Obscured Narrow-Line Seyfert 1 Galaxy Candidate Mrk 1388 with Nonthermal Jets
aa r X i v : . [ a s t r o - ph . GA ] N ov PASJ:
Publ. Astron. Soc. Japan , 1– ?? , h in press i c (cid:13) Obscured Narrow-Line Seyfert 1 Galaxy Candidate Mrk 1388with Nonthermal Jets
Akihiro D OI The Institute of Space and Astronautical Science, Japan Aerospace Exploration Agency,3-1-1 Yoshinodai, Chuou-ku, Sagamihara, Kanagawa 229-8510 (Received 2014 September 1; accepted 2014 October 21)
Abstract
Mrk 1388 has an unusual Seyfert nucleus that shows narrow emission-line components without broad ones,but shows a strong featureless continuum and strong iron-forbidden high-ionization emission lines. The apparentcoexistence of type-1/2 characteristics is potentially attributed to a heavily obscured broad-line region or to anintermediate-mass black hole with a broad-line component intrinsically narrower than those of typical narrow-lineSeyfert 1 (NLS1) galaxies. Our observation using very-long-baseline interferometry (VLBI) reveals high-brightnessradio emission from nonthermal jets from an active galactic nucleus (AGN) with a significant radio luminosity.Furthermore, we investigate the radial profile of the host galaxy using a Hubble Space Telescope (HST) image,which shows a S ´e rsic index suggestive of a pseudobulge. Using the VLBI and HST results, which are essentially notaffected by dust extinction, three individual methods provide similar estimates for the black hole mass: (0 . – . × M ⊙ , . × M ⊙ , and . × M ⊙ . These masses are in a range that is preferential for typical NLS1 galaxiesrather than for intermediate-mass black holes. Based on the estimated masses, the full width at half maximum F W HM (H β ) of approximately 1200–1700 km s − should have been seen. The scenario of a heavily absorbedNLS1 nucleus can explain the peculiarities previously observed. Key words: galaxies: active — galaxies: supermassive black holes — galaxies: jets — radio continuum: galax-ies — galaxies: Seyfert — galaxies: bulges — galaxies: individual (Mrk 1388)
1. Introduction
Mrk 1388 has an unusual Seyfert nucleus that shows thecharacteristics of both type-1 and type-2 Seyfert galaxies (here-inafter Sy1 and Sy2, respectively). Permitted lines suchas the hydrogen Balmer emission lines of H α and H β arenarrow ( and km s − , respectively) with essentiallythe same widths as forbidden lines at the lower end ofthe line-width distribution for Seyfert galaxies (Osterbrock,1985). The ratio of an oxygen-forbidden line and the Balmerline, [O III ] λ /H β = 11 . , is much more typical of Sy2([O III ] λ /H β > for a Sy2s criterion). The iron emissionline of Fe II is very weak, as is typical for Sy2s (Osterbrock& Pogge, 1985). On the other hand, Osterbrock (1985)observed a strong featureless continuum (as for Sy1) andstrong iron-forbidden high-ionization emission lines such as([Fe VII ] λ + [Fe X ] λ )/[O III ] λ . , which ishigher than typical values for Sy1s but lower than the mostextreme examples. Mrk 1388 was included by Osterbrock& Pogge (1985) as a narrow-line Seyfert 1 (NLS1) galaxyon the basis of the strong high-ionization emission lines of[Fe VII ] λλ , , and [Fe X ] λ , whereas Osterbrock& Pogge (1984) noted that Mrk 1388 may be most simplydescribed as an unusually high-ionization Sy2 galaxy . Thecurrent classification for NLS1 galaxies (Pogge, 2000) re-quires (1) narrow permitted lines only slightly broader thanforbidden lines, (2) F W HM (H β ) < km s − , and (3) Mrk 1388 is listed as a type-1.9 Seyfert galaxy in the catalogue of quasarsand active nuclei: 13th edition (V´eron-Cetty & V´eron, 2010), which refersto Osterbrock (1985). [O III ] λ /H β < ; however, exceptions are allowed if strong[Fe VII ] and [Fe X ] are also present, unlike what is seen in Sy2galaxies. Thus, Mrk 1388 lacks the broad-line property to beclassified as an NLS1 galaxy. The narrow component Pa β ofthe hydrogen Paschen emission line is detected in the J -band inthe near infrared; however, no broad component appears in Pa β (Goodrich et al., 1994; Veilleux et al., 1997), which suggests aheavily obscured ( A V > ∼ mag) broad-line region (BLR) or abroad-line component intrinsically narrower than those of typi-cal NLS1 galaxies because of an intermediate-mass black hole( – M ⊙ ) for active galactic nucleus (AGN) activity.Radio observations are not affected by dust extinction to-ward the nucleus through our line of sight. Based on obser-vations by the Very Large Array (VLA) A-array configurationat 1.49 and 4.86 GHz (Ulvestad et al., 1995), Mrk 1388 is anunresolved radio source of nonthermal emission with a power-law spectrum with an index α = − . ( S ν ∝ ν α , where S ν isthe flux density at frequency ν ). The standard infrared/radioratio q (Condon et al., 1995) for Mrk 1388 is significantlysmall, which suggests that AGN dominates star-forming activ-ity (Dennefeld et al., 2003). No polycyclic aromatic hydrocar-bon (PAH) feature, which also provide excellent diagnostics todistinguish starburst and AGN energy sources, is detected forMrk 1388 (Dennefeld et al., 2003); this is consistent with thelow infrared/radio ratio q . The appreciable jet activity is pow-ered by proportionate mass accretion with limits lesser than theorder of the Eddington luminosity, which is proportional to theblack hole mass. Thus, radio jet clues to the mass of the centralblack hole.The properties of the bulge component correlate with the Obscured NLS1 Candidate Mrk 1388 [Vol. ,mass of the central black hole (Ferrarese & Merritt, 2000;Gebhardt et al., 2000), although they do not correlate witha pseudobulge or galaxy-disk component (Kormendy et al.,2011). Mrk 1388 was described as “very compact” (Zwickyet al., 1963). Markaryan et al. (1979) discovered this galaxyand described it as spherical, with diffuse edges, and appar-ently with a star-like nucleus. However, its morphology isactually slightly elongated, and thus is elliptical rather thanspherical (Osterbrock & Pogge, 1985). The Hubble type is“E?” in the NASA/IPAC Extragalactic Database (NED); the -mag/arcsec radius is only . kilo-parsecs (kpc). Mrk 1388is an isolated galaxy with almost no asymmetry or distortion(Xanthopoulos & De Robertis, 1991), suggestive of long sincea last merging event. Thus, the host of Mrk 1388 may be asmall and isolated elliptical galaxy. This peculiar galaxy mayprovide important clues for extreme-parameter regions con-necting bulges and central black holes in the scenario of co-evolution.The present paper discusses the mass of the central blackhole in Mrk 1388 based on very-high-angular-resolution radioand optical studies that used the very-long-baseline interferom-etry (VLBI) technique and the Hubble Space Telescope (HST),respectively. In Section 2, we describe our VLBI observationand data reduction and analysis of an HST image. The resultsare presented in Section 3, and their implications are discussedin Section 4. Throughout this paper, we use Lambda cold darkmatter ( Λ CDM) cosmology with H = 70 . km s − Mpc − , Ω M = 0 . , and Ω Λ = 0 . . The redshift is z = 0 . ± . (Sloan Digital Sky Survey Data Release 7); the lu-minosity distance is Mpc, and the angular-size distance is . Mpc; 1 milliarcsecond (mas) corresponds to a projectedlinear scale of . pc at the distance to Mrk 1388.
2. Data and Data Analysis
We observed Mrk 1388 on May 9, 2005 in the L -band us-ing ten antennas of the Very Long Baseline Array (VLBA)at the National Radio Astronomy Observatory (NRAO). Thedata were obtained with observation code BD106, in whichseven objects of nearby radio-quiet NLS1 galaxies were alsoobserved. The results of these observations were previouslypublished (Doi et al., 2013). A left-circular polarization wasreceived at a center frequency of . GHz with a total band-width of MHz. Because Mrk 1388 is a faint radio source,we observed in phase-referencing mode, which allowed us toderive calibration parameters for instrumental and atmosphericeffects from observations of the nearby strong compact radiosource (“calibrator”) (Beasley & Conway, 1995) J1455+2131separated by . ◦ from the target. The period of the antennanodding cycle was min, and the total on-source time was ap-proximately min for the target. Standard calibration proce-dures for VLBA phase referencing were applied during datareduction using the Astronomical Image Processing System( AIPS ; Greisen 2003) software, developed by the NRAO. Thedetails of data reduction are the same as for Doi et al. (2013).We set the phase-tracking center for Mrk 1388 to the positiondetermined using the VLA A-array at GHz with an accu-racy of approximately . ′′ (Ulvestad et al., 1995). An emis- sion was initially found at a position offset by less than sev-eral tens of mas. After shifting the mapping center, we decon-volved the images using the task IMAGR (CLEAN), and wemeasured the astrometric positions of the emission using thetask
JMFIT . Subsequently, we self-calibrated in phase with athreshold signal-to-noise ratio of . using the task CALIB . Wesmoothed and interpolated the solutions before applying to thedata. The deconvolution and self-calibration algorithms wereinteractively applied several times.We used the
Difmap software (Shepherd et al., 1994) tomake the final image using the CLEAN algorithm from the cal-ibrated visibilities. To retrieve both compact and diffuse com-ponents to the extent possible, we used step-by-step uniform,natural, and ( u , v )-tapered weighting functions for CLEAN.The final VLBA image is shown in Figure 1. Image parame-ters are listed in Table 1. We used an HST archival data set (Dataset = U2E62101P),which contains an the observation obtained August 25, 1994using a s exposure with the F606W filter in the field-of-view of Planetary Camera 1 (PC1) in the Wide Field andPlanetary Camera 2 (WFPC2). The archival data were cal-ibrated by the On-The-Fly Reprocessing (OTFR) system de-veloped at the Space Telescope Science Institute. The two-dimensional radial profile of Mrk 1388 was analyzed using
GALFIT (Peng et al., 2002). A point spread function (PSF)was made using the web-based modeling tool Tiny Tim withfour-times oversampling, which was then downloaded and ap-plied to the GALFIT analysis. The magnitude Zeropoint wascorrected for F606W ( . mag). We initially tried model-ing with a S ´e rsic + sky bias model, a two-S ´e rsic + sky biasmodel, and a S ´e rsic + exponential disk + sky bias model, butthese combinations failed. However, a two-S ´e rsic + exponen-tial disk + sky bias model provided a set of plausible solutions(Table 2).The first and second S ´e rsic components have parameterswith S ´e rsic indices n = 1 . and n = 0 . , respectively. Thelatter is significantly resolved ( R e = 1 . pixels) with respectto the PSF. This component appeared clearly at the center ofgalaxy in the residual image of the single S ´e rsic case; if we usea PSF component instead we cannot represent this central ex-cess. The exponential disk has the scale length R s = 1 . kpc.Figure 2 shows the radial profiles of the data and the fits.
3. Result
The detected radio emission of Mrk 1388 is compara-ble or lower than radio powers emanated from the mostradio-luminous starbursts, which are approximately . – . W Hz − at GHz (equivalent to – . ergs s − )(Smith et al., 1998). However, the VLBI detection gives abrightness temperature of . K (Table 1), which is too highto attribute to any stellar origin. We can conclude, therefore,that the detected radio emission is relevant to the activity ofAGN, which is presumably nonthermal synchrotron jets on thepc or sub-pc scales. o. ] Akihiro Doi 3 Table 1.
Results of VLBA observation.Object I . S . σ θ maj θ min P.A. log( T B / K) L Astrometric position (J2000.0)(mJy/b) (mJy) (mJy/b) (mas) (mas) ( ◦ ) (erg s − ) R.A. Decl.(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)Mrk 1388 . ± . . ± . − . × h m . s + ◦ ′ . ′′ α = − . (Ulvestad et al., 1995); Cols. (10)–(11) astrometric positions determined by VLBA observation.
50 0 -50Right Ascension (mas) -50 0 50
10 pc
MRK 1388
Fig. 1.
VLBA image for Mrk 1388 at . GHz in the region of × mas . Contours are separated by factors of √ beginningat three times the rms noise. Negative and positive contours are repre-sented by dashed and solid curves, respectively. The half-power beamsize is given in the lower-left corner. Angular scale corresponding to alinear scale of pc is given in the lower-right corner. No clearly elongated structure suggestive of jets is evident inthe limited dynamic range of the image (Figure 1). However,the deconvolution shows a significantly resolved structure,which is also supported by the significant discrepancy be-tween peak intensity and integrated flux density (Table 1) inthe self-calibrated VLBA image. The VLBA retrieved %of the expected 1.7 GHz total flux density estimated from a1.4 GHz VLA flux density of . mJy and a spectral indexof α = − . (Ulvestad et al., 1995). The remaining por-tion indicates the presence of low-brightness components withlinear scales greater than pc, which presumably originatedin extended AGN jets and/or in supernova remnants. Thus,we confirm that, in the radio regime, approximately half ormore of the energy source in Mrk 1388 originates in the AGN.Previous suggestions of the predominance of the AGN com-pared with star-forming activity, which were based on the lowinfrared/radio ratio q and the negative detection of PAH fea-tures (Dennefeld et al., 2003), are confirmed by our VLBI ob-servation.The GALFIT modeling determined well the two- dimensional profile of the host galaxy of Mrk 1388 (Figure 2).The notable feature characterizing the structure of this galaxyis the first S ´e rsic component with an index n = 1 . , whichis fairly small and less than the transition between classicalbulges and pseudobulges at n ≈ (Fisher & Drory, 2008;Mathur et al., 2011). For the secondary S ´e rsic component,although the parameter determined ( n = 0 . ) for the centralexcess might not be so robust, any larger index or a PSF modelinstead could hardly represent this sizable component. Itsphysical scale (approximately 30 pc) is suggestive of a nuclearstar cluster, as reported at the centers of a majority of spiral andlower-mass elliptical galaxies (B¨oker et al., 2002; Cˆot´e et al.,2006). While a large fraction of total magnitude was modeledwith a low-brightness exponential disk, no spiral-like patternwas seen in a residual image. Thus, Mrk 1388 comprises onlyspheroidal components.The VLBA observation provides a very accurate astromet-ric position (Table 1) with an uncertainty < ∼ mas, which de-pends predominantly on the accuracy of the cataloged positionof calibrator source in the International Celestial ReferenceFrame-2 (ICRF2; Ma et al. 2009); ∆R . A . = 0 . mas and ∆Dec . = 0 . mas for J1455+2131. The HST position ofMrk 1388 was apparently offset by 0 . ′′ ′′ – ′′ or worse. Thus, the comparisons with radio-astrometricpositions of VLA (approximately . ′′ ) or VLBA ( < ∼ mas)are meaningless. No bright star or galaxy to refer an astro-metric position against Mrk 1388 exists in the field of view ofWFPC2.
4. Discussion
The discussion herein focuses on the mass of the cen-tral black hole in Mrk 1388. The black hole mass is oneof key remaining parameters to discriminate between an un-usual Sy2 and a Sy1 with an intermediate-mass black hole forMrk 1388. An example of VLBI detection exists in the Sy1nucleus of NGC 4395 with an intermediate-mass black hole( . × M ⊙ ; Filippenko & Ho 2003; Wrobel & Ho 2006),which indicates nonthermal jets associated with a central en-gine of intermediate mass.The radio jet is relevant to mass accretion onto the cen-tral black hole. A fundamental plane of black hole activ-ity was discovered in three-dimensional (logarithmic) spacecomprising the nuclear radio luminosity L r (in ergs s − ) at GHz, the X-ray luminosity L X (in ergs s − ) at – keV,and the black hole mass M BH (in M ⊙ ) ranging from to M ⊙ for samples of X-ray binaries, our Galactic Center, Obscured NLS1 Candidate Mrk 1388 [Vol. , RA = 14h50m37s RA = 14h50m37sDec = 22:44:00 Dec = 22:44:20
Radius (arcsec) I n t en s i t y ( C oun t s ) Fig. 2.
Top panel shows HST image of Mrk 1388. The image is ori-ented at − . ◦ from North. Bottom panel shows slice pro-file along semi-major axis of bulge component. Open circles representpixel data in counts in a pixel size of 0 . ′′ ´e rsic index of n = 1 . andan effective radius of R e = 0 . kpc. Dotted curve represents nuclearstar cluster of a fitted function with n = 0 . and R e = 0 . kpc.Dashed curve represents a fit to an exponential disk with a scale lengthof R s = 1 . kpc. Thin solid line represents a sky bias fit. Gray solidcurve represents the sum of them of dotted, dashed, solid, and thin solidcurves. As a reference, the dot-dashed curve represents PSF used fordeconvolution (no PSF component was used in GALFIT). low-luminosity AGNs, Seyfert galaxies, and quasars: log L r =0 .
60 log L X + 0 .
78 log M BH + 7 . (Merloni et al., 2003). Thistendency of jet activity can be attributed to nonlinear depen-dencies on the black hole mass and the accretion rate (Heinz& Sunyaev, 2003). Low-luminosity AGNs with low-accretionrates shows a relatively tighter correlation (K¨ording et al.,2006) and well-investigated by many studies (e,g, Yuan et al.,2009; de Gasperin et al., 2011; Plotkin et al., 2012). G¨ultekinet al. (2014) showed the capability of the fundamental planeto estimate masses even for less than approximately M ⊙ athigh accretion rates. According to this relationship, the blackhole mass of Mrk 1388 can be constrained by observed radioand X-ray emissions.The ROSAT All-Sky Survey (ASS; . – . keV) detectedX-rays from Mrk 1388 with an intrinsic luminosity L (0 . – . × ergs s − and a photon index Γ = 1 . ± . with an absorption column density N H = 9 . × cm − , as-suming H = 50 km s − Mpc − (Dennefeld et al., 2003). Weconvert their result into the H -corrected 2–10 keV luminosity Table 2.
Results of radial profile fit for HST image.Component Bulge Nuclear star cluster Exp. diskS ´e rsic index n R e or R s (Pixel) ◦ ) − . − . − . Axis ratio ( b/a ) 0.65 0.87 0.81Magnitude (mag) 17.32 17.73 15.13 The effective radius R e for the S ´e rsic profiles or the scale length of R s ( = 1 . R e for the n = 1 case of a S ´e rsic profile) for the exponential diskprofile. L X = 1 . × ergs s − . However, we must verify the poorstatistics of the photon index Γ (i.e., the absorption columndensity N H ). The ROSAT ASS Faint Source Catalog (Vogeset al., 2000) provides the hardness ratios HR . ± . (corresponding to Γ = − . − . − . at . – . keV) and HR . ± . ( Γ = − . +0 . − . at . – . keV), which suggesta significant N H with a negligible thermal component. Thenegative detection of the broad component of Pa β (Goodrichet al., 1994; Veilleux et al., 1997) indicates A V > ∼ mag,which is equivalent to N H > ∼ × cm − if we simplyadopt the Galactic ratio N H /A V ≈ . × cm − mag − (Predehl & Schmitt, 1995). Therefore, the ROSAT ASS lumi-nosity presented above must be an underestimate, and shouldgive a lower limit for the intrinsic X-ray luminosity. On theother hand, the negative detection with the RXTE XSS of < . × − ergs cm − s − at a hard X-ray regime of 3–20 keV (Heckman et al., 2005) suggests L (2 –
10 keV) < . × ergs s − . By combining them, we obtain . × 10 keV) > . × ergs s − inferred from the soft X-rayregime and the range L (2 – 10 keV) < . × ergs s − in-ferred from the hard X-ray regime are not so unreliable.According to the fundamental plane, the – keV luminos-ity range and the 5-GHz luminosity L r = 8 . × ergs s − from our VLBI observation (Table 1) impose a black hole massof . × < M BH < . × M ⊙ . This result includes theboundary between the intermediate-mass and the supermassiveblack hole near the lower end of the AGN mass function (Zhouet al., 2006; Greene & Ho, 2007; Doi et al., 2012).o. ] Akihiro Doi 5Alternatively, based on our HST image analysis(Section 2.2), we also estimate the black hole mass using therelationship between the black hole mass and the S ´e rsic index n (e.g., Graham & Driver 2007). The S ´e rsic index of the majorcomponent in Mrk 1388 is n = 1 . . Graham & Driver (2007)presented log M BH = 7 . 98 + 3 . 70 log( n/ − . n/ with a total absolute scatter of . dex from galax-ies; this empirical relationship predicts a black hole mass M BH = 1 . × M ⊙ for Mrk 1388, which is consistent withthe value derived based on the fundamental plane discussedabove.The third method we use is the improved version of the re-lationship between the black hole mass and the bulge luminos-ity (G¨ultekin et al., 2009); namely, log ( M BH /M ⊙ ) = 8 . 95 +1 .