The Star Formation and Nuclear Accretion Histories of Normal Galaxies in the AGES Survey

These expectations are largely borne out by observations. There are strong correlations between the hard (> 2 keV) X-ray luminosity (presumably from HMXBs) and other star-formation indicators, such as radio, far-IR, B-band, and UV flux for both local (e.g., Fabbiano 1989, David, Jones, & Forman 1992) and higher redshift (Seibert et al. 2002, Bauer et al. 2002) galaxy samples. Many studies have found a linear correlation between X-ray luminosity and the SFR (e.g., Ranalli et al. 2002, Grimm et al. 2003, Gilfanov et al. 2004, Persic et al. 2004, and Colbert et al. 2004). In particular, Grimm et al. (2003) found that the hard (2-10 keV) HMXB X-ray luminosity scales with the star formation rate according to L HMXB

x,hard ≃ 0.67 × 10 40 SFR M ⊙ yr -1 ergs s -1 .

(1)

In contrast, the (total band: 0.5-8 keV) X-ray emission from LMXBs is well-correlated with the K-band flux of galaxies, which in turn is a good tracer of stellar mass. Kim and Fabbiano (2004) find that

where L K * = ν K L νK , * = 2.6 × 10 43 ergs s -1 corresponds to M K * = -23.4 mag (Kochanek et al. 2001).

Because the stellar mass density evolves slowly between z = 0 and z ≃ 0.5 (e.g., Bell 2004 and references therein), we expect little change in the number of LMXBs and assume that Eqn. (2) holds out to z ≃ 0.5. On the other hand, the star formation rate density rises rapidly with look back time, (e.g., ∝ (1 + z) 2.7±0.7 Hogg 2001), so we should see a correspondingly rapid evolution in the X-ray flux of star-forming galaxies due to the increasing number of HMXBs.

The accretion luminosity density of AGN also rises rapidly with redshift (e.g., ∝ (1 + z) 3.2±0.8 , Barger et al. 2005) with higher densities of more luminous sources at higher redshifts (e.g., Barger et al. 2001, Cowie et al. 2003, Ueda et al. 2003, Miyaji 2004, Hasinger 2004, Hasinger, Miyaji, & Schmidt 2005). The observed trends suggest that faint AGN with characteristic X-ray luminosities approaching those of bright starbursts, L x < ∼ 10 41 ergs s -1 , may be 1 most abundant at z < ∼ 1 and, if so, they could produce a substantial fraction of the “normal” galaxy X-ray flux at these redshifts.

Given their similar evolution, spectral shapes (e.g., Ptak et al. 1999), and potentially similar luminosities, it may be difficult to disentangle the X-ray emission from AGN and star formation at z < ∼ 1. Probing the X-ray evolution of normal galaxies between z ≃ 0.1 and z ≃ 1 is also difficult because it requires relatively deep observations over relatively wide areas. In fact, most of our knowledge about the X-ray properties of galaxies still comes from either largearea, local surveys or high-redshift, small-volume deep fields. This is the second in a series of papers in which we attempt to bridge this gap by examining the X-ray evolution of galaxies within the 9.3 square degree Boötes field of the NOAO Deep Wide-Field Survey (NDWFS; Jannuzi & Dey 1999, Jannuzi et al., in preparation, Dey et al. in preparation). The XBoötes survey (Murray et al. 2005) obtained a 5 ksec Chandra mosaic of the entire 9.3 square degree field based on 126 ACIS-I pointings. These data are too shallow, however, to detect typical galaxies even at modest redshifts. Fortunately, the wide area of the survey allows us to measure the mean properties of a large, representative population of galaxies by “stacking” (averaging) their X-ray emission (Brandt et al. 2001;Nandra et al. 2002;Hornschemeier et al. 2002, Georgakakis et al. 2003, Lehmer et al. 2005, Laird et al. 2005, Laird et al. 2006, Lehmer et al. 2007).

In the first paper in the series, Brand et al. (2005; hereafter B05) employed the stacking technique to examine the X-ray luminosity evolution of ≈ 3300 optically luminous (∼ L * ), red galaxies with photometric redshifts 0.3 < z < 0.9. By constraining the sample to have the same evolution-corrected, absolute R-band magnitude distribution at all redshifts (M R < -21.3, M R ≃ -22.0), it

…(Full text truncated)…