TThe High-redshift Universe with Spitzer
Maruˇsa Bradaˇc , Department of Physics, University of California, Davis, CA 95616, USA
When did galaxies start forming stars? What is the role of distant galaxies in galaxy forma-tion models and the epoch of reionization? What are the conditions in typical star-forminggalaxies at z (cid:38) ? Why is galaxy evolution dependent on environment? The Spitzer
SpaceTelescope has been a crucial tool for addressing these questions. Accurate knowledge of stel-lar masses, ages, and star formation rates (SFRs) requires measuring rest-frame optical (andUV) light, which only
Spitzer can probe at high-redshift for a sufficiently large sample of typ-ical galaxies. Many of these science goals are the main science drivers for James Webb SpaceTelescope, and
Spitzer afforded us their first exploration.
Very little was known about the high-redshift universe when the
Spitzer Space Telescope [1] was being planned in the 90’s. The then optimistic science cases that were put forward werepredicting Spitzer could observe ordinary galaxies at z ∼ [2, Chapter 15]. Given that the highest-redshift galaxy discovered at that time was z spec = 4 . [3], this was quite an extraordinary claim.It was not until 1998 that the redshift five barrier was broken [4]. While spectroscopic searcheswith Lyman- α were proposed in the 1960’s, and radio galaxies were discovered beyond z > inthe late 1980’s [5, 6], it was the powerful new photometric method [7] to search for “Lyman BreakGalaxies” (LBGs) that opened a new window for efficient discoveries of high-redshift galaxies.This meant that high redshift galaxies were being discovered in large numbers, in particular with1 a r X i v : . [ a s t r o - ph . GA ] M a y ubble Space Telescope (
HST ).Spitzer, after its launch, soon followed these discoveries. While ordinary galaxies at z (cid:38) (i.e., those with characteristic luminosities L ∗ ) needed large time investment before they could bedetected, there are now many surveys that are deep enough to observe those galaxies and studytheir properties (Fig. 1). Among the first studies of z > galaxies with Spitzer were observationswith Spitzer’s Infrared Array Camera (IRAC, 8) of objects gravitationally lensed by foregroundclusters. This allowed early detection of galaxies at z = 6 . [9] and z = 6 . [10]. This wasonly possible through the considerable magnification of these galaxies afforded by the presence ofthe massive cluster. Gravitational lensing resulted in a significant increase in depth and resolution,a feature that Spitzer has continued to benefit from ever since. Larger samples at z (cid:38) werediscovered concurrently by virtue of deep observations of the HUDF [11, 12]. The exposure timeneeded to reach the required depth was many hours and for the first time this was achieved with Spitzer
Great Observatories Origins Deep Survey (GOODS, PI Dickinson) program for a largeenough area.The present landscape has changed considerably since early
Spitzer observations. While
HST is still the prime telescope to search for and identify the highest redshift galaxies,
Spitzer hasallowed us to study these galaxies in detail. For cosmic dawn sources (i.e., galaxies at z > ) Spitzer became a key telescope not only to help establish redshifts, but also in the study of stellarmasses, star formation rates, optical emission line strengths, and the identification of old stellarpopulations. 2or intermediate redshift sources ( < z < in this review), Spitzer has also been crucialin robustly determining their stellar masses. In addition, although not discussed in detail here (butsee Comment to Nature Astronomy by Daniela Calzetti), at z < observations with the MultibandImaging Photometer for Spitzer (MIPS, 13) have been central to measurements of cosmic starformation history. Finally,
Spitzer was not only succesful in finding high redshift galaxy clusters( z > ), but also in studying the stellar properties of their cluster members. In this review, we givea brief overview of how Spitzer has provided measurements confirming our expectations, as wellas of new puzzles that are changing the paradigm of galaxy formation at high redshifts.
Tracing star formation to the earliest times has been a long-standing goal of extragalactic astron-omy. In particular, studying the onset of star formation is of importance not only for galaxy forma-tion models, but also for studies of the early universe. The cosmic Dark Ages - when the Universewas filled with neutral hydrogen that was opaque to ultraviolet (UV) light - are thought to haveended around 500 million years after the Big Bang, when early light sources produced enoughenergetic photons to ionize the neutral hydrogen [14]. This phase is referred to as the epoch ofreionization, and is also the era of the formation of the first galaxies. It is now clear that the pro-cess was completed by z ∼ [15, 16, 17]. However, a direct link between early light sources andreionization requires a detailed understanding of when and how galaxies first formed and built uptheir stellar content. 3 pitzer played (and is continuing to do so with the archival data) a unique role in advancingour understanding of the formation and evolution of galaxies at z (cid:38) . Deep observations at . µ m and . µ m with IRAC probe rest-frame optical properties of these galaxies, hence Spitzer data iscritical for age (Fig. 2) and stellar mass determination at high-redshift. While observations with
HST measure UV light emitted by young stars,
Spitzer
IRAC measures the rest-frame optical lightfrom long-lived stars in galaxies. In addition, light from some of the most prominent rest-frameoptical emission lines (e.g., [O
III ]4959, 5007 ˚A+H β ) enter Spitzer . µ m and . µ m bands at z > , allowing us to measure their contributions.Deep observations with Spitzer were first undertaken in the HST deep fields. Of the mostambitious projects with
Spitzer designed (in part) for the purposes of observing high-redshift ( z > ) galaxies was the deep IRAC imaging of the HUDF field (IRAC Ultradeep Field IUDF programPI Labb´e, 18). Many large Spitzer surveys followed, the deepest blank-field survey to date beingGREATS [19], with a near-homogeneous observing depth of 200 hours over ∼ arcmin (seealso Fig. 3). Results quickly revealed that the galaxies at these redshifts have different rest-frameoptical properties from their lower-redshift counterparts.From the very beginning Spitzer benefited greatly from the magnification due to gravita-tional lensing. While lensing decreases the effective area surveyed, it more than compensates byincreasing the depth and resolution (Fig. 1). Many surveys have recently been executed with thisin mind. Among the largest are the Cluster Lensing And Supernova survey with Hubble (CLASH;20), Hubble Frontier Fields (HFF; 21), Reionization Lensing Cluster Survey (RELICS; 22). Many4argets for these surveys have been selected from Massive Cluster Survey (MACS; 23, 24). Eachof these surveys has its
Spitzer counterpart with IRAC Lensing Survey (PI Egami), iCLASH (PIBouwens, 25), SURFSUP (PI Bradaˇc, 26), SHFF (PI Capak), SRELICS (PI Bradac, 27). Thesesurveys have delivered many interesting results, especially at the lower end of the galaxy stellarmass function. For the largest among them, catalogs have been published as well and stellar prop-erties investigated [28, 29, 30, 31, 32, 33]. Perhaps the highest-redshift galaxy so far detected by
Spitzer to date is a lensed galaxy. In [34] the authors report a detection of MACS0647-JD z ∼ galaxy candidate strongly lensed by a cluster. It is likely a massive and rapidly star-forming galaxy.Already the first samples of galaxies that were detected in Spitzer /IRAC images (e.g., 18, 35,36, 37) showed stellar masses in the range from ∼ to ∼ M (cid:12) . Galaxies with stellar massescomparable to the Milky Way masses seemed surprisingly large for a universe younger than 1 Gyrold (e.g., 11). However, it quickly became clear that at least for some of these galaxies their masseshave likely been overestimated, due to their IRAC fluxes being boosted by strong nebular emissionlines like [O III ]4959, 5007 ˚A+H β (38, 39, 40, 41, 42, 43). Spitzer broad-band photometry can provide indirect measurements of the nebular lines. Forexample,
Spitzer /IRAC colors can be used to measure strengths of [O
III ]+H β lines for . (cid:46) z (cid:46) . galaxies, for which these lines are expected to fall in the . µ m band while the . µ m bandis relatively free of line contamination (Fig. 4; 41, 44, 45, 46). Similarly, [O III ]+H β lines land in . µ m for galaxies at (cid:46) z (cid:46) [19, 27, 27, 43, 47, 48], while the . µ m band remains relativelyfree of strong emission lines. 5he results from these works show that we are now faced with a puzzle. Taken the Spitzer colors at face value, the line strengths required to fit the data for many of these galaxies is extremelyhigh, with rest-frame equivalent widths
EW > ˚A. But there is also a degeneracy in thesemeasurements, as the color can also be boosted by an older stellar population in the form of aBalmer break and/or dust [49]. Thus, rest-frame optical emission lines are likely not to be the solecause of excess flux in the rest-frame optical bands. Unfortunately, spectroscopic observationsof rest-frame optical emission lines of the highest redshift galaxies are out of reach for the currentinstruments. The sensitivity of the InfraRed Spectrograph (IRS) on
Spitzer [50] was mostly limitedto observations of z < IR luminous galaxies [51, 52], and we will have to wait for JWST to allowus to perform efficient spectroscopic follow-up in the rest-frame optical.However, with
Spitzer we are already able to study some exceptional cases where we canmitigate or break these degeneracies. In particular, when spectroscopic redshifts are known, wecan break the degeneracy by considering particular redshift ranges where lines are shifted out ofthe filters and/or by adding external data. For example, one can, in principle, use Lyman- α fluxesto estimate the nebular emission line contribution [42] or can constrain the amount of dust usingmeasurements using ALMA [53, 54].The best case where the degeneracy has been lifted is a z = 9 . galaxy MACS1149-JDbehind the cluster MACS J1149.5 + . µ m Spitzer data in [55]. It was later detected in both, . µ m and . µ m , Spitzer bands using deeper data (Fig. 5; 26, 46, 56) and its redshift was spectroscopically measured by653]. For this galaxy the nebular emission lines are redshifted out of these
Spitzer bands, yet ithas a strong color excess. In addition, the cold dust content of the galaxy was constrained to bemodest from observations taken with ALMA, making at least cold dust an unlikely cause of the red
Spitzer color [53]. It is therefore highly likely the old ( ∼ Myr) stellar population is causing thered rest-frame optical color (46, 53, 57). This is surprising, given the galaxies would need to startforming significant amounts of stars shortly after the Big Bang. At z = 9 . , when the Universe isonly (cid:46) Myr old, the presence of a strong Balmer break can thus provide the timing of the firststar formation. In the case of MACS1149-JD the dominant stellar component formed about 250million years after the Big Bang, corresponding to a redshift of about 15.There are several other objects with
Spitzer detections at z (cid:38) where this experiment can berepeated. The highest spectroscopically confirmed galaxy detected by Spitzer (GN-z11, 58) doesnot show a red color. There are others that do, e.g. GN-z10-3 and GN-z9-1 [59] and Abell1763-1434 [27]. They all have . µ m excesses that are consistent with an evolved stellar population, butunfortunately they currently lack spectroscopic confirmations. If their redshifts are confirmed, it islikely that a re-analysis of these galaxies would indicate that their star formation occurred within Myr of the Big Bang.While the sample is still small, observations of galaxies at z > with Spitzer nonethelessshow that galaxies have unexpected properties. A large fraction of high-redshift galaxies haveeither unusually strong nebular emission lines, pronounced Balmer breaks indicating old stellarpopulations, or the large amounts of dust. All three possibilities are difficult to reconcile given the7ge of the Universe. E.g., old stellar population in MACS1149-JD puts formation redshift of themajority of stellar population ( ∼ M (cid:12) ) at z ∼ which is high given that simulations predictfirst stars only started forming at z ∼ [60]. Other objects are routinely showing large rest-frame equivalent widths ( EW ∼ ˚A) of nebular emission lines, these have not been observedat lower redshifts except for in some extreme cases (e.g., 61, 62). IRAC color excess could alsobe explained by the presence of dust, but it is difficult to produce a significant amount of dustneeded to explain the observations [49]. While more precise answers await the James Webb SpaceTelescope ( JWST ), it is clear even at present, the star formation models used in simulations at thehighest redshift are being constrained as a result of observations by
Spitzer . Studies of intermediate redshift ( < z < ) have also thrived thanks to Spitzer observations. Oneof the main diagnostics in galaxy evolution models is the evolution of the Star Formation RateDensity (SFRD, 63). However, robust determination of dust attenuation is essential to transformFUV luminosity densities into total SFRDs. Prior to
Spitzer observations star formation historywas determined out to z ∼ using mostly HST data [64, 65, 66]. It was only after Spitzer data(along with Herschel) was obtained, that the history of cosmic star formation could be robustlymeasured, and the finding that the majority of star formation density comes from dust-obscuredsources was established (63 and references therein). MIPS also played an important role in thesediscoveries at lower redshifts (e.g., [67, 68]). 8he true power of
Spitzer comes with robust measurements of stellar masses at these red-shifts. The stellar-mass function measurements were first undertaken for SCANDELS, SCOS-MOS, SPLASH, and UDF fields (e.g., 69, 70, 71, 72, 73, 74). But the largest of surveys to explorethis are The Spitzer Matching Survey of the UltraVISTA Ultra-deep Stripes (SMUVS, PI Caputi,75, 76) and Euclid/WFIRST
Spitzer
Legacy Survey (Moneti et al. in prep.). With SMUVS, a largefraction of galaxies ( (cid:38) ) which were previously detected in the optical, were, for the first time,also detected with
Spitzer . This allowed for a precise measurement of the stellar mass function.The latest results indicate that massive and intermediate-mass galaxies have different evolutionarypaths in the early universe [77].One of the key properties to describe galaxy growth is the ratio between star formation rate(SFR) and stellar mass ( M ∗ ), also known as specific star formation rate (sSFR = SFR /M ∗ ; 63, 78).This ratio depends heavily on good estimates of both stellar masses and SFR. In addition, sSFR isparticularly well-determined in the lensing fields, as its value is independent of magnification, andyet gravitational lensing allows us to study sub- L ∗ galaxies at high redshift [31]. Its evolution withredshift is one of the key questions in galaxy formation studies and we are still unclear as to exactlyhow sSFR evolves. While pioneering studies predicted a constant sSFR at high redshifts [79, 80],the later results, which included nebular emission lines in SED fitting [81] and better Spitzer datafind an increase at < z < and beyond (e.g., 31, 71, 73, 82, 83, 84, 85, 86, 87).With Spitzer also SFR using H α was investigated in detail for the first time at < z < ,which is regarded as one of the most reliable among the easily accessible nebular SFR tracers9e.g.,88). Just like with [O III ], at intermediate redshifts there are ranges where emission linestrength of H α can be measured with Spitzer (44, Fig. 4). One of the caveats is that H α is blendedwith N II . Still, this has been used to indirectly measure the H α strength at . (cid:46) z (cid:46) . anddeduce the star formation rates [73, 76, 87, 89, 90, 91, 92]. More recently, in [93] authors measuredburstiness of star formation by comparing ultra-violet, H α luminosity, and H α equivalent-widthof z ∼ − main sequence galaxies, indicating that for at least half of their sample the starformation history is not smooth. Access to H α also allows modelling of Lyman-continuum photonproduction efficiency ξ ion . In [94] the authors used stacking of 300 < z < galaxies to estimateH α equivalent widths and found that ξ ion is not strongly dependent on luminosity and is thussimilar to the values derived for brighter galaxies. This has important implications, in particularfor establishing that faint galaxies are able to produce the Lyman-continuum photons needed forcosmic reionization.While Spitzer has improved our knowledge of stellar masses and SFRs considerably, severalissues remain. In particular, the evolution of sSFR is still not a completely solved problem. Thesituation will likely improve soon with the data taken from
Spitzer in the last year, as well as in thefuture with missions like
JWST , Euclid and the Wide-field Infrared Survey Telescope (WFIRST).
The present epoch is one of a rapid decline in the global star formation rate, but clusters experiencean evolution in star formation activity over this time that is even stronger than in the field. Identi-10ying the processes that trigger and terminate star formation in cluster galaxies (e.g. ram-pressurestripping, starvation, merging, harassment; 95, 96), and contrasting them to those operating in thefield is key to understanding the causes of the general decline. Furthermore, studying clustersat earlier times provides new constraints on both the evolution of the cluster abundance and theevolution of early-type galaxies over a substantial look-back time [97].
Spitzer plays a large rolein identifying both high-redshift ( z > ) clusters and protoclusters. Red-sequence overdensitiesand color searches can efficiently be done with Spitzer and have been utilized in both high-redshiftcluster and protocluster searches. In particular, the inclusion of
Spitzer data dramatically increasesthe accuracy/precision of photometric redshifts at z > , which is crucial for finding structures andto robustly estimate the environment (e.g., 98).Early on, Spitzer helped in discoveries of the protoclusters. In [99] the authors present thediscovery of a z = 5 . protocluster that dates back to Gyr after the Big Bang (Fig. 6). The fieldcontains a luminous quasar as well as a large galaxy rich in molecular gas. The Clusters AroundRadio-Loud AGN (CARLA) survey used this fact, as CARLA targeted radio-loud quasars with
Spitzer at z > . , discovering several proto-clusters surrounding them [100, 101]. Spitzer data isused to improve redshift information, as well as to estimate stellar masses, SFR and ages as theseare the only bands available that probe the optical/NIR rest-frame for high-redshift protoclusters.
Spitzer played an even more prominent role in finding the highest redshift galaxy clusters.Shallow yet wide surveys, in particular the IRAC Shallow Survey, enabled detections of manyhigh-redshift clusters (102, 103, 104). More discoveries were made in the
Spitzer
Wide-Area11nfrared Extragalactic (SWIRE) Survey using a simple color excess in
Spitzer bands [105]. Thelarge area is needed as clusters of galaxies are extremely rare, and one requires such a survey tofind the most massive examples (using numbers from [106] ∼ deg are needed to detect onehigh-redshift z (cid:38) , massive M > M (cid:12) cluster; Fig. 1).The Spitzer
Adaptation of the Red-sequence Cluster Survey (SPARCS; 107, 108; Fig. 6) wasthe first to perform a comprehensive study of z > galaxy clusters with all IRAC and MIPS bands.At z = 1 . , the 1.6 micron bump [109], due to infrared emission of stars, is shifted into Spitzer . µ m band, making it a useful tool to detect cluster members. Finally, surveys like IRAC DistantCluster Survey [110], MaDCoWS (using WISE, but utilizing Spitzer for confirmation; 106) and
Spitzer
South Pole Telescope Deep Field [111, 112] have also detected many rich clusters at z > .Even more importantly, both IRAC and MIPS enabled the study the stellar properties of clustermembers [113, 114]. The main conclusion from these surveys is that while low-redshift clustersare considered mostly star formation graveyards, at the earlier times ( z > ), galaxies in galaxyclusters were star forming and active [113]. This is thus the star-forming epoch of galaxy clusters,the study of which was enabled by Spitzer . When
Spitzer was first planned, nobody was expecting it to do the groundbreaking discoveries atthe cosmic dawn. This was mostly due to the mirror size, yet the 85cm diameter mirror (onlyslightly larger than a wine-barrel) has surpassed a lot of predictions. It was meant to explore “the12ld, the Cold, and the Dusty”, and yet it also explored the starry, the many, and the first.
Spitzer let us study early star formation in the first galaxies at the epoch of reionization, it delivered newinsights on star formation at high-redshift, and allowed us to find and study the earliest galaxyclusters and protoclusters. While the next space telescope (JWST) will revolutionize these fieldsin many ways, it will not surpass
Spitzer in its ability to do survey science thanks to
Spitzer ’scombination of a large field of view, excellent sensitivity, and long life. So long
Spitzer , andthanks for all the photons.
Acknowledgements
Based on observations made with the
Spitzer
Space Telescope, which is operated bythe Jet Propulsion Laboratory, California Institute of Technology under a contract with NASA. Support forthis work was provided by NASA through ADAP grant 80NSSC18K0945, NSF grant AST 1815458 andthrough an award issued by JPL/Caltech. The author would like to thank Brian Lemaux and Victoria Straitfor their help with the manuscript.
Financial Interest
The authors declare no competing financial interests.
Correspondence
Correspondence should be addressed to M.B.
1. Werner, M. W. et al.
The Spitzer Space Telescope Mission.
Astrophys. J. Suppl. , 1–9(2004). astro-ph/0406223 .2. Werner, M. & Eisenhardt, P.
More Things in the Heavens: How Infrared Astronomy isExpanding our View of the Universe (Princeton University Press, 2019).13. Schneider, D. P., Schmidt, M. & Gunn, J. E. PC 1158+4635: an Optically Selected Quasarwith a Redshift of 4.73.
Astron. J. , 1951 (1989).4. Dey, A., Spinrad, H., Stern, D., Graham, J. R. & Chaffee, F. H. A Galaxy at z=5.34. Astro-phys. J. Lett. , L93 (1998). arXiv:astro-ph/9803137 .5. Lilly, S. J. Discovery of a Radio Galaxy at a Redshift of 3.395.
Astrophys. J. , 161 (1988).6. Chambers, K. C., Miley, G. K. & van Breugel, W. J. M. 4C 41.17: A Radio Galaxy at aRedshift of 3.8.
Astrophys. J. , 21 (1990).7. Steidel, C. C., Giavalisco, M., Pettini, M., Dickinson, M. & Adelberger, K. L. SpectroscopicConfirmation of a Population of Normal Star-forming Galaxies at Redshifts Z > Astrophys.J. Lett. , L17+ (1996). arXiv:astro-ph/9602024 .8. Fazio, G. G. et al.
The Infrared Array Camera (IRAC) for the Spitzer Space Telescope.
Astrophys. J. Suppl. , 10–17 (2004). astro-ph/0405616 .9. Egami, E. et al.
Spitzer and Hubble Space Telescope Constraints on the Physical Propertiesof the z ∼ Galaxy Strongly Lensed by A2218.
Astrophys. J. Lett. , L5–L8 (2005).10. Chary, R.-R., Stern, D. & Eisenhardt, P. SpitzerConstraints on the z= 6.56 Galaxy Lensed byAbell 370.
The Astrophysical Journal , L5–L8 (2005).11. Yan, H. et al.
Rest-Frame Ultraviolet-to-Optical Properties of Galaxies at z ˜6 and z ˜5 inthe Hubble Ultra Deep Field: From Hubble to Spitzer.
Astrophys. J. , 109–127 (2005). astro-ph/0507673 . 142. Eyles, L. P. et al.
Spitzer imaging of i’-drop galaxies: old stars at z ∼ Mon. Not. R. Astron.Soc. , 443–454 (2005). astro-ph/0502385 .13. Rieke, G. H. et al.
The Multiband Imaging Photometer for Spitzer (MIPS).
Astrophys. J.Suppl. , 25–29 (2004).14. Planck Collaboration et al.
Planck 2015 results. XIII. Cosmological parameters.
Astron.Astrophys. , A13 (2016). .15. Fan, X., Carilli, C. L. & Keating, B. Observational Constraints on Cosmic Reionization.
Annu. Rev. Astron. Astrophys. , 415–462 (2006). astro-ph/0602375 .16. Robertson, B. E., Ellis, R. S., Furlanetto, S. R. & Dunlop, J. S. Cosmic Reionization andEarly Star-forming Galaxies: A Joint Analysis of New Constraints from Planck and the Hub-ble Space Telescope. Astrophys. J. Lett. , L19 (2015). .17. Mason, C. A. et al.
Inferences on the timeline of reionization at z ∼ Mon. Not. R. Astron. Soc. , 3947–3969 (2019). .18. Labb´e, I. et al.
Ultradeep Infrared Array Camera Observations of Sub-L* z ∼ ∼ Astrophys. J. Lett. , L26–L31 (2010). .19. Stefanon, M. et al.
The Brightest z (cid:38)
Astrophys. J. , 99 (2019). .150. Postman, M. et al.
The Cluster Lensing and Supernova Survey with Hubble: An Overview.
Astrophys. J. Suppl. , 25 (2012). .21. Lotz, J. M. et al.
The Frontier Fields: Survey Design and Initial Results.
Astrophys. J. ,97 (2017).22. Coe, D. et al.
RELICS: Reionization Lensing Cluster Survey.
Astrophys. J. , 85 (2019). .23. Ebeling, H. et al.
A Complete Sample of 12 Very X-Ray Luminous Galaxy Clusters at z > . . Astrophys. J. Lett. , L33–L36 (2007). arXiv:astro-ph/0703394 .24. Repp, A. & Ebeling, H. Science from a glimpse: Hubble SNAPshot observations of massivegalaxy clusters.
Mon. Not. R. Astron. Soc. , 844–864 (2018). .25. Bouwens, R. J. et al.
A Census of Star-forming Galaxies in the z ∼ − Universe basedon HST+Spitzer Observations over 19 Clash Clusters: Three Candidate Z ∼ − Galaxiesand Improved Constraints on the Star Formation Rate Density at z ∼ . . Astrophys. J. ,126 (2014). .26. Bradaˇc, M. et al.
Spitzer Ultra Faint SUrvey Program (SURFS UP). I. An Overview.
Astro-phys. J. , 108 (2014). .27. Strait, V. et al.
Stellar properties of z (cid:38)
The Astrophysical Journal , 124 (2020).168. Merlin, E. et al.
The ASTRODEEP Frontier Fields catalogues. I. Multiwavelength photom-etry of Abell-2744 and MACS-J0416.
Astron. Astrophys. , A30 (2016). .29. Castellano, M. et al.
The ASTRODEEP Frontier Fields catalogues. II. Photometric redshiftsand rest frame properties in Abell-2744 and MACS-J0416.
Astron. Astrophys. , A31(2016). .30. Di Criscienzo, M. et al.
The ASTRODEEP Frontier Fields catalogues. III. Multiwavelengthphotometry and rest-frame properties of MACS-J0717 and MACS-J1149.
Astron. Astrophys. , A30 (2017). .31. Santini, P. et al.
The Star Formation Main Sequence in the Hubble Space Telescope FrontierFields.
Astrophys. J. , 76 (2017). .32. Shipley, H. V. et al.
HFF-DeepSpace Photometric Catalogs of the 12 Hubble Frontier Fields,Clusters, and Parallels: Photometry, Photometric Redshifts, and Stellar Masses.
Astrophys.J. Suppl. , 14 (2018). .33. Bradaˇc, M. et al.
Hubble Frontier Field photometric catalogues of Abell 370 and RXCJ2248.7-4431: multiwavelength photometry, photometric redshifts, and stellar properties.
Mon. Not. R. Astron. Soc. , 99–107 (2019). .34. Lam, D. et al.
Detection of a Lensed z ≈
11 Galaxy in the Rest-Optical with Spitzer/IRAC andthe Inferred SFR, Stellar Mass, and Physical Size. arXiv e-prints arXiv:1903.08177 (2019). . 175. Yan, H. et al.
The Stellar Masses and Star Formation Histories of Galaxies at z ∼ : Con-straints from Spitzer Observations in the Great Observatories Origins Deep Survey. TheAstrophysical Journal , 24 (2006).36. Capak, P. et al.
Spectroscopy of Luminous z > Galaxy Candidates and Sources of Con-tamination in z > Galaxy Searches.
Astrophys. J. , 68 (2011). .37. Labbe, I. et al.
The Spectral Energy Distributions of z ∼ Galaxies from the IRAC UltraDeep Fields: Emission Lines, Stellar Masses, and Specific Star Formation Rates at 650 Myr.
Astrophys. J. Lett. , L19 (2013). .38. Finkelstein, S. L. et al.
A galaxy rapidly forming stars 700 million years after the Big Bangat redshift 7.51.
Nature , 524–527 (2013).39. Smit, R. et al.
EVIDENCE FOR UBIQUITOUS HIGH-EQUIVALENT-WIDTH NEBU-LAR EMISSION IN z ∼ GALAXIES: TOWARD A CLEAN MEASUREMENT OF THESPECIFIC STAR-FORMATION RATE USING A SAMPLE OF BRIGHT, MAGNIFIEDGALAXIES.
The Astrophysical Journal , 58 (2014).40. De Barros, S., Schaerer, D. & Stark, D. P. Properties of z 3–6 Lyman break galaxies.
A&A , A81 (2014).41. Roberts-Borsani, G. W. et al. z (cid:38) Astrophys. J. , 143 (2016). .182. Jiang, L. et al.
Physical Properties of Spectroscopically Confirmed Galaxies at z ≥
6. III.Stellar Populations from SED Modeling with Secure Ly α Emission and Redshifts.
Astrophys.J. , 16 (2016). .43. De Barros, S. et al.
The GREATS H β + [O III] luminosity function and galaxy propertiesat z ∼ : walking the way of JWST. Mon. Not. R. Astron. Soc. , 2355–2366 (2019). .44. Shim, H. et al. z ∼ H α Emitters in the Great Observatories Origins Deep Survey: Tracingthe Dominant Mode for Growth of Galaxies.
Astrophys. J. , 69 (2011). .45. Smit, R. et al.
High-precision Photometric Redshifts from Spitzer/IRAC: Extreme [3.6] -[4.5] Colors Identify Galaxies in the Redshift Range z ∼ . − . . Astrophys. J. , 122(2015). .46. Huang, K.-H. et al.
Detection of Lyman-alpha Emission from a Triply Imaged z = 6.85Galaxy behind MACS J2129.4 − Astrophys. J. Lett. , L14 (2016). .47. Laporte, N. et al.
The first Frontier Fields cluster: 4.5 µ m excess in a z ∼ galaxy candidatein Abell 2744. Astron. Astrophys. , L8 (2014). .48. Bridge, J. S. et al.
The Super Eight Galaxies: Properties of a Sample of Very Bright Galaxiesat < z < . Astrophys. J. , 42 (2019). .49. Katz, H., Laporte, N., Ellis, R. S., Devriendt, J. & Slyz, A. Probing cosmic dawn: modellingthe assembly history, SEDs, and dust content of selected z ∼ Mon. Not. R. Astron.Soc. , 4054–4068 (2019). .190. Houck, J. R. et al.
The Infrared Spectrograph (IRS) on the Spitzer Space Telescope.
Astro-phys. J. Suppl. , 18–24 (2004). astro-ph/0406167 .51. Teplitz, H. I. et al.
Measuring PAH Emission in Ultradeep Spitzer IRS Spectroscopy ofHigh-Redshift IR-Luminous Galaxies.
Astrophys. J. , 941–949 (2007). astro-ph/0701409 .52. Pope, A. et al.
Mid-Infrared Spectral Diagnosis of Submillimeter Galaxies.
Astrophys. J. , 1171–1193 (2008). .53. Hashimoto, T. et al.
The onset of star formation 250 million years after the Big Bang.
Nature , 392–395 (2018). .54. Laporte, N. et al.
Dust in the Reionization Era: ALMA Observations of a z = 8.38 Gravita-tionally Lensed Galaxy.
Astrophys. J. Lett. , L21 (2017). .55. Zheng, W. et al.
A magnified young galaxy from about 500 million years after the Big Bang.
Nature , 406–408 (2012). .56. Zheng, W. et al.
Young Galaxy Candidates in the Hubble Frontier Fields. IV. MACSJ1149.5+2223.
Astrophys. J. , 210 (2017). .57. Hoag, A. et al.
Spectroscopic confirmation of an ultra-faint galaxy at the epoch of reioniza-tion.
Nature Astronomy , 0091 (2017). .58. Oesch, P. A. et al. A Remarkably Luminous Galaxy at z=11.1 Measured with Hubble SpaceTelescope Grism Spectroscopy.
Astrophys. J. , 129 (2016). .209. Oesch, P. A. et al.
The Most Luminous z ˜9-10 Galaxy Candidates Yet Found: The Luminos-ity Function, Cosmic Star-formation Rate, and the First Mass Density Estimate at 500 Myr.
Astrophys. J. , 108 (2014). .60. Dayal, P. & Ferrara, A. Early galaxy formation and its large-scale effects.
Phys. Rep. ,1–64 (2018). .61. Du, X. et al.
Searching for z > . Analogs Near the Peak of Cosmic Star Formation.
Astrophys. J. , 65 (2020). .62. Li, Z. & Malkan, M. A. Extreme [O III] Emitters at z ∼ . . Astrophys. J. , 83 (2018). .63. Madau, P. & Dickinson, M. Cosmic Star-Formation History.
Annu. Rev. Astron. Astrophys. , 415–486 (2014). .64. Madau, P. et al. High-redshift galaxies in the Hubble Deep Field: colour selection and starformation history to z˜4.
Mon. Not. R. Astron. Soc. , 1388–1404 (1996). astro-ph/9607172 .65. Madau, P., Pozzetti, L. & Dickinson, M. The Star Formation History of Field Galaxies.
Astrophys. J. , 106–116 (1998). astro-ph/9708220 .66. Lilly, S. J., Le Fevre, O., Hammer, F. & Crampton, D. The Canada-France Redshift Survey:The Luminosity Density and Star Formation History of the Universe to Z approximately 1.
Astrophys. J. Lett. , L1 (1996). astro-ph/9601050 .217. Le Floc’h, E. et al.
Infrared Luminosity Functions from the Chandra Deep Field-South: TheSpitzer View on the History of Dusty Star Formation at (cid:46) z (cid:46) . Astrophys. J. ,169–190 (2005). astro-ph/0506462 .68. Magnelli, B. et al.
Evolution of the dusty infrared luminosity function from z = 0 to z = 2.3using observations from Spitzer.
Astron. Astrophys. , A35 (2011). .69. Davidzon, I. et al.
The COSMOS2015 galaxy stellar mass function . Thirteen billion years ofstellar mass assembly in ten snapshots.
Astron. Astrophys. , A70 (2017). .70. Song, M. et al.
The Evolution of the Galaxy Stellar Mass Function at z = 4-8: A SteepeningLow-mass-end Slope with Increasing Redshift.
Astrophys. J. , 5 (2016). .71. Grazian, A. et al.
The galaxy stellar mass function at 3.5 ≤ z ≤ Astron. Astrophys. , A96 (2015). .72. Caputi, K. I. et al.
Spitzer Bright, UltraVISTA Faint Sources in COSMOS: The Contributionto the Overall Population of Massive Galaxies at z = 3-7.
Astrophys. J. , 73 (2015). .73. Salmon, B. et al.
The Relation between Star Formation Rate and Stellar Mass for Galaxiesat . ≤ z ≤ . in CANDELS. Astrophys. J. , 183 (2015). .74. Steinhardt, C. L. et al.
Star Formation at < z < from the Spitzer Large Area Survey withHyper-Suprime-Cam (SPLASH). Astrophys. J. Lett. , L25 (2014). .75. Ashby, M. L. N. et al.
Spitzer Matching Survey of the UltraVISTA Ultra-deep Stripes22SMUVS): Full-mission IRAC Mosaics and Catalogs.
Astrophys. J. Suppl. , 39 (2018). .76. Caputi, K. I. et al.
Star Formation in Galaxies at z ∼ α Emitters on the SFR-M* Plane.
Astrophys. J. , 45 (2017). .77. Deshmukh, S. et al.
The Spitzer Matching Survey of the UltraVISTA Ultra-deep Stripes(SMUVS): The Evolution of Dusty and Nondusty Galaxies with Stellar Mass at z = 2-6.
Astrophys. J. , 166 (2018). .78. Guzm´an, R. et al.
The Nature of Compact Galaxies in the Hubble Deep Field. II. Spectro-scopic Properties and Implications for the Evolution of the Star Formation Rate Density ofthe Universe1,.
Astrophys. J. , 559–572 (1997). astro-ph/9704001 .79. Stark, D. P. et al.
The Evolutionary History of Lyman Break Galaxies Between Redshift 4and 6: Observing Successive Generations of Massive Galaxies in Formation.
Astrophys. J. , 1493–1511 (2009). .80. Gonzalez, A. H., Papovich, C., Bradaˇc, M. & Jones, C. Spectroscopic Confirmation of a z= 2.79 Multiply Imaged Luminous Infrared Galaxy Behind the Bullet Cluster.
Astrophys. J. , 245–251 (2010).81. Schaerer, D. & de Barros, S. The impact of nebular emission on the ages of z ∼ galaxies. Astron. Astrophys. , 423–426 (2009). .232. Davidzon, I., Ilbert, O., Faisst, A. L., Sparre, M. & Capak, P. L. An Alternate Approachto Measure Specific Star Formation Rates at < z < . Astrophys. J. , 107 (2018). .83. Tasca, L. A. M. et al.
Evidence for major mergers of galaxies at (cid:46) z < in the VVDS andVUDS surveys. Astron. Astrophys. , A10 (2014). .84. de Barros, S., Schaerer, D. & Stark, D. P. Properties of z ∼ − Lyman break galaxies.II. Impact of nebular emission at high redshift.
Astron. Astrophys. , A81 (2014). .85. Duncan, K. et al.
The mass evolution of the first galaxies: stellar mass functions and starformation rates at 4 < z < Mon. Not. R. Astron.Soc. , 2960–2984 (2014). .86. Gonz´alez, V. et al.
Slow Evolution of the Specific Star Formation Rate at z >
2: The Impactof Dust, Emission Lines, and a Rising Star Formation History.
Astrophys. J. , 34 (2014). .87. Stark, D. P. et al.
Keck Spectroscopy of < z < Faint Lyman Break Galaxies: The Im-portance of Nebular Emission in Understanding the Specific Star Formation Rate and StellarMass Density.
Astrophys. J. , 129 (2013). .88. Moustakas, J., Kennicutt, R. C., Jr. & Tremonti, C. A. Optical Star Formation Rate Indicators.
Astrophys. J. , 775–796 (2006). astro-ph/0511730 .249. Smit, R. et al.
Evidence for Ubiquitous High-equivalent-width Nebular Emission in z ∼ Galaxies: Toward a Clean Measurement of the Specific Star-formation Rate Using a Sampleof Bright, Magnified Galaxies.
Astrophys. J. , 58 (2014). .90. Bouwens, R. J. et al.
The Lyman-Continuum Photon Production Efficiency ξ ion of z ∼ − Galaxies from IRAC-based H α Measurements: Implications for the Escape Fraction andCosmic Reionization.
Astrophys. J. , 176 (2016). .91. M´armol-Queralt´o, E. et al.
The evolution of the equivalent width of the H α emission line andspecific star formation rate in star-forming galaxies at 1 < z < 5. Mon. Not. R. Astron.Soc. , 3587–3597 (2016). .92. Faisst, A. L. et al.
A Coherent Study of Emission Lines from Broadband Photometry: Spe-cific Star Formation Rates and [O iii]/H β Ratio at < z < . Astrophys. J. , 122 (2016). .93. Faisst, A. L., Capak, P. L., Emami, N., Tacchella, S. & Larson, K. L. The Recent Burstinessof Star Formation in Galaxies at z ∼ . from H α Measurements.
Astrophys. J. , 133(2019). .94. Lam, D. et al.
The mean H α EW and Lyman-continuum photon production efficiency forfaint z ≈ Astron. Astrophys. , A164 (2019). .95. Dressler, A. et al.
A Spectroscopic Catalog of 10 Distant Rich Clusters of Galaxies.
Astro-phys. J. Suppl. , 51–80 (1999). arXiv:astro-ph/9901263 .256. Poggianti, B. M. et al.
The Star Formation Histories of Galaxies in Distant Clusters.
Astro-phys. J. , 576–593 (1999). astro-ph/9901264 .97. Rosati, P., Borgani, S. & Norman, C. The Evolution of X-ray Clusters of Galaxies.
Annu.Rev. Astron. Astrophys. , 539 (2002).98. Lemaux, B. C. et al. The VIMOS Ultra-Deep Survey: Emerging from the dark, a massiveproto-cluster at z 4.57.
Astron. Astrophys. , A77 (2018). .99. Capak, P. L. et al.
A massive protocluster of galaxies at a redshift of z˜5.3.
Nature ,233–235 (2011). .100. Galametz, A. et al.
The Mid-infrared Environments of High-redshift Radio Galaxies.
Astro-phys. J. , 169 (2012). .101. Noirot, G. et al.
HST Grism Confirmation of 16 Structures at . < z < . from the ClustersAround Radio-Loud AGN (CARLA) Survey. Astrophys. J. , 38 (2018). .102. Stanford, S. A. et al.
An IR-selected Galaxy Cluster at z = 1.41.
Astrophys. J. Lett. ,L129–L132 (2005). astro-ph/0510655 .103. Stanford, S. A. et al.
The XMM Cluster Survey: A Massive Galaxy Cluster at z = 1.45.
Astrophys. J. Lett. , L13–L16 (2006). astro-ph/0606075 .104. Eisenhardt, P. R. M. et al.
Clusters of Galaxies in the First Half of the Universe from theIRAC Shallow Survey.
Astrophys. J. , 905–932 (2008). .2605. Papovich, C. et al.
A Spitzer-selected Galaxy Cluster at z = 1.62.
Astrophys. J. , 1503–1513 (2010). .106. Gonzalez, A. H. et al.
The Massive and Distant Clusters of WISE Survey. I. Survey Overviewand a Catalog of > Galaxy Clusters at z (cid:39) . Astrophys. J. Suppl. , 33 (2019). .107. Wilson, G. et al.
Spectroscopic Confirmation of a Massive Red-Sequence-Selected GalaxyCluster at z = 1.34 in the SpARCS-South Cluster Survey.
Astrophys. J. , 1943–1950(2009). .108. Muzzin, A. et al.
Spectroscopic Confirmation of Two Massive Red-Sequence-SelectedGalaxy Clusters at z ∼ . in the SpARCS-North Cluster Survey. Astrophys. J. , 1934–1942 (2009). .109. Sawicki, M. The 1.6 Micron Bump as a Photometric Redshift Indicator.
Astron. J. ,3050–3060 (2002). astro-ph/0209437 .110. Stanford, S. A. et al.
IDCS J1426.5+3508: Discovery of a Massive, Infrared-selected GalaxyCluster at z = 1.75.
Astrophys. J. , 164 (2012). .111. Rettura, A. et al.
Candidate Clusters of Galaxies at z > . Identified in the Spitzer SouthPole Telescope Deep Field Survey.
Astrophys. J. , 109 (2014). .112. Ashby, M. L. N. et al.
SEDS: The Spitzer Extended Deep Survey. Survey Design, Photome-try, and Deep IRAC Source Counts.
Astrophys. J. , 80 (2013).2713. Brodwin, M. et al.
The Era of Star Formation in Galaxy Clusters.
Astrophys. J. , 138(2013). .114. Nantais, J. B. et al.
Stellar mass function of cluster galaxies at z ∼ . : evidence for reducedquenching efficiency at high redshift. Astron. Astrophys. , A161 (2016). .115. Bruzual, G. & Charlot, S. Stellar population synthesis at the resolution of 2003.
Mon. Not.R. Astron. Soc. , 1000–1028 (2003). arXiv:astro-ph/0309134 .28 -2 -1 Survey Area (deg ) . µ m D ep t h ( M A B , σ ) SWIREFLS SHELA SSDFSPIESSDWFS SERVS+DDE/WFIRSTSIMPLE S-COSMOSSpUDSE-CDFS SPLASHSEDSSMUVSEGSS-CANDELSGOODSNEPUDF GREATS D ep t h Area
SURFSUPHFFSRELICS-DEEP SRELICSICLASHIRACLS
Magniffication µ = 3 L* (z=4)L* (z=9)Area needed to detect1 high-z massive cluster
Spitzer Surveys
Figure 1:
Sensitivities and area of Spitzer Surveys.
The . µ m σ point-source sensitivitiesvs area surveyed for surveys executed during the Spitzer mission. Recent
Spitzer
Surveys havereached the depths that allow us to push observations to observe typical ( L ∗ ) galaxies at e.g., z > and z > (horizontal grey lines). Plotted are depths and areas for both field (green) and lensing(purple) surveys as symbols. The lensing surveys have a typical magnifications of µ ∼ , the effectof such a magnification (increased depth and reduced area) are indicated with arrows. The typicalarea needed to detect ∼ high-redshift ( z (cid:38) ) massive cluster ( M > M (cid:12) ) is indicated by thevertical grey line. Adapted from [75]. 29 [ ] M ( A B ) IRAC3.6 m IRAC4.5 mWFC3 = . , == . , == , == . , == . , = z=8 [ ]
Figure 2:
Spitzer data is crucial for determining stellar ages.
Five different spectral energy dis-tributions (SEDs) for starburst galaxies of different ages (from 115) redshifted to z = 8 . Whereasall these galaxies have similar colors in HST/WFC3 bands (blue shaded region) relative to typicalphotometric uncertainties, the different ages can be easily distinguished once . µ m and . µ m Spitzer /IRAC imaging is added (the . µ m and . µ m IRAC bands are shown with the light anddark red shaded regions), as their m H − m . µ m and m H − m . µ m colors are very different.We plot a Z = 0 . Z (cid:12) starburst galaxy SED with stellar population at t = 290 Myr after the burst(green),
Myr (orange), Myr (pink) and Myr (yellow). Also shown is a t = 100 Myr SEDwith Z = Z (cid:12) (purple), to show the effect of metallicity degeneracy with age, which is small.30igure 3: High-redshift galaxies detected in
Spitzer images. (a) Color composite image of thecentral deepest region of the GREATS (GOODS Re-ionization Era wide-Area Treasury). A subsetof z ∼ galaxies from [43] are circled in red. (b) An example of such a galaxy is shown in theinset. Image credit: NASA/JPL-Caltech/ESA/Spitzer/P. Oesch/S. De Barros/ I.Labbe.31 Redshift + [ ] [ ] [ ] [ ] [ ]
Figure 4:
Spitzer data can be used to measure (combined) strengths of several prominentoptical lines.
Lines visible in different
Spitzer channels are plotted as a function of redshift.32igure 5:
MACS1149-JD is the best example with an evidence of old stellar population at z ∼ . a Spitzer image of MACS1149-JD from SURFS UP survey. MACS1149-JD was first dis-covered by [55] and first detected in both . µ m and . µ m in SURFS UP survey ( 26, image from Spitzer press release feature14-13). b SED modeling of the object.
SED modeling provided anevidence of old stellar population once the redshift was determined to be at z = 9 . ± . . bI In particular,
Spitzer fluxes can not be fit without an inclusion of old ( +190 − Myr old) stellar pop-ulation. Black squares show (from left to right) F125W, F140W and F160W data from HST, a σ upper limit for the Ksband from VLT/HAWK-I, and . µ m and . µ m fluxes from Spitzer /IRAC.The red solid line indicates the SED model and the corresponding magnitudes are shown by yel-low crosses. Blue and black lines represent the contributions from the young ( +2 − Myr) and oldpopulations, respectively. bII , The black square is the observed [O Roman3] 88 µ m emission lineflux and its 1 σ uncertainty, while the yellow cross indicates the model prediction. bIII , The blacksquare shows the 2 σ upper limit for the dust continuum flux density, and the yellow cross indicatesthe model prediction. [53]. 33igure 6: Spitzer data has been very effective in searching for and characterizing z > galaxy clusters and protoclusters. a Examples of high-redshift clusters from SpARCS survey .Tri-color gz[3.6] images of the central regions of the four southern z > . SpARCS clusters.Spectroscopic cluster members are marked with white squares [114]. b Image of a proto-clusterof galaxies at z = 5 . . The cluster was discovered by a suite of multi-wavelength observationsmade with