Steven Fellows ([email removed]) wrote in the last
digest about star information.
Okay. First off: you aren't going to get an 'accurate' location out
of that kind of data. Right ascension and declination can currently
be made accurate to less than a tenth of a second of arc, otherwise
we wouldn't be able to detect parallax. This means that a position
error in the plane perpendicular to our field of vision is quite
small. Unfortunately, the same cannot be said of the distance to
the star. Most of the stars we've observed close by have a parallax
error on the order of 10%. Some are less; some are much more. This
means that at a range of 25 parsecs the average error approaches
five parsecs toward or away from us. If, taking a 2300ad-style
approach to star travel, a starship cannot travel more than 2.3
parsecs between stars, the errors swamp our data.
Fine. There is no way to correct this. So, accept the data as if
it were without error and you get a star map. Probably not the true
one, by any means, but a star map. Once you have done
that, teeny corrections for the difference in R. A. and Dec. between
the 1950 epoch (that for which that data was compiled) and, say, the
2200 epoch become meaningless. Stars just don't move _that_ much in
250 years or so. (UVW velocities and box-orbit parameters are just
other notations for radial velocity.)
Declination is positive or negative because the original data was in
terms of north or south declination. By using +/- signs the
astronomers made a continuous range from +90 degrees to -90 degrees.
B-V and U-B color are ratios involving the magnitude of the star
viewed through different filters. It's a more direct way of
determining the color of a star than using the spectral class
(B/A/F/G/K/M), but I haven't figured out a good way to use them
either for that or for a bolometric magnitude correction yet.
It is not a value for red shift here.
The conversions for RA/Dec/Px to X/Y/Z coordinates are relatively
simple. Before you start, however, you need to decide if you want
to realign the axes. For example, it might be 'cool' to have X
point toward the galactic center and Y point to spinward or
something. Then you need to pick two points, one being the galactic
center and the other being a point 90 degrees away that is directly
spinward of us, and rotate the coordinates for all the stars. Note:
GDW didn't do this.
Once that is done, or not, plug these numbers in:
Z: 1/Px * sin Dec
Y: 1/Px * cos Dec * sin RA
X: 1/Px * cos Dec * cos RA
This will give you rectangular coordinates in parsecs away from the
celestial pole, the vector of the vernal equinox, and the line
perpendicular to both of those. If you want light years, multiply
by 3.27.
Note: In Excel it will be easier to do these if you have a column
for RA in radians and a column for Dec in radians. There doesn't
seem to be a DMS->Radians conversion function included, so you'll
have to write some big function that multiplies and adds them
together and stuff.
By the way, you can rather easily check your data against the Near
Star List. I found out that the base catalogs for both this catalog
and the one that GDW used have the same catalog numbers, so if the
catalog number has three digits chances are it'll also be in the
Near Star List. This came in handy trying to pin down some of the
2300ad stars that have either names or oddball catalog numbers
listed.
Jack Dietz
// [email removed]
// [email removed] -- Evil Geniuses for a Better Tomorrow