2024-01 Jan Science Readings

---

Well, what have I got this month?

2024-01-03 - Leftovers.

I've decided to try moving to one post per paper (or topic). These posts are getting unmanageably long.

Contents

Articles studied this January - some of which might go to Slashdot. But … I'm losing the will to contribute to Slashdot.

Short-lived Repeating FRBs from White Dwarf disruption

Little Bits of Language

Negative Findings

Always a provocative Idea - does the Sun contain a black hole?

Crash, Bang, Wallop

New Year, New Mond updates

Duck! OK, it missed. Didn't it?

Flag Washing

HTML Learnings for the month.

Happenstance, Circumstance, Enemy Action?

The Role of Giant Impacts in Planet Formation

Large planets may not form fractionally large moons

End of document

EDIT - May : I knew I'd been writing about "Thorne-Żytkow" objects (normal-ish stars which take a meal of a compact object - neutron star, or black hole, and particularly a tiny "primordial" black hole - which as a significant effect on the star's subsequent life.) recently, but I couldn't find the damned stuff when i needed it. So I've modified the titles and labels a bit.

---

Short-lived repeating fast radio bursts from tidal disruption of white dwarfs by intermediate-mass black holes

https://arxiv.org/pdf/2312.03107.pdf

FRBs - Fast Radio Bursts - remain problematic. They're at isotropic distances (outside the Galaxy, randomly distributed), very energetic, but from very small source regions (because of their rapid variability). Being a new (first detected 2007), "shiny" phenomenon, they continue to attract warranted attention. Obviously much kudos will accrue to whoever comes up with a workable solution. A big problem to such a solution is that, unlike most energetic events, at least some in this class can repeat - irregularly, and more rarely regularly.

This proposed class of solutions envisages moderately common objects interacting in a moderately unusual way. Using common objects is good for an explanation - little is known about the digestive tracts of rocking-horses, since rocking-horse shit is so rare. Ordinary horseshit on the other hand … The proposal here is to take a fairly normal black hole [an IMBH - "Intermediate-Mass Black Hole" formed by stellar collapse, massing a fistful of solar masses, not a planet mass (Primordial Black Holes, PBH - still theoretical) and not a large portion of a galaxy (SMBH - Super-Massive BLack Holes). Truely a pedestrian object - we've even detected their formation and interaction in recent years, using gravity wave observatories such as LIGO, KAGRA and VIRGO.] and throw a fairly normal White DWarf (WD, wiki link) at it ; as the BH (a few km across) tears chunks off the WD (a few thousand km across) and swallows them, the torn-up remnants may well emit radiation on time scales, frequencies, and with variability consistent with FRBs.

Even better, compact bodies (Neutron Stars, IMBHs, WDs) should fairly naturally gravitate (literally, gravitate) towards similar areas such as the centres of galaxies, globular clusters etc, increasing the likelihood of interactions.

The non-total destruction of a WD in the first encounter provides a natural mechanism for the recurrence of an FRB in a particular system. The figure that describes the model implies that the outputs are sensitively dependent on the details of a particular encounter. If any significant amount of mass transfer takes place, the details of the next encounter (and thus, a repeated FRB) are very unlikely to be identical to the previous encounter.

A particular wrinkle of this model is that the material from the accreting WD should cut across the BH's magnetic field lines, which would lead to pulsations in the accretion - as seen in FRBs. Of course, other compact body mergers would also potentially suffer similar effects, which would make deciding the cause of a particular event a matter of fairly finely balanced observations. But that was probably always going to be the case. Magnetic fields are to be expected on WDs, and NSs … but how much remains "frozen into" a BH after it's mass disappears beyond the event horizon … Good question!

Nice model. Might even be true!

Back to Contents List.

---

Little bits of Language - Semi-Random notes.

Number 1 in a possibly repeating series.

When I stumble across a new word (to me) in English (or any of the other languages I study), I like to learn a bit more about it. Making this sort of note helps to fix things in my memory. It's the same logic as behind my Who Said What collection - to which I have a few new things to add. Haven't touched that since April last year, so that needs a little more fettling.

What is the origin of "kudos"?

Oh, all the way from Ancient Greek κῦδος (kûdos, “praise, renown”). I'm slightly surprised not to have guessed that myself. See also, "nous"

Wasn't kudos a big element in the Illiad? Or didn't they call that "kleios"? Sigh - another one for the re-reading list. Something about Achilles saying rude things about Ancient Greek life from the perspective of being dead. So, Odyssey, not Iliad.

Back to Contents List.

---

Negative Findings

Search for planets in hot Jupiter systems with multi-sector TESS photometry. IV. Null detections in 12 systems

One of the things non-scientists frequently don't get is that scientist are continually testing their theories by looking for things, and not finding them. This is a classic of the (not so common) publication of a negative result. (More often, they negative results are incidental remarks in the "Methods" section of a paper.)

This is perfectly simple : the data from a dozen systems observed by the TESS satellite was examined for short period, high frequency luminosity variations. They didn't find any. Which doesn't mean that these systems don't have planets - any planets could be further out (requiring longer periods of observation ; or the planet's orbital planes could not intersect with Earth ; or the periods of observation could have just been unlucky. Or, indeed, the systems may not have planets.

Negative results are a bedrock of science. They're just not sexy. If your data analysis pipeline always produces a positive result, you don't know if the result is a product of external reality, or an artefact of the processing pipeline. If you get positive results with some targets, and negative results with other targets, there probably is some difference between the two systems.

Back to Contents List.

---

A Provocative Idea - does the Sun contain a black hole?

Is there a black hole in the center of the Sun?

Ohhh, this sounds like fun! Beats the LHC eating the Earth any day of the week. It'd cure (for certain values of "cure") Anthropogenic Global Warming as something to worry about. And that's just the title.

Where could they be going with this? Astroseismology? Or perhaps (yet another) solution to the Solar neutrino problem? Or just blatant redneck-baiting (I have a sudden image of modifying the DREADCO "buttered toast/cat" perpetual motion machine to run off the inexhaustible indignation of rednecks about things with too-long words. Cynical? Moi?)

So, speculation aside, let's RTFP. Starting with the author list - three long, not an Avi Loeb solo lunacy special (sorry, that's being cruel ; Loeb's lunacy is generally well reasoned.), enough people there to keep somewhere in contact with reality. No names I recognise. Three institutions. None of the strong indicators of the ... fringe.

Abstract There is probably not a black hole in the center of the sun. Despite this detail, our goal in this work to convince the reader that this question is interesting … Oh, it's one of those papers! OK, let's play!

After introducing the idea of a Hawking star (which has an asteroid-mass black hole 10^-16~10^-10×M_☉ accreting internally in addition to normal nuclear fusion, the authors justify modelling it's evolution in some detail.

The assertion that "Despite the claimed constraints, [primordial black holes ("PBH"s, in the mass range cited above)] remain a compelling dark matter candidate for several reasons" is a bit of a surprise to me. I thought that had been excluded by OGLE, MACHO and the like, but maybe I wasn't paying close enough attention. Meh. Accepting that assertion - because it sounds a real fun idea for testing redneck PM machines - how do these authors think the presence of an accreting PBH would affect the evolution of a solar-mass star?

This sounds somewhat familiar. Did I RTFP, without making notes? Naughty Aidan! Or maybe I read one of the papers they cite - several earlier in 2023. This argument "Stars are unlikely to capture PBHs after formation. A PBH with finite velocity and falling from infinity will be accelerated up above the escape velocity and rapidly transit the star." sounds very much like an argument I made on Slashdot not-long ago. Contrary to the Physics of Star Trek ("many readers", as the authors … politesse it), the accretion of PBHs isn't an all-enveloping maw, but strongly limited by the size of the event horizon. The 10^-10 M_☉ PBH would have a radius of about 30000 pm while the 10^-16 M_☉ PBH would have a radius 0.03pm. For comparison, the (Van der Waals) radius of a hydrogen atom is about 120pm. So the number of atoms that will be absorbed at a time is in the order of zero to a few million - in the order of 10^-20 kg per position. (That's my estimate. Their models, incorporate compression and gravitational radius to give a (reasonably simple) expression for the luminosity in L_☉ (incorporating BH mass accretion rate (M_☉/yr) and radiative efficiency (dimensionless).) Their model give accretion rates 10^-9 to 10^-23 M_☉/yr I can't meaningfully assess the plausibility of these numbers, but a rate of 10^-9 would have consumed the Sun 3.5 billion years ago. Over those accretion rate ranges, the inferred luminosities due to the BH are well below 1% of Solar luminosity - which would be within the "noise" in nuclear fusion models, given uncertainty about core metallicity and physical conditions. In short, on luminosity grounds, we can't rule out a present-day in-Sun accreting PBH.

As Telly "Kojak" Savalas used to sing, "If a picture paints a thousand data points …", so here's the graphs. Zoom in to read the axes. Their model changes behaviour (convection-limited to radiation-limited) in the middle of the M_BH range considered, but still leaves Sol as not-excluded from being a Hawking star.

Section 3 of the paper examines variations on this "toy" model, including how convection pressure or radiation pressure ("Bondi" or "Eddington" domains) throttles accretion onto the BH. Two main points stand out :

1. The overall luminosity (and evolution over time of luminosity) are not greatly changed by the presence of the BH ;
2. but, the interior of the star is considerably better stirred by the BH, and this potentially could yield a measurable signal by astroseismography, and the early mixing of the core of accumulated fusion products (for a Solar mass star, essentially helium) into the outer body of the star.

The larger BH mass models cause the end of nuclear fusion at ... well, not far off the Sun's present age. Which is exactly the reason that "Hawking stars" were proposed as a "solution" to the "solar neutrino problem" in the 1970s. (The correct solution is that neutrinos have mass, and so can experience time, and so can oscillate "flavours" during their flight from Sun to neutrino telescopes on Earth. There isn't a "solar neutrino problem" now.)

The "Kippenhahn" diagrams (Figs 2 below and 4) are a nice way of summarising the evolution of stellar structure with time. I'd not met them previously. Wiki doesn't use these, but they do allow one to visualise and compare evolutionary tracks well. I suspect they'll appear at some point.

Note that the radius and luminosity vertical scales are logarithmic. The final size of a solar-mass BH is less than one-millionth of the star's radius for most of it's evolution. (If you wanted to go to longer timescales, you might make the time axix logarithmic, and a lot of Kippenhahn diagrams around the explosion of supernovae do exactly that. But these don't - just a linear time axis.

Figure 4 uses a different time scale to figure 2 : decreasing time to stellar destruction, from billions of years to the last few seconds. Things happen fast at the end.

And despite my warning above about logarithmic scales, I've slipped up above where I said that overall phenomena are not greatly changed by the BH. The authors say Once in its post-main sequence phase the Hawking star becomes fully convective and slowly swells over multiple Gyr to about 10 R_☉ and about 10 L_☉ causing them to appear as a sub-subgiant star, joining a relatively sparse population on the HR diagram. This phase of stellar evolution is qualitatively insensitive to the initial seed mass, and lasts for multiple Gyr in our simulations.

That's somewhat smaller but longer lasting than the typical solar-mass red giant phase, so you might be able to find Hawking stars in the lower-luminosity "sub-giant" area of the Hertzsprung-Russell diagram. A moderate difference, but it might be enough to perform a pre-cut on candidate Hawking stars. (Proposed TZOs have been reported from exactly this sort of search.)

Getting into the summary, indeed the authors consider astroseismology as a potential way of detecting such stars due to the fully-convective nature of the star. Another good question : Are there X-ray transients associated with stellar destruction that are distinct from other accreting binary systems, and if so, what are the lifetimes of this transient?

It's a fun subject, and it's got lots of potential for ha-ha-but-serious blowing of minds. And Hollywood is absolutely going to love the idea of the Sun being eaten from the inside by a BH. OK, They'll have to invent some pseudo-physics for James Tiberias Hulk and Spock Man to save the universe (well, Earth) with. But fun nonetheless.

This is strongly related to this post I made a few months ago.

Back to Contents List.

---

Crash, Bang and Wallop

Outcomes of Sub-Neptune Collisions

Who doesn't love a bit of random destructiveness at a suitably large range? In the other carriageway for a little car bump ; in a different part of the galaxy for slightly bigger bangs, as discussed here ; in a galaxy far, far away for whatever causes Gamma Ray Bursts. This is medium-scale rubber-necking.

The recently published 6-planet system in (I can still remember the "phone number"!) the HD 110067 system (which has also been reported to be a widely-spaced multiple star) is very much the sort of system they're talking about. Though I don't think they mention it by name.

Why does this class (size range) of planets deserve separate treatment (from "rocky" planets, "small bodies", "Jupiters" (hot or not) and bigger bodies? FTFAbstract : For sub-Neptunes, the dominant type of observed exoplanets, the planetary mass is concentrated in a rocky core, but the volume is dominated by a low-density gaseous envelope. For these, using the traditional sticky-sphere assumption is questionable. While they can get mergers (for certain ranges of energy and impact factor), loss of large amounts of atmosphere are also a common outcome (which doesn't really have an analogue in smaller collisions, and complete disruption of both impactors is also common. That latter might suggest that growth of planets rather stalls in this mass-range, leading to the sub-Neptune group being the commonest type of planet discovered. (Though how much of that is a discoverability issue remains open.)

Rather than the more common computational strategy of using an array of different test objects, they only consider one pair of planets (masses m₁,₂ = 4.51,7.55 M_⊕, radii r₁,₂ = 3.56,3.69 R_⊕) around a sun-like star. But they inspect the parameter space by putting the two bodies in an unstable pair of orbits, then letting them evolve. With computational noise, that generates a … realistic (hmmm - questionable?) range of impact conditions for investigation. Unusual strategy, but I can see reasons for using it.

Those "test planets" are 9.99 % and 15.0 % of the Earth's density. I guess those numbers aren't so random as they look. 3:2 ratio of masses too. I wonder if they "select" from the variety of impacts they get : yes, they select on v_impact÷v_escape. And high values of v_impact÷v_escape lead to complete disruption of both planets - now that shouldn't be a surprise.

Here's another non-surprise. The star would get in the way of studying the interactions of the planets, so We treat the host star as a non-interacting point-mass particle, exerting only gravitational influence over the planets. If a particle approaches within one solar radius of the star, it is accreted into the star, conserving mass and linear momentum. Again, perfectly reasonable, but it's fun to see people saying it. hubris, thy name is throwing planets at each other for a living. Must make for fun dinner parties. Or am I channelling Rick'n'Morty from the telly? (episode: "The ABCs of Beth" - hey, the Krootabulan male has an appropriate three nipples!).

Hmmm, this description of the planets is … not over-detailed : We assume that 15% of the total planet mass is in the H-He envelope, and the rest is in the iron core and rocky mantle in 30 : 70 ratio. We truncate the planet profile at a negligible density of 0.01g cm⁻³ and consider this as the planet’s surface. It's not wildly different from "Earth". But using an Equation of State (EOS) for a granite to describe a mantle? Are there no suitable peridotite EOSs?

More hubris in the planet-smashing community : We tackle [problem] by utilizing StarSmasher’s ability to handle unequal mass particles. yes, that is the name of one of their software packages, which they use for smashing stars together. But it's obviously a game's name. Sick! They're sick, I tell you!

OK, I learn something new every day. Today is the use of a CGS pressure unit, the "barye" (symbol "Ba"). Which is 0.1 Pascal (SI), a millionth of a bar (non-SI). That's unhelpfully obscure. I read quite a bit of astronomical work (and specifically, planetary science), and I had to look that up. Not a helpful choice of units - but probably also in "progress one funeral at a time" territory. The pressure and density profiles in Fig 2 are otherwise useful - yes, these are very different planets to Earth ("terrestrial" planets) and Jupiter (gas giants, see for example, the Juno data about Jupiter in Figure 1 of this Icarus paper of 2022). At the limit, these sub-Neptune models start to approach the "dilute core" models for Jupiter, but that theory (and the disk instability formation model underlying it) are moving into the realm of theory, not (Solar system) reality.

All that comment on the work of the paper, and I finally get to the Results. They're not too surprising : they get three types of outcome : "Hit and Run" (the most common ; two planets at the end ; small mass transfer, mostly in the atmosphere component) ; "Merger" (one planet, and some debris) ; and "Catastrophic" (no large fragments of either planet result ; high mutual velocity compared to escape velocity). The boundaries between "Hit and Run" and "Merger" are rather blurred, as different workers seem to use different degrees of mass loss (to the star ; to distribution away from the "inner" planetary system) to classify interactions. "Mergers" can lose up to 40% of the two planet's mass, and "Hit and Runs" can lose up to 25% of the system's mass. Which is not exactly a trivial consequence. "Mergers" that loses 40% of the mass of the planets could largely "stall" planetary growth at the "sub-Neptune stage" without there being a meaningful "reason" for sub-Neptunes to be favoured over (say) Jupiters. Which is rather less significant than the raw observational data might seem to suggest.

Back to Contents List.

---

New Year, New MOND data

A personal itch, being scratched in public

On a regular basis one hears assertions that "Big Science suppresses non-standard theories" - often followed up by "Free energy" schemes, Climate change scepticism, Anti-vaccination screeds and the like. It's a feature particularly of Slashdot, but more generally of the Internet. If I "did" Facebook, I'd probably cite them too. Kooks are everywhere and are vocal, and they love a "my voice is being suppressed" narrative.

So, a few months ago, I conducted an (admittedly crude) survey of a "controversial" idea in "non-standard" science using Arχve. It was nothing complex (I searched for various terms ["Mordehai Milgrom", "MOND", "Non-Newtonian Gravity", "MOG", "Dark matter" in Arχve's database (anywhere in article title, abstract, or body text, or figures) from Jan 01 to Dec 31 for each year, recording the counts. And now, it's time to update the numbers. They're not good reading for the conspiracy-of-suppression" theorists (well, is any reading good for them? It's not as if they like actual evidence).

Date of search	Mordehai Milgrom	MOND	Non-Newtonian Gravity	MOG	Dark matter
1991 - 2001-12-31	75	1072	735	180	43348
2001-12-31	4	38	46	16	3404
2002-12-31	2	12	13	2	693
2003-12-31	1	22	17	2	765
2004-12-31	1	12	20	2	885
2005-12-31	2	35	22	2	1005
2006-12-31	2	35	27	4	1068
2007-12-31	2	49	24	2	1179
2008-12-31	3	61	20	3	1329
2009-12-31	4	51	23	6	1635
2010-12-31	5	50	38	4	1586
2011-12-31	4	60	35	5	1643
2012-12-31	5	42	23	6	1765
2013-12-31	6	56	33	3	1802
2014-12-31	3	58	33	8	1986
2015-12-31	3	40	33	5	2123
2016-12-31	6	51	32	6	2150
2017-12-31	2	55	39	17	2239
2018-12-31	3	48	35	16	2231
2019-12-31	4	55	34	8	2419
2020-12-31	3	51	46	19	2525
2021-12-31	2	43	47	9	2651
2022-12-31	4	63	44	11	2836
2023-12-31	4	85	51	24	3429

How long is it since I made a snide comment about conspiracists and their data aversion? Too long. See table above. For those who aren't familiar with the field, "Mordehai Milgrom" is a prominent researcher in "modified gravity" ; "MOND" is a popular theory of "modified gravity" ; "MOG" is a different such theory ; "non-Newtonian Gravity" is more general term for the field ("Newtonian Gravity" being the "Big Science" conventional theory in favour of which all the other theories are being suppressed for (whatever) reason(s) ; "Dark Matter" is just there for a marker of overall theoretical activity in astronomy.

Lots of data. Would a picture help?

Much clearer. Publications in these two particular theories of non-Newtonian gravity have continued to happen at comparable rates to general astronomical activity, though the relative popularity of these two theories varies a little, "MOG" having picked up in the last few years. Mr (Professor? Probably.) Milgrom continues to publish at a fairly consistent rate - which isn't so surprising, since he's been doing so for about 40 years.

That's not the sign of a field of research that is being "suppressed", "forced underground", or even "harmful to researcher's careers". It's a sign of a relatively unpopular topic within a field. Now, "unpopular" may not be a particularly nice state to be in (anthropomorphising "theories"), but it's not a sign of effective suppression. Effective suppression is samizdat publication on midnight press runs, and the occasional publisher's head boiled in tar and spiked over the entry gates to Physics & Astronomy Departments pour encourager les autres (a short-form Voltaire-ism [Candide, ch.xxiii]; the long form is darker than many people realise.)

Having done the leg-work, I'll continue to update the data file yearly (until I get bored). If you think I should be looking at different data, that's what the "Comments" are enabled for.

Back to Contents List.

---

It missed. Didn't it?

HD 7977 and its possible influence on Solar System bodies.

What caught my attention about this? Well, "influence on Solar System bodies" means something really close, but the HD catalogue is of relatively bright stars. Just from the number, I'd think - that is likely a star about 7th magnitude (to 6th magnitude - naked eye limit - there are about 6000 stars). So ... that's not a normal combination. "Attention grabbing", as a newspaper editor might say (followed by Shome mishtake, shurley? from Lord Gnome).

The abstract is more informative : In the latest Gaia third data release one can find extremely small proper motion components for the star HD 7977. This, together with the radial velocity measurement lead to the conclusion that this star passed very close to the Sun in the recent past. Ohhh, right. That's a fun one. See my choice of title. Next question - does the WWW (Wonderful World of Wiki) know about this? Yes. No significant additions needed there.

This paper is about how much effect the passage of such a star (see 'properties' table) has on Solar system bodies, and particularly Long Period Comets (LPCs, because they're relatively loosly bound to the Sun). It's not immediately obvious that passing a Solar mass through the Solar system would have a big effect - the star is currently 250 LY away making for an average speed of 1 LY every 11,200 years. 1 LY is a reasonable approximation to the outer edge of the Oort cloud of LPCs. (Buried in the paper, they imply their "edge" to the oort cloud is at 0.5 pc, which is 1.6 LY. Meh.) More correctly, I should be talking about the sun's Hill sphere. But that's more complicated (partly because for a small "miss distance", the Hill sphere becomes very small. Guess I'd better RTFP.

My inference of brightness from the low HD-number is fair. SIMBAD has 13 identifiers for it, only 3 from GAIA. It's brightness is (as normal) variable across different colour bands, from 9.63 in the B band to 8.89 in the G band. It is situated in Cassiopeia, so is circumpolar from most of Europe and North America. The record has astrometry as far back as 1840 - pre-photographic - which cannot be ignored. The GAIA data releases have a low uncertainty, but they assume the star is a single body. If it is multiple bodies whose images overlap, differently during different observations, then the positional uncertainty could be much higher than the relatively short time span of GAIA observations can detect. Because of this, and the star's brightness, visual and older photographic data must be considered too. (Table 3) Argelander catalogued the star in 1842 and 1875 - at different declinations, from different locations (Finland and Bonn, when he was compiling the Bonner Durchmusterung where it is catalogued as BD +61 250. Further observations were recorded in 1903 (whose plates were re-measured by the author team),the 1920s, 1930s, 1940s and 1950s, then a hiatus until the 1980s, 1990s, 2000s and then the various satellite data. None of the data shows the star to definitely be a multiple star, but the multiplicity hypothesis cannot be ruled out either (remember - about half of all stars are found in multiple systems, so at least one in three of observed stars are probably multiple. No companion has been located on the sky, but a barely resolvable component remains possible.

All this positional work gives revised, and larger, uncertainties on the star's proper motions (in Right Ascension and in Declination), which project back to how closely the star's path would have intersected with the Sun's position. 5 of their 27 position estimates had to be discarded due to inconsistencies in the resultant dataset - which isn't terrible, but doesn't reduce the final uncertainty in the data. Putting it all together, they get a nominal closest approach between HD 7977 and the Sun to be equal 0.011 pc (∼2,300 au) during its passage 2.47 Myr ago, but with uncertainties on all factors (±0.024 pc in distance and ±0.03 Myr (30,000 yr, about 3 times the transit period of the star through the Oort cloud).

The issues identified here were also applied to the rest of the GAIA set of stars which appeared to have approached within a parsec of the Sun, identifying another two possible encounters. One of these, "P0533" (in the Stellar Potential Perturbers Database - "StePPeD") was quite close (0.223 pc) at 3.19 Myr ago, but weighs only an estimated 0.87 M_☉. But the other new discovery (or identification really - it too was catalogued in the 1920s) "P3509" approached to (nominally) 0.402 pc at 3.94 Myr ago, but is estimated to weigh 3.50 M_☉. That is much more likely to have had a noticeable effect.

Having established several new potential close-approaches, and some factors to consider in searching for more, the authors then move on to consider the effects of such a close encounter. Far and away most likely is that the star will pass close to a long-period comet, particularly at it's aphelion when it is moving most slowly and so is exposed to the passing star's gravity for the longest period of time. This could lead to ejection from the Soalr system (which we then wouldn't see) or it could deflect the comet to a closer aphelion). To further reduce the dataset, they only considered comets observed with a good orbit (well, obviously) and those with perihelion outside 3.5 AU (keeping them away from interactions with most of the asteroid belt and inner planets, but not avoiding Jupiter). Their interpretation is that close stellar approaches do increase, appreciably (maybe 40%, for the HD7977 close approach, alone). Several of their detailed study comets show major changes in orbital parameters due to interactions with these stars. But to me, the important point is that they're seeing multiple interactions within the period of one orbit - up to 4 or 5 times in one orbit. So, these interactions become the matter of noise in the record, not a cataclysmic single event. And we know form the geological record that - while major impacts do happen, life as a whole can handle them. It's a deal, but not a big deal.

Do these close encounters affect the planet's orbits?

The most likely planet to be affected is Neptune. It's furthes from the Sun, furthest from Jupiter, and (potentially) closest to the perturbing body. And, yes, it can be disrupted, but not very often. And we already know that it hasn't been. So ... what's going on? Is there an error in the analysis, or in interpretation of positional data (and hence closest approaches, resulting from unidentified multiple stars. And that reminds me that we're still not sure that the Sun is a single star.

I've got a headache now. The issues in astrometry are quite clearly explained. The results of the close aproaches are much harder to interpret. But we know - from the geological record - that bad things don't happen often enough to be important. Time for a break.

Back to Contents List.

---

Flag-washing

link

Totally by coincidence, I'm watching Anthony Hopkins eating people's faces as "Hannibal Lecter". Was the posing of a skinned+gutted cop with Red/White/Blue bunting in a "blood eagle" some way of prodding at American sensibilities? Are they really so sensitive about .. those three colours? The colours of who-knows how many flags around the world? In the words of Marvin (The Paranoid Philosopher Android ; the "clearest thinker I ever knew" (DentArthurDent), "It gives me a headache to think down to that level."

Truely a great comedy - for how it exposes people's irrationality.

Back to Contents List.

---

HTML learnings for this month

So, I'm getting a bit ... ticked off with clicking the "open in new window" option on Blogger's "add link" dialog. So ... I should be able to put ' target="_blank" ' into the CSS style header for my "a" link. No? Need to check on a different page!

OK, well I tried that - at least simplistically with 'a {target:"_blank";}' in the "style" header block . Didn't work ; page opened in the same tag.

Now, why didn't it work?

Table fidding

Do I have unstructures ways to add a comment to a table. Or would a caption do? I've got "caption-side: bottom;" set in the style sheet.

I set up a TFOOT line below the TBODY section, including a TR with COLSPAN set to occupy the whole table.

I need to find a better way to set columns to R-align - better than cell-by-cell, at least. (Stack-Overflow suggests CSS of ... Which looks workable. But it's complicated. It's easier to change the default to , because that's what I'm going to use most often. And it's mucking up updating. Or is the system just broken? This is getting horribly complex - and not working.

Back to Contents List.

---

Happenstance, Circumstance, Enemy Action

Memory in the Burst Occurrence of Repeating FRBs

One of the consequences of having something repeat is that you start to ask "how often does it repeat?". "is the rate changing", "is the rate not changing?", "is the rate noisy?", "is the rate not noisy?". As I used to say to my trainees, with a datum, you've got a datum ; with two data points, you've got a difference ; but with three data points, you can start to do meaningful sadistics." These authors agree. FTFAbstract :

In this work, we analyze the statistics of waiting time between bursts of three repeating FRBs from four data sets.

So, about the sparsest dataset you can hope to treat the numbers with sadistics and get something meaningful out. One of the most significant discoveries in the study of FRBs is the identification of a repeating FRB known as FRB 121102. Wait a minute - that has been repeatiung for a few years now. So how many events does it's sequence have? Quite a bit more data than the sparse set I just described. Sounds like a job for Wiki! Which gives me the 2012-11-02 first-identified burst, ten more identified from archived radio data in 2015. The next batch report was in December 2016 (one having been received on 13 November 2015, four on 19 November 2015, and one on 8 December 2015) - so that's giving a nice range of intervals already. On 26 August 2017, astronomers using data from the Green Bank Telescope detected 15 additional repeating FRBs coming from FRB 121102 More grist. Active little begger, this one! This also sounds like a well-planned observation campaign, because they wrung a lot more information from the data - polarization, and rotation of polarization. Hundreds more burst in 2018 through 2019, with a noticeable periodicity near 157 days. (This is coming back to me. The process of writing notes helps fix things in long-term memory. As any teacher could tell you.) So the data situation is by no means as dire as the introduction made out. Wiki isn't very good at separating the repeaters from the non-repeaters, and the data pile gets a bit confusing. Let's get back to the paper.

Since the identification of the first FRB in 2007, the tally of discovered FRBs has increased to over 700, with more than 60 shown repeating burst phenomena. Right, so the data situation is, indeed, far less serious than I'd inferred from the introduction. People are debating if all FRBs are repeating, but we don't happen to have the right telescope in the right mode and orientation at the right time. I'm sure some people are incorporating testing that into other observation programmes.

Working into the paper, the total data available is actually quite a lot :

FRB data

Source FRB	Burst (count)	Observed (hrs)	Observing run (d)
121102	1652	59.5	47
20201124A(A)	1863	82	54
20220912A	1076	8.67	55
20201124A(B)	881	~4	4
Totals	5472	154.17	160
Two distinct sequences of events at the location "20201124A" were recorded, designated here 'A' and 'B'. The event "20201124A" occurred on that date, at a location on the sky ; it repeated at other dates and times.

The relative abundance of data, spread across a range of locations, fluences (brightnesses) allows these scientists to pick out a scaling law where short intervals are followed by short intervals, and long intervals are followed by long intervals. This is decidedly different to the behaviour of a Poisson process - where the length of intervals is independent of previous behaviour.

This behaviour isn't unique to FRBs. Clustering of events following similar laws is also seen in earthquakes and solar flares (both events where there is a continuous accumulation of strain (mechanical, magnetic) leading to an event that reduces that strain. Which suggests models for what is happening in an FRB. Contrary to intuitive view, the probability of next burst occurring does not increase with the time t elapsed since the last burst; rather, it decreases and the expected residual time to the next burst ̂λ is also increased. […]Furthermore, the fluence of the subsequent burst does not increase with the waiting time but appears to be independent of it. Now that last part is really contra-intuitive. By no means unique (in thinking about earthquakes, a model of the fault "annealing" and getting stronger with time is often used. Of course, being buried in rock renowned for transparency, it is probably harder to see what is happening in an earthquake fracture zone than in something half-way across the universe (FRBs with high apparent redshift have been detected).

This work is one of the first I've seen (which doesn't say much) that uses a lot of data from the new Chinese Five-hundred-metre Aperture Spherical Telescope (FAST), to which I assume the authors are attached.

Back to Contents List.

---

The Role of Giant Impacts in Planet Formation

https://arxiv.org/pdf/2312.15018.pdf

This is why I use Iarxiv : Regardless of the sequence that papers were submitted to Arχiv on the day in question, this one got filtered to the top of my pile by Iarxiv, on the basis that I'd followed links to other papers with similar keywords in the past - clearly a topic I'm interested in. Whatever their scoring algorithm, this one came in at a value of 0.65, when few papers make it above 0.4. If I continued to use the daily email (of all papers, in certain categories in Arχiv), then the papers would be ordered in sequence of submission (or acceptance - yes, Arχiv does do some refereeing of it's submissions). I may block most scripts on most sites, and block cross-site scripting to the likes of doubleclick and facebook, and auto-block a large list of advertising sites ; that's because I don't like advertisers distracting me, not because I don't understand what they're doing. If they do something of value to me, I might choose to accept their services. But if they don't provide anything of value to me, why should I provide them with information they can sell. But that's politics, not science. (Mentioning Iarxiv's use of "keywords", I guess I'd better add a few here.)

"Giant Impacts" were a bit of a shock when the concept was spreading in the 1970s. The appaling power of even regular impacts (Meteor/ Barringer Crater ; then Chicxulub - the "dinosaur killer") was disturbing to people who were getting somewhat used to the idea of thermonuclear annihilation (I grew up within cycling distance of a foreign nuclear base on "Aircraft Carrier 1". Imminent death for other people's gain changes your thinking.)

But the concept of a "Giant Impact" (the capitals seem appropriate) is inherent in the model of planet formation by growth of smaller bodies aggregating onto each other. "Aggregation" means either a relatively small body impacting a relatively large body (e.g. Chicxulub, regardless of it's effects of extinguishing ammonites, and some of the dinosaurs), or two nearly-equal size bodies impacting. Which latter is a "Giant Impact". And hence the interest in the "Giant" part of the impact range.

Worse - if you find the idea discomforting - is that most of the growth in size of planetesimals to become planets, comes from the mutual impact of similar-size bodies. If two bodies collide, one ten times the diameter of the other, then their ratio of masses is approximately 1000:1. So, to get the growth effect of a giant impact (with 50% mass loss) on the larger body, you need O(500) ("order of 500" of the smaller impacts? (Yes, the efficiency of accretion is an important factor. At least, if you're in the business of building planets.)

OK, enough general maundering. You can probably get much the same from Wiki (but see my caveat). Let's see what the "Annual Review of Earth and Planetary Sciences" has to say on the matter. This year.

First, RTFAbstract : their first substantive point is that the Giant Impact (hereafter, "GI") that formed the Moon is probably the one we have most information about. We have samples definitely from the resultant debris (a programme called Apollo, IIRC) ; we have more arguable samples from the target body (Earth has active geology ; it's accessible surface is almost certainly different to the surface - and depths - during the GI) ; we even know where the target body and debris are now, probably 4+ Gyr later (though they may have moved, significantly, in that time interval). Contenders for the second best known example might be the Pluto-Charon forming impact, or the (suggested) GI that gave Mars it's hemispheric dichotomy. (My caveat : the Wiki article is about specifically the "Moon"-forming impact, while it's title suggests it's more general application. Somewhere below I'll detail my compilation of data about which Solar system bodies have evidence suggesting a late-stage GI in their formation, but the total is that about 6 of the 11 largest bodies in the Solar system have GI-suggestive features.

Section 1 - Introduction

In the "Introduction" they introduce alternative terminology, and point at an interesting wrinkle : the common monikers are pairwise accretion and similar-sized collisions; aside from gas giants, there is little distinction between the terms similar size and similar mass. The wrinkle is that, because of internal compression, the mass of gas giants is complexly related to their mass. Contrary to intuition, as mass increases through Saturn size to Jupiter size, the radius increases, but somewhere not far above Jupiter mass, internal compression becomes more important and as the mass increases, the radius decreases. This continues (probably, by modelling) until about 80 Jupiter masses, when the core reaches the point of thermonuclear ignition. (What happens with temporary deuterium-powered fusion is an uncertain question.) That's an interesting wrinkle, but not really important to GI. The paper discussed above, about the impact of "sub-Neptunes" on each other is also very relevant in this - as you move from "rocky" planets ("planetesimals", "protoplanets", or even "small bodies") into the realm of "ice-giants" then "gas giants" the changing density profile through the bodies does result in changing impact behaviour.

Figure 1 : Pretty picture. For certain values of "pretty", probably best involving a very good telescope. Credit Andrew Gonzalez

The introduction mentions the important role of the (mutual) impact velocity (and hence kinetic energy and angular momentum) in analysing these particular impacts. And, as anyone who did ball-bearing on ball-bearing calculations at school (or played snooker after school), they're essential in rigid body collisions too. In planetary calculations, the velocities are generally normalised compared to the (mutual) escape velocity of the impacting bodies, v_esc.

It is probably worth noting that most computational work in this area (all? possibly ; I can't remember hearing of another technique) is done with "smoothed particle hydrodynamics" (SPH) where the bodies are treated as collections of (relatively small) particles which have their own mass, volume, density, "stickiness", rigidity and gravity and they are thrown about in the computational arena until they either adhere to each other, or go beyond the computational limits of the modle. Mathematically simple (literally, you probably did this at school), but involving millions or billions of calculations per step of the model. Big number crunching like this used to need supercomputers ; modern recreational graphics hardware (GPUs) makes it a much cheaper task. I wonder if someone has tried to farm such work out to BOINC?

Giant impacts can result in many outcomes (see the "sub-Neptunes" paper discussed above), with impact factor (offset from a head-on collision) and collision velocity (comapred to v_esc) as major determinants). The impacting bodies can merge almost fully (Venus, Uranus?), merge with debris (Earth-Moon, Pluto-Charon?), or be completely disrupted (myriads of asteroid belt examples?). Part of one body can be lost, and the other either lost or disrupted (Mercury, and it's missing mantle?). And the largest range of outcomes is classified as "hit and run", which results in some loss from both impactors, but both surviving (this is not much help when you're trying to build a larger planet).

Section 1 of the review discusses these ranges of outcomes, and the methematical underpinnings of the analyses. It's a review.

Section 2 - modelling

Section 2 discusses aspects of modelling. One thing that is reported that hasn't been much discussed previously is that if the impact velocity is higher than the speed of sound in the materials, thne the collision is supersonic, and understanding these events demands accurate shock equations of state - which is potentially amenable to laboratory investigation, involving big guns and exciting booms. Figure 4 puts this into context of impactor size and material - collisions involving Earth-size bodies are probably supersonic for forsterite (an olivine mineral, an approximation to mantle materials) but Mars or Mercury-size impactors are probably subsonic for forsterite, but supersonic for solid or liquid water. Asteroid belt size bodies, on the other hand are probably subsonic even for bodies rich in liquid water. That is news (to me).

This last is more relevant as collisions in the early Solar system are likely to have involved bodies that are relatively warm (compared to today) due to the presence (in the early Solar system ; this may not be a common situation) of short-lived radioisotopes such as ²⁶Al from a recent nearby supernova. The isotopic signatures of this can be detected in Solar system specimens, but detecting such evidence remotely would be ... "challenging".

The presence of liquid water and the potential melting of solid water is important because it is relatively effective at converting impact energy to thermal energy. But it doesn't dispose of angular momentum. That makes the modelling of the impacts of particles in "SPH" further from inelastic "ball-bearings" and into the realm of soft, deformable bodies - needing a lot more computational grunt in your simulation. 20 years ago it was a problem that had to be ignored ; these decades, it's a problem that has to be faced - billion-fold.

Section 3 - Consequences

1- Well, duh, that's going to have consequences. Of course.

The contra-implication is that planets didn't form by myriads of "small" accretions. Which is something that you could reasonably argue for one, or a few, planets. But at some point it has to be faced for medium-size planet(-esimals). But, at some point, even if there were a mystical way of avoiding similar-size bodies from interacting, you'd end up with some "similar-size" bodies interacting. The issue can't be avoided.

Giant impacts in pebble accretion theory. Yes, certainly "pebbles" will interact, and accrete. The effects of within-particle cohesion and viscosity will lead to larger particles, eventually resulting in the similar-size interactions which are the subject.

Frankly, "pebble accretion" isn't a problem.

The unavoidably stochastic nature of "hierarchical growth" also, naturally, leads to a considerable diversity of outcomes from each collision, and each chain of collisions.

It may be disturbing to some that seemingly simple interactions following well-understood laws of physics can lead to uncertain outcomes. But that is the case. The word "seemingly" is important in that description.

Section 4 - Conclusions

Giant impacts are ubiquitous. As planetary systems evolve, the proportion of protoplanetary embryos (considered as approximately Moon-mass bodies, grown by pebble accretion or some other process, including smaller-scale GI) rapidly decreases until effectively no non-GI bodies are left. Bodies that are the products of a chain of ten, twenty or more GI are intrinsically possible. Reduction of numbers of planetary bodies is the only thing that stops this process.

Back to Contents List.

---

Large planets may not form fractionally large moons

Arχiv Paper

In this context a "fractionally large moon" is one, like Earth's Moon, which is a significant fraction of the planet's size (3474.2 vs 12742 km diameter, 0.272 or 1/3.66 ratio ; 0.0202 oe 1/49 in volume).

THey start with an assertion I've always treated with considerable suspicion : that The Moon stabilizes the Earth’s spin axis at least by several degrees and contributes to Earth’s stable climate.Well, since Mars - without that "stabilisation" has obliquity (rotation axis to orbital plane angle) rate of change which can reach a degree per 10 kyr - which is should be within the capability of life to evolve (or migrate) around. As for how "necessary" it is ... well both Venus and Mercury are thought to be stabilised by interaction with the Sun. So how "necessary" is the Moon to life on Earth? (While checking my prejudices on this, I notice that significant Earth-obliquity: Moon resonances may occur in the next few Gyr, which also challenges just how important a "large Moon" is to life on Earth. (reference : Ward, W.R. (1982). "Comments on the Long-Term Stability of the Earth's Obliquity". Icarus. 50: 444 - 448 doi:10.1016/0019-1035(82)90134-8. ; It's open access, so go get it.)

Sorry, but I'm used to SF authors asserting that "The Moon is necessary for life to have evolved on Earth" or even that a "freakishly large Moon is essential to the development of life anywhere". Well, that may be good enough for an SF author, but I am far from convinced that it's a securely-founded opinion. Sorry - pet rant.

The gist of this work is that vapour rich planets involved in a giant impact (GI, see previous discussion) will produce a vapour-rich debris disk, within which many of the debris fragments from the collision will experience gas drag and re-accrete to the primary rather than forming a "fractionally large" satelite. To get a "fractionally large" satellite, you need a relatively large impact energy or low-Vapour Mass Fraction (VMF) impactors.

Which is well enough. There probably isn't enough information to say exactly what happened in the formation of the Solar system - but we do know that (1) it happened, and (2) models that can describe it happening are reasonably easy to construct.

Back to Contents List.

---

End of Document

I think that maybe I should try a different construction and make one page (post) per paper.

Back to Contents List.

2023-09 September - 12 December Science Readings

Yeah, well for several months, I've done damned-all. Try and get back into harness. I've been thinning out the Arχiv paper listings, even if I couldn't bear to get to work on them. So let's see what remains in the pile.

2023-12-20

Over a week without doing anything. Try to get back into the habit. The HD110067 paper stuck it's nose out at me because I recognised the HD number. Which isn't normal, even for me. It's the system which was recently reported as having 6 planets at a "perfect spacing". Which is a rather problemtic description. Anyway, follow the link into the main record.

New Year

Get a few more done.

Contents

Articles studied this September to December - some of which might go to Slashdot.

Forming Asteroid Moons

MOND Count

Star Formation in the Cartwheel Galaxy

December HTML things

The "Cosmic Train Wreck" Galaxy, A520

The abundance discrepancy in ionized nebulae: which are the correct abundances?

6-planet systemis a multi-star system.

Dating of a Latin Astrolabe.

Early Galaxies initial Initial Mass Function

End of document

---

2023-09 September to 12 December Science Readings.

Revisiting Dimorphos formation: A pyramidal regime perspective and application to Dinkinesh’s satellite

https://arxiv.org/pdf/2311.07271.pdf

Last year NASA hit an asteroid moon with a spacecraft to test the efficiency of momentum transfer into asteroids. Being an (astronomically) small body orbiting in an (astronomically) weak gravitational field, the (astronomically) small amount of energy the NASA can afford to pump into such a system could produce detectable effects. The "mainstream media" didn't generally understand the point of the exercise, which was not to test if there would be an effect - everyone involved was sure there would be an effect - but to test how efficiently momentum was transferred to the body. That was a very open question, because astronomers were (and still are) pretty sure that many PHAs (Potentially Hazardous Asteroids) are "rubble piles" of many smaller bodies (from dust, up to mountains) held together by their own gravity, and were expected to be very weak. So - if hit by (say) the force of a megatonne nuke, it was very unclear if 90% of the energy would go into accelerating 99% of the asteroid's mass, or if that 90% would go into accelerating 1% of the asteroid's mass, and only 10% MT go into accelerating the main part of the asteroid away for an impact situation. That's quite an important question if we were ever faced with solving the problem in the coming year.

Anyway, that was the "why" of the mission, and a reasonably useful reesult was obtained (IIRC, about 1/3 of the momentum applied stayed in the "rubble pile, resulting in the moon's orbit being changed by about 30s - which was easily detectable. But this paper in't about that - it's about how such asteroid moons can form.

Available scenarios include : forming together (in which case, why didn't the last parts of the accumulating debris go onto the larger body, with it's stronger gravity field), capture (which Darwin (George) showed is actually quite difficult, unless you've got a "third body" to take away some of the energy and angular momentum from the encounter), impact on the main asteroid (as for the Earth's Moon, most likely ; but why then are asteroid moons so common?). Or, as this paper proposes :

[An asteroid can] deposit material into orbit via landslides.

Now, there's an idea. It doesn't sound right. If the energy comes from the movement of material under gravity, how can that material get into orbit? The trick here is that this was happening while the primary was having it's spinn increased by the Yarkovsky–O'Keefe–Radzievskii–Paddack (YORP) effect, and that effectively reduces the weight of materials near it's surface (far from the centre of spin). And that is enough to allow some of the materials disturbed by such movements.

Which sounds very weird. But that is how weak the gravitational fields are on the "small bodies" of the solar system.

The paper has more work about the time sequence of events they infer leading to the formation of this moon, and in this shape. We're probably going to see a number more examples of this as the "Lucy" mission works through it's list of asteroid fly-bys.

Back to List.

---

MOND Count

Formation and Growth of the First Supermassive Black Holes in MOG

A while ago, I did a count of papers posted to Arχiv over the last few years of papers on the subject of "MOND" (MOdified Newtonian Dynamics") because someone had posted (again) the false claim that scientists don't look at alternatives to the main theories of science. They do get attention - just not a lot. Previously I'd searched for "Mordehai Milgrom", "MOND" and "Non-Newtonian Gravity". "MOG" is another search term to watch out for. My previous searches have looked at yyyy-12-31 as a cut-off date, so I'll carry on with that.

Hmmm, moderately interesting - at current publishing rates, Mordecai Milgrom looks likely to retire (or stop publishing - same thing, really) about 2030~31.

Back to List.

---

Formation of the Cartwheel Galaxy

Star formation history of the post-collisional Cartwheel galaxy using Astrosat/UVIT FUV images

The Cartwheel Galaxy is one of the prettier - and weirder - sights on the sky. To the telescopic eye, one can see an outer ring of bright, bluish stars, and an inner ring of slightly redder stars around the margin where approximately the edge of the nucleus would be, and a set of slightly curved "spokes" faintly link the two main structures.

It's a pretty thing, shouting a really loud "How?" across the universe across to us.

The general understanding of this galaxy's structure is that another galaxy passed, nearly centrally, through the "Cartwheel" galaxy. In 1996 radio telescopy detected a discontinuous band of neutral hydrogen emission stretching from the Cartwheel towards a nearby galaxy, "G3", which is interpreted as the impacting or "bullet" galaxy. Critically, this emission structure extends beyond the other nearby galxies, "G1" and "G2". A wider view of the system than the "pretty one" shows the relationship of the galaxies: (this is an image from the DSS - Deep Sky Survey - but as a negative image) :

This paper reports studies on UV emissions and brightness to estimate the star-forming rates of the Cartwheel's brighter regions, and infer the time since their formation. So we now know that the galaxies were in collision at about the start of Earth's Carboniferous period, 300 million years ago. With some modest error bars.

Back to List.

---

December HTML things

link

What got me prompted to try to get back to work was seeing a reference to an unusual bit of HTML CSS - the "rotate" property. With this, you can ... go on, guess ... rotate text (well, strictly, a division of a page) about a point.

So, naturally, I wondered, will it work in Blogger's sandbox? Let's try.

I'll start by using my handy-dandy bit of table template :

Heading	Column
a	Α

And now, I wrap it up in the "div-Wol-Rotate" division I've defined up in the page's style header.

div-Wol-Rotate { height: 100px; width: 200px; background-color: lightgrey; rotate: 1 1 0 60deg;

Not working. 3 axes? no Angle? No (and the bgnd-colour isn't working either) Ah, whoops, missing trailing ";" Size? Nope. Still doesn't seem to be working. Quick dive over to the "try it" sandbox at W3schools.com ... and the code above works there, but not here. Deduction : Blogger doesn't like rotation. Oh well

I picked up this question from reading Elf Sternberg's blog, about recreating the 1994 alt.sex FAQ website in modern coding. Now, I don't know how often Mr Sternberg gets people linking to the programming bits of his website, but just so he knows, from the above - people don't only read his website one-handed.

I can't claim to have fully understood the "rotate" property - the height, width and definition of the rotation axis interact in some way I need to look at to shear the image before the rotation is applied. I need to work on that. but it doesn't seem to work here, so I'll shelve it for now.

Fiddled with the CSS for HR elements though.

A bit more mucking around with tables in IMF article.

Back to List.

---

The Cosmic Train Wreck Galaxy, Abell-520

How Zwicky already ruled out modified gravity theories without dark matter,

This 2017 paper just jumped out at me while doing the counts of "MOND" and "MOG" and "dark matter" papers while checking out the question above.

It has interesting bits. Who could resist a phrase like "The Cosmic Train Wreck Galaxy"? Not me - that's for sure. It sounds like something "fun" happened, in the sense of "may you live in interesting times", and "best observed using a telescope". Even better that it's in a context of overheated extravagant astronomers, not staid, restrained stage actors who couldn't emote if you pulled the stick out of their back, leaving splinters. Talk dirty to me, like a central 3-4 × 10¹³M_☉ mass clump with mass-to-light ratio 800 × M_☉/L_R☉. i don't know how dirty, but that's definitely dirty talk! (There's probably a typo - "L_R☉" for "L_☉" ; I'm sure I'll find a resolution somewhere.)

OK - how dirty are we? 800 Solar masses for each Solar luminosity? Let's see - Sol is heavier than about 90% of stars, but the decrease in luminosity with decreasing star mass is steep. A half-M_☉ star has a luminosity of 0.063 × L_☉ (about 1/16th) while at the other end of the mass scale, the luminosity ramps far more steeply than the star mass. A 20 × M_☉ star shies with luminosity 53600 × L_☉. So, calculating the mean brightness of matter on a gelactic scale is .. hairy. It's not clear to my limited maths if it'll be a figure above or below 1 × M_☉/L_☉. Probably not wildly off from 1 × . So 800 × is a galaxy with remarkably little light for it's mass. Something seems to have wrecked the galaxy's mechanism for turning mass into luminosity. No wonder people noticed them (once they'd done their accounting).

The "Train-Wreck Galaxy" has 4 references linked ( 13 M. Jee, A. Mahdavi, Hoekstra et al. The Astrophysical Journal 747(2), 96 (2012).
14 D. Clowe, M. Markevitch, Bradaˇc et al. The Astrophysical Journal 758(2), 128 (2012).
15 M. J. Jee, H. Hoekstra, A. Mahdavi, and A. Babul The Astrophysical Journal 783(2), 78 (2014).
16 Q. Wang, M. Markevitch, and S. Giacintucci arXiv preprint arXiv:1603.05232 (2016).) but the rather more accessible Wiki article suggests that later observations have picked up considerably more luminosity, reducing the seeming unusualness of the galaxy cluster by quite a lot. It's still odd, but less odd.

The other list of cosmological oddities is worth looking at too. As they phrase it, so Lambda;CDM is rightfully a good effective theory. But is it a fundamental theory? Many puzzling observations make this nonevident. :

1. The CDM particle, the WIMP, if it exists, keeps on hiding itself 6 more years since its “moment of truth” ;
2. One observes 19 quasars with spins aligned with their hosts large-scale structures on a scale of almost 1 Gpc (ref.10) and dozens of radiojets from AGNs aligned on a scale of 30 Mpc (ref.11) ;
3. A ring (actually, a spiral) of 9 gamma ray bursts extends over nearly 2 Gpc (ref.12) ;
4. the "train wreck galaxy" that caught my attention above (which may not be so weird) ;
5. in the cluster A3827 the offset between baryonic and dark mass (ref.17) is an order of magnitude ‘too large”(ref.18) ;
6. Puzzles in galaxies include: the brightness fluctuations in the Twin Quasar allow a DM interpretation in terms of a large halo of rogue planets in the lensing galaxy (ref.19) ;
7. the observed satellites of the Galaxy lie in a plane, not in random ΛCDM directions (ref.20) ;
8. the predicted transition for the most massive galaxies to transform from their initial halo assembly at redshifts z = 8 − 4 to the later baryonic evolution known from star-forming galaxies and quasars is not observed (ref.21) ;
9. the galaxy power spectrum deduced from SDSS-III observations fits well to the stretched exponential exp[−(k/k_b)^1/2] from turbulence (ref.22) ;
10. Various further arguments can be found in our investigations (refs.23,24) ;

That's an interesting list of "problems" with the ΛCDM model, while acknowledging that it is still a good model. No joy there for Archimedes Plutonium and the Electric Universe crowd of the deluded. Frankly, while they're interesting, they're not interesting enough. Just worth me noting down, to counterblast people claiming "censorship" and "suppression" of unorthodox thinking.

Why did Zwicky think you needed dark matter, not MOND/MOG, in the 1940s?

On a third reading, I still don't get the argument the authors are making. They seem to think that, for the cluster they study (Abell 1689 ; the mention of Abell 520 is just in the "problems with cosmology" section of the introduction ; it doesn't help that this is a fairly early draft, with several arguments re-written in the text.) their calculation under several generic "modified gravity" models, still need to add some non-visible mass to the system to match the observed lensing [strong ("SL" with multiple arcs for identifiable background galaxies) and weak ("WL" with shape distortion of individual background galaxies)]. Their preferred (why?, not clear) DM candidate is neutrinos. Particularly "heavy" neutrinos with a mass of ≦ 2.0 eV. Most experimental estimates for conventional neutrinos gives them a mass of ≈ 0.1 eV - a substantial difference, but some particle physics ideas produce "sterile" neutrinos in this mass range.

I think they need to make their point on Zwicky clearer - the title is attractive, attention-getting. But I can't follow their reasoning.

Back to List.

---

Element Abundances

The abundance discrepancy in ionized nebulae: which are the correct abundances?

A long-standing question I've had is : what is the average composition of a planet (asteroid, "small body"), versus this particular surface, or an ore body. There are a lot of slightly different answers in the literature, and not a lot of clues pointing to the most-accepted (and most detailed) answers. Slightly different, but related, answers often given for this question include :

• "What Is the Composition Of ("WICO") the Sun?" (not necessarily the same as the average planets)
• "WICO the Sun, if you lose the H and He?" (ditto)
• "WICO a CI meteorite? Is it the same as their parent bodies before impact heating?"
• "WICO the Earth (including the core)?"
• "WICO the Earth's (or any other planet's) surface?"

They're all quite similar questions, with not quite the same answers. And the differences may (or may not) be significant when you're trying to consider the changes implied by formation processes. Obviously, I'm looking for a full, 92 element composition. How it's expressed is an issue to be wary of - a short while ago I had one of those "drowning goldfish" moments when someone was talking about "atmospheric carbon di-oxide being absorbed by the lime (calcium oxide) present in basalts". It took a time to realise that he was reading the traditional geochemist's way of reporting compositions as "oxides" (e.g. 40% SiO₂, 10% CaO, 10% Al₂O₃ …), and taking that to mean that a basalt contains grain or molecules of actual CaO, with all the chemical properties of a lump of quicklime. Which ... well, I can see why this relic from gravimetric "wet chemical analysis" of the 18th century is confusing. But it obviously can lead to profound misunderstandings. Gotta watch that banana skin.

I'm slightly surprised that my confused correspondent didn't notice that the trace elements are normally reported as "ppm of [Element]", rather than in the "oxides" presentation. I suspect I'm going to have to address this in more detail one day. It may explain a lot of confusion about geochemistry.

So, interesting paper title, given that context. Does it deliver?

OK - first thing is to see the paper as a contribution to a programme called "Planetary Nebulae: a Universal Toolbox in the Era of Precision Astrophysics" - which isn't terribly optimistic for my "geochemistry" interests. Still useful enough though - what a star spits out this gigayear is likely to affect the composition of next Gyr's molecular cloud, hence stars and attendant planets.

From TFA : "However, the heavy-element abundances derived from collisional excited lines (CELs) and recombination lines (RLs) do not align." - which is what they mean by an "abundance discrepancy". Going back to the "oxides" presentation discussion above, the technology of measurement influences the precise data collected, and how it's expressed. Specifically, since they're measuring line strengths in spectra, I'd be surprised if they presented their abundances as "oxides".

Actually, it's quite a big effect. From the paper's introduction :

However, there is a problem, and it is a big one: we are not certain about the correct abundance of heavy elements like O, C, N, Ne, collectively known as “metals”. Since the pioneering works of Bowen & Wyse (1939) and Wyse (1942), it has been known that the abundances of these elements determined from the bright collisionally excited lines (CELs) (e.g., [O III] 5007, 4959) are systematically lower than those inferred from the faint recombination lines (RLs) (e.g., O II 4649, 4650). The ratio between both estimates is known as the ”Abundance Discrepancy Factor” (ADF), and it has been found to be around a factor of 2-4 in H II regions (García-Rojas & Esteban 2007) but can reach values of more than 500 in some PNe (Wesson et al. 2003).

A discrepancy of 2~4×, but up to 500× - yep, that's a problem. But, it's an astronomer's problem, not really a geochemist's (or planetary scientist's) problem, so I think I'll just file that as a detail to watch out for.

Annnnd … the paper is discussing the various ideas for explaining those discrepancies, but not presenting an example composition for any of their target objects. Ho hum. That just leaves me with the materials I've already collected in my "compositions" collection ( http://ArXiv.org/pdf/2105.01661v1.pdf which gives Solar photosphere and CI chrondite compositions for 83 elements ; 54 terrestrial rocks from "Science Snippets" on my HDD ; … ; ; ; ) and anything I add to it. Back to the grindstone. I still don't have estimates for the average composition of the (local) universe. Granted that the outer arms of the Milky way are second (or third) generation stars, enriched by at least one (possibly two or even three) previous generations of stars burning, exploding (for the larger ones) and contaminating the (primordial) interstellar medium ("ISM", H, He, trace Li) with "metals" which change the structure and nuclear processes of the stars. Probably I'd be looking for a local Milky Way composition, a Solar system one, and maybe a composition for an estimate first generation ISM.

Back to List.

---

6-planet system is a multi-star system.

HD 110067 is a wide hierarchical triple system

A couple of weeks ago there was a moderate amount of news fuss over the announced discovery of a planetary system containing 6 planets whose orbital periods were at small-integer ratios (2:3 ; 3:4, etc). News coverage was all over the place. Which was all very well and good, but the source paper was a deal more nuanced than the write-ups in the popular press (well colour me sideways and brush me with a Blackadder-shaped paint pot!) Most was made of the small-integer orbital period ratios - which to me (and most people) talks of the system having gone through a phase of planetary migration that carved these ratios - probably when there was some degree of "viscosity" in the system to damp interactions to some degree. (all that means is - considerable remaining gas, or many "small bodies", being present to diffuse energy and angular momentum less violently than by collisions or close approaches). The more interesting thing, to me, is how a relatively sparse set of observations were translated into "six planets". Although the reviewers at Nature must have challenged this, I'm sure they were convinced by the data in a way I can't (and am not interested) to challenge. I accept that they've got at least some of the planets adequately located.

This original paper was published in Nature, (here, titled, "A resonant sextuplet of sub-Neptunes transiting the bright star HD 110067", which contains substantial "Supplementary Information"), but the paper was also released on Arχiv without the access restrictions of Nature (but also without the "SI"). You should be able to judge the paper on it's own merits, but the SI is probably in large part a response to the reviewer's criticisms (or their anticipated criticisms - same difference really). I'll refer to this as the "Nature paper", with the titular paper of the "triple star" system being referred to as the "Arχiv paper".

The Nature paper reports several long-period observations of the target star (HD110067 - "HD"= the Henry Draper Catalogue and a reference number) using the TESS (Transiting Exoplanet Survey Satellite) observatory in two sessions, two years apart. The first observing session recorded 4 transits (3 of which had a common spacing - indicating one planet ; one unconstrained transit). The second session recorded 9 more transits, which confirmed the first candidate (HD110067 b, period 9.113678 days) and a second (HD110067 c, period 13.673694 days)candidate observed in both sessions. Leaving 8 unexplained transit events - which was sufficient justification for observing periods on CHEOPS (CHaracterising ExOPlanets Satellite) of fractions of a day around possible repetitions of one or more of those potential objects. In considering potential observing times, the negative evidence - that neither of the TESS sessions had recorded transits except at the noted times - which constrained the possible orbital periods associated with those transits. This series of observations confirmed HD110067 d (period of 20.519 days). Those periods (bc ; cd) were both close to a "Mean Motion Resonance" (MMR) ratio of 3:2, justifying further data collection.

This is where the "chain" of logic starts to get a bit thin. There was adequate justification for targetting more observation time, and some of that justification is in the "SI" of the Nature paper. Another planet was discovered by analysis of one of the transits, for which there was an absence of a second transit in the other TESS data set. This was identified as HD110067 e (period 30.7931 days, in a 3/2 MMR with planet d) - but only on the basis of two transits. That left two unexplained transits in the TESS data. With the assumption that these represent more planets, which continue the chain of low-integer period ratios. Which is how they proposed two additonal planets, which should have been seen in the original TESS data. But the periods they should have been there were also when the Moon or Earth were also near the satellite's line of sight (TESS orbits between a little above geocentric orbit, and a little below Lunar orbit), so originally these records weren't analysed because of glare. When this study gave adequate reason for study, ... there were the expected transits, in the (rather noisy) data. So that gave confirmation of HD110067 f (41.05854 days) and HD110067 g (54.76992 days). Which is sufficient for the reviewers and for science - but it's a lot thinner than the stereotype of multiple transits per planet.

That is how future planet identifications are going to go. If we're looking for evidence of "Earth-like" planets, you're not going to get the stereotypical multiple tranists at a rate faster than 1 observation per planetary "year". So inevitably, we're going to see increasing numbers of single-transit detections, and longer chains of inference from raw observation to deduced planet (and their properties).

Some work was also done using ground-based observations, which revealed a fair amount of magnetig and radial-velocity noise in the system, but some support for the planetary system too.

The inferred planet properties are nothing wildly unusual. They're close to the star (necessary, to find multiple transits of the first few members of the chain), at average radii (of 0.0793, 0.1039, 0.1362, 0.1785, 0.2163, 0.2621 AU) all within Mercury's orbit compared to the Sun. The deduced surface temperatures (from the star's temperature of 5266 K, and the orbital distances) are high, but not extreme (800, 699, 602, 533, 489, 440 K); a short distance out from the detected planets would be the "habitable zone" of the star. The star is estimated to be about 8 billion years old - about half way through it's estimated 17 billion year lifespan, from it's 0.798 solar-mass mass. The planetary masses (5.69, <6.3, 8.52, <3.9, 5.04, <8.4) are in a range not found in the Solar system, but actually very common in exoplanet studies.

The triple-star aspect

The second, "Arχiv" paper is also an event likely to recur. While the first group had been doing their transit-based study, other data was present in already-taken databases. The Nature paper took the primary as being a single star, it had been classified as a "wide binary" since at least 2001 (references in the Arχiv paper) using positional records from 1893 to 2015 (that must be a later revision of the "Washington Double Star Catalogue" than the 2001 reference). To my moderate surprise, the listed companion is also in the HD Catalogue, and in the original numbering system (number 110106), which implies it is quite bright, quite close, and could be spectroscopically investigated at the turn of the last century. That clearly shows that the two stars are indeed, a wide binary, separated by 415 seconds of arc, which at the star's distance (32.25pc, 105.2ly, from Gaia data) translates to some 13394 AU - which is very comparable to the spearation of Proxima Centauri and Alpha Centauri A+B. To further the similarity of the two systems, both Alpha Centauri and HD110106 are double stars - though Alpha Centauri can be resolved by telescope, while HD110106 hasn't (yet).

Since these secondary stars are well out of the line of sight to the primary, with it's resonant planetary system, they don't interfere with the detection of the primary's planets. The presence of a wide binary companion (s) do raise questions about the planet-forming process around the primary, so it's a safe bet that more investigations of this system are planned. You heard it here first.

Back to List. 2023-12-20

---

Dating of a Latin Astrolabe

https://arxiv.org/pdf/2311.17966.pdf quote mark

This is just a piece of fun. Ha-ha, but serious fun.

An astrolabe is an astronomical observation instrument consisting of a plate (the "mater"), a targeting sight bar (the "alidade"), a "plate" engraved with the coordinates of the celestial sphere, a movable metal net ("rete") which indicates the position of various stars, and a suspension point and vertical bar ("rule", which keeps it aligned with local vertical). It can be used for surveying, for measuring the altitude of heavenly bodies (and hence, local time), and of course identifying heavenly bodies (particularly since they normally had a celestial chart engraved on one face). Mediæval users suggested that there were a thousand uses, many of which were listed in a manual on the device written by Geoffrey Chaucer.

One of these days I'm going to find one in a second hand shop (I've got a sextant somewhere a working toy) and learn how to use it.

By pointing the sight ("alidade") at an identified star, and aligning the net ("rete") to match the star's indicator to it's altitude, one can work out the local time from the celestial grids enmgraved on the plate ("mater"). Other ways of solving two knowns for an unknown can also be done with the instrument - including (for the many Islamic users) locating the direction of Mecca in order to pray towards it (the "quiblah". [Totally tangential, I wonder if that was the inspiration behind Rowling's "Potterverse" equivalent of the "Fortean times", the "Quibbler".]

That's the mechanics. The astronomy, by which the astrolabe's manufacture (or design, or calibration) date was determined is that the precession of the equinoxes (rotation of the Earth's spin axis compared to the celestial sphere) moved the position of the stars relative to the celestial grids, allowing the determination of the manufacture (or design, or calibration) to approximately 1550 (CE ; all dates are CE, not AH).

The reason I specify "manufacture (or design, or calibration)" is that once soneone has made one of these, then the additional work to make a second is relatively small, separating "design" and "manufacture" events. The way of indicating star positions (extensions of the "rete" as long pointers, with the star being at the tip of the pointer) is potentially amenable to being fine-tuned with a pair of pliers, so a 1550 design might still be modifiable for use in 1650 (or a 1450 design for use in 1550). In fact, if the "rete" were made by casting in bronze, the raw casting would very likely need some cleaning up, and adjustment because castings - particularly reticulate ("net-like" ; same Latin root) castings often distort as the cool from molten bronze to room temperature. It's very unlikely that the raw casting would have been flat to the plane of the "mater" without some tweaking.

The pictures in the paper aren't high enough resolution to see if the "pointers" have been tweaked. But I bet they have. (when I posted the full-resolution image, I looked more closely - the "rete" is made of several pieces, which look to be cast and have then been soldered together. Yep - that almost certainly needed "calibration" after soldering together. Whether that was higher precision than this naked eye instrument would have needed ... maybe not. But it's at the limit.

Just a fun bit of sideline stuff. For me. Isn't this a thing of beauty?

Caption : Figure 6: The front of the astrolabe, with a plate, the rete, the alidade (diagonal bar), the rule (vertical bar).

Back to List.2023-12-21

---

Initial IMF

Nebular dominated galaxies in the early Universe with top-heavy stellar initial mass functions

That's a convoluted name for a relatively simple idea, with an awful lot of consequences.

As JWST has rapidly improved our state of knowledge of galaxies in the early universe, there has been an increasing tension between the inferred properties of those galaxies. What we can observe are luminosities at different wavelengths. But to convert those observations into a model of stellar populations, ages, brightnesses and interactions, we need a way to convert those luminosites into a number of stars, at a selection of brightnesses. This is not simple, because the brightness of a star does not relate simply to it's mass. The relationship is far from linear.

Which, frankly, has been known since the start of astrophotometry - the (reasonably) accurate measurement of star brightness - in the 1870s. When that was combined with positional measurement (revealing parallax, for relatively close stars, and from that the distance to the stars), and analysis of eclipsing binaries (which gives disc sizes, relative brightness in a pair, and the absolute masses of the stars in a pair), a relationship between a star's mass and it's brightness (when on the main sequence - where stars spend most of their lifetimes) emerges. But it's not a simple relationship.

In my handy-dandy pile of astronomical data and calculations, I put together a handful of ranges of mass to cover the (main sequence) brightness of stars in the Milky Way.

Stellar Mass - Brightness relation (referenced to Sol)

Star Mass (M☉)	Luminosity (L☉)	Main Sequence lifetime (T, Gyr)	Comment
0.1	0.001	1000	Quite approximate
0.2	0.006	333	Luminosity ≃ 0.23 × (mass ratio)2.3
0.3	0.014	214
0.5	0.063	79	Luminosity ≃ (mass ratio)4.0
0.7	0.24	29
0.9	0.7	13
1.0	1.0	10.0	(solar estimate)
1.3	2.9	4.48
1.8	9.4	1.86
2.0	16.0	1.25	Luminosity ≃ 1.5 × (mass ratio)3.5
2.5	37.1	0.6739
5.0	419.3	0.1192	Gyr
7.5	1733	43	Myr
10	4743	21.1	Myr
15	19600	7.5	Myr
20	56670	3.73	Myr, Luminosity ≃ 3200 × (mass ratio)
30	84000	3.57	Myr
40	152000	2.63	Myr
50	240000	2.08	Myr
75	550000	1.37	Myr
100	980000	1.02	Myr
I have two references for this : arxiv 1710.11134.pdf and "The Astronomical Journal, 154:115 2017, Table 1" ; it's likely that the exact numbers will change with time, but not by much.

Already the lower rows of that table are decidedly uncertain. Massive stars (more than a few solar masses) typically lose more than a few percent of their mass during their evolution (through "solar wind", strong coronal mass ejections and other processes) meaning thir luminosity and lifetime will vary considerably from the initial expectation, which is based on nuclear physics. But you can clearly see that the total luminosity varies a lot more than the stellar mass. 100 solar masses of material will shine (approximately) as brightly as 100 Sol if in individual stars, but nearly 10,000 times brighter if collected into one body.

A significant uncertainty in the nuclear physics part of this model is how much non-primordial (H, He, a touch of Li) matter there is in the mix. Since nuclei like C and N take part in catalytic series of nuclear reactions at lower temperatures than without them, even small amounts of such nuclei can considerably alter the rate of energy release form the cores of medium-size stars, and hence the mass- luminosity relationship.

When all the data you have about a distant object is it's total luminosity and distance (from it's redshift), you don't have enough data to directly infer the mass of the body (galaxy) from that data. You have to, somehow, incorporate some relationship between luminosity and mass - such as the table given above. The conventional way of doing this - in the absence of any better information - is to assume that the remote object has a similar relationship to that present in the Milky Way. It's an assumption, but it's an explicit assumption, and very open to challenge. This paper challenges that assumption for high red-shift galaxies.

(It is worth noting that, because the evolution of bright stars releases non-primordial nuclei (heavier than Li) into the interstellar medium of a galaxy, it is expected that the brightness-mass relationship will change with time - and that always has been expected. Exactly how the mass-luminosity relationship changes with time remains a matter of debate though.)

This paper looks at the spectrum of the light from three of these high red-shift galaxies (GS-NDG-9422, at red-shift z = 5.943 [reference 14]; the "Lynx arc" [from "Massive Star Formation in a Gravitationally Lensed H II Galaxy at z = 3.357. reference 15 from the paper] ; and A2744-NDG-ZD4 at red-shift z = 7.88, reference 16) and sees a distinct break in the spectrum that they infer as meaning there is a lot of gas heated to tens of thousands of Kelvin, producing significant "free-bound" electron interactions. Those temperatures are a lot hotter than the hydrogen ionisation event which gave the Cosmic Microwave Background (CMB) radiation at a temperature of approximately 3000 K, allowing the inference of considerable numbers of stars of high brightness that "pump up" the galaxy's gas temperature to these levels. Which means that the "Initial Mass Function" of the galaxies is richer in heavy stars than the Milky way of today. And in turn, that means that the inferred mass of the galaxies - which had been stressing models of the evolution of the early universe would be considerably lower.

Having a "top heavy" initial mass function in the early universe will require a lot of re-calibration of models of the evolution of galaxies. But at least there is now some data to suggest in which direction, and by how much, to adjust the models.

I see that Dr Becky Smethurst has done a video-blog on this paper on her "Night Sky News" channel.

(The heaviest species detected in the spectra is argon (lines at λ4711 and λ4740 &Angstrom;, which would suggest some "silicon burning" nuclear reactions.)

End of Document
Back to List.

---

And that's New Year, so on to the next page.