[Aavso-photometry] Cluster Photometry

[Arne Henden]

On 5/24/07, Arne Henden <arne at aavso.org> wrote:
> On 5/24/07, Eric Broens <Eric.Broens at skynet.be> wrote:
> > Hi Arne,
> >
> > Isn't NGC6811 a candidate?
> > I don't know if you could observe it on a sufficient number of
> > photometric nights from SRO during the campaign.
> > If so wouldn't these observations be suitable for determining the
> > transformation coefficients?
> >
> There are lots of potential candidates for "standard" fields.
> I've observed many fields on hundreds of nights, for example.
> Such fields are good for rought calculation of transformation
> coefficients, but the problem is how well do they actually
> match the Landolt standard system.  There are a number of
> tests that need to be made before you can trust the absolute
> quality of the data.  I haven't done all of the tests to satisfy me
> for more than a few such fields.
> Arne
>
I ought to explain this comment in more detail.  When dealing
with transformations, there are 3 things to remember:

1.  It is actually the Landolt System, not Johnson/Cousins.
While Arlo tries extremely hard to match the results from
Johnson and Cousins, it is not a perfect match (including
systematics in their original papers).  However, *everyone*
then uses Landolt's measures for determining their own
transformation coefficients.  Hence, we are trying to
emulate Landolt.

2.  Landolt never observed any clusters - he does single stars
scattered all over the sky, using a photometer with a fixed
aperture size.  So all cluster work is coming from someone
else, such as Henden, who is trying to emulate Landolt.
In general, we are not trying to do true standards work, so
are not using nonlinear fits to the transformation coefficients,
not doing extinction with as much care, or were using
less-than-perfect nights for our work.

3. CCD measures of wide fields are great for getting magnitudes
and colors for large numbers of stars, but such observations
have their own problems.  The dynamic range of the chip means
bright stars are saturated and faint ones underexposed.  Imperfect
flatfielding means you might have systematic errors across a field.
Blending of stars means an object might appear in one catalog
with one magnitude, but another catalog with a different magnitude
depending on resolution and aperture size.  You should also note
that red stars in clusters are usually intrinsically red, whereas the
red stars from Landolt are usually reddened objects that don't
have the molecular features.   Using them for transformations will
introduce small systematic effects.

So in general, you have to hand-inspect your data if you want it
to be truly standards-quality.  You want to compare your results
with others who have observed the same field.  These things
take time, and so I don't do it often.   Using "secondary standard"
clusters will work for most transformation determination needs,
but you should always be aware that such determinations usually
are systematically different than the Landolt system.  They may
be the only way you can get coefficients with your sky conditions,
and so don't avoid clusters, just consider using the primary
Landolt calibrators if conditions are favorable.
Arne