Draft version October 20, 2018 A Preprint typeset using LTEX style emulateapj v. 08/22/09 INFRARED AND HARD X-RAY DIAGNOSTICS OF AGN IDENTIFICATION FROM THE SWIFT/BAT AND AKARI ALL-SKY SURVEYS K. Matsuta1, 2, P. Gandhi2, T. Dotani2, 1, T. Nakagawa2, N. Isobe2, Y. Ueda3, K. Ichikawa3, Y. Terashima4, S. Oyabu5, I. Yamamura2, 1, andL. Stawarz2, 6 1Department of Space and Astronautical Science, The Graduate University for Advanced Studies, 3-1-1 Yoshinodai, Chuo-ku, Sagamihara, Kanagawa 252-5210, Japan 2Institute of Space and Astronautical Science, Japan Aerospace Exploration Agency, 3-1-1 Yoshinodai, Chuo-ku, Sagamihara, Kanagawa 252-5210, Japan 3Department of Astronomy, Kyoto University, Kitashirakawa-Oiwake-cho, Sakyo-ku, Kyoto 606-8502, Japan 4Department of Physics, Ehime University, 2-5, Bunkyo-cho, Matsuyama, Ehime 790-8577, Japan 5Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Aichi 464-8601, Japan 6Astronomical Observatory, Jagiellonian University, ul. Orla 171, Krak´ow 30-244, Poland Draft version October 20, 2018 ABSTRACT We combine data from two all-sky surveys in order to study the connection between the infrared and hard X-ray (> 10 keV) properties for local active galactic nuclei (AGN). The Swift/Burst Alert Telescope all-sky survey provides an unbiased, flux-limited selection of hard X-ray detected AGN. Cross-correlating the 22-month hard X-ray survey with the AKARI all-sky survey, we studied 158 AGN detected by the AKARI instruments. We find a strong correlation for most AGN between the infrared (9, 18, and 90 µm) and hard X-ray (14–195 keV) luminosities, and quantify the correlation for various subsamples of AGN. Partial correlation analysis confirms the intrinsic correlation after removing the redshift contribution. The correlation for radio galaxies has a slope and normalization identical to that for Seyfert 1s, implying similar hard X-ray/infrared emission processes in both. In contrast, Compton-thick sources show a large deficit in the hard X-ray band, because high gas column densities diminish even their hard X-ray luminosities. We propose two photometric diagnostics for source classification: one is an X-ray luminosity vs. infrared color diagram, in which type 1 radio-loud AGN are well isolated from the others in the sample. The other uses the X-ray vs. infrared color as a useful redshift-independent indicator for identifying Compton-thick AGN. Importantly, Compton- thick AGN and starburst galaxies in composite systems can also be differentiated in this plane based upon their hard X-ray fluxes and dust temperatures. This diagram may be useful as a new indicator to classify objects in new and upcoming surveys such as WISE and NuSTAR. Subject headings: galaxies: active — infrared: galaxies — X-ray: galaxies 1. INTRODUCTION is not expected according to pioneering AGN dust Active galactic nuclei (AGN) emit broadband emission torus models (Pier & Krolik 1993), which assumed over the entire electromagnetic spectrum. Their typical homogeneous and smooth dust distributions, because IR spectral energy distributions (SEDs) display a strong in- emission from the warm-inner clouds is expected to be frared (IR) bump around 1–100 µm, where a significant dramatically scattered and reduced when viewed from portion of their power emerges (e.g., Antonucci 1993; an edge-on orientation. On the contrary, observations Elitzur 2008). Its source is considered to be a pc-scale show a tight correlation with little scatter and no IR torus of dust and gas clouds, which is heated by radia- depletion, largely independent of orientation (which ought to correlate approximately with dust reddening) arXiv:1205.0140v1 [astro-ph.HE] 1 May 2012 tion from the nucleus and reprocesses this power into the IR, at least for the majority of AGN which are not dom- at all AGN luminosities. Under the torus paradigm, inated by beamed emission from relativistic jets (e.g., the correlation may be consistent with models having Elitzur 2008). The nuclear X-ray emission from AGN is a clumpy obscuring geometry (e.g., Krolik & Begelman produced by Comptonization of the disk emission in a hot 1988; Dullemond & Dominik 2005; H¨onig et al. corona above an accretion disk (Sunyaev & Titarchuk 2006; Nenkova et al. 2008; Schartmann et al. 2008; 1980; Haardt & Maraschi 1993; Zezas et al. 1998), again H¨onig et al. 2010; H¨onig & Kishimoto 2010), in which with a possible exception of jet-dominated sources. the inner torus regions can be visible even at edge-on It has been shown that there is a good linear orientations. correlation between the logarithms of the observed However, previous works are based on small and in- continuum emission in the mid-IR and the intrinsic complete samples, with selection criteria biased towards soft X-ray (< 10 keV) band luminosities of local bright sources visible in the mid-IR from the ground, or AGN, which appears to be largely independent of the sources in which the intrinsic soft X-ray power could be amount of dust reddening (e.g., Krabbe et al. 2001; determined. Below 10 keV, AGN spectra can require Lutz et al. 2004; Horst et al. 2006, 2008; Gandhi et al. detailed modeling due to the presence of several compo- 2009; Levenson et al. 2009; Asmus et al. 2011). This nents including the underlying AGN power law, obscura- tion, scattering, non-thermal emission of relativistic jets Electronic address: [email protected] (in case of radio-loud sources), and eventually thermal 2 emission of hot gaseous interstellar medium. Good qual- 2007). The observed soft X-ray and IR appearance is ity X-ray spectroscopy is necessary to disentangle these. not easily distinguishable in these cases. Our hard X-ray The intrinsic X-ray fluxes of Compton-thick (CT) AGN selection now allows us to do this. The simple photomet- 24 −2 (NH > 10 cm ), especially, are fully absorbed in soft ric diagnostics can be useful tools for new and upcoming X-ray band energies. hard X-ray surveys with NuSTAR, ASTRO-H and in the In this paper, our aim is to explore the X-ray vs. IR IR with WISE. correlation using unbiased samples, and then define new This paper is organized as follows. In section 2, we and simple photometric diagnostics for AGN classifica- present details of the sample selection and catalog cross- tion. We use the Swift/Burst Alert Telescope (BAT) matching. Section 3 shows the main results regarding the hard X-ray (> 10 keV) survey as the base AGN selection hard X-ray and IR luminosity correlation and its statis- catalog. The hard X-ray energy band is rather insen- tics. Section 4 discusses the implications of the results sitive to the intervening column density up to mildly and provides new diagnostic plots for AGN classification. CT columns. The survey reflects the intrinsic lumi- Section 5 lists the main conclusions. This work is closely nosity of the source and provides samples largely unbi- related to that of Ichikawa et al. (2012), in which the re- ased by obscuration. We use the Swift/BAT 22-month lation between the IR/X-ray spectra of Seyferts and new Source Catalog herein (Tueller et al. 2010). Several re- type AGN are studied by using the Swift/BAT 9-month cent works explored the connection of hard X-rays with AGN Catalog (Tueller et al. 2008) and the AKARI/PSC near-IR powers and with mid-IR emission line proper- complemented by those with IRAS and WISE to en- ties (e.g., Mushotzky et al. 2008; Diamond-Stanic et al. sure high completeness of cross identification. The larger 2009; Rigby et al. 2009; Weaver et al. 2010), but corre- 22-month BAT sample used as the starting catalog for lation with an all-sky mid-IR survey at high angular res- the present work includes several heavily CT AGN and olution is still lacking. radio-loud (RL) sources, and we are able to compare the High angular resolution is crucial in order to regions occupied by these various AGN classes in the properly separate AGN from stellar emission in luminosity correlation plane. We also provide detailed the IR (e.g., Gandhi et al. 2009; Horst et al. 2009; statistical tests and correlation fits, and construct AGN Vasudevan et al. 2010; Mullaney et al. 2011). For many photometric classification plots. years, IRAS provided the most complete all-sky sur- A flat Universe with a Hubble constant H0 = −1 −1 vey (Joint IRAS Science 1994), but with an angular 71 km s Mpc ,ΩΛ =0.73 and ΩM =0.27 is assumed resolution only of order arcmins. The AKARI satel- throughout this paper. lite (Murakami et al. 2007) has now completed its all- 2. DATA SELECTION sky survey which is several times more sensitive than IRAS, and at a much higher angular resolution of or- We used the following two all-sky survey catalogs to der arcsecs, (7′′ and 48′′ for the mid- and far-IR in- study the correlation in a complete, flux-limited sam- struments, respectively). The AKARI survey thus pro- ple and examine the luminosity correlation between the vides the best IR sample for cross-correlation with observed IR and hard X-ray bands of local AGN (z < Swift AGN. We use data available in the AKARI Point 0.1); the Swift/BAT 22-month Source Catalog and the Source Catalogs (AKARI/PSC; Ishihara et al. 2010; AKARI/PSC. Yamamura et al. 2010) over a range of wavelengths from 2.1. The Swift/BAT 22-month Source Catalog 9–160 µm. Although AKARI’s angular resolution is lower than large ground telescopes (e.g., the diffraction The prime objective of the Swift/BAT is the study of limit of the Very Large Telescope/VISIR ∼ 0.3′′ in the gamma-ray bursts, covering the hard X-ray band of 14– N-band), our study has the capability of probing pre- 195 keV.