Giovanelli & Haynes (1991) discuss the general techniques and limitations of the use of the HI 21-cm line as a redshift indicator, as well as reviewing HI redshift survey work through 1991. We refer the reader there for a more general review. Here we will discuss some of the major previous surveys in an attempt to illustrate the advantages of using HI observations for studying specific samples of galaxies. We divide our comments into two categories, depending on whether or not they are targeted toward galaxies identified in advance by their emission at some other wavelength.
3.1. Targeted Surveys
Nearby Galaxies and the Structure of the Local Supercluster Using the NRAO 91-m and 43-m telescopes and the Bonn 100-m antenna, Fisher & Tully (1981) measured the redshifts, HI fluxes and velocity widths for roughly 1000 nearby galaxies, supplementing their earlier survey of nearby DDO dwarf galaxies (Fisher & Tully 1975). The majority of these objects are located in the volume of the Local Supercluster cz < 3000 km s-1, and these observations served as the backbone for the development of the Nearby Galaxy Catalog (Tully 1988). These data have played an important role in determining the three dimensional structure of the Local Supercluster (Tully & Fisher 1987).
The highly flattened disk-like structure of the Local Supercluster can be seen in Figure 1, which shows the sky distribution of galaxies on at Aitoff equal area projection in celestial coordinates of ~ 5700 galaxies with known redshift cz < 3000 km s-1. The figure is centered on 6h, 0°. The large concentration of galaxies in the left side of the diagram (northern galactic cap) is the Virgo cluster. The supergalactic plane (e.g., Tully 1982) runs through Virgo, and extends northward, passing close to the north celestial pole. It can be traced around to the southern galactic cap, although the density of galaxies in the anti-Virgo region is significantly reduced compared to the north.
Figure 1. All-sky Aitoff projection map showing the positions of ~ 5700 galaxies with measured redshifts less than 3000 km s-1. The figure is plotted in equatorial coordinates, with the center of the figure at 6h and 0°. The dotted lines denote the location of the galactic plane and lines of b = ±20°. The Virgo cluster is the densest collection of objects on the left side of the figure. The supergalactic plane runs through Virgo, and can be traced most of the way around the sky.
HI redshifts have also contributed significantly to the understanding of the dynamical structure of the Virgo Cluster. Hoffman and collaborators (e.g., Hoffman et al. 1989a) have used the Arecibo telescope to measure the HI emission from spiral and dI galaxies included in the Virgo cluster catalog (Binggeli et al. 1985). Many of these objects are of such low optical surface brightness that optical redshift measurements are impractical. In using HI redshifts to map the distribution of blue compact dwarfs (BCDs) in Virgo, Hoffman et al. (1989b) have found that those objects are found in two main clumps, one near the cluster core but believed from its lower heliocentric velocity to be infalling from behind, and a second, more diffuse aggregate, located to the south and behind the core, coincident with the so-called W group. Most recently, Hoffman et al. (1995) have confirmed that the cluster membership assignment of Binggeli and collaborators based on the optical appearance of the galaxies works very well. "Background" galaxies as catalogued by Binggeli et al. are indeed located behind Virgo.
Dwarf and Low Surface Brightness Galaxies A number of HI redshift surveys have targeted samples of galaxies which were selected from optical catalogs as likely having significant HI content while at the same time being optically faint. Such galaxies would be difficult targets in optical redshift surveys, and in fact are often skipped over in programs that claim to represent "complete magnitude-limited" samples of galaxies. We stress that the galaxies included in the studies mentioned below represent a very heterogeneous sample, ranging from relatively nearby gas-rich dwarf irregular (dI) galaxies to more distant low surface brightness (LSB) disk galaxies with luminosities comparable to the Milky Way. As stressed by McGaugh (1996b), LSB does not necessarily imply low luminosity.
Schneider et al. (1990; 1992) used both Arecibo and the NRAO telescopes to observe a large sample of galaxies selected from the Uppsala General Catalogue (UGC, Nilson 1973) that were classified as dwarfs or had low surface brightness. The spatial distribution of this sample is discussed by Thuan et al. (1991). Bothun et al. (1986) also selected LSB galaxies from the UGC and obtained 21-cm redshifts for 375 objects using Arecibo. Specific areas of the sky were targeted in redshift surveys for dwarf galaxies by Hoffman et al. (1989a) in the Virgo cluster, and in two low density regions of space by Eder et al. (1989) and Salzer et al. (1990). Schombert et al. (1994) have obtained HI redshifts for 134 ultra-LSB galaxies selected from searches of the POSS-II plates (e.g., Schombert et al. 1992). A few recent studies have focused on observations of dwarf and LSB galaxy samples at southern declinations using the NRAO 43-m telescope (Maia et al. 1993; Gallagher et al. 1995; Matthews et al. 1995).
The Zone of Avoidance Efforts to delineate the large-scale structure in the local universe are hampered by the obscuration at optical wavelengths close to the galactic plane. HI redshift surveys have thus contributed significantly to the mapping of the galaxy distribution at low galactic latitudes, including the region of the Great Attractor. Some surveys have targeted galaxies identified through careful examination of optical or infrared plates or images. For example, Seeberger et al. (1994) and Pantoja et al. (1994) looked at newly discovered optically selected galaxies behind the northern zone of avoidance, while Kraan-Korteweg & Huchtmeier (1992) used Bonn to observe faint galaxies in the Puppis region. A number of surveys have targeted sources discovered by the Infrared Astronomical Satellite (IRAS) using color criteria to separate stars and galaxies close to the galactic plane., e.g., Lu & Freudling (1995), Chamaraux et al. (1995), Yamada et al. (1994), Takata et al. (1994). In combination, these surveys are contributing to our understanding of the large-scale structure behind low latitude regions.
General Redshift Surveys As indicated in the preceding paragraphs, most programs for obtaining HI redshifts have focused on specific subsets of the galaxian population (e.g., LSB disk galaxies, dwarf irregulars, etc.). In contrast, a few HI survey projects were done according to the schemes of some of the well known optical surveys: obtain redshifts for all cataloged galaxies in a specific area of the sky down to a specific magnitude or diameter limit. The largest such program was carried out by Giovanelli, Haynes, and collaborators (e.g., Giovanelli et al. 1986; Bicay & Giovanelli 1987; Freudling et al. 1992; Wegner et al. 1993) using the NRAO and Arecibo telescopes. In particular, this group took advantage of the sensitivity of the Arecibo reflector, which allowed for the acquisition of redshifts in a fraction of the time required to obtain optical spectra for the same galaxies. Of course, their samples were restricted to disk galaxies (type Sa or later); the impact of this bias is discussed below. When combined with optical redshifts for the early-type galaxies, the regions of the sky surveyed by this group (most notably the southern galactic cap between 0° and 40° declination) have very complete coverage for galaxies in the UGC and CGCG (Zwicky et al. 1961-1968) down to apparent magnitudes of 15.5. Most recent surveys have focused on extending the HI database to southern declinations (e.g., Bottinelli et al. 1993) or to faint galaxies in clusters (Haynes et al. 1996).
3.2. Blind Surveys
Current catalogs of optical galaxies identify objects not only by their luminosity but also by their surface brightness. Numerous authors (e.g., Disney 1976) have pointed out that large numbers of galaxies that do not meet such selection criteria might exist. Blind HI surveys can contribute significantly to the identification of gas-rich, low luminosity and/or low surface brightness objects. To date, blind surveys have been limited to relatively small volumes (e.g., Henning 1995; Weinberg et al. 1992; Szomoru et al. 1994), and their conclusions are not definitive. Schneider (this volume) describes in detail past and current blind HI survey programs.
A new survey being conducted with the Dwingeloo 25-m telescope by Kraan-Korteweg et al. (1994) has already found a large, previously unrecognized nearby galaxy, Dwingeloo I. Note that this galaxy and another object also in Cassiopeia were detected in HI independently by Huchtmeier et al. (1995) with the Bonn telescope.
The development of multi-feed array receivers offers the opportunity to cover large volumes in much less time. The Parkes Multibeam Survey Working Group (Staveley-Smith et al. 1996) plan to use the new 13-beam multifeed receiver on the Parkes 62-m telescope to conduct two surveys. The first will be a deep HI survey for optically obscured galaxies in the zone of avoidance. The survey will target the portion of the galactic plane |b| < 5° over 213° < l < 33°, simultaneously producing not only the search for hidden galaxies, but also a survey of the Galactic HI and a 20-cm pulsar survey. 1500 hours of observing time have been requested to begin in August 1996. The second survey will cover the entire southern sky, surveying for galaxies in a six million Mpc3 volume. A complementary survey using a four beam array receiver will be conducted at Jodrell Bank (Disney 1995). Further discussion of the potential for future redshift surveys is found in Vanden Bout & Haynes (this volume).