2 Source Catalogues

 

2.1 The Astrographic Catalogue

  Astrographic Catalogue resulted from the international cooperation "Carte du Ciel'' (CdC), initiated in 1887 in order to construct an all-sky photographic catalogue of all stars down to $m_{\rm pg}=11$ mag (AC), and a survey down to $m_{\rm pg}=14$ mag (CdC). Detailed description of the AC and CdC projects can be found elsewhere (Eichhorn 1974; Debarbat et al. 1988).

The Astrographic Catalogue provides rectangular coordinates as measured on the plate and magnitude estimates of about 8.6 million star images. The limiting magnitude of AC is roughly $m_\mathrm{pg}=12$ mag, though some zones give coordinates of stars as faint as 13 mag. The basic information on the zonal catalogues comprising the AC is provided in Table 1. It should be noted that by chance the plates centred at $+11\hbox{$^\circ$}$ were taken and measured twice, by the Bordeaux and Toulouse observatories. The Hyderabad Observatory was responsible for two zones. The Potsdam Observatory, originally responsible for the zone $+32\hbox{$^\circ$}$ to $+39\hbox{$^\circ$}$, had stopped the work soon after the First World War. Only about 30 per cent of the original plates taken and measured at Potsdam were published. Later on, this zone was reobserved at Oxford, Hyderabad and Uccle.

  

Table 1: The zones of the Astrographic Catalogue

2.1.1 Observation and measurement

  The Astrographic Catalogue was observed with so-called normal astrographs - a type of telescope built especially for the purpose, with an aperture of 33 cm and a focal length of 3438 mm (which gives a plate scale of $1\hbox{$^\prime$}=1$ mm). The photographs were taken on $16~{\rm cm}\times16~{\rm cm}$ plates with a useable field of $13~{\rm cm}\times 13~{\rm cm}$, which is equivalent to $2.1\hbox{$^\circ$}\times 2.1\hbox{$^\circ$}$.Normal astrographs were used by all participating institutions except the Hyderabad Observatory (see Sect. 3.6.3).

Primary coordinate system on the AC plates was realized by the réseau - a silvered glass plate with a quadratic net of fine lines at 5 mm spacing, which was copied on every plate. The measurements of the positions of star images were made with respect to the photographic images of the réseau lines. A secondary coordinate system was realized by a micrometer or a cross of scales in the eyepiece of the measuring microscope used to measure image coordinates relative to réseau lines. The mode of measurement employed by the individual observatories is given in Table 1; S means an eyepiece scale, M a micrometer screw, and E an eyepiece grid, a special measuring technique used by the Vatican Observatory only.

Brightness measures were derived by a number of methods: by measuring the image diameter, by matching the image with standard ones or by eye estimation (Gureeva 1992). The diversity of the AC brightness measures substantially impedes both their reduction to a modern magnitude system and the removal of the magnitude equation present in the measured coordinates. The worst cases in this sense are the Sydney, Cape and Perth zones due to the two distinct estimation methods used for every plate, one for the bright stars and another one for the faint stars. Consequently, the analysis of the magnitude equation in the data of these observatories was the most complicated of all. This situation could only be improved by using modern magnitudes for the stars.

2.1.2 On the precision of published AC measurements

  The accuracy of the published measurements is defined by the measurement errors and the digitization errors. The latter is $D/2\sqrt{3}$, D being the last decimal place of the published measurements. D varies from observatory to observatory, ranging from 0$.\!\!^{\prime\prime}$006 to 0$.\!\!^{\prime\prime}$3 (cf. Table 1). In the latter case the digitization error of 0$.\!\!^{\prime\prime}$09 becomes comparable to the measurement and instrumental errors, estimated as 2 $\mu$m or 0$.\!\!^{\prime\prime}$12 each.

Furthermore, an analysis of the last digit of the published coordinates suggests that its distribution is far from uniform. In the Tacubaya zone, for example, only the figures 0, 2, 5, 7 or 0, 3, 5, 8 occur as the last digit, and on any given plate only one of these two sets is present. Other figures are rare, probably being misprints. Moreover, the last digit in those zones where it corresponds to 0$.\!\!^{\prime\prime}$3 was found to be dominated by even numbers (or 0 and 5, in case of the Vatican zone). A similar effect for the Hyderabad zone was previously reported by (Eichhorn & Gatewood 1966). In such cases the digitization error is at least two times larger than the 0$.\!\!^{\prime\prime}$09 value predicted from a uniform distribution of the last decimal place, reaching as much as 0$.\!\!^{\prime\prime}$43 in the case of the Vatican zone. Thus for at least half of all AC measurements (i.e. the zones where the measurements are given to 0$.\!\!^{\prime\prime}$3) the accuracy of the published coordinates is defined by the digitization error rather than by the measurement accuracy. Obviously, a remeasurement of zones like Vatican or Tacubaya would nearly double the precision.

2.1.3 Published data and the machine-readable version

  Publication of the AC measurements proceeded from 1902 to 1964 and resulted in 254 printed volumes of raw data. The machine-readable version of the Astrographic Catalogue was constructed at Sternberg Astronomical Institute (SAI) in 1987-94 (Nesterov et al. 1991; Gulyaev & Nesterov 1992). Plate headers, running identification number, rectangular plate coordinates and brightness measures were keypunched. All the misprints reported in published erratum lists were corrected in the data during keypunching. The machine-readable version of the Astrographic Catalogue includes all published measurements of the 19 completed zones plus measurements of 406 published plates of the original Potsdam zone.

Verification of the keypunched AC data included manual and automated procedures which assured that the data comply to the formats and record sequencing used in the published volumes, that all the fields are present and all the exceptions (incomplete or uncertain data) are marked with special flags. A substantial number of keypunching errors and 5-10 times more frequent unreported misprints were detected and corrected on the basis of an identification of the AC data with the HST Guide Star Catalogue (GSC, Lasker et al. 1990; Russell et al. 1990; Jenkner et al. 1990). The final error rate of the machine-readable AC, i.e. of keypunching errors and misprints missing in the published errata lists and not found from AC-GSC matching, is estimated to be less than 0.1%.

2.2 The Tycho Catalogue

The Tycho Catalogue is based on the observations with the star mapper on board the ESA Hipparcos astrometric satellite. A detailed description of the Tycho observational campaign, data processing and production of the catalogue can be found in the printed introduction (ESA 1997, Vol. 4).

2.2.1 General characteristics

The Tycho Catalogue provides positions at mean epoch J1991.25, proper motions, parallaxes and two-colour photometry (in $B_{\mathrm{T}}$ and $V_{\mathrm{T}}$ bands) of 1052031 stars brighter than $V_{\mathrm{T}}=11.5$ mag. The Tycho astrometric data is referred to the ICRS system. The median standard error for stars at the median magnitude $V_{\mathrm{T}}=10.5$ mag and colour index $(B_{\mathrm{T}}-V_{\mathrm{T}})=0.7$ mag is estimated as 25 mas in position and 0.06 mag in photometry. Standard errors of Tycho astrometry and photometry as a function of $V_{\mathrm{T}}$ magnitude are shown in Fig. 1. The catalogue is more than 99 per cent complete down to $V_{\mathrm{T}}\sim10$ mag, the incompleteness basically occurring in dense fields.

  

Figure 1: Astrometric (left) and photometric (right) accuracy of the Tycho Catalogue, as a function of $V_{\mathrm{T}}$ magnitude (ESA 1997, Vol. 4)

2.2.2 Tycho stars selection criteria

  The Tycho Catalogue is intended to be a complete sky survey, and therefore not all Tycho stars are plausible candidates for an astrometric reference catalogue like TRC. For the sake of reliability of the TRC stellar content, it was decided to exclude Tycho stars with unreliable astrometry or photometry, along with the high-proper-motion stars. The latter constraint was imposed in order to limit the number of misidentifications of the Tycho Catalogue stars in the AC (cf. Sect. 4.2). Specifically, the following entries of the Tycho Catalogue were rejected:

• Hipparcos stars not observed by Tycho;
• stars with positions derived from the Revised Tycho Input Catalogue;
• stars with standard error of $B_{\mathrm{T}}$ or $V_{\mathrm{T}}$ exceeding $0.3^{\rm m}$;
• stars with Tycho proper motion modulus exceeding 180 mas/yr at a $3\sigma$ level.

The total number of Tycho stars surviving the rejection and consequently accepted for TRC was 1018531.