spectro-aras.com

Hello all,

Some days ago, I made a (good) observation of Be star, from home, with Alpy 600 & C8 telescope. I'm in very poor conditions (urban lights, installed on my balcony...), but it worked fine (spectra are now in BeSS :>).

During this observation I did a very simple skill : take several spectra from reference stars, and build several times the instrumental response curve - normally I should get exactly the same result. ... I write this message, because it is not really the case, and I need your help to go further.

Here is exactly what I did.
- I selected two stars as reference, known to have no reddening issue (taken from LISA Pack documentation) : zet Ari and HD51104.
- For zet Ari, I took 2 spectra, with 1 hour difference. For the first one, the star was at 38° above the horizon, the second was at 49°. HD51104 was observed at 52°.
- In all cases, I observed roughly in the same direction - I do not think that there is any light pollution effect change to be taken into account.
- Because of the building heating and the poluted sky, I had an enormous seeing (more than 10 pixels FWHM), and I've an old mount which tracks hardly (I worked in autoguidig, however). All this is not good for the exposure time (I lose for sure a lot of efficiency), but it has the advantage to prevent any chromatic effect that I could have if I were is very good conditions.

To build the response curve of zet Ari (A1V, mag 4,8), I took a spectrum of HD000319 (Miles Library in ISIS database - there is no A1V in Pickles database).
For HD51104 (B8Vn, mag 5,9), I used the Pickles B8V spectrum.

For each spectrum (zetAri1, zetAri2, HD51104), I made the response curve using ISIS tools (response button, in the profile tab, then continuum calculation, by removing the balmer lines and smoothing the profle).
For zetAri2, I did the calculation two times at different moments, to see how consistent is my own contribution (buliding a response curve has some subjective effect). Here is the result : ... I consider this is really consistent !

Now, here is the comparison of both zet Ari spectra (1 and 2) : Again, this is consistent, but we see a gap, mainly in blue part. I suspect this is because of the height of the star in the sky (atmospheric effect). This is why we must take reference stars as close as possible of the observed star (at least close in elevation).

Now, here is my real problem. The picture below shows a comparison of the zet Ari (2nd observation) and HD51104. Spectra were not taken at the same moment - about 6 hours difference between observations - but stars were at the same elevation (49° vs 52°). This time, I cannot say anymore that data are consistent.
And I have, of course, the question ; which one is the right one ?
I remind there are several differences between these observations :
- Not the same star (zet Ari - A1V, mag 4,8 and HD51104 - B8Vn, mag 5,9), but both are considered as reference stars (stable, no reddening...)
- Not the same catalog for theoritical spectrum (Miles a Pickles)
- Not the same time (19h30 vs 1h50), but same date and conditions (telescope, observer, direction of observation...)

My feeling is that main difference come from the catalog used for theoritical spectra - I know that the hardest problem is to find THE reference, to calibrate our data.

Note that for all my Be observations, I used zetAri1 observation to correct the response curve.

Friendly,

François

Any idea how I could go further ?

Hi François,

First, for generate a A1V reference from Pickles, compute the mean of A0V and A2V Pickles profiles.
The result is a very good approximation of an A1V star.

Concerning response curves, the situation sound like a large variation
of atmospheric transsmission during the run, or more probably, an
atmospheric chromatism effet. Ok the elevation is nearly the same for the two dates,
but that about the hour angle ? Could you precise if the star are observed near
local meridian.

If possible, publish one of the 2D spectrum for evaluate also instrumentale
chromatism in the blue part).

Christian

Thanks for your reply, Christian.

Ok for Pickles AOV or A2V - I agree : profiles are very close.

Regarding the position in the sky, for zet Ari, the slit was about 45° compared to horizon, and the observed direction was close to 120° (180° = south). For HD51104, it was about 20°, and observed direction was 160° (rough estimation). Note that in my site, I see Grenoble in the direction of the south - there is then more pollution at 160° than at 120°.
Atmospheric dispersion change between both observations ? I would say no, because of the seeing and the bad tracking of my mount, which make mix permanently this kindof defect - Am I wrong ?

Here is a 2D image of zet Ari - I don't think there is a too big instrumental effect (no fish tail). I use a focal reducer (F/6.3). Friendly,

François

Hello Francois,

as I know from my own experience, instrument response is a tricky business with a mount which is not tracking well. Looking at the difference between the response from zet Ari and HD51104 I notice that the difference is correlated to the signal intensity. This may hint at a problem with background subtraction or an offset in signal level (before dividing by the flat), but of course it may have a completely different cause. I noticed this with echelle spectra, but with an old Audela version, which has been corrected a while ago.
During my latest observation of NovaDel2013 low over the horizon at the big telescope at Falera I carefully aligned the slit vertically and got good results, but this should not be a problem at 50° altitude.
I noticed another problem with my LISA. When I adjust the focus before a session at room temperature (20°C) and then mount the LISA to the telescope at around 0° I notice a rather large shift in focus (approx. 0.2 mm). I put a mark on the focusing ring so that I can focus according to the expected temperature. Is this behaviour normal or is there something wrong with my mounting of the CCD to the LISA? Surprisingly there does not seem to be a wavelength shift with the temperature change, only a change in resolution.

Regards, Martin

Hi Francois,

Even though there is no obvious chromatism, I think I would still suspect a drift/change in focus of the telescope. I have seen significant effects with my C11 and a focal reducer.
http://www.spectro-aras.com/forum/viewt ... ?f=8&t=662
Does the thickness of the spectrum look the same for both spectra ?

Cheers
Robin

Here are the spectrum images showing the change in focus which produced the change in instrument response.

Robin