X-ray spectral modelling of the AGN obscuring region in the CDFS: Bayesian model selection and catalogue

TL;DR

The paper develops a Bayesian framework for X-ray spectral analysis of AGN to jointly estimate parameters and perform model comparison among ten physically motivated obscuration models, explicitly propagating Poisson counting uncertainty and photometric redshift PDFs. Using nested sampling via MultiNest and Sherpa/BXA integration, the authors compute evidences and posterior distributions to discriminate between geometries (wabs, sphere, torus) with scattering and reflection components, revealing that an intermediate, geometrically extended obscurer with significant Compton scattering best describes the 4 Ms CDFS AGN population. For the exemplar source 179, simple absorption is insufficient, and models including scattering and reflection against an obscuring medium yield the highest evidence, with $N_H$ and $\Gamma$ lying around $10^{22}$ cm$^{-2}$ and $\sim1.9$, respectively; the inferred parameters vary with geometry, emphasizing the importance of proper model selection. Across the full sample, the analysis supports a four-component spectral model, rules out simplistic geometries, and points toward density gradients or disk reflection for the additional Compton reflection; the authors also release a catalogue of $L_{2-10\mathrm{keV}}$ and $N_H$ estimates for the AGN in CDFS, enabling improved population studies of AGN obscuration and luminosity function evolution.

Abstract

AGN are known to have complex X-ray spectra that depend on both the properties of the accreting SMBH (e.g. mass, accretion rate) and the distribution of obscuring material in its vicinity ("torus"). Often however, simple and even unphysical models are adopted to represent the X-ray spectra of AGN. In the case of blank field surveys in particular, this should have an impact on e.g. the determination of the AGN luminosity function, the inferred accretion history of the Universe and also on our understanding of the relation between AGN and their host galaxies. We develop a Bayesian framework for model comparison and parameter estimation of X-ray spectra. We take into account uncertainties associated with X-ray data and photometric redshifts. We also demonstrate how Bayesian model comparison can be used to select among ten different physically motivated X-ray spectral models the one that provides a better representation of the observations. Despite the use of low-count spectra, our methodology is able to draw strong inferences on the geometry of the torus. For a sample of 350 AGN in the 4 Ms Chandra Deep Field South field, our analysis identifies four components needed to represent the diversity of the observed X-ray spectra: (abridged). Simpler models are ruled out with decisive evidence in favour of a geometrically extended structure with significant Compton scattering. Regarding the geometry of the obscurer, there is strong evidence against both a completely closed or entirely open toroidal geometry, in favour of an intermediate case. The additional Compton reflection required by data over that predicted by toroidal geometry models, may be a sign of a density gradient in the torus or reflection off the accretion disk. Finally, we release a catalogue with estimated parameters such as the accretion luminosity in the 2-10 keV band and the column density, $N_{H}$, of the obscurer.

Figures (6)

• Figure 1: Illustration of the typical shapes of the discussed spectral features. Emission lines and absorption edges have been omitted for simplicity.
• Figure 2: Cartoon illustrations of the geometries associated with each model. The wabs model (b) represents an absorbing slab in the line of sight, and can also be interpreted as the case of a torus with extreme opening angle. While torus (d) uses a intermediate opening angle, the sphere (c) represents the other extreme, a vanishing opening angle. The +scattering extension (e) of the named models correspond to Thomson scattering from outside the line of sight, which does not experience any energy-dependent effects. Finally, the reflection component (+pexmon) corresponds to either disk reflection (f) or additional reflection if the torus is not viewed through the same column density as the reflection (g, h, i). For the sphere it should be noted that scattering is not physically possible, as no unabsorbed radiation can escape.
• Figure 3: Models considered. For model comparison, we compute the evidence (see Section \ref{['sub:Model-selection']}) for each source and model. We then start by assuming a power law (powerlaw) and move along the arrows to more complex models if model comparison justifies the preference. The three obscurer models, wabs, sphere and torus, are compared against each other, as well as the introduction of additional features (+scattering, +pexmon). See text for details.
• Figure 4: The redshift and number of counts in the $0.5-8\text{ keV}$ band of the AGN sample. For the redshift, the spectral redshift is shown where available and otherwise the median on the photometric redshift distribution is used.
• Figure 5: Observed (convolved) spectrum of object 179, binned for plotting to 10 counts per bin. Shown are analyses using various models and their individual components: powerlaw (upper left), wabs (upper right), torus+scattering (lower left) and wabs+pexmon+scattering (lower right). The posterior of the parameters are used to compute the median and 10%-quantiles of each model component.