Constraints on the Redshift Evolution of the L_X-SFR Relation from the Cosmic X-Ray Backgrounds

Astronomy Department, Harvard University, 02138, Cambridge, MA, USA
Monthly Notices of the Royal Astronomical Society (Impact Factor: 5.23). 08/2011; 421(1). DOI: 10.1111/j.1365-2966.2011.20292.x
Source: arXiv

ABSTRACT Observations of local star forming galaxies have revealed a correlation
between the rate at which galaxies form stars and their X-Ray luminosity. We
combine this correlation with the most recent observational constraints on the
integrated star formation rate density, and find that star forming galaxies
account for 5-20% of the total soft and hard X-ray backgrounds, where the
precise number depends on the energy band and the assumed average X-ray
spectral energy distribution of the galaxies below ~20 keV. If we combine the
L_X-SFR relation with recently derived star formation rate function, then we
find that star forming galaxies whose X-ray flux falls well (more than a factor
of 10) below the detection thresholds of the Chandra Deep Fields, can fully
account for the unresolved soft X-ray background, which corresponds to ~6% of
its total. Motivated by this result, we put limits on the allowed redshift
evolution of the parameter c_X \equiv L_X/SFR, and/or its evolution towards
lower and higher star formation rates. If we parametrize the redshift evolution
of c_X ~ (1+z)^b, then we find that b \leq 1.3 (95% CL). On the other hand, the
observed X-ray luminosity functions (XLFs) of star forming galaxies indicate
that c_X may be increasing towards higher redshifts and/or higher star
formation rates at levels that are consistent with the X-ray background, but
possibly at odds with the locally observed L_X-SFR relation.

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Evolution in the X-ray luminosity -- star formation rate ($L_X$-SFR) relation could provide the first evidence of a top-heavy stellar initial mass function in the early universe, as the abundance of high-mass stars and binary systems are both expected to increase with decreasing metallicity. The sky-averaged (global) 21-cm signal has the potential to test this prediction via constraints on the thermal history of the intergalactic medium, since X-rays can most easily escape galaxies and heat gas on large scales. A significant complication in the interpretation of upcoming 21-cm measurements is the unknown spectrum of accreting black holes at high-$z$, which depends on the mass of accreting objects and poorly constrained processes such as how accretion disk photons are processed by the disk atmosphere and host galaxy interstellar medium. Using a novel approach to solving the cosmological radiative transfer equation (RTE), we show that reasonable changes in the characteristic BH mass affects the amplitude of the 21-cm signal's minimum at the $\sim 10-20$ mK level --- comparable to errors induced by commonly used approximations to the RTE --- while modifications to the intrinsic disk spectrum due to Compton scattering (bound-free absorption) can shift the position of the minimum of the global signal by $\Delta z \approx 0.5$ ($\Delta z \approx 2$), and modify its amplitude by up to $\approx 10$ mK ($\approx 50$ mK) for a given accretion history. Such deviations are larger than the uncertainties expected of current global 21-cm signal extraction algorithms, and could easily be confused with evolution in the $L_X$-SFR relation.
    Monthly Notices of the Royal Astronomical Society 06/2014; 443(2). DOI:10.1093/mnras/stu1193 · 5.23 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper we investigate the power spectrum of the unresolved 0.5-2 keV cosmic X-ray background (CXB) with deep Chandra 4-Msec (Ms) observations in the Chandra Deep Field South (CDFS). We measured a signal that, on scales >30 arcsec, is significantly higher than the shot noise and is increasing with angular scale. We interpreted this signal as the joint contribution of clustered undetected sources like active galactic nuclei (AGN), galaxies and the intergalactic medium (IGM). The power of unresolved cosmic source fluctuations accounts for ˜12 per cent of the 0.5-2 keV extragalactic CXB. Overall, our modelling predicts that ˜20 per cent of the unresolved CXB flux is produced by low-luminosity AGN, ˜25 per cent by galaxies and ˜55 per cent by the IGM. We do not find any direct evidence of the so-called 'warm hot intergalactic medium' (i.e. matter with 105 < T < 107 K and density contrast δ < 1000), but we estimated that it could produce about 1/7 of the unresolved CXB. We placed an upper limit on the space density of postulated X-ray-emitting early black holes at z > 7.5 and compared it with supermassive black hole evolution models.
    Monthly Notices of the Royal Astronomical Society 11/2012; 427(1):651-663. DOI:10.1111/j.1365-2966.2012.21867.x · 5.23 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We present the most complete study to date of the X-ray emission from star-formation in high redshift (median z=0.7; z<1.5), IR-luminous (L_IR=10^10-10^13 L_sun) galaxies detected by Herschel's PACS and SPIRE instruments. For our purpose we take advantage of the deepest X-ray data to date, the Chandra deep fields (North and South). Sources which host AGN are removed from our analysis by means of multiple AGN indicators. We find an AGN fraction of 18+/-2 per cent amongst our sample and note that AGN entirely dominate at values of log[L_X/L_IR]>-3 in both hard and soft X-ray bands. From the sources which are star-formation dominated, only a small fraction are individually X-ray detected and for the bulk of the sample we calculate average X-ray luminosities through stacking. We find an average soft X-ray to infrared ratio of log[L_SX/L_IR]=-4.3 and an average hard X-ray to infrared ratio of log[L_HX/L_IR]=-3.8. We report that the X-ray/IR correlation is approximately linear through the entire range of L_IR and z probed and, although broadly consistent with the local (z<0.1) one, it does display some discrepancies. We suggest that these discrepancies are unlikely to be physical, i.e. due to an intrinsic change in the X-ray properties of star-forming galaxies with cosmic time, as there is no significant evidence for evolution of the L_X/L_IR ratio with redshift. Instead they are possibly due to selection effects and remaining AGN contamination. We also examine whether dust obscuration in the galaxy plays a role in attenuating X-rays from star-formation, by investigating changes in the L_X/L_IR ratio as a function of the average dust temperature. We conclude that X-rays do not suffer any measurable attenuation in the host galaxy.
    Monthly Notices of the Royal Astronomical Society 07/2014; 443(4). DOI:10.1093/mnras/stu1441 · 5.23 Impact Factor


1 Download
Available from