[Aavso-photometry] survey photometry

[Arne Henden]

Michael,
   Your basic idea is good, but the practical implementation is difficult.
The problem is in comparing apples to oranges.  NSVS, for example,
is an unfiltered survey while TASS and ASAS use standard Johnson/Cousins
filters.  Some may be transformed; some may use local Tycho stars
(which had to be transformed onto the standard Johnson/Cousins system),
etc.  Usually when you combine multiple site observations, they are
all transformed using Landolt standards, and combining the datasets
will improve the basic calibration by sqrt(n) to first order.  The
same goes for data from a single site, obviously; Tycho for example
does this to beat down those 100-400 single observations to a mean
magnitude (you should look at the individual observations of a 10th
magnitude Tycho star some day!).
   Bottom line: I don't think you would improve anything by combining
TASS, ASAS and NSVS.
   Useful range for a catalog is pretty easily determined by just looking
at the Landolt stars that appear in that catalog.
Arne

Michael Koppelman wrote:
> A hidden gem about these surveys (TASS, ASAS, NSVS, etc) is their 
> measurements of *constant* stars. Once we know how to transform them 
> all, we have many independent measurements by which we could probably 
> determine fairly if not highly accurate photometry of a lot of stars. If 
> you average 3 measurements each with errors in the neighborhood of 0.1m 
> you get an uncertainty of 0.06 or so. All the surveys probably beat 0.1 
> so we could maybe get in the 0.03 range?
> 
> Has anyone ever done this?
> 
> 1. Create (or find) survey-to-Johnson/Cousins transforms for TASS, ASAS 
> and NSVS. Determine the uncertainty in the transforms.
> 2. Average the measurements for a single star using many, many 
> measurements from a given survey. Determine the uncertainty.
> 3. Determine the total error in the measurement based on the error in 
> the average and the error transformations for a given star in a given 
> survey.
> 4. Create a weighted average of all three survey measurements for a 
> given star.
> 
> If you did that for all stars which all surveys have observed, you could 
> then plot the uncertainty as a function of color and magnitude and get a 
> pretty good idea of the useful range of the catalog. Something like that 
> may be more useful than, say, Tycho.
> 
> Cheers,
> Michael Koppelman
> http://www.lolife.com/astronomy/
> 
> _______________________________________________
> Aavso-photometry mailing list
> Aavso-photometry at mira.aavso.org
> http://www.aavso.org/mailman/listinfo/aavso-photometry
> 
>