[Aavso-photometry] TC Question

[arne]

Keith Graham wrote:
> Hi,
> 
> 
> 
> I have a question concerning transformation coefficients. I just completed a
> 3 night run of M67. I took 3 images each night through BVRI filters. I
> averaged the 3 mag results for each filter for each night and then did TC
> calculations for each night to see how they might differ from one another.
> Two of the nights were pretty close, but the third night was quite different
> (sometimes off by as much as .4). The conditions on that “off” night
> included some high clouds. This made me think that we cannot possibly match
> the conditions under which our TCs were taken to the any or all other
> conditions under which data is collected. Since nightly conditions vary, how
> can we be certain that the TCs we use will actually standardize the data for
> any given night? Is there a method we need to be using to get the TCs as
> close as possible to make them useful for any given situation?
> 
My understanding from your description is that you can actually obtain
*9* coefficient determinations: 3 each on 3 nights.  That is what I
would do, rather than averaging each night's images before extracting
photometry.

In general, the slopes should be consistent even if the conditions are
not great.  The *zeropoints* will change, but they are unimportant unless
you are performing all-sky photometry, and if you are doing this, you
will determine new slopes each night anyway.  The slopes (the transformation
coefficients themselves) should be reproducible at the 10% level or so;
I find, for example, that the Tv value at Sonoita might average to 0.044
with our older filters, with a standard deviation of 0.008, using an
all-sky ensemble of standards and not just from a single cluster.  I'd
assume better results from a cluster as that removes the extinction unknown
from the analysis.

You might determine the slopes from the 9 data sets, and report the
values to the mail-list so that we can publicly analyze the results
for everyone's benefit.
Arne