Variable star V467 Dra two nights wasted

[Dave Smith]

I was in two minds as to whether I should post this or not. I have decided that it is worth showing that everything does not always go smoothly with variable star photometry. This particular star has a period of nearly 5 hours and so to get a decent range requires most of the night.  Recently the sky has been clear all night and day so a good opportunity to get results on this long period HADS star. Here are the results for the first night, Sunday night 22/23 March 2020.

It started well but then YUCK. The comparison star plots also showed increased noise. There was a little mist that night and so I put it down to that.

The next night was also clear so I repeated measurements on the same star.

Equally YUCKY !!! That night there was no cloud (that I could see) and again no jet stream. Wondering if it was a software problem I did the analysis again using MaximDL but it gave the same result.

By the way, in both cases the exposures were for 60 seconds binned 2x2.

Last night was also clear but I couldn't face doing it again so I had a play with the guiding of my scope. So far with variable stars I have not used guiding with my EQ8 mount but there is a very slow drift which I wanted to eliminate, although not urgent for my variable star work.

Ho Hum I do need a good night of readings. I have not submitted the above results to the BAA or AAVSO.

Dave 

[andrew s]

@Dave Smith  happens to all of us. I get exactly the same issue when monitoring flare stars all night. If you did not see the variation in the comparison stars I could easily conclude I had lots of small flares!

Regards Andrew 

[Whirlwind]

This looks like too much high cloud that wasn't easily observable.  It's difficult to use the comparison stars as a metric in this case because we don't what was being used and its magnitude.  A much brighter star will show less variability because the flux is so much higher and it still gives a very strong signal to noise despite the cloud.  If you have a background of 10 and a flux of 1,000,000 losing 10% flux because of cloud isn't that significant from a S/N perspective.  If you have a flux of 100 and lose 10% of your signal variably it is much more of an issue from an S/N perspective.  I would suspect if you use increasingly fainter comparisonthen you will see the scatter increase similarly to the target star.   Indeed clouds can be useful for photometry for very bright stars as they can scale back the flux and reduce the risk of saturation (though generally defocussing is more consistent).

[coatesg]

Yeah - high cloud (or worse, passing cloud that affects the comparisons and the target unequally) will definitely give weird results like this. It's usually right at the end of the period as well! 

[Dave Smith]

20 hours ago, Whirlwind said:

This looks like too much high cloud that wasn't easily observable.  It's difficult to use the comparison stars as a metric in this case because we don't what was being used and its magnitude.  A much brighter star will show less variability because the flux is so much higher and it still gives a very strong signal to noise despite the cloud.  If you have a background of 10 and a flux of 1,000,000 losing 10% flux because of cloud isn't that significant from a S/N perspective.  If you have a flux of 100 and lose 10% of your signal variably it is much more of an issue from an S/N perspective.  I would suspect if you use increasingly fainter comparisonthen you will see the scatter increase similarly to the target star.   Indeed clouds can be useful for photometry for very bright stars as they can scale back the flux and reduce the risk of saturation (though generally defocussing is more consistent).

Thank you very much for your reply. My initial thought was that your possible explanation was not the reason because I had tested the signal strength at the beginning of measurements but then I examined the output file and found that a very large proportion of my readings were saturated but not at the beginning. It looks as if the sky became more transparent rather than high cloud appearing. It is a lesson learnt although rather severe to have two nights wasted due to the same error.  I am  rather relieved that the cause has been revealed. Without your post I would not have thought to check that out, so thanks again.

Dave

[Dave Smith]

7 hours ago, coatesg said:

Yeah - high cloud (or worse, passing cloud that affects the comparisons and the target unequally) will definitely give weird results like this. It's usually right at the end of the period as well! 

Thanks Graeme. See my response to Whirlwind. At least some good news, Patrick has been able to extract a period from the results which fits in with the previous pattern.

Dave

[Whirlwind]

55 minutes ago, Dave Smith said:

Thank you very much for your reply. My initial thought was that your possible explanation was not the reason because I had tested the signal strength at the beginning of measurements but then I examined the output file and found that a very large proportion of my readings were saturated but not at the beginning. It looks as if the sky became more transparent rather than high cloud appearing. It is a lesson learnt although rather severe to have two nights wasted due to the same error.  I am  rather relieved that the cause has been revealed. Without your post I would not have thought to check that out, so thanks again.

Dave

Cool, glad it helped solve the problem.  It could be an improvement in seeing as well.  However, you shouldn't worry too much...this impacts professional observations as well with comparison stars.  You start with a 10 second exposure and you are far below saturation (usually 40-50% saturation is targeted) at 2.5" seeing and if it improves to <1" then the signal in the comparison explodes and saturates, commencing a lot of cursing and rapid change of the exposure time!

[coatesg]

Ah that's good - and the wrong way round compared to usual!

I tend to shoot for 10-20k ADU which gives good headroom and space to drop while still having decent snr...of course it can always go to pot!