Ju l 1 99 7 Deep Near - Infrared Surveys — Understanding Galaxy

Deep near-infrared (NIR) surveys are critical to our current, and even more to our future, understanding of galaxy evolution in the early universe. In this review I will be discussing the relevance of deep NIR surveys and looking at the information provided by different types of survey: broad-band imaging, spectroscopic observations and narrow-band imaging. In particular I will be looking at the future possibilities for faint galaxy work provided by forthcoming, innovative, NIR instrumentation being developed for the next generation of 8 m telescopes. The initial interest in the NIR for studies of galaxy evolution came from the realisation that at low redshift (z < 1) the NIR K-corrections for different morphological classes of galaxy are very similar (Figure 1). Thus by selecting in the NIR the morphological mix of the sample, i.e. the ratio of galaxies with young and old stellar populations, would be insensitive to redshift if there was no evolution. Thus any evolutionary signature would be more clearly seen. This led to many early NIR surveys at relatively bright magnitudes (e.g. [13],[15]) which confirmed a general picture of very little evolution in early-type galaxies along with a general increase in the space density of star-forming systems at low redshift. Such surveys are reviewed elsewhere in these Proceedings [14]. Here we will consider deeper surveys, with IR magnitudes typically K > 19–20 in which a significant number of galaxies lie beyond a redshift of unity. At these redshifts the prominent optical emission and absorption spectra features are shifted into the NIR. For example the [OII] and Ca H and K lines are in the J-band for z > 1.6 and for z > 2 the Hα line lies in the K-band. This has so far provided a natural limit for redshift surveys in the optical, the deepest