Spectrum of Altair

[athornett]

Spectroscopy of Altair (and Vega) in Lichfield, UK 11/10/2018:
A new star tonight for me – Altair. I have not taken a spectrum of this star before. Being also an A type star (A7V), it was going to be similar to Vega (A0V). Therefore, I took a spectrum of Vega for comparison too..
Taken using CCDSPEC spectrometer with Sky Watcher Equinox Pro 800mm Telescope on alt-az undriven Manfrotto mount – hand-guided for 50 second exposures.
RSPEC software has some great reference spectra which I was able to use to calibrate these images this evening. Cross checked using my own image of Vega as reference spectrum.
Note Vega = A0V spectral type
Altair = A7V spectral type

More information at https://roslistonastronomy.uk/spectroscopy-of-altair-and-vega-in-lichfield-uk-11-10-2018
Andy

[Demonperformer]

You seem to be getting some great data, Andy. During my (brief) foray into the world of spectroscopy, I totally failed to get anything like this.

[robin_astro]

Looking good Andy. Was this wavelength calibrated using your RELCO lamp?  It looks like you have a few hot pixels/cosmic ray hits (The sharp "emission lines" 

You might be interested to know that the reference spectrum you used  in Rspec (and also Visual Spec, ISIS and BASS) comes from the Pickles library, downloadable from here for example

http://www.eso.org/sci/facilities/paranal/decommissioned/isaac/tools/lib.html

It is one of many libraries of spectra though the pickles spectra are generic of the spectral type rather than of a particular star so do not always perfectly match actual spectra

Vega is nice and easy to identify features as there  is not many strong lines other than the Balmer and telluric lines. Christian Buil has a nice annotated higher resolution (~3A)spectrum of Vega here which can help you identify some  of the weaker lines, including all the Telluric bands 

http://www.astrosurf.com/buil/us/vatlas/vatlas.htm

( Spectra of cooler stars like the G/K stars are much more complex so most lines cannot be uniquely identified at the resolution of your spectrograph as they are blended)

Next step is to correct your spectrum for the effect of the instrument and our atmosphere. You can use your Vega or Altair spectra for this and then use it for other stars provided they are at similar height in the sky. The continuum will then show the same overall shape as your reference spectrum. Doing this for a few known stars and comparing them with professionally measured spectra is a good test of your technique, as I have done here for example using spectra of stars from the MILES database

http://www.threehillsobservatory.co.uk/astro/spectroscopy_21.htm

 

Cheers

Robin

[athornett]

7 hours ago, Demonperformer said:

You seem to be getting some great data, Andy. During my (brief) foray into the world of spectroscopy, I totally failed to get anything like this.

Thanks DP - that is really kind of you to say that. 

Andy

[athornett]

27 minutes ago, robin_astro said:

Looking good Andy. Was this wavelength calibrated using your RELCO lamp?  It looks like you have a few hot pixels/cosmic ray hits (The sharp "emission lines" 

You might be interested to know that the reference spectrum you used  in Rspec (and also Visual Spec, ISIS and BASS) comes from the Pickles library, downloadable from here for example

http://www.eso.org/sci/facilities/paranal/decommissioned/isaac/tools/lib.html

It is one of many libraries of spectra though the pickles spectra are generic of the spectral type rather than of a particular star so do not always perfectly match actual spectra

Vega is nice and easy to identify features as there  is not many strong lines other than the Balmer and telluric lines. Christian Buil has a nice annotated higher resolution (~3A)spectrum of Vega here which can help you identify some  of the weaker lines, including all the Telluric bands 

http://www.astrosurf.com/buil/us/vatlas/vatlas.htm

( Spectra of cooler stars like the G/K stars are much more complex so most lines cannot be uniquely identified at the resolution of your spectrograph as they are blended)

Next step is to correct your spectrum for the effect of the instrument and our atmosphere. You can use your Vega or Altair spectra for this and then use it for other stars provided they are at similar height in the sky. The continuum will then show the same overall shape as your reference spectrum. Doing this for a few known stars and comparing them with professionally measured spectra is a good test of your technique, as I have done here for example using spectra of stars from the MILES database

http://www.threehillsobservatory.co.uk/astro/spectroscopy_21.htm

 

Cheers

Robin

Hi Robin,

Thanks again for brilliant advice - please keep it coming - I really appreciate having a mentor on the spectroscopy side!

So, now to try and correct my spectrum for instrument performance.......

Andy

[athornett]

29 minutes ago, robin_astro said:

Looking good Andy. Was this wavelength calibrated using your RELCO lamp?  It looks like you have a few hot pixels/cosmic ray hits (The sharp "emission lines" 

You might be interested to know that the reference spectrum you used  in Rspec (and also Visual Spec, ISIS and BASS) comes from the Pickles library, downloadable from here for example

http://www.eso.org/sci/facilities/paranal/decommissioned/isaac/tools/lib.html

It is one of many libraries of spectra though the pickles spectra are generic of the spectral type rather than of a particular star so do not always perfectly match actual spectra

Vega is nice and easy to identify features as there  is not many strong lines other than the Balmer and telluric lines. Christian Buil has a nice annotated higher resolution (~3A)spectrum of Vega here which can help you identify some  of the weaker lines, including all the Telluric bands 

http://www.astrosurf.com/buil/us/vatlas/vatlas.htm

( Spectra of cooler stars like the G/K stars are much more complex so most lines cannot be uniquely identified at the resolution of your spectrograph as they are blended)

Next step is to correct your spectrum for the effect of the instrument and our atmosphere. You can use your Vega or Altair spectra for this and then use it for other stars provided they are at similar height in the sky. The continuum will then show the same overall shape as your reference spectrum. Doing this for a few known stars and comparing them with professionally measured spectra is a good test of your technique, as I have done here for example using spectra of stars from the MILES database

http://www.threehillsobservatory.co.uk/astro/spectroscopy_21.htm

 

Cheers

Robin

Hi Robin,

Re your question regarding how I calibrated - I used my Vega spectrum (hence why I took the spectrum last night) and (because I knew main lines wavelengths on Vega spectra) calibrated CCDSPEC/QHY6 using that - was that OK to do?

Re: hot pixels - I do seem to have a lot of hot pixels on the QHY6 which become evident on exposures over 30 secs - the exposure time last night was 50 secs per spectrum. Should I be concerned?

Andy

[athornett]

My attempt today to produce an instrument response curve:

https://roslistonastronomy.uk/attempt-to-generate-instrument-response-curve-in-rspec-software-for-ccdspec-using-altair-spectrum-from-10-10-2018

Andy

 

 

[robin_astro]

4 hours ago, athornett said:

Hi Robin,

Re: hot pixels - I do seem to have a lot of hot pixels on the QHY6 which become evident on exposures over 30 secs - the exposure time last night was 50 secs per spectrum. Should I be concerned?

Andy

Hi Andy,

A dark subtraction will probably deal with them.  (The first step with processing spectrum images is the same as with astro imaging ie bias, dark subtraction, flat field correction)  You then make any geometric corrections needed eg slant and tilt, remove the sky background and bin the region containing the spectrum to produce the digital data.  You then wavelength calibrate and correct for the instrument response/atmospheric extinction to produce the final fully calibrated spectrum.  You can see examples of these steps in the presentation "Low Resolution Slit spectroscopy (ALPY) - Confirming and classifying a Supernova " I gave at a BAA workshop a couple of years back for example. 

Cheers

Robin 

 

[robin_astro]

1 hour ago, athornett said:

My attempt today to produce an instrument response curve:

 

Hi Andy,

You have sort of got the idea but there are a couple of things you need to do first.

1.  You need to subtract the sky  background by selecting areas above and below the spectrum and  subtracting them from the region where the spectrum is.  It is important how you select these regions. Make sure the region you select for the spectrum is wide enough to cover the full height of the spectrum, turn up the brightness in the image to make sure. Then select regions above and below this for background subtraction, far enough away that there is no contamination from the star spectrum.

2.  your camera/spectrograph is only sensitive between about 3700 and 8000A (After background subtraction your spectrum should read ~zero below 3600A and the spectrum above 8000A is not valid because it overlaps with the 2nd order spectrum)  So you first need to crop your spectrum from 3700-8000.   EDIT: - 3700-7400 in your case as this is the edge of your camera field. 

You can then divide this by the library spectrum, remove any features which did not exactly cancel and smooth the result to produce your instrument response.  As a check you can then apply it to your uncorrected spectrum. The result should look like the library version.

Again you can see the steps in my workshop tutorial

Cheers

Robin

[robin_astro]

4 hours ago, athornett said:

Re your question regarding how I calibrated - I used my Vega spectrum (hence why I took the spectrum last night) and (because I knew main lines wavelengths on Vega spectra) calibrated CCDSPEC/QHY6 using that - was that OK to do?

Yes that is ok but you could use your relco lamp to get a more accurate calibration.

Robin

[robin_astro]

5 hours ago, athornett said:

the exposure time last night was 50 secs per spectrum. Should I be concerned?

50 sec is an extremely long exposure on such bright stars at this resolution.  With my similar resolution ALPY setup and a 280mm aperture, a spectrum of Vega is saturated in about a second or less.  Was the star well focused and positioned on the slit? Have you checked that no part of the spectrum image was saturated ?

Robin