Mg ii absorption systems and their neighbouring galaxies from a background-subtraction technique

Abstract

We estimate the absolute magnitude distribution and luminosity function of galaxies which lie within a few hundred kiloparsecs of Mg II absorption systems. The absorption systems themselves lie along 1880 lines of sight to Sloan Digital Sky Survey (SDSS) Data Release 3 quasi-stellar objects (QSOs), have rest equivalent widths greater than 0.8 A and redshifts that range from 0.37 ≤ z ≤0.82. Our measurement is based on all galaxies identified by the SDSS photometric pipeline which lie within a projected comoving distance of about 1 h -1 Mpc of each QSO demonstrating absorption. The redshifts of these projected neighbours are not available, so we use a background-subtraction technique to estimate the absolute magnitude distribution and luminosity function of true neighbours. Our method exploits the fact that although we do not know the redshifts of the neighbours, we do know the redshift of the absorbers. The luminosity function of Mg II neighbours isolated by our method is well approximated by that of earlier type galaxies (Es-Sa's) at these redshifts. However, we note that the SDSS magnitude limit of r ∼ 22 means that we are not sensitive to lower luminosities L < 0.56L * where later types dominate; thus, we do not readily detect the later type neighbours of these Mg II systems. The number of neighbours we find between 20 and 100 h -1 kpc is only about 20 per cent of the number of absorbers in our sample. Accounting for absorbers within 20 h -1 kpc will increase this number slightly, indicating that at least 70-75 per cent of the absorber host galaxies are fainter than 0.56L * ; they may well be of a later type. We carry out simulations which indicate that the actual fraction of absorber host galaxies which were too faint for our survey may be even higher. When the sample is divided into half on the basis of rest-frame equivalent width (at REW = 1.28 A), we find that weaker systems have, within 0.02-1 h -1 Mpc, 1.5 times as many neighbours as do stronger systems and that these neighbours tend to be more luminous than the neighbours of stronger systems. However, on scales smaller than 50 h -1 kpc it is the stronger systems that have more neighbours; this suggests that stronger systems tend to be closer to their host galaxy's centre than weaker systems. We provide an analytic description of the method which helps build intuition and show that it is generally applicable to any data set in which redshifts are only available for only a small sub-sample of objects. Hence, we expect it to aid in the analysis of galaxy near-environments based on photometric redshift data sets. In this respect, if Mg II systems are not in particularly unusual environments, then our analysis represents a determination of the bright end of the z ∼ 0.7 galaxy luminosity function from SDSS data alone.