[Aavso-photometry] V filter only - transforms possible?

[arne]

robertjmodic at att.net wrote:

>> My intent is to concentrate on observing techniques in 2009, in an effort
>> to get observers producing even higher scientific quality data.
> 
> One thing that might encourage more observers to report transformed
> observations would be to have the AAVSO provide a spreadsheet template 
> that would take the flux measurements from any photometry software and 
> show how to obtain differential transformed magnitudes ready for 
> reporting to the AAVSO.  Most software and spreadsheets I've seen only 
> deal with the case of transforming all-sky photometry.  But, most of the 
> CCD observations submitted to the AAVSO are differential.  So, a 
> spreadsheet template along with some worked examples might save 
> observers from having to 'roll their own' and help make sure that 
> everyone is doing this correctly.
> 
Yup, that is one feature that we are considering.  We already offer
transformation support for the photoelectric photometry observers,
and should be able to do something similar for CCD observers.  It is
just a matter of limited staff time and other priorities (like getting
funding!).

Michael Newberry wrote:
 > This only works approximately if you know or assume the color index of
 > the target object. So for followup observations of known objects, I can
 > see it. I have been considering the general photometry case where you
 > don't know the B-V of the target. In that case, you are talking about
 > systematic errors on the order of 0.05 to perhaps 0.1 magnitude in some
 > cases, independent of brightness.

I fully agree with Michael here regarding the general photometry
case with unknown colors.  However, most observations reported
to the AAVSO are of objects on our "program", where we actually know
the color index.  If you don't know the color index, or cannot make
a reasoned guess as to what it should be, then you are better off
not transforming your data.

It should be noted that, if you use two filters (such as B and V), and
have determined your transformation coefficients by using standard
stars or clusters, you can transform any differential data - you don't
need to know the standard color indices of either the target or the
comparison star.  We give a worked example in the Astronomical Photometry
book.

Michael Koppelman wrote:
 > The reason I'm surprised is -- it's complicated and easy to screw up
 > transforms. Also, if people sometimes do and sometimes don't do the
 > transform, we lose the homogeneity of their data set (although you
 > suggest a good method for undoing it). The transforms are also
 > unnecessary if you have color matched comp stars.

If people are submitting time series, and if their time series overlap,
and if the star does not change color during the total time series
window, then you can adjust zeropoints and get a nice light curve
without transformation.  If the target and comparison star have
similar color, and if the observation is at small airmass, and if
the instrumental system has a small transformation coefficient, then
you can plot individual datapoints and get a nice light curve without
transformation.  For the more general case, such as a Mira where the
comparison stars are of very different color and the observer might
submit only a handful of datapoints throughout the 300-day period,
approximate transformation is the only way to get a nice light curve.
We provide a check-box on data submission so that an observer can
transform one observation and not transform the next - it is up to
the observer to correctly indicate what they have done to their data.

My intent over the next year is to get everyone to determine their
transformation coefficients, and to either apply them or else let
HQ know what they are so that researchers can apply them.  We'll
give clear instructions, and probably provide the service where you
send us your images of specific fields and we'll tell you your
coefficients (much like I did with the M67 experiment).  PEP observers
do this - you can, too!

Arne