In this Chapter, we have reviewed how deep imaging is a fundamental tool in the study of the outermost structure of galaxies. Three main sources of imaging are currently used to detect and characterize the outskirts of galaxies: (1) surveys such as the Sloan Digital Sky Survey's Stripe 82 project, (2) very long exposures on small telescopes, including by amateurs, and (3) long exposures on the largest professional telescopes. The technical challenges in overcoming systematic effects are significant, and range from the treatment of light scattered by the atmosphere and the telescope and instrument, via flat fielding, to the accurate subtraction of non-galaxy light in the images. We have reviewed recent results on galaxy disks and haloes obtained with deep imaging, including the detection and characterization of thick disks, truncations of stellar disks, tidal streams, stellar haloes, and satellites. We have shown how each of these interrelated aspects of the faintest detectable structure in galaxies can shed light on the formation and subsequent evolution of the galaxies and our Universe.
The future is promising in terms of discovering the “low surface brightness Universe” through deep imaging. Current techniques using small and large ground-based telescopes, as reviewed in Sect. 3, will continue to be exploited while new facilities will become available. The most promising of these new facilities are poised to be the ground-based LSST, and the new space telescopes JWST and Euclid. LSST (Ivezic et al 2008) will repeatedly image the whole of the Southern sky in 6 passbands (ugrizy), with a planned start date for surveys of around 2021. While each 30 s exposure with the 8.2 m telescope and optimized camera will yield a depth of 24.7 r-mag (5-sigma point source depth) over the 9.6 square degree field of view, combining all imaging obtained over the approximately 10 year survey duration could yield a depth of 27.5 r-mag (5-sigma point source depth). This roughly means as deep as the current Stripe 82 in one exposure, and close to three magnitudes deeper over the whole survey duration. The scientific possibilities offered by this depth of imaging, over an area of sky of over 20000 deg2, are tremendous. All the science discussed in this paper could essentially be done with one or a few exposures, or easily with the data of say one hour of observation. This obviously depends critically on whether systematic effects, including but not limited to those discussed in Sect. 2, can be properly controlled, modelled, and corrected for. This is not trivial, and may hinder the full exploitation of the data to their theoretical limits.
The JWST, to be launched in 2018, will allow deep imaging but, when compared to, e.g., the HST, won't be quite as revolutionary as LSST. The field of view will be limited to just over 2×4 arcmin2 in the near-IR which will all but exclude deep imaging of the nearest galaxies. Where significant progress can be expected is in the deep imaging of galaxies at redshifts of, say, 0.2 and higher. In particular, near-IR imaging will allow the observation of galaxies at redshifts beyond 1 in rest-frame red passbands, necessary to reduce the effects of both young stellar populations and dust extinction.
Euclid will provide, among other data products, imaging in a very wide optical band (R + I + Z) over an area of 15000 deg2 to a depth of 24.5 mag (10σ for extended source). Compared to LSST, advantages of Euclid imaging will be its higher spatial resolution (of ∼ 0.2 arcsec, due to the relatively small telescope aperture of 1.2 m) and better-behaved PSF, thanks to the absence of the Earth atmosphere. A disadvantage is that the visual imaging is done through a very wide filter which excludes the use of colour information. The depth of imaging will be comparable with LSST, though, for Euclid's Deep Survey (over an area of around 40 square degree), which will allow the comparison of high-resolution Euclid imaging with the colour information obtained from the LSST images.
The area of ultra-deep imaging is still very much unexplored territory, and future work in this area will be stimulated by the availability of revolutionary new data sets and the continued understanding of the systematics affecting them. It will be a huge technical challenge to properly treat and analyze the upcoming deep imaging, in particular those from LSST and Euclid. As we enter previously unexplored territory with ultra-deep imaging, the systematic effects we know about will be challenging to characterize and correct for, and additional difficulties will almost certainly present themselves. But overcoming these issues will definitely pay off in terms of increased understanding of the formation and evolution processes which have led to the Universe and the galaxies as we observe them now. The future of imaging ultra-faint structures is very bright.
Acknowledgements JHK thanks Sébastien Comerón and Carme Gallart for comments on sections of the manuscript. IT has benefitted from multiple conversations with and the hard work of many members of his team. In particular, he thanks Jürgen Fliri, María Cebrián and Javier Román. JHK and IT acknowledge financial support from the Spanish Ministry of Economy and Competitiveness (MINECO) under grants number AYA2013-41243-P and AYA2013-48226-C3-1-P, respectively. JHK acknowledges financial support to the DAGAL network from the People Programme (Marie Curie Actions) of the European Unions Seventh Framework Programme FP7/2007-2013/ under REA grant agreement number PITN-GA-2011-289313.