On contamination and completeness in z≥ 5 Lyman-break galaxy surveys

A large population of z > 5 Lyman-break galaxies has been identified in recent years. However, the high-redshift galaxies selected by different surveys are subject to a variety of selection effects – some overt, others more subtle. We present an analysis of sample completeness and contamination issues in high-redshift surveys, focusing on surveys at z ≈ 5 and using a spectroscopically confirmed low-redshift sample from the DEEP2 survey in order to characterize contaminant galaxies. We find that most surveys underestimate their contamination from highly clustered galaxies at z ≈ 1 and stars. We consider the consequences of this for both the rest-frame ultraviolet luminosity function and the clustering signal from z ≈ 5 galaxies. We also find that sources with moderate strength Lyman α emission lines can be omitted from dropout surveys due to their blue colours, again effecting the derived luminosity functions. We discuss the points of comparison between different samples, and the applicability of survey-specific results to the population at z > 5 in general.     

The SCUBA Half-Degree Extragalactic Survey (SHADES) – VIII. The nature of faint submillimetre galaxies in SHADES, SWIRE and SXDF surveys

We present the optical-to-submillimetre spectral energy distributions (SEDs) for 33 radio and mid-infrared (mid-IR) identified submillimetre galaxies discovered via the SHADES 850-μm SCUBA imaging in the Subaru- XMM Deep Field (SXDF). Optical data for the sources come from the SXDF and mid- and far-IR fluxes from SWIRE. We obtain photometric redshift estimates for our sources using optical and IRAC 3.6- and 4.5-μm fluxes. We then fit SED templates to the longer wavelength data to determine the nature of the far-IR emission that dominates the bolometric luminosity of these sources. The IR template fits are also used to resolve ambiguous identifications and cases of redshift aliasing. The redshift distribution obtained broadly matches previous results for submillimetre sources and on the SHADES SXDF field. Our template fitting finds that active galactic nuclei, while present in about 10 per cent of our sources, do not contribute significantly to their bolometric luminosity. Dust heating by starbursts, with either Arp220 or M82 type SEDs, appears to be responsible for the luminosity in most sources (23/33 are fitted by Arp220 templates, 2/33 by the warmer M82 templates). 8/33 sources, in contrast, are fitted by a cooler cirrus dust template, suggesting that cold dust has a role in some of these highly luminous objects. Three of our sources appear to have multiple identifications or components at the same redshift, but we find no statistical evidence that close associations are common among our SHADES sources. Examination of rest-frame K -band luminosity suggests that 'downsizing' is underway in the submillimetre galaxy population, with lower redshift systems lying in lower mass host galaxies. Of our 33 identifications six are found to be of lower reliability but their exclusion would not significantly alter our conclusions.     

The Millennium Galaxy Catalogue: 16 ≤BMGC < 24 galaxy counts and the calibration of the local galaxy luminosity function

The Millennium Galaxy Catalogue (MGC) is a 37.5 deg², medium-deep, B -band imaging survey along the celestial equator, taken with the Wide Field Camera on the Isaac Newton Telescope. The survey region is contained within the regions of both the Two Degree Field Galaxy Redshift Survey (2dFGRS) and the Sloan Digital Sky Survey Early Data Release (SDSS-EDR). The survey has a uniform isophotal detection limit of 26 mag arcsec⁻² and it provides a robust, well-defined catalogue of stars and galaxies in the range  16 ≤ B _MGC < 24 mag  .
Here we describe the survey strategy, the photometric and astrometric calibration, source detection and analysis, and present the galaxy number counts that connect the bright and faint galaxy populations within a single survey. We argue that these counts represent the state of the art and use them to constrain the normalizations (φ*) of a number of recent estimates of the local galaxy luminosity function. We find that the 2dFGRS, SDSS Commissioning Data (CD), ESO Slice Project, Century Survey, Durham/UKST, Mt Stromlo/APM, SSRS2 and NOG luminosity functions require a revision of their published φ* values by factors of  1.05 ± 0.05, 0.76 ± 0.10, 1.02 ± 0.22, 1.02 ± 0.16, 1.16 ± 0.28, 1.75 ± 0.37, 1.40 ± 0.26  and  1.01 ± 0.39  , respectively. After renormalizing the galaxy luminosity functions we find a mean local b _J luminosity density of     . ¹     

The cluster galaxy luminosity function at z= 0.3: a recent origin for the faint-end upturn?

We derive deep luminosity functions (LFs) (to   M _z =−15  ) for galaxies in Abell 1835  ( z = 0.25)  and AC 114  ( z = 0.31)  , and compare these with the local z ' LF for 69 clusters. The data show that the faint-end upturn, the excess of galaxies above a single Schechter function at   M _z < −17  , does not exist in the higher redshift clusters. This suggests that the faint-end upturn galaxies have been created recently, by infall into clusters of star-forming field populations or via tidal disruption of brighter objects.     

The galaxy luminosity function in clusters and the field

We present the results from a CCD survey of the B -band luminosity function of nine clusters of galaxies, and compare them to published photographic luminosity functions of nearby poor clusters like Virgo and Fornax, and also to the field luminosity function. We derive a composite luminosity function by taking the weighted mean of all the individual cluster luminosity functions; this composite luminosity function is steep at bright and faint magnitudes and is shallow in-between.
All clusters have luminosity functions consistent with this single composite function. This is true both for rich clusters like Coma and for poor clusters like Virgo.
This same composite function is also individually consistent with the deep field luminosity functions found by Cowie et al. and Ellis et al., and also with the faint end of the Las Campanas Redshift Survey R -band luminosity function, shifted by 1.5 mag. A comparison with the Loveday et al. field luminosity function, which is well determined at the bright end, shows that the composite function, which fits the field data well fainter than M _B=−19, drops too steeply between M _B=−19 and −22 to fit the field data there.     

The galaxy luminosity function from MR=−25 to MR=−9

Redshift surveys such as the Sloan Digital Sky Survey (SDSS) have given a very precise measurement of the galaxy luminosity function down to about   M_R =−17 (≈ M_B =−16)  . Fainter absolute magnitudes cannot be probed because of the flux limit required for spectroscopy. Wide-field surveys of nearby groups using mosaic CCDs on large telescopes are able to reach much fainter absolute magnitudes, about   M_R =−10  . These diffuse, spiral-rich groups are thought to be typical environments for galaxies, so their luminosity functions should be the same as the field luminosity function. The luminosity function of the groups at the bright end  ( M_R < −17)  is limited by Poisson statistics and is far less precise than that derived from redshift surveys. Here we combine the results of the SDSS and the surveys of nearby groups, and we supplement the results with studies of Local Group galaxies in order to determine the galaxy luminosity function over the entire range  −25 < M_R < −9  . The average logarithmic slope of the field luminosity function between   M_R =−19  and   M_R =−9  is  α=−1.26  , although a single power law is a poor fit to the data over the entire magnitude range. We also determine the luminosity function of galaxy clusters and demonstrate that it is different from the field luminosity function at a high level of significance; there are many more dwarf galaxies in clusters than in the field, due to a rise in the cluster luminosity function of  α∼−1.6  between   M_R =−17  and   M_R =−14  .     

The properties of Lyα emitting galaxies in hierarchical galaxy formation models

We present detailed predictions for the properties of Lyα-emitting galaxies in the framework of the Λ cold dark matter cosmology, calculated using the semi-analytical galaxy formation model galform . We explore a model that assumes a top-heavy initial mass function in starbursts and that has previously been shown to explain the sub-millimetre number counts and the luminosity function of Lyman-break galaxies at high redshift. We show that this model, with the simple assumption that a fixed fraction of Lyα photons escape from each galaxy, is remarkably successful at explaining the observed luminosity function of Lyα emitters (LAEs) over the redshift range  3 < z < 6.6  . We also examine the distribution of Lyα equivalent widths and the broad-band continuum magnitudes of emitters, which are in good agreement with the available observations. We look more deeply into the nature of LAEs, presenting predictions for fundamental properties such as the stellar mass and radius of the emitting galaxy and the mass of the host dark matter halo. The model predicts that the clustering of LAEs at high redshifts should be strongly biased relative to the dark matter, in agreement with observational estimates. We also present predictions for the luminosity function of LAEs at   z > 7  , a redshift range that is starting to be be probed by near-infrared surveys and using new instruments such as the Dark Ages Z Lyman Explorer (DAzLE).     

The SCUBA HAlf Degree Extragalactic Survey (SHADES) – VII. Optical/IR photometry and stellar masses of submillimetre galaxies

We present estimates of the photometric redshifts, stellar masses and star formation histories of sources in the Submillimetre Common-User Bolometer Array (SCUBA) HAlf Degree Extragalactic Survey (SHADES). This paper describes the 60 SCUBA sources detected in the Lockman Hole covering an area of ∼320 arcmin². Using photometry spanning the B band to 8 μm, we find that the average SCUBA source forms a significant fraction of its stars in an early period of star formation and that most of the remainder forms in a shorter more intense burst around the redshift it is observed. This trend does not vary significantly with source redshift. However, the sources show a clear increase in stellar mass with redshift, consistent with downsizing. In terms of spectral energy distribution types, only two out of the 51 sources we have obtained photometric redshifts for are best fitted by a quasar-like spectrum, with approximately 80 per cent of the sources being best fitted with late-type spectra (Sc, Im and starburst). By including photometry at 850 μm, we conclude that the average SCUBA source is forming stars at a rate somewhere between 6 and 30 times the rate implied from the rest-frame optical in a dust obscured burst and that this burst creates 15–65 per cent of the total stellar mass. Using a simplistic calculation, we estimate from the average star formation history that between one in five and one in 15 bright  ( L _*+ 2 < L _optical < L _*− 1 mag)  galaxies in the field over the interval  0 < z < 3  will at some point in their lifetime experience a similar energetic dusty burst of star formation. Finally, we compute the evolution of the star formation rate density and find it peaks around   z ∼ 2  .     

A census of metals at high and low redshift and the connection between submillimetre sources and spheroid formation

Deep surveys in many wavebands have shown that the rate at which stars were forming was at least a factor of 10 higher at redshifts >1 than today. Heavy elements ('metals') are produced by stars, and the star formation history deduced by these surveys implies that a significant fraction of all metals in the Universe today should already exist at   z ∼ 2–3  . However, only 10 per cent of the total metals expected to exist at this redshift have so far been accounted for (in damped Lyman α absorbers and the Lyman forest). In this paper, we use the results of submillimetre surveys of the local and high-redshift Universe to show that there was much more dust in galaxies in the past. We find that a large proportion of the missing metals are traced by this dust, bringing the metals implied from the star formation history and observations into agreement. We also show that the observed distribution of dust masses at high redshift can be reproduced remarkably well by a simple model for the evolution of dust in spheroids, suggesting that the descendants of the dusty galaxies found in deep submillimetre surveys are the relatively dust-free spiral bulges and ellipticals in the Universe today.