Talk:Sun's mood swings not so strange after all, say scientists

@Robertinventor: I'm going to let this one sit for a little bit but if you want a crack at it, I'm sure we'd all be glad for your expertise. Do you know of a good source on the sun-identity controversy? I don't want to leave it for more than a few hours without making it available for review, though. Darkfrog24 (talk) 21:31, 15 July 2017 (UTC)

@Darkfrog24: Okay I did some minor edits of the first paragraph. Had a look for background on the controversy, and I found this thesis about the problem [1]. It's discussed in section 5.8 Summary of Results (page 181) onwards. There Earth is a G type star. Seems that the hypothesis of a connection with the Rossby number is one that they had already at least before the time of this thesis April 2017. But it was difficult to test it because of absence of data about rotation periods of many sun like stars. On page 189 he mentions several puzzles about our sun's period - here are a couple of them - that some models predict a slower cycle with faster rotation, but they observe a faster rotation with high quality observations, and that the models predict faster rotation at the poles than at the equator for the Sun, the opposite of what is observed:

"However, the latest generation of numerical hydrodynamic simulations show that meridional flow speed is inversely related to rotation, so that faster rotation leads to slower meridional flows. Therefore, for a single-cell meridional flow, we would conclude that cycle period increases with faster rotation. This is contrary to previous observations, and to our results when we restrict cycle selection only to those of the highest quality. ... Testing the observation of a critical Rossby number threshold Rocrit = 1.5 for producing smooth, monoperiodic cycles in numerical experiments is made difficult by the lack of a correspondence between the semi-empirical Rossby number of Noyes et al. (1984a) and the results of convective MHD simulations. Gastine et al. (2014) found that a wide variety of MHD and HD-only models produce solar-like differential rotation (fast equator; slow poles) when Roc < 1, and anti-solar differential rotation when Roc > 1 ... Now consider that the Sun’s Rossby number is 2.15 according to the formula of Noyes et al. (1984a), yet it clearly has solar-like differential rotation. A reconciliation of numerical and empirical Rossby numbers would aid the interpretation of our results."

Not exactly a neat "sound bite" and it would be too specialist to use as a source. There is no way we can expect a reviewer to read it. But it gives an idea of the background. Perhaps it might be adding this thesis as an "external source" for specialist readers??? Directing them to pages 181 and following? Basically what they have done is to build a good model that works, and validated it with 25 sun like stars. Sadly, I can't read their paper at AAAS - it's behind a paywall for me, and you only get free access one year after publication. Robertinventor (talk) 01:15, 16 July 2017 (UTC)

Okay your quote is only showing up in Wikicode. I swear this page was blank literally a minute ago when I went to find a source for sunspots and the auroras. I realized there's a whole other story in the Bastille Day Northern Lights. Darkfrog24 (talk) 01:18, 16 July 2017 (UTC)

I had the same problem - I've been doing minor edits and it should be fixed now. Robertinventor (talk) 01:19, 16 July 2017 (UTC)

For the controversy, I was asking more about the people side of it. How long have people been fighting about this? Are there just two sides? What do the people on the [two] sides think, and what's their reasoning? From the other sources, I've inferred that one side thinks "The Sun is very different from solar stars" and these new findings are "No it's not. It's similar enough that we can say it is a solar star." But there might be more detail out there. Darkfrog24 (talk) 01:21, 16 July 2017 (UTC)

To toss out something specific, you say ours is a G-type star. Does that mean these findings suggest it's a solar-type star instead of a G-type star or do these categories cover two different aspects of star-ness, like how your ABO blood type has nothing to do with whether you're RH+ or RH-? Darkfrog24 (talk) 01:27, 16 July 2017 (UTC)

Oh, it's nothing like that. It's a very particular debate only about the solar cycles. The sun is a perfectly normal G type star. A solar type star, a solar analog, or solar twin, are stars that are increasingly similar to our sun. The most similar in that list is the "solar twin".

A solar twin is a G type star that in addition is very similar to our sun in its mass, chemical composition, temperature, age and other features. Though all G type stars are hydrogen burning, most of them also, like our sun, have minute but easily detectable amounts of just about all the other elements in the periodic table (apart from some short lived radioactive ones). It's those trace elements that condensed out of the nebula to form the terrestrial planets - any elements we have on Earth that have been here since its birth are also in the sun in minute quantities. So it's things like that which they take into account when they classify solar analogues.

It got those from the gas cloud that it condensed out of, which got seeded by nearby supernova explosions, by ejections of material as in the planetary nebula and so on from older stars. We can actually detect several different supernova explosions that predated Earth from study of interstellar grains in some of the meteorites in our connection. So - other stars from our birth nebula will have similar compositions - but they are now scattered though the galaxy. Generally older stars have less of the heavy elements, and younger stars have more, as those concentrations build up in the gas clouds as the galaxy gets older. So gas clouds have more in them now than they did in the past. That's what is called the "metallicity" but it means all elements heavier than helium not just metals (strange name for it I know).

So, there isn't any dispute about that. There are many stars that are similar to our sun, and several that are so similar they count as "solar twins" though none that are totally identical twins of our sun. That's why I rewrote those two sentences in the first paragraph.

It's a cosmic oddball only as far as the solar cycles are concerned. Also, this is a very specialist debate. I didn't know about this puzzle until looking up the cites. I think few people interested in astronomy will know that it ever was a puzzle. I didn't find any popular articles about this puzzle - may well be, but I think I'd have turned up something if it was frequently reported. However when I searched the scholarly literature I found that it was a longstanding puzzle amongst some specialists who work on models of our sun and its convection and magnetic structure etc. Even then - not that much material on it. That thesis was the best source I found and it referred to earlier specialist papers on the topic. Hope that's a bit clearer. I'd say that the article as it is explains it fine on the basis of what I found. I suppose you could say "some astronomers specializing in the study of solar cycles" or some such, though I don't think you really need to do that. Robertinventor (talk) 02:47, 16 July 2017 (UTC)

BTW I had another look, didn't find anything else to fix, made a wiki link to the 11 year solar cycle article. Robertinventor (talk) 02:47, 16 July 2017 (UTC)

Notes for reviewers

I didn't do my usual in-Wikicode tagging this time, so I'll note here that the Washington Post article is here solely to corroborate the perhaps-not-common knowledge that sunspots rev up the auroras. Darkfrog24 (talk) 01:25, 16 July 2017 (UTC)

Review of revision 4331064 [Passed]

Revision 4331064 of this article has been reviewed by Pi zero (talk · contribs) and has passed its review at 02:05, 17 July 2017 (UTC). Copyright: Passed Newsworthiness: Passed Verifiability: Passed NPOV: Passed Style: Passed Comments by reviewer: Our goal isn't to "dumb down" the science into a summary, but the (very difficult) challeng to accurately and successfully explain. (Here's a favorite of mine: link.) Describing a geomagnetic storm from a sunspot as "rays" arriving at Earth struck me as inaccurate, and I struggled with whether-and-how I could fix the problem within my purview as an independent reviewer. I had similar difficulties in the last paragraph over the Rossby number, as I felt it important not to mislead the reader into thinking of the Rossby number as directly incorporating luminosity. Rossby number also clearly imo needed to be mentioned in the following sentence, otherwise why brining it up, but that got tangled up with verification problems (next bullet). I was surprised and frustrated to find, on close scrutiny, that I could not tell from the sources whether brighter meant higher or lower Rossby, whether faster spinning meant higher or lower Rossby, nor details of how cycle period related to sunspots (the timing of sunspots correlates with the cycle, of course, but what else is going on with the sunspots was not spelled out). Possibly this stuff is in the paper, but, paywall. The reviewed revision should automatically have been edited by removing {{Review}} and adding {{Publish}} at the bottom, and the edit sighted; if this did not happen, it may be done manually by a reviewer.

Revision 4331064 of this article has been reviewed by Pi zero (talk · contribs) and has passed its review at 02:05, 17 July 2017 (UTC). Copyright: Passed Newsworthiness: Passed Verifiability: Passed NPOV: Passed Style: Passed Comments by reviewer: Our goal isn't to "dumb down" the science into a summary, but the (very difficult) challeng to accurately and successfully explain. (Here's a favorite of mine: link.) Describing a geomagnetic storm from a sunspot as "rays" arriving at Earth struck me as inaccurate, and I struggled with whether-and-how I could fix the problem within my purview as an independent reviewer. I had similar difficulties in the last paragraph over the Rossby number, as I felt it important not to mislead the reader into thinking of the Rossby number as directly incorporating luminosity. Rossby number also clearly imo needed to be mentioned in the following sentence, otherwise why brining it up, but that got tangled up with verification problems (next bullet). I was surprised and frustrated to find, on close scrutiny, that I could not tell from the sources whether brighter meant higher or lower Rossby, whether faster spinning meant higher or lower Rossby, nor details of how cycle period related to sunspots (the timing of sunspots correlates with the cycle, of course, but what else is going on with the sunspots was not spelled out). Possibly this stuff is in the paper, but, paywall. The reviewed revision should automatically have been edited by removing {{Review}} and adding {{Publish}} at the bottom, and the edit sighted; if this did not happen, it may be done manually by a reviewer.

@Pi zero: First, sorry, missed the "rays" - I'm not that good at copy editing :). Yes it's charged particles from the sun, in the form of a "coronal mass ejection" which hits the Earth's magnetic field causing its magnetic fields to oscillate, these in turn cause previously trapped particles close to Earth to hit the Earth's atmosphere and form the Aurora and the magnetic effects can in the most severe very rare cases affect power grids. Details here [2]. The Coronal mass ejections in turn are usually associated with groups of sunspots. Good to have corrected it but it's not quite right as it is the solar plasma of charged particles that hits Earth, not so much the magnetic field although of course there's a magnetic field associated with the plasma but the plasma is the primary cause of the effects on Earth's magnetic field. I'll give it some thought, and suggest a correction there

Good question about the Rossby number. The answer from the literature is that it's complex. It is defined for stars here [3] (equation 1.9, page 14) as P_rot/τ_c where P_rotis the rotation period of the star.

There, quoting from page 4 of another thesis [4], says

". The dependence of magnetic activity upon rotation rate was first quantified by Noyes et al. (1984), who used these chromospheric emission lines to establish a close correlation between the magnetic activity of a star and its inverse inverse Rossby number, Ro₋₁ = Ωτ_c, where Ω is the angular velocity of the star and τ_c is the star’s (theoretically predicted) convective turnover time – the larger the value of Ro₋₁, the more active the star."

Basically a star with a high Rossby number is less active. And τ_c is the star’s (theoretically predicted) convective turnover time.

Back to the newer thesis[5], on page 15 he says it's hard to determine for the sun, is thought to be a lot greater than 1 for the surface, a lot less than 1 deep down, because τ_c varies a lot.

So anyway on page 28 he describes a "semi empirical" Rossby number. Basically they use the measured angular rates from star observations. The rotation data is found mainly by things such as seeing optical changes in brightness, so that would be an average over the entire surface of the star and surface rotation only of course (see the comments on table 5.3 in the thesis). The empirical guess at τ_c for this "semi empirical" Rossby number is estimated using log (R'_HK) where R'_HK is defined on page 23 - give a formula but they explain it also in words as "This ratio can be understood as the fraction of the total energy flux of the star that is caused by magnetic heating in the chromosphere leading to HK emission"

There HK emission means n Ca ii H & K emission lines - emission lines in the spectrum for Calcium. Why that is especially important for convection studies I don't know, but that's what they use.

So - in short - I can't read it either because it's behind a paywall and none of the popular articles I read explained. But they have to be using some empirical Rossby number. Perhaps this one. It will be proportional to the period, so inversely proportional to the measured angular speed, so faster stars have lower Rossby numbers. (Main sequence stars also slow down as they age so it's not a fixed property of a star but that changes very slowly).

But it will also depend on some empirical measure of the convection timescale, such as the ratio of the strength of the Ca ii H and K bands to the total energy flux from the star.

Hope that helps. I can give it some thought and suggest an edit to the last paragraph to try to encapsulate some of that in a nutshell in one sentence. Robertinventor (talk) 11:46, 17 July 2017 (UTC)

(edited sorry) Robertinventor (talk) 11:50, 17 July 2017 (UTC)

Oh this explains why the Calcium emission lines are important, says right at the beginning, old paper on these lines, that they are linearly proportional to the magnetic field intensity at the surface. [6] So a high log (R'_HK) means more magnetic field intensity, and so (likely to be) a more active star (so lower Rossby number). So makes sense it would be a good surrogate. Robertinventor (talk) 12:15, 17 July 2017 (UTC)

@Darkfrog24: and @Pi zero: I've just found the paper as a recent submission to arxiv.org here [7]. So not behind a paywall and will read it later today and that should help resolve this question of how they obtain the Rossby number, presumably some kind of empirical Rossby number and maybe there will be other points of interest too. Robertinventor (talk) 12:29, 17 July 2017 (UTC)

@Robertinventor: Just to remind, there's the 24-hour horizon after publication; substantive changes have to be before that. --Pi zero (talk) 12:40, 17 July 2017 (UTC)

@Pi zero:. Oh okay, I didn't know that. Should be able to do it in that time-frame.

Anyway, I had a look. It's a little indirect so needs a few paragraphs to explain properly. It's no wonder most of the news stories didn't try to explain it in detail. So anyway I'll summarize what the paper says. I think I'll need to then come back to it later and see if I can find a nice simple way of presenting the most essential points in a final paragraph, or if either of you or anyone else has any ideas about that do say. So for anyone who is interested, here is my summary of the main result in the paper [8]:

---

First, they found (Fig 2A) that in both their models and the real stars, the cycle gets shorter as the star spins faster at constant convective luminosity (brightness of the surface convective regions of the star). They found (Fig 2B) that the cycle gets shorter as the convective luminosity increases.

When they compensated for the luminosity dependence, then their results follow a straight line - and the stars roughly follow the same line as well (Figure 2D). Figure 2C is "normalized" but they don't explain that in detail and it's not quite so neat looking a correlation. So I think Figure 2D is the best one to look at to understand it.

So the blue dots match part of the data at the top - the longer cycle similar to our 11 year cycle. There's a lot less scatter than there was before they factored out the luminosity.

So there is no mention of the Rossby number yet, but that all suggests that their simulation is a reasonably good one. So then they look at correlation of magnetic cycle (measured in multiples of the rotation period) with the Rossby number in their figure 3. This shows that in their models, the length of the magnetic cycle (measured in rotation periods) is inversely correlated with the Rossby number.

They define the Rossby number in the section "Estimation of the Rossby number" but it's defined in terms of their models, where they have perfect knowledge of the properties. They have two definitions there, the fluid Rossby number (complicated definition related to the coriolis force) and the stellar Rossby number as I described above. They used both and came to the same conclusion of an inverse correlation of the magnetic cycle with the Rossby number.

So, the Rossby number is still defined as the ratio of the rotation period to the convective turnover-time - it's not defined in terms of luminosity. But their model predicts that more luminous sun like stars (of the same mass and other properties) will have a shorter magnetic and sunspot cycle. And in their model the Rossby number is inversely correlated to the period.

---

I'll come back to this a bit later when I have more perspective as right now I'm caught up in the details after reading the paper and suggest a change. Robertinventor (talk) 16:21, 17 July 2017 (UTC)

┌─────────────────────────────────┘
@Pi zero: I've just tried a minor change, but I'd forgotten how it works, that after you do a minor change it warns you that there are pending changes on the article when you try to edit it again. Since I have a lot to do - not lots of words but several parts to fix, and time is short, I think I'll just do them all as one edit, so ignore that pending minor change as I'll put it in the next one. I will

• Fix the magnetic field sentence
• Minor edit the "at least 25" is in fact 27, so why not say 27?
• new version of the Rossby section at the end.

I'll put the edits here first and then copy them over in one go. That way if anyone wants to comment they can do so here and it's easy to edit them. I'll put the planned edits between the lines below

Robertinventor (talk)

---

Planned edits

To build this model, they incorporated data from 27 other stars .

• A. Strugarek, P. Beaudoin, P. Charbonneau, A. S. Brun, J.D. do Nascimento Jr.. "Reconciling solar and stellar magnetic cycles with nonlinear dynamo simulations (preprint on arxiv.org)" — arxiv.org, 13 Jul 2017

Astronomers have known for some time that more active stars have shorter cycles. However, until now this correlation has seemed to be only approximate. In principle it should be easy to estimate how active a star is. It is more active if it spins faster, and it is also more active if the surface layers convect more. If you multiply the spin rate by the convection time, you get the "Inverse Stellar Rossby number" which is a measure of how active it is. (The stellar Rossby number can be defined as the period of rotation of the star divided by the convection time, so the higher this is, the less active a star is)

It's relatively easy to find the spin rate of distant stars, through precise light curves, and through the doppler broadening of spectral lines. However the problems start when you try to observe how fast the surface of a star churns over. We can't do this accurately even for our Sun. Various approaches have been tried to estimate it (including the ratio of the intensity of the calcium ii H and K spectral lines to the star's luminosity), but none was totally convincing. Previous research had suggested that there were two different branches of solar type stars, with our sun in an uneasy and hard to explain transition state between the two branches.

In this study, they used a new approach. They modeled seven different types of star, differing in rotation period and a parameter that describes how they convect heat, and then calculated various parameters for the models including the luminosity, cycle length and the Rossby number . As expected, the cycle got shorter, in their models, as the star spins faster. However they found a new correlation from their luminosity calculations, that the cycle is shorter for slightly brighter stars - these were stars that were brighter only because of increased convection.

That lead them to the idea of measuring the luminosity of the stars in question very precisely. They did these calculations for 27 stars including many that had been studied previously. When they plotted these on a graph (of cycle length against rotation period) along with their modeled stars, and adjusted for luminosity, they found that the empirically observed stars all lay in a broad diagonal band on the graph, with the model stars at the top of the band (figure 2D of the paper). This meant that the predicted luminosity and rotation rates of their models were in correlated with the cycle length in the same way as the observed stars. They then looked at the Rossby numbers for the models. Since these were models, they could just derive the convection time from the model. They found that the length of the cycle (in star rotations) decreases in a close to linear fashion as the Rossby number increases.

With this new research, by taking account of luminosity as well as rotation period, they can place them all within a single branch. This removes the Sun from its "oddball" status it had before, and our Sun now can be treated as a normal star like any of the others in their study. It also suggests that the length of the cycle would turn out to be closely correlated with the Rossby number of the star, if we could only know its real convection time.

---

I know the last few paragraphs are a bit long, but you wanted scientific accuracy - and that's what you have here now. It would be hard to trim it by much without losing accuracy. I'll do the magnetic field sentence next, and then return to this, and see if I can see a neater way to express anything, and double check it for accuracy. If there are no more comments by then, I'll just go ahead and add them given the time constraint. Anyway now I'll try a suggested correction for the magnetic field, not started to think about that but it shouldn't be that hard. Robertinventor (talk) 21:13, 17 July 2017 (UTC)

@Robertinventor: You'll find I'd already dealt with the one pending edit (if you're referring to the exactly 27 stars thing), and in the process I added an external link to the arXiv paper. --Pi zero (talk) 21:23, 17 July 2017 (UTC)

@Pi zero: Oh that's quick, didn't expect it so soon. Okay I'll do the magentic field one directly in the article next after I have a bit of time to think it over. I want to return to the Rossby number paragraphs in case I find a way to improve them before adding them and double check them. Just did a bit of work just now between the lines above. I may do some more editing here before adding them, see how it goes. Probably I should have done it as a sandbox in my user space but I tend to forget about that capability, sorry Robertinventor (talk) 21:31, 17 July 2017 (UTC)

@Pi zero: Okay done the eleven year cycle section. I'll do the Rossby number part later, maybe an hour or two from now.

@Robertinventor: Time is getting tight, for this article. I'm looking over the cycle edit now. --Pi zero (talk) 23:32, 17 July 2017 (UTC)

Okay I'll check and add those paras now. Robertinventor (talk) 23:34, 17 July 2017 (UTC)

┌─────────────────────────────────┘
@Pi zero: Okay done now. I did a bit more copy editing before adding it in, and I've now got html comments throughout if you look at the source, saying which paragraph or section the sentences are based on, which should help with verifying. Some of the references to the paper are done as figure or table numbers visible to the reader, who if interested may want to find the table or figure in the paper and look at it themselves. Hope that leaves you time to review, it thanks! Robertinventor (talk) 00:03, 18 July 2017 (UTC)

@Robertinventor: I feel awful about this, but I'm totally unable to verify that first paragraph, from which all the rest flows — in fact, from my study of the source I think key parts of it aren't true — and we've run out of time. I'll comment further here (which will take a few minutes for me to compose), but, that's the bottom line. --Pi zero (talk) 01:38, 18 July 2017 (UTC)

Fwiw, my run-down on that paragraph:

• The paragraph starts "Astronomers have known for some time that more active stars have shorter cycles. However, until now this correlation has seemed to be only approximate." I tried and failed to verify that.
• It then talks further about how "active" a star is. I didn't find this terminology in the sources. But then, when it goes into the Rossby number, one of the two terms inverts. Having just described the "activity" as the product of rotation rate and convection rate, which is plausible although, as I say, it's not terminology I found in the source(s), it then describes Rossby number — correctly, I believe — as rotation rate times convection time, which is inverse to convection rate. This is, from what I can see, at the heart of the great befuddlement surrounding the Rossby number, that it isn't "activity" at all, but something far more weird.
• With things out of step at the start, afaics the whole structure of the edit is lost. Realizing that there was simply not time for a whole additional round of revise-resubmit-rereview, I made all the effort I could to find some way of rescuing it by some surgical change that would be within my purview as an independent reviewer — such things usually consist mostly of cutting out words and bits of passages and such, since cutting is less author-like than adding — but, no dice.

I'm sad that this hasn't worked out. Ordinarily, of course, anything this tricky would be undertaken before publication, with — ideally — a little more scheduling flexibility (supposing other factors go well: submitted soon after the event, reviewed promptly, no other problems with the article that soak up review time). --Pi zero (talk) 02:06, 18 July 2017 (UTC)

┌─────────────────────────────────┘
@Pi zero: Oh shame it is too late. No problem,, thanks for trying! It's my silly mistake. I thought we had an extra 12 hours for some reason, now I understand why it was so tight. If I'd known. I wouldn't have landed you with all that with just an hour or two to assess it, but would have done something much much shorter.

Also - I think neither you nor @Darkfrog24: nor myself of us need to feel we've done too badly here. After all none of the sources, except the paper itself which is highly technical, explains their research at all clearly. Ours could have been perhaps the only news article out there that really explained what they did, at least of the ones we discovered.

Anyway - yes some of those things I got wrong and would have spotted with another review, but we were out of time and I didn't spot them, sorry. I'll answer anyway, even though it is too late. Also I wonder if I can suggest an edit to the last paragraph which has a few inaccuracies?

So anyway, To answer your points first. I'll do the second one first as it's the most important I think, rather central. The first one is just a mistake I made, will come to that later.

• It then talks further about how "active" a star is. I didn't find this terminology in the sources. But then, when it goes into the Rossby number, one of the two terms inverts. Having just described the "activity" as the product of rotation rate and convection rate, which is plausible although, as I say, it's not terminology I found in the source(s), it then describes Rossby number — correctly, I believe — as rotation rate times convection time, which is inverse to convection rate. This is, from what I can see, at the heart of the great befuddlement surrounding the Rossby number, that it isn't "activity" at all, but something far more weird.

Oh sorry. I was using a different source here. It's from this talk page. I didn't think to give those sources as I was focusing on getting the right sources for the paper itself. It's from page 4 of this thesis [9], says

". The dependence of magnetic activity upon rotation rate was first quantified by Noyes et al. (1984), who used these chromospheric emission lines to establish a close correlation between the magnetic activity of a star and its inverse inverse Rossby number, Ro₋₁ = Ωτ_c, where Ω is the angular velocity of the star and τ_c is the star’s (theoretically predicted) convective turnover time – the larger the value of Ro₋₁, the more active the star."

Basically a star with a high Rossby number is less active. It's an empirical thing. That when you observe the magnetic activity of a star and then you find its rotation rate and work out its theoretically predicted convection turnover time, then the Rossby number tells you how active it is - smaller Rossby number means it is more active.

And - as I said "If you multiply the spin rate by the convection time". The source says "convection turnover time" the extra "turnover" there might have made it clearer. But as you say - the paper itself doesn't give the definition in this form. But it's clearly the same number, it's just that the paper uses the Rossby number instead of its inverse and period instead of angular velocity.

And yes I know that when I define the Rossby number I use periods and not rates. The convection turnover time is the same. The Rossby number though is defined using the rotation period. So it's the rotation period divided by the convection turnover time. That's just because that's how the two sources define it.

It works out fine. The first is rotation rate * convection turnover time. When you invert it, you get 1/ (rotation rate * turnover time) = period of the star / convection turnover time. So they are talking about the same thing.

• The paragraph starts "Astronomers have known for some time that more active stars have shorter cycles. However, until now this correlation has seemed to be only approximate." I tried and failed to verify that.

Sorry - this was a mistake. This is what it was based on:

"Historically, two distinct branches in the rotation-cycle period diagram(3, 5) have been favored in the literature. The Sun lies in between these branches (see Fig 2A), requiring conjecture that it is in a transition state between the branches(24). A third branch showing an anti-correlation between cycle period and rotation period was also identified for slower rotators

If you look at the Fig 2A which they say describes the two distinct branches - you can see those two branches going up from bottom left to top right (I think), with the sun (purple) in between, if I read it right.

So the old hypothesis was that the cycle period correlated to the rotation period - more slowly rotating stars would have longer cycles. Though they did have the idea already for some unusually slow rotators that they were anti correlated. A shame they didn't mark the branches on that diagram to make it clear.

Anyway so their new result is in figures 3C and 3D where you see their blue model values that slope from top left to bottom right correlating with the stars in a rough way once you take account of the luminosity.

So their new result is that the cycle is always anticorrelated with the rotation period. A slower rotating star with a longer period has a shorter cycle.

So it should be

"Astronomers have thought up to now that more active stars had longer cycles"

So you are right to pull me up on this, thanks!

• With things out of step at the start, afaics the whole structure of the edit is lost. Realizing that there was simply not time for a whole additional round of revise-resubmit-rereview, I made all the effort I could to find some way of rescuing it by some surgical change that would be within my purview as an independent reviewer — such things usually consist mostly of cutting out words and bits of passages and such, since cutting is less author-like than adding — but, no dice.

No, there was nothing you could have done. It just wasn't ready, sorry! I wonder if we can fix the last paragraph though, based on what I've found out from reading the paper, which we didn't have before?

I think this is a more accurate version of it:

"They found the Sun operated on the same basic principles as sun-like other stars: The cycle time of the 27 stars studied was found to be driven by their luminosity, and their rotation period. The rotation period and convection turnover timescale are used together to determine a star's Rossby number (= P_rot/τ_c). Small Rossby numbers correlate to more magnetically active stars. The convection turnover timescale can't be observed directly, but their models confirmed this correlation of the Rossby number with the magnetic cycle period and predicted that the cycle period decreases with increasing luminosity, and also with increasing rotation speed. Data from the stars confirmed this."

Amongst other things.

• The 27 stars don't form a stellar system but are from many locations, just a list of sun like stars wherever they happen to be
• The luminosity is not used to define the Rossby number when you check the paper itself, which uses convection time and rotation period
• The smaller the Rossby number, the more active the star, not less active
• Correlate seems better because the Rossby number is not a direct measure of how magnetically active they are but correlates to magnetic activity
• The last sentence then explains what they discovered, in short.

How is that, is that sufficiently minor for a fix now or is it too substantial?

Also, I wonder, just a thought, can I perhaps add some of the more detailed material I prepared (once corrected) as an opinion instead? Not sure how those work? Thanks.

Also - I think there is a value in trying to do something even if one doesn't succeed at it :). Robertinventor (talk) 04:22, 18 July 2017 (UTC)

@Robertinventor: Various thoughts

• Wikinews is very much based on the principle that one learns from each article how to do better on future articles. This is a corollary of our treating each article as a "snapshot in time", to be worked on in a flurry of activity at the time and then archived and so-to-speak set under glass; even if we succeed in publishing an article, we can't continue indefinitely to apply lessons learned from it to further improve that article, so we have to apply the lessons forward.
• From time to time, we manage to have pretty clearly the best coverage of a story anywhere. We can't get our hearts set on it, of course, as it doesn't always work out; individual articles don't even always make it to publication (as one Wikinewsie memorably put it, once upon a time, "Don't marry the article"). We generally have to be satisfied with subtler advantages to our coverage — like our neutrality (I suspect we often have exceptionally neutral coverage, but that's not showy), accuracy here and there, some increased visibility for stories either poorly covered in the msm (mainstream media), or poorly covered in the English msm, our permanently free archives. But, yeah, it would have been good to achieve that, and the bar was tantalizingly within reach since, in this case, the technical aspects were so very hard to get a grip on that other news sources weren't managing it (which was also why we had trouble, of course).
• We're really not supposed to make any substantive changes after the 24-hour horizon. After that horizon, if there's anything wrong (other than typos and such that don't change the meaning), we issue a {{correction}}, which we hate to do while, at the same time, it's a point of pride that we own up to our mistakes when they happen. I hope we don't have actual inaccuracies, but, let's see...

• The "system". Hm. I'd taken the word in a generic sense, rather than in some technical sense of a system of stars. The question becomes whether that's worthy of correction.
• The luminosity is not used to define the Rossby number. My understanding was that it is used in determining the Rossby number. My hope, during the original review, was that adding the words "relating to plasma flow" at the end of that sentence would weaken the suggested connection between luminosity and Rossby number: luminosity is (I thought/think) used to determine the Rossby number, but the number itself relates to plasma flow (I thought/think).
• The word "correlate" might be clearer; I'm hoping it isn't strictly needed.

• There seem to be two different senses of "active" here. There's plasmic activity (I suppose I might have been a bit better off saying that rather than "plasma flow", but perhaps that's overthinking), and then there's magnetic activity. The trouble is that the two don't coincide, because (always supposing I understand what's going on here) the Rossby number correlates with magnetic activity but is a ratio between two kinds of plasmic activity.
• I think it would be fine to offer a remark on the opinions page, if you wish to do so. Comment space is primarily for opinions about the topic of an article (we sometimes call it "troll space", partly affectionately, partly in frustration, and partly just as a reminder that readers are allowed to express non-neutral opinions rather more forcefully there than elsewhere on the wiki); but other sorts of remarks are sometimes made too. There should be a button on the opinions page called "start a new dicussion"; click that.

--Pi zero (talk) 11:22, 18 July 2017 (UTC)

Ouch! You're quite right, though, the last bit is said backward, smaller Rossby versus less/more magnetically active. We're going to need a {{correction}} for that. --Pi zero (talk) 12:19, 18 July 2017 (UTC)

Okay. That would be good. If you are doing a correction I think good to add an extra sentence too, see the end of this reply (which was mid edit when you posted your reply just now).

On your other comments - first - okay I see, if we are only supposed to do typos at this stage, then it's too late to fix the last paragraph except as a correction.

• I'd like to add one other thing to your list of advantages of wikinews. It is the only news outlet I know that as a matter of course always provides its sources. Almost nobody else does that. I'm in the UK and the BBC is one of the most reliable in science (though it sometimes gets things wrong). But it almost never provides sources, except in very rare cases. Sometimes it provides links to Nature articles under the Nature open sharing initiative. That's about it. You have to search the academic literature using clues from the story such as the researcher's names if they are mentioned. Phys.org provides sources. But many even of the more specialist news sources, e.g. most of the online astronomy websites don't. So even if we've got something wrong in this story, an expert can go to the source, and check it, and then - post about it here too.

• The "system". I agree not worthy of correction given the situation that we'd have to issue it as a correction. Anyone with good background in astronomy will know that the idea of a system of 27 sun like stars is improbable to say the least. I doubt if it was ever intended that way. Just me being niggly.

• luminosity - no it is not used to determine the Rossby number directly at least not in this paper. But perhaps it is close enough not to need correction. Their seven star simulations from their model had the property that the stars which had more convective luminosity had shorter cycle periods. I think it would follow that stars with more convective luminosity also have lower Rossby numbers and perhaps it can be used as an empirical proxy for the convection turnover time in the future (increasing that decreases the Rossby number) but they didn't actually use it as an empirical proxy in the paper because they were using models, so they could just work out what the convection turnover time is in their model directly (which you can't do with the Sun of course or stars, so you need proxies). Also they didn't try to calculate the convection turnover time or the Rossby number for the 27 stars in their study. They just did that calculation for their models, and then observed a correlation between the data points for their model stars with the data points for the real stars in their graphs. One suggestion for a minimal, not far off type type edit here would be to change "are used together" to "could be used together". I.e.that potentially their work could be used in the other direction to derive Rossby numbers from the luminosity plus rotation period of a star. But it's a bit original research because they don't say that. They could have added "This may lead to a way of determining the Rossby number empirically from the luminosity and rotation period of a star" - but they don't say that. I'm not sure what one can do about that sentence. Maybe just leave it as is, as a minor thing that's rather hard to fix with a simple correction.

• On "activity" - yes magnetic and plasma activity, go together. The plasma consists of ions and electrons and is a movement of electrical currents. Electricity generates magnetism and vice versa so they go together. So when you get a coronal mass ejection for instance, of course there are distortions of the magnetic field lines. And the magnetic fields guide the charged particles and that forms a lot of the structure of the sun's photosphere. However the Rossby number is defined only in terms of movement of fluid or gas. There is nothing about magnetism or plasma in its definition. The general definition involves coriolis forces, but as applied to stars as the "stellar Rossby number", it's the ratio of the rotation period of the star to its convection turnover time. The convection is thermally driven convection, similar to convection in boiling water, though in a gas in the case of the Sun. It's physical movement of the surface of the sun in convection cells which happens because it is hot underneath and cold above. For more on this see Convection_zone. And rotation of course is physical movement and nothing to do with plasma or magnetism. What happened though was that back in 1984 Noyes et al established an empirical correlation between this number which is just a property of the sun as a rotating heated ball of gas and the activity of the sun in a plasma / magnetic activity sense. That's the cite I gave above from page 4 of this thesis [10].

And as the cite from that thesis says, it's an inverse correlation "the larger the value of Ro⁻¹, the more active the star." (as you agree above :) ).

Also as I said above, the old hypothesis was that the cycle period correlated to the rotation period - more slowly rotating stars would have longer cycles. Though they did have the idea already for some unusually slow rotators that they were anti correlated. (sorry seem to be missing a sentence, this fixes it)

This however is not what the paper discovered. What they found (in their modeled stars where they can calculate it directly) is that the smaller the Rossby number (and so, the more active the star), the longer the cycle. That's their figure 3a. Their figure 2a is similar - it shows that the shorter the rotation period the longer the cycle, when you factor out the luminosity effects and that fits in, because a shorter rotation period means a more active star by that correlation that Noyes et al discovered back in 1984. I hope that's all a bit clearer. It's a little surprising and rather interesting, a star that rotates faster has a longer rather than a shorter cycle.

As for issuing a correction then, yes, the less to more. And then if it is possible to add a last sentence "What they found (in their modeled stars where they can calculate it directly) is that the smaller the Rossby number (and so, the more active the star), the longer the cycle."? That then would fix most of it apart from the luminosity thing where they definitely didn't use luminosity to calculate the Rossby number in the paper, though in principle it might be a useful future empirical proxy (though they don't suggest it as such in the paper either). Do you think we should correct that as well? I corrected it in my suggested new version of the last paragraph above and we could work on that a bit if necessary.

* Okay I'll have a go at saying some of this as an opinion later, will also be interesting to find out how the feature works. Robertinventor (talk) 12:53, 18 July 2017 (UTC)

┌─────────────────────────────────┘
@Pi zero: @Darkfrog24: Just a thought, is there a possibility that the "less" was a typo? I wonder if we can do that one at least as a minor correction? Or did one of the sources say this (if so it's a mistake in the source and so I suppose has to be done as a correction...) Robertinventor (talk) 15:32, 18 July 2017 (UTC)

Sorry was missing a sentence above, would have been confusing. I've just written an "opinion" which is my best understanding now of what the article says, see [11] Robertinventor (talk) 17:38, 19 July 2017 (UTC)

@Robertinventor: I've just been re-examining the last sentence of the article, and comparing it to the abstract of the paper. It looks right to me; not magnificently well-worded, but right.

Our article:

The smaller the Rossby number, the less active the star with respect to magnetic changes.

The paper abstract:

We find that the magnetic cycle period is inversely proportional to the Rossby number[.]

So, the smaller the Rossby number, the longer magnetic cycle period, which is described in our article as "less active with respect to magnetic changes" — "changes" meaning, reversing polarities. Admittedly, the meaning of "active with respect to magnetic changes" could be confusing. I'm wondering if we could justify a very slight tweak to the wording, there. --Pi zero (talk) 14:43, 20 July 2017 (UTC)

@Pi zero: Oh, that's a possibility, interesting thought. As you say, the word "active" is potentially confusing as in the literature on this topic it usually means having more active magnetic and plasma disturbances. But I think it would be accurate to say something like: "The smaller the Rossby number, the longer its activity cycles with respect to magnetic changes.". There the actual observations of distant stars are based mainly on measuring visual changes of activity such as sun spots and effects of the activity on the spectrum. It's harder to measure the magnetic field directly but we do have at least one example, in a faint naked eye F type star (so not actually a solar type star), Tau Boötis - [12], [13], with a proposal to monitor many stars in a dedicated observation program [14]. The observations this study is based on though are visual ones. But their models do model magnetic reversals if I understand right, so the last sentence would be an accurate statement about their models, and a hypothesis they have for the actual stars. So I think in that form it would be reasonably accurate. Could that be a minor edit, or something along those lines? Robertinventor (talk) 07:37, 21 July 2017 (UTC)

@Robertinventor: Thinking of this in terms of minimal changes, what if we were to change the final word of the sentence from "changes" to "reversals", so that the sentence reads

The smaller the Rossby number, the less active the star with respect to magnetic reversals.

(This has a minimality-of-edit advantage over changes to the word "active", which is more structurally central to the sentence.) --Pi zero (talk) 11:02, 21 July 2017 (UTC)

@Pi zero: Okay. It's a compromise but not seriously misleading. Given the constraints, that we don't want to issue a correction if we can avoid it, I'm okay with that. Robertinventor (talk) 22:16, 21 July 2017 (UTC)