[Aavso-photometry] transformations

[arne]

Wm. Mack Julian II wrote:
> Since its monsoon time here in New Mexico I've been reading stuff on BVRcIc and how to get my measurments on the standard system. I have BVR filters and just ordered a I based on a message from Arne saying its easier to compute v-i than v-r since there is a greater color difference.
> 
> Message to group from Arne from Feb 16 2006
> 
>> (B-V) and (V-Ic) are the two color indices that most professionals
>> use.
>> They have about the same size, and give the most range as a function
>> of star temperature.  (V-Rc) is has a much smaller size (about half
>> of
>> the other two indices), so transformations are less accurate.
>> V and Ic have completely separate passbands, so transforming to
>> other systems is cleaner.  Ic is less affected by dust and
>> atmospheric
>> extinction than is Rc.  Ic is easier to measure for red stars.
> 
> Bah, hit send before I got to my question. I think I understand that to find transforms you plot:
> 
> B-V vs b-v and take 1/m to get Tbv
> V-R vs v-r and take 1/m to get Tvr
> R-I vs r-i and take 1/m to get Tri
> V-R vs V-v and m is Tv
> R-I vs R-r and m is Tr
> 
> is there a Ti? Would you normally report using a standard mag say V and V-I as color inidice something like B and B-V?
> 
There is a Ti, just as there is a Tx for any bandpass x.  Normally, I'd
use (V-Ic) vs. (Ic-ic) and then the slope is Ti.

Now, "normal" report.  Professionals typically always measure V, and
then report all other bandpasses in color index form: (B-V), (V-Rc), etc.
This is because the V magnitude sets the base brightness of the star, and
the color indices tell you something about the physics of the star: its
basic spectral type, any reddening, unusual brightness in some filter, like
Rc, indicating a bright emission line, etc.  Any other filter could be
used for the base magnitude, but professionals have standardized on V
for the Johnson/Cousins system.

The AAVSO has had the tradition of reporting V, and as people started using
other filters, the database was just extended to include B, Rc, Ic.  Reporting
each filter's measurement separately has the advantage of properly
timestamping each observation.  If you use Ti to transform to Ic, for
another example, you get a better error estimate than if you separately
transform V and (V-Ic) and then form Ic from Ic = V - (V-Ic).

On the other hand, having the magnitudes from a BVRI dataset reported
separately makes it difficult to determine upon download which measures
are part of the dataset.  That is why we added the "Group" field in the
Extended Format - with this field, it is possible to recreate the
appropriate color indices.  Reporting each filter individually also
tends to promote untransformed submissions, and I'm going to be
changing that approach and pushing transformation shortly.

Professionals and amateurs alike tend to be pretty sloppy with respect
to photometry.  Unfiltered observations are assumed to be V-like or R-like
and reported as if they were filtered measures; use of USNO-A photometry
and colors is often made, and even without reporting which USNO-A star(s)
was used as the comparison; extinction is ignored; transformation is ignored;
I've even seen professional reports that were not flatfielded.  It is not
difficult to get high-quality photometry; you just have to be attentive
to details.
Arne