What to do with Data Set 19

[Tom Droege <tdroege@veriomail.com>]

Most of you that asked for it will soon have Data Set 19.  What to do with it?

Here is my suggestion.

1) DS191 contains material to make darks and flats.  Make darks and flats.

2) DS192 contains 50 exposures of the same V field.  DS193 contains 50 
exposures of the same I field.

I took DS193 and Co Added two and four frames.  This gave the expected 
result, the noise decreased by nearly sqrt 2 and 2.  Since 50 frames is 
more than is needed to make a plot, I would Co Add the frames in groups of 
four.  This will give 12 measurements on each star.

When I did this with the I data, I found 6700 stars on a single frame, 8200 
with a Co Add of two, and 10,000 with a Co Add of four.

3) Co Add the frames in groups of 4.  This should give about 10,000 stars 
to a sigma of .3 mag by my rough test.

Now read Bohden Paczynski's paper "The Future of Massive Variability 
Searches" which you can find on the tass home page.  This paper suggests 
that there are lots of Contact Binaries to be found.  These are called W. 
Ursae Majoris stars.  This paper suggests that there is a contact binary 
per 1000 or 10,000 stars depending on how you read it.  So there should be 
1 to 10 W. Ursae Majoris stars in our sample.  Why not find them?  These 
have a period of 0.2 to 1.0 days and a typical amplitude of 0.6.  By the 
time one folds in all the factors of where we are looking on the curve, the 
magnitude, and the error in measurement, etc., we have a fighting chance to 
find one.  Our sample covers nearly 0.1 day so there is a good chance to 
see variation at the 0.2 to 1.0 day level.

4) Convert the frames to star lists.  Match them to a catalog, perform 
color corrections, etc..  One should get nearly 10,000 12 item star lists.

5)  Get the sigmas of the 10,000 lists.  Look for the items with the 
largest sigma.

6)  Plot the interesting items.

There is a good chance to find an new variable star.  Reading BP's paper, 
you will find other things to try to find.

I already have 20 similar data sets.  The plan is to take data in a form 
that provides both long and short period measurements.  For example, the 
same data that is used to form the flat field contains once a night 
measurements of a larger area of sky.  I already have data spanning a 
month.  So there is data for 10 or so nights over a month.  This should be 
good for finding Pulsating Stars, Miras, Cepheids, and RR Lyrae stars.

I have hopes that one of you will perform this analysis in a way that I can 
copy and use to process the data I am taking.  I am still open to produce a 
suitable bribe.  I notice that one needs a fast computer when you start 
Co-Adding frames.

I can still make more Data Set 19s as it is still loaded on my 
computer.  Just ask for it with an address.

Happy data munching.

Tom Droege