[vsnet-chat 2169] Re: Supernovae estimates using USNO SA2.0

[aah@nofs.navy.mil]

Taichi Kato wrote:
>   Some softwares apparently don't explicitly use -2.5log(count) formula,
>but look to experimentarily determine the term k in -k log(count) from
>a matching with catalogs.  This is probably incorrect in the usual
>photometric sense, but I don't know whether this prescription can be more
>reliable particularly in amateur SN photometry.  This prescription has been
>apparently brought into reality by the availability of massive faint star
>photometry like USNO A and USNO SA catalogs.  I'd like to know your opinion
>on the introduction of the variable -k log(count) term.
  I assume, having never seen such software, that what the writers
are doing is the following.
  Say that you have a sequence of reference (photographic) magnitudes
as well as the instrumental counts/magnitudes from your CCD system.
You can fit your data like
   Mref = a + b * Mccd
        = a + b * (-2.5 log (counts))
        = a + k * log(counts)

where you might use a different k-value depending on the magnitude range
you are trying to fit.

  This is a perfectly adequate way of transforming raw data into a reference
system.  In fact, you use a similar technique when transforming your instrumental
CCD data into the Johnson-Cousins system with Landolt standards as your reference.
However, beware of using this technique with USNO-A or GSC magnitudes, since
you are taking a highly linear detector like a CCD and transforming it into
the nonlinear photographic regime.  That seems like the wrong direction!  Plus,
the photometric errors in the photographic catalogs far exceeds the errors
in the CCD frame, and so you are taking good data and trying to fit it using
poor reference data.  Use USNO-A or GSC to set a zero point for the field,
and then just use the adjusted CCD instrumental values.
Arne

VSNET Home Page

Return to Daisaku Nogami

---

vsnet-adm@kusastro.kyoto-u.ac.jp