Light curve deblending for a candidate Variable star

Science Data Analysis
1

Hi everyone.

First of all I introduce myself, since this is my first post. :grinning:

I am Angelo Gambino I live in Catania (Italy) and I am starting to study variable sources using my equipment by some months. I analyse variable sources together with some guys of the astronomy association I belong to here in Catania: the ā€œGruppo Astrofili Catanesiā€ (GAC) .
I have an astrophysics background, since I was a research fellow in Astrophysics until some years ago, with interests in Low Mass X-ray binary systems hosting Neutron Stars, that I studied both from a temporal and spectral point of view. The question I would like to bring to your attention is quite general and concerns the deblending technique. I searched through the topic archive but couldnā€™t find an answer. I hope this subject hasnā€™t already been addressed somewhere and I just wasnā€™t able to find it.
In the old version of the AAVSO forum I learned that there is a deblending routine that allows isolating the light curve of a candidate variable source, in order to validate its variability.
Could I ask if someone could point me to the procedure, manual, or tool to use for this analysis? Is there a manual or precise step-by-step instructions?

Thanks in advance.

Regards,

Angelo

2

Hi Angelo,

When two stars are blended, you can find two scenarios:

  • two variables stars
  • a variable star and a constant star

In the first case, you will need to subtract one signal from the other in order to study each objectā€™s variability separately.
It would be the same as studying a single star with more than one variability frequency (e.g. a double mode pulsator).

In the second case, you will only have to correct the magnitudes for light contamination, in order to obtain a correct amplitude and zero point. This implies that you know the magnitude of the contaminating source. Nowadays, we are able to obtain such value for most stars using catalogs like Gaia DR3.
You can also correct the magnitudes for the 1st case objects but it will be more difficult because both are variable and you will have to use a mean magnitude for the contaminating source.
When light contamination is strong a very small error in the magnitude of the companion may cause a large discrepancy in the corrected result so you have to be careful. Also care has to be taken not to mix different passbands when you correct magnitudes.

To study multiple variability frequencies from a single object, I use Period04, it is freeware and it is a powerful tool.
To correct magnitudes for light contamination from a companion, I use a spreadsheet that we have also uploaded to our server, and is linked from the VSX FAQ page.
It is here.

I hope this is useful for you.

Cheers,
Sebastian

2 Likes
3

Hi Sebastian!
Thanks for your kind availability and for your explanation. It seems quite clear. I have to study some cases following your suggestions and the tools you suggested.
I will send you a feedback on my first deblending soon.

Thank you again. :blush:

Angelo

1 Like
4

Sebastian is absolutely correct. In most cases those methods would be sufficient.

However, if you want to get into some more detailed methods you can use something like DAOPhot that does Point Spread Function fitting to each star on the frame. This provides a way to separate close stars and is particularly good in star clusters or close stars of similar brightness. There are multiple versions of the software with little support. But it is powerful. I think I have a student thesis that gives step-by-step instructions for the stand-alone version.

A second option is to use an image subtraction software. There is one out of France with a really bad name right now. Do a search on ā€˜astronomy image subtraction softwareā€™ and you will likely find one from www.iap.fr. Iā€™ve seen this used to work on RR Lyrae variables in globular clusters. You effectively subtract everything that is constant in the frame and then measure what is left after the subtraction. The results I saw were some of the best RR Lyr light curves in clusters Iā€™ve ever seen. But it again has a steep learning curve.

Both methods are best for working on fields of stars over individual stars, but you might find them useful.

Eric

2 Likes
5

Hi Eric!
Thank you very much for your suggestions. I really appreciate it.
Could I ask you where to find the thesis containing the step-by-step instructions for the stand-alone version of DAOPhot? I would be very interested to try also this method.
In addition, I tried to look around in the web to understand how to download this software. I am not able to fine a source code download address.

Thank you in advance! :slight_smile:

Angelo

6

I didnā€™t find the thesis from the original student, but I found a thesis from a later student who gave a DAOPhot manual. This one is the version inside IRAF, but it should work for the stand-alone version as well.

andythesis.pdf (1.4 MB)

Hope that at least gives you some ideas. It is Chapter 3 you will want to look at.

1 Like
7

Thanks a lot for your suggestion, @Sebastian_Otero and @doctor13 ! :slight_smile:
I will try to follow them!