8. REDSHIFT-DISTANCE CATALOGS

As redshift measurements accumulated in the 1970s and 1980s, it was widely recognized that there was a need to assemble these data into comprehensive catalogs. Beginning with the publication of the CfA redshift survey in 1983 (Huchra et al. 1983), all major redshift surveys (see the Chapter by Strauss in this volume) led to electronically available databases in fairly short order.

Comparable efforts involving redshift-independent distance measurements have been slower in coming. This is largely due to the issue of uniformity. Whereas redshift measurements by different observers rarely exhibit major differences, redshift-independent distances obtained by different observers can, and generally do, differ systematically for any number of reasons. In some cases the origin of such differences is different calibrations of the DI. In others, the calibrations are the same but the input data differ in a subtle way. Finally, the way statistical bias effects are treated (Section 9) often differs among those involved in galaxy distance measurements. For all these reasons, it is not possible simply to go to the published literature, find all papers in which galaxy distances are reported, and lump them together in a single database. Instead, individual data sets must be assembled, their input data and selection criteria characterized, their DI relations recalibrated if necessary, and the final distances brought to a uniform system. Only then can the resultant catalog be relied upon - and even then, caution is required.

The first steps toward assembling homogeneous redshift-distance catalogs were taken in the late 1980s by David Burstein. His goal was to combine the then newly-acquired D_n- data from the 7-Samurai group (Section 4) with the extant data on spiral galaxy distances, especially the infrared TF data obtained by the Aaronson group (Section 3). Burstein's efforts produced two electronic catalogs, the Mark I (1987) and Mark II (1989) Catalogs of Galaxy Peculiar Velocities. ⁽³⁾ Burstein's chief concern was matching the TF and D_n- distance scales. As there are, by definition, no galaxies that have both kinds of distances, this matching could be carried out through a variety of overlapping approaches. The approach decided upon by Burstein, in consultation with the other 7 Samurai, was to require the Coma cluster spirals and ellipticals to have the same mean distances. Although this procedure was imperfect, the Mark II catalog was considered reliable enough to be used in the first major effort to constrain the density parameter ₀ by comparing velocities and densities (Dekel et al. 1993).

With the publication of a number of large, new TF data sets in the early 1990s, the need for a greatly expanded redshift-distance catalog became apparent. An important development was the superseding of the majority of the older infrared TF data, obtained by the Aaronson group, by CCD-based (R-and I-band) TF data. Han, Mould and coworkers (Han 1991, 1992; Han & Mould 1992; Mould et al. 1991, 1993) obtained a full-sky cluster TF sample, based on I -band magnitudes and 21 cm velocity widths, comprising over 400 galaxies. Willick (1990, 1991) and Courteau (1992; Courteau et al. 1993) gathered R-band TF data in the Northern sky for over 800 galaxies in total. The largest single contribution was that of Mathewson et al. (1992) who published an I-band TF sample of 1355 galaxies in the Southern sky. Despite the influx of the new CCD data, one portion of the infrared TF database of the Aaronson group was not rendered obsolete: the sample of over 300 local (cz 3000 km s^-1) galaxies first observed in the late 1970s and early 1980s (Aaronson et al. 1982). This local sample was, however, subjected to a careful reanalysis by Tormen & Burstein (1995), who rederived the H-band magnitudes using a more homogeneous set of galaxy diameters and inclinations than was available to the original researchers a decade earlier.

In 1993, a group of astronomers (myself, Burstein, Avishai Dekel, Sandra Faber, and Stéphane Courteau) began the process of integrating these TF data and the existing D_n- data into a new redshift-distance catalog. Our methodology is described in detail in Willick et al. (1995, 1996), and portions of the catalog are presented in Willick et al. (1997). The full catalog, known as the Mark III Catalog of Galaxy Peculiar Velocities, is quite large (nearly 3000 spirals and over 500 ellipticals, although this includes several hundred overlaps between data sets) and is available only electronically, as described in Willick et al. (1997).

Building upon the foundation laid by Burstein in the Mark I and II catalogs, the Mark III catalog was assembled with special emphasis placed on achieving uniform distances among the separate samples it comprises. Four specific steps were taken toward this goal. First, the raw data in all of the TF samples underwent a uniform set of corrections for inclination and extinction (cf. Section 3.2). Second, the TF relations for each sample were recalibrated using a self-consistent procedure that included correction for selection bias (Section 9). Third, final TF zero points were assigned by requiring that the TF distances of objects common to two or more samples agree in the mean. This step ensures that the different samples are on similar relative distance scales. The global TF zero point was determined by the fully-sky Han-Mould cluster TF sample. (As explained in Section 3, this zero point was such that the distances are given in units of km s^-1, not Mpc.) Fourth, the spiral and elliptical distance scales were matched by applying the POTENT algorithm (see the chapter by Dekel in this volume) to each separately, and requiring that they produce statistically consistent velocity fields.

In parallel with the efforts of the Mark III group, similar enterprises have been undertaken by two other groups. Brent Tully has also assembled and recalibrated much of the extant TF data. Riccardo Giovanelli, Martha Haynes, Wolfram Freudling, Luiz da Costa, and coworkers have acquired new I band TF data for ~ 2000 galaxies, and have combined it with the sample of Mathewson et al. 1992). Initial scientific results from each of these efforts have been published (Shaya et al. 1995; Giovanelli et al. 1996; da Costa et al. 1996), and the catalogs themselves will soon become publically available.

New distances for elliptical galaxies, now mostly from the FP rather than D_n- (Section 4), continue to be obtained as well. Jorgensen et al. (1995a,b) have published distances for E and S0 galaxies in 10 clusters out to 10,000 km s^-1. The EFAR group (Burstein, Colless, Davies, Wegner, and colleagues) are now finishing an FP survey of over 80 groups and clusters at distances between 7000 and 16,000 km s^-1 (Colless et al. 1993; Wegner et al. 1993, 1996; Davies et al. 1993).

Implicit in all this ongoing work is that the Mark III catalog, like its predecessors, is just one step along a path still being traveled. Just as the Mark III data consists in part of recalibrated data alreay present in the Mark II, so will future catalogs incorporate, partially recalibrate, and expand upon the Mark III. Of particular note are the distances coming from the SBF survey of Tonry and coworkers (Tonry et al. 1997; cf. Section 5). The SBF distances are much more accurate than either TF or D_n- and can provide important checks on them. Tonry et al. (1997) have taken initial steps toward such an intercomparison, and the preliminary results, which suggest mutally consistent results among SBF, D_n-, and TF, are encouraging. Little comparison of SNe and BCG distances with other DIs has yet been carried out, but will be in the coming years. It is reasonable to hope that, by the turn of the century at the latest, the available redshift-distance catalogs will be superior, in terms of sky coverage, accuracy, and homogeneity, to the best we have today.