Alma Redshifts of Millimeter-Selected Galaxies From the SPT Survey: The Redshift Distribution of Dusty Star-Forming Galaxies

Abstract

Using the Atacama Large Millimeter/submillimeter Array, we have conducted a blind redshift survey in the 3 mm atmospheric transmission window for 26 strongly lensed dusty star-forming galaxies (DSFGs) selected with the South Pole Telescope. The sources were selected to have S 1.4 mm > 20 mJy and a dust-like spectrum and, to remove low-z sources, not have bright radio (S 843 MHz < 6 mJy) or far-infrared counterparts (S 100 μm < 1 Jy, S 60 μm < 200 mJy). We robustly detect 44 line features in our survey, which we identify as redshifted emission lines of 12 CO, 13 CO, C i, H 2 O, and H 2 O +. We find one or more spectral features in 23 sources yielding a ∼90% detection rate for this survey; in 12 of these sources we detect multiple lines, while in 11 sources we detect only a single line. For the sources with only one detected line, we break the redshift degeneracy with additional spectroscopic observations if available, or infer the most likely line identification based on photometric data. This yields secure redshifts for ∼70% of the sample. The three sources with no lines detected are tentatively placed in the redshift desert between 1.7 < z < 2.0. The resulting mean redshift of our sample isz = 3.5. This finding is in contrast to the redshift distribution of radio-identified DSFGs, which have a significantly lower mean redshift ofz = 2.3 and for which only 10%-15% of the population is expected to be at z > 3. We discuss the effect of gravitational

Dusty star-forming galaxies at high redshift

“…Lensing does impact the observed redshift distribution since lower redshift galaxies (z < 2) are less likely to be gravitationally lensed by a foreground object than higher redshift galaxies. Weiß et al (2013) argue that this effect is strong at z < 2, especially at z < 1, but it has little impact on lensing of higher redshift galaxies. In Figure 18 this lensing-bias is indicated by hashed gray areas at z < 2.…”

“…The distribution in infrared photometric redshifts for the same GOODS-S sample is illustrated in hashed light blue to demonstrate its overall consistency with the optical spectroscopic/photometric sample. (Right:) The redshift distribution of 1.4 mm-selected DSFGs detected with the South Pole Telescope Weiß et al, 2013). The redshifts are all spectroscopically confirmed in the millimeter.…”

“…First, submm galaxy number counts are steep with a sharp intrinsic cut-off, thus flux magnification by gravitational lensing moves sources from the the steep faint end to the bright end and produce counts that are flatter at the bright flux densities (this was first discussed as an interesting sub-population in Blain, 1996). Second, due to negative K-correction, longer wavelength observations probe higher redshift galaxies where the optical depth to lensing is rapidly increasing (see Figure 6 of Weiß et al, 2013). At wavelengths shorter than 500µm, while lensed DSFGs do exist, the bright-end counts are dominated by star-forming galaxies at low red-shifts where the lensed counts do not make up an appreciable fraction (Negrello et al, 2007;Bussmann et al, 2013;Wardlow et al, 2013).…”

“…Figure 18 (right) shows the observed FIR-spectroscopically-confirmed redshift distribution of 1.4 mm-selected DSFGs from Weiß et al (2013). This sample is unique in that it consists of galaxies that are most certainly gravitationally lensed because the 1.4 mm imaging from the South Pole Telescope is too shallow to detect distant DSFGs unlensed.…”

“…It is important to stress that, although the SPT sample is small, these galaxies are the most spectroscopically complete DSFG sample of any; 26 targets were blindly chosen, for which 23 millimetric spectroscopic redshifts were measured from ALMA (their composite millimeter spectrum is shown in Figure 16). The Weiß et al (2013) work also includes an interesting discussion of size evolution and its potential impact on the observed redshift distribution. If the mean physical size of DSFGs evolves with redshift (i.e.…”