3.2. Submillimeter Detections of Distant Galaxies

Submillimeter (sub-mm) identifications and observations of distant galaxies has recently been energized by array detectors, most notably the Submillimeter Common User Bolometer Array (SCUBA; Holland et al. 1999) on the 15 m James Clerk Maxwell Telescope on Mauna Kea. SCUBA has enabled the first deep, unbiased surveys to be made of the sub-mm sky (generally implemented at 850 µm). Sub-mm observations are particularly sensitive to dust emission, associated with reprocessed light from young, high-mass stars in galaxies undergoing massive bursts of star formation. The thermal dust spectrum peaks at a rest-frame wavelength of around 100 µm which has the interesting result of negative sub-mm k-corrections: the redshifted, modified blackbody spectrum is sufficient to offset cosmological dimming for ₀ = 1 and 1 z 10! Even for low values of ₀, the 850 µm flux density is only expected to decrease by a factor of a few over this redshift range (Blain & Longair 1993; Hughes et al. 1998). Dusty, starburst galaxies can therefore be selected at z 1 in an almost distance-independent manner. Several relatively bright sources have recently been identified (e.g., Smail, Ivison, & Blain 1997; Barger et al. 1998, 1999; Hughes et al. 1998) and the flux densities may occasionally imply immense star formation rates in excess of 1000 M yr^-1 (e.g., Dey et al. 1998)! A new question is what fraction of young galaxies are dusty enough to reemit strongly at _rest 100 µm? In particular, do the sub-mm galaxies constitute an orthogonal population to the star-forming, UV-bright galaxies considered in Section 4?

The main difficulty in the current interpretation of weak extragalactic sub-mm sources is the relatively large (15") beam size of SCUBA at 850 µm. The issue is reminiscent of the early days of follow-up work on radio sources: sub-mm field sources have uncertain identifications. For example, the Hughes et al. (1998) deep 850 µm observations of the Hubble Deep Field (HDF) detects five sources to a limiting flux density of 2 mJy, corresponding to a surface density of ~ 0.9 sub-mm sources arcmin^-2 to this limiting brightness. The sources are likely to be associated with individual (starburst) galaxies over a large redshift range, with some contribution of AGNs likely. The critical point is, of course, which galaxies? And what are their redshifts? Hughes et al. (1998) suggest that one source (HDF 850.4; S_{850 µm} = 2.3 ± 0.5 mJy) is associated with an extreme starburst galaxy (HDF 2-339.0; I₈₁₄^AB = 23) at z 1, whilst the other four sub-mm sources could be identified with more distant galaxies, perhaps out to z 4. However, recent deep high angular resolution radio observations of the HDF by Richards (1999) suggest alternate identifications of the sub-mm sources: translating the astrometric center of the sub-mm map by 6" increases the number of radio identifications of sub-mm sources from one out of five to four out of five. If correct, this would then suggest that the presently detectable sub-mm sources in the HDF are restricted to z 2.

Carilli & Yun (1999) show that the radio-to-sub-mm spectral index is a viable redshift indicator for star-forming galaxies. Using semianalytic models and the well-studied local starburst galaxies M82 and Arp 220, they show that the 1.4 GHz to 850 µm (350 GHz) spectral index _{1.4 GHz}^{850 µm} increases with redshift; galaxies with _{1.4 GHz}^{850 µm} 0.5 are likely to be at z 1. Examining the Hughes et al. (1998) sub-mm identifications in the HDF, Carilli & Yun (1999) conclude that most of the sources are at z 1.5. Carilli & Yun (1999) also show that the Richards (1999) 6" offset yields low-redshift identifications with inconsistently large radio-to-sub-mm spectral indices, implying that the Hughes et al. (1998) astrometry is likely accurate.

Smail et al. (1997, 1998) report on sub-mm selected sources from a SCUBA cluster lens survey. Deep sub-mm imaging of clusters offers two advantages: first, gravitational lensing magnifies any background sources (a median amplification of ~ 2.5 is reported for this survey), making spectroscopic follow-up easier. Second, gravitational lensing increases the mean separation of background sources, thus diminishing source confusion. Barger et al. (1999) reports on a spectroscopic follow-up study of the possible optical counterparts and suggest that the majority of the sub-mm sources reside at z < 3. This implies that the far-infrared and sub-mm background light recently measured from the FIRAS and DIRBE experiments on the COBE satellite (e.g., Schlegel, Finkbeiner, & Davis 1998) are emitted by sources at z < 3. Furthermore, this work suggests that the peak activity in heavily obscured (dusty) sources (both AGN and starbursts) lies at relatively modest redshift.

In the absence of high angular resolution sub-mm capabilities, this new field enjoys considerable uncertainties. Further research is clearly needed, and we can look forward to rapid progress in understanding the redshift distribution, host galaxy properties, dust properties, and evolution of the sub-mm population in the near future as the next generation of mid-IR/sub-mm instruments (e.g., SCUBA+, BOLOCAM) become available.