Starship Asterisk*

Recently I posted a graph, showing Alpha Centauri on Sep 7th as compared to Sep 4th.



It was suggested that when a known constant star was shown to indeed be constant between two nights, those nights could likely be used to check other stars.

I chose to look at three other stars on those nights, not exactly randomly but almost.
Bet Aqr
Alp Gru
Eps Peg

These are the plots





These graphs show quite a bit of variation.
Of course it is possible that, through dumb luck, I chose 3 variable stars.
I think the explanation is the presence of the moon on Sep 4, on closer inspection I should not have used Sep 4th as a comparison night. I believe the counts are lower on Sep 4th do to an increased background level, and that 3 days later as the moon faded the background counts went down and the star counts back up. The question that still remains however is why does Alp Cen show almost completely constant on these two nights?
Alp Cen remained fairly low in the sky, and mostly away from the light concentrated near the center.

Also seen on the image is Alp PsA (Fomalhaut) which seems to be at least, if not more, free of the bright backgrounds as Alp Cen.

This is the graph of Alp PsA



Which does not seem to support the idea Alp Cen shows constant because it is distant from increased background levels.

Anyway, I don't think Sep 4th was a good night to use for comparison. I am going to look for a better combination of nights, Sep 16th and 17th from Haleakala seems promising. If anyone can explain what might be occuring in these graphs, let me know.

Vic,

I think the problem is that Alpha Cen maxes out the central pixels at 180 seconds of exposure. The odd thing is that a max out will appear as a C1 central pixel of anything above about 60,000 counts, not just 2^16 = 65,536 counts as is the theoretical maximum. When two frames are maxed out, their ratio is about 1, hence the constancy of your first plot.

Visual inspection of the thumbnails on South Africa (SA) here:
http://nightskylive.net/sa/sa040904.html
indicates some clouds and central blurring due to condensation, even in the beginning of the night.

I therefore do not think this was a good night for SA CONCAM photometry. The other non-Alpha-Cen stars tell the more correct story. The moral is to watch out for stars that are above 60,000 for C1. For Alpha Cen, one of the brightest stars in the sky, you might try photometry during 20-second moon exposures. The counts could be up to 180/20 = 9 times lower.

- RJN

That is exactly what I was seeing, thank you RJN.

Proceeding backwards and looking for nights with an adequete collection of 20 second exposures, I have come to Jan 12.
Curiously it appears that Alp Cen saturates on some 20 second exposures also.
For this image (20 second exposure) C1-B is listed as 64,461 for Alp Cen





Jan 12 shows 5 of 18 counts that are likely saturated.
http://nightskylive.net/sa/sa040112/HD252838.html

Jan 14th shows 7 of 21 likely saturated.
http://nightskylive.net/sa/sa040114/HD252838.html

Should these data points be discarded?
And am I wrong in thinking that if Alp Cen saturates on 20 second exposures it must be way too bright for us on 180 second exposure? Even taking the lowest believable count from Jan 12, nine times it would be over 150,000. Even if the 180 counts were only expected to be twice the 20 sec, it would saturate continually.

Hey Vic,

Well if there is only an occasional 20-sec exposure that saturates, the cause might be a Moon "ghost" that got internally reflected on top of the Alp Cen position. One has to be careful to ignore / delete the odd saturated pixel obsevation when that happens. The surrounding observations, particularly when the Moon is low in the sky, should be OK.

If all of the 20-sec exposures are saturated regardless of altitude -- which I do not think is the case for Alp Cen -- then getting a light curve for Alp Cen is pretty well impossible with that CONCAM. I think this may be the case with Sirius. Sirius can only be done with 20-sec exposures with CONCAM2s I think.

- RJN