8 TRC & ACT

  In parallel with the TRC production, the US Naval Observatory undertook a similar project based on the proprietary machine-readable version of the AC (Urban et al. 1998a). The resulting "Astrographic Catalogue and Tycho'' (ACT) Catalogue (Urban et al. 1998b) has a lot in common with TRC:

• Both catalogues have the same aim, viz. to extend the ICRS/Hipparcos system to fainter stars.
• The two catalogues are based on exactly the same observational material: the Astrographic Catalogue and the Tycho Catalogue.
• Both groups had to solve exactly the same set of problems: AC reduction onto the Hipparcos system, identification of Tycho stars in AC, etc.

This means that TRC and ACT might be expected to be of very similar stellar content and quality. A first comparison (Høg et al. 1998a) roughly confirmed this expectation. However, this does by no means say that the two catalogues are essentially identical: there are lots of differences due to the different catalogue construction strategies. We illustrate this by just two examples.

The basic difference between the identification techniques used by TRC and ACT is that the latter adopted the proper motions given in the Tycho Catalogue. Prior to identification, the position of a Tycho star was propagated to the AC plate epoch using the published proper motion. This difference led to discrepant identifications in ACT and TRC and, in consequence, to large disagreements of the proper motion of some stars in the two catalogues.

Concerning the AC reduction there is a major difference in the sequence of the individual steps, which influenced the results for all stars. For TRC, the intermediate reference catalogue (ACRS) was first transformed to the ICRS/Hipparcos system, and then used for the derivation of the zone-specific corrections to the AC data. The Hipparcos Catalogue served as a reference at the final reduction step (cf. Sect. 3.1) to put individual AC plates onto the ICRS system. For ACT, the AC reduction was done on the original ACRS, and the resulting AC positions were subsequently transformed to the Hipparcos system by means of a filtering approach (Urban et al. 1998a).

There is also an important difference in the meaning of the published data that any potential user of either catalogue should be aware of. While the standard errors given in ACT (1.6-1.8 mas/yr per proper motion component) represent the internal errors only, the corresponding TRC data attempt to incorporate at least some of the remaining systematic errors. For bright stars, the standard deviation of the ACT-Hipparcos differences is 3.5-3.8 mas/yr (Urban et al. 1998b).