Re: Tycho 2 catalog astrometry

[Stupendous Man <richmond@a188-l009.rit.edu>]

  Andrew Bennett discovered that Tycho-2 produces better astrometry
with Mark IV images than Tycho-1:

> This makes quite a big difference. For example, for
> file H3R1659.840 (CD15) my program using Tycho:
> Matched 235, accepted 215 sd 0.918 pe 0.745 arc sec
> For Tycho 2:
> Matched 352, accepted 324 sd 0.686 pe 0.541 arc sec

  I followed his example and ran astrometric routines on the
same image, H3R1659.840 (the image containing M5).  The results
follow his ... under one condition:

  Tycho-1      match 215 of 224       mean offset 0.76  median 0.47  arcsec

  Tycho-2  (all)     449    545                   0.86         0.58
  Tycho-2  (bright)  277    329                   0.58         0.38

  If I use _all_ the possible Tycho-2 stars, my routines yield residuals
which are slightly larger than those from Tycho-1.  On the other hand,
if I restrict my input to the Tycho-2 stars brighter than VT = 11.5,
which are very roughly the brighter half of all stars in Tycho-2,
then the results are significantly better than those from Tycho-1.

  For photometry, my results are again the same as Andrew's:

  Tycho-1      match 215 stars   mean offset 3.24  stdev 0.14   median 0.07

  Tycho-2 (all)      449                     3.20        0.24          0.11
  Tycho-2 (bright)   277                     3.22        0.17          0.09

  In this case, the Tycho-2 values agree slightly less well with the
Mark IV values than the Tycho-1 values do.

  Now, I need to find enough disk space here to hold the Tycho-2
catalog ....

                                                  Michael Richmond