4. An independent comparison
One of the consequences of the differences in the construction of the TRC and the ACT catalogues is that Hipparcos stars are more like reference stars in the TRC and field stars in the ACT. This is mainly due to the moment at which the respective catalogues are put onto the Hipparcos reference system. In the construction of the ACT catalogue the ACRS 2 was used as the reference catalogue for the plate reductions, and only after the whole catalogue was completed was it transformed from the FK5 to the Hipparcos reference frame. Furthermore, this transformation was done in a zonal sense, i.e., not for each individual plate. In the TRC construction the ACRS was also used as the reference catalogue for the plate reductions. However, before starting the reduction process the ARCS was first transformed to the Hipparcos reference frame.
To make an evaluation of the ACT and TRC independent of the Hipparcos Catalogue we compared both catalogues with stars from the Southern Proper Motion Program 3 (SPM) (Girard et al. 1998; Platais et al. 1998). This catalogue is the only one available at this moment with sufficient accuracy (2.5 mas yr-1) and number of stars ( HIP stars and TYC stars are contained in the SPM). We calculated the normalized proper-motion differences , , , and for several magnitude intervals. The results are summarized in Table 1.
Table 1. Standard deviations in right ascension and declination for the and distributions for four different magnitude intervals.
The table shows that the and yield similar results: . Despite the similar standard deviations we still notice that the distribution is not properly fitted by a single Gaussian (see Fig. 4). The fact that the standard deviations of the distribution are smaller is mostly due to the typical errors in the SPM which are 2.5 mas yr-1 instead of 1 mas yr-1 for the HIP. The larger errors dilute the characteristics we clearly see in the distribution.
Based on this comparison of the TRC and the ACT with an independent catalogue (the SPM) we confirm the conclusion we reached earlier: a fraction of the ACT stars has underestimated errors.
© European Southern Observatory (ESO) 2000
Online publication: July 27, 2000