Re: Ensemble analysis: "new variables"

[Andrew Bennett <andrew.bennett@ns.sympatico.ca>]

On Wed, 30 Jan 2002 23:08:25 -0600, Tom Droege
<tdroege2@earthlink.net> wrote:

>>#1 at 69.9225 6.9974 showed a remarkably smooth
>>dip in a series of 8 consecutive images (2199902 -
>>2199915) in MV but not in MI.
>>...
>This is somewhat worrisome.  What was the variation?
+0.7 mags. The star was actually detected as being
double on a couple of images where it straddled the
bad pixels.

This one is a bit of a horror to spot automatically
as it barely shows on either the Darks or the Flats
but only on star images. Well - it's easy enough to
spot when you know what to look for. With a human eye
in the loop. I've no idea how to automate detection.

My "pipeline" has now grown a manually created file of 
bad pixels over and above the ones automatically flagged 
as "hot" or "noisy" when the Dark file is created.
Results Real Soon Now!

>
>>#2 at 69.2319 3.5466 shows a brightening by 0.8 mags
>>on one V image (2199939) with no corresponding
>>change in MI.
>>...
>Hmmm!  You can usually recognize a cosmic ray hit when you see one out in 
>the open.  But one mixed in with a star may be just tough to sort 
>out.  Hmmm!  At one per sq cm per min we should get an average of 15 per 
>image.  So the probability one of several thousand stars having a cosmic 
>ray hit mixed in is sort of in the 0.1 range.  So every few images will 
>have this problem.  This is lots of false novas.

Right. It certainly is not a single pixel cosmic ray; something
like 5 or 6 pixels are brighter than they should be, by several
hundred
to a thousand adu ...

>just don't get mad at the data, be amused by it.

I just mustn't pray too hard to detect variables -
my prayers are too likely to be answered by these
embarrassing fakes!

Andrew Bennett, Avondale Vineyard