50:01Speaker 1: runs out and the territory continues into darkness.