|WikiProject Astronomy / Astronomical objects||(Rated B-class, Top-importance)|
|Threads older than 3 months may be archived by.|
Limited Sample Basis / Sampling Error?
Regarding the last column in the table under 'Harvard spectral classification'. The source used for the frequency of each type was a survey of our 'stellar neighborhood'. As such, it may or may not represent the GENERAL frequency that each type may occur. For instance G and K are listed as 7.6% and 12.1% respectively. Other sources (see wiki entry for Orange Dwarf/K-type main-sequence star) indicate K types may be 3-4 times more common than G type.
'3 to 4 times...' is conversational and not statistical as the chart attempts to be, however many authoritative sources indicate a higher frequency of K/Orange Dwarfs than simply the 12.1% value listed in the chart. I suspect the stats used are skewed by the fact that the source used was constrained to those stars within a sphere of about 13 parsecs in diameter (read the article), and thus nearby. As such it is a very poor source to use as a citation for a general statement about all stars in the galaxy.
Just looking at the G vs K values, they seem wrong: in the rest of the chart the frequency at least doubles as you go up the scale, but K frequency is only 80% or so more than G. This page supports much of the data in the chart but lists frequency/abundance values that seem more in line. I cannot determine the source of their data, but it does comport with other info from Nature, Science and other publications.
— Preceding unsigned comment added by 184.108.40.206 (talk) 16:38, 22 October 2012 (UTC)
- When you've got some better data ("better" meaning more recent, more comprehensive, and statistically more valid) then feel free to cite it and change the table. Until then, the table reflects the data from *a* peer-reviewed source, while your opinion is just an opinion. The Wikipedia K Main Sequence page states that K stars are 3-4 times more frequent than G and even quotes a source but the source is not peer-reviewed nor does it appear to include that information. Good source though, lots of fun data there :)
- Incidentally there are some anomalies that could unnaturally suppress the counted numbers of K stars. For example, some of the K sub-divisions do not have representative standard stars and hence are almost never assigned. Thus many stars that would be K8 or K9 on an even scale will be assigned to M0. Might not seem much, but that might be around half the expected K stars.
- Another curiosity is the stated number of white dwarfs, high enough to make the total number of A stars higher than F or G stars. Lithopsian (talk) 19:22, 22 October 2012 (UTC)
I dont really have anything against the article - it is good for what it is. But all it is really, is a *newsletter* article and according to it, the author "Glenn LeDrew has been an avid amateur astronomer...". A survey of 800 (or 824 depending...) of the very nearest stars is a flawed methodology for galactic generalities. Sadly, I can find nothing with citations which is any better, but all things considered, passing comments in Nature and the like seem more reliable.
I'm still looking for something more reliable. (Incidentally, I too am sure that many erstwhile K7+ stars are sometimes confused with M0 or M1 - lots of articles point out the problem of sorting 'true' Ms from degraded K class. I think that is where some of the very high values come from.) You'd think someone could run a star catalog thru a program to tally things up, huh? — Preceding unsigned comment added by 220.127.116.11 (talk) 21:32, 23 October 2012 (UTC)
- Not quite as simple as tidying things up. If the defined examples of what features of a spectrum make for a K9 star don't exist then how can people assign stars to that category? Although the scale is intended, and to some extent has been manipulated, to provide a regular and even mapping of spectral class numbers to temperature, the lack of distinguishing features between certain steps of the scale has led to a number of well documented "gaps". Some have been closed in various ways but there are still spectral classes that are relatively poorly populated. Lithopsian (talk) 21:52, 23 October 2012 (UTC)
- From the (Third Catalogue of Nearby Stars (Gliese/Jahreiss), including a little over 3,500 stars out to 25 parsecs. This is not expected to include all stars out to even that small distance, mostly missing M and fainter dwarfs. I ignored anything fainter than M.
- B 3
- A 73
- F 266
- G 501
- K 887
- M 1721
Don't get too literal with those numbers. I included k-m in K, and the small number of g-k in G. I ignored 2 giants, 12 sub-dwarfs, and a couple of hundred white dwarfs that are classified by their own spectroscopic scheme. The numbers at decimal classes are extremely variable, for example 27 K6, 94 K7, 50 K8, and 2 K9 (don't compare to total K stars since many do not have a decimal classification). Lithopsian (talk) 21:16, 25 October 2012 (UTC)
- A sample to 5 parsecs should be more complete, and yet some stars cooler than about M5 (absolute magnitude below 15, apparent magnitude below 13) have probably still been missed. Ignoring brown and white dwarfs, 49 out of 60 stars are class M, over 80%. 6 (10%) are class K and 3 (5%) are class G. Statistically a little small but you can see the trend. — Preceding unsigned comment added by Lithopsian (talk • contribs) 09:58, 26 October 2012 (UTC)
Why are L, T and Y stars >= 100% abundance?
In the table for the last column labelled "Fraction of all main-sequence stars", the last three rows, labelled L, T and Y stars all say they are >= 100% abundance. This makes no sense to me. What is going on here? FrankH 01:21, 14 November 2012 (UTC) — Preceding unsigned comment added by FrankH (talk • contribs)
- They're not stars, and there are more of them than there are stars. Hence > 100%. Even if you want to argue with the definitions, they aren't included within counts of main sequence stars so you still get more than 100%. Mostly guesswork though since we've found nowhere near as many L/T/Y combined as we have M stars even in the closest surveys. Lithopsian (talk) 21:47, 14 November 2012 (UTC)
- Thanks Lithopsian, that now makes sense, but I was totally confused by a claim that it was more than 100%. Wouldn't it make more sense to leave those three cells blank, or better fill those cells with something like "Not a Star" or "Not a Main Sequence Star". The current ">=100%" makes no sense to me or I bet to most non experts reading this chart. - FrankH 06:11, 15 November 2012 (UTC) — Preceding unsigned comment added by FrankH (talk • contribs)
- Not my choice to include them. I'd rather they weren't there, or at least very clearly separated out to indicate that they don't fall within the classical definition of a main sequence star or even a star. But they're a sort of a trending topic at the moment and people are fond of putting the latest greatest news items into Wikipedia articles without a lot of context. Even in the science world there is vigorous discussion about how we should define stars, sub-stellar objects, brown dwarfs, planemos, etc., but don't expect a clear resolution any time soon. The Wikipedia brown dwarf article covers the subject fairly well. Lithopsian (talk) 12:29, 15 November 2012 (UTC)
- The more I look at those three lines in the table, the more I dislike them. The colours are largely unsubstantiated opinion and certainly not accurate colours of the objects. The mass, radius, and luminosity are given as "unknown" when in fact they are known and the author just couldn't be bothered looking them up. And the fraction column manages to be misleading without conveying any verifiable information. It is currently speculation to claim there are more L class stars than all other main sequence stars combined, and the same is true for T and Y. The first Y class objects weren't even discovered until the last year or so, and this very article claims there are only six (tentatively) known. I'm tagging all three entries as needing citations. If nobody provides them then I'll edit the table, but that isn't in anyone's best interest because this isn't my area of expertise. Lithopsian (talk) 14:00, 15 November 2012 (UTC)
- The plot thickens. It seems that observation and theory agree that there are actually relatively few L type brown dwarfs, due to the way the classes are formulated and to the relatively short lifetime of these objects as they cool without fusing hydrogen. Even cooler objects are thought to be more common, partly because the same objects spend a longer amount of time in the cooler state, but observational support is limited. Lithopsian (talk) 14:53, 15 November 2012 (UTC)
I mean Stellar_classification#cite_note-proportions-32 and Stellar_classification#cite_note-40. What do they mean? I get the impression they are trying to qualify the quoted proportions of stars in each spectral type, yet they appear to be both incorrect and meaningless. I find no evidence that the quoted fractions are of a sample down to absolute magnitude 16. Even if they were, the statement that samples to a lower absolute magnitude would only include M stars and exclude earlier types seems remarkably pointless. From my own counts from various catalogs, the given reference does not appear to be close to a complete volume-limited sample to that magnitude, and the footnote might (with a reference?) better point that out. The second footnote, at best, needs a reference for the precise number it gives. Again, the number does not match my own expectations for a complete volume-limited sample of stars, which would be over 80% M class. The fractions in the article perhaps deserve to be qualified, although there is a reference given for them and that might be sufficient, but any qualification needs to be both correct and explanatory. Maybe someone can edit them to make sense before I remove them. Lithopsian (talk) 19:19, 13 May 2013 (UTC)
There are no refs in this section. Also the sentence below needs to be recast. I'm not going to recast the sentence due to my own confusion about what the previous editor was trying to say.
"The giants among those stars are presumed to produce this carbon themselves, but not too few of this class of stars are believed to be double stars whose odd atmosphere once was transferred from a former carbon star companion that is now a white dwarf."
I think the main problem is: "but not too few of this class of stars are believed to be double stars" - I want to interpret it as "but many of this class are believed to be double stars".
Without the refs its hardly worth touching since without refs you can detail or claim anything in an article. Its a Percival Lowell canals on Mars situation. Quite often I look at things and want them to be something they are not - a bit like the sentence above. lol
I'm just copy editing so can't touch the section. If any astronomy editors know where to find the refs then please add them in.
Classification systems confusion
As a novice to astronomy I had a really difficult time understanding the differences between the classification systems discussed in the article (and referenced inconsistently):
- Morgan - Keenan system
- Harvard spectral classification
- Yerkes classification (MKK sytem)
- MK system (as named in the Yerkes classification paragraph)
Here's my current summarized understanding, so correct me if I'm wrong: the Morgan-Keenan system (aka MK system) combines the Harvard spectral classification system with a luminosity class to produce a two-dimensional classification. For example, a star classified "M2V" has spectral classification "M2" (two-tenths from M to L) and luminosity class V (meaning it is a dwarf star).
It took me a long time to get to this understanding from the current article. The problems I ran into, in no particular order:
- At a high level, it is not at all clear which system is the current, state-of-the-art, standard.
- Why is the paragraph which discusses the Morgan-Keenan system--the current standard system--titled the Yerkes Classification, a system which has been replaced???
- From the paragraph on the Yerkes classification it is not at all clear that the Morgan-Keenan system uses the same spectral classifications as discussed in the Harvard spectral classification system above. The only hint of this comes from looking at the examples.
- It is very confusing that the current classification system is buried underneath the paragraph on "conventional and apparent colors"--it makes it sound like the MK system is just another alternative, when in fact it is the current one.
The big fundamental problem I think is that this article discusses the history of classification systems interleaved with the description of the current classification system. I think it would make more sense to present Morgan-Keenan first, with both spectral and luminosity classes defined, and then have a later section discuss the history of the classification systems.
Anyway I don't want to rant since I'm obviously glad this page exists, I just think it could be easier to grok on the first couple reads. I'm happy to help out on anything, I just want to make sure I don't pollute the article with my own misunderstandings in the process. — Preceding unsigned comment added by Bpodgursky (talk • contribs) 10:06, 2 October 2013 (UTC)
- Go for it. I've already rewritten the rambling and confused header so that it hopefully summarises the current usage of the spectral class without confusing and irrelevant detail. Now the rest of the article should be reordered to match, with Secchi classes and such relegated to a historical background. I suggest starting with the MKK system, bringing the separate "spectral types" section under it, and sending the older systems into a section that is clearly identified as historical. Lithopsian (talk) 11:12, 2 October 2013 (UTC)
- Further, it seems that the sections on brown dwarf classes have got way out of hand, perhaps before there were separate articles for those stars. These should be brought back to brief summaries in line with other classes. Should the section on habitability even be in this article? Certainly not stuck in the middle like it is. A whole section for magnetic stars which are just one of many peculiarities? I don't know what to do with the section on apparent colours. It is background information of marginal relevance, but given the big table comparing conventional and apparent colours within the Harvard classification system, it really needs to come earlier. Lithopsian (talk) 11:12, 2 October 2013 (UTC)