The Arxiv[0] paper abstract seems much clearer and more informative than the linked article.
In particular, it points to WHY the mass has been revised down so far -- there's seemingly a lot less mass at 19+ kiloparsecs (about 62,000 lightyears) from galactic center than we expected the galaxy to have. I'm not an expert so it isn't clear to me if the discussions in the paper about dark-vs-baryonic matter are a theoretical explanation of the gap.
Here's a 2019 article[1][1b] describing the mass of the galaxy at 1.5 trillion solar masses. This current research lowers that number to just 200 billion solar masses with a 3 sigma confidence.
The implication is that there may be significantly less dark matter in the Milky Way than previously estimated. From [0]:
“Using this newly derived rotation curve, a new mass was deduced for our Milky Way. And surprise... it is much lighter. It is now estimated to be only two hundred billion times that of the Sun (~2.06 x 10^11), so about four to five times lower than the previous estimates. The amount of ordinary matter did not change, thus, there must be a lot less dark matter in the Milky Way than previously thought. The new expectation is that the ordinary matter like stars and gas in the Milky Way now makes up about 1/3 of the Milky Way mass and the other 2/3 are accounted to dark matter.”
> Is it possible that another turn of the crank or two takes dark matter to zero?
Not sure about the Milky Way. But we already found galaxies out there that don't seem to have any dark matter; ie where their rotation curves match what you would predict from the ordinary matter alone.
Interestingly enough that is evidence in favour of dark matter: it's easy to conceive of mechanisms that can separate dark matter from visible matter with some low probability in a galaxy merger or near-merger. (Similarly for galaxies having more than the average amount of dark matter.) But alternatives like Modified Newtonian Dynamics (MOND) would have a much harder time accommodating those differences.
No. MOND predicts a large external field effect in the milky way due to satellite galaxies (like the large magellanic clouds). This would manifest itself in as "a reduced DM halo", and an keplerian decline (see figure in TFA). LCDM typically has a difficult time with this because in order to have a flat curve, the dark matter must be distributed in a halo that is MUCH bigger than the galaxy, and that leaves no explanation for the decline.
Not having run the numbers, at first glance it looks like these new data are STRONG support for MOND.
MOND also generally predicts that denser fast rotating galaxies should appear to have ~no dark matter (in these cases centripetal acceleration is > a0 constant, so it's in the Newtonian regime) , which is highly consistent with observations of galaxies.
Naively (I have not run simulations) thinking about it it seems puzzling that the denser galaxies would have no DM -- since DM is supposed to be the nucleus for galactic formation in the early universe.
Conversely, the galaxies that are highly diffuse (UDGs) seem to have extra dark matter, which is very consistent with MOND (since being diffuse, gravitational acceleration a < a0). Again, this makes little sense with LCDM as you would expect a pocket of extra DM to attract mass and nucleate the formation of a very dense galaxy.
> separate dark matter from visible matter with some low probability in a galaxy merger
Generally these observations have been done with weak lensing, and unfortunately there is not really yet a good model that reconciles GR with MOND. Apparently the math is hard. It is entirely possible that when you combine the two the calculated spacetime curvature solves to what you would expect from the observed weak lensing effects.
> .. at the outskirts of the disk of the Galaxy, this [rotation] curve begins to decrease rapidly, following the prediction known as the Keplerian decline
Yes. MOND predicts the keplerian decline, due to the EFE, hence:
> these new data are STRONG support for MOND
In standard cosmology keplerian decline is predicted because other galaxies have it. It makes no sense at all from the structure of DM halos that galaxies are supposed to have.
You have it completely backwards. Keplerian decline, i.e. the decline of the rotation speed the further away you get from the center of your solar system, galaxy or whatever, is predicted by traditional cosmology.
What is detected in other galaxies, and what was thought to be detected in our own galaxy is well, is a lack of Keplerian decline. That is where both dark matter and MOND stepped in trying to explain. See e.g. https://link.springer.com/article/10.12942/lrr-2012-10
What was detected now, is that our galaxy shows Keplerian decline after all. Which, at least for our own galaxy, removes the need for introducing MOND or dark matter.
No. Keplerian decline doesn't mean that the entire rotation curve is keplerian. It means that it regresses to keplerian after a certain radius. You cannot explain the plateauing part within that radius without MOND or LCDM.
Without knowing anything about MOND, this never made sense to me.
If MOND can explain away all cases where m > 0[*], it'd be a bit strange it can't explain m = 0.
If MOND can only accommodate a certain range, wouldn't that also present a problem for dark matter? If all previously known galaxies had dark matter in that range, why would every galaxy have roughly the same amount of dark matter except the one that doesn't have any?
[*] Where I guess m would be the ratio of dark to ordinary matter.
MOND can most easily explain a certain fixed value of what you call m. It has a much harder time explaining lots of different m at the same time.
> If all previously known galaxies had dark matter in that range, why would every galaxy have roughly the same amount of dark matter except the one that doesn't have any?
They don't all have the same amount of dark matter. There's a distribution.
You were right, that a weird distribution would invite investigation.
Actually, dark matter isn't the only game in town.
The whole need for dark matter arose from the fact that our theories (on gravity) didn't match some of our observations - while matching other observations phenomenally well.
Abstractly speaking, when theory doesn't match observations, either the theory is lacking, or the observations are. Dark matter is the hypothesis that the observations are lacking. The alternative is to come up with modifications of Einsteinian gravity that preserve accordance with observations where it's extremely good (e.g. perihelion shift of Mercury), and modify only the results for the parts where the match with observations is lacking. To the best of my knowledge, this category is collectively called MOND.
There are some good reasons to think that there exists a type of matter that we cannot observe - not that we didn't happen to see it like some small black hole in the middle of nowhere, but that our current instruments literally cannot see it, only its effects. There are also some reasons to think otherwise. The main one for that is that dark matter, by its nature, should be something. So not only should dark matter theories make predictions that can be proven, evidence for dark matter should eventually answer the question "what is dark matter?" That is a bit of a problem: right now, it might as well be unicorn farts.
IANAP, but I thought is that MOND is a modification of classical gravity and that GR-compatible extensions to MOND end up introducing dark-matter like terms.
Dark Matter is the opposite of String Theory imho.
String Theory was about desperately seeking a Grand Unifying Theory that everybody wants and mashing math together until you figured out something that matched the evidence but was so overcomplicated it was completely untestable.
Dark Matter is about having evidence that something is terribly wrong with the math, and the evidence pointing to something goddamned absurd that everybody hates, and reluctantly finding more and more evidence that this thing everybody hates might be true.
The main thing they have in common is that they're both incredibly ugly solutions to problems in physics, but one is invented to reconcile a an inconsistency in theory, and the other is invented to reconcile an inconsistency in evidence.
And honestly, that's why I'm comfy thinking dark matter is probably real: I'm a pessimist. The ugly solution that everybody finds kinda gross is probably the truth.
We've been in this situation before a few times. E.g. when geology was new, the earth was much too old, compared to what the heath radiating from it allowed. Chemical reactions were simply not up to the job over such long time spans.
We needed wild new physics, nuclear reactions, before things started to make sense. Everybody involved knew something was very wrong. A lot of ink was used trying to explain the error. But finally, the error was settled as a sidekick result from a completely different branch of science.
The dark matter hypothesis has always seemed like BS to me, like the simple product of a math error. I think the observational results are caused by something else that's not understood or seen yet.
What's funny to me is how people can be so confident about astrophysics theories, when we're just a primitive race that hasn't even ventured beyond our own moon and fighting brutal wars over whose imaginary god is correct.
We could also call it the "Jesus Christ effect" too. Certainly that wouldn't suggest any sort of mechanism or bias interpretations of what could cause it
If you're interested in learning more I recommend you watch "dark matter is not a theory"[1]. We consistently, across many measures, observe the effect of substantially more mass than we expect using traditional understandings of dynamics. People have been aware of it, under different names, for about 90 years now.
This is terrible. Of course dark matter is a theory. Saying it's just a bundle of observations is like saying "1) lee Harvey Oswald had interactions with the CIA and 2( he killed JFK", and claiming "hey man, just making two innocent factual observations". No, of course not. By shoving different things together you are implying a causal connection, which is exactly a hypothesis.
This kind of a wordplay shell game is really bad for science
Things fall towards the earth. You can check that they do right now by dropping something. That things fall towards the earth isn't a theory - it's an observation. Gravity - the most popular explanation for why that happens - is a theory!
"Dark matter" comes from a group of observations that consistently suggest there is more mass in the universe than we can account for. The fact that we keep observing it is not the same thing as an explanation for the observations. There are also theories that try to explain dark matter! But the fact that people observe it is not a theory.
"he killed JFK" isn't an observation: there's no witnesses that ever testified to this. You don't know that he killed JFK, and in fact new evidence casts serious doubt on it (it was here on HN recently I think).
There's a good cosmology YouTuber who basically explains: dark matter is not a theory, it's a set of observations. Galaxies behave like there's a halo of invisible mass surrounding them, one that varies in density per-galaxy. The halos even perturb as expected in galaxy collisions. But nobody's got a coherent theory on what that halo is.
This is maybe half true. Dark matter is required to explain these under the current lambda-cdm model, but the behavior of galaxies is not evidence for it in a constructive sense. It's only evidence that the current model is untenable.
Galaxy rotation curves are better explained by applying general relativity without the severely restrictive assumptions required in lambda cdm. You don't even need the full thing, just the first order linear approximation that allows for gravitational waves (and thus is causal), as Ludwig showed a few years ago. You need at least this because gravitational waves exist, and those cannot occur in the singular newtonian limit used mostly for convenience.
It doesn't take much to then question the need for dark matter as if it is compensating for poor models in one case, it probably is doing the same in others.
> Galaxies behave like there's a halo of invisible mass surrounding them,
No they don't. Galaxies exhibit anomalous velocity curves according to existing gravitational theory as applied in a specific cosmological model. That's the observation. This could be explained by a halo of indivisible stuff or it could be explained by a misunderstanding of gravity or a different cosmological model.
I'm general, we should not explain observations in terms of speculative theories, because this carries a presumptive bias.
> The halos even perturb as expected in galaxy collisions.
They actually don't. Dark matter would explain part of the bullet cluster (lensing), but it can't explain the high collision velocities observed.
I think almost everyone realizes dark matter is a placeholder for the unknown. It doesn't mean we shouldn't explore the possibility of special dark matter particles that don't interact with regular matter. In fact, there's not much else to do.
I think this is the most honest-sounding reply to a layperson.
Our data doesn’t fit GR with the boundary conditions that make sense more locally. Positing a bunch of non-interacting mass/energy is closer to observations than doing nothing.
Calling it a placeholder seems honest, reasonable, and productive.
Getting pushy about WIMPs and stuff (which is deeply intertwined with the refusal to give up on supersymmetry because, careers) in a world where Michio Kaku is on TV talking about quantum computing and AI in a way that’s less scientific than Matt Gaetz yelling about UFOs on C-SPAN to distract from the child-trafficking charges is icky.
There are two historical dark matter theories in early cosmology. Neptune and Vulcan. One of them turned out to be real. It would do to remember that the other one did not.
Many Christians vehemently disagree that the Muslim god is the same as theirs. What a bunch of academics have to say on the matter is irrelevant; religion is just someone's personal belief and completely unfalsifiable, so you can't say they're "wrong".
No, I'm criticizing the people who claim that one group is less wrong about their fake deity than another. I see this all the time when religious violence flares up: "they're not following their religion correctly!", coming from people who aren't even in that religion but think they're experts on that religion's teachings (and ignoring the fact that so many believers have incredibly different ideas of what those teachings are). It really annoys me, so I feel the need to call it out. If someone claims that their religion tells them to murder children, then I believe them; it's all made-up anyway, so it's ridiculous for some outsider to say they're doing their religion wrong.
With the original in french, and not a physicist, the three pieces of information i wanted:
- What did they revise it from/to?
- What is the confidence?
- What is the suspected reason?
1.5 trillion solar masses seems like an insane number for a galaxy that's over the hill on star formation, has about 100 billion stars, over half of which are red dwarfs.
The vast majority of the mass in most galaxies is dark matter anyway, so the number of stars is not so important. Remember, normal matter is only about 5% of all the stuff in the universe.
Given that they just revised the mass of the Milky Way by a factor of five, we should not regard all the other numbers as being absolutely certain. They may be the best numbers we have now, but they are subject to revision, sometimes drastically.
There's already a lot of variation in the proportion of dark matter to visible matter in different galaxies. Revising one galaxy doesn't change the statistics that much.
(However you are right that from a Bayesian point of view, the Milky Way is the galaxy we have the most data on. So should count for a lot more than a random galaxy.)
Yes, that’s the basically the reason this paper gives for the large discrepancy between their estimate of the mass and prior estimates. The Gaia satellite measured the positions and velocities of billions of stars in the Milky Way, information we’ve never had before, so a new estimate being significantly different from the prior estimate is not _that_ surprising.
But the 5% number that I quoted above should not be taken as an extremely precise number. You can plainly see that it has only one significant figure; if the real number turns out to be 4% or 6% then I don’t think cosmologists will be too surprised.
This observation clearly contradicts multiple other measurements of the MW mass which is not really explained. Given quite a complex set of assumptions that went into this particular derivation, I am personally quite sceptical about this measurement, but no one have identified a specific flaw of the study.
If I understood this correctly, this mass was derived from the rotation curve, which was itself derived from the motions of stars. The other mass estimate was from (a rotation curve derived from?) neutral gas. So there's maybe some room for the relationship between the motion of the stars and the motion of the neutral gas to be not what we expect.
Or, there's room for there to be a flaw in this study. Or for a flaw in the studies (I assume plural) that used neutral gas, but it would have to be a flaw in the assumptions, or in the techniques common to all of the studies.
Most of mass estimates of the MW in the outskirts come from measuring motions of stars (just not in the disk of the MW) dwarf galaxies, globular clusters or stellar streams. These measurements have their own issues, but now many of these measurements broadly agree and this new one very is an outlier.
Measurements using gas have their own issues, and also they do not go as far away from the galactic centre as the new rotation curve from the paper.
IMHO the main point of the French article is that the rotation speed of the galaxy decreases according to Kepler laws for the outer parts of our galaxy. They're us a graph for that. This contradicts previous observations. They ask why our galaxy should be special and they suggest that we (they) measured rotation speed of stars in the Milky Way while we measure rotation of gas in the other galaxies. I'm not sure why gas and stars would have different orbital speeds, but well, who am I to actually understand this matter.
WTF is the picture in the linked press release? They've multiplied Kepler curve by an arbitrary coefficient of ~2 to make it match the measurements from the paper in 16-25 kpc range, while it clearly won't match in < 16 kpc range. What is this picture supposed to tell us? Compare it to figure 5 from original paper to see the actual Kepler curve used in the study.
In particular, it points to WHY the mass has been revised down so far -- there's seemingly a lot less mass at 19+ kiloparsecs (about 62,000 lightyears) from galactic center than we expected the galaxy to have. I'm not an expert so it isn't clear to me if the discussions in the paper about dark-vs-baryonic matter are a theoretical explanation of the gap.
Here's a 2019 article[1][1b] describing the mass of the galaxy at 1.5 trillion solar masses. This current research lowers that number to just 200 billion solar masses with a 3 sigma confidence.
[0] https://arxiv.org/abs/2309.00048 [1] https://www.sciencealert.com/the-most-accurate-measurement-y... [1b] https://arxiv.org/abs/1804.11348