Draftversion January4,2017 PreprinttypesetusingLATEXstyleemulateapjv.5/2/11 A NEW COMPTON-THICK AGN IN OUR COSMIC BACKYARD: UNVEILING THE BURIED NUCLEUS IN NGC 1448 WITH NuSTAR A. Annuar1, D. M. Alexander1, P. Gandhi2, G. B. Lansbury1, D. Asmus3, D. R. Ballantyne4, F. E. Bauer5,6,7, S. E. Boggs8, P. G. Boorman2, W. N. Brandt9,10,11, M. Brightman12, F. E. Christensen13, W. W. Craig8,14, D. Farrah15, A. D. Goulding16, C. J. Hailey17, F. A. Harrison12, M. J. Koss18, S. M. LaMassa19, S. S. Murray†,20,21, C. Ricci5,22, D. J. Rosario1, F. Stanley1, D. Stern23 and W. Zhang19. 1CentreforExtragalacticAstronomy,DepartmentofPhysics,DurhamUniversity,SouthRoad,Durham,DH13LE,UK 7 2DepartmentofPhysics&Astronomy,FacultyofPhysicalSciences andEngineering,UniversityofSouthampton, Southampton, SO17 1BJ,UK 1 3EuropeanSouthernObservatory,AlonsodeCordova,Vitacura,Casilla19001,Santiago, Chile 0 4CenterforRelativisticAstrophysics,School ofPhysics,GeorgiaInstitute ofTechnology, Atlanta, GA30332, USA 2 5InstitutodeAstrof´ısicaandCentrodeAstroingenier´ıa,FacultaddeF´ısica,PontificiaUniversidadCat´olicadeChile,Casilla306, Santiago22,Chile n 6MillenniumInstitute ofAstrophysics(MAS),NuncioMonsen˜orS´oteroSanz100,Providencia,Santiago, Chile a 7SpaceScienceInstitute, 4750WalnutStreet,Suite205,Boulder,Colorado80301 J 8SpaceSciences Laboratory,UniversityofCalifornia,BerkeleyCA94720, USA 2 9DepartmentofAstronomyandAstrophysics,ThePennsylvaniaState University,525DaveyLab,UniversityPark,PA16802,USA 10InstituteforGravitationandtheCosmos,ThePennsylvaniaStateUniversity,UniversityPark,PA16802,USA 11DepartmentofPhysics,ThePennsylvaniaState University,525DaveyLab,UniversityPark,PA16802,USA ] 12CahillCenter forAstronomyandAstrophysics,CaliforniaInstituteofTechnology, Pasadena,CA91125,USA E 13DTUSpace, NationalSpaceInstitute, Technical UniversityofDenmark,Elektrovej327,DK-2800Lyngby, Denmark H 14LawrenceLivermoreNationalLaboratory,Livermore,CA94550,USA 15DepartmentofPhysics,VirginiaTech,Blacksburg,VA24061, USA h. 16DepartmentofAstrophysicalSciences,PrincetonUniversity,Princeton,NJ08544,USA 17ColumbiaAstrophysicsLaboratory,ColumbiaUniversity,NewYork,NY10027, USA p 18Institute forAstronomy,DepartmentofPhysics,ETHZurich,Wolfgang-Pauli-Strasse27,CH-8093Zurich,Switzerland - 19NASAGoddardSpaceFlightCenter,Greenbelt, Maryland20771, USA o 20Harvard-SmithsonianCenterforAstrophysics,60GardenStreet, Cambridge,MA02138, USA r 21Department ofPhysicsandAstronomy,JohnsHopkinsUniversity,3400NorthCharlesStreet,Baltimore,MD21218, USA st 22KavliInstituteforAstronomyandAstrophysics,PekingUniversity,Beijing100871, Chinaand a 23JetPropulsionLaboratory, CaliforniaInstituteofTechnology, Pasadena, CA91109, USA [ Draft version January 4, 2017 1 ABSTRACT v 7 NGC 1448 is one of the nearest luminous galaxies (L8−1000µm > 109L⊙) to ours (z = 0.00390), and yet the active galactic nucleus (AGN) it hosts was only recently discovered, in 2009. In this 9 paper, we present an analysis of the nuclear source across three wavebands: mid-infrared (MIR) 4 0 continuum, optical, and X-rays. We observed the source with the Nuclear Spectroscopic Telescope 0 Array (NuSTAR), and combined this data with archival Chandra data to perform broadband X-ray . spectral fitting (≈0.5–40 keV) of the AGN for the first time. Our X-ray spectral analysis reveals 1 that the AGN is buriedunder a Compton-thick (CT) column of obscuringgas alongour line-of-sight, 0 with a column density of N (los) & 2.5 × 1024 cm−2. The best-fitting torus models measured an 7 H 1 intrinsic 2–10 keV luminosity of L2−10,int = (3.5–7.6) × 1040 erg s−1, making NGC 1448 one of the : lowestluminosityCTAGNsknown. InadditiontotheNuSTARobservation,wealsoperformedoptical v spectroscopy for the nucleus in this edge-on galaxy using the European Southern Observatory New i X Technology Telescope. We re-classify the optical nuclear spectrum as a Seyfert on the basis of the Baldwin-Philips-Terlevich diagnostic diagrams, thus identifying the AGN at optical wavelengths for r a thefirsttime. WealsopresenthighspatialresolutionMIRobservationsofNGC1448withGemini/T- ReCS, in which a compactnucleus is clearly detected. The absorption-corrected2–10keV luminosity measured from our X-ray spectral analysis agrees with that predicted from the optical [Oiii]λ5007˚A emission line and the MIR 12µm continuum, further supporting the CT nature of the AGN. Subject headings: galaxies: active — galaxies: nuclei — techniques: spectroscopic — X-rays: galaxies — X-rays: individual (NGC 1448) 1. INTRODUCTION an active galactic nucleus (AGN) at the center of the At a distance of 11.5 Mpc (z = 0.00390),1 NGC 1448 galaxy was only discovered less than a decade ago by Goulding & Alexander (2009) through the detection of is one of the nearest bolometrically luminous galaxies (L8−1000µm > 109L⊙) to our own. Yet, the presence of tohfeahisgahm-ipolneizoaftilounm[Nineovu]sλ1g4a.l3a2xµiems eombsiesrsvioendlibnye,Sapsitpzaerrt. †Deceased. Basedonthe[Nev]/f8−1000µmand[Nev]/[Neii]λ12.82µm 1 Thequoteddistanceisthemetric/properdistancecalculated luminosity ratios, they found that the AGN contributes basedupontheMouldetal.(2000)cosmicattractor model,using a significant fraction (>25%) of the infrared (IR) emis- H0=75kms−1 Mpc−1 andadopting flatcosmology(ΩM =0.3, sion. So far,the focus of the majorityof studies ofNGC ΩΛ=0.7,q0=0.3)(Sanders etal.2003). 2 A. Annuar et al. 1448 has been on a number of supernovae occuring in new highangularresolutionmid-infrared(MIR) and op- thegalaxy(e.g.;Wang et al.2003;Sollerman et al.2005; tical observations of the source by Gemini/T-ReCS and Monard et al. 2014). NTT/EFOSC2, respectively, which also reveal the pres- BasedonitstotalK-bandmagnitudefromthe2MASS ence of a buried AGN. We organized the paper as fol- Large Galaxy Atlas, K = 7.66 (Jarrett et al. 2003), lows: In Section 2, we describe details of the multiwave- Tot the total mass of NGC 1448 is log (Mgal/M⊙) = 10.3 length observations and data reduction procedures for (using the stellar mass-to-light ratio versus B−V rela- NGC 1448. We presentthe X-rayspectralmodeling and tion of Bell et al. 2003.) The star formation rate (SFR) results in Section 3, followed by the data analysis and of the galaxy estimated from its far-IR luminosity mea- resultsof the opticalandMIRobservationsin Section4. suredbytheInfraredAstronomicalSatellite (IRAS),log Finally, we discuss and summarize the overall results in Lfir = 9.70 L⊙ (Sanders et al. 2003), is SFR ∼ 1 M⊙/yr Section 5 and Section 6, respectively. (Kennicutt 1998). This is consistent with the SFR ex- 2. OBSERVATIONS pected at this redshift, given the mass of the galaxy (Dav´e 2008). The nucleus is classified as an Hii region Inthissection,wedescribethenewNuSTAR(Sections in the optical by Veron-Cetty & Veron (1986) using the 2.1) and archival Chandra (Sections 2.2) observations of Hα to [Nii]λ6583˚A line ratio (Hα/[Nii] > 1.7), without NGC 1448. The NuSTAR data are essential for tracing any clear evidence for an AGN. The [Oiii]λ5007˚A and theintrinsicemissionfromtheburiedAGNathighX-ray energies (E & 10 keV). The Chandra data were used to Hβ emissionlineswerenotdetectedintheirobservation, aid our X-ray spectral analysis of the AGN at lower en- which could be attributed to obscuration by the highly inclinedScdhostgalaxy(i ≈86◦ relativetothe planeof ergies (E . 3 keV) where NuSTAR is not sensitive, and to reliably account for the emission from contaminat- the sky).2 Based on the [Oiv]λ25.89µm to [Oiii]λ5007˚A ing off-nuclear X-ray sources in the NuSTAR spectrum. lower limit emission line ratio, Goulding & Alexander Combining both NuSTAR and Chandra data together (2009) found an extinction of A > 5 mag within the V allows us to analyze the X-ray spectrum of the AGN in galaxy, suggesting high obscuration towards the AGN. NGC1448overabroadbandrangeofenergyforthefirst In recent years, many studies have been done to find time. We also describe new optical data (spectroscopy themostheavilyobscuredAGNs,particularlyCompton- andimaging)obtainedwiththeESONTT(Section2.3), thick (CT) AGNs, in the local universe. CTAGNs are and the new high angular resolution MIR observations AGNs that are obscured along our line-of-sight by gas with the Gemini-South telescope (Section 2.4), to pro- with column density of N ≥ 1/σ = 1.5 × 1024 cm−2 H T vide a multiwavelength view of the AGN. (whereσ istheThomsonscatteringcross-section). The T In addition to these data, we note that NGC 1448has obscuring gas is predominantly attributed to the AGN also been observed by Suzaku in X-rays for an exposure circumnuclear torus posited by the AGN unification timeof≈53ksin2009(2009-02-17;PID.M.Alexander; model(Antonucci1993;Urry & Padovani1995),butcan ObsID 703062010). However, it is only significantly de- also be contributed by larger-scalemolecular clouds and tected in the X-ray Imaging Spectrometer (XIS) instru- dustlanes(e.g.,Elvis2012;Prieto et al.2014). Thehigh mentup to ∼10keV.Analysis ofthe deepestSwift-BAT columnofgasseverelyabsorbsthe directX-rayemission 104-month maps using custom detection techniques to from the AGN, even at E & 10 keV in cases where the look for faint sources (Koss et al. 2013), show no signif- absorptionis extreme (N > 1025 cm−2; see Figure 1 of H icant excess (signal-to-noise ratio, SNR = −0.1) in the Gilli et al. 2007). This is what makes it very challeng- area near NGC 1448. The source was only detected in 1 ing to identify CTAGNs. Indeed, a hard X-ray study out of 5 Swift X-ray Telescope (XRT) observations with (E > 10 keV) by Ricci et al. (2016) using a large AGN ∼12counts in∼10ks exposuretime (2009-11-28;ObsID sample from the Swift Burst Alert Telescope (BAT) 70- 00031031001). We do not include these data in our X- month survey catalog, found an observed CTAGN frac- rayanalysisastheydonotprovideadditionalconstraints tion of ∼8% at z ≃ 0.055. This is significantly lower beyond those already achieved with our NuSTAR and than that expected from the synthesis of the cosmic Chandra data. X-ray background spectrum (10%–25%; e.g., Gilli et al. 2007; Treister et al. 2009; Draper & Ballantyne 2010; 2.1. NuSTAR Akylas et al. 2012; Ueda et al. 2014), and what is pre- TheNuclearSpectroscopicTelescopeArray (NuSTAR; dicted from multiwavelength studies of nearby AGN Harrison et al. 2013), launched in 2012 June, is the first (∼30%; e.g., Risaliti et al. 1999; Goulding et al. 2011). orbiting X-ray observatory which focus light at high en- Thissuggeststhatwearestillmissingasignificantnum- ergies (E & 10 keV). It consists of two co-aligned focal ber of CTAGN, even in the local universe. Forming a plane modules (FPMs), which are identical in design, complete census of their population is important in un- and referred to as FPMA and FPMB. Each FPM cov- derstanding the cosmic X-ray backgroundspectrum and ers the same 12′ × 12′ portionof the sky, and comprises the growth of supermassive black holes. four Cadmium-Zinc-Telluride detectors placed in a 2 × In this paper, we present NuSTAR and Chandra ob- 2 array. NuSTAR operates between 3 and 79 keV, and servations of NGC 1448 in which the AGN is detected provides100×improvementinsensitivityascomparedto in X-rays for the first time. We find that our direct previoushardX-rayorbitingobservatories(E &10keV; view towards the AGN is hindered by a Compton-thick e.g., Swift-BAT and INTEGRAL). In addition, it has column of obscuring gas. We also report the results of anangularresolutionof 18′′ full width at half maximum (FWHM), with a half power diameter of 58′′, resulting 2 WeobtainedthehostgalaxyinclinationfromtheHyperLeda in 10× improvement over previous observatories operat- website(http://leda.univ-lyon1.fr/). ingatE &10keV.Thesecharacteristics,particularlyits Unveiling the buried nucleus in NGC 1448 with NuSTAR 3 Figure 1. Chandra images of NGC 1448 in the 0.5–3 keV and 3–8 keV bands (top panel) and NuSTAR FPMA+B images in the 3–8 keV, 8–24 keV and 24–40 keV bands (bottom panel). The color scale for Chandra and NuSTAR images are shown with magenta and red representing the lowest and highest counts in each image, respectively. The black circle marks a 20′′-radius region centered on the Chandra position of the AGN that was used to extract the X-ray spectra. Images are smoothed with a Gaussian function of radius 5 pixels,correspondingto2.47′′ and12.3′′ forChandraandNuSTAR,respectively. Northisupandeastistotheleftinallimages. Thetwo off-nuclear X-ray sources detected within the extraction region in the Chandra image are at the north-east and north-west of the AGN, labeledasNEandNW,respectively. highenergycoverageandbettersensitivity,haveallowed TAR spectrum of NGC 1448 from each FPM using a ustoperformaccuratebroadbandX-rayspectralmodel- circular aperture region of 20′′-radius (corresponding to ingsofheavilyobscuredAGNsinthelocaluniverse(e.g., ∼30%NuSTARencircledenergyfraction,ECF)centered Balokovi´cet al. 2014; Puccetti et al. 2014; Bauer et al. on the Chandra position of the AGN (see Section 2.2). 2015; Annuar et al. 2015; Gandhi et al. 2016). Theaperturesizewaschosentominimizecontamination NGC 1448 was observed by NuSTAR in 2015 (2015- from off-nuclear sources observed in the Chandra data. 07-12; ObsID 60101101002) with an effective exposure Thebackgroundphotonswerecollectedfromanannulus time of 58.9 ks (60.3 ks on-source time) for each FPM. region centered on the AGN with inner and outer radii The sourcewas observedas partof our programto form of 40′′ and 70′′, respectively. The extracted spectrum themostcompletecensusoftheCTAGNpopulationand from each FPM were then co-added using the addas- the N distribution of AGN in the local universe, us- caspec script to increase the overall SNR of the data. H ing a volume-limited (D < 15 Mpc) AGN sample from We detected significant counts up to ∼40 keV from this Goulding & Alexander (2009).3 combined spectrum, and measured a net count rate of WeprocessedtheNuSTARdataofNGC1448withthe 2.79 × 10−3 counts s−1 in the 3–40 keV band. NuSTAR Data Analysis Software (nustardas) v1.4.1 We note that in the 24–40 keV band image, where we within heasoft v6.15.1 with CALDB v20150316. The expectthe AGNto completelydominate, the peakemis- nupipeline v0.4.3 script were used to produce the cali- sionappearstobeoffsetfromthecenteroftheextraction bratedandcleanedeventfiles using standardfilter flags. region. The offset is not observed in the 3–8 and 8–24 We extracted the spectra and response files using the keVbandimageswhere NuSTAR is moresensitive. Per- nuproducts v0.2.5 task. forming our spectral fits up to only 24 keV gave consis- The AGN is detected in both of the NuSTAR FPMs. tentresultswiththatobtainedbythe0.5–40keVspectral standard filter flags. The combined images of the AGN fits. We therefore attribute this offset due to NuSTAR from the two FPMs in the 3–8, 8–24, and 24–40 keV statistical uncertainty due to lower SNR at this higher bands are shown in Figure 1. We extracted the NuS- energy band. We note that centering the NuSTAR ex- traction region according to this offset, or enlarging the 3 The results of the first source in the sample observed by extractionregionto accountfor this apparentoffset also NuSTAR as part of this program, NGC 5643, was reported in donotsignificantlyaffectourfinalresultsontheanalysis Annuaretal.(2015). 4 A. Annuar et al. Figure 2. Optical R-bandimageofNGC 1448taken withtheESO NTT.Theapertureusedtoextract the optical spectraofthe AGN, optical peak and the total galaxy areplotted using cyan rectangle regions, and labeled as 1, 2 and3, respectively. The zoom-inimage of the central 20′′ × 20′′ of the galaxy is shown in the bottom-right panel. “×” marks the sources detected within a 20′′ circular radius of theAGNintheChandra2–8keVbandimage. of the AGN in NGC 1448. 2.2. Chandra NGC 1448 was observed by Chandra in 2014 with an exposure time of 49.4 ks (50.1 ks on-source time) using the ACIS-S detector as part of the Chandra HRC-GTO program (2014-03-09; PI S. Murray; ObsID 15332). We reprocessed the data to create event files with updated calibrationmodifications,followingstandardprocedures. WedeterminedthecentroidpositionoftheAGNinthe Chandrahardenergybandof2–8keVusingthewavde- tect tool within ciao with the thresholdparameter set to1×10−7. Wedetectedthreesourceswithinthecentral 20′′-radius of the galaxy in this energy band (see Figure 1). The brightest source was detected at position of RA = 3:44:31.83, and Dec. = −44:38:41.22, with errors of 0′.′17 and 0′.′11, respectively. This is consistent with the 2MASS and Gemini/T-ReCS (see Section 2.4) positions of the nucleus within ∼1′′. Therefore, we adopted this Chandra position as the AGN position. Figure 3. Top: HighspatialresolutionMIRimageofNGC1448 We extracted the source spectrum using the specex- takenbyGemini/T-ReCS.Bottom: Imagefitsperformedusingthe tract task in ciao from a circular region of 20′′-radius mirphottaskinidlfollowingAsmusetal.(2014). Shownarethe centered on the detected position of the AGN to match central 4′.′5 × 4′.′5 of the image. The left column is the original the NuSTAR extraction region. A 50′′-radius circular image, the middle column is the fit performed on the data, and the right column is the residual after subtracting the fit from the aperturewasusedtoextractthebackgroundcountsfrom data. The top row shows the results from a Gaussian fit to the an offset, source-free region. The total net count rate total emission, and the bottom row shows the results from fitting within the 20′′-radius extraction regionin the 0.5–8keV thestandardstar(asaPSFreference)tothetotalemission. Inall band is 5.69 × 10−3 counts s−1. The net count rate images,Northisupandeastistotheleft. measured by wavdetect for the AGN is 1.39 × 10−3 countss−1 inthe0.5–8keVband. Thetwoothersources detected within the extraction region are located to the Unveiling the buried nucleus in NGC 1448 with NuSTAR 5 north-east(NE)andnorth-west(NW)oftheAGN.They 2.4. Gemini/T-ReCS do not have counterparts at other wavelengths, and are NGC 1448 was observed at MIR wavelengths in 2010 likely to be X-ray binaries within NGC 1448 (see Sec- with high spatial resolution using the Thermal-Region tion 3.2.1). These sources have 0.5–8 keV count rates of Camera Spectrograph (T-ReCS; field of view 28′.′8 × 6.67 × 10−4 counts s−1 and 8.61 × 10−4 counts s−1, for 21′.′6; 0.09 arcsec pixel−1; Telesco et al. 1998), mounted the NE and NW sources, respectively, as measured by on the Gemini-South telescope. The observation was wavdetect. carried out on 2010-08-19 (Program ID GS-2010B-Q- 2.3. NTT/EFOSC2 3) for ≈319 s on-source time using the N-band filter (λ = 7.4–13.4 µm) in parallel chop and nod mode. The Atopticalwavelengths,weperformedspectroscopyfor data were reduced using the midir pipeline in iraf pro- NGC 1448 using the European Southern Observatory videdbytheGeminiObservatory,andtheimageanalysis (ESO) New Technology Telescope (NTT) on 2015-12- was performed using the idl package mirphot, follow- 07, using the Faint Object Spectrograph and Camera ing Asmus et al. (2014). The data were flux-calibrated v.2 (EFOSC2) instrument (Program ID 096.B-0947(A); with a standard star, HD22663,which was observed im- PI G. B. Lansbury). The source was observedfor four 5 mediately before NGC 1448. The resolution of the ob- mtioinne2x.2p)o.suTrheescselinttaerdeodpotendthweaCsh1′a′n(d7r5appco)siitniownid(stehe,Saencd- servation is ∼0′.′4 as measured from the FWHM of the standard star. the adopted grism yielded a spectral coverage of 4085– A compact nucleus is clearly detected in the MIR 7520˚A, and spectral resolution of FWHM = 12.6˚A. The continuum at position RA = 3:44:31.75 and Dec. = seeingatthetimeoftheobservationwas∼0′.′7. Standard −44:38:41.8(telescope astrometric uncertainty is on the iraf routines were followed to reduce the spectra, and order of 1′′) with a SNR ∼5. The source appears to be spectrophotometric standard star observations from the embeddedin∼0′.′7extendedemissionalongthehostma- same night were used for calibration. As shown in Fig- jor axis (position angle PA ∼ 44◦). However, owing to ure2,weextractedspectrafromthreedifferentapertures the small extent and often unstable point spread func- along the slit, correspondingto: (1) the AGN (using the tion (PSF) in ground-based MIR observations, a second Chandrapositionasareference);(2)the “opticalpeak”; epoch of high spatial resolution MIR images is required and (3) the total galaxy. The optical peak corresponds in order to confirm this extension. The source is best- to the position where the optical emission is the bright- fitted with a 2-dimensional Gaussian model (see Figure est in the imaging data and in the 2-dimensional spec- tra, and is offset from the AGN position by ∼3′′ (RA 3),whichmeasuredaflux density off12µm =12.5± 2.3 mJy. = 3:44:31.99 and Dec. = −44:38:42.33). The spatial aperturesadoptedwere2′.′4for the AGNandthe optical 3. X-RAYSPECTRALFITTING peak, and 36′.′9 for the total galaxy (see Figure 2). The 3.1. Basic Characterization apertures for the AGN and the optical peak correspond totwospatiallydistinct,prominentline-emittingregions Inthis section,wedescribethe broadbandX-rayspec- whichareseparatedbyadustlane,fromwhichwedonot tral analysis of the AGN in NGC 1448, performed us- see any prominent line emission. The aperture extents ing xspec. Given the non-negligible contribution of the for the AGN and the optical peak were chosen based background flux to the weak source flux, particularly in on the light profile of the Hα and [Oiii]λ5007˚Aemission the NuSTAR data at high energies, we binned both the lines in spatial direction, to ensure that we included the ChandraandNuSTAR spectra to a minimum of1 count total emission of these lines. per bin and optimized the fitting parameters using the Beforewecanmeasurethe fluxesofthe emissionlines, Poisson statistic, C-statistic (Cash 1979).5 We included we first need to subtract stellar emission from the ex- a fixed Galactic absorption component, NGal = 9.81 × H tracted spectra. This was done by fitting the emission 1019cm−2 (Kalberla et al.2005),usingthexspecmodel line-free regions of each spectrum with a combination “phabs” in all spectral fits, and assumed solar abun- of a small subset of galaxy template spectra from the dances for all models. The redshift was fixed at z = Bruzual & Charlot (2003) stellar library. This library 0.00390inallanalyses. Wequotedallerrorsat90%con- consists of model templates for 39 stellar populations fidence,unlessstatedotherwise. Wesummarizethemain with ages between 5.0 × 106 and 1.2 × 1010 yr, and results of the analysis in Table 1, and present the X-ray metallicitiesbetween0.008and0.05,withaspectralsam- data and best-fitting models in Figure 4. plingof3˚A.Thislibraryhasbeenusedtofitthecontinua We began our spectral modeling by examining the and measure the emission line fluxes of the Sloan Digi- NuSTAR and Chandra data separately. At this point tal Sky Survey (SDSS) galaxy spectra (Tremonti et al. wearenotyetconsideringnon-AGNcontributionstothe 2004). Weperformedourspectralsynthesisforthethree X-ray spectrum, such as from the two off-nuclear point spectra in xspec v12.8.2,4 assuming z = 0.00390, solar sources within the extraction region. We modeled the metallicity (Z⊙ = 0.02), and the Cardelli et al. (1989) two spectra in the 3–40keV and 2–8 keV bands, respec- extinction law. The best-fitted stellar spectra were then tively,using asimplepower-lawmodelwithGalacticab- subtracted from the observed spectra, which resulted in sorption. The bestfitting photonindices areveryflat: Γ residual spectra with emission lines only. We then ana- = −0.29 ± 0.38 for NuSTAR (C-stat/d.o.f = 229/237), lyzed the detected emission lines using the splot tasks in iraf. 5 We note that grouping the Chandra andNuSTAR data to a minimumof20and40countsperbin,respectively,andperforming 4 The XSPEC manual is available at thespectralfitsusingχ2minimizationyieldsconsistentresultswith http://heasarc.gsfc.nasa.gov/xanadu/xspec/XspecManual.pdf theC-statisticapproach. 6 A. Annuar et al. andΓ=0.82+0.88 forChandra(C-stat/d.o.f=125/126). 3.2. Physical Modeling −0.77 Aslightexcessofcountsjustabove6keVwasobservedin We proceeded to model the X-ray broadband spec- both the NuSTAR and Chandra spectra. This is likely trum of the potentially heavily obscured AGN in an indication of a fluorescent Fe Kα line emission, as- NGC 1448 using more physically motivated obscura- sociated with a spectral component produced by AGN tion models to better characterize the broadband spec- emission that is being reflected off a high column of gas. trum. The two models used are the torus model by The fluxes of the two spectra within the common en- Brightman & Nandra 2011 (Model T), and MYTorus ergy range of 3–8 keV measured by the model are f3−8 model by Murphy & Yaqoob 2009 (Model M). Details = 2.00+−00..2537 × 10−14 erg s−1 cm−2 for NuSTAR, and and results of the two models are described in Sections f3−8 = 3.50+−01..5364 × 10−14 erg s−1 cm−2 for Chandra. 3.2.2 and 3.2.3, respectively. In addition to these mod- These fluxes are consistent with each other within the els, we added extra components required to account for measurementuncertainties,indicating that there wasno non-AGN contributions to the X-ray spectrum, and to significant variability between the two observations. providea goodfitto the data asdescribedbelow,andin We then fitted the NuSTAR spectrum simultaneously Section 3.2.1. with the Chandra spectrumbetween 3 and40keV using At low energies,typically at E . 2 keV, where the di- a simple power-lawmodel anda Gaussiancomponentto rectemissionfromaheavilyobscuredAGNiscompletely model the possible Fe Kα emission line, with the line absorbed,otherprocessescandominate. Theseprocesses energy and width fixed to E = 6.4 keV and σ = 10 eV, includeX-rayemissionradiatedbyunresolvedoff-nuclear respectively. This model measured a photon index of Γ X-ray sources, thermal emission from a hot interstellar = 0.02+0.27 and Fe Kα line equivalent width of EW = medium, gas photoionized by the AGN, and scattered −0.28 2.1+1.0 keV (C-stat/d.o.f = 315/331). The flat photon emission from the AGN. Our data are not of sufficient −0.8 quality to accurately distinguish between these different indexandlargeEWmeasuredforthe FeKαline(EW& physical processes. Therefore, we simply parametrized 1 keV), are characteristic signatures for CT absorption. thelowenergypartofthe spectrumcoveredbyChandra To test for this, we model both spectra simul- using the thermal model “apec” (Smith et al. 2001),7 taneously from 3–40 keV with the pexrav model and a power-law component to simulate the scattered (Magdziarz & Zdziarski 1995), which has historically emission from the AGN. been used to model reflection-dominated AGN spectra. A constant parameter, C is often included when an- This model produces an AGN continuum which is re- alyzing data from multiple X-ray observatories simulta- flected from a slab-geometry torus with an infinite col- neouslytoaccountforthecross-calibrationuncertainties umn density, which is highly unlikely to represent the betweenthe variousinstruments, and to accountfor any actual AGN torus. However, it can provide a useful ini- significantvariabilityofthetargetsbetweentheobserva- tial test to investigate the AGN spectrum for CT obscu- tions. As detailed earlier, we do not find any significant ration. Because it does not self-consistently model the differences between the Chandra and NuSTAR spectra, fluorescence emission lines expected from a CTAGN, we indicating that there has not been significant variabil- includedaGaussiancomponentinthemodeltosimulate ity between the two observations. Based on the NuS- Fe Kα narrow line emission, which is the most promi- TAR calibration paper (Madsen et al. 2015), the cross- nent line produced by CTAGNs. The centroid energy calibrationuncertainty of Chandrawith respectto NuS- and width of the line were fixed to E = 6.4 keV and TARis≈1.1. We thereforedecidedto fixthis parameter σ = 10 eV, respectively. Due to the limited number of to the value found by Madsen et al. (2015) as a conser- counts, the inclination of the reflector was fixed to the vative approach. However, we note that allowing this default value set by the model; i.e., θ ≈ 63◦. How- inc parameter to vary in all models, returns results that are ever, we note that fixing θ to other values (e.g., near inc consistent with this value within the statistical uncer- the lower and upper limits of the model; θ ≈ 26◦ and inc tainties. 84◦, respectively) have insignificant effects on the pa- Our two models are described in xspec as follows: rameters obtained. We also fixed the reflection scaling factortoR =−1,tosimulateapurereflectionspectrum. Model T = constant ∗ phabs ∗ (apec In addition, we included the absorbed transmitted com- ponent of the AGN, simulating the Compton scattering + cons ∗ zpow + 2 × tbabs ∗ zpow + and photoelectric absorption by the torus using cabs and zphabs models, respectively. This model provides torus), (1) a decent fit to the data (C-stat/d.o.f = 318/330), and measured a column density of N (los) = 2.7+u × 1024 H −1.2 cm−2,consistentwithCTcolumn.6 Theintrinsicphoton Model M = constant ∗ phabs ∗ (apec indexinferredfromthebest-fitmodelisΓ=1.57±0.35, inagreementwiththe typicalintrinsicvalueforanAGN + cons ∗ zpow + 2 × tbabs ∗ zpow + (e.g., Burlon et al. 2011; Corral et al. 2011). The reflec- tion component of the model dominates the transmit- zpow ∗ mytz + myts + mytl). (2) tedcomponentatallspectralenergiesprobed,indicating that the spectra is reflection-dominated, consistent with The two additional absorbed power-law components being heavily obscured. (tbabs∗zpow) are included in the models to take into 7 We note that the apec component is mainly contributed by 6 “u”isunconstrained. theunresolvedemissionwithinthe20′′-radiusextraction region. Unveiling the buried nucleus in NGC 1448 with NuSTAR 7 account the two off-nuclear sources detected within the source component free to vary, and found that the value extractionregionin the Chandradata. We discuss these is consistent with zero, suggesting that the source does sources in the following section. not significantly contaminate our broadbandAGN spec- trum. We thereforedidnotincludethis source,andcon- 3.2.1. Off-nuclear X-ray Sources sequentlyanyofthefaintersourcesoutsidetheextraction In this section, we present the analysis of the two off- region, in our modeling of the AGN X-ray spectra. nuclear point sources detected in the 0.5–8 keV Chan- dra band within the 20′′-radius extraction region (see 3.2.2. Model T Figure 1). We extracted their spectra using extraction The torus model by Brightman & Nandra (2011) regionsof3′′ and2′′ fortheNEandNWsources,respec- (Model T) simulates obscuration by a spherical torus tively. Due to limited counts (∼30 counts at 0.5–8 keV with variable biconical polar opening angle (θ ) rang- tor for each source), we binned their spectra to a minimum ing between 26–84◦. The line-of-sight column density, of 5 counts per bin and optimized the fitting parameters N (los), through the torus, which is equal to the equa- H using C-statistic. We model both sources using a sim- torial column density N (eq), is independent of the in- H plepower-lawmodelabsorbedbyhostgalaxyabsorption clination angle (θ ), and extends up to 1026 cm−2, (tbabs), in addition to the Galactic absorption. inc an order of magnitude higher than that allowed by the The photon indices and intrinsic absorption columns MYTorus model (see Section 3.2.3). The model also measured by the fits are Γ = 1.02+1.84 and N (int) −1.28 H self-consistentlypredictstheComptonscatteringandflu- ≤ 2.19 ×1022 cm−2 (C-stat/d.o.f = 8.1/7) for the NE orescent Fe Kα and Fe Kβ lines, as well as Kα emission source, and Γ = 2.15+2.03 and N (int) ≤ 2.11 ×1022 from C, O, Ne, Mg, Si, Ar, Ca, Cr, and Ni, which are −1.30 H cm−2 (C-stat/d.o.f = 7.4/10) for the NW source. The commonly seen in CTAGNs. The model is available be- 0.5–8keV intrinsic luminosities for both sources,assum- tween 0.1 and 320 keV. ingthattheyarelocatedwithinNGC1448,are2.23+1.83 In addition to the torus model, we included several −0.85 othermodelcomponentsasdescribedearlierinSection3 × 1038 erg s−1 and 1.56+1.77 × 1038 erg s−1 for the NE −0.09 and 3.2.1 to provide a good fit to the data. Initially, we and NW sources, respectively. These luminosities are left both θ and θ free to vary; however, they were almost an order of magnitude lower than the threshold inc tor both unconstrained. We then fixed the torus inclination luminosityforultraluminousX-raysources(ULXs;LX & angle to the upper limit of the model (θ = 87◦), to 1039 erg s−1). inc simulate an edge-on inclination torus, and to allow for The best-fitted photon index for the NE source is po- the full exploration of θ (Brightman et al. 2015). The tor tentially flat (although with large uncertainties), and column density measurement is insensitive to θ when inc may suggest that it is a background obscured AGN. We θ >θ (Brightman et al.2015),andthereforeshould inc tor estimatetheprobabilityoffindinganAGNwithinaran- not significantly affect the N (los) measured. In addi- dom 20′′-radius region with at least the observed flux of H tion, we fixed the photon index to Γ = 1.9 as it could the NW source(f0.5−8 & 4× 10−15 ergs−1 cm−2) using not be constrained simultaneously with θ . tor theAGNnumbercountsofthe4MsChandraDeepField The model implied a column density of N (los) = H South (CDFS) Survey (Lehmer et al. 2012). Based on 4.2+u ×1024cm−2(C-stat/d.o.f=431/445). Thelower this, the probability of finding two or more background −1.7 limitofthis columndensity iswellabovethe CTthresh- AGNs within a 20′′-radius circular region is < 1%. In old, and therefore this model confirms that the AGN is addition, we do not find counterparts to these Chan- CT. The best-fit torus opening angle is θ = 45.9+33.1 drasourcesatotherwavelengths(e.g.,MIRandoptical). tor −18.9 Given their high Galactic latitudes (|b| ∼ 51◦), it is un- degrees, suggesting a geometrically thick torus. The model measured a small scattering fraction of 0.2+0.3 % likely that these are Galactic sources. Based on these −0.2 with respect to the intrinsic power-law. This is con- arguments, we conclude that the two off-nuclear X-ray sistent with that found in other obscured AGN (e.g., sourcesaremore likely to be X-raybinaries within NGC Noguchi et al. 2010; Gandhi et al. 2014; Gandhi et al. 1448. Given the stellar mass and SFR of the galaxy, we 2015). However we note that given the modest quality would expect to find ∼3 X-ray binaries with L0.5−8 > 1038 erg s−1 in NGC 1448 (Gilfanov 2004; Mineo et al. ofourdata,thescatteredpower-lawcomponentwillalso include contribution from other processes such as unre- 2012). This is consistent with our detections of two po- tential candidates within the central 20′′ circular radius solved X-ray binaries. Therefore, the true AGN scatter- ing fraction could be lower than this value. We inferred of the galaxy. We therefore included the two absorbed the intrinsic luminosities of the AGN in three different power-lawcomponents detailed above into Model T and bands based upon the best-fit parameters obtained, as M to account for their maximum contributions to our presented in Table 1. X-rayspectraofNGC 1448. We fixedallthe parameters at the best-fit values measured from the Chandra data. 3.2.3. Model M Due to the NuSTAR PSF, sources that lie outside the 20′′-radius extraction region could also contaminate our The MYTorus model by Murphy & Yaqoob (2009) spectrum. We tested whether these sources contribute (Model M) simulates a toroidal absorber geometry with a significant fraction of our X-ray broadband spectrum a fixed opening angle of θtor = 60◦ and variable inclina- by including the best-fit power-law model of the source tion angle. The line-of-sight column density, NH(los), is ∼30′′ westoftheAGNtoourX-rayspectralmodelingof derived from the measured inclination angle and equa- NGC 1448. This source is the brightest in a 50′′-radius torial column density, NH(eq), which is simulated only circular region around the AGN which corresponds to a up to a column density of N (eq) = 1025 cm−2. There H ∼70% NuSTAR ECF. We left the normalization of the is more freedom in exploring complex absorbing geome- 8 A. Annuar et al. Figure 4. Best-fittingmodelstothecombinedNuSTARandChandradataofNGC1448-ModelT(top)andModelM(bottom). Model Twas fitted between 0.5and 40keV,and ModelMwas fitted between 0.6and 40keV duetothe strongresidualsfoundat∼0.5keV for thismodel. Thedatahavebeenrebinnedtoaminimumof3σsignificancewithamaximumof25and100binsforNuSTARandChandra, respectively, for visual clarity. Color scheme: black (NuSTAR FPMA+B), red (Chandra). Plots on the left show the model components fitted to the data (dotted lines), with the total model shown in solid lines. We included an apec component and a scattered power-law component to model the emission at the softest energies, and two power-law components to model the two off-nuclear sources located within the extraction region in all models. The iron line modeled in Model T is labeled as “Fe Kα”, and the direct, scattered and line components of Model M arelabeled as mytz, mytsand mytl, respectively. The top panels of the left plots show the data and unfolded modelinE2FE units,whilethebottompanelsshowtheratiobetweenthedataandthefoldedmodel. Plotsontherightshowtheobserved (dashed lines) and intrinsicspectra (solid line) of the AGN component for each model. The slight offset between the red and black lines areduetothecross-calibrationuncertainties betweenChandraandNuSTAR(Madsenetal.2015). TheseplotsshowthatevenatE≥10 keV,thespectrathatweobservedforCTAGNsarestillsignificantlysuppressed(upto∼2ordersofmagnitudeforthecaseofNGC1448), demonstratingtheextremeofCTabsorption. Table 1 X-rayspectralfittingresultsforNGC1448. Component Parameter Units ModelT ModelM (1) (2) (3) (4) (5) apec kT keV 0.63+0.14 0.63+0.14 −0.33 −0.42 Absorber/Reflector NH(eq) 1024 cm−2 4.2+−u1.7 4.9+−u2.0 NH(los) 1024 cm−2 4.2+−u1.7 4.5+−u1.8 θinc deg 87f 78.6+−61.15.8 θtor deg 45.9−+3138..19 60f AGNContinuum fscatt % 0.2+−00..32 0.1+−00..21 Γ 1.9f 1.9f L0.5−2,obs 1039 ergs−1 0.2 0.2 L2−10,obs 1039 ergs−1 0.9 0.9 L10−40,obs 1039 ergs−1 11.7 11.8 L0.5−2,int 1040 ergs−1 2.6 5.6 L2−10,int 1040 ergs−1 3.5 7.6 L10−40,int 1040 ergs−1 3.5 7.6 C-stat/d.o.f. 431/445 429/440 Note. —(1)Modelcomponent;(2)parameterassociatedwitheachcomponent;(3)unitsofeachparameter;(4)best-fittingparameters forModelT(torusmodelbyBrightman&Nandra2011);(5)best-fittingparametersforModelM(MYTorusmodelbyMurphy&Yaqoob 2009). “f”isfixedparameterand“u”isunconstrainedparameter. DetailsofeachmodelaredescribedinSection3. Unveiling the buried nucleus in NGC 1448 with NuSTAR 9 try in the MYTorus model as it allows the user to dis- The interstellar extinction measured towards the stel- entangle the direct (mytz), scattered (myts) and line- lar population at the AGN position from the spectral emission(mytl)componentsfromeachother. Themytl synthesis modeling is much higher than at the optical componentsimulates the neutralFe Kα and Fe Kβ fluo- peak position, AAGN = 2.15+0.16 mag and Apeak = V −0.06 V rescencelines,andtheassociatedComptonshouldersfor 0.59+0.04 mag, respectively, further suggesting that the theAGN.Themodelisdefinedbetween0.5and500keV. −0.05 opticalemissionfromtheAGNisobscuredalongourline- However,we noticed strong residuals at ∼0.5 keV which of-sight. The total extinction measured from the whole could be attributed to the fact that we are probing the galaxy spectra, Agal = 0.73 ± 0.06 mag, is ∼3× lower lower limit of the model. Therefore for this model, we V thanthatmeasuredattheAGNposition,demonstrating restricted our fit to above 0.6 keV. thatthe extinctionacrossthe galaxyis non-uniformand Forsimplicity,wefittedtheAGNspectrumbycoupling inhomogeneous. all the parameters of the scattered and fluorescent line We measured the emission line fluxes and ratios for components to the direct continuum component. The the AGN, optical peak and the whole galaxy from relative normalizations of myts (A ) and mytl (A ) S L the extinction-corrected residual spectra (see Table 2). with respect to mytz (A ) were set to 1. At first, we Z Firstly, we calculated the extinctions toward the AGN left the photon index free to vary; however, it reached narrow line region (NLR) from the Balmer decrements the lower limit of the model, suggesting that it is not at the three positions. We assumed an intrinsic Balmer well constrained. Therefore, we fixed the value of this decrement of (Hα/Hβ) = 2.86, which corresponds to parameter to Γ = 1.9. int ModelMalsogivesasgoodafittothedataasModelT, a temperature of T = 104 K and an electron density withC-Stat/d.o.f=429/440. Usingthismodel,wemea- ne = 102 cm−3 for Case B recombination (Osterbrock sured an equatorial column density of NH(eq) = 4.9+−u2.0 1989). Based on these, we found AAVGN =1.89 ± 0.03 × 1024 cm−2. The model measured a high inclination mag, ApVeak = 0.74 ± 0.09 mag, and AgVal = 1.03 ± 0.04 angle of θ = 78.6+6.5 degrees, close to the maximum mag. The high extinction measured at the AGN posi- inc −11.8 tion provides strong evidence that the AGN is heavily valuefixedinModelT.ThecorrespondingN (los)value H obscured at optical wavelength by the host galaxy. is well within the CT regime, within the uncertainties, N (los)=4.5+u ×1024cm−2,andagreesverywellwith Using the emission line ratios tabulated in Table 2, H −1.8 we constructed the Baldwin-Philips-Terlevich(BPT) di- thatmeasuredbyModelT.8 Thismodelalsomeasureda agnostic diagrams (Baldwin et al. 1981) for NGC 1448. small scattering fraction; i.e., 0.1+−00..21 %, consistent with This is shown in Figure 6. Based on these diagrams, all Model T. As with Model T, we determined the intrin- three spectra fall within the region of Seyfert galaxies, sic luminosity of the AGN in different bands from the withtheAGNhavingslightlyhigheremissionlineratios. best-fitting parameters (see Table 1). These results provide the first identification of the AGN 4. MULTIWAVELENGTHRESULTS in NGC 1448 at optical wavelengths, and also confirm the position of the AGN in the galaxy. The total galaxy In this section, we present the results from the optical spectrumis locatedcloseto the Hii/lowionizationemis- andMIRobservationsofNGC1448asdetailedinSection sion line regions (LINERs) of the BPT diagrams. This 2.3 and 2.4, respectively. We use these data to calculate might explain why the AGN was misidentified as an Hii the properties of the AGN at different wavelengths to region in Veron-Cetty & Veron (1986); i.e., due to host compare with the results of our X-ray spectral analysis. galaxy contaminations as a result of the larger slit size 4.1. Optical used to extract the spectrum. TheseparationbetweentheAGNandtheopticalpeak, As described inSection 2.3,we extractedopticalspec- ∼225 pc, is consistent with the NLR scale observed tra from three different regions along the slit; i.e, the in AGNs with comparable luminosities as NGC 1448 AGN, optical peak and across the whole galaxy. The (Bennert et al.2006). Theapparentseparationobserved extracted spectra, along with the fits obtained from the between the two can be attributed to the presence of a spectral synthesis modeling (see Section 2.3), and the dust lane which obscures part of the system along our fit residuals, are shown in Figure 5. As shown in this line-of-sight. figure, no significant residuals were left after fitting the The intrinsic [Oiii]λ5007˚A flux of the AGN corrected observedspectrawithstellarpopulationtemplates,with- for the Balmer decrement is about an order of mag- out the need ofadditional AGN continuum components. nitude higher than the observed luminosity (see Table This suggests that the AGN optical continuum is highly 2). The [Oiii] line emission in AGN is mostly pro- obscured. The spectral analysis revealed that the stel- duced in the NLR due to photoionization from the cen- lar emission from all three regions is due to both young tral source. Since this region extends beyond the torus, (5 Myr) and old (5 Gyr) stellar populations, dominated it does not suffer from nuclear obscuration like the X- by the latter with ∼75–84% contributions measured be- ray emission. However as we demonstrated earlier, it tween the three spectra. Although the optical peak is can suffer from significant optical extinction from the where the total optical emission is the brightest, the emission lines are the strongest at the Chandra posi- host galaxy. Indeed, in extreme cases, the host galaxy obscuration can be so high that the optical Balmer tion, providing independent evidence of the AGN loca- decrement only provides a lower limit on the extinction tion within the galaxy. (Goulding & Alexander 2009). 8 The line-of-sight column density for Model M was derived Using our intrinsic [Oiii] luminosity measured for the usingequation 3.1of Murphy&Yaqoob (2009), and assumingan AGN, L = (6.89 ± 0.08) × 1038 erg s−1, we pro- inclinationangleatthebest-fitvalue. [OIII] 10 A. Annuar et al. Figure 5. The optical spectra for the AGN (top), optical peak (middle) and the whole galaxy (bottom) extracted from the aperture regionsshowninFigure2. Blackaretheobservedspectra,bluearethebest-fittedstellartemplatespectra,andredaretheresidualspectra obtained after subtracting the best-fitted stellar template spectra from the observed spectra. Detected emission and absorption lines are labeledwithdottedlines. Theabsorptionlineslabeledwith⊕areTelluric. ceeded to compare the intrinsic X-ray luminosity deter- Table 2 minedfromourX-rayspectralfittingwiththatpredicted OpticalemissionlinefluxesandratiosforspectrashowninFigure from the X-ray:[Oiii] intrinsic luminosity relationship of 5. Panessa et al. (2006). Based on this correlation, we in- EmissionLine AGN Optical Total fer an intrinsic X-ray luminosity of L2−10,int = (0.2–4.5) Peak Galaxy × 1040 erg s−1 (see footnote 9). This is consistent with Hα 3.79±0.08 1.51±0.17 15.2±0.50 thatdeterminedfromourX-rayspectralfittings,provid- Hβ 0.65±0.05 0.40±0.07 3.60±0.32 [Oiii]λ4959 1.85±0.05 0.74±0.10 3.51±0.38 ing confidence that our analysis is reliable. [Oiii]λ5007(obs.) 5.38±0.06 2.07±0.09 9.78±0.37 4.2. Mid-Infrared [Oiii]λ5007(int.) 43.7±0.49 16.8±0.73 79.4±3.00 [Oi]λ6300 0.38±0.08 <0.34 2.22±0.60 The high spatial resolution MIR observation by [Nii]λ6549 1.51±0.08 0.58±0.19 3.44±0.55 [Nii]λ6583 4.71±0.06 1.73±0.10 11.7±0.39 Gemini/T-ReCSmeasureda12µmluminosityofλL12µm [Sii]λ6717 1.36±0.09 0.40±0.27 3.41±0.87 =(4.90±0.93)×1040ergs−1forNGC1448(seeSection [Sii]λ6731 1.12±0.09 0.57±0.29 3.14±0.93 2.4). MIRluminosityispredictedtoprovideanaccurate LineRatio estimate for the intrinsic luminosity of the AGN. This is Hα/Hβ 5.83±0.47 3.78±0.79 4.22±0.40 [Oiii]λ5007/Hβ 8.28±0.64 5.18±0.93 2.72±0.26 because, the absorbed X-ray radiation from the central [Oi]λ6300/Hα 0.10±0.02 <0.23 0.15±0.04 engine are mostly re-emitted in the MIR by the torus. [Nii]λ6583/Hα 1.24±0.03 1.15±0.15 0.77±0.04 Tofurthertesttheabsorption-correctedX-rayluminosity [Sii](λ6717+λ6731)/Hα 0.65±0.04 0.64±0.27 0.43±0.08 measured from our spectral analysis, we also compared Note. —Thefluxes aregiveninunits of10−15 ergs−1 cm−2. theluminositymeasuredfromouranalysiswiththatpre- Theintrinsic[Oiii]λ5007fluxwascorrectedfortheBalmerdecre- dictedfromtheX-ray:MIRcorrelationconstructedbased ment. upon high angular resolution MIR observations of local Seyferts, which has been shown to trace the intrinsic X- ray luminosity of AGN very well (e.g., Horst et al. 2008; Gandhi et al. 2009; Asmus et al. 2015). The correlation with respect to the recent Asmus et al. (2015) relation 9 Thegiven luminosityrange accounts for the mean scatter of is shown in Figure 7. The figure shows that the ob- thecorrelation. served 2–10 keV luminosity of the AGN in NGC 1448 is