The Herschel Multi-tiered Extragalactic Survey: SPIRE-mm photometric redshifts

Abstract

International audienceWe investigate the potential of submmmm and submmmmradio photometric redshifts using a sample of mm-selected sources as seen at 250, 350 and 500 mu m by the SPIRE instrument on Herschel. From a sample of 63 previously identified mm sources with reliable radio identifications in the Great Observatories Origins Deep Survey North and Lockman Hole North fields, 46 (73 per cent) are found to have detections in at least one SPIRE band. We explore the observed submm/mm colour evolution with redshift, finding that the colours of mm sources are adequately described by a modified blackbody with constant optical depth tau=(/nu0)beta, where beta=+1.8 and nu 0=c/100 mu m. We find a tight correlation between dust temperature and IR luminosity. Using a single model of the dust temperature and IR luminosity relation, we derive photometric redshift estimates for the 46 SPIRE-detected mm sources. Testing against the 22 sources with known spectroscopic or good quality optical/near-IR photometric redshifts, we find submm/mm photometric redshifts offer a redshift accuracy of vertical bar Delta z vertical bar/(1+z)= 0.16 (\textlessvertical bar Delta z vertical bar \textgreater=0.51). Including constraints from the radiofar-IR correlation, the accuracy is improved to vertical bar Delta z vertical bar/(1+z)=0.15 (\textlessvertical bar Delta z vertical bar \textgreater=0.45). We estimate the redshift distribution of mm-selected sources finding a significant excess at z \textgreater 3 when compared to similar to 850 mu m selected samples

Dusty star-forming galaxies at high redshift

“…The unique advantage of 450µm mapping with Scuba-2 is its superb spatial resolution, ∼7 , which is significantly better than the spatial resolution of Herschel at 500µm, ∼36 . Figure 19 (right panel) plots the redshift distribution for galaxies selected at 450µm from data described by Chen et al (2013a), Casey et al (2013), Geach et al (2013) and Roseboom et al (2013) and compares it to some statistical determinations of the redshift distribution for 500µm-selected galaxies from Herschel (Béthermin et al, 2012b,a) . The differences could be caused by the different field depths, where the Casey work extends over larger areas and is able to detect rare high-z systems while the Roseboom work is more sensitive to fainter galaxies at z ∼ 1 − 2.…”

“…see Carilli & Yun, 1999;Hughes et al, 2002;Aretxaga et al, 2007;Chapin et al, 2009;Yun et al, 2012;Roseboom et al, 2012;Barger et al, 2012;Chen et al, 2013a). These 'millimetric redshifts' are determined from the shape of the far-infrared or submillimeter SED or its colors rather than any properties indicated by its stellar emission characteristics in the optical/near-IR or emission line signatures.…”

“…The Scuba-2 Cosmology Legacy Survey (S2CLS) is the Guaranteed Time project on Scuba-2 which is dedicating the most observing time to blank-field extragalactic mapping at both 450µm and 850µm. Several initial works (Chen et al, 2013a;Geach et al, 2013;Casey et al, 2013;Chen et al, 2013b;Roseboom et al, 2013) describe galaxies selected at 450µm and compare and contrast them to galaxies selected at longer wavelengths.…”

“…First, several works fit the longwavelength data ( > ∼ 50µm) to a single temperature modified blackbody and then cut-off the SED at short wavelengths and attach a power-law SED such that S (ν) ∝ ν −α where dS /dν < ∼ − α Younger et al, 2007Younger et al, , 2009bRoseboom et al, 2013). In other words,…”

“…Roseboom et al (2010) introduced a cross-matching technique, dubbed 'XID,' which uses 24µm and radio positional priors to determine both accurate positions as well as deboosted flux densities for Herschel-selected galaxies (HSGs). The technique (updated in Roseboom et al, 2012) identifies significant Herschel detections and investigates nearby 24µm or radio sources as possible counterparts using a likelihood estimator. Since Herschel maps are confusion limited, this multi-wavelength counterpart matching must be done iteratively across the entire map, and not individually source by source.…”