Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Pushing the limits of the ALPY 600 spectrograph -- Be binary stars

  • Please log in to reply
15 replies to this topic

#1 Organic Astrochemist

Organic Astrochemist

    Viking 1

  • -----
  • topic starter
  • Posts: 794
  • Joined: 10 Jan 2015

Posted 02 January 2018 - 03:21 PM

Classical Be stars are fast rotating stars that have a disk of circumstellar material that is irradiated by the star, resulting in emission lines that can be seen spectroscopically.

 

Often the brightest emission line (H-alpha) is asymmetric. The shorter wavelength side is called violet and the longer wavelength side is called red, the ratio is V/R and is variable. This has been explained by both binary companions and also by a spiral arm distribution of material in the disk.

 

A recent article by Miroshnichenko suggests that at least half of all Be stars might be binary systems.

 

Sometimes the period of the V/R variations and the binary companion are the same (phase locked) and sometimes, like in the case of Zeta Tauri, the two are different.

 

When Classical Be stars are observed edge-on (high inclination angle) material in the disk along the line-of-sight absorbs light emitted by the disk (and star) and causes a deep absorption line. In high-resolution spectroscopy, this is a typical example of Zeta Tauri (taken from the BeSS database taken by Guarro Fl√≥ on 12/18/2017)

zeta tau.png

 

Often such detail is not visible with low resolution spectroscopy, like that obtained with my ALPY 600, but two recent articles by Panaglou et al 2016 and Panaglou et al 2017, if I understand them correctly, suggest to me that some binary Be stars might offer some opportunities. Briefly, the companion eats away at the outermost portion of the disk (they call it truncation) and this is where the slowest, sharpest, highest peak of H-alpha is formed. When combined with the absorption of the shell geometry, the deep absorption splits the H-alpha peak enough that the two peaks can become visible with the ALPY (although not like in high resolution spectroscopy where the absorption goes below the continuum).

 

Here's my spectrum of Zeta Tauri

Zeta Tauri 12 30 2017.png

and here's a zoom in on the H-alpha region, clearly showing the split peak.

Zeta Tauri halpha.png

 

Finally, here's a comparison between just averaging 16 sequential spectra (in black) vs averaging 19 spectra that I selected because they had especially clear double-peaked structure. I noticed that the degree of double-peak structure wasn't constant, but varied. This could be because of my mount/guiding or because of seeing. I'm not sure, but the selection did produce a slightly deeper trough and a slightly higher R peak.

Zeta Tauri average vs selected.png

 

I think this adds another exciting aspect to observing Be stars with low resolution spectroscopy: looking for flat or double peaked H-alpha peaks. While this feature, and the binary nature of the system, won't be confirmed by low resolution spectroscopy, I think the ALPY 600 can give results consistent with the spectrum with high resolution spectroscopy and can suggest that further inspection of the system is warranted.

 

Thanks for reading.


Edited by Organic Astrochemist, 02 January 2018 - 03:32 PM.

  • old_frankland, Ed Wiley, Anthony236J and 1 other like this

#2 robin_astro

robin_astro

    Viking 1

  • -----
  • Posts: 534
  • Joined: 18 Dec 2005

Posted 03 January 2018 - 01:01 PM

An interesting observation but I would advise caution when attempting to interpret variations which are smaller in wavelength scale than the intrinsic resolution of the instrument. For example the  wavelength range covered in the zoomed in plot falls within the width of the slit so any distortion of the point spread function of the star image (eg from secondary mirror shadow) or systematic guiding errors favouring the edges of the slit over the centre (as PHD tends to do when guiding on a split image) will appear as features in the spectrum smaller than the width of the slit.

 

Cheers

Robin


  • old_frankland likes this

#3 robin_astro

robin_astro

    Viking 1

  • -----
  • Posts: 534
  • Joined: 18 Dec 2005

Posted 03 January 2018 - 09:04 PM

There is a long history of amateur monitoring of the H alpha line profile of Zeta Tau. For example Ernst Pollman has just published an analysis of the variability of the central absorption on the ARAS forum

http://www.spectro-a...php?f=14&t=1934

 

Cheers

Robin


  • old_frankland likes this

#4 Organic Astrochemist

Organic Astrochemist

    Viking 1

  • -----
  • topic starter
  • Posts: 794
  • Joined: 10 Jan 2015

Posted 04 January 2018 - 10:51 AM

An interesting observation but I would advise caution when attempting to interpret variations which are smaller in wavelength scale than the intrinsic resolution of the instrument. For example the  wavelength range covered in the zoomed in plot falls within the width of the slit so any distortion of the point spread function of the star image (eg from secondary mirror shadow) or systematic guiding errors favouring the edges of the slit over the centre (as PHD tends to do when guiding on a split image) will appear as features in the spectrum smaller than the width of the slit.

 

Cheers

Robin

I think I did observe such spurious images in some of the spectra I collected, and yet, when averaged out, the errors seemed to cancel out and the double peak feature endured.

 

Obviously this is not adequate for quantitative measurements, but qualitatively it does seem consistent with the high resolution spectroscopy. In addition to measuring the depth of the central absorption as Ernst Pollman did, some have found variations in the heights of V/R in Zeta Tauri (period=1503days!). See Stefl et al. especially Figure 3. Presently the V peak is greater than the R peak. It will be interesting to see if I can see evolution of this ratio over time, at least in a qualitative way.


  • old_frankland likes this

#5 robin_astro

robin_astro

    Viking 1

  • -----
  • Posts: 534
  • Joined: 18 Dec 2005

Posted 04 January 2018 - 01:31 PM

The problem is what you are seeing at this scale are potentially just systematic artifacts due to the small star image size produced by your fast small aperture scope . When following features at this scale you lose the advantages of the slit and are effectively operating slitless (ie you are relying on the star image being smaller than the slit to increase the resolution) so you are very much at the mercy of the focus, seeing and guiding. (With your 6 inch f6 scope the standard 23um slit is ~6 arc sec wide on the sky which is twice the FWHM of your star image for typical 3 arc sec seeing).  If you are genuinely seeing features at smaller scales than the intrinsic resolution as defined by the slit width then a better approach would be to reduce the slit width which would give a reproduceable increase in resolution. Although they do not advertise them I believe  Shelyak can supply different width mirror slits which might better match the size of the star image and give you higher resolution without significant light loss. 

 

Cheers

Robin


  • old_frankland and Anthony236J like this

#6 Organic Astrochemist

Organic Astrochemist

    Viking 1

  • -----
  • topic starter
  • Posts: 794
  • Joined: 10 Jan 2015

Posted 04 January 2018 - 03:38 PM

Thank you Robin for your wise and knowledgeable comments.

Here in Florida it is quite common for us to get at least 1 arc sec seeing and I easily split subarcsecond doubles. However, when I look at my guiding camera, the star image of these bright stars always looks much bigger than the slit. Is that because the camera is capturing much more of the point spread function than just the FWHM? Thanks for the tip about Shelyak. I will contact them.

I never thought of f/6 as a fast scope. I was going to get a 6" f/4, but I read somewhere that f/5 was optimal for the ALPY 600. I think it is a testament to the design of the ALPY 600 that I can use it (350-750 nm) with the relatively small camera I have, compared to larger CCD cameras. I guess at f/4, the spectrum would fit on an even smaller chip.

 

I captured a few other spectra that night, but I'm finding that the naming protocol that I use isn't very helpful for working with ISIS. Can you please suggest how you name your files and what directories you store them in. Do you copy your spectra into a working directory? Can you keep the bias, dark and flat files in another directory? Seems like small things, but although I have processed a few spectra with ISIS, it is like black magic to me. It seems rather particular about file names and directories.



#7 robin_astro

robin_astro

    Viking 1

  • -----
  • Posts: 534
  • Joined: 18 Dec 2005

Posted 05 January 2018 - 09:30 AM

Provided the psf of the star is roughly Gaussian in shape the FWHM should include around 75% of the total flux. If the star image is overexposed though the star can look much larger. When designing a spectrograph the slit is typically sized to match the FWHM. (slightly more flux could be included by making the slit wider but at the expense of loss of resolution, similarly a narrower slit would increase resolution but at the loss of efficiency).  There is discussion on the effect of slit size on throughput here for example

https://spectroscopy...lator/#more-525

 

 

The length of the spectrum depends only on the spectrograph optics (specifically the line density of the grating and the focal length of the camera lens.) The resolution in that length of spectrum then depends on the width of the slit. (or more precisely the width of the slit image in the spectrum). A narrower slit will give higher resolution (up to a point, the slit image still needs to be at least 2-3 pixels width) but we still need to fit the starlight though the slit so we would need a smaller star image. (ie good seeing and a short focal length)  A faster telescope will give a smaller star image but the spectrograph optics need to be at least as fast to accept the full light cone from the starlight passing through the slit. The ALPY optics are f4 but like most optics, give a sharper image when operating slightly stopped down hence the recommended f5 for the ALPY.  The 23um standard slit for the ALPY equates to a 1600mm focal length with 3 arcsec seeing

 

There are various ways of using ISIS but  I make a new directory for each observation and assemble all the files I need there before running ISIS. I then archive that directory with all the input and output files. This leads to some duplication as I typically may use the same master darks, flats, wavelength calibration, instrument response etc for several obervations but it does mean that I can always return to that directory at any time and rerun the data reduction again. (The html file that ISIS generates for each run is very useful in that respect as it contains all the setup configuration used for that run)

 

Cheers

Robin


Edited by robin_astro, 05 January 2018 - 09:56 AM.


#8 Organic Astrochemist

Organic Astrochemist

    Viking 1

  • -----
  • topic starter
  • Posts: 794
  • Joined: 10 Jan 2015

Posted 06 January 2018 - 11:34 PM

Thanks Robin. You were so right about me effectively performing slitless spectroscopy within my ALPY 600 slit.

I have admired the Beta Monocerotis triple system many times. But I am amazed to learn that this is the mother-of-all Be binary systems -- three stars, all Be stars with H-alpha emission! Each star had a distinctive spectrum.

In the guide camera, system looked like this. I could resolve all three stars, but to me the seeing was far from excellent. 

Beta Mon.jpg

AB are separated by 7.4" and BC are separated by 2.8"

This is the spectrum of Beta Monocerotis A (which I note hasn't had a recorded observation in BeSS database since 12/2016)

A was on the slit and B and C were not

Beta Mon A.png

B and C are only separated by 2.8", but they were oriented nearly parallel to the slit, so I put them both on the slit and acquired the following spectrum (click to zoom):

Beta Mon BC.jpg

To my surprise B and C, although they were both on the slit, were separated far enough vertically that I could get two different spectra -- each was different.

I couldn't find any spectra in the BeSS database.

This is the spectrum for Beta Mon B

Beta Mon B.png

This is the spectrum for Beta Mon C

Beta Mon C.png

For this star, H-beta emission was also visible. I think that is a lot of detail for effectively a slitless system.

Beta Mon C h-beta.png


Edited by Organic Astrochemist, 06 January 2018 - 11:34 PM.

  • old_frankland and robin_astro like this

#9 robin_astro

robin_astro

    Viking 1

  • -----
  • Posts: 534
  • Joined: 18 Dec 2005

Posted 07 January 2018 - 08:38 AM

That Beta Mon B/C spectrum split image is really cool !  Are you submitting observations to BeSS? 

I envy your seeing. With that seeing you could put an ALPY on a really big aperture scope and go seriously deep.  Anyone got a 30 inch f5 standing around doing nothing ? 

 

Robin



#10 robin_astro

robin_astro

    Viking 1

  • -----
  • Posts: 534
  • Joined: 18 Dec 2005

Posted 07 January 2018 - 08:49 AM

Provided the psf of the star is roughly Gaussian in shape the FWHM should include around 75% of the total flux. If the star image is overexposed though the star can look much larger. 

Also the guider optics are achromatic in the ALPY so the image in an unfiltered guide camera are probably not as sharp as they could be. (I wondered in the past about adding a UV/IR block to the guide camera but currently run unfiltered to maximise the brightness for faint objects)

 

Robin


  • old_frankland likes this

#11 James Foster

James Foster

    Lift Off

  • -----
  • Posts: 11
  • Joined: 10 Sep 2011

Posted 25 January 2018 - 06:15 PM

RE:"I captured a few other spectra that night, but I'm finding that the naming protocol that I use isn't very helpful for working with ISIS. Can you please suggest how you name your files and what directories you store them in."

Yes I was confused at first when using IS.  IS accepts most naming convention in the prefix (objectname), but requires the user to adhere to strict convention in the suffix (end of filename).   Here is a screen shot of how I organize files:

IsisPROscreenShot.jpg

 

Note that for the object EG And I have 2 sets of files, one (Maxim DL Generated...Eg And-001) vs two (Isis renamed file Eg And-1).  All you need to to do is have it in this format for IS processing.  I use the Isis function on tab Misc, sub-tab, MaxImDL -> ISIS, to rename my object and object calibration (say Eg And_neon-001) files.  Also note that I have a sub-directory called "calib" that contains the 3 major calibration files, cosme, dark900_9, and offset_9 .fit files.  I also use the Masters tab, to generate flats which have the suffix _tung at the end.  Just try processing while following Christian's guide to Alpy600 processing seen here:

http://www.astrosurf...lpy/tuto_en.htm

 

James Foster

Los Angeles, CA


Edited by James Foster, 25 January 2018 - 06:17 PM.

  • Organic Astrochemist likes this

#12 Organic Astrochemist

Organic Astrochemist

    Viking 1

  • -----
  • topic starter
  • Posts: 794
  • Joined: 10 Jan 2015

Posted 27 January 2018 - 10:48 AM

Thank you. That was helpful.

After climbing a mountainous learning curve of imaging and spectral acquisition, I still have much to learn about ISIS. I find that RSpec was easier to learn because one could see what was happening every step of the way. ISIS gives a better result, but it is still like magic to me. I am so unsophisticated that I use Nebulosity (rather than MaximDL) for my acquisitions, but that renaming tool in ISIS will be useful for me too. I know I just have to practice processing with ISIS.

Great spectra you posted of Mira, by the way. 


Edited by Organic Astrochemist, 27 January 2018 - 10:48 AM.


#13 James Foster

James Foster

    Lift Off

  • -----
  • Posts: 11
  • Joined: 10 Sep 2011

Posted 03 February 2018 - 02:22 AM

RE:"ISIS gives a better result, but it is still like magic to me."

Yes it is painful at first, but just like any sophisticated program like Adobe Photoshop, AutoCad, PixInsight, etc....the power released with these tools more than make-up for "learning-curve."

I used AutoCad way back in the late 1990s and when I needed it 2 years ago (2016) I had forgotten 90% of what I did to make it work on 3D objects.  Took a few weeks, but I was able to use it to design another I-beam tripod

as seen here (middle white tube telescope):

3TelescopeMTPinos.jpg

 

James


  • old_frankland and PrestonE like this

#14 old_frankland

old_frankland

    Mercury-Atlas

  • *****
  • Posts: 2945
  • Joined: 28 Mar 2005
  • Loc: San Francisco Bay Area

Posted 04 February 2018 - 08:41 PM

Talk about multi-tasking!!!



#15 Organic Astrochemist

Organic Astrochemist

    Viking 1

  • -----
  • topic starter
  • Posts: 794
  • Joined: 10 Jan 2015

Posted 05 February 2018 - 12:21 PM

Certainly something to look forward to in retirement...

Strange how out of the thousands of Americans on this website, three who post frequently on spectroscopy have the same name! shocked.gif Sockpuppet anyone?lol.gif

Personally, I blame credit Jim Ferreira for enticing me down this delightful rabbit hole of spectroscopy. Thanks to James Foster and especially Robin Leadbeater for enabling me all the way down.


  • old_frankland likes this

#16 James Foster

James Foster

    Lift Off

  • -----
  • Posts: 11
  • Joined: 10 Sep 2011

Posted 07 February 2018 - 05:03 PM

I'm in South California. I believe Jim Ferreira is in Livermore in Northern California.

 

There are not a lot of Americans doing spectroscopy with the Alpy600 or LhiresIII.    Most users and heavy contributors are Europeans

in France, England, Germany, etc. 

 

When I retire, I plan to get an Echelle system to capture amateur high (R= 10K) resolution from 3900A-8540A; professional "high" resolution spectroscopy starts at R=30K!

Christian Buil's recent upgrades to the Shelyak Echelle make this possible; see: spectro-aras.com/forum/viewtopic.php?f=8&t=1860&start=50

 

James


  • old_frankland likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics