Using the 1.6um Bump to Study Rest-frame NIR Selected Galaxies at Redshift 2

We explore the feasibility and limitations of using the 1.6um bump as a photometric redshift indicator and selection technique and use it to study the rest-frame H-band galaxy luminosity and stellar mass functions at redshift z~2. We use publicly available Spitzer/IRAC images in the GOODS fields and find that color selection in the IRAC bandpasses alone is comparable in completeness and contamination to BzK selection. We find that the shape of the 1.6um bump is robust, and photometric redshifts are not greatly affected by choice of model parameters. Comparison with spectroscopic redshifts shows photometric redshifts to be reliable. We create a rest-frame NIR selected catalog of galaxies at z~2 and construct a galaxy stellar mass function (SMF). Comparisons with other SMFs at approximately the same redshift but determined using shorter wavelengths show good agreement. This agreement suggests that selection at bluer wavelengths does not miss a significant amount of stellar mass in passive galaxies. Comparison with SMFs at other redshifts shows evidence for the downsizing scenario of galaxy evolution. We conclude by pointing out the potential for using the 1.6um technique to select high-redshift galaxies with the JWST, whose lambda > 0.6 um coverage will not be well suited to selecting galaxies using techniques that require imaging at shorter wavelengths.