Although this chapter is focussed on the XRB populations that we can now resolve and study with Chandra in galaxies as distant as ~ 20 Mpc, the study of the integrated emission properties of samples of galaxies (either more distant, or observed at lower resolution) can also give useful information on the average properties of their XRB components. We will summarize here some of these studies, that were pursued mostly by using the samples of galaxies observed with Einstein and ROSAT.
Most of the early work in this area was done by Fabbiano and collaborators, using the first sample of galaxies ever observed in X-rays, the Einstein sample (see reviews in Fabbiano 1989; 1995). Besides suggesting the baseline XRB emission in E and S0 galaxy, that is now confirmed with Chandra (Section 4), these results suggested a general scaling of the integrated X-ray emission with the optical luminosity (and therefore stellar population) of the galaxies, and pointed to a strong association of the XRB populations of disk/arm-dominated spirals with the far-IR emission, i.e. the younger component of the stellar population (e.g., Fabbiano, Gioia & Trinchieri 1988; see also David, Jones and Forman 1992/). More recent work on the Einstein sample (Shapley, Fabbiano & Eskridge 2001; Fabbiano & Shapley 2002 1), on ROSAT-observed galaxies (Read & Ponman 2001), and on Beppo-SAX and ASCA data (Ranalli, Comastri & Setti 2003) has examined some of these correlations afresh. Given the different pass-bands of these observatories, these studies have a varied sensitivity to the effect of hard XRB emission and soft hot ISM emission in the galaxies.
The Einstein sample is the largest, consisting of 234 S0/a-Irr galaxies observed in the 0.2-4. keV band. The X-ray luminosities are compared with B, H, 12 µm, 60 µm, 100 µm, global FIR, and 6 cm luminosities (Shapley, Fabbiano & Eskridge 2001; Fabbiano & Shapley 2002). Both fluxes and upper limits were used in this work, to avoid obvious selection biases. This work provides baseline distributions of LX and of LX / LB for the entire Hubble sequence (including E and S0 galaxies), and a critical compilation of distances for the sample. Multi-variable correlation analysis shows clear dependencies of the emission properties on the morphological type of the galaxies (and therefore indirectly on the stellar population and star formation activity). In Sc-Irr galaxies, all the emission properties (including the X-rays) are tightly correlated, suggesting a strong connection to the stellar population. This is not true for S0/a-Sab, where there is a general connection of the X-ray luminosity with the B and H-band emission (stellar population), but not with either radio or FIR. In Sc-Irr galaxies the strongest link of the X-ray emission is a linear correlation with the FIR, suggesting a connection with the star-forming stellar component. This conclusion is reinforced by a correlation between LX / LB and L60 / L100, which associates more intense X-ray emission with hotter IR colors.
The X-ray emission / star-formation connection is also discussed as a result of the analysis of a small sample (17 nearby spirals) observed with ROSAT in a softer energy band (0.1-2.0 keV; Read & Ponman 2001), and more recently from the analysis of another small sample (also 17 galaxies) observed in the 2-10 keV band (Ranalli, Comastri & Setti 2003). The advantage of this harder band is that the emission is predominantly due to the XRB population (if the sample does not include AGNs). These authors suggest that the hard X-ray emission can be used as a clean indicator of star formation, because extinction is not a problem at these energies.
These correlation analyses are now being extended to the XRB populations detected with Chandra. Colbert et al. (2003) report good correlations between the total point source X-ray luminosity in a sample of 32 galaxies of different morphological type extracted from the Chandra archive and the stellar luminosity (both B and K bands). While correlations are still present in the spiral and merger/irregular galaxies with FIR and UV luminosities, the ellipticals do not follow this trend and show a clear lack of FIR and UV emission, consistent with their older stellar populations. This results is consistent with the conclusions of Fabbiano & Shapley (2002; see above), which were however based on the analysis of the integrated x-ray luminosity of bulge dominated and disk/arm dominated spiral and irregular galaxies.
In summary, there is a correlation between X-ray emission and SFR in star-forming galaxies, that may lead to a new indicator of the SFR. However, one has to exercise caution, because this conclusion is only true for star-forming galaxies. In old stellar systems (bulges, gas-poor E and S0s), the X-ray emission is connected with the older stellar population of these systems. This conclusion is also in agreement with the recent studies of XLFs (Section 2; Section 4.3).
1 probably the last paper to be published on the Einstein data Back.