255 replies to this topic

[mwr]

https://www.astronom...org/?read=15800

 

RW Cep could be an interesting target for low and high resolution spectroscopy:

 

[robin_astro]

Some PEP photometry in the IR could be interesting,  particularly H and J bands. The "great fade" of Betelgeuse was mainly just in the visible. If you looked in the IR the luminosity  was hardly changing at these wavelengths.

 

 

Cheers

Robin

[flt158]

Thank you for giving us the heads up on this variable star, mwr and Robin!

 

I'm noticing on www.aavso.org that RW Cephei is a wonderfully complex star from the point of view of its spectral class. 

It can vary from G8 to M2. 

Presumably it is at its faintest when it reaches M2. 

 

I must seek this star soon. 

We're supposed to have clear skies over Ireland tomorrow (Sunday) night. 

 

Just over a week ago, one observer said that the star was magnitude +7.8. 

Normally the RW Cep varies between 6 to 7.6. 

So maybe it's gone fainter than what is the norm. 

 

Clear skies from Aubrey. 

Edited by flt158, 10 December 2022 - 04:17 PM.

[robin_astro]

Here is a spectrum tonight (R=1000 using an ALPY 600 with a narrower than usual 10um slit)
https://britastro.or...hp?obs_id=13110

It is very red but clearly heavily reddened by interstellar or circumstellar dust

 

VSX suggests a very wide range of spectrum classifications ranging from G to M2 but Keenan suggests K2i in his 1989 catalogue.
https://articles.ads...ApJS...71..245K

Using the B,V brightness values in SIMBAD, this would imply for a K2i star an E(B-V) = 0.99. Dereddening my spectrum by this amount indeed gives a good match to the Pickles standard for K2i both in terms of the continuum shape and more importantly in the spectral line details (My spectrum is a slightly higher resolution which would explain the higher apparent intensity of the lines in my spectrum)
https://britastro.or...hp?obs_id=13110

 

I suspect the M2 classification may have come from the colour index which would match an M star without reddening. The broad continuum slope of the spectrum reddened by dust approximately matches that of an M star but the lack of molecular bands in the spectrum immediately shows we are dealing with a hotter star here (significantly hotter than Betelgeuse)

 

Cheers

Robin

Edited by robin_astro, 11 December 2022 - 05:58 PM.

[mwr]

Here is a spectrum tonight (R=1000 using an ALPY 600 with a narrower than usual 10um slit)
https://britastro.or...hp?obs_id=13110

It is very red but clearly heavily reddened by interstellar or circumstellar dust

 

VSX suggests a very wide range of spectrum classifications ranging from G to M2 but Keenan suggests K2i in his 1989 catalogue.
https://articles.ads...ApJS...71..245K

Using the B,V brightness values in SIMBAD, this would imply for a K2i star an E(B-V) = 0.99. Dereddening my spectrum by this amount indeed gives a good match to the Pickles standard for K2i both in terms of the continuum shape and more importantly in the spectral line details (My spectrum is a slightly higher resolution which would explain the higher apparent intensity of the lines in my spectrum)
https://britastro.or...hp?obs_id=13110

 

I suspect the M2 classification may have come from the colour index which would match an M star without reddening. The broad continuum slope of the spectrum reddened by dust approximately matches that of an M star but the lack of molecular bands in the spectrum immediately shows we are dealing with a hotter star here (significantly hotter than Betelgeuse)

 

Cheers

Robin

Hi Robin,

 

great spectrum (the first spectrum in the BAA database of RW Cep!). James Foster took a spectrum of RW Cep back in 2020 (see AAVSO database: https://app.aavso.org/avspec/obs/5722).

 

Have you had a chance to compare your spectrum with James' one and to look for differences? I'm travelling right now and don't have access to my PC with all the spectroscopy software installed.

 

Best regards

 

Matthias

[robin_astro]

 

 

Have you had a chance to compare your spectrum with James' one and to look for differences? 

 

Interesting. I don't know what to make of it. The features are similar but the continuum shape is very different. The ratio curve (in red) does not look like additional dust extinction for example  which I would  expect to have a convex shape (flatter in the red steepening towards the blue)

 

Cheers

Robin 

[mwr]

May be James did not deredden his spectrum?

[robin_astro]

May be James did not deredden his spectrum?

My dereddened spectrum in my original post 

https://britastro.or...hp?obs_id=13110

is just a hypothesis as we don't have an independent measurement. Neither spectrum in the above plot is corrected for IS extinction.  Both are as measured (ie just corrected for atmospheric extinction) straight from the respective databases.  If dimming was due to say an outburst of dust as was suggested for Betelegeuse the ratio should resemble the characteristic absorption curve of the material. It does not match typical interstellar dust here.

 

I have had a quick look at James' spectrum in more detail. I was not able to find a good fit to any combination of spectral type and interstellar extinction so the validity of the continuum shape there perhaps needs checking and of course we should get some independent confirmation of my spectrum too.

 

Cheers

Robin

Edited by robin_astro, 12 December 2022 - 07:24 AM.

[robin_astro]

There is a single very high resolution  spectrum of RW Cep from 2005 in the ELODIE archive

http://atlas.obs-hp.fr/elodie/

Here it is, filtered to roughly match the resolution of my spectrum and overlaid  (I have cropped the blue end where the flux calibration of ELODIE spectra is notoriously inaccurate)

 

 

 

 

 

[Organic Astrochemist]

I don’t think that dereddening is necessary because interstellar extinction will be divided out in the ratio.

Bands of TiO are much more apparent in James’s spectrum than in Robin’s. So something further away from the star is blocking these. Actually I think these show up as bumps in the ratio spectrum.

The ratio spectrum shows that bluer light is being selectively blocked, consistent with extinction caused by scattering due to dust.

Robin correctly notes that it doesn’t look like interstellar extinction, which is generally caused by single scattering events because dust grains are far apart and scattered light is scattered out of the line of sight and unlikely to be scattered into the line of sight.

But with circumstellar dust scattering the concentration of dust is high and light can be scattered many times into the line of sight. Apparently this produces different extinction curves.

https://iopscience.i...X/823/2/104/pdf

https://iopscience.i...4357/ac86d9/pdf

I’m no expert on circumstellar dust but it seems that small grains would form first and grow. Extinction by dust is very sensitive to grain size, so maybe some evolution will be seen.

Edited by Organic Astrochemist, 12 December 2022 - 08:44 AM.

[Organic Astrochemist]

The ratio curve (in red) does not look like additional dust extinction for example which I would expect to have a convex shape (flatter in the red steepening towards the blue)

Cheers
Robin

As I drove to work today a very dense fog was lifting. Most of the time the sun was invisible but there were patches of reduced optical depth. The disk always appeared white but the light around the sun was a pale but distinct shade of blue. Red was nowhere to be seen. Assuming the fog was roughly isotropic around the sun, although red light was less likely to be scattered than blue, red light that was scattered out of my line of sight was less likely to be scattered back into my line of sight.

In this situation blue light that had been scattered out of my line of sight was very likely to encounter another scatterer and be scattered back into my line of sight around the sun. At sunset however, blue light scattered below my line of sight to the sun is blocked by the earth and blue light scattered above my line of sight to the sun encounters progressively fewer scatters and is less likely to be scattered back into my line of sight. The light is also refracted. The result is a red sun at sunset.

Edited by Organic Astrochemist, 12 December 2022 - 09:42 AM.

[yuzameh]

I suspect the M2 classification may have come from the colour index which would match an M star without reddening. The broad continuum slope of the spectrum reddened by dust approximately matches that of an M star but the lack of molecular bands in the spectrum immediately shows we are dealing with a hotter star here (significantly hotter than Betelgeuse)

 

It was Keenan himself who first classified the star as M0:Ia as shown in the original MKK book/paper in 1942.  Changed his mind over the years thanks to improved spectra or did the thing change since the early 1940's?  Probably never know, but the former is the more parsimonious interpretation.

 

Here's an image of the photographic spectrum he likely used from a paper of his a year later

 

https://articles.ads...000464P015.html

[yuzameh]

I don’t think that dereddening is necessary because interstellar extinction will be divided out in the ratio.

Bands of TiO are much more apparent in James’s spectrum than in Robin’s. So something further away from the star is blocking these. Actually I think these show up as bumps in the ratio spectrum.

The ratio spectrum shows that bluer light is being selectively blocked, consistent with extinction caused by scattering due to dust.

Robin correctly notes that it doesn’t look like interstellar extinction, which is generally caused by single scattering events because dust grains are far apart and scattered light is scattered out of the line of sight and unlikely to be scattered into the line of sight.

But with circumstellar dust scattering the concentration of dust is high and light can be scattered many times into the line of sight. Apparently this produces different extinction curves.

https://iopscience.i...X/823/2/104/pdf

https://iopscience.i...4357/ac86d9/pdf

I’m no expert on circumstellar dust but it seems that small grains would form first and grow. Extinction by dust is very sensitive to grain size, so maybe some evolution will be seen.

There is a big difference for circumstellar dust, and as you imply forward scattering is a 3D thing, not just 360 degrees isotropic.  There are lots of different types of circumstellar, the usually main physical and chemical composition difference lies between accretion and decretion generated discs/haloes.

 

Some conventions and historic reasons lead to terms and ideas being named backwards (current flows opposite to how the later discovered actual electrons flow, and evolution is the result of descent via natural selection but people, even Uncle David, always express it in terms sounding of designed advancement in what often sound like Lamarkian terminology) which can mislead thoughts sometimes, or cause them to be boxed in outlook.

 

Reddening is a misnomer, it is in fact deblueing (debluening?).

 

Post AGB stars will throw off material.  UXORs, UX Orionis variables which are young stellar objects, on the other hand might do a bit of both.  They have accretion discs, but they are hot Ae/Be stars so may be able to throw off material as well.  This isn't a hot star though, but it is a rapidly evolving hypergiant it seems and in that case will have some analogous behaviour.  I suppose it could be said hypergiants are all technically kind of YSOs as they don't live very long at all, but this one could also be said to be post AGB because it's a cool red one rather than a hot blue one.  That's why these sorts of objects can be fascinating.  Hypergiants can include things like rho Cas, FS CMa and VY CMa, all similar and individually unique, as indeed this one is.

 

The reason for mentioning UXORs is because the deeper their minima, which are caused by circumstellar ejecta, the bluer the colour becomes.

 

The current conjecture for this blue bias, or at least one of the conjectures, is that this is due to the material being clumpy.  Again, I think this is being said backwards in that clumps at least imply holes/gaps for the un-debluened light to show.  Dense enough clumps would otherwise block the light altogether.  Compare this to "sooty" RCB dimmings.

 

There was a lot of hype about alpha Orionis at first but it seemed at the time that the multiple periodicities only had to sync, harmonize, every now and again for TiO and other materials to drop out when all the periodicities summed lead to a cooler than usual state.  This kind of seems to have happened from latest reports, however they suggest that whatever the material it was also expelled somehow, whether just by radiation pressure or otherwise I've no idea.

 

As already mentioned, J and possibly H band photometry would map the photospheric lightcurve changes.  Visual ones too, probably with V-Ic being better than B-V, although B-V should give a hint of any blueing at minima.  J-H would probably show photospheric colour whilst J-K would show up as infrared excess due to the circumstellar material.  However it is overexposed in 2MASS and would probably be so in most professional observatories.

 

There seems, according to adsabs, have been a fairly long history of polarimetry measures of this star over time by Russian astronomers that appeared in the original PZ, but that's in Russian and I can't find any scanned or archived online source for those.

 

Otherwise there doesn't seem to have been much past work on this star with which to compare modern two colour photometry, least of all infrared photometry, or polarimetry (which could be very useful) with.

 

This means that the amateur variable star groups will need to keep following it, especially with photometric and spectroscopic equipment, for many a year after this event ends.

 

Assuming the star goes back to normal.

 

IRSA at IPAC seems to be down at the moment, but various infrared source images can be found below showing their no obvious nebulosity, though with the glare from this relatively bright star that's no big surprise.

 

https://darts.isas.j...ataset=dataset1

 

Hopefully someone at NOEMA will be tempted to look at it, it's likely too far north for ALMA.  There are probably other mm wave interferometers about but those are the two that seem to be in the news most and are new and big.

 

GCVS has it as Lc traditional and Lb in their newer prototype variability nomenclature.  I doubt it has ever really been a Mira as some books claim, it looks to have always been of no higher amplitude than a Semi-Regular.  I used the BAAVSS archival visual data, as it is far more homogeneous than AAVSO's, especially since they started lumping in every other groups' data given they all use slightly different estimation methods, some of which work in decimal steps which don't fit well with the logarithmic response of the eye.  Unfortunately, despite using the fractional method, most BAAVSS data is only reported to one decimal, which is a big problem when trying to find periods in red semiregular low amplitude long period variables.  I couldn't find any obvious period, and certainly couldn't confirm any 343 day period or any period in that range.  Even though this star is circumpolar from the UK such stars can still end up with aliasing in their period determinations as fewer visual observers tend to follow stars when they are below the Celestial Pole.  Nevertheless 343 is a fair bit off from an annual alias.  There were strange peaks at 400 and 500 days, very coincidental sounding numbers, even if just as likely to appear as any other numbers.  The visual lightcurve does look more like a multiperiod Semi-Regular variable than an Lc, which means long period supergiant of irregular periodicity.

Edited by yuzameh, 13 December 2022 - 10:47 PM.

[robin_astro]

Two further repeat spectra from last night using two different reference stars for flux calibration to double check the validity of the continuum in my spectra.

 

 

I am satisfied we now have a reliable spectrum to track any future changes. 

 

Cheers

Robin

[Organic Astrochemist]

Looks good Robin. But I’m curious, if you take a ratio of two of these do you see any difference between nights or at different elevations (I think the reference star is supposed to take care of this).

[yuzameh]

Whilst reading up on this star here and there, on and off, in the few papers available online that actually deal with it rather than just list it in a table or mention it in passing relative to another star, I've noticed a couple of little things.

 

The first M0 I spectrum was given by Keenan in 1942 in the MKK seminal writeup, as already mentioned.  Morgan, the M of MKK, was the one who classified it as G8 Ia (G8.5 Ia) in 1950 with co-author Nancy Roman.  Then Keenan, again, decided it was the latest favourite of K0 Ia around 1979 or 1980.  In other words, these  classifications came from a small group of people who probably could be called the fathers of modern astronomical spectroscopy (as opposed to the great grandfather, "The Angel of Death", and the grandmothers, the ladies working on the HD).  I don't think it is just to do with them having better equipment and resolution over time either.

 

There's a habit for people to describe it as a yellow supergiant (although no one outright calls it an SRd), and of comparing it with VV Cephei (which is a VV Cep variable rather than a symbiotic one, despite having a hot blue dense companion).  Nearly everyone calls the spectrum, and the "continuum", which there isn't really much of in terms of a smooth length of curvature in the plots, complex.  Possibly there's a chance that it isn't so much differing and increasingly better equipment over time, nor even marked variation in the star over time, but more of a case of what bit of the spectrum is noticed at the different times.

 

Probably any reference spectrum is as good as any other due to complexity and stochastic small variation even when taken by the same kit.

 

There's a shortage of photometry for this one in the online sources, I couldn't find any ZTF data, there was no CRTS, so only ASAS-SN gives a lightcurve.  This has a dip in it, a well formed dip covering a period of time, with a suggestion of a relaitevly short lasting minimum in relatively recent times when it went below V = 8.  However, there are two camera lightcurves from ASAS-SN for this object and the second one doesn't confirm it, but it doesn't negate it either as it has no data for that same duration of time.  Simplest interpretation then is that that dip is a glitch in their system somewhere because there is no confirmation of it anywhere, least of all independent confirmation.  The lightcurve at the top of the thread might contradict the ASAS-SN one in question, but its gapped at just the wrong place, although it is a very short gap.  My bias is towards a glitch.

 

The longer the wavelength the brighter it gets, right up to the Q band at around 21 microns, and on into the far infrared, so plenty of circumstellar stuff it seems.

 

IPHAS DR2 shows a sizeable Halpha minus SDSS r' excess but there's no hint of Halpha emission in the above.

 

Far too little data, some contradictory, and not really saying much when taken together as a whole, I think the study of this star starts now and goes on into the future, rather than comparing with the past.

[robin_astro]

Looks good Robin. But I’m curious, if you take a ratio of two of these do you see any difference between nights or at different elevations (I think the reference star is supposed to take care of this).

Yes all taken care of in the (relative) flux calibration using reference stars though the telluric bands have been left in as is standard for the BAA and other databases. The spectrum is unchanged within the uncertainties. Here are the two nights (red,blue) overlaid and the ratio in green

 

 

The noise increases rapidly into the violet where the star is faint and the sensitivity low. The increasing difference  in the continuum  towards the blue is likely due to a combination of systematics (chromatism in the telescope optics combined with focus shift but also atmospheric dispersion and extinction differences, all of which I try to minimise using reference star measurements before and after but  +-10% uncertainty in the relative SED at the blue end is typical for my setup.

 

Cheers

Robin

[robin_astro]

The increasing difference  in the continuum  towards the blue is likely due to a combination of systematics (chromatism in the telescope optics combined with focus shift 

This plot shows the difficulties of getting an accurate continuum with achromatic telescope optics that I mentioned in this thread.

https://www.cloudyni...-ha/?p=12333379

Red and Blue are the measured system response on the consecutive nights using the same reference star under nominally the same atmospheric conditions. Green is  the ratio. 

 

 

The difference is due to a difference in focus of the star on the slit between the two nights, giving selective sampling by wavelength of the chromatic star image. With achromatic telescope optics getting the correct continuum for the target depends on the focus being kept the same for both target and reference measurements. All reflective optics do not suffer from this affect, though atmospheric dispersion can still cause a similar affect, reduced by running the slit at the parallactic angle ie vertical relative to the horizon)

 

Cheers

Robin

[VY Canis Majoris]

I would like to thank mwr greatly for reporting this for me. I had noticed this dimming as early as the 20th of November, and reported it to the AAVSO forums on the 2nd of December.

I have noticed a couple things that might be of interest.

Firstly the historical light curve data from AAVSO shows that the star's average brightness has increased slightly over time since the 1950s, up until the peak in the start of 2020.

Secondly there is a second spectrum from AAVSO taken in June of 2021, coinciding with a visual dip in the light curve, which shows the star much bluer than usual. This strikes me as odd since pulsating variable stars are usually hottest at their maxima and coolest at their minima. I suggest this spectrum should be looked into in great detail.

Has the possibility of RW Cep being a binary been ruled out yet?

[mwr]

Secondly there is a second spectrum from AAVSO taken in June of 2021, coinciding with a visual dip in the light curve, which shows the star much bluer than usual. This strikes me as odd since pulsating variable stars are usually hottest at their maxima and coolest at their minima. I suggest this spectrum should be looked into in great detail.

 

Hi VY,

 

Tim Stone's spectrum (taken in June of 2021) has been recorded using a slitles SA-200 grism setup and most probably has not been response corrected. Robin is an expert in relative response correction issues and might give further comments and explanations on that. (Information on wether a response correction was applied or not is generally not given in the AAVSO spectrum database; the BAA database is more detailed in this respect and provides this helpful information).

Edited by mwr, 15 December 2022 - 11:45 AM.

[robin_astro]

 Robin is an expert in relative response correction issues and might give further comments and explanations on that. 

The fits headers do not contain details of exactly what processing was done and  I can only speak for the accuracy of my spectra but the continuum of both spectra in the AAVSO database are certainly very different from each other and from mine (which compares well with the professional ELODIE spectrum).

 

However the most  important information about the star, such as temperature, which is used to classify it  is contained in the lines, not in the shape of the continuum (particularly here as we have significant dust extinction) and all spectra have sufficient resolution to show some important features so rectifying all the spectra (ie normalising the continuum to 1 everywhere)and comparing them against each other and against MK standards could be a useful exercise.

 

Cheers

Robin

[yuzameh]

... mine (which compares well with the professional ELODIE spectrum).

So you're saying there's actually not much significantly different in the optical spectral spectrum at the moment compared to normal maximum?

 

I've been looking at some of these ELODIE and what I think are some PolarBase spectra (uncertain provenance, I found them via VizieR "obscore") and I've decided the NaI D doublet is from the star and not atmospheric or interstellar as the radial velocity of the object appears to be variable, double checked with Halpha when possible.  I think I see the D line as one trough in your spectrum?

 

thanks

[VY Canis Majoris]

By the way, we need to make an effort into monitoring this star soon, by taking photometry (hopefully in multiple bands) and spectra every couple days.

[robin_astro]

So you're saying there's actually not much significantly different in the optical spectral spectrum at the moment compared to normal maximum?

 

 

I am talking specifically about the differences in the continuum here. The large differences in the continuum shown in the AAVSO spectra and compared with mine and  the ELODIE spectrum are surprising to me. The changes in the Betelgeuse spectrum from maximum to minimum and during the great fade for example were much more subtle so I am suggesting we should perhaps treat those spectra showing significant differences in the continuum with caution and just concentrate on any differences the spectral features they show until we can confirm the reliability of the flux calibration (Getting good flux calibration (even relative) is challenging for amateurs.)

 

 

 

I've been looking at some of these ELODIE and what I think are some PolarBase spectra (uncertain provenance, I found them via VizieR "obscore") and I've decided the NaI D doublet is from the star and not atmospheric or interstellar as the radial velocity of the object appears to be variable, double checked with Halpha when possible.  I think I see the D line as one trough in your spectrum?

 

thanks

This is the ELODIE spectrum I was referring to

 

http://atlas.obs-hp....um&c=o&o=RW Cep

 

Here is the NaD line overlaid on my 2022-12-12 spectrum (barely resolved at the much lower resolution of my spectrum)

 

 

The D lines are broad and although saturated show significant structure. Doesn't this imply an interstellar origin ?

(To put a maximum value on it, The total galactic IS extinction E(B-V) in this direction from NED is 0.746)

 

Cheers

Robin

Edited by robin_astro, 16 December 2022 - 05:54 AM.

[robin_astro]

Here for comparison is how ELODIE sees the same region in Betelegeuse