The first generation of extragalactic submm-wave surveys have provided an important complement to more traditional optical and radio searches for distant galaxies, and discovered a cosmologically significant population of very-luminous, high-redshift dusty galaxies.
We have found that is very hard to study a complete sample of submm galaxies at other wavelengths (Smail et al., 2002). The similar experience of other groups involved in deep mm/submm surveys (Barger et al., 1999a; Eales et al., 2000; Carilli et al., 2001; Scott et al., 2002; Webb et al., 2002b) is reflected in the relatively few papers describing the individual multi-waveband properties of the almost 200 galaxies detected. The most sensitive follow-up observations are required in the near-IR, radio and optical wavebands to identify and study them (Frayer et al., 2000; Ivison et al., 2001), that is very faint detection thresholds of order 10 µJy at 1.4-GHz, K 23 and B >> 26 respectively. Much more time has been devoted to multi-waveband follow-up observations than was spent on the initial submm detections. Typical examples of the submm population can be detected in imaging-mode SCUBA observations in about 10 h of integration. However, at least 2 h of near-IR observations at the 10-m Keck telescope and about 24 h of integration at the VLA are then required in order to find likely counterparts to typical submm galaxies. The advantage of the VLA radio observations over those at optical and near-IR wavelengths is the very large field of view, which allows many galaxies to be detected simultaneously. The very brightest optical counterparts to submm galaxies can be identified spectroscopically in about 7 h of integration using 4-m class telescopes (Ivison et al., 1998) and higher-quality spectra can be obtained in a comparable time using 8-m class telescopes (see the results of a 5-h integration using the UVES spectrograph at the European Southern Observatory (ESO) VLT by Vernet and Cimatti, 2001). In all cases, identifying a plausible counterpart, where possible, is only a first step; finding a redshift for these typically faint, red galaxies is much more challenging. In this context, the unusual sensitivity of submm surveys to the most distant galaxies is almost a drawback, making it very hard to detect a complete sample of submm galaxies at other wavelengths.
Key questions for understanding submm galaxies in the future include:
What are the properties of typical submm galaxies in other wavebands, and what is their relationship to other high-redshift galaxy samples? The submm-selected galaxies appear to a diverse mixture of types, including bright merging systems (Ivison et al., 1998a, 2000a, 2001), optical QSOs (Kraiberg Knudsen et al., 2001), EROs with K < 20 (Smail et al., 1999, 2002; Gear et al., 2000; Lutz et al., 2001), and much fainter IR-detected galaxies (Frayer et al., 2000), which may also turn out to have very red colors. It seems that the overlap between the 850-µm submm galaxy population and both the LBGs and faint Chandra X-ray sources is small. Note that some of this apparent diversity is sure to be due to the very wide redshift distribution of the submm galaxies.
What is the redshift distribution of the submm galaxies? Models of the evolution of submm galaxies that do not grossly violate basic observational constraints on the source counts and cosmic background radiation are easy to generate. However, it is vital to predict a plausible redshift distribution, with only a small fraction at redshifts less than unity, and a probable median redshift of at least 2-3. It is easy to generate a redshift distribution that is biased too high. The observational determination of a redshift distribution for a well-defined sample of submm galaxies remains a crucial goal. This will be easy with ALMA. In the meantime, concerted and time-consuming campaigns of optical and near-IR spectroscopy will pay off gradually, while observations of cm-wave megamasers and the development of wide-band mm- and cm-wave spectrometers may offer alternative routes. The forthcoming (sub)mm interferometers CARMA and SMA, and developments of the IRAM PdBI will also provide accurate positions and some CO redshifts for submm galaxies.
What are the details of the astrophysics responsible for the luminosity of the submm galaxies? This is very important, as the submm galaxies appear to be signposts to some of the most luminous and violent phases of galaxy evolution, and could be associated with the formation of the bulk of galactic bulges, elliptical galaxies and supermassive black holes (Lilly et al., 1999; Dunlop, 2001). Whether these galaxies are formed in a single event, or as a series of lesser bursts, is a key question for our understanding of the process of galaxy formation and evolution. Detailed comparisons of the luminosity derived from dust continuum emission, the dynamical mass inferred from molecular line profiles, the evolved stellar mass inferred from near-IR observations, and the spatial extent of the activity from various high-resolution observations will all be important for disentangling the complex astrophysics of these systems.
When can submm instruments be used to resolve and study high-redshift galaxies in detail? This is already practical given enough observing time at the OVRO MMA and the IRAM PdBI. The CARMA and SMA interferometers will soon have important roles to play in these studies. In about 10 years, ALMA will provide the first real chance to detect and study galaxies rapidly and in great detail using submm observations alone. Luminosities, redshifts, dynamical masses and metallicities could all be determined without needing to resort to radio, optical and near-IR observations as a matter of course. However, because ALMA has a relatively small field of view, the most efficient survey strategy may be to detect large numbers of galaxies using wide-field mm/submm cameras like BOLOCAM, SCUBA-II and their successors on single-antenna 10-50-m aperture survey telescopes, and the Herschel and Planck Surveyor space missions, and then use ALMA to provide detailed images and spectra of all the detected galaxies.
What is the fundamental limit to making submm observations of distant galaxies? Submm observations rely on the presence of metals, in the form of molecular gas or dust grains in order to detect galaxies. While submm radiation is able to travel unattenuated across the Universe from prior to the epoch of reionization, it is possible that a large fraction of pre-reionization `first-light' sources are insufficiently dusty and metal rich to be detectable as continuum sources. Low-metallicity galaxies should be detectable by fine-structure C and O far-IR line emission, however. It would be tremendously exciting to see the birth of the first metal-enriched dusty systems with ALMA, and so perhaps to determine directly the redshift limit for submm surveys. Of course, even if this were possible, ALMA would still have a long and fruitful career studying the detailed astrophysics of galaxies out to and beyond redshift 5, while the search for the most primitive galaxies in the second and third decades of the century is taken up by space-based mid-IR interferometers and the SKA radio telescope (see Fig. 8).
Submm observations of the distant Universe are a new tool for probing the earliest and most dramatic stages of the evolution of galaxies. Over the years to come, the capabilities of submm-wave observatories, and our understanding of the Universe in this new window, should continue to advance dramatically.
This work is heavily based on results obtained from the SCUBA Lens Survey. The following have all been involved with aspects of the SCUBA lens survey: Lee Armus, Amy Barger, Jocelyn Bezecourt, Leo Blitz, Len Cowie, John Davies, Alastair Edge, Aaron Evans, Andy Fabian, Allon Jameson, Tom Kerr, Jean-Francois Le Borgne, Malcolm Longair, Leo Metcalfe, Glenn Morrison, Frazer Owen, Naveen Reddy, Nick Scoville, Genevieve Soucail, Jack Welch, Mel Wright and Min Yun. We thank the staff of the JCMT for operating and the UK ATC for providing SCUBA.
We thank Omar Almaini, Vicki Barnard, Frank Bertoldi, Jamie Bock, Chris Carilli, Helmut Dannerbauer, Darren Dowell, Steve Eales, Jason Glenn, Sunil Golwala, Dean Hines, Kate Isaak, Kirsten Kraiberg Knudsen, Attila Kovacs, Andrew Lange, Simon Lilly, Ole Möller, Priya Natarajan, Max Pettini, Tom Phillips, Kate Quirk, Enrico Ramirez-Ruiz, Nial Tanvir, Neil Trentham, Paul van der Werf, the editor Marc Kamionkowski, an anonymous referee and Roberta Bernstein for useful conversations and comments on the manuscript.
AWB was supported in Cambridge by the Raymond and Beverly Sackler Foundation as part of the Foundation's Deep Sky Initiative Program at the IoA. IRS is supported by the Leverhulme Trust and the Royal Society. JPK is supported by CNRS. Full references and acknowledgement to the instruments and telescopes used in this research can be found in Smail et al. (2002). This research has made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration.