The $N_H$ Distribution of Hard X-ray Selected AGN in the NEP Field

TL;DR

This study analyzes 60 hard X-ray sources in the JWST NEP field using NuSTAR and overlapping XMM-Newton data to derive the obscuration properties of AGN. Employing two spectral models, a baseline absorbed powerlaw and the uxclumpy torus model, within a Bayesian framework (BXA/UltraNest), the authors determine both the observed and intrinsic distributions of hydrogen column density $N_{ m H}$ and quantify the Compton Thick fraction $f_{ m CT}$. They find an observed $f_{ m CT} = 0.13^{+0.15}_{-0.04}$ (and $0.30^{+0.21}_{-0.08}$ intrinsic) down to an 8–24 keV flux of $6.0 imes 10^{-14}$ erg s$^{-1}$ cm$^{-2}$, consistent with population synthesis models, while suggesting possible redshift evolution with higher CT incidence at $z > 1$. The NEP field data, including deep, simultaneous soft and hard X-ray coverage, provide a high-signal measure of AGN obscuration and underscore the need for broader spectral models and more complete redshift information for robust evolutionary trends.

Abstract

X-ray surveys are one of the most unbiased methods for detecting Compton Thick (CT; $N_{\mathrm{H}} \geq 10^{24}$ cm$^{-2}$) AGN, which are thought to comprise up to $60\%$ of AGN within $z \lesssim 1.0$. These CT AGN are often difficult to detect with current instruments, but the X-ray data within the JWST-North Ecliptic Pole (NEP) Time Domain Field (TDF) present a unique opportunity to study faint and obscured AGN. The NEP contains the deepest NuSTAR survey to date, and Zhao et al. (2024) detected 60 hard X-ray sources from the combined exposure of NuSTAR's Cycle 5 and 6 observations. In this work, we utilize the NuSTAR Cycle 5+6+8+9 data and simultaneous XMM-Newton observations in order to perform the first spectroscopic analysis of the 60-source catalog. We present this analysis and measure the $N_{\mathrm{H}}$ distribution of the sample. We measure an observed CT fraction of $0.13_{-0.04}^{+0.15}$ down to an observed $8-24$ keV flux of $6.0 \times 10^{-14}$ erg/s/cm$^{2}$, and we correct our analysis for absorption bias to estimate an underlying CT fraction of $0.32_{-0.08}^{+0.23}$. The derived obscuration distribution and CT fraction are consistent with population synthesis models and previous surveys.

Figures (11)

• Figure 1: The redshifts (left) and observed 3-24 keV X-ray luminosities (right) for sources with detected multi-wavelength counterparts. Shown are quasars (dotted; pink), galaxies (hashed; purple), and unclassified (solid; gray) objects.
• Figure 2: A histogram of the net spectral counts in NuSTAR (pink; dotted) and XMM-Newton (purple; hashed) for the sources in our sample.
• Figure 3: A schematic demonstrating the spectral components (unabsorbed (purple; dashed) and absorbed (purple; dotted) primary powerlaw, scattered powerlaw (red; dot-dashed), and galactic emission (gold; solid)) of the baseline model described in Section \ref{['sec:models']}. This model is shown for a typical Compton Thin AGN with $N_{\mathrm{H}} = 10^{23}$ cm$^{-2}$. The total spectrum is shown in bolded black. The background is shaded to demonstrate the energy range of XMM-Newton (gray) and NuSTAR (purple).
• Figure 4: A comparison of the best-fit column densities ($N_{\mathrm{H}}$) for the baseline and clumpy models. The shaded regions mark the Compton Thin (purple) and Compton Thick (gray) regimes. Sources without redshift measurements are shown as red squares.
• Figure 5: The effective number $N$ and fraction $f_{N_\mathrm{H}}$ of AGN observed in each log($N_{\mathrm{H}}$) bin within $1 \sigma$ for the baseline (pink; hatched) and clumpy (purple; dotted) models, as derived by integrating the summed PDF shown in Figure \ref{['fig:nh_pdf']} (Appendix \ref{['apendix-fits']}).