Perhaps this is mainly for Bill, but maybe there'll be some other
interest.

I perhaps should have mentioned that the corrected Tycho-2 photometry
forms an excellent supplement to the 'loneos.phot' file, or even the main
one, with my file as the supplement. The reason is that the fainter "good"
Tycho photometry falls in the middle range of magnitudes in the GSC. The
GSC magnitudes are fine in the mag. 10-11 range; it is only brighter than
about mag. 8 or so that they are squirrelly. In fact, I would avoid using
any stars brighter than 8th or 9th for the calibration since that could skew
things for the fainter ones. As best I can tell, there is no scale error in
the GSC data at the faint end, so that a zero-point fix in the middle range of
brightness should be fine at the faint end.

One thing not really made clear in your description is that the "color"
one gets after the GSC calibration is the one corresponding most closely to
the plate material used for a particular field. The adjustment is not always
to V, in fact usually not, but instead either B or R. Thus for most of the sky
south of +3 Dec, the corrected GSC magnitudes will be B magnitudes, and mostly
R north of that, with numerous special cases, such as along the entire southern
galactic plane, where they will be V magnitudes, around M31 (B), etc. Thus
depending on where you are observing, the magntiudes will shift around as a
function of the source plate material. This should make no difference to the
MPC since it will be specified in the observation report, and since asteroids
are roughly all the same color at the sort of error level we're talking about
here. (As an aside, it is fairly ridiculous for folks to report a magnitude
with every observation. If you take a set of three or four measurements in a
single session, simply report the average of these.)

I would tighten up the Tycho-2 error limit. Accepting stars up to 0.1
mag. error really lets you in for some truly bad photometry. Remember that
the Tycho-2 errors are _underestimated_ by 50 percent(!); see Figure 6 and
accompanying text of the main Tycho-2 construction paper, where the authors
admit that the error budget is incomplete for reasons unknown. But basically,
where the entry says the error is 0.10, you're already up to external errors
of 0.15 mag., and that's a mean error of the mean, not 1-sigma. I would avoid
the photometry of any Tycho-2 star where the errors in _either_ VT or BT exceed
0.050. As you've probably seen for small areas, the GSC photometry is at least
as consistent as that. Yeah, I know it gets rid of a lot of stars, but the
Tycho photometry on the faint stars is garbage, and the GIGO law still obtains.
Then as per the first paragraph, I would limit the Tycho-2 photometry sample to
stars having corrected V > 9.0 _and_ with errors on _both_ VT and BT <= 0.050.

To obtain B and R magnitudes when the necessary colors were not available,
did you use the Caldwell et al. transformations, which I think I sent you code
for some months ago?

To get standard V and B from Tycho-2, it is probably best to use the
relation shown by Mike Bessell in the July 2000 PASP, which is not out yet in
print, but can be obtained from the PASP Web page. Non-subscribers will not
be able to read the paper on-line, however. Even worse, Bessell does not give
an algebraic relation, but instead shows a cubic spline fit with a look-up
table. In a recent MNASSA (Monthly Notices of the Astronomical Society of
South Africa) Alan Cousins (he of Cousins VRI) has also published a relation
based on the E-region standards, as did Bessell. Unfortunately this is
expressed in terms of the standard colors, not the ones observed by Tycho, so
it is not useful for converting Tycho data. I have copied out Bessell's table
below as a flat ASCII list.

External tests. One obvious test is to match the results against some
of Arne Henden's files. Where I have used Arne's data in 'loneos.phot', it
is typically only 10 percent of the stars present in any given field, usually
selected so that there are 3-5 stars per magnitude interval, and uncrowded in
USNO-Ax.x. To get enough GSC stars, you'll have to choose a low-latitude
field (avoid the Sagittarius region however, since things are extremely
crowded there). Arne has also posted a bunch of results from the last couple
of weeks of his observing that I have not looked at at all, so on those you
could use any star. In addition, I have a "secret" file of unpublished BV
photometry that I was asked not to include yet in the public copy. This
includes several hundred stars mostly bright enough to appear in the GSC.
I can send this list to you for testing purposes. Yet another source of solid
photometry is the Landolt stars that I did not include in 'loneos.phot' (again
I only sampled that data). Plenty of mag. 10-12 stars there that can be used
as an external check.

Is the GSC plate without any Tycho-2 stars one of the special ones? There
certainly is _not_ a full 6x6-deg Schmidt field without Tycho-2 stars.

All for now.

\Brian

===============================================================================

from: Bessell 2000, PASP 112, 961 (July)

RELATION BETWEEN BT-VT AND HIPPARCOS/TYCHO DATA FOR
B-G MAIN-SEQUENCE STARS AND K-M GIANTS

[note that this means you don't have to worry about what kind of star is
involved, since other types are statistically quite rare.]


BT-VT V-VT del(B-V) V-Hp
-0.250 0.038 0.031 0.066
-0.200 0.030 0.021 0.051
-0.150 0.022 0.011 0.036
-0.100 0.015 0.005 0.021
-0.050 0.008 0.002 0.006
-0.000 0.001 -0.005 -0.011
0.050 -0.005 -0.010 -0.025
0.100 -0.012 -0.017 -0.038
0.150 -0.018 -0.020 -0.048
0.200 -0.024 -0.021 -0.058
0.250 -0.029 -0.023 -0.069
0.300 -0.035 -0.025 -0.079
0.350 -0.040 -0.025 -0.087
0.400 -0.045 -0.026 -0.094
0.450 -0.050 -0.030 -0.101
0.500 -0.054 -0.035 -0.108
0.550 -0.059 -0.045 -0.114
0.600 -0.064 -0.051 -0.120
0.650 -0.068 -0.060 -0.127
0.700 -0.072 -0.068 -0.131
0.750 -0.077 -0.076 -0.134
0.800 -0.081 -0.085 -0.137
0.850 -0.085 -0.094 -0.142
0.900 -0.089 -0.104 -0.147
0.950 -0.093 -0.113 -0.151
1.000 -0.098 -0.122 -0.155
1.050 -0.102 -0.131 -0.158
1.100 -0.106 -0.142 -0.157
1.150 -0.110 -0.154 -0.160
1.200 -0.115 -0.166 -0.162
1.250 -0.119 -0.178 -0.164
1.300 -0.124 -0.189 -0.166
1.350 -0.128 -0.199 -0.166
1.400 -0.133 -0.210 -0.165
1.450 -0.138 -0.222 -0.164
1.500 -0.143 -0.234 -0.161
1.550 -0.148 -0.245 -0.157
1.600 -0.154 -0.256 -0.153
1.650 -0.160 -0.266 -0.148
1.700 -0.165 -0.277 -0.143
1.750 -0.172 -0.288 -0.137
1.800 -0.178 -0.299 -0.131
1.850 -0.185 -0.309 -0.125
1.900 -0.191 -0.320 -0.119
1.950 -0.199 -0.331 -0.112
2.000 -0.206 -0.342 -0.106
=end=