Talk:List of largest stars

NGC 1313-310 and NGC 253-222

Two new stars at over 1650 solar radii. Should we keep them? Diamantinasaurus (talk) 09:37, 7 September 2024 (UTC)

they are newly discovered and they are located in distant galaxies outside of the local group (they are further from the sun than NGC 2363-V1) so their sizes might be inaccurate in some ways Hoanghao314159 (talk) 13:10, 7 September 2024 (UTC)

the problem is that i'm not an expert on stars in distant galaxies so you can keep them until an expert explains everything Hoanghao314159 (talk) 13:11, 7 September 2024 (UTC)

Should we really include all these stars?

The list contains a lot of stars which do not have well-constrained radii. For the stars that have their radii measured with the Stefan-Boltzmann law, many of them (especially the intergalactic ones) do not have properly modelled dust envelopes and assume spherical symmetry or, if in a catalogue, use the exact same extinction correction for each entry, which often lead to significant over- or underestimations in the luminosity as most RSG dust envelopes are asymmetric.

There are also stars on the list that use Gaia data, which is not accurate for RSGs, due to their very large convection cells and the fact that their angular diameters are greater than the stars' measured parallaxes. This causes the star to appear to wobble to a degree comparable to or even greater than the parallax itself.
Therefore, a large portion of the radii on this list are tentative.

Should we remove them until they have better constrained radii? SpaceImplorerExplorerImplorer 15:42, 7 September 2024 (UTC)

No. These radii are still better than nothing, and this list never had the objective to be over-accurate. This list can't strict in accuracy because all estimates are uncertain and are still our best guesses. If all star sizes that we think that are inaccurate are removed, this list would have no more than three entries, or would be empty. 21 Andromedae (talk) 18:39, 7 September 2024 (UTC)

I would say that the vast majority of the star sizes here are reliable. 21 Andromedae (talk) 18:46, 7 September 2024 (UTC)

Explain why SpaceImplorerExplorerImplorer 19:14, 7 September 2024 (UTC)

Because it is difficult to determine whether an estimate is correct or not, so we assume it is correct until we can determine whether it is (or not). The distances determined by Gaia may be imprecise, but in many cases they are the only estimate available, and are better than nothing. And there are cases where the Gaia distances are consistent with other publications (e.g. PZ Cassiopeiae, R Cancri, V354 Cephei, V509 Cassiopeiae, RW Cephei, BC Cygni) or when the Gaia DR2/DR3 distances are similar (e.g. UY Scuti). 21 Andromedae (talk) 19:44, 7 September 2024 (UTC)

Consistency doesn't always imply accuracy. Also, you can't use Gaia to say that Gaia data is accurate, which you did with all of these stars except for PZ Cas. SpaceImplorerExplorerImplorer 20:11, 7 September 2024 (UTC)

Also, we are already citing the "best" sources. There are no longer stars based on oudated/unreliable sources like De Beck et al. 2010, or the TESS Input Catalog, or the Messineo & Brown paper, which means we are cutting off less likely estimates and prioritizing better ones. 21 Andromedae (talk) 20:08, 7 September 2024 (UTC)

Just remove the tentative ones with blind and potentially inaccurate assumptions. If there are too few, simply lower the list's cutoff. Only keeping the well-defined radii would still leave at least 23 entries above 700 R_☉. SpaceImplorerExplorerImplorer 19:13, 7 September 2024 (UTC)

Just noting that this is a perennial request, with the most recent discussion being about a year ago. Are there new arguments for change, or is this just asking the question because it hasn't been discussed in a while and it's time to bring it back up (again)? Primefac (talk) 19:23, 7 September 2024 (UTC)

Yes, there are new arguments and this a different request as the linked discussion is about two specific entries, not the list as a whole. I'm not going on about limits this time. I'm instead commenting on radii derived from automated calculations and unreliable Gaia data, which have been used extensively throughout the page and have made it a mess. SpaceImplorerExplorerImplorer 20:25, 7 September 2024 (UTC)

HR 5171 downsize?

HR 5171 has a radius of 1,060 - 1,160 R_☉ in the list, but since combining angular diameter and radius from different sources is apparently allowed, this star can be like 550 - 650 R_☉ assuming a distance of 1.5 kpc and the angular diameter from other papers. Diamantinasaurus (talk) 17:33, 12 September 2024 (UTC)

combining angular diameter and radius from different sources is apparently allowed... it is? When did we come to that consensus? Primefac (talk) 10:54, 13 September 2024 (UTC)

WP:NOTJUSTANYSYNTH states that Wikipedia synthesis is only original research by synthesis. Since calculations are not original research, it shouldn't violate the policy. SpaceImplorerExplorerImplorer 08:19, 14 September 2024 (UTC)

I am uncomfortable with the idea of taking a distance from one paper and combining it with a radius from another paper in order to get a size calculation, and I do not think that is the sort of exception that your link is meant to describe. Primefac (talk) 12:14, 14 September 2024 (UTC)

NML Cygni

NML Cygni on the list uses a distance and angular diameter from two different sources, and attempts to remove it have been reverted. What do we do about this? VY Canis Majoris (talk) 07:00, 16 September 2024 (UTC)

They were reverted because apparently Wikipedia synthesis only refers to unverifiable and original research, which routine calculations are not. SpaceImplorerExplorerImplorer 17:12, 16 September 2024 (UTC)

Well, i felt that NML Cygni deserved to be in the list, since it well-known, has an accurate distance and hence accurate properties. I don't mind if two sources are being used for calculating a radius, and it is still better than nothing. Many other articles use two sources for calculating radii in the infobox as well. Just keep the radius as it is. 21 Andromedae (talk) 22:35, 16 September 2024 (UTC)

I think the best thing to do is come to a consensus about how to deal with this situation (and likely "using multiple sources for one fact" in general). I'll cross-post to WT:AST to get more opinions. Primefac (talk) 12:17, 17 September 2024 (UTC)

EV Carinae is too small?

The 1,165 R_☉ estimate uses a distance of 2.4 kpc, smaller than the real one (2.96 kpc). Due to this, the luminosity and radius were underestimated as well. The actual luminosity was estimated t be 288,000 L_☉, which combined with the normal M4.5Ia RSG temperature (3,535 K) from Levesque (2005) gives a radius ~1,430 R_☉. I don't think this can be added as it is WP:OR (although with that NML Cygni estimate on the list it could be possible). Still, 1165 R_☉ is too small. Diamantinasaurus (talk) 19:05, 29 September 2024 (UTC)

If there is any estimate based on the "true" distance of 2.96 kpc, it can be added. 21 Andromedae (talk) 20:22, 29 September 2024 (UTC)

I think that if we had an angular diameter measurement we could calculate a size. 21 Andromedae (talk) 20:31, 29 September 2024 (UTC)

Credibility

Do you think we should cut down on using stars that are just taken from large databanks/tables? There's a reason WOH G64 has a consensus amongst scientific literature as one of, if not the largest known star because its properties are well defined from a whole paper dedicated to calculating its parameters. Most of these stars aren't even given a direct radius from their sources and even some of those NGC stars which do (one being bigger than WOH G64) have luminosities that are enormously above the limit. Faren29 (talk) 01:04, 2 October 2024 (UTC)

E.g.: there's two stars with larger radii than WOH G64 that were just chucked into the list. Guess what? Those two stars' luminosities are on the order of half a million solar luminosities. Stars like some the RSGC stars are fine because they are at least in the confines of what is achievable for RSGs but I think we'll need to cut some of the junk like this out of the list, Faren29 (talk) 01:16, 2 October 2024 (UTC)

Those luminosities are possible for RSGs with an unusually high initial mass. However, I do think that these catalogues should be removed from the list as they rely on assumptions and often don’t even use SEDs, which can easily lead to a large amount of inaccurate measurements of their properties. SpaceImplorerExplorerImplorer 06:11, 2 October 2024 (UTC)

The luminosity estimate of NGC1313-310 uses the SED of the star, LGGS J013339.28+303118.8 doesn't, however it appears to be very luminous based on its apparent brightness and it doesn't seem to be a foreground star either and RSGs this luminous are possible as you said.

Plus this paper which is used A LOT on this page also doesn't use the SED data but I don't think we should entirely remove it. Diamantinasaurus (talk) 09:36, 2 October 2024 (UTC)

No, I would still be hesitant. This star is very distant, lacks observations in various spectral banda, has no data from Gaia, PS1 releases, and there is a high chance it could be a multiple star system rather than a single bright red supergiant. SkyFlubbler (talk) 17:41, 5 October 2024 (UTC)

Which star? NGC1313-310 or LGGS J013339.28+303118.8? Faren29 (talk) 20:05, 6 October 2024 (UTC)

Agreed with Diamantinasaurus. 21 Andromedae (talk) 13:24, 2 October 2024 (UTC)

SED fitting

Should we favor estimated derived via Spectral Energy Distribution data over those who don't? For example if there is a star with two estimates, one that used SED and another one that didn't, we favor the one that did over the one that did not. Diamantinasaurus (talk) 19:49, 3 October 2024 (UTC)

Agreed. 21 Andromedae (talk) 20:59, 3 October 2024 (UTC)

MU Cephei

972 R_☉ should be added back to the Mu Cephei entry. The reason to remove it was solely because the distance is "inconsistent with the OB association where the star is". This is not a convincing reason to remove the whole estimate, Mu Cephei is not a confirmed member of any OB association, it is just assumed based on its position in the sky. Nothing disallows it to be a foreground object, and Jim Kaler even says Oddly, one study shows the star NOT to be a member of the association. Recently the blue supergiant Sher 25 was found to lie in the foreground of its cluster, so the possibility that Mu Cephei is closer than previously believed is not remote. 21 Andromedae (talk) 11:09, 5 October 2024 (UTC)

Yeah, but the 641 pc distance used for the 972 R_☉ estimate is just based on the 222 pc distance to Betelgeuse but upscaled, which isn't a reliable method of estimating the distance. Plus its extinction is in agreement with the larger ~900 pc distance. Diamantinasaurus (talk) 09:32, 6 October 2024 (UTC)

As Perin et al. (2005). explains, the extinction reaches a plateau at around 500 parsecs, which encompasses both 940 and 641 pc. Hence the extinction agrees with every distance a bit larger than 500 parsecs. This list isn't strict with reliabilities of star sizes, so adding this estimate is acceptable, and the method used to estimate the distance isn't bad. If Montarges et al. (2019) and Perin et al. (2005) chose this method, there is a reason, of course. 21 Andromedae (talk) 17:02, 6 October 2024 (UTC)

the reason is that they're both M2 stars, but I don't like this reasoning honestly. W60 B90 is M2, V354 Cephei is M2.5 and MSX LMC 597 is M2.5 and yet all three are over 1200 solar radii.

mu Cephei is M2Ia and Betelgeuse is M1-2Ia-Iab, so their spectral types aren't even that similar. RSGC1-F04 might be M1I too but no one uses Betelgeuse to estimate its distance.

Also, the 640 pc distance uses an overestimated distance to Betelgeuse as well. Diamantinasaurus (talk) 18:57, 7 October 2024 (UTC)

Mu Cephei is either M1I (infrared spectral type from Levesque et al. 2005) or M2Ia, not so far from Betelgeuse, and it could be even hotter, with a surface temperature of 4,022 Kelvin (see Perrin et al. 2005), perhaps a K-type supergiant. The mass-loss rate of Mu Cephei is also much lower than that of similar-sized RSGs, it is 100 times lower than that of VY Canis Majoris and 10 times lower than that of other red supergiants (Shenoy et al. 2015) far too low for a large RSG with a radius of 1,400 solar units. 21 Andromedae (talk) 20:51, 12 October 2024 (UTC)

mu Cephei recently had a higher mass-loss rate at around 5 * 10^-6 solar masses per year per the link, which doesn't differ too much from that of several RSGC1 RSGs (RSGC1-F06 has a smaller one, F10 has a similar one, F05 has a similar one, even the extremely luminous F04 has a not much higher one etc.) and the 1,000 - 1,300 R_☉ SU Persei has an even lower mass-loss rate. The 1,245 - 1,520 R_☉ V354 Cephei also has a similar low mass-loss rate. Diamantinasaurus (talk) 21:15, 12 October 2024 (UTC)

RSGC1 is an extra case. Take RSGC1-F01 as an example, it is larger than VY Canis Majoris, same spectral type and present maser emission, but its estimated mass-loss rate is 33 times lower, pretty pretty odd. The mass-loss rates could have been underestimated as well. SU Persei is could only 800–900 solar radii based on another distance of 2137 parsecs in table 2.1 here. Indeed the mass-loss can't rule out a radius up to 1,200 R_☉, but a very large radius e.g. 1,426 R_☉ is unlikely given the estimated mass loss. 21 Andromedae (talk) 00:01, 13 October 2024 (UTC)

Fair point for 1,400+ R_☉. Maybe keep it at 1,259 R_☉ then? Diamantinasaurus (talk) 15:32, 3 November 2024 (UTC)

Removal of some LGGS stars

Some stars with sizes bearing WOH G64 calculated using suspiciously large luminosities have been added, and an attempt to remove them has been reverted based on very poor reasons. These high luminosities, much higher than the luminosity limit of RSGs of 316,000 L_☉ are almost certainly inaccurate, and the paper itself give a little comment on these estimates casting doubt on their reliability. These stars appear to be inconsistent with the new generation S4 models and both the S0 and S3 old generation models, all of which loop back to the blue portion of the HR diagram without extending to such cool temperatures. [...] 21 Andromedae (talk) 18:16, 7 October 2024 (UTC)

Hence, i am proposing the removal of LGGS J013312.26+310053.3, LGGS J013418.56+303808.6 and LGGS J013312.26+310053.3 due to they being larger than the limits of luminosity and/or radius of RSGs. 21 Andromedae (talk) 18:31, 7 October 2024 (UTC)

The 316,000 L_☉ limit isn't an exact limit. AH Scorpii is slighty more luminous and HD 269551 A is in the list at 389,000 L_☉, more than even LGGS J013418.56+303808.6. The two other one have more extreme luminosities though.

All three stars (including the two more luminous ones) actually fit well in the HR diagram on the bottom of figure 20 in the paper. RSGs can archieve higher luminosities with an initial mass of 40 M_☉ and lower metallicity (this is in the list).

All three stars are rank 1 stars, which are "highly likely supergiant" and both spectroscopic and kinematic data suggests that they're not foreground stars but actual red supergiant members of the Triangulum Galaxy. Diamantinasaurus (talk) 19:07, 7 October 2024 (UTC)

The luminosity of AH Scorpii is overestimated, using a more apporopriate temperature for a M4-5 supergiant (from Levesque et al. 2005) and the radius of 1,411 R_☉ i got only 254–280,000 L_☉. HD 269551 itself has a close OB companion which could overestimate its luminosity measurement, if the combined luminosity is atribuited to a single object (this could be the case of some of these LGGS stars BTW). Some stars can be larger under certain conditions, but it requires subsolar metallicites, which is unlikely considering that RSGs are born in metal-rich places (e.g. open clusters, OB associations). These stars also aren't fully consistent with the H-R diagram, as the first figure in Fig. 20 shows. The second figure isn't very relevant as it just assumes a null rotational velocity, which is unlikely given that many O-type stars are fast rotators. 21 Andromedae (talk) 19:42, 7 October 2024 (UTC)

fair for AH Scorpii, but I disagree with HD 269551 A, it does seem that luminous. However, that 316,000 L_☉ limit applies more to cooler red supergiants like VY CMa or WOH G64, I think early K-type and early M-type stars can get to larger luminosities. For example the M0-M1 RSGC1-F04 has a luminosity of 380,000 L_☉, SMC 18592 has a luminosity of 355,000 L_☉, the 3800 K HD 269551 A has a luminosity of 389,000 L_☉, Stephenson 2 DFK 49 has a luminosity of 390,000 L_☉ and RW Cephei has a luminosity of 339,000 - 409,000 L_☉ (based on 1,100 R_☉ and the temperature of 4,200 - 4,400 K).

Also from that paper about M33 red and yellow supergiants:

"Indeed, we see that the new generation S0 tracks (Figure 20, bottom panel) proceed to cooler temperatures than the S4 tracks (Figure 20, top panel) before looping back to the blue. Although the distribution of stellar rotational velocities has yet to be fully decoupled from the effect of inclination (since one can only measure vsin i), there is increasing evidence that some massive stars are born as genuine slow rotations (e.g., Huang et al. 2010). Thus, although the location of our five most luminous supergiants could likely be explained by a change in the mass-loss prescriptions used by the Geneva models, it is possible that they are simply slow rotators. If this were the case, their locations would be fully consistent with the new generation models presented here." Diamantinasaurus (talk) 19:57, 7 October 2024 (UTC)

So i would agree that LGGS J013339.28+303118.8 (1,566 R_☉) might still be in the list. LGGS J013418.56+303808.6 (10^5.76 L_☉) is a foreground object based on recent research. Also, i edited my answer above. 21 Andromedae (talk) 20:23, 7 October 2024 (UTC)

But how was that distance in there calculated? Diamantinasaurus (talk) 04:46, 8 October 2024 (UTC)

Photometry. That is not very great, but there is more evidence that it is a foreground object: The position in the sky is far away from the Triangulum Galaxy itself, unusual considering that RSGs form within the galactic disk and not the halo, and the spectroscopically-determined surface gravity and metallicity are similar to that of a metal-poor giant star, suggesting that the star belongs to the Milky Way halo. 21 Andromedae (talk) 20:53, 12 October 2024 (UTC)

Why is Stephenson 2 DFK 1 not listed?

Genuine question, why is Stephenson 2 DFK 1 not listed? I've heard that it has a claimed radius of 2.99 billion kilometres, which is also often claimed to be unreliable, I'm mostly confident that that is the reason, but I haven't had confirmation. Newaccount33333 (talk) 22:02, 17 October 2024 (UTC)

Check through Archive 4, there are a half-dozen threads about the matter. Primefac (talk) 22:08, 17 October 2024 (UTC)

Consensus in the last year resulted in the removal of this star. In short, the radius of Stephenson 2 DFK 1 is highly unreliable and must not be there. 21 Andromedae (talk) 23:33, 17 October 2024 (UTC)

The reason why it isn't here is because the distance used to estimate the radius had a >50% uncertainty and it was derived assuming a spherical dust envelope which can often lead to overestimations of the luminosity given that most red supergiant dust envelopes are asymmetric. SpaceImplorerExplorerImplorer 07:27, 18 October 2024 (UTC)

So yes, in short, unreliable estimates. I appreciate the explanations. Newaccount33333 (talk) 20:08, 19 October 2024 (UTC)

NGC 1313-310

This star has a bunch of notes for its uncertainty, like for its temperature and its luminosity. For the temperature, it’s the Titanium(II) oxide lines that usually result in higher temperatures, and the luminosity which apparently hasn’t been constrained enough to confirm its large size. To be honest, this just seems like Stephenson 2 DFK 1 but without the distance uncertainty. And that brings the question I want to ask, should it be removed from the list for these reasons? Atlantlc27Lol (talk) 11:21, 31 October 2024 (UTC)

It does seem like a lot of caveats and addenda. Primefac (talk) 11:52, 31 October 2024 (UTC)

You could use the T_eff scaling relation in page 13. It's quite uncertain but works very well for other stars with a TiO-derived effective temperature which could make it moderately acceptable. SpaceImplorerExplorerImplorer 16:13, 31 October 2024 (UTC)

I removed it along with two other stars from the list. The rest are acceptable but those three simply have too many caveats. Faren29 (talk) 14:35, 1 November 2024 (UTC)

Likely a very good choice. They shouldn’t be placed on the list until they have accurate radii or atleast better radii without as many caveats and potential errors. Atlantlc27Lol (talk) 02:54, 2 November 2024 (UTC)

I think that gets like 1,500 solar radii for it, iirc. Diamantinasaurus (talk) 15:22, 3 November 2024 (UTC)

"To be honest, this just seems like Stephenson 2 DFK 1 but without the distance uncertainty."

Isn't distance literally the biggest issue with the large estimate for St2-18? Diamantinasaurus (talk) 15:26, 3 November 2024 (UTC)

I meant that with the temperature and luminosity Atlantlc27Lol (talk) 15:40, 5 November 2024 (UTC)

I think that the caveats in the side are already enough. I agree that these stars could be removed but some people still disagree, so it may be better to include them for now. 21 Andromedae (talk) 16:50, 3 November 2024 (UTC)