[vsnet-sequence 167] (fwd) Re: Observation Report for Nova V4740 Sgr

[Taichi Kato <tkato@kusastro.kyoto-u.ac.jp>]

Date: Sat, 16 Mar 2002 13:44:02 -0700 (MST)
From: Brian Skiff <Brian.Skiff@lowell.edu>
Subject: Re: Observation Report for Nova V4740 Sgr

     I see several problems here.  Among them is the use of Tycho-1 data
instead of Tycho-2.
     The Tycho-1 numbers, besides having the color term, also have a scale
error, such that they are systematically too bright at the faint end by
several tenths of a magnitude.  This _not_corrected_for_ in the reported
Tycho-1 standard V/B-V!  In addition, the Tycho-2 data are based on a careful
summing of all the detections, whereas this was not done in Tycho-1.  Yes,
I know the Tycho-1 catalogue files conveniently show standard V and B-V
explicitly, but this is based on only a rough calibration.  I would recommend
avoiding them.  You will do much better using Tycho-2 VT/BT, and adjusting
those to the standard system using Bessell's look-up table, which I have
copied out here:

ftp://ftp.lowell.edu/pub/bas/starcats/bessell.tyc

If you wish you might try fitting a smoothing spline to the Bessell values
so as to yield a smooth series in, say, 0.01-mag. increments (instead of
0.05 as in the table).
     The first star on your list, TYC 7392 3006 1 = HD 166424, shows Tycho-2
VT = 9.974 and BT = 11.465.  Using the Bessell look-up tables, these correct
to V = 9.83 +/- 0.08 and B-V = 1.26 +/- 0.17.  V-I will be around 1.4, so the
I mag. ~8.4, nearly 0.4 brighter than you have assumed.
     Also, be sure to account for the uncertainties in the Tycho-2 or other
photometry.  For the example above, the nominal error on VT is 0.04, but if
you read the Tycho-2 construction paper (2000A&A...357..367H), and have a
look at Fig. 6, you see that below about mag. 9, the errors are
_underestimated_ by 50%, so really the error is already 0.06 on VT.  (The 
authors admit in the accompanying text that they don't know why their error
budget is underestimated.)  Then add about 0.02 for the error introduced in
transforming to standard V (Bessell suggests this value in his paper), and the
uncertainty becomes +/-0.08 mag.  The error on BT is 0.086, which becomes 0.15
with the fudge factor (50% + 0.02); the error on B-V is then +/-0.17.  This
is a 1-sigma mean error, so there is a 1-in-3 chance that the real value lies
even farther from the catalogue figure.  Whoa!  These uncertainties are
distressing if you're used to thinking that such a "bright" star should have
reliable Tycho photometry.  Similar problems will affect the other stars,
possibly worse since they are even fainter.

     A look at the 2MASS photometry errors shows uncertainties of 0.037 and
0.039 mag. on J and K, resp.  Thus the uncertainty on J-K is +/-0.054 mag.
Add on whatever slop you get from transformation and derived colors, and pretty
quickly the uncertainty becomes ~0.1 mag.  

     Sorry to make the Tycho-2 data look so unappealing, but I'm suggesting
that just problems with the values for the calibration stars could be messing
you up more than any errors in reductions or data-taking.

\Brian

VSNET Home Page

---

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