Gemini IFU, VLA and HST observations of the OH Megamaser galaxy IRASF23199+0123: the hidden monster and its outflow
Carpes Hekatelyne, Rogemar A. Riffel, Dinalva Sales, Andrew Robinson, Jack Gallimore, Thaisa Storchi-Bergmann, Preeti Kharb, Christopher ODea, Stefi Baum
aa r X i v : . [ a s t r o - ph . GA ] N ov MNRAS , 1–12 (2017) Preprint 14 September 2018 Compiled using MNRAS L A TEX style file v3.0
Gemini IFU, VLA and HST observations of the OHMegamaser galaxy IRASF23199+0123: the hidden monsterand its outflow
C. Hekatelyne, ⋆ Rogemar A. Riffel, Dinalva Sales, Andrew Robinson, Jack Gallimore, Thaisa Storchi-Bergmann, Preeti Kharb, Christopher O’Dea, , Stefi Baum, , Departamento de F´ısica, CCNE, Universidade Federal de Santa Maria, 97105-900, Santa Maria, RS, Brazil Instituto de Matem´atica, Estat´ıstica e F´ısica, Universidade Federal do Rio Grande, Rio Grande 96203-900, Brazil School of Physics and Astronomy, Rochester Institute of Technology, 84 Lomb Memorial Drive, Rochester, NY 14623, USA Department of Physics, Bucknell University, Lewisburg, PA 17837, USA Departamento de Astronomia, Universidade Federal do Rio Grande do Sul. 9500 Bento Gon¸calves, Porto Alegre, 91501-970, Brazil National Centre for Radio Astrophysics, Tata Institute of Fundamental Research, S. P. Pune University Campus, Post Bag 3,Ganeshkhind, Pune 411 007, India Department of Physics and Astronomy, University of Manitoba, Winnipeg, MB, R3T 2N2, Canada School of Physics & Astronomy, Rochester Institute of Technology, 84 Lomb Memorial Dr., Rochester, NY 14623, USA. Center for Imaging Science, Rochester Institute of Technology, 84 Lomb Memorial Dr., Rochester, NY 14623, USA.
Accepted XXX. Received YYY; in original form ZZZ
ABSTRACT
We present Gemini Multi-Object Spectrograph (GMOS) Integral field Unit (IFU),Very Large Array (VLA) and Hubble Space Telescope (HST) observations of theOH Megamaser (OHM) galaxy IRASF23199+0123. Our observations show that thissystem is an interacting pair, with two OHM sources associated to the eastern(IRAS 23199E) member. The two members of the pair present somewhat extendedradio emission at 3 and 20 cm, with flux peaks at each nucleus. The GMOS-IFU ob-servations cover the inner ∼ α and [N ii ] λ ii ]+H α narrow-band image, being more extended along the northeast-southwest di-rection, as also observed in the continuum HST F814W image. The GMOS-IFU H α flux map of IRAS 23199E shows three extranuclear knots attributed to star-formingcomplexes. We have discovered a Seyfert 1 nucleus in this galaxy, as its nuclear spec-trum shows an unresolved broad (FWHM ≈ − ) double-peaked H α compo-nent, from which we derive a black hole mass of M BH = 3.8 + . − . × M ⊙ . The gaskinematics shows low velocity dispersions ( σ ) and low [N ii ]/H α ratios for the star-forming complexes and higher σ and [N ii ]/H α surrounding the radio emission region,supporting interaction between the radio-plasma and ambient gas. The two OH masersdetected in IRASF23199E are observed in the vicinity of these enhanced σ regions,supporting their association with the active nucleus and its interaction with the sur-rounding gas. The gas velocity field can be partially reproduced by rotation in a disk,with residuals along the north-south direction being tentatively attributed to emissionfrom the front walls of a bipolar outflow. Key words: galaxies: ISM – galaxies:ULIRGs – galaxies: dynamics – galaxies: indi-vidual (IRASF23199+0123) ⋆ E-mail: [email protected] (Ultra)luminous infrared galaxies ([U]LIRGs) are among themost luminous objects in the universe showing infrared (IR) c (cid:13) C. Hekatelyne et al. luminosities of L IR > L ⊙ . These objects are believed torepresent a key stage in the evolution process of galaxies inwhich tidal torques associated with mergers drive gas intothe galaxy core, leading to the feeding/triggering of nuclearstarbursts or the fuelling of embedded active galactic nuclei(AGN - e.g. Sales et al. 2015).These merging systems provide a conducive environ-ment for OH maser emission and approximately 20% of[U]LIRGs contain extremely luminous OH masers, emittingprimarily in the 1667 and 1665 MHz lines with luminosi-ties 10 − L ⊙ (Lo 2005; Darling & Giovanelli 2002). TheOH megamasers (OHMs) are commonly associated to merg-ing systems, but the environment that produces this phe-nomenon is still not completely understood. Many OHMhosts present a composite spectrum, showing both AGNand starburst features. A possible explanation for these fea-tures is that they originate in a central AGN, contami-nated by emission of circum-nuclear star-forming regions,as the sampling of the observations usually correspondsto more than one kpc at the galaxies. Alternatively, theOHM galaxies could represent a transition stage betweena starburst and the eruption of an AGN, as suggested byDarling & Giovanelli (2006).Considering the above scenario, it becomes relevant toinvestigate the nature of the gas ionization source of OHmegamaser galaxies. In this paper, we present Gemini Multi-Object Spectrograph (GMOS) Integral Field Unit (IFU) ob-servations, VLA continuum data and Hubble Space Tele-scope (HST) narrow and broad band images of the galaxyIRASF23199+0123, which is an interacting pair of ULIRGsthat presents OH megamaser emission. This galaxy is part ofa sample of 15 OH Megamaser galaxies, for which we have al-ready HST images from a project that has the overall goal ofrelating the merger state and OH maser properties to AGNand Starburst nuclear activity. We have selected targets forIFU observations from the 15 galaxies observed with HST,on the basis of the morphology revealed by the images. Thepresent paper is a pilot study based on multiwavelength ob-servations, aimed to study the gas kinematics and excitationof one OH Megamaser galaxy and that we hope to extendto the whole sample.IRASF23199+0123 has a redshift z = . (Darling & Giovanelli 2006), corresponding to a distance of558 Mpc for which 1 ′′ corresponds to 2.7 kpc at the galaxy,assuming a Hubble constant of H = km s − Mpc − .Its OH maser emission was first detected in the Arecibosurvey, that observed 52 objects with . < z < . (Darling & Giovanelli 2001). Darling & Giovanelli (2006)used spectroscopic data obtained with the Palomar 5 mtelescope with the Double Spectrograph in order to performan optical spectroscopic study of the properties of thesample of the Arecibo survey and identified the nuclearemission of IRASF23199+0123 as being due to a Seyfert 2nucleus, based on emission-line ratios.Our GMOS-IFU data comprise observations of the cen-tral region of the eastern galaxy of the IRASF23199+0123pair and our aim is to map the distribution and kinematicsof the optical line emitting gas and investigate the excita-tion mechanism of the nuclear emission. This is the firsttime that OH megamaser galaxies have been observed withan Integral Field Spectrograph, allowing a two-dimensionallook at the gas excitation and kinematics in detail. This paper is organized as follows. Section 2 describes the obser-vations and data reduction procedure and section 3 explainsthe emission-line fitting process and present maps for theemission-line flux distributions and kinematics, as well asthe HST and VLA radio continuum images, while in section4 the results are discussed. Finally, the conclusions of thiswork are presented in section 5. We observed IRASF23199+0123 with the Karl G. JanskyVery Large Array (VLA) on Apr 20, 2014. The observa-tions included X-band (8–10 GHz) continuum, L-band (1–2 GHz) continuum, and L-band spectral line observationsof the redshifted OH (1665/1667 MHz) maser lines. The L-band and X-band observations comprised respectively threeand one 10-minutes scans, alternating with 3-minute scansof the phase calibrator, J2320+0513. We observed the source3C48 in both X-band and L-band for flux and bandpass cal-ibration.The VLA pipeline in CASA (McMullin et al. 2007),was used for data reduction. This includes initial data flag-ging and phase, flux and bandpass calibrations. The contin-uum images were generated using multi-frequency synthesis(e.g. Conway et al. 1990; Rau & Cornwell 2011) with nat-ural weighting and deconvolved using the Cotton-Schwabvariant of the CLEAN algorithm (Schwab 1984). Imagingincluded simultaneous deconvolution of neighbouring ra-dio sources within the primary beam. We applied threerounds of phase-only self-calibration based on CLEAN mod-els for the radio continuum (self-calibration is reviewed byPearson & Readhead (1984)). For the L-band continuumimage, the restoring beam is 1 . ′′ × . ′′
28, PA 14 ◦ , and thebackground rms is 0.024 mJy beam − . The restoring beamof the X-band continuum image is 0 . ′′ × . ′′ ◦ ,and the background rms is 0.0093 mJy beam − . The radiocontinuum images are presented in Figure 2.The spectral line visibilities were continuum subtractedin two steps. First, we produced CLEAN continuum mod-els based on line-free channels, and the CLEAN mod-els were subtracted from the observed visibilities. Second,we removed any residual continuum using the CASA task uvcontsub ; the continuum was determined by averagingvisibility spectra over line-free channels. The continuum-subtracted, spectral line data cube was produced us-ing standard Fourier inversion and CLEAN deconvolu-tion. The expected line width of the 1667 MHz featureis 0.68 MHz (Darling & Giovanelli 2001); therefore, to im-prove the signal-to-noise, we binned the spectral line data to0.23 MHz channels (roughly / line width). The restoringbeam of the OH spectral line cube is 1 . ′′ × . ′′
36, PA 14 ◦ ,and the typical background rms for a single 0.23 MHz chan-nel is 0.44 mJy beam − . HST images of IRASF23199+0123 were acquired using theAdvanced Camera for Surveys (ACS) with the broad-band
MNRAS , 1–12 (2017) utflows from the nucleus of IRASF23199+0123 filter F814W, the narrow-band filter FR656N and medium-band FR914M filter as part of a snapshot survey programto observe a sample of OHMGs (Program id 11604; PI: D.J.Axon). The total integration time was 600 s in the broadband (I) F814W filter, 200 s in the medium-band filter and600 s in the narrow band H α FR656N filter. The band-pass of the FR656N narrow band filter includes both H α and [N ii ] λ , emission lines. We have processed the fi-nal images in order to remove cosmic rays using IRAF task lacos im (van Dokkum 2001). The continuum free H α + [N ii ]image of IRASF23199+0123 was constructed according tothe following the steps: (i) the count rates of a few starswere obtained in both the medium (FR914M) and narrowband (FR656N) ramp filter images; (ii) from the count rateratios the mean scaling factor was computed and applied tothe medium band FR914M image; (iii) the scaled FR914Mimage was then subtracted from the narrow-band FR656Nimage. We next visually inspected our continuum subtractedH α + [N ii ] image to assure that the residual fluxes of fore-ground stars were negligible within the uncertainties. Thisprocedure results in typical flux uncertainties of 5-10% (seeHoopes et al. 1999; Rossa & Dettmar 2000, 2003).The HST images (see Fig. 4) show for the first time thatIRASF23199+0123 is indeed an interacting pair and we haveobtained IFS of the eastern member of the par (hereafterIRAS23199E). IRAS23199E was observed using the GMOS (Hook et al.2004) IFU (Allington-Smith et al. 2002) as part of the pro-gram GS-2013B-Q-86 (PI: D. Sales). Only the eastern nu-cleus was observed, as it presents a steep continuum flux dis-tribution and with bright guide stars available in the GMOSpatrol field. The major axis of the IFU was oriented alongposition angle PA = ◦ , approximately along the major axisof the galaxy. The total on source exposure time was 4800s divided into 4 individual exposures of 1200 s each. Theobservations were performed on August 29, 2013 using theB600 grating with the IFU in the one slit mode, in com-bination with the GG455 filter. This setup resulted in anangular coverage of 5 . ′′ × . ′′
5, covering the spectral regionfrom 450 nm to 750 nm at a spectral resolution of 1.7 ˚A,as obtained from the measurement of the full width at halfmaximum (FWHM) of typical emission lines of the Ar lampspectrum used for the wavelength calibration.The data reduction process was performed using rou-tines of the GEMINI package in the Image Reduction andAnalysis Facility (IRAF, Tody 1986, 1993) software and fol-lowed the standard procedure of spectroscopic data reduc-tion (Lena 2014). First, we subtracted the bias level fromeach image, performed flat-fielding and trimming. Then, weapplied wavelength calibration to the data using the spectraof arc lamps as references and subtracted the sky emission.Finally, we performed flux calibration using a sensitivityfunction obtained from the spectrum of the H600 photo-metric standard star observed during the same night of theobject observations.Finally, datacubes for each individual exposures werecreated at an angular sampling of 0 . ′′ × . ′′
1, which were thenmedian combined using the IRAF gemcombine task to ob-tain the final datacube for the object. The peak of the con- tinuum emission was used as a reference during the mosaick-ing of the individual datacubes and we used the avsigclip algorithm for bad pixel/cosmic ray removal. We estimatedthe angular resolution as 0 . ′′
85 from the measurement of theFWHM of the flux distribution of field stars present in theacquisition image of the galaxy. This angular resolution cor-responds to ∼ d = Mpc).In order to remove noise from the final datacube, weperformed a spatial filtering using a Butterworth bandpassfilter (Gonzalez & Woods 2002; Menezes et al. 2014, 2015)via the IDL routine band pass − f ilter . pro , which allows thechoice of the cut-off frequency ( ν ) and the order of the fil-ter n . A low value of n (e.g. 1) is close to a Gaussian filter,while a high value (e.g. 10) corresponds to an ideal filter.We used n = and ν = . Ny, chosen by comparing thefiltered cube with the original one. For lower values of ν ,besides the removal of spatial noise, the filter excludes alsoemission from the galaxy, while for larger values of ν thefiltering procedure is not efficient. The filtering process doesnot change the angular resolution of the data and all mea-surements presented in the forthcoming sections were doneusing the filtered cube. α +[N ii ] GMOS-IFU spectra In order to map line fluxes, line-of-sight velocity ( V LOS )and velocity dispersion ( σ ) of the emitting gas, we fittedthe emission-line profiles of H α and [N ii ] λλ profit ) rou-tine (Riffel et al. 2010), which provides as outputs theemission-line flux, the centroid velocity, the velocity dis-persion and their corresponding uncertainties. Besides the[N ii ]+H α emission lines, the nuclear spectrum includes alsothe [O i ] λ ii ] and H α narrow components, being wellreproduced by a single Gaussian curve.The fitting process for the H α and [N ii ] λλ . ′′ . ′′ α +[N ii ]complex using one Gaussian per line, adopting the follow-ing constraints: (i) we kept tied the kinematics of the [N ii ]lines, such that the two lines have the same velocity andvelocity dispersion, and (ii) fixed the [N ii ] λ /[N ii ] λ intensity ratio to its theoretical value (3). The underlyingcontinuum was fitted by a linear function.Within 0 . ′′ α +[N ii ] complex. In the first, we fitted theline profiles using four Gaussian curves, in order to include a The routine is available at htt ps : // . harrisgeospatial . com / docs / bandpass − f ilter . html MNRAS , 1–12 (2017)
C. Hekatelyne et al.
Figure 1.
Fit of the nuclear spectrum (within 0 . ′′ α +[N ii ] complex. The observed profile is shown in black, while the blue dottedlines represent the broad and narrow components. The red line isthe result of the fit and the green dotted line shows the residualof the fit plus an arbitrary constant. broad component to represent H α . The resulting fit does notreproduce the observed profiles adequately. In the second fit,the [N ii ] lines were fitted by a single Gaussian component,while the H α profile was fitted by the same narrow compo-nent plus two broad components. This procedure resulted ina better fit to the profiles. As the presence of two broad com-ponents is restricted to the nucleus and the correspondingemission is not resolved, we propose that they actually rep-resent a single double-peaked line originating in the BroadLine Region (BLR). Such double-peaked components are notuncommon in AGN (e.g. Storchi-Bergmann et al. 2017).In Figure 1 we show the resulting fit of the nuclear spec-trum, where the observed profiles are shown in black, thebest fit model in red and the individual components as dot-ted blue lines. As the emission of the BLR is not resolved,the width and central wavelength of each broad componentas well as their relative fluxes were kept fixed for all spax-els, while their amplitudes were allowed to vary to enablefor smearing by the seeing. The values of the centroid ve-locities and velocity dispersions were obtained by fitting anintegrated spectrum of the inner 0 . ′′
8. The centroid veloc-ities relative to the systemic velocity are −
416 km s − and668 km s − for the blueshifted and redshifted components,respectively. The corresponding velocity dispersions are σ = km s − for the blue component and σ = km s − for the red component. In the hypothesis described abovethat these two components represent a single double-peakedline, this line profile has a full-width at half maximum of2 170 km s − . The systemic velocity adopted in this paperis v s =37 947 ± − (corrected for the heliocentric restframe), as derived by the modelling of the gas velocity fieldas discussed in Sec. 4.3. Figure 3 shows the VLA OH spectra extracted at the posi-tions of the eastern (IRAS23199E) and western nuclei. OHmasers are detected only towards the eastern nucleus. Two
Figure 2.
Top panel: VLA L-Band (1.6 GHz) continuum imageof IRASF23199+0123, shown as filled contours. The contours are(black) 0.071 ( σ ), 0.15, (white) 0.32, 0.69, & 1.5 mJy beam − .The OH1 and OH2 labels identify the locations where the OHmaser sources were detected (see Fig. 3). Bottom panel: VLAX-Band (8 GHz) continuum image of IRASF23199+0123, shownas filled contours. The contours are (black) 0.0278 ( σ ), 0.0647,(white) 0.150, & 0.349 mJy beam − . line features appear in the spectrum, which we label ‘OH1’,detected at . σ significance, and ‘OH2’, a . σ detection.To within the measurement uncertainties, OH1 matches thecentral frequency and peak flux density of the maser featuredetected by Darling & Giovanelli (2001). To our knowledge,OH2 has never been detected before; OH2 falls outside thebandpass of the Darling & Giovanelli (2001) Arecibo obser-vations. The position of OH1 and OH2 defections are iden-tified in the top panel of Fig. 2.Fig. 3 shows that OH1 appears to identify with the1665 MHz line at the redshift of the eastern nucleus. As-suming OH1 is indeed the 1665 MHz line, its centroid ve-locity would be v = + ± km s − relative to systemicin the rest frame of the eastern nucleus. OH2 is howeversignificantly offset from the expected heliocentric frequen-cies of possible emission lines. It may identify either with1665 MHz line at v = − ± km s − or the 1667 MHzline at v = − ± km s − relative to systemic. Figure 4 presents the HST broad-band continuum F814Wimage (top panel) and the narrow-band H α +[NII] im-age (bottom panel) of the inner 20 ×
20 arcsec ofIRASF23199+0123, which reveal that this system is an in-teracting pair. The images in the left panels show both galax-ies of the pair, while the right panels show a zoomed in viewof IRAS23188E within the same field of view of the GMOS MNRAS , 1–12 (2017) utflows from the nucleus of IRASF23199+0123 Figure 3.
VLA OH maser spectra of the eastern (top) and west-ern (bottom, offset by -4 mJy) nuclei. The vertical, dashed greylines mark the expected, redshifted frequencies for the 1612 MHz,1665 MHz, 1667 MHz, and 1712 MHz maser features. Two spec-tral features are detected at the position of the eastern nu-cleus, marked OH1 and OH2. OH1 was originally detected byDarling & Giovanelli (2001); OH2 is a new detection.
IFU observations. The green box in the left panels corre-sponds to the field of view of our GMOS data. The HSTimages were rotated to the same orientation of the GMOSdata.The F814W continuum image of IRAS23199E presentsthe highest intensity levels in an elongated structure at PA ≈ ◦ suggesting that the galaxy is highly inclined. Thewestern galaxy of the pair shows a less elongated flux dis-tribution, suggesting a more face-on orientation, althoughthe flux distribution towards the center is not uniform, butpresents a complex structure. The F814W image also revealsa structure that resembles a spiral arm to the north thatseems to connect the two galaxies, that may be a tidal tailconnecting the two galaxies. The linear distance between thegalaxy nuclei, projected in the plane of the sky is 24 kpc,assuming a distance to the galaxy of 558 Mpc. This is alower limit as we do not know the orientation of the planecontaining both galaxies or their nuclei relative to the planeof the sky.The bottom left panel of Fig. 4 shows the continuum-free HST H α +[N ii ] narrow-band image of the two galaxies.The bottom right panel shows a zoom of the region observedwith GMOS-IFU, covering the central part of IRAS23199E.The H α +[N ii ] flux distribution is similar to that of the con-tinuum, and suggests the presence of two tidal tails, one tothe northeast and another to the southwest of the nucleus. The top panels of Figure 5 present the flux distributions inthe narrow components of H α (left) and [N ii ] λ − cm and the grey regions represent maskedlocations where the uncertainty in the flux is larger than30%. The central cross marks the location of the nucleus,defined as the position of the peak of the flux distribution ofthe broad H α component and is labelled with the letter Nin Fig. 5. The two lines show similar flux distributions, withthe emission within the inner ∼ ′′ being elongated in thenortheast–southwest direction. At least three extranuclearknots of emission, labelled A, B and C in Fig. 5, are observedin the H α flux map, at (x,y) angular distances relative to the nucleus of ( − . ′′ − . ′′ − . ′′ . ′′
8) and (1 . ′′ − . ′′ ii ]+H α image. The central panels of Fig. 5 present the line-of-sight veloc-ity ( V LOS ) fields for the H α (left) and [N ii ] λ V s =37 947 km s − was subtracted fromthe observed velocities, as derived from the modelling of theH α velocity field with a rotating disk model (see Sec. 4.3).The velocity fields derived from H α and [N ii ] emis-sion lines are similar, presenting redshifts to the west andblueshifts to the east of the nucleus with velocities reach-ing up to 200 km s − . The zero velocity line presents an S shape and values close to the systemic velocity are observedat ∼ ′′ to the north of the nucleus.The bottom panels of Fig. 5 show the velocity disper-sion ( σ ) maps for H α (left) and [N ii ] λ ∼ to 150 km s − , with the highest valuesobserved co-spatially with the S shaped zero velocity curveof the V LOS maps (central panels). In addition, some high σ values are seen for H α at 1 . ′′ σ are co-spatial with the extranuclear knots ofemission observed in the emission line flux distributions inthe top panels of Fig. 5. In this work, we present for the first time high quality HSTimages of IRASF23199+0123. These images allowed us toidentify that this system is indeed composed by two mem-bers and discern tidal structures. Thus, this is the first timethat this system has been established to be an interactinggalaxy pair.
The [N ii ] λ α flux ratio can be used to map thegas excitation (e.g. Baldwin, Phillips & Terlevich 1981;Cid Fernandes et al. 2010), with values [N ii ] λ α ≤ ii ] λ α ratio map forIRAS23199E. Grey regions correspond to locations maskeddue to poor fits. The map shows values ranging from ∼ σ values (bottom panels ofFig. 5), while the lowest values are observed in the knots ofenhanced H α emission (top-left panel of Fig. 5). We interpretthese latter locations as complexes of star forming regions.The fact that the locations with the highest[N ii ] λ α ratios are co-spatial with the highest σ val-ues suggests that shocks contribute to the gas excitation. MNRAS , 1–12 (2017)
C. Hekatelyne et al.
Figure 4.
Top panels: Left – Large-scale image (ACS/HST F814W - i band). Right – Zoom of the region observed with GMOS. Bottompanels: Left – HST large-scale continuum free H α +[N ii ] image. Right – Zoom at the region observed with GMOS IFU. The green boxesshow the GMOS IFU field of view (3 . ′′ × . ′′
5) and the color bars show the fluxes in arbitrary units.
We have extracted a spectrum within a circular aperture of0 . ′′ α from this spectrum, to plot this region in the diagnos-tic diagram WHAN (Cid Fernandes et al. 2010). This dia-gram has been proposed as an alternative to the BPT dia-grams of Baldwin, Phillips & Terlevich (1981), and is a plotof the the H α equivalent width against the [N ii ] λ α flux ratio. While BPT diagrams need four emission linesto separate the regions ionized by AGN or starburst, theWHAN diagram requires just H α and [N ii ]. The WHANdiagram enables a separation between Starbusts, Seyfertgalaxies (sAGN) and low-luminosity AGNs (wAGN). It isalso possible to separate wAGN population, where the emit-ting gas is excited by a central AGN from Retired Galax- ies (RG), where the gas emission may be due to excitationby hot, evolved (post-asymptotic giant branch - post-AGB)stars, in which case the H α equivalent width is smaller than3˚A(e.g. Belfiore et al. 2016; Brum et al. 2017).The WHAN diagram for IRAS23199E is shown in theleft panel of Figure 7, while the right panel of this figureshows the corresponding excitation map, with the distinctexcitation classes identified. The red triangle represents thenucleus, as obtained from the fitting of the line profiles froman integrated nuclear spectrum within 0 . ′′ α broad component is seen to up to 0 . ′′ . ′′ . ′′ MNRAS , 1–12 (2017) utflows from the nucleus of IRASF23199+0123 Figure 5.
Top panels: flux maps in the H α (left) and [N ii ] λ − cm − . Central panels: line-of-sight velocityfields for the H α (left) and [N ii ] (right) emitting gas. The colorbars show the velocities in units of km s − , after the subtractionof the systemic velocity of the galaxy. Bottom panels: velocitydispersion maps for the H α (left) and [N ii ] (right) emission lines,corrected for the instrumental broadening. The color bars showthe σ values in units of km s − . The central cross in all panelsmarks the position of the nucleus and grey regions in the flux and σ maps and white regions in the V LOS maps represent masked lo-cations, where the signal-to-noise was not high enough to obtainreliable fits to the emission line profiles or locations with no linedetection. The green contours in the velocity dispersion map ofH α line are from the 3 cm radio image. The grey asterisks labelledas OH1 and OH2 mark the locations where the maser emissionhas been detected. triangles correspond to the extra-nuclear locations identifiedin Fig. 5. The green triangle corresponds to the spectrumfrom region E described above where there is an enhance-ment of both the line ratio and the velocity dispersion σ . Allpoints are located within the region expected for emissionof gas excited by an AGN or Starburst. In particular, thepoint corresponding to the nucleus is clearly located withinthe sAGN region (strong AGN), indicating that the nuclearemission originates in a Seyfert type AGN, in agreementwith the presence of the broad H α component. In additionto the nucleus, region E can also be classified as sAGN. Atthis location, the [N ii ] λ α ratio is even larger thanfor the nucleus, possibly that shocks contribute to the gasexcitation.. The blue triangles corresponding to the extra-nuclear regions identified in Fig. 5 are located very close tothe line that separates starburst from sAGN excitation, in-dicating that both the central AGN and star forming regionscontribute to the gas excitation at these locations.We note that there is a general trend for higher[N ii ] λ α to be associated with larger σ values. Someof these regions surround the contours of the 3 cm radio emis-sion, as shown in Fig. 6. We thus attribute the gas emissionat high- σ and [N ii ]/H α ratio as at least being partly pro-duced by excitation of the gas by shocks associated withproduced by the radio-emitting plasma. There are also someregions to the south of the nucleus where there is no radioemission but there is still enhancement of the σ and lineratios, which we attribute to the presence of additional per-turbations possibly due to the interaction between the twogalaxies in IRASF23199+0123. As discussed in previous sections, we identify several knotsof enhanced H α emission, associated to star forming regions.These knots are labelled A, B and C in Fig. 5. In order tocharacterize the star formation at these locations, we usethe integrated fluxes within the circular apertures shownin Fig. 5. These fluxes have been used to estimate physicalproperties of the star forming regions that are listed in Ta-ble 1. In order to estimate the mass of ionized gas we used(Peterson 1997): M M ⊙ ≈ . × L ( H α ) n , (1)where L (H α ) is the H α luminosity in units of 10 erg s − and n is the electron density ( N e ) in units of 10 cm − . Wehave assumed N e = cm − , which is the mean value ofelectron density of circumnuclear star formation regions de-rived from the [S ii ] λ λ × M ⊙ .We estimated the rate of ionizing photons Q [H + ] andstar formation rate (SFR) under the assumption of a con-tinuous star formation regime. The rate of ionizing photonsfor each star formation region was derived using Osterbrock(1989): Q [ H + ] = α B L H α α EFFH α h ν H α , (2) MNRAS , 1–12 (2017)
C. Hekatelyne et al.
Figure 6. [N ii ]/H α flux ratio map for IRAS23199E. Grey regionscorrespond to masked locations where no reliable measurementsare available. The green contours in the map are from the 3 cmradio image. The black asterisks labelled as OH1 and OH2 markthe position of the maser emission. where α B is the hydrogen recombination coefficient to allenergy levels above the ground level, α EFFH α is the effectiverecombination coefficient for H α , h is the Planck’s constantand ν H α is the frequency of the H α line. Using α B =2.59 × cm s − and α EFFH α =1.17 × − cm s − (Osterbrock1989) we obtain: (cid:18) Q [ H + ] s − (cid:19) = . × (cid:18) L H α s − (cid:19) . (3)The star formation rate (SFR) was computed using thefollowing relation (Kennicutt 1998): SFR M ⊙ yr − = . × − L H α ergs − (4)SFRs derived for the star formation regions ofIRAS23199E are in the range 0.05 – 0.12 M ⊙ yr − , con-sistent with a moderate star-forming regime. These SFRsfall within the range observed for circumnuclear star form-ing regions in nearby Seyfert galaxies, derived using optical(Dors et al. 2008) and near-infrared (Falc´on-Barroso et al.2014; Riffel et al. 2016; Hennig et al. 2017) emission-linesand are consistent with the average value of SFR = . M ⊙ yr − for a sample of 385 galaxies.The values of the ionizing photons rate are in therange log Q[H + ] = (51.87 - 52.23) s − and are in agreementwith previous reported values for circumnuclear star form- ing regions in nearby galaxies (e.g. Wold & Galliano 2006;Galliano & Alloin 2008; Riffel et al. 2009, 2016).The masses of ionized gas derived for the star-formingcomplexes in IRASF23199E are in the range of (1.78 -4.08) × M ⊙ , and agree with those previously obtainedfor star forming regions surrounding Seyfert nuclei (e.g.Riffel et al. 2016).We can also use the Far-Infrared luminosity of galaxyto calculate the SFR (Kennicutt 1998): SFR ( M ⊙ yr − ) = . × − L FIR ( ergs − ) (5)where L FIR is the IR luminosity integrated over the mid-and far-IR spectrum (8 - 1000 µ m). Using L FIR = . × erg s − (Darling & Giovanelli 2006), we obtain SFR ≈ M ⊙ yr − . This value is much higher than those obtained foreach star-forming region (shown in Tab. 1) and suggests thatmost of the FIR luminosity may not be due to star forma-tion, but to the AGN, or most of the star formation is em-bedded in dust. The velocity fields shown in Fig. 5 are complex, but they sug-gest the presence of a rotation pattern with the line of nodesoriented approximately along the east-west direction. In or-der to describe analytically this behaviour, we used a simplerotation model (van der Kruit & Allen 1978; Bertola et al.1991),which assumes that the gas moves in circular orbitsin the plane of the galaxy, within a central gravitational po-tential. In this model, the rotation velocity field is given by: V mod ( R , ψ ) = v s + AR cos ( ψ − ψ ) sin ( i ) cos p ( i ) { R [ sin ( ψ − ψ ) + cos ( i ) cos ( ψ − ψ )] + c cos ( i ) } p (6)where R and ψ are the coordinates of each pixel in the planeof the sky, v s is the systemic velocity of the galaxy, A is thevelocity amplitude, ψ is the major axis position angle, i isthe disc inclination relative to the plane of the sky ( i = for face-on disc), p is a model fitting parameter (for p = 1the rotation curve at large radii is asymptotically flat whilefor p = 3/2 the system has a finite mass) and c is a concen-tration parameter, defined as the radius where the rotationcurve reaches 70 % of its velocity amplitude.The [N ii ] and H α emission-lines present similar velocityfields (Fig. 5), so we have chosen the H α velocity field toperform the fit, as this line is stronger than [N ii ] λ atmost locations.The observed velocities were fitted with the equa-tion above using the MPFITFUN routine (Markwardt et al.2009) in IDL , that performs a non-linear least-squares fit,after initial guesses for the parameters. During the fit, theposition of the kinematical center was kept fixed to the po-sition of the peak of the flux distribution of the broad H α component, adopted as the location of the nucleus of the htt p : // . harrisgeospatial . com / ProductsandSolutions / GeospatialProducts / IDL . aspx MNRAS , 1–12 (2017) utflows from the nucleus of IRASF23199+0123 Figure 7.
Left: the WHAN diagram for IRAS23199E showing the different excitation regions (Cid Fernandes et al. 2010). Each crosscorresponds to an individual spaxel of the IFU datacube; the red triangle represents the nucleus; blue triangles are for regions A, B, Cand D identified in the H α flux map of Fig. 5 and the green triangle corresponds to the region E identified in the [N ii ] σ map of Fig. 5.Right: Excitation map identifying the regions within the FoV corresponding to different excitation mechanisms: strong AGNS (sAGN),weak AGN (wAGN), Starforming (SF) and Retired Galaxy (RG). Table 1.
Physical properties of the star formation regions in IRAS23199E.Region L H α (10 erg s − ) EqW[H α ] M (10 M ⊙ ) log Q[H + ] (s − ) SFR (M ⊙ yr − )A 0.12 44.41 3.06 52.10 0.09B 0.07 28.41 1.78 51.87 0.05C 0.16 45.84 4.08 52.23 0.12 Figure 8.
Observed H α velocity field (left), rotating disk model (center) and residual between the two (right). The central cross marksthe position of the nucleus, the white regions are masked locations where we were not able to fit the emission line profiles and the dottedlines represent the orientation of the line of nodes. The black contours in the residual map are from the 3 cm radio image with the sameflux levels as shown in Fig. 2 and the white contours show radio contours at the 1.5 σ level. The black asterisks labelled OH1 and OH2mark the position of the maser sources.MNRAS , 1–12 (2017) C. Hekatelyne et al. galaxy and the parameter p was kept fixed at p = . , asdone in previous works (e.g. Brum et al. 2017).The resulting best fit model is shown in the centralpanel of Figure 8 and its parameters are A =349 ±
26 kms − , v s =37 947 ± − (corrected for the heliocentric restframe), ψ =95 ± ◦ , c =1 . ′′ ± . ′′
1, i=41 ◦ ± ◦ . The systemicvelocity of the galaxy can be compared with previous mea-surements. Lawrence et al. (1999) described the construc-tion of the QDOT survey, which consists of infrared proper-ties and redshifts of an all-sky sample of 2387 IRAS galaxies.They obtained a systemic velocity for IRASF23199+0123of 40 981 km s − . On the other hand, Darling & Giovanelli(2006) performed an optical spectrophotometric study of re-solved spectra of multiple nuclei merging systems that hostsOHM sources and obtained v s =40 679 ± − . We notethat the systemic velocity derived here is smaller than thoseobtained in previous studies. We speculate that this differ-ence may be due to the fact that previous works possiblehave observed the western nucleus, which is brighter, sinceat that time it was not known that IRASF23199+0123 iscomposed by two members.Besides the disk rotation model, Fig. 8 shows also theobserved H α velocity field in the left panel and the residualmap between the observed velocities and the model in theright panel. The residual velocity map shows values muchsmaller than the observed velocity amplitude, but residualsof up to 100 km s − are present at some locations. Blueshiftsare seen to the south and north of the nucleus, while redshiftsare observed at the nucleus and its surroundings.In order to investigate the origin of the velocity residu-als, we have overlaid the contours from the 3 cm radio imageon the residual map shown in the right panel of Fig. 8. Inthis figure, we have plotted also the radio contours at the1.5 σ level (in white), as these levels show that faint radioemission is elongated towards the region where blueshifts areobserved to the north of the nucleus. At the 3 σ level thereis just a hint of this elongation. This apparent association ofthe extended radio emission (although at faint levels) withthe blueshifts in the map of velocity residuals suggests thatthe radio-emitting plasma may play a role in the gas kine-matics and a possible interpretation is that the blueshiftedresiduals are due gas pushed away from the nucleus by aradio jet. Other indications of the interaction between theradio plasma and the emitting gas are the higher velocitydispersion and [N ii ]/H α values surrounding the radio struc-tures, as seen in Figs. 5 and 6. Similar results have beenfound for other galaxies and interpreted as originating fromthe interaction of the radio jet with the ambient gas (e.g.Riffel et al. 2006, 2015).We speculate that the observed blueshifts originate inoutflows along a bi-cone oriented in the north-south direc-tion with its axis approximately in the plane of the sky.The blueshifts would come from the front walls (nearside)of the cones to both sides of the nucleus, while the redshiftfrom the back (farside) walls are not observed probably dueto obscuration. The redshifts that are observed surround-ing the nucleus could be due to inflows towards the nucleus,probably due to gas motions associated with the interactionbetween the two galaxies of the pair.Finally, it is interesting to note that the OH masersare observed in the vicinity of the active nucleus ofIRASF23199E, close to regions of enhanced velocity disper- sion and [N ii ]/H α ratio in the emitting gas. This suggeststhat the maser sources are associated with the AGN, perhapsproduced in gas compressed in an interaction with expand-ing radio plasma. We also notice that the redshifted masersource (OH1), is located in a region with redshifted residu-als, while the blueshifted maser source (OH2), is located ina region of blueshifted residuals, suggesting that the OH2source is participating in the outflow. Darling & Giovanelli (2006) presented a study of the opticalproperties of the Arecibo Observatory OH megamaser sur-vey sample, with the aim of investigating the types of nuclearnuclear environments that produce OH megamasers. Theydetermined that IRASF23199+0123 harbours a Seyfert 2nucleus, based on an optical spectrum but the spectra usedfor the classification includes both nuclei, thus it was notpossible to reveal the nature of each nucleus separately.With our data we have discovered that IRAS23199Eharbours a Seyfert 1 nucleus, as we clearly detected a broadunresolved H α component at the nucleus and the WHANdiagram is consistent with Seyfert-like gas excitation there.The double-peaked nature of this broad component sug-gests it is due to unresolved disk rotation in the BLR, asmany recent studies have supported a flattened geometry forthis region (Storchi-Bergmann et al. 2017). In the NLR, jet-cloud interaction may give rise to double-peaked lines (e.g.Capetti et al. 1999; O’Dea et al. 2002; Kharb et al. 2017)and another possibility is that some interaction with theradio emission may be happening in the BLR is this partic-ular case. The discrepancy between Darling’s classificationand ours can be understood if Darling & Giovanelli (2006)observed the western nucleus, which appears brighter in ourHST image and in the Sloan Digital Sky Survey (SDSS) im-age (Bundy et al. 2015; Albareti et al. 2016).We estimate the mass of the central black hole usingthe empirical relation given by Greene & Ho (2005): M BH M ⊙ = ( . + . − . ) × (cid:18) L H α ergs − (cid:19) . ± . (cid:18) FW HM H α kms − (cid:19) . ± . (7)where M BH is the black hole mass, L H α is the luminosityand FW HM H α is the full width at half maximum of thebroad component. The luminosity was calculated as thesum of the luminosities of both components, resulting in L H α ≈ . × erg s − . We obtained the FW HM H α of 2170km s − directly from the observed profile. Using these valueswe have estimated a black hole mass of . + . − . × M ⊙ . We present a multi-wavelength study of the OH megamasergalaxy IRASF23199+0123 using the HST, VLA and GeminiNorth telescope. Our HST images show that this system isan interacting pair of galaxies and used integral field spectro-scopic data obtained with the Gemini GMOS-IFU to observe
MNRAS , 1–12 (2017) utflows from the nucleus of IRASF23199+0123 its eastern nucleus, which we call IRASF23199E. Our obser-vations cover the inner 9.5 ×
13 kpc of the galaxy at a spatialresolution of 2.3 kpc and velocity resolution of ∼ km s − .Our main conclusions are: • We show that IRASF23199+0123 is an interacting pairwith a tidal tail connecting the two galaxies and detect twoOH maser sources associated to the eastern member. • Both nuclei present extended radio emission at 3 and20 cm, with intensity peaks at the each nucleus. The 20 cmradio emission of the eastern nucleus is elongated in the di-rection of the most extended emission in the HST continuumimage (northeast-southwest), while in the western nucleusthe 20 cm radio emission is tilted by about 45 ◦ eastwardsrelative to the orientation of the most extended continuumemission. In the better spatially resolved 3 cm observations,some elongation is observed at low brightness level towardsthe north in the eastern nucleus. • One of the main results of this paper is the discoveryof a Seyfert 1 nucleus in IRASF23199E, via the detection ofan unresolved broad (FWHM ≈ − ) double peakedcomponent in the H α emission-line from the BLR. This isimportant in regard to the OH maser emission, because thetwo masing sources are detected in this galaxy which hostsan AGN, rather than the other member of the pair. In addi-tion, the masing sources are observed in the vicinity of en-hanced velocity dispersion and higher line ratios, suggestingthat they are associated with shocks driven by AGN out-flows. The blue and redshifted maser sources are associatedwith the blue and redshifted ionized gas velocity residuals.This combination of evidence from HST images, VLA linespectroscopy and IFU spectroscopy strongly indicates thatin this system, the OH megamaser sources are associatedwith the AGN, rather than star formation. • Using the width and luminosity of the broad H α profile,we estimate a mass of M BH = 3.8 + . − . × M ⊙ for the centralSMBH. • The comparison between the HST [N ii ]+H α im-age and GMOS-IFU emission-line flux distributions ofIRASF23199E shows that they are similar, being more elon-gated in the northeast-southwest direction, following thecontinuum emission. In addition, the GMOS H α flux mapreveals the presence of three extra-nuclear knots, attributedto star forming regions. • From the measurement of the H α fluxes from the starforming regions, we obtain: (1) star formation rates in therange (0.05 – 0.12) M ⊙ yr − ; (2) ionized gas content in therange (1.8 – 4.1) × M ⊙ ; and (3) ionized photons rate logQ[H + ]=51.9 to 52.2 s − . From the FIR luminosity we obtainobtain SFR ≈ M ⊙ yr − suggesting that most of the FIRluminosity may be due to the AGN or due to star formationembedded in dust. • The [N ii ] λ α flux ratio map of IRAS23199Epresents the highest values coincident with the regions withhighest σ values, which surround the radio source and sug-gests that the radio-emitting plasma interacts with the am-bient gas via shocks that seem to play a role in the gasexcitation. The lowest [N ii ] λ α values are co-spatialwith the star forming regions detected in the H α emission. • The velocity fields of IRASF23199E show a disturbedrotation pattern with the line of nodes oriented along PA = ◦ , as derived by the fit of the H α velocities with a ro- tation disk model. The residuals between the observed andmodelled velocity field combined with the velocity dispersionmaps suggest the presence of non circular motions, possiblydue to outflows from the nucleus along the directions northand south and inflows towards the nucleus in its vicinity. ACKNOWLEDGEMENTS
We thank an anonymous referee for useful suggestions whichhelped to improve the paper. This work is based on obser-vations obtained at the Gemini Observatory, which is oper-ated by the Association of Universities for Research in As-tronomy, Inc., under a cooperative agreement with the NSFon behalf of the Gemini partnership: the National ScienceFoundation (United States), the Science and Technology Fa-cilities Council (United Kingdom), the National ResearchCouncil (Canada), CONICYT (Chile), the Australian Re-search Council (Australia), Minist´erio da Ciˆencia e Tecnolo-gia (Brazil) and south-eastCYT (Argentina). This researchhas made use of the NASA/IPAC Extragalactic Database(NED) which is operated by the Jet Propulsion Labora-tory, California Institute of Technology, under contract withthe National Aeronautics and Space Administration. TheNational Radio Astronomy Observatory is a facility of theNational Science Foundation operated under cooperativeagreement by Associated Universities, Inc. We acknowledgethe usage of the HyperLeda database (http://leda.univ-lyon1.fr). C. H. thanks for CAPES financial support. R.A.R.acknowledges support from FAPERGS and CNPq.
REFERENCES
Albareti, F. D. et al., 2016, arXiv:1608.02013Allington-Smith, J.et al. 2002, PASP, 114, 892.Baldwin, J. A., Phillips, M. M., Terlevich, R., 1981, PASP, 93, 5.Belfiore, F., et al. 2016, MNRAS, 461, 3111.Bertola, F., Bettoni, D., Danziger, J., Sadler, E., Sparke, L., deZeeuw, T., 1991, ApJ, 373, 369Brum, C., Riffel, R. A., Storchi-Bergmann, T., Robinson, A.,Schnorr Muller, A., Lena, D., 207, MNRAS, 469, 3405.Bundy, K., et al., 2015, ApJ 798, id. 7Capetti, A., Axon, D. J., Macchetto, F. D., Marconi, A., Winge,C., 1999, ApJ, 518, 187.Cid Fernandes, R., Stasinska, G., Mateus, A., Vale Asari, N.,2011, MNRAS, 413, 1687.Cid Fernandes, R., Stasinska, G., Schlickmann, M. S, Mateus, A.,Vale Asari, N., Schoenell, W.; Sodre, L., 2010, MNRAS, 403,1036.Conway, J. E., Cornwell, T. J., & Wilkinson, P. N. 1990, MNRAS,246, 490Darling, J., Giovanelli, R. 2001, AJ, 121, 1278Darling J., Giovanelli, R., 2002, APJ, 572, 810Darling J., 2006, Giovanelli, R., APJ, 132,2596D´ıaz, ´A. I., Terlevich, E., Castellanos, M., & H¨agele, G. F. 2007,MNRAS, 382, 251.Dors O. L., Storchi-Bergmann T., Riffel R. A., Schmidt A. A.,2008, A&A, 482, 59Elitzur M., Shlosman I., 2006, ApJ, 648, L10.Elitzur, M. 2012, ApJL, 747, L33Falc´on-Barroso, J., Ramos Almeida, C., Boker, T., et al. 2014,MNRAS, 438, 329.Galliano, E., & Alloin, D. 2008, A&A, 487, 519MNRAS , 1–12 (2017) C. Hekatelyne et al.
Gonzalez R.C., Woods R.E., 2002, Digital Image Processing, 2ndedition, Prentice-Hall, Englewood Cliffs, NJ.Greene, J. E., & Ho, L. C., 2005, ApJ, 630, 122.M. G. Hennig, Riffel, R. A., Dors O. L., (in preparation)H¨onig, S. F., Kishimoto, M., Tristram, K. R. W., et al. 2013,MNRAS, 771, 87Hook, I., Jorgensen, I., Allington-Smith, J. R., Davies, R. L., Met-calfe, N., Murowinski, R. G., Crampton, D., 2004, PASP, 116,425Hoopes, C. G., Walterbos, R. A. M., & Rand, R. J. 1999, ApJ,522, 669Ivezi´c, Z., Elitzur, M. 2010, MNRAS, 404, 1415Kauffmann, G. et al. 2003a, MNRAS, 346, 1055Kennicutt, R. C. 1998, ARA&A, 36, 189Kewley, L. J., Dopita, M. A., Sutherland, R. S., Heisler, C. A.,Trevena, J. 2001a, ApJ, 556, 121Kewley, L. J, Groves, B, Kauffmann G., Heckman, T., 2016, MN-RAS, 372Lena, D., 2014, arXiv:1409.8264Lena, D., Robinson, A., Storchi-Bergmann, T., Schnorr M¨uller,A., Seelig, T., Riffel, R. A., Nagar, N. M., Couto, G.S., &Shadler, L., 2015, ApJ, 806, 84Li, J., Ostriker, J., Sunyaev, R., 2013, ApJ, 767, 105L.Lo, K. Y., ARA&A, 2005, 43, 625Lawrence, A., et al. 1999, MNRAS, 308, 897McMullin, J. P., Waters, B., Schiebel, D., Young, W., & Golap,K. 2007, Astronomical Data Analysis Software and SystemsXVI, ASP Conf. Ser. 376, ed. R. A. Shaw, F. Hill, & D. J.Bell, San Francisco, CA: ASP, 127Markwardt C. B., 2009, in Bohlender D. A., Durand D., DowlerP., eds, ASP Conf. Ser. Vol. 411, Astronomical Data AnalysisSoftware and Systems XVIII. Astron. Soc. Pac., San Fran-cisco, p. 251Menezes, R. B., Steiner, J. E., Ricci, T. V., 2014, MNRAS, 438,2597.Menezes, R. B., da Silva, P., Ricci, T. V., Steiner, J. E., May, D.,Borges, B. W., 2015, MNRAS, 450, 369.Mor, R., Netzer, H., & Elitzur, M. 2009, ApJ, 705, 298Nenkova, M., Sirocky, M. Ivezi´c, Z., Elitzur M., 2008, ApJ, 608,147.O’Dea, C. P. et al., 2002, AJ, 123, 2333.Osterbrock, D. E., 1989, Astrophysics of Gaseous Nebulae andActive Galactic Nuclei, University Science Books, Mill Valley,California.Pearson, T. J., & Readhead, A. C. S. 1984, ARAA, 22, 97Peterson B. M., 1997, An Introduction to Active Galactic Nuclei,Cambridge, New York Cambridge University PressKharb, P.; Subramanian, S.; Vaddi, S.;Das, M.; Paragi, Z., 2017, ApJ acceptedhttp//esoads.eso.org/abs/2017arXiv170800618KRau, U. & Cornwell, T.J. 2011, AA, 532, 71Riffel, R. A., Storchi-Bergmann, T., & Winge, C., Barbosa, F. K.B., 2006, MNRAS, 373, 2.Riffel, Rogemar A., Storchi-Bergmann, T., Dors, O. L., Winge,C., 2009, MNRAS, 393, 783Riffel R. A., 2010, ApSS, 327, 239.Riffel, R. A., Storchi-Bergmann, T., & Riffel, R. 2014, ApJL, 780,L24Riffel, R. A., Storchi-Bergmann, T., & Riffel, R. 2015, MNRAS,451, 3587Riffel, Rogemar A., Colina, L., Storchi-Bergmann, T., Piqueras,L´0pez J., Arribas, S., Riffel, R., Pastoriza, M., Sales, DinalvaA., Dametto, N. Z., Labiano, A. & Davies, R. I., 2016, MN-RAS, 461, 4192.Rossa, J., & Dettmar, R.-J. 2000, A&A, 359, 433Rossa, J., & Dettmar, R.-J. 2003, A&A, 406, 505Sales, D. A., Robinson, A., Axon, D. J., Gallimore, J. Kharb, P.,Curran, R. L., O’Dea, C., Baum, S., Elitzur, M.,Mittal, R. 2015, ApJ, 799Schwab, F. R. 1984, AJ, 89, 1076-1081Shi, L., Gu, Q. S., & Peng, Z. X. 2006, A&A, 450, 15.Storchi-Bergmann, T., Dors, O. L. Jr., Riffel, R. A., Fathi, K.,Axon, D. J., Robinson, A., Marconi, A., Ostlin, G., 2007, ApJ,670, 959.Storchi-Bergmann, T.; Schimoia, J. S.; Peterson, B. M.; Elvis, M.;Denney, K. D.; Eracleous, M.; Nemmen, R. S., 2017, ApJ, 835,236.Tody, D. 1986, The IRAF Data Reduction and Analysis Systemin Proc. SPIE Instrumentation in Astronomy VI, ed. D.L.Crawford, 627, 733Tody, D. 1993, IRAF in the Nineties” in Astronomical Data Anal-ysis Software and Systems II, A.S.P. Conference Ser., Vol 52,eds. R.J. Hanisch, R.J.V. Brissenden, J. Barnes, 173.van der Kruit, P.C., Allen, R.J., 1978, ARA&A, 16, 103P. G. 2001, PASP, 113, 1420Wold, M., Galliano, E., 2006, MNRAS, 369, 47.This paper has been typeset from a TEX/L A TEX file prepared bythe author. MNRAS000