Over the last decade we have witnessed a revolution in our knowledge of galaxies in the first billion years of cosmic time. Arguably the next 10 years should be even more exciting.
The accurate measurement of the bright end of the evolving galaxy UV luminosity function should soon be improved by combining the data over ≃ 0.2 deg2 from the HST CANDELS project (Grogin et al. 2011; Finkelstein et al. 2012; Oesch et al. 2012) with the brighter but even larger-area multi-colour imaging being produced by the new generation of ground-based near-infrared surveys, such as UltraVISTA (Fig. 27; Bowler et al. 2012; McCracken et al. 2012). Beyond this, the EUCLID satellite, as part of the "deep" component of its mission, is expected to survey several tens of square degrees down to J ≃ 26 mag. (Laureijs et al. 2011). Near-infrared narrow-band imaging surveys can also be expected to continue to expand in scope (e.g. ELVIS, at z ≃ 8.8; Nilsson et al. 2007). Crucially, these wide-area near-infrared imaging surveys will be complemented by well-matched wide-area optical imaging provided, for example, by Hyper-Suprime CAM on Subaru (Takada 2010).
Figure 27. A small sub-region of the first-year UltraVISTA near-infrared image of the COSMOS field, presented as a Y + J + Ks colour image (McCracken et al. 2012). This 1.5 deg2 image reaches Y + J ≃ 25 AB mag, and has already revealed ≃ 200,00 galaxies, of which ≃ 5 appear to be massive galaxies at z ≃ 7 (Bowler et al. 2012). The final UltraVISTA imaging should reach 5-10 times deeper, and enable the search for rare massive galaxies out to z ≃ 10 (courtesy UltraVISTA/Terapix/CNRS/CASU).
At the faint end, attempts will continue to exploit the power of WFC3/IR on HST to the full, with further ultra-deep near-infrared imaging planned over several square arcmin of sky, and at the end of the decade NIRCam on the JWST should extend this work out to z > 10. At the same time the angular resolution limitations of IRAC on Spitzer will be overcome with MIRI on JWST, which should deliver rest-frame optical imaging of the highest-redshift galaxies of a quality comparable to that currently achieved with WFC3 on HST (Fig. 28). This should enormously improve our knowledge of the rest-frame optical morphologies, and the stellar masses of the highest-redshift galaxies. It should also enable much more accurate measurement of the UV-optical SEDs of faint galaxies at z ≃ 5 - 10, including accurate determination of their UV slopes (with resulting implications for age, metallicity and escape fraction as discussed above).
Figure 28. HST WFC3 and Spitzer IRAC observations of a small sub-region of the HUDF field, alongside a simulated image of the same region as expected from MIRI on the JWST. The HST image is the product of 22 hours of on-source integration with WFC3 in the H160 band. The simulated 5.6 µm JWST image was produced using the H-band morphology and the IRAC fluxes as input, and assumes 28 hours of intergation with MIRI. The unconfused mid-infrared imaging which will be delivered by MIRI at IRAC wavelengths will enable the study of the rest-frame optical morphologies out beyond z ≃ 10, and will also allow much more robust determinations of the stellar masses of the most distant galaxies (courtesy A. Rogers).
Spectroscopic follow-up of the brighter z > 7 galaxies (perhaps down to J ≃ 28) is set to be transformed by the new generation of ground-based multi-object near-infrared spectrographs including FMOS on Subaru (Kimura M. et al., 2010), KMOS on the VLT (Sharples et al. 2006), and MOSFIRE on Keck (McLean et al. 2008), before this work should be extended to even fainter magnitudes with NIRSpec on JWST (Birkmann et al. 2011). This should clarify the currently confused picture of Lyman-α emission from LBGs, with important implications for our understanding of the progress of reionization. Wide-field near-infrared GRISM spectroscopy with EUCLID may enable the first meaningful study of the clustering of LAEs relative to LBGs at z > 7, where an enhancement of the apparent clustering of Lyman-α emitters is a prediction of some models for reionization (e.g. McQuinn et al. 2007). The next generation of giant ground-based near-infrared telescopes equipped with sophisticated adaptive-optics systems (TMT, E-ELT, GMT) will also enable detailed near-infrared high-resolution spectroscopy of the most distant galaxies.
Finally, in the rather near future, we can expect a revolution in the search for and study of galaxies at z > 5 at sub-millimetre wavelengths. We already know that the most distant quasars are detectable in the sub-mm, so we can anticipate that a significant population of rare high-mass dusty galaxies should be uncovered by combining existing Herschel SPIRE imaging with longer-wavelength data from the SCUBA-2 Cosmology Legacy Survey now underway at the JCMT. Crucially, detailed millimetre spectroscopy of such objects will be relatively straightforward with ALMA. Over the next few years ALMA can also be exploited to undertake the first sub-mm surveys of sufficient depth and angular-resolution to complement the Ultra Deep Field studies previously only possible at shorter wavelengths with HST. This work should enormously clarify our understanding of the role of dust and molecules at the highest redshifts, completing our census of cosmic star-formation history at early times, and transforming our understanding of the production-rate of the first metals in the Universe.
James Dunlop gratefully acknowledges the support of the Royal Society through a Wolfson Research Merit award and the support of the European Research Council via the award of an Advanced Grant. He also wishes to acknowledge the many substantial contributions of his collaborators in the study of high-redshift galaxies, and the (mostly) good-natured and productive rivalry engendered by the various competing groups working at this exciting research frontier.