|Annu. Rev. Astron. Astrophys. 2006. 44:
Copyright © 2006 by . All rights reserved
Chandra observations of galaxies at high redshift (z > 0.1), either from identification of deep survey sources or from stacking analysis of distant galaxy fields, have been reviewed recently in the literature (see Fabbiano & White 2006; Brandt & Hasinger 2005) and will not be discussed in detail here. In summary, the emission from normal galaxies becomes an increasingly greater component of the X-ray emission at the deepest X-ray counts (Bauer et al. 2004; Ranalli, Comastri & Setti 2005); moreover, the hard X-ray emission is a direct diagnostic of star formation, as demonstrated by the good FIR-X-ray correlations and by the work on the XLFs of star-forming galactic populations discussed earlier in this review (Fabbiano & Shapley 2002; Grimm, Gilfanov & Sunyaev 2003; Ranalli, Comastri & Setti 2003; Colbert et al. 2004; Gilfanov, Grimm & Sunyaev 2004a; Persic et al. 2004). It is clear that the study of the global properties and luminosity functions of galaxies at different redshifts can give information in this area, and this work is beginning to gather momentum (e.g., Georgakakis et al. 2003; Norman et al. 2004; Hornschemeier et al. 2005; Ranalli, Comastri & Setti 2005), given the availability of XMM-Newton surveys of the nearby universe and the increasingly deep Chandra surveys.
Enhanced star formation early in the life of a galaxy is expected to produce enhancements in its X-ray emission at different epochs, related to the formation and evolution of HMXB and LMXB populations (Ghosh & White 2001). Lehmer et al. (2005) report such an effect in their stacking analysis of Lyman break galaxies in the HST GOODS fields covered by deep (1-2 Ms exposure) Chandra fields (Figure 11). Conversely, if the SFR is independently known, the relation between the integrated luminosity of galaxies and the SFR can be used to measure the maximum luminosity of a HMXB and the presence of a very high luminosity IMBH population not related to stellar sources (Gilfanov, Grimm & Sunyaev 2004b). These authors, based on the XLF-SFR connection (Grimm, Gilfanov & Sunyaev 2003), explore the statistical properties of a population of discrete sources and demonstrate that a break is expected in the relation between the total X-ray luminosity LX of the galaxies and the SFR, which depends on the high luminosity cut-off of the XRB population. Comparing the local galaxy sample with the Hubble Field North galaxies, they suggest a cut-off luminosity ~ 5 × 1040 erg s-1 for HMXBs. They also suggest that a population of very luminous IMBHs (LX > 1040 erg s-1) would reveal itself with a steeper LX-SFR relation at higher luminosities and star-formation regimes.
Figure 11. Evolution of the X-ray-to-optical ratio of galaxies with redshift, peaking at z ~ 1.5-3 (from figure 5 of Lehmer et al. 2005).
The Chandra observations of nearby galaxies have significantly increased our understanding of X-ray source populations, and will deepen this understanding in the next several years. These results and the tantalizing possibility of studying the X-ray evolution of galaxies show that future very deep, high-resolution X-ray observations are essential. Unfortunately, there are no plans for X-ray missions with comparable angular resolution to follow Chandra. These types of studies may have to wait for several decades for a large area, subarcsecond resolution X-ray mission (e.g., the Generation-X mission, for which a NASA-funded vision study is in progress).
This review has benefited from comments and discussions with several colleagues, including Dong-Woo Kim, Martin Elvis, Andrew King, Mitch Begelman, Andreas Zezas, Albert Kong, Sergey Trudolyubov, Konstantin Postnov, Massimo Persic, Tom Maccarone, Curt Struck, Tim Roberts, Phil Kaaret, Wolfgang Pietsch, Xiang-Dong Li, and John Kormendy. I thank A. Tennant, M. Gilfanov, R. Soria, E. Kim, Z. Wang, A. Prestwhich and B. Lehmer for providing figures. I am indebted to the NASA ADS for making my job of searching the literature so much easier.