[vsnet-chat 5413] Re: Some very red standard stars

[aah@nofs.navy.mil]

All transformation coefficient determinations just extend
as far as the reddest standard star you measure; you should
always be wary of extrapolation.  So observing and calibrating
stars of, say, (B-V)<2.5 is pretty straightforward.
  Where you lose quickly is observing stars outside of the
Landolt/Cousins range.  As Brian stated, transformations
become nonlinear, and even splitting between two (B-V) ranges
is not the full solution.  Landolt did *not* observe redder
stars, so how can you be sure whatever you measure exactly
matches his photomultiplier/filter system?  So observing, say,
Miras, will have errors between observers even if they
properly transform their data.  These errors will have
a systematic component since Miras change color during their
cycle.  Luckily, Miras and other LPVs have large amplitude
so that such systematic errors are often lost in the noise.
  Allyn Smith is trying to solve this problem a little more
carefully for the SDSS ugriz system, observing some very
red stars with the original equipment so that you rarely
have to extrapolate.  Even so, for truely accurate photometry,
you should consider nonlinear solutions for both red stars
and also for filters that deviate significantly from the standard
system under normal conditions.
Arne

VSNET Home Page

Return to Daisaku Nogami

---

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