Forum der Vereinigung der Sternfreunde

Hi all,

Here in Montreal, we've begun the work of combining and analyzing the WR campaign data from this summer and the delta Ori data from last winter. The data are great, and we appreciate the efforts and time of our valuable colleagues.

We have another project we are hoping for some assistance with another project. We analyzed data from the MOST satellite and found what we believe to be spots on the surface of the O giant. These results are somewhat un-expected, so we are repeating the experiment with MOST this October-November (starting in about 2 weeks) in order to verify this. Amateur spectroscopy can aid our understanding by helping us look for minor changes in the wind that are thought to be tied to surface features and looking for pulsations. This does mean we need high S/N (> 150 per pixel) with high spectral resolution (R~10,000 or more) of several spectral lines. We've detailed the needs in the accompanying document. Thanks in advance for your help!

This analysis will be led by Montreal grad student Tahina Ramiaramanantsoa along with Tony Moffat.

Cheers,

Noel

Hi Noel,

thank you for inviting me!

As you know, I have back my eShel now and he oprates again in my small observatory.

I would be rather glad to give you some assistance for your project. The star is rather bright and it is rather good reachable. for me. So I think I can reach the desired R and the desired S/N. In the next clear night I will try to do so.

cheers

berthold
From: Noel Richardson (fg-spek-convento@vds-astro.de)
Sent: Monday, October 07, 2013 5:44 PM
To: fg-spek-convento@vds-astro.de (fg-spek-convento@vds-astro.de)
Subject: [fg-spek] Spots on xi Persei



Hi all,

Here in Montreal, we've begun the work of combining and analyzing the WR campaign data from this summer and the delta Ori data from last winter. The data are great, and we appreciate the efforts and time of our valuable colleagues.

We have another project we are hoping for some assistance with another project. We analyzed data from the MOST satellite and found what we believe to be spots on the surface of the O giant. These results are somewhat un-expected, so we are repeating the experiment with MOST this October-November (starting in about 2 weeks) in order to verify this. Amateur spectroscopy can aid our understanding by helping us look for minor changes in the wind that are thought to be tied to surface features and looking for pulsations. This does mean we need high S/N (> 150 per pixel) with high spectral resolution (R~10,000 or more) of several spectral lines. We've detailed the needs in the accompanying document. Thanks in advance for your help!

This analysis will be led by Montreal grad student Tahina Ramiaramanantsoa along with Tony Moffat.

Cheers,

Noel

Hi Noel,

is there a limit to the exposure time?

Cheers,
Siegfried

This sounds exciting Noel!
It will be very interesting to find out what those spots are.

Looking forward to seeing the documentation on the observation requirements.

I just hope we can get some clear weather here in the UK so that we can contribute...

best regards

John

Apparently, the document didn't attach before because it is too large. Here it is, detailing the star and the observing requirements:

http://www.astro.umontreal.ca/~richardson/xi-Persei.pdf

As in previous campaigns, we would like a dense time-series of the star at high resolution and high signal-to-noise.

Cheers,

Noel

Hi,

This night, I see the moon , not many stars, ......) it' s slightly cloudy here! Nevertheless I tried getting some spectra of xi persei. I hope I took the correct star. The echellespectra have nearly no lines , so it could be the right star. As soon I have reduced the spectra I 'll show them here. I am afraid, that the S/N is not good enough, but we will see.....

cheers berthold

Had the good fortune of getting some clear skies ( if a little misty ) on the 16th. Working with Jack Martin using the LHIRES on a C11 we managed to get seven twenty minute exposures of the star before the fog came in. Anyone got any ideas of how to calculate the SNR from our spectra - there aren't that many flat points on the continuum. Using 6589-6594 range in vspec I get SNRs ranging at the start when the star was lowest at 135 then it rises up to over 200.

The text against the first attachment got truncated in my previous post

So here is the text along with the picture again

For fun I did this analysis of the time series of the spectra. In the top half is the average spectra below it is the results of each spectra divided by the average - which have been changed into "synthetic" spectra using vspec and then I applied the gaussian blur to remove the sharp lines. You can clearly see some patterns in the later spectra (time runs from bottom to top). Whether these patterns represent something real or not I am not sure I expect we would need a lot more data... which is difficult to get at the moment as it is cloudy. If anyone has got any code out there to do this sort of analysis properly I would love to get access to it - from reading the papers fullerton's TVS or the more recent smTVS (Sudnik) would be worth getting if anyone has a copy of it?

Hi John,

I expect Noel has some nice tools for this but ISIS has a tool to generate these sorts of 2D plots and animations. I used it on the eps Aur data for example
http://www.threehillsobservatory.co.uk/ ... iginal.pdf
You would have to divide by the mean from each spectrum but you could possibly use a script in ISIS to do that. I would probably also try to remove the tellurics first before generating the plot as they are likely to vary with altitude and sky conditions during the run. If you have not measured a telluric standard, the tool in VSpec (or ISIS) works pretty well, allowing you to tune a generic telluric profile to match the resolution and intensity in the spectrum. It is a very strict test of wavelength calibration of course too.

I find the number of electrons counted/bin is a reasonable indicator of the SNR you are likely to end up with.

ie SNR ~sqrt(total ADU * camera gain e-/ADU)

so for example if your camera gain is 0.3, you need ~75k counts/bin to give 150 SNR (This should underestimate somewhat due to additional noise from the camera and sky background but it seems to work out reasonably close)

Cheers
Robin

Hi,

now there is my echellespektrum to find in the adnex. I know, it is not so good, but as already said, it was rather cloudy in this night. I hope I can do it betterin future.

You can get there the hole spectrum and a *.png where I have extracted the H-alpha and helium region. All this lines are rather large. Is the rotation velocity of this star so fast?


cheers
berthold

Thanks for the tip on calculating the SNR Robin - I have over 100,000 counts per bin even in the lowest quality spectra so the SNR should be ok.

I'll have a look for that functionality in ISIS

best regards

John

Hi John,

Here is an example of the use of ISIS animation feature
http://www.astrosurf.com/buil/isis/anim ... ion_en.htm
(as well as producing the animation it also generates a 2D image)
This example also shows how to calculate the differences from the mean by the looks too

Cheers
Robin

----- Original Message ----- From: John Strachan (fg-spek-convento@vds-astro.de)
To: fg-spek-convento@vds-astro.de (fg-spek-convento@vds-astro.de)
Sent: Saturday, October 19, 2013 8:18 PM
Subject: [fg-spek] Re: Spots on xi Persei



Thanks for the tip on calculating the SNR Robin - I have over 100,000 counts per bin even in the lowest quality spectra so the SNR should be ok.

I'll have a look for that functionality in ISIS

best regards

John

Thanks for the link Robin.

Its clear in the south east at the moment - captured the telluric standard 22 per and am now capturing menkib

Hi,

there is a new spec of xi_persei from 131030. I am at time not able, to eliminate the waves along the continuum.

cheers
berthold

Hi,

now I reduced the spectrum of xi persei131030 once more. But first I defined the instrumentell reponse in each order using zeta aql as telluric standard and then I used these results to reduce the echellespektrum.

cheers
berthold

Hello Berthold,
does the wave shape depend on the distribution of the spectrum over different orders?
Cheers
Christian

Here are two spectra of Menkib from 19th Oct (red line ) and tonight 4th Nov (blue line). There appear to be clear differences at the edge of the wings of the Halpha line especially the red side. Anyone else see this?

John

Hello John,

excuse me, but I'm not able to see this changes. The red spectrum has stronger H2O-lines.
I think the spectra have to bee dried. The water lines disturb the judging of profile variations.

_________________
Herzliche Grüße / best regards

Lothar

https://lotharschanne.wordpress.com/

Hi,

Lothar says what I think also. The spectra in this region should be dried. Otherwise it is not possible to see any changes

Cheers
Berthold.

Ok Lothar and Berthold I have started removing the telluric lines from my spectra. Here are the first two I have done from 16th Oct (blue line) and 19th Oct (green line). Can you spot any differences?

If you have any spectra of the Halpha line for Menkib I would interested in seeing them

cheers

John

Ok John,

you asked for spectra from me! All my echellespectra of this object are available on this page, you can dowmload them and then select the area of H-alpha.

cheers
berthold

Hello John,

yes, now I can see the profile variation. Thank you.

_________________
Herzliche Grüße / best regards

Lothar

https://lotharschanne.wordpress.com/

Hi John,

there's my last profile for comparison.
R = 12600 in 6678
S/N = 180 in 6570-6610

cheers
Siegfried

Thanks very much Hold for sending me your spectra.
I compare your one with one I took approx 4 hours after your one. I normalised your spectra to the same part of the continuum as I had been using.

You can see the comparison - your spectra is in blue and mine is in pink. They look pretty close to me.

Have you taken enough spectra to see the 2.1 day period in the line yet?

cheers

John

Even with data from a small number of spectra am picking up the periodic behaviour in the blue side of the Halpha line. See chart for lomb-scargle analysis of Equivalent Widths - giving a 2.1day peak, and a chart for EW versus phase assuming period 2.1 days

Hi all! Please note this important message from Tony Moffat:

Thomas,
Had a clear night last night in the uk - so Jack and I got a full nights worth of spectra. Its going to be difficult tonight though with the weather turning cloudy...

John

Great, John! Please proceed. This is important, as Tony highlighted it.

Cheers, Thomas