Extragalactic surveys made by SCUBA and MAMBO have spatial resolutions of 10 to 15 arcsecs, with positional errors for the detected sources of 1 to 3 arcsecs. It is difficult to associate the submillimeter galaxies with any optical counterparts, not only because of the ambiguity in identifying the correct counterpart due to the positional uncertainty, but also because of the possibility that the optical emission from the counterpart may be completely absorbed by dust. Infrared observations, which are less sensitive to the presence of dust, have proved to be more successful in identifying the correct counterpart.

The submillimeter source SMMJ04431+0210 was originally detected in the SCUBA Lens survey, which used foreground clusters as massive gravitational lenses to efficiently search for high-redshift submillimeter galaxies. SMMJ04431+0210 provides a good example of the ambiguities that arise during the efforts to identify the correct optical and IR counterpart. Smail et al. 1998 originally identified the optically-bright spiral at z=0.18 as the counterpart to the submillimeter source SMMJ04431+0210 (see figure). Subsequent IR observations by Smail et al. 1999 then identified an Extremely Red Object (ERO), an IR source at z>2, that was invisible at optical wavelengths and which offered a more plausible counterpart. It was not until recently that Frayer et al. 2003 obtained an IR spectroscopic redshift (z=2.51) for the ERO. The identification of the ERO as the true counterpart to SMMJ04431+0210, however, was not confirmed until millimeter-wavelength spectroscopic observations detected redshifted CO(3-2) at z=2.51 (Neri et al. 2003).

It is still possible that optical or IR spectroscopically-determined redshifts may be completely unrelated to those of the submillimeter galaxies. Future millimeter-wavelength spectroscopy, with ultra-wideband heterodyne receivers, will accurately measure redshifted CO-lines without ambiguity. But for a limited number (hundreds) of submillimeter galaxies, BLAST offers an immediate alternative solution. Despite the poorer resolution of BLAST (30 to 60 arcsecs), the simultaneous observations at 250, 350 and 500 μm will provide an estimate of the photometric redshift for luminous submillimeter galaxies (L_FIR > 3x10¹² L_sun) identified in the BLAST surveys, with a conservative average error of Δz<=±0.5 over the redshift range 0<z<6 (see Hughes et al. 2002 for details). With this photometric-redshift accuracy and large statistically significant samples (>2000 galaxies), BLAST will provide a robust measure of the global star formation history for dusty, optically-faint (or invisible) starburst galaxies.