|Annu. Rev. Astron. Astrophys. 2014. 52:
Copyright © 2014 by Annual Reviews. All rights reserved
The very first steps to use far-infrared emission as a tool to study galaxy evolution date back to the pioneering IRAS mission, when 60 μm source counts in the ecliptic pole region were found to exceed no-evolution models (Hacking, Condon & Houck 1987). The Infrared Space Observatory (ISO) obtained the first deep surveys at both mid- and far-infrared wavelengths, detecting strong evolution of the luminosity function out to z ~ 1 and supporting by plausible extrapolation that such dusty galaxies constitute the cosmic infrared background. ISO also pioneered the application of the rich mid-infrared spectra as a tool of studying energy sources and physical conditions in dusty galaxies (see Genzel & Cesarsky 2000 for a review). The Spitzer Space Telescope revolutionized mid-infrared surveys and opened the window to direct mid-infrared spectroscopy of faint high-z galaxies (review by Soifer, Helou & Werner 2008). The Akari mission (Murakami et al. 2007) provided uniquely detailed mid-infrared photometric coverage. Independently and at a similar time, ground-based (sub)millimeter surveys at wavelengths 850 μm - 1.2 mm detected luminous `SCUBA galaxies' (Blain et al. 2002), which are among the most actively star forming systems at redshifts z ~ 2.
For most redshifts of interest, both spaceborne mid-infrared and groundbased submillimeter surveys miss the rest frame far-infrared peak that dominates the spectral energy distributions of most galaxies. Extrapolation via template spectral energy distributions (SEDs) is then needed to characterize the energy budget of samples that were selected at these wavelengths. Also, such samples can be significantly biased compared to a truly calorimetric selection by infrared luminosity. When aiming for the crucial deep far-infrared surveys, fully cryogenic space missions with small 60-80 cm diameter primary mirrors, such as ISO, Spitzer, and Akari, were strongly limited by source confusion. At 250-500 μm, results from the 2 m BLAST balloon telescope (Pascale et al. 2008) provided a first deep glimpse at the extragalactic sky, but are now superseded by Herschel data. The European Space Agency's Herschel Space Observatory (Pilbratt et al. 2010), in operation 2009-2013, for the first time met the need for a combination of sensitivity and large aperture (3.5 m) over the full far-infrared and submillimeter wavelength range, substantially reducing confusion levels. At 350 μm and longer wavelengths, low spatial resolution Planck all-sky maps and catalogs (2032013aPlanck Collaboration et al. ) support studies of galaxy evolution.
Much of this review is based on Herschel photometric surveys and pointed observations using the camera modes of the PACS (70, 100, 160 μm, Poglitsch et al. 2010) and SPIRE (250, 350, 500 μm, Griffin et al. 2010) instruments. We will refer to this range as the `far-infrared', leaving the `(sub)millimeter' terminology to ground based surveys at typically 850 μm or longer wavelengths. Similarly, `Submillimeter galaxy (SMG)' will refer to the type of galaxy detected in these surveys, while equivalent Herschel selected sources will be called `dusty star forming galaxy (DSFG)' where appropriate.
Traditionally, studies in the local universe make use of a terminology of `luminous infrared galaxies (LIRGs)' defined by their total 8-1000 μm infrared (IR) luminosity LIR > 1011 L⊙, and their `ultra-' and `hyper-' luminous ULIRG and HYLIRG equivalents above 1012 and 1013 L⊙, respectively (e.g. Sanders & Mirabel 1996). These are handy acronyms, but for the purpose of galaxy evolution studies it is important to recall that connotations of these classifications that were carefully calibrated in the local universe may not apply at high redshift. For example, local ULIRGs are found to be major mergers with unusually dense and warm interstellar medium. The same cannot necessarily be assumed for their higher redshift equivalents at same infrared luminosity. Where used in this review, the (U)LIRG acronyms should be seen as pure infrared luminosity classifications, without further implications for the nature of the galaxy under study.
1.1. An inventory of Herschel surveys
Figure 1. Survey area and point source depth reached by some extragalactic surveys with Herschel PACS (top, shown for 100 μm but 160 μm data are availabe for the same fields) and SPIRE (bottom, shown for 250 μm but 350 and 500 μm data are available). Exposure is computed from total observing time (including overheads) and survey area. The point source depths (including confusion noise) shown on the right axis should hence be seen as indicative only, since no attempt was made to capture detailed effects of different observing layouts in the various projects. Some sets of cluster observations are included, at the total area summing all objects. Surveys shown are from the projects listed in Section 1.1 as well as other projects for the Akari-NEP (PI S. Serjeant), Akari deep field south (PI T. Takagi), SPT deep field (PI J. Carlstrom) and SDSS stripe 82 (Viero et al. 2013a).
The Herschel general purpose observatory served a wide range of science goals, using its capabilities that include both photometric imaging and spectroscopy over the 55-672 μm wavelength range. Guaranteed and open time for in total about 15% of Herschel's observing time was devoted to imaging surveys of galaxy evolution, not counting pointed imaging or spectroscopic studies of individual high redshift sources. The list below summarizes some of the major efforts, from the initial guaranteed time and open time Herschel key programmes as well as from similar scale programs that were allocated after later calls for proposals. In addition, a good fraction of the science discussed below is based on smaller scale imaging projects as well as pointed imaging and spectroscopic studies which are not listed explicitly. Herschel extragalactic surveys exceed 1000 square degrees in total but do not reach the full sky coverage that Planck provides at lower spatial resolution for λ ≥ 350 μm and Akari at lower resolution and sensitivity for λ≤ 160 μm. Extragalactic confusion limits and instrument design imply a focus for PACS surveys on small area deep surveys, while SPIRE surveys typically emphasize large area. Below we list some major surveys, and Figure 1 gives a graphical overview of area and depth of fields covered.
HerMES (http://hermes.sussex.ac.uk/content/hermes-project, Oliver et al. 2012) is a wedding cake type multipurpose survey of blank and cluster fields with focus on SPIRE, covering 70 square degrees in its shallowest tier. A shallow 270 square degree extension has been obtained in the HeLMS project (PI M. Viero).
PEP (http://www.mpe.mpg.de/ir/Research/PEP/index.php, Lutz et al. 2011) is a PACS survey of popular multiwavelength fields such as GOODS, COSMOS, EGS, ECDFS, Lockman-XMM and cluster fields, coordinated with SPIRE observations of the same fields from HerMES.
GOODS-Herschel (http://hedam.oamp.fr/GOODS-Herschel/, Elbaz et al. 2011) provides deep PACS and SPIRE observations of GOODS-North and ultradeep PACS coverage of part of GOODS-South. A combined PEP/GOODS-Herschel dataset including all PACS data of the two GOODS fields is published in Magnelli et al. (2013).
The Herschel Lensing Survey (HLS)
(http://herschel.as.arizona.edu/hls/hls.html, Egami et al. 2010) provides deep PACS and SPIRE data of 44 X-ray luminous clusters as well as SPIRE snapshots of another 527 clusters. HLS aims at both lensed background objects and at cluster members.
H-CANDELs (PI M. Dickinson) provides deep
Herschel data of the CANDELs
(http://candels.ucolick.org/index.html) subregions of the COSMOS and UKIDSS-UDS fields, that were not yet covered at equivalent depth by the projects mentioned above.
Figure 2 visualizes an example for the post-Herschel status of deep surveys over the full mid-infrared to submillimeter wavelength range. Surveys with beams of width 5" to 30" are now available over the full range. In the deepest fields, they reach the confusion limit for all wavelengths except 70 μm, where a fully cryogenic 3 m class telescope such as the SPICA project will be needed.
Figure 2. Current status of deepest 24-870 μm infrared surveys, visualized by 4' × 4' cutouts in the HUDF region. Data are from the GOODS project (24 μm), PEP and the combined PEP and GOODS-Herschel data (70-160 μm), HerMES (250-500 μm), and the groundbased LESS survey (870 μm, Weiß et al. 2009).