I have been spending a lot of time lately writing up a formal paper on high redshift galaxies, so haven’t had much time to write here. The paper is a lot more involved than I told you so, but yeah, I did. Repeatedly. I do have a start on a post on self-interacting dark matter that I hope eventually to get back to. Today, I want to give a quick note about the MHONGOOSE survey. But first, a non-commercial interruption.
Triton Station joins Rogue Scholar
In internet news, Triton Station has joined Rogue Scholar. The blog itself hasn’t moved; Rogue Scholar is a community of science blogs. It provides some important capabilities, including full-text search, long-term archiving, DOIs, and metadata. The DOIs (Digital Object Identifiers) were of particular interest to me, as they have become the standard for identifying unique articles in regular academic journals now that these have mostly (entirely?) gone on-line. I had not envisioned ever citing this blog in a refereed journal, but a DOI makes it possible to do so. Any scientists who find a post useful are welcome to make use of this feature. I’m inclined to follow the example of JCAP and make the format volume, page be yearmonth, date (YYMM, DD), which comes out to Triton Station (2022), 2201, 03 in the standard astronomy journal format. I do not anticipate continuing to publish in the twenty second century, so no need for YYYYMM, Y2K experience notwithstanding.
For everyone interested in science, Rogue Scholar is a great place to find new blogs.
MHONGOOSE
In science news, the MHONGOOSE collaboration has released its big survey summary paper. Many survey science papers are in the pipeline. Congratulations to all involved, especially PI Erwin de Blok.
Erwin was an early collaborator of mine who played a pivotal role in measuring the atomic gas properties of low surface brightness galaxies, establishing the cusp-core problem, and that low surface brightness galaxies are dark matter dominated (or at least evince large mass discrepancies, as predicted by MOND). He has done a lot more since then, among them playing a leading role in the large VLA survey of nearby galaxies, THINGS. In astronomy we’re always looking forward to the next big survey – its a big universe; there’s always more out there. So after THINGS he conceived and began work on MHONGOOSE. It has been a long road tied to the construction of the MeerKAT array of radio telescopes – a major endeavor on the road to the ambitious Square Kilometer Array.
I was involved in the early phases of the MHONGOOSE project, helping to select the sample of target galaxies (it is really important to cover the full dynamic range of galaxy properties, dwarf to giant) and define the aspirational target sensitivity. HI observations often taper off below a column density of 1020 hydrogen atoms per cm2 (about 1 solar mass per square parsec). With work, one can get down to a few times 1019 cm-2. We want to go much deeper to see how much farther out the atomic gas extends. It was already known to go further out than the stars, but how far? Is there a hard edge, or just a continuous fall off?
We also hope to detect new dwarf galaxies that are low surface brightness in HI. There could, in theory, be zillions of such things lurking in all the dark matter subhalos that are predicted to exist around big galaxies. Irrespective of theory, are there HI gas-rich galaxies that are entirely devoid of stars? Do such things exist? People have been looking for them a long time, and there are now many examples of galaxies that are well over 95% gas, but there always seem to be at least a few stars associated with them. Is this always true? If we have cases that are 98, 99% gas, why not 100%? Do galaxies with gas always manage to turn at least a little of it into stars? They do have a Hubble time to work on it, so it is also a question why there is so much gas still around in these cases.
And… a lot of other things, but I don’t want to be here all day. So just a few quick highlights from the main survey paper. First, the obligatory sensitivity diagram. This shows how deep the survey reaches (lower column density) as a function of resolution (beam size). You want to see deeply and you want to resolve what you see, so ideally both of these numbers would be small. MHONGOOSE undercuts existing surveys, and is unlikely to be bettered until the full SKA comes on-line, which is still a long way off.
And here are a couple of individual galaxy observations:
These are beautiful data. The spiral arms appear in the HI as well as in starlight, and continue in HI to larger radii. The outer edge of the HI disk is pretty hard; there doesn’t seem to be a lot of extra gas at low column densities extending indefinitely into the great beyond. I’m particular struck by the velocity dispersion of NGC 1566 tracking the spiral structure: this means the spiral arms have mass, and any stirring caused by star formation is localized to the spirals where much of the star formation goes on. That’s natural, but the surroundings seem relatively unperturbed: feedback is happening locally, but not globally. The velocity field of NGC 5068 has a big twist in the zero velocity contour (the thick line dividing the red receding side from the blue approaching side); this is a signature of non-circular motion, probably caused in this case by the visible bar. These are two-dimensional examples of Renzo’s rule (Sancisi’s Law), in which features in the visible mass distribution correspond to features in the kinematics.
I’ll end with a quick peak at the environments around some MHONGOOSE target galaxies:
This is nifty on many levels. First, some (presumptively satellite) dwarf galaxies are detected. That in itself is a treat to me: once upon a time, Renzo Sancisi asked me to smooth the bejeepers out of the LSB galaxy data cubes to look for satellites. After much work, we found nada. Nothing. Zilch. It turns out that LSB galaxies are among the most isolated galaxy types in the universe. So that we detect some things here is gratifying, even in targets that are not LSBs.
Second, there are not a lot of new detections. The halos of big galaxies are not swimming in heretofore unseen swarms of low column density gas clouds. There can always be more at sensitivities yet unreached, but the data sure don’t encourage that perspective. MHONGOOSE is sensitive to very low mass gas clouds. The exact limit is distance-dependent, but a million solar masses of atomic gas should be readily visible. That’s a tiny amount by extragalactic standards, about one globular cluster’s worth of material. There’s just not a lot there.
Disappointing as the absence of zillions of new detections may be discovery-wise, it does teach us some important lessons. Empirically, galaxies look like island universes in gas as well as stars. There may be a few outlying galaxies, but they are not embedded in an obvious cosmic network of ephemeral cold gas. Nor are there thousands of unseen satellites/subhalos suddenly becoming visible – at least not in atomic gas. Theorists can of course imagine other things, but we observers can only measure one thing at a time, as instrumentation and telescope availability allows. This is a big step forward.
Triton Station 
I am glad to see you joining Rogue Scholar. One major fault with this, is that Rogue Scholar does NOT provide a RSS feed for all its posts. Since I am a heavy user of RSS, for my science news this is a big deal. Could you send a hint to them that RSS is still alive and flourishing and could they add a feed. It is extreme easy to do.
LikeLike
Oh Dear, I need to ask a stupid question. For the velocity dispersion (2nd moment) do the higher numbers (red areas) mean more dispersion? (a wider range of velocities.)
LikeLike
Yes. The color bar at the side shows what color corresponds to what velocity dispersion. I’m so used to these that I forget how confounding they can be the first hundred times you see such a plot.
LikeLike
Could you say a little more about the significance of not finding detectable cold neutral hydrogen beyond the galaxy perimeters? If one aim of MHONGOOSE was to confirm a prediction to the contrary, the failure to detect such gas should have its own implications. The significance is perhaps enhanced when one considers that there do exist detectable amounts of (hot?) ionised gas around the galaxies – e.g. Peebles ‘Cosmic fog and smog’ in Nature a few years ago.
LikeLike
An obvious hope at one juncture was that many of the predicted subhalos would light up when observed this way, so we’d see thousands instead of tens of such satellites. That hope has long since faded, and there is no clear prediction to test.
There is certainly a fair amount of hot, ionized gas in the circumgalactic medium around bright galaxies. Exactly how much is unclear. This is sometimes thought to be associated with the halo of the host galaxy rather than its satellites. One reason for this is that galaxies like the Milky Way apparently have rather fewer baryons than they should – maybe 1/3 of what should be there. So this circumgalactic medium is the reservoir of requirement for where these missing baryons might reside. While there are certainly some baryons in this component, it is rather sanguine to hope it adds up to the right amount. This problem gets worse as you look to lower mass galaxies, where the missing baryon problem is an order of magnitude or more. See https://tritonstation.com/2016/08/06/missing-baryons-in-lcdm-and-mond/
LikeLike
I will have to look more carefully at the basis on which Shull et al infer a deficit of baryonic matter. I was thinking more about how, if at all, galaxies accrete whatever mass they are known to have. Di Teodoro & Peek (2021) found little evidence of any sustained radial inflow of HI gas. The De Blok paper supports this finding. The edges of the discs are hard.
The evidence is that galaxies do not and cannot grow in mass. And if groups and clusters are actually the products of supergiant proto-galaxies splitting – most frequently at high redshifts – then, far from gaining mass over time, galaxies get smaller. The Bullet Cluster would be a perfect example of galaxy generation by multiplication, since no one doubts that the subclusters are diverging. The superhot gas in the middle is the residue of this process.
Galaxies become less compact as their nuclei (their ‘black holes’) shed mass. Even in the expanding universe scenario there is little evidence that galactic mass increases through time. If the universe is not expanding, the masses of galaxies at higher redshift would have to be revised upward.
The hot CGM, of course, is not primordial gas.
LikeLike
There is a lot here to keep straight, and doing so is not straightforward. There are two distinct missing baryon problems: one global and one local. Shull addresses the global one, that the sum of known baryons does not add up to the amount expected from BBN. I have opinions about that, but here I was referring to the local deficit: individual galaxies do not appear to have all the baryons that they should for their DM halo mass. There are easy ways to explain that in a hand-waving way, but it is harder to account for the systematic variation of the detected baryon fraction with mass along with the small scatter around this trend. So, situation normal, a complete mess.
I concur that there is considerable evidence that galaxies are not growing rapidly at the present time. There is always room for a little accretion, but it is typically not as great as anticipated in many models. To say more requires wading into the messy details of individual models.
LikeLike
<i>The evidence is that galaxies do not and cannot grow in mass. </i>
What about the recent studies that have identified stellar populations in the Milky Way that appear to be the product of galactic mergers?
LikeLike
Mergers certainly happen sometimes, so “no mass growth” is an overstatement. What isn’t obvious is that they are accreting large amounts of mass via gas inflows at the present time.
Similarly, the stellar pops that are the remnants of mergers into the Milky Way do appear legitimate, but also represent a rather small amount of mass that accreted a very long time ago.
LikeLike
There are a few unequivocal examples of mergers in the nearby universe, so yes, it is a bit of an overstatement. Gravity will re-assert itself if the offshoot does not exceed the escape velocity.
LikeLike
To be clear, if, as I suspect, genuine mergers represent re-integrations of originally unified material, there is still no net growth.
LikeLike
Re the paucity of dwarf satellites, these are now also being detected at redshifts of 0.5 to 3.0 (Suess et al 2023). Across a sample of 161 quiescent primaries, a total of 629 companions were identified, 94% of them smaller than 1/10th the mass of the primary. This is ‘not unexpected’ – though why so many low-mass entities should converge on high-mass galaxies from beyond their gravity wells is not clear to me. In a non-expanding universe, the primaries and companions would be genetically related and their masses one to two orders of magnitude greater. The ‘dwarfs’ would then be normal-sized galaxies. It seems significant in this regard that even when lumped together there is a noticeable discontinuity between the masses of the primaries and the companions (Fig. 2 in the paper).
LikeLike
I have a question about cluster collisions, which as you’ve said, may reveal baryonic matter – which would ease the missing baryons problem, and also help MOND indirectly.
Whatever the collisionless material in clusters is, it travels through with the visible matter, leaving the gas behind, as lensing data shows extra gravity there with the visible matter.
I’m sure this has been ruled out, but how do we know MOND doesn’t boost the gravity of both the visible matter and the gas independently? They get separated, but if we don’t know how much gas there is, MOND could boost the gravity of both. So perhaps we have two separate ways to estimate the mass of the gas, one of them being lensing data. (I know there are questions about which versions of MOND show up in lensing data, which might complicate things.)
LikeLike