[Aavso-photometry] Differential transformation

[Radu Corlan]

On Wed, 29 Dec 2004, Arne Henden wrote:

> Radu Corlan wrote:
> >>   Second, please don't use my normal field calibrations for
> >>determining your coefficients.  Please use Landolt (and Cousins)
> >>standards only.  While I am a careful observer, I will not guarantee
> >>that my field calibrations are exactly on the Landolt system, and
> >>if you use my calibrations, you basically remove yourself one more
> >>step from being on the true standard system.  My field calibrations
> >>are primarily for determining the magnitude and color zeropoints
> >>for differential photometry.  I know I am pretty darn close, and
> >>for rough values you will be fine.  For precise measures, a reviewer
> >>will question the use of Henden photometry in this manner as systematics
> >>have not been investigated.
> > 
> > I see this as a "lesser of two evils" situation. If i have to work in 
> > non-photometric conditions, either using transformation coefficients 
> > obtained on a different night or getting them at a significatly different 
> > airmass (or widely spaced in time) is likely to introduce some systematic 
> > errors. Are these less than the systematics of one of your sequences? 
> > Maybe, maybe not. But, although i cannot prove it with data, i have a 
> > strong hunch that the averaged sytematics of several sequences are pretty 
> > small. 
> > 
> The problem here is that you, as an end user, do not know where
> any systematics in my photometry might lie.  As a simple example,
> some of my field calibration files have many nights of data; others
> might only have one or two nights, and those nights might have been
> marginal to start with.  That is why it is better to use published
> photometry that was meant to be used for standard measures, and which
> have been tested for the normal systematic variations such as color,
> declination, RA, time_of_year, etc.  There are plenty of Landolt
> standards, with several dozen fields that have multiple standards
> that can be imaged simultaneously.  No need to use Henden photometry
> for this, except as a place-holder to be replaced as soon as you
> get the chance to observe true standards.

I understand. It's certainly easier to "trace" data when reffered to 
primary standards.

Just for fun, i have data from two different nights on 18 fields with your 
sequences. I reduced transformation coefficients from them using either 
all the fields together, or in groups of 3-4. The results are (for 
Kv(B-V))

Night I (mean AM=1.15)

18 fields		0.062/0.002

4 fields(group1)	0.058/0.004
4 fields(group2)	0.060/0.006
4 fields(group3)	0.059/0.006
3 fields(group4)	0.069/0.008
3 fields(group5)	0.067/0.004


Night II (a subset of 14 fields from the above, mean AM=1.24)

all fields		0.061/0.002

4 fields (group1)	0.056/0.003
4 fields (group2)	0.062/0.005
3 fields (group3)	0.064/0.011
3 fields (group4)	0.069/0.004

What i can see here is that the various field groups yield pretty 
consistent values. I have only one night of m67 data (at AM=1.6), and it 
wasn't very good. But FWIW, it gave a coefficient of 0.048/0.004

Now this means a difference of 3 sigmas, which i would say is borderline 
significant -- if i trusted my m67 fields. in any case, the difference in 
extinction coefficient is about 0.014 (is that significant?)


>    Transformation coefficients remain constant over long periods of
> time; I am not quite sure what you mean about not using coefficients
> obtained on different nights since this is a very common practice.
> In fact, the best coefficient determination method is to use the
> handful of really good, photometric nights for your coefficient
> determination and then average these determinations to give you
> mean coefficients.  These can then be applied to future nights for
> long periods of time.

I always wondered how much would the extinction coefficients change with 
airmass, or between a good photometric night and a run-of-the-mill average 
one. There must be a small variation, but i have no feeling for the order 
of magnitude. 

Radu


> Arne
> 
> _______________________________________________
> Aavso-photometry mailing list
> Aavso-photometry at mira.aavso.org
> http://www.aavso.org/mailman/listinfo/aavso-photometry
> 

-- 

-------------
Radu Corlan       Snail Mail: Bucuresti sect. 1, 
rcorlan at pcnet.ro  str. Argentina nr. 28, Romania

   You can still escape the "Gates" of Hell!   
                 Use Linux!