Gemini NIFS survey of feeding and feedback in nearbyActive Galaxies - III. Ionized versus warm molecular gasmasses and distributions
Astor J. Schonell Jr., Thaisa Storchi-Bergmann, Rogemar A. Riffel, Rogério Riffel, Marina Bianchin, Luis G. Dahmer-Hahn, Marlon R. Diniz, Natacha Z. Dametto
aa r X i v : . [ a s t r o - ph . GA ] M a r MNRAS , 1–19 (2019) Preprint 20 March 2019 Compiled using MNRAS L A TEX style file v3.0
Gemini NIFS survey of feeding and feedback in nearbyActive Galaxies - III. Ionized versus warm molecular gasmasses and distributions
Astor J. Sch¨onell Jr. , ⋆ , Thaisa Storchi-Bergmann , Rogemar A. Riffel ,Rog´erio Riffel , Marina Bianchin , Luis G. Dahmer-Hahn , Marlon R. Diniz ,Natacha Z. Dametto Instituto de F´ısica, Universidade Federal do Rio Grande do Sul, Av. Bento Gon¸calves 9500, 91501-970Porto Alegre, RS, Brazil Universidade Federal de Santa Maria, Departamento de F´ısica, Centro de Ciˆencias Naturais e Exatas, 97105-900,Santa Maria, RS, Brazil Instituto Federal de Educa¸c˜ao, Ciˆencia e Tecnologia Farroupilha, BR287, km 360, Estrada do Chapad˜ao, 97760-000,Jaguari - RS, Brazil
Accepted XXX. Received YYY; in original form ZZZ
ABSTRACT
We have used the Gemini Near-Infrared Integral Field Spectrograph (NIFS) inthe J and K bands to map the distribution, excitation and kinematics of the ionizedH ii and warm molecular gas H , in the inner few 100 pc of 6 nearby active galaxies:NGC 788, Mrk 607, NGC 3227, NGC 3516, NGC 5506, NGC 5899. For most galaxies,this is the first time that such maps have been obtained. The ionized and H gas showdistinct kinematics: while the H gas is mostly rotating in the galaxy plane with lowvelocity dispersion ( σ ), the ionized gas usually shows signatures of outflows associatedwith higher σ values, most clearly seen in the [Fe ii ] emission line. These two gas speciesalso present distinct flux distributions: the H is more uniformly spread over the wholegalaxy plane, while the ionized gas is more concentrated around the nucleus and/orcollimated along the ionization axis of its Active Galactic Nucleus (AGN), presentinga steeper gradient in the average surface mass density profile than the H gas. Thetotal H ii masses cover the range × − × M ⊙ , with surface mass densitiesin the range 3–150 M ⊙ pc − , while for the warm H the values are 10 − times lower.We estimate that the available gas reservoir is at least ≈
100 times more massive thanneeded to power the AGN. If this gas form new stars the star-formation rates, obtainedfrom the Kennicutt-schmidt scalling relation, are in the range 1–260 × − M ⊙ yr − .But the gas will also – at least in part – be ejected in the form of the observed otflows. Key words:
Galaxies: active – Galaxies: Seyfert – Galaxies: nuclei – Galaxies: exci-tation
The growth of super-massive black holes (SMBH) and theirhost galaxies are connected by the AGN feeding and feed-back processes, that can presumably explain the correla-tion between the mass of the SMBH and the mass of thegalaxy bulge (Ferrarese & Ford 2005; Somerville et al. 2008;Kormendy & Ho 2013). The feeding via gas accretion is re-quired to trigger the nuclear activity, while the feedback ⋆ E-mail: juniorfi[email protected] provided by the AGN radiation and outflows is fundamen-tal to constrain galaxy evolution models, since without theAGN feedback the models predict that the most massivegalaxies form too many stars and grow more than observed(Springel et al. 2005; Fabian 2012; Terrazas et al. 2016)).Gas distribution, excitation and kinematics in the vicin-ity of Active Galactic Nuclei (AGN) ( ≈
100 pc scales) provideimportant constraints on the physics of the AGN feedingand feedback processes. The near-infrared (hereafter, near-IR) integral field spectroscopy (IFS) of nearby galaxies is aneffective method to quantify these processes. Eight to ten- © Astor J. Sch¨onell Jr. meter telescopes IFS with adaptive optics (AO) can providetwo-dimensional coverage with spatial resolution of a fewto tens of parsecs in nearby galaxies at spectral resolutionsthat allow to resolve gas inflows and outflows at such scales(M¨uller S´anchez et al. 2009; Davies et al. 2009, 2014). AOsystems are available mainly in the near-IR, a spectral re-gion that also has the advantage of being less affected bydust extinction (usually high in the central region of galax-ies) than optical observations. The use of AO assisted IFS ofnearby galaxies in the near-IR thus allows to resolve the cen-tral regions down to a few parsecs, and also simultaneouslymap two distinct gas phases: the ionized and molecular (H )gas emission. The latter is not available in the optical butcan be observed in the near-IR K band.The near-IR line-emission at ≈
100 pc scales in AGNhosts is originated by the heating and ionization of am-bient gas by the AGN radiation and by shocks producedby radio jets (Riffel et al. 2006, 2010a). Recent observationsby our group AGNIFS - AGN Integral Field Spectroscopy(e.g. Riffel et al. 2018) and others show that the molecularand ionized gas have distinct spacial distributions and kine-matics at these scales: the former is usually more restrictedto the plane of galaxies, with the kinematics being domi-nated by rotation in the disk in most cases, and presentingalso signatures of inflows in some cases; the latter traces amore disturbed medium, usually associated to outflows fromthe AGN, but frequently showing also a disk rotation com-ponent (e.g. Riffel et al. 2010a, 2013; Mazzalay et al. 2014;Barbosa et al. 2014; Diniz et al. 2015). Our previous studiesled to the conclusion that, while the ionized gas emission canbe considered a tracer of the AGN feedback, the moleculargas emission is usually a tracer of its feeding.In this paper we present maps of the ionized andmolecular gas distribution, excitation and kinematics of theinner 3 ′′ × ′′ of a sample of 6 nearby Seyfert galaxies. Thediscussion is restricted to the the gas mass distributionsand total ionized and molecular gas masses as well as tothe presentation of the global gas kinematics, pointingout signatures of rotation and outflows. The analysis anddiscussion of the gas excitation as well as the modellingof the gas kinematics and quantification of outflows willbe deferred to a forthcoming paper (hereafter identified asPaper B). This work is the third paper with the results ofa large Gemini proposal (P.I. Storchi-Bergmann) in whichour group AGNIFS aims to map and quantify the feedingand feedback processes of a sample of 29 nearby Seyfertgalaxies (Riffel et al. 2018), selected for their proximity andX-ray luminosity, as described in Sect. 2. Our ultimate goalis to investigate possible correlations between measuredproperties (as gas masses and densities, mass inflow andoutflow rates and kinetic power of the outflows) and theAGN luminosity. Results for individual galaxies of thesample have been already presented in previous papers byour group: NGC 4051 (Riffel et al. 2008a, 2017), NGC 4151(Storchi-Bergmann et al. 2009, 2010; Riffel et al. 2009a),Mrk 1066 (Riffel et al. 2010b; Riffel & Storchi-Bergmann2011; Ramos Almeida et al. 2009; Riffel et al. 2017),Mrk 1157 (Riffel & Storchi-Bergmann 2011; Riffel et al.2011, 2017), NGC 1068 (Storchi-Bergmann et al. 2012;Riffel et al. 2014a; Barbosa et al. 2014), Mrk 79 (Riffel et al.2013), Mrk 766 (Sch¨onell et al. 2014; Riffel et al. 2017),NGC5929 (Riffel et al. 2014b, 2015, 2017), NGC 2110 (Diniz et al. 2015), NGC 5548 (Sch¨onell et al. 2017;Riffel et al. 2017), NGC 788, NGC 3227, NGC 3516,NGC 4235, NGC 4388, NGC 5506, NGC 1052, NGC 5899and Mrk 607 (Riffel et al. 2017).This paper is organized as follows. In Section 2 wepresent the sample and in Section 3 the description of the ob-servations and data reduction procedures, while the fittingprocedure of the emission lines is discussed in Section 4. Theresults are shown in Section 5 , we discuss them in Section 6, and in Section 7 we present our conclusions. Our AGN sample was selected from the Swift-BAT 60-month catalogue adopting three criteria: (I) 14 −
195 keV lu-minosities L X ≥ . erg s − , (II) redshift z ≤ . , and(III) being accessible for NIFS ( − o < δ < o ). The selec-toin according to the hard (14 −
195 keV) band emission ofthe Swift-BAT survey is justified by the fact that it measuresdirect emission from the AGN rather than scattered or re-processed emission, and is much less sensitive to obscurationin the line-of-sight than soft X-rays or optical observations,allowing a selection based only on the AGN properties. Inorder to assure that we will be able to probe the feedingand feedback processes we further selected the galaxies forhaving previously observed extended [O iii ] λ We used J and K band data obtained with theGemini Near-Infrared Integral-Field Spectrograph (NIFS;McGregor et al. (2003)) with the AO system ALTAIR be-tween 2008 and 2016. NIFS has a square field of view of3 ′′ × ′′ , divided into 29 slitlets 0 . ′′
103 wide with a spatialsampling of 0 . ′′
042 along them. We used the standard Sky–Object–Object–Sky dither sequence in the observations,with off-source sky positions since all targets are extended.The individual exposure times varied according to the tar-get and are listed, together with further information on theobservations in Table 1. The filter ZJ G0601 was used withthe J-band observations, while the K-band observations wereperformed using the HK G0603 filter, as shown in Table 2together with further details of the instrument configuration.Data reduction followed standard procedures and wasaccomplished using tasks specifically developed for NIFSdata reduction, as part of gemini IRAF package, as well
MNRAS , 1–19 (2019) eeding and Feedback in AGN Table 1.
Log of the observations together with basic information on the sample galaxies. (1) Galaxy name; (2) Project ID; (3) J and(4) K-band exposure times (s); (5) Distance (Mpc); (6) Nuclear Activity; (7) Hubble type from NED; (8) Scale; (9) AGN bolometricluminosity. 1 2 3 4 5 6 7 8 9Galaxy Project ID J Exp-time K Exp-time D Act. Hub. type Scale logL
AGN (sec) (sec) (Mpc) (pc/arcsec) (erg/s)NGC788 GN-2015B-Q-29 ×
400 11 × ×
500 12 × ×
400 6 × ×
450 10 × (s) 184 44.2NGC5506 GN-2015A-Q-3 ×
400 10 × ×
460 10 × as generic IRAF tasks and IDL scripts. The procedures in-cluded trimming of the images, flat-fielding, sky subtrac-tion, wavelength and s-distortion calibrations. The telluricabsorptions have been removed using A-type standard starsobservations. These stars were also used to flux calibrate thespectra of the galaxies by interpolating a black body func-tion to the spectrum of each star in order to generate thesensitivity function. Finally, calibrated datacubes were cre-ated for each individual exposure at an angular sampling of0 . ′′ × . ′′
05 and combined in a final datacube for each galaxy.All datacubes cover the inner ≈ . ′′ × . ′′ − and the angular resolution rangesfrom 0 . ′′
12 to 0 . ′′
18, corresponding to a few tens of parsecsat the galaxies.
We have used the
PROFIT routine (Riffel 2010) to fitthe profiles of the following emission lines at each pixelover the whole field-of-view (FOV): [P ii ] λ µ m,[Fe ii ] λ µ m, Pa βλ µ m, H λ µ m andBr γλ µ m. This was done using Gauss-Hermite series,which were chosen to preserve most of the gas velocityinformation by fitting also the emission line wings via themoments h and h , besides returning the line-of-sightvelocity (V LOS ) and velocity dispersion ( σ ).The h Gauss-Hermite moment measures asymmetricdeviations from a Gaussian profile, such as blue (negativevalues) or red (positive values) wings, while the h momentquantifies the peakness of the profile, with positive valuesfor a more peaked profile and more extended wings thana Gaussian and negative values for a broader profile (moreflat-topped) and with less extended wings than that of aGaussian curve.The underlying continuum was fitted using a first degreefunction, since the spectral range used in the fit of each emis-sion line was small. The routine uses the mpfitfun routine(Markwardt 2009) to perform a non-linear χ minimization.The fit of the line plus continuum involves 7 free parameters– line amplitude, central wavelength, σ , h , h plus another2 from the first degree function used to fit the continuum.The only restrictions placed are that | h | and | h | are < profit rou-tine also outputs 1-sigma errors for each of the parameters,computed from the covariance matrix.In the case of NGC 3227, NGC 3516 and NGC 5506 wealso fitted a broad component to the Pa β and Br γ emis-sion line profiles in the central region, as these galaxies hosttype 1 AGN. This was done through a modification of the PROFIT routine to fit the broad component and subtractits contribution from the profiles in order to generate a datacube only with the narrow components. The steps for suchtask were: ( i ) fit of only one Gaussian to the broad com-ponent, by masking strong narrow emission lines; ( ii ) itssubtraction from the spectra in which it is present, and ( iii )fit of the narrow components. We achieved very satisfac-tory fits in all cases, with no constraints placed for the fit.The spatial region within which the fit of a broad compo-nent was necessary is shown by a cyan square in the thirdpanel of the first row in Figs. 4, 5 and 6, respectively forNGC 3227, NGC 3516 and NGC 5506. As the broad com-ponents are from the unresolved Broad Line Region (BLR),during the fit we kept fixed the central wavelength and widthof the broad components to the value measured from the in-tegrated spectrum within the square mentioned above andallowed only the variation of its amplitude. From the mea-surements presented in Figs. 4, 5 and 6, we conclude thatthe subtraction of the broad components was satisfactory,as we found no traces of them in the subtracted data cube. From the fits of the line profiles and resulting parameterswe have constructed maps of: flux distributions, V
LOS , ve-locity dispersions, h and h moments as well as of thereddening E ( B − V ) and emission line ratio maps, pre-sented in Fig. 2 for NGC 788, Fig. 3 for Mrk 607, Fig. 4 forNGC 3227, Fig. 5 for NGC 3516, Fig. 6 for NGC 5506 andFig. 7 for NGC 5899. We chose to show the results onlyfor [Fe ii ] λ µ m, Pa βλ µ m (or Br γλ µ m inone case) and H λ µ m because these are the strongestemission lines allowing the mapping of the ionized gasproperties via the [Fe ii ] and Pa β (or Br γ ) emission lines,and the warm molecular gas properties through the H emission line above. Although the nuclear spectra of some MNRAS000
LOS , ve-locity dispersions, h and h moments as well as of thereddening E ( B − V ) and emission line ratio maps, pre-sented in Fig. 2 for NGC 788, Fig. 3 for Mrk 607, Fig. 4 forNGC 3227, Fig. 5 for NGC 3516, Fig. 6 for NGC 5506 andFig. 7 for NGC 5899. We chose to show the results onlyfor [Fe ii ] λ µ m, Pa βλ µ m (or Br γλ µ m inone case) and H λ µ m because these are the strongestemission lines allowing the mapping of the ionized gasproperties via the [Fe ii ] and Pa β (or Br γ ) emission lines,and the warm molecular gas properties through the H emission line above. Although the nuclear spectra of some MNRAS000 , 1–19 (2019)
Astor J. Sch¨onell Jr.
Table 2.
Configuration of the observations: (1) spectral band, (2) grating, (3) filter, (4) filter central wavelength, (5) filter spectral range,(6) galaxies observed in each configuration.Band Grating Filter Central Wav. Spec. Range Galaxies observed( µ m) ( µ m)J J G5603 ZJ G0601 1.25 1.14 – 1.36 AllK K G5605 HK G0603 2.20 1.98 – 2.40 NGC 788, NGC 3516 and NGC5506K l Kl G5607 HK G0603 2.30 2.08 – 2.50 Mrk 607, NGC 3227 and NGC 5899
Figure 1.
J-band (left) and K-band (right) pectra obtained within a 0 . ′′ × . ′′
25 aperture centred at the nucleus of the galaxy indicatedin the top left corner of the left panels. Flux units are 10 − erg s − cm − ˚A − . galaxies show further coronal lines, as [S ix ] λ µ m and[Ca viii ] λ µ m, their flux distributions are unresolvedor barely resolved and are not shown here, leaving a de-tailed discussion about these lines together with that of thegas excitation to be presented in Paper B.In the next subsections we discuss separately the resultsfor each map, where we have masked out pixels with bad linefits, flagged according to the following criteria, applied to allmaps: (1) relative uncertainties in the line fluxes larger than30%; (2) uncertainties in velocity and velocity dispersionslarger than 50 km s − . Flux maps are shown in the first row of Figs. 2 – 7, wherewe have drawn the Position Angle (PA) of the majoraxis of the galaxy as given in the HYPERLEDA database(Prugniel et al. 2001) over the continuum images (leftmostpanel) and H flux and velocity maps. In all emission-linemaps, the peak flux is observed at the location of the con-tinuum peak, defined as the origin of the coordinates andidentified with the galaxy nucleus. The flux distributions forthe selected emission lines cover most of the FOV, corre- MNRAS , 1–19 (2019) eeding and Feedback in AGN NGC 788
Figure 2.
Maps of properties derived from the emission-line profiles of NGC 788 over the Gemini NIFS FOV (3 ′′ × ′′ ), shown as the smallcyan square over the continuum image, in the top-left corner. First column, from top to bottom: HST-WFPC2 F606W continuum image(Malkan et al. 1998) and in the insert at the top right corner the NIFS J-band continuum image; the grey line marks the photometricmajor axis; reddening E ( B − V ) map obtained from the Pa β /Br γ line ratio; and line ratio maps identified on the top of each panel. Secondcolumn, from top to bottom: Flux F , line-of-sight velocity V LOS , velocity dispersion σ , h Gauss-Hermite and h Gauss-Hermite momentsfor the [Fe ii ] λ . µ m emission line. Third and fourth columns: same as previous column for the Pa β and H λ . µ m emission lines,respectively. Fluxes are shown in logarithmic units of erg s − cm − , V LOS in km s − , relative to the systemic velocity of the galaxy and σ values are shown in km s − , after correction for the instrumental broadening. This galaxy is individually discussed in Sec. 6.1.MNRAS000
Maps of properties derived from the emission-line profiles of NGC 788 over the Gemini NIFS FOV (3 ′′ × ′′ ), shown as the smallcyan square over the continuum image, in the top-left corner. First column, from top to bottom: HST-WFPC2 F606W continuum image(Malkan et al. 1998) and in the insert at the top right corner the NIFS J-band continuum image; the grey line marks the photometricmajor axis; reddening E ( B − V ) map obtained from the Pa β /Br γ line ratio; and line ratio maps identified on the top of each panel. Secondcolumn, from top to bottom: Flux F , line-of-sight velocity V LOS , velocity dispersion σ , h Gauss-Hermite and h Gauss-Hermite momentsfor the [Fe ii ] λ . µ m emission line. Third and fourth columns: same as previous column for the Pa β and H λ . µ m emission lines,respectively. Fluxes are shown in logarithmic units of erg s − cm − , V LOS in km s − , relative to the systemic velocity of the galaxy and σ values are shown in km s − , after correction for the instrumental broadening. This galaxy is individually discussed in Sec. 6.1.MNRAS000 , 1–19 (2019) Astor J. Sch¨onell Jr.
Mrk 607
Figure 3.
As in Fig. 2 for Mrk 607. This galaxy is individually discussed in Sec. 6.2 sponding to maximum distances from the nucleus varyingfrom 100 pc for NGC 3227 to 410 pc for NGC 788.The direction of the largest extent of the [Fe ii ] λ . µ mand Pa β flux distributions follows the orientation of the ion-ization axis, and which in many cases approximately coin-cides with the orientation of the major axis of the galaxy inour sample. Exceptions are the cases of NGC 788 where theyare oriented at a small angle relative to the major axis, andNGC 5506, where the largest extent of the [Fe ii ] emission is perpendicular to the major axis. Although also showingthe highest emission levels following the major axis orienta-tion, the H λ . µ m flux maps usually show emission morespread over all directions, being less collimated than the ion-ized gas emission. The only exception is NGC 788, for whichthe strongest H emission is more extended perpendicularlyto the major axis of the galaxy. MNRAS , 1–19 (2019) eeding and Feedback in AGN NGC 3227
Figure 4.
As in Fig. 2 for NGC 3227. The cyan square in the Pa β flux map shows the region in which we had to subtract the contributionof a broad component. This galaxy is individually discussed in Sec. 6.3 Here we describe briefly the main features observed in theline-ratio maps; a more in-depth discussion of the gas exci-tation will be presented in Paper B.The reddening and line-ratio maps are shown in the firstcolumn of Figs. 2 – 7, below the galaxy continuum image. The E ( B − V ) map was obtained from: E ( B − V ) = .
74 log (cid:18) . F Pa β / F Br γ (cid:19) , (1)where F Pa β and F Br γ are the corresponding line fluxes andwe have adopted in the derivation of the expression abovethe Cardelli et al. (1989) extinction law and F Pa β / F Br γ the-oretical ratio of . (Osterbrock & Ferland 2006). MNRAS000
74 log (cid:18) . F Pa β / F Br γ (cid:19) , (1)where F Pa β and F Br γ are the corresponding line fluxes andwe have adopted in the derivation of the expression abovethe Cardelli et al. (1989) extinction law and F Pa β / F Br γ the-oretical ratio of . (Osterbrock & Ferland 2006). MNRAS000 , 1–19 (2019)
Astor J. Sch¨onell Jr.
NGC 3516
Figure 5.
As in Fig. 2 for NGC 3516. The cyan square in the Pa β flux map shows the region in which we had to subtract the contributionof a broad component. This galaxy is individually discussed in Sec. 6.4 The E ( B − V ) values are mostly in the range 1–3, reachingthe highest values at the nucleus of NGC 5506 ( E ( B − V ) ≈ )and NGC 5899 ( E ( B − V ) ≈ ) while for the other galaxiesthe highest values are observed outwards.The [Fe ii ] λ µ m/Pa β ratio maps can beused to investigate the excitation mechanism of [Fe ii ](Rodr´ıguez-Ardila et al. 2005; Riffel et al. 2008b, 2009b;Storchi-Bergmann et al. 2009, e.g.), with typical values for Seyfert galaxies ranging from 0.6 to 2. Most values areobserved within this range, usually increasing from thecentre outwards. In the case of NGC 3227 its value increasesto ≈ ii ] emission lines is observed. Highervalues than 2 are also observed in NGC 3516 and NGC 5899,also in association with regions of higher velocity dispersion MNRAS , 1–19 (2019) eeding and Feedback in AGN NGC 5506
Figure 6.
As in Fig. 2 for NGC 5506. The cyan square in the Pa β flux map shows the region in which we had to subtract the contributionof a broad component (although less broad than in the previous Seyfert 1 galaxies). This galaxy is individually discussed in Sec. 6.5 than the surroundings, suggesting the contribution ofshocks for the gas emission.The [Fe ii ] λ µ m/[P ii ] λ µ m line ratio can alsobe an indicator of shocks (Storchi-Bergmann et al. 2009), ifits value becomes larger than ≈
2. The corresponding mapsare compact, due to the small extent of the [P ii ] emission,with typical values of ≈ ≈
10 at ≈ . ′′ , thus supporting the presence of shocks. In the cases of NGC 3227 and NGC 5899, the off-nuclear valuesreach values larger than 10, also in association with high[Fe ii ] σ values, supporting even stronger contribution fromshocks.The H λ µ m/Br γ , used to investigate the origin ofthe H excitation, is usually in the range 0.6 to 2 for Seyfertgalaxies (Rodr´ıguez-Ardila et al. 2005; Riffel et al. 2008b,2009b; Storchi-Bergmann et al. 2009). Most measured val- MNRAS000
10 at ≈ . ′′ , thus supporting the presence of shocks. In the cases of NGC 3227 and NGC 5899, the off-nuclear valuesreach values larger than 10, also in association with high[Fe ii ] σ values, supporting even stronger contribution fromshocks.The H λ µ m/Br γ , used to investigate the origin ofthe H excitation, is usually in the range 0.6 to 2 for Seyfertgalaxies (Rodr´ıguez-Ardila et al. 2005; Riffel et al. 2008b,2009b; Storchi-Bergmann et al. 2009). Most measured val- MNRAS000 , 1–19 (2019) Astor J. Sch¨onell Jr.
NGC 5899
Figure 7.
As in Fig. 2 for NGC 5899, but with the large-scale image from SDSS at the z-band (Baillard et al. 2011). This galaxy isindividually discussed in Sec. 6.6. ues are indeed in this range, with a general behavior of show-ing the lowest ratios at the nucleus and increasing outwards,that we attribute to the destruction of the H molecule bythe strong radiation field close to the nucleus. The low-est values – lower than 2 – are observed for NGC 788 andNGC 5506, while the highest, reaching values in the range 8–10, are seen again in NGC 3227, NGC 3516 and NGC 5899,at similar locations as those where an enhancement is seen in [Fe ii ]/Pa β and [Fe ii ] σ values, supporting again a contri-bution from shocks. The line-of-sight velocity (V
LOS ) fields are shown in the sec-ond line of Figs. 2 – 7, after the subtraction of the systemicvelocity obtained through the fit of a rotating disk model to
MNRAS , 1–19 (2019) eeding and Feedback in AGN the H velocity field. Again, here we just describe the gen-eral features of these velocity fields, as a proper discussionis deferred to Paper B.The most common characteristic of the velocity fieldsis the rotation pattern, that is clearer in the H veloc-ity maps (presenting the typical “spider diagram” structure(Binney & Tremaine 1987)). The H velocity amplitudes issimilar to those also seen in the [Fe ii ] and Pa β velocity fields,except in the case of NGC 788, for which the H rotationamplitude is smaller. The rotation pattern can also be seenin the [Fe ii ] and Pa β velocity fields, but in these cases itis usually disturbed – more in the case of [Fe ii ] than in thecase of Pa β – indicating the presence of additional kinematiccomponents, usually associated with enhanced σ values. These maps are shown in the third line of Figs. 2 – 7.The [Fe ii ] σ maps usually show the highest values,reaching up to ≈ − , while the H maps show thelowest values, in the range ≈ − . The Pa β σ mapsare similar to those of [Fe ii ] although reaching somewhatoverall lower values, in the range ≈
40 km s − to 250 km s − .An exception is the case of NGC 5506 that shows regions ofenhanced σ to ≈
350 km s − ) for all emission lines, extendedperpendicularly to the major axis for [Fe ii ] and Pa β and inpatches parallel to the major axis in H ; such patches arealso seen in [Fe ii ]. h and h Gauss-Hermite moments
The h and h maps show values ranging from –0.3 to 0.3for all emission lines. An inverse correlation between the h map and the velocity fields is observed for all emission-lines:positive values of h are seen at the locations where negativevelocities are observed in the velocity fields, while negativevalues are seen where positive velocities are observed. Thismeans that red wings are observed in blueshifted emissionlines, while blue wings are observed in redshifted emissionlines.We also observe in some cases (e.g. for the three emis-sion lines in NGC 3227 and some of the lines in the othergalaxies) an inverse correlation between the h and the ve-locity dispersion maps. Positive values and highest values of h are seen at the locations showing the lowest velocity dis-persion values, while values zero or most negative ones areseen in the regions with the highest velocity dispersion. The measurement of gas masses and surface mass densitieswithin the inner few hundred pc of the host galaxies of AGNcan be used to evaluate the gas reservoir available to triggerand maintain the nuclear activity, as well as the formationof new stars in the circumnuclear region. The presence ofrecent star formation in the circumnuclear region of activegalaxies has been evidenced via the observation of low stellarvelocity dispersion structures (Riffel et al. 2017) and theirassociation with young stars (e.g. Storchi-Bergmann et al.2012). We can also use gas mass estimates in the vicinity of AGN to look for a possible correlation between the amountof available gas and the power of the AGN.We have calculated the ionized gas masses, inunits of solar masses (M ⊙ ) using the following ex-pression (e.g. Scoville et al. 1982; Riffel et al. 2008a;Storchi-Bergmann et al. 2009; Sch¨onell et al. 2014, 2017): M HII ≈ . × (cid:18) F Pa β erg s − cm − (cid:19) (cid:18) D Mpc (cid:19) [ M ⊙ ] , (2)where F Pa β is the flux in the Pa β line and D is the distanceto the galaxy.This equation was obtained from Scoville et al. (1982)assuming an electron temperature of T = K, elec-tron density of N e = cm − and case B recombination(Osterbrock & Ferland 2006), values applicable to the innerkiloparsec of active galaxies. The flux of Br γ used in the orig-inal equation was replaced by the stronger Pa β one assumingthe theoretical ratio F Pa β / F Br γ =5.88 (Osterbrock & Ferland2006).For NGC 5506, we did use the flux of Br γ emissionline instead of Pa β , as the signal to noise in the J-bandspectrum is worse than in the K-band one, particularly inthe extranuclear regions.The mass of the warm molecular gas (in M ⊙ ) was ob-tained using (Scoville et al. 1982): M H ≈ . × (cid:18) F H λ . erg s − cm − (cid:19) (cid:18) D Mpc (cid:19) [ M ⊙ ] , (3)where F H λ . is the flux for the corresponding emissionline. In the derivation of the above equation by Scoville et al.(1982) it was assumed that the vibrational temperature is T = K, indicating a thermalized gas which is validfor n H > . cm − . Previous studies by members of ourgroup have found that the assumption of thermal equi-librium is indeed applicable for nearby AGNs and thederived temperatures are very close to the above value(Storchi-Bergmann et al. 2009; Diniz et al. 2015).Uncertainties in the derived mass values were obtainedconsidering that the mass derived from both Eq. 2 and 3are directly proportional to the line fluxes. All the mass un-certainties are thus lower than 30%, as we have masked outregions with flux uncertainties larger than this value. Theactual uncertainties may be somewhat higher due to furtheruncertainties in the absolute flux calibration. In addition, wecannot derive precise physical parameters such as gas den-sity and temperature, thus, we have to rely on the aboveassumptions even though they are just “educated estimates”based on our previous similar studies.We show the surface mass density Σ distributions of ion-ized and warm molecular gas – Σ HII and Σ H , respectively –in the first two columns of panels of Fig. 8, and the cor-responding azymuthally averaged profiles in the third col-umn, in units of solar masses per parsec square. The surfacemass densities were obtained by calculating the gas massesin each spaxel using the equations above and dividing themby the surface area of each spaxel, taking into account pro-jection effects via the use of the disk inclinations quoted inRiffel et al. (2018). The resulting ionized gas surface massdensity distributions are in some cases more compact whencompared with that of the warm molecular gas, that seems MNRAS000
The h and h maps show values ranging from –0.3 to 0.3for all emission lines. An inverse correlation between the h map and the velocity fields is observed for all emission-lines:positive values of h are seen at the locations where negativevelocities are observed in the velocity fields, while negativevalues are seen where positive velocities are observed. Thismeans that red wings are observed in blueshifted emissionlines, while blue wings are observed in redshifted emissionlines.We also observe in some cases (e.g. for the three emis-sion lines in NGC 3227 and some of the lines in the othergalaxies) an inverse correlation between the h and the ve-locity dispersion maps. Positive values and highest values of h are seen at the locations showing the lowest velocity dis-persion values, while values zero or most negative ones areseen in the regions with the highest velocity dispersion. The measurement of gas masses and surface mass densitieswithin the inner few hundred pc of the host galaxies of AGNcan be used to evaluate the gas reservoir available to triggerand maintain the nuclear activity, as well as the formationof new stars in the circumnuclear region. The presence ofrecent star formation in the circumnuclear region of activegalaxies has been evidenced via the observation of low stellarvelocity dispersion structures (Riffel et al. 2017) and theirassociation with young stars (e.g. Storchi-Bergmann et al.2012). We can also use gas mass estimates in the vicinity of AGN to look for a possible correlation between the amountof available gas and the power of the AGN.We have calculated the ionized gas masses, inunits of solar masses (M ⊙ ) using the following ex-pression (e.g. Scoville et al. 1982; Riffel et al. 2008a;Storchi-Bergmann et al. 2009; Sch¨onell et al. 2014, 2017): M HII ≈ . × (cid:18) F Pa β erg s − cm − (cid:19) (cid:18) D Mpc (cid:19) [ M ⊙ ] , (2)where F Pa β is the flux in the Pa β line and D is the distanceto the galaxy.This equation was obtained from Scoville et al. (1982)assuming an electron temperature of T = K, elec-tron density of N e = cm − and case B recombination(Osterbrock & Ferland 2006), values applicable to the innerkiloparsec of active galaxies. The flux of Br γ used in the orig-inal equation was replaced by the stronger Pa β one assumingthe theoretical ratio F Pa β / F Br γ =5.88 (Osterbrock & Ferland2006).For NGC 5506, we did use the flux of Br γ emissionline instead of Pa β , as the signal to noise in the J-bandspectrum is worse than in the K-band one, particularly inthe extranuclear regions.The mass of the warm molecular gas (in M ⊙ ) was ob-tained using (Scoville et al. 1982): M H ≈ . × (cid:18) F H λ . erg s − cm − (cid:19) (cid:18) D Mpc (cid:19) [ M ⊙ ] , (3)where F H λ . is the flux for the corresponding emissionline. In the derivation of the above equation by Scoville et al.(1982) it was assumed that the vibrational temperature is T = K, indicating a thermalized gas which is validfor n H > . cm − . Previous studies by members of ourgroup have found that the assumption of thermal equi-librium is indeed applicable for nearby AGNs and thederived temperatures are very close to the above value(Storchi-Bergmann et al. 2009; Diniz et al. 2015).Uncertainties in the derived mass values were obtainedconsidering that the mass derived from both Eq. 2 and 3are directly proportional to the line fluxes. All the mass un-certainties are thus lower than 30%, as we have masked outregions with flux uncertainties larger than this value. Theactual uncertainties may be somewhat higher due to furtheruncertainties in the absolute flux calibration. In addition, wecannot derive precise physical parameters such as gas den-sity and temperature, thus, we have to rely on the aboveassumptions even though they are just “educated estimates”based on our previous similar studies.We show the surface mass density Σ distributions of ion-ized and warm molecular gas – Σ HII and Σ H , respectively –in the first two columns of panels of Fig. 8, and the cor-responding azymuthally averaged profiles in the third col-umn, in units of solar masses per parsec square. The surfacemass densities were obtained by calculating the gas massesin each spaxel using the equations above and dividing themby the surface area of each spaxel, taking into account pro-jection effects via the use of the disk inclinations quoted inRiffel et al. (2018). The resulting ionized gas surface massdensity distributions are in some cases more compact whencompared with that of the warm molecular gas, that seems MNRAS000 , 1–19 (2019) Astor J. Sch¨onell Jr. to extend farther and more uniformly from the nucleus thanthe ionized gas.The azimuthaly-averaged profiles of Σ HII and Σ H ,shown in the last column of Fig. 8, reveal that typical ra-tios between the two surface mass densities within the inner ≈
100 pc are of the order of 10 and that Σ gradients aresteeper for the ionized than for the molecular gas.In Table 3 we show the integrated gas mass values (forthe whole FOV) as well as the average surface mass den-sities (in units of M ⊙ pc − ) obtained as the ratio betweenthe integrated masses and the area over which they are dis-tributed, listed also in the Table. We also include in Ta-ble 3 an estimate of the cold molecular gas mass. A num-ber of studies have derived the ratio between the cold andwarm H gas masses by comparing the masses obtained us-ing the cold CO molecular lines observed in millimetric wave-lenghts with that of the warm H observed in the near-IR.Dale et al. (2005) obtained ratios in the range –10 , whileM¨uller S´anchez et al. (2006), using a larger sample of 16 lu-minous and ultra-luminous infrared galaxies, derived a ra-tio M cold /M warm = 1 – 5 × . More recently, Mazzalay et al.(2013) compiled from the literature values of M cold derivedfrom CO observations and H (2.12 µ m) luminosities for alarger number of galaxies, covering a wider range of lu-minosities, morphological and nuclear activity types. Fromthose data, they propose that an estimate of the cold H gasmass can be obtained from the flux of the warm H λ M H cold ≈ (cid:18) L H λ . L ⊙ (cid:19) , (4)where L H λ . is the luminosity of the H (2.12 µ m) line.The error associated to the factor β = ≈ gas(within typically 100 – 400 pc from the nucleus) ranges from106 M ⊙ for Mrk 607 to 820 M ⊙ for NGC 5506, while the esti-mated mass of cold molecular gas ranges from 8 × M ⊙ forMrk 607 to 6 × M ⊙ for NGC 5506. The masses of ionizedgas range from 1.8 × M ⊙ for NGC 3516 to 1.9 × M ⊙ forNGC 5506. We point out again that the uncertainties in thecalculated masses, presented in Tab. 3, were obtained onlyby propagating the line flux uncertainties and are thus lowerlimits, considering that there are additional uncertainties inadopted physical properties – e.g. on the M H cold / M H ratio,H vibrational temperatures and electron densities. In this section we discuss individually each galaxy presentinga brief review of previous results from the literature, tryingto relate them with ours. We also discuss and analyze themasses of ionized and molecular gas, as well as their sur-face mass density distributions and the global properties ofthe gas kinematics, leaving the kinematic modelling and theanalysis of the gas kinematics and excitation to Paper B.
NGC 788 (Fig. 2) was identified as a Seyfert galaxyby Huchra et al. (1982), and has been observed inthe optical (Hamuy & Maza 1987; Wagner 1988; Kay1994; Cruz-Gonzalez et al. 1994), radio (Ulvestad & Wilson1989), and millimetre (Heckman et al. 1989) wavebands.This galaxy is an early-type spiral with faint arms visible upto ≈ . ′′ from the nucleus (Evans et al. 1996), most conspic-uous to the north-west of the nucleus with a string of brightcompact H ii regions, while a complex of fainter H ii regionsis associated with a southern arm (Evans et al. 1996).The kinematics major axis PA ≈ ◦ was determinedfrom the V LOS fields (Fig. 2), with a value close to that ob-tained by Riffel et al. (2017) for the stellar kinematics of ≈ ◦ . The velocity amplitude observed for [Fe ii ] and Pa β ,of ≈
150 km s − is much higher than that for H , of only ≈
50 km s − . These higher velocities are associated with amore collimated emission and higher velocity dispersionssuggesting the presence of a bipolar outflow along the east-west direction, as suggested from the morphology of the gasemission in the [Fe ii ] flux map.The E ( B − V ) map from Fig. 2 has the highest values tosouthwest of the nucleus, which coincide with a dust laneseen in the HST - F606W optical continuum image fromMartini et al. (2003). Mrk 607 (Fig. 3) is an Sa galaxy hosting a Seyfert 2 nucleus.A continuum image from Ferruit et al. (2000) shows a highellipticity of ≈ ≈ ◦ ), withmajor axis PA of − ◦ , which agrees with our estimate ofthe kinematic major axis orientation of ≈ − ◦ from the H velocity field (Paper B).Our data reveal that the three velocity fields – [Fe ii ],Pa β and H – show what seems to be a rotation pattern,very steep in the center in H and less steep in the first twolines. Spiral dust lanes are clearly visible in the HST F547continuum image of the inner 8 ′′ (1.4 kpc) (Ferruit et al.2000). The high inclination of the galaxy may explain thelarge velocity dispersions observed. In the central 5 ′′ (875pc) it presents weak radio emission extended along PA ≈ ◦ (Colbert et al. 1996; Nagar et al. 1999) which is notaligned with the axis of the most extended gas emission,which is observed along the galaxy major axis.The E ( B − V ) map from Fig. 3 has high values ( ≈ NGC 3227 (Fig. 4) is a well studied barred galaxy, witha Seyfert 1.5 nucleus (Ho et al. 1997) and in interactionwith the dwarf elliptical galaxy NGC 3226 (Mundell et al.2004). The galactic disc has an inclination of 56 ◦ , with anouter photometric major-axis at a position angle PA = 158 ◦ (Mundell et al. 1995). This value agrees with the PA ≈ ◦ we have obtained from the fit of the H V LOS velocity field(Paper B), and with the value of 156 ◦ found by Riffel et al.(2017) from the stellar kinematics. The three emission-line MNRAS , 1–19 (2019) eeding and Feedback in AGN Figure 8.
Surface gas mass density distributions Σ : in the first column we show Σ HII in units of M ⊙ pc − ); in the second column Σ H inunits of 10 − M ⊙ pc − , and in the third the corresponding azymuthaly averaged gradients, showing more centraly peaked distributionsfor the ionized gas.MNRAS000
Surface gas mass density distributions Σ : in the first column we show Σ HII in units of M ⊙ pc − ); in the second column Σ H inunits of 10 − M ⊙ pc − , and in the third the corresponding azymuthaly averaged gradients, showing more centraly peaked distributionsfor the ionized gas.MNRAS000 , 1–19 (2019) Astor J. Sch¨onell Jr.
Table 3.
Areas, masses and average surface mass density Σ values for the ionized, warm and estimated cold molecular gas within theGemini NIFS FOV.Galaxies Area (H ) Area (H ii ) M (H ) hot M (H ) cold M (H ii ) Σ (H ) hot Σ (H ) cold Σ (H ii )10 pc pc M ⊙ M ⊙ M ⊙ − M ⊙ /pc M ⊙ /pc M ⊙ /pc NGC 788 6.2 5.5 5.9 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
12 1.8 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± velocity fields show a rotation pattern, that is neverthe-less distorted due to the presence of additional components,mainly to the north-east, east and south-east. These compo-nents are most conspicuous in [Fe ii ] and Pa β velocity mapsand are associated with regions of enhanced velocity disper-sion.The central region has been mapped in CO(1-0) and CO(2-1) by Schinnerer et al. (2000), who detected molec-ular gas very close to the nucleus (within ≈
13 pc), inagreement with our results that show a large moleculargas concentration towards the centre (see the surface massdensity profile in the third line of Fig. 8). In addition,Schinnerer et al. (2000) found an asymmetric nuclear ringwith a diameter of about 3 . ′′ ii ]/Pa β and H /Br γ of Fig. 4.The inner kiloparsec hosts a radio jet at PA ≈ − ◦ anda conical NLR outflow at PA ≈ ◦ (Mundell et al. 1995),while Arribas & Mediavilla (1994) report an H α outflow atPA ≈ ◦ . The extent of the [Fe ii ] emission to the north,as well as its enhanced velocity dispersion may be relatedto the radio emission, while the outflows observed at PAsbetween 15 ◦ and 50 ◦ can explain the deviation from rotationobserved in the velocity fields of [Fe ii ] and Pa β to the north-east. Signatures of gas outflows in NGC 3227 have also beenobserved by Davies et al. (2014) in the inner ′′ − ′′ usingnear-IR IFS observations with the instrument SINFONI atthe VLT. A V − H colour map shows dust lanes to the south-west of the nucleus (Martini et al. 2003; Davies et al. 2014),which are co-spatial with the highest E ( B − V ) values seen inFig. 4. The Seyfert 1 nucleus of this SB0 galaxy (Fig. 5)shows variable ultraviolet absorption lines (Voit et al.1987; Kriss et al. 1996) as well as variable broad emis-sion lines and continuum (Koratkar et al. 1996). Spec-troscopic studies of the gaseous kinematics using eitherlong-slit (Ulrich & Pequignot 1980; Goad & Gallagher 1987,1988; Mulchaey et al. 1992) or integral field spectroscopy(Veilleux et al. 1993; Aoki et al. 1994; Arribas & Mediavilla1994) have outlined multiple spectral components display-ing strong deviations from “normal” galaxy rotation. Twogeneral models have been proposed to explain the mor-phology and the kinematics of the emission-line gas thatpresents a Z-shaped structure covering the inner 5 . ′′
0: a bentbipolar mass outflow, first suggested by Goad & Gallagher (1987) and further developed by Mulchaey et al. (1992) andVeilleux et al. (1993), and a precessing twin-jet model byVeilleux et al. (1993).The stellar rotation curve obtained byArribas et al. (1997) yields a systemic velocity of 2593 ± − , close to that derived by Riffel et al. (2017) of ≈ − . The kinematic major axis obtained by Arribas et al.(1997) has PA= 53 ◦ ± ◦ and is consistent with our resultsfor the stellar kinematics (Riffel et al. 2017) and also withthe orientation of the kinematic major axis of the moleculargas of ≈ ◦ . The nucleus of NGC 5506 (Fig. 6) is classified as a Sy1.9based on the detection of broad wings in the Pa β profile(Blanco et al. 1990). However, more recently, Nagar et al.(2002) presented evidence that NGC 5506 is an ob-scured narrow-line Sy1, via the detection of the permittedO i λ µ m line, together with a broad pedestal of Pa β andrapid X-ray variability. The galaxy nucleus is very compactin the mid-IR, with an apparent optical depth of τ app µ m ≈ iii ] and [N ii ] narrow lines (Zaw et al.2009). The log([O iii ]/H β ) and log([N ii ]/H α ) line ratios, 0.88and − . respectively (Kewley et al. 2001), place it firmlyin the Seyfert region of the BPT diagram.NGC 5506 is close to edge-on, with an inclination of75 ◦ , above and below which ionized gas is in outflow withincones with an opening angle of ≈ ◦ (Maiolino et al. 1994).These outflows are consistent with the distribution of en-hanced sigma values we observe in the ionized gas in a ver-tical structure crossing the galaxy plane and opening to thenorth and south of it (Fig. 6). The outflow to the north isalso well traced by the [Fe ii ] flux distribution (its highestemission shows an approximately conical shape), while theH flux distribution seems to better trace the gas in thegalaxy plane, as it is oriented along this direction.As observed in Fig. 1, the nucleus of NGC 5506 shows asteep red continuum over the . − . µ m spectral range thatcan be attributed to a blackbody source with temperature ≈ V − H dust map from Martini et al. (2003), theorigin of the K-band continuum is most probably hot dustemission from the AGN torus, as already observed for other MNRAS , 1–19 (2019) eeding and Feedback in AGN active galaxies (e.g. Riffel et al. 2009c; Burtscher et al. 2015;Diniz et al. 2018). This inclined SAB(rs)c galaxy (Fig. 7) is reported to bein a pair and presents an optical Seyfert 2 spectrum(de Vaucouleurs et al. 1991). The stellar kinematics derivedby Riffel et al. (2017) gives a PA ≈ ◦ , that is somewhat dis-tinct from that obtained by us for the molecular gas (PaperB), of ≈ ◦ . The kinematics of the ionized gas is very differentfrom that of the molecular gas (Fig. 7): while the later seemsto be dominated by rotation in the inclined galaxy plane,with blueshifts to the north and redshifts to the south, theionized gas kinematics shows velocities that are opposite tothat observed in the H velocity field. A possible interpreta-tion is that the ionized gas is tracing an outflow at PA ≈ ◦ ,as supported also by the north-south elongation observedmostly in the [Fe ii ] flux map and by the enhanced velocitydispersion observed at these locations. Fig. 8 shows that the molecular gas seems to be more evenlydistributed over the field of view than the ionized gas, whichis more concentrated towards the nucleus and shows a morepatchy and sometimes collimated mass distribution. Themore peaked distribution of the ionized gas is clearly seenin the average surface mass density profiles shown in thethird column of Fig. 8 for the galaxies NGC 788, Mrk 607,NGC 3227 and NGC 5506. In the cases of NGC 3516 andNGC 5899, the average profiles of the molecular and ionizedgas surface mass densities are similar to each other. Theprevalence of more concentrated ionized gas profiles thanthose in H can be attributed to the fact that the neutralgas rapidly absorbs the ionizing photons from the AGN, thusconcentrating in its vicinity and/or along the ionization axis– the preferred direction of escape of the AGN radiation.The H , on the other hand, has another source of exci-tation, as shown in our previous works (e.g. Ilha et al. 2016),where the typical temperature is ≈ molecule canbe attributed to X-rays originating in the AGN. These X-rays penetrate deep along the galaxy plane in every directionheating the region and exciting the H molecule. In addition,in the regions closest to the AGN, where the temperaturecan reach ∼ . K, the H molecules are destroyed, andwe only see those more deeply embedded in the circumnu-clear dust, where they are shielded from the strongest AGNradiation.We thus conclude that the distinct nature of the ex-citation of the ionized and the molecular gas and differentphysical conditions of the regions where they originate canexplain the difference in the gas surface mass density pro-files. In this section we discuss briefly the global properties of the gas kinematics, with the goal of just highlighting apparentcorrelations among properties and differences between themolecular and ionized gas kinematics. As we already pointedout, we will leave the modeling of the velocity fields, thequantification of possible inflows and outflows, as well asthe determination of the impact of the outflows on the hostgalaxy (feedback) to Paper B.A global property of the gas kinematics, as seen inFigs. 2–7, is that the warm H velocity field is dominatedby rotation in the plane of the galaxy. Although the ion-ized gas velocity fields show also signatures of rotation, therotation pattern is distorted due to the presence of addi-tional components that can be attributed to outflows andwhich are usually associated with an increase in the velocitydispersion.The above results are in agreement with those of previ-ous studies by our AGNIFS group, in which we have addi-tionally reported, for H , the presence of inflows in a numberof galaxies. We have also found that outflows are most easilyseen in the ionized gas (Storchi-Bergmann & Schnorr-M¨uller2019) and that this gas shows a more centrally concentratedflux distribution than that of H Riffel et al. (2018), also inagreement with what we found in the present paper.The most frequent presence of outflows in the ionizedgas than in the molecular gas can be understood as due tothe fact that the ionized gas is mostly concentrated along theionization axis, where it can be easily pushed by an AGNoutflow. The molecular gas, on the other hand, is usually de-stroyed by the hard AGN radiation at these locations. Thus,although the H flux distributions appear also surroundingthe nucleus, it is not co-spatial with the ionized gas, as itoriginates in gas at temperatures much lower (2000K) thanthose characteristic of the ionized gas (10000K). This is alsosupported by the lower velocity dispersion of H . Clear ex-amples are the cases of NGC 4151 (Storchi-Bergmann et al.2009) and NGC 1068 (Barbosa et al. 2014) that show H flux distributions surrounding the nucleus but avoiding theionization axis. On the other hand, molecular gas outflowsare observed in a number of cases, and in distant AGN hosts(Emonts et al. 2017). A possibility in these cases is that anuclear outflow can push the surrounding gas including theH gas which may be shielded from the ionizing radiationby dust, for example. We hope to re-visit this point in PaperB. Regarding the presence of outflows, we have found sig-natures of them in most galaxies of our sample, most clearlyobserved in the [Fe ii ] kinematics: (1) in NGC 788, along theeast-west direction; (2) in NGC 3227, towards the northeast,as suggested by distortions in the velocity field and patternsin h and h ; (3) in NGC 3516, from the south-east to thenorth-west, as suggested by the disturbed velocity field andenhanced velocity dispersion; (4) in NGC 5506, along north-south, as suggested by the enhanced velocity dispersions; (5)and in NGC 5899, towards north and south, as indicated bythe opposite velocity field between the ionized and moleculargas. Another global property of the sample is the fact thatthe h Gauss-Hermite moment shows an inverse correlationwith the V
LOS in a number of cases: (1) in the Pa β line forNGC 788; (2) in Pa β and [Fe II] for Mrk 607; (3) in Pa β andH in NGC 3227; (4) In Pa β and H for NGC 3516; (5) in Br γ and H in NGC 5506; (6) in H in NGC 5899. This means MNRAS , 1–19 (2019) Astor J. Sch¨onell Jr.
Table 4.
Mass accretion rates Û m , average SFR surface densities Σ SFR , and total SFRs.Galaxy Û m < Σ SFR > SFR10 − M ⊙ yr − − M ⊙ yr − kpc − − M ⊙ yr − NGC788 49.2 2.5 ± ± ± ± ± ± ± ± ± ± ± ± that there are red wings in centrally blueshifted profiles andblue wings in redshifted ones. One possible explanation is theeffect known as “asymmetric drift” (Westfall et al. 2007): gasrotating in the galaxy plane gives origin to the “main” ve-locity field – corresponding to the emission-line peaks, whiletenuous gas at higher latitudes rotating with lower velocity– thus lagging behind the rotation in the plane – gives originto the wings of the profiles.We also observe in some cases, an inverse correlationbetween the h and the σ values – mostly positive h valuesat the locations with low σ , and only a few negative valuesat locations with high σ . This has been observed in: (1) H for NGC 788; (2) in the three emission lines for NGC 3227;(3) in Pa β and H for NGC 3516; (4) in Br γ for NGC 5506;(5) in the three emission lines for NGC 5899.Positive h values indicate profiles more “peaky” than aGaussian curve, but with broader wings, while negative val-ues indicate profiles less peaky and with less extended wingsthan a Gaussian. We tentatively interpret the positive h val-ues as being mostly due to the emission of gas rotating inthe disk, with low σ values, while the wings could originatein diffuse gas emission, that may extend to high latitudes,increasing the range of velocities probed by the emission.The high σ and negative h values are rarer, and could arisein spatially unresolved double components, and may be re-lated to the presence of an outflow, as seems to be the casefor the [Fe II] line in NGC 3227, for example (Fig. 4). In order to evaluate if the gas reservoirs accumulated in theinner few hundred pc of these galaxies have enough mass tofeed the AGN, we first estimate the AGN accretion rates Û m using: Û m = L bol c η , (5)where L bol is the bolometric luminosity of the AGN, c isthe light speed, and η is the conversion efficiency of restmass of the accreted material into radiation, that we haveadopted as η =0.1. The bolometric luminosities were deter-mined by Riffel et al. (2018) based on the hard X-ray lu-minosities. The only exception is the case of Mrk 607, forwhich the [O iii ] luminosity was used instead, adopting abolometric correction factor of 3500 (Heckman et al. 2004).The resulting accretion rates are presented in Table 4.As the estimated mass accretion rates to the AGN are in the range 0.1 – 50 × − M ⊙ yr − , considering an AGN ac-tivity cycle of 10 – 10 yr, and assuming that most of theionized and molecular gas are concentrated within the in-ner few 100 pc of the galaxies (Fig. 8), it can be concludedthat the ionized gas mass alone would be enough to feed theAGN. Nevertheless, Table 3 shows that the estimated massesof the cold molecular gas are larger, and range from 10 to10 M ⊙ ; therefore, there seems to be at least ≈ timesmore gaseous mass in the inner few 100 pc of these galaxiesthan that necessary to feed the AGN.The fate of the gas that is not used to feed the AGNcan be: (1) be consumed by star formation; (2) be pushedaway by AGN feedback; (3) be pushed away by stellar feed-back (e.g. Hopkins et al. 2016). Most probably more thanone process will occur.Here we will discuss only the possibility that the gasaccumulated in the nuclear region will lead to the formationof new stars, thus calculating the star formation rate (SFR)in the inner few 100 pc. The feedback due to the observedoutflows and possibility of stellar feedback will be discussedin Paper B.Schmidt (1959) has shown that the SFR is directly re-lated to the gas density, and later Kennicutt (1998) pro-posed a relation between the SFR surface density Σ SFR andthe ionised gas mass surface density Σ HII , as follows: Σ SFR M ⊙ yr − kpc − = ( . ± . ) × − (cid:18) Σ HII M ⊙ pc − (cid:19) . , (6)Using the relation above, we obtain the mean values for Σ SFR listed in Table 4, which range from 1 × − to 0.26 M ⊙ yr − kpc − . Using the areas quoted in Table 3 for the regionsoccupied by the ionized gas we estimate total star formationrates for the area covered by our observations (radius of ≈
300 pc) in the range 10 − – 10 − M ⊙ yr − (Table 3). Thesevalues are within the range of values observed for the innerfew 100 pc of galaxies and circumnuclear star-forming re-gions (Shi et al. 2006; Dors et al. 2008; Falc´on-Barroso et al.2014; Tsai & Hwang 2015; Riffel et al. 2016).We therefore conclude that the mass reservoirs in theinner 300 pc of the sample galaxies can not only power thecentral AGN but also form new stars at low SFR ( ≤ − M ⊙ yr − ). The presence of recently formed stars in the inner few100 pc of AGN is supported by the observation of low-stellarvelocity dispersion ( σ ∗ ) structures in 10 of 16 galaxies of oursample for which we could measure σ ∗ (Riffel et al. 2017). We have mapped the ionized and molecular gas flux dis-tributions, excitation and kinematics in the inner kpc of 6nearby active galaxies using adaptative optics assisted NIRJ- and K-band integral field spectroscopy obtained with theGemini NIFS instrument. The main conclusions of this workare listed below. • The flux distributions are usually distinct for the ion-ized and molecular gas: while the former is more concen-trated and sometimes collimated along a preferred axis, thelatter is distributed more uniformly over the galaxy plane.These flux distributions lead to azimuthally averaged surfacemass density profiles steeper for the ionized gas than for the
MNRAS , 1–19 (2019) eeding and Feedback in AGN molecular gas. We attribute this difference to the differentexcitation mechanisms: while the ionized gas is excited bythe AGN radiation in regions with temperatures of about10000 K, close to the AGN, the molecular gas is thermallyexcited in regions of lower temperatures of about 2000 K,that extend farther from the nucleus; • The gas kinematics is also distinct: while the molec-ular gas is mostly rotating in the galaxy plane with lowvelocity dispersions, the ionized gas frequently shows othercomponents associated with higher velocity dispersions anddistorted velocity fields suggesting outflows; • Signatures of outflows are mostly observed in the [Fe ii ]kinematics of NGC 788, NGC 3227, NGC 3516, NGC 5506,NGC 5899. The modelling of the gas kinematics and quan-tification of the mass outflow rates and powers will be pre-sented in a forthcoming paper (Paper B); • There is usually an inverse correlation between the h Gauss-Hermite moment and the velocity field: positive val-ues (red wings) are associated to blueshifts and negativevalues (blue wings) are associated to redshifts. This can beunderstood as due to the gas rotating closer to the galaxyplane originating the “main” velocity field – correspondingto the emission-line peaks, while tenuous gas at higher lat-itudes rotating with lower velocities originate the “lagging”wings; • There is in some cases also an inverse correlation be-tween h and the velocity dispersion maps: low values of σ correspond to positive values of h (peaky profiles withextended wings), while high σ values correspond to nega-tive values of h (boxy profiles). We see mostly positive h values, which we attribute to gas rotating in the disk (low σ ), on which less luminous emission of hotter gas at higherlatitudes gives origin to the broad wings; • Although the excitation will be further analised in Pa-per B, general trends observed in the emission-line ratiosare: (1) an increase in [Fe ii ] λ µ m/Pa β in associationwith higher velocity dispersion indicating contribution fromshocks, and an increase in H λ . µ m/Br γ outwards, at-tributed to destruction of the H molecule close to the AGNdue to its strong radiation; • The integrated mass of ionized gas within the inner ≈
300 pc radius ranges from 1.8 × M ⊙ to 1.9 × M ⊙ ,while that of warm molecular gas is ∼ − times lower,and the estimated mass in cold molecular gas is ∼ timeshigher; • The average ionized gas surface mass density rangesfrom 2.8 M ⊙ pc − to 140 M ⊙ pc − , while for the warm molec-ular gas it is ∼ − times lower, and that estimated for thecold molecular gas is ∼ times higher. • The AGN accretion rates in our sample range from0.1 × − M ⊙ yr − to 49 × − M ⊙ yr − ; considering an ac-tivity cycle of duration 10 – 10 yr, it can be concludedthat there are ≈ times more gas in the inner few hun-dred pc of the galaxies than needed to feed the AGN over aduty cycle; • If most of this gas will lead to star formation in the innerfew 100 pc, we estimate star formation rates in the range1–260 × − M ⊙ yr − which is within the range of typicalvalues observed for circumnuclear star-forming regions innearby galaxies; • The mass reservoirs in the inner few 100 pc of thesegalaxies are thus enough to power both the central AGN and star formation. But our observations show also that atleast part of this gas is being pushed away by an AGN-driven outflow or supernovae winds, which we will furtherinvestigate via the analysis of the gas kinematics in PaperB. This paper is the thirteenth of a series by our group AG-NIFS in which we have been mapping in detail AGN feedingand feedback processes in nearby galaxies using adaptativeoptics assisted integral field observations with NIFS. And itis the third in which we aim to characterize a global sampleof 20 nearby active galaxies, the first one in which we char-acterized the stellar kinematics (Riffel et al. 2017) and thesecond in which we presented the sample and mass densityprofiles (Riffel et al. 2018). The observations are scheduledto be concluded in 2019. Further and more detailed analysisof the the gas kinematics and excitation for the six galaxiespresented in this work will be presented in the forthcomingPaper B.
ACKNOWLEDGMENTS
We thank an anonymous referee for valuable suggestionswhich helped to improve the paper. This study was financedin part by the Coordena¸c˜ao de Aperfei¸coamento de Pes-soal de N´ıvel Superior - Brasil (CAPES) - Finance Code001, Conselho Nacional de Desenvolvimento Cient´ıfico e Tec-nol´ogico (CNPq) and Funda¸c˜ao de Amparo `a pesquisa doEstado do RS (FAPERGS). 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-east CYT (Argentina).
REFERENCES
Aoki K., Ohtani H., Yoshida M., Kosugi G., 1994, PASJ, 46, 539Arribas S., Mediavilla E., 1994, ApJ, 437, 149Arribas S., Mediavilla E., Garc´ıa-Lorenzo B., del Burgo C., 1997,ApJ, 490, 227Baillard A., et al., 2011, A&A, 532, A74Barbosa F. K. B., Storchi-Bergmann T., McGregor P., Vale T. B.,Rogemar Riffel A., 2014, MNRAS, 445, 2353Binney J., Tremaine S., 1987, Galactic dynamicsBlanco P. R., Ward M. J., Wright G. S., 1990, MNRAS, 242, 4PBurtscher L., et al., 2015, A&A, 578, A47Cardelli J. A., Clayton G. C., Mathis J. S., 1989, ApJ, 345, 245Colbert E. J. M., Baum S. A., Gallimore J. F., O’Dea C. P.,Lehnert M. D., Tsvetanov Z. I., Mulchaey J. S., Caganoff S.,1996, ApJS, 105, 75Cruz-Gonzalez I., Carrasco L., Serrano A., Guichard J., Dultzin-Hacyan D., Bisiacchi G. F., 1994, ApJS, 94, 47Dale D. A., Sheth K., Helou G., Regan M. W., H¨uttemeister S.,2005, AJ, 129, 2197Davies R. I., Maciejewski W., Hicks E. K. S., Tacconi L. J., GenzelR., Engel H., 2009, ApJ, 702, 114Davies R. I., et al., 2014, ApJ, 792, 101MNRAS000
Aoki K., Ohtani H., Yoshida M., Kosugi G., 1994, PASJ, 46, 539Arribas S., Mediavilla E., 1994, ApJ, 437, 149Arribas S., Mediavilla E., Garc´ıa-Lorenzo B., del Burgo C., 1997,ApJ, 490, 227Baillard A., et al., 2011, A&A, 532, A74Barbosa F. K. B., Storchi-Bergmann T., McGregor P., Vale T. B.,Rogemar Riffel A., 2014, MNRAS, 445, 2353Binney J., Tremaine S., 1987, Galactic dynamicsBlanco P. R., Ward M. J., Wright G. S., 1990, MNRAS, 242, 4PBurtscher L., et al., 2015, A&A, 578, A47Cardelli J. A., Clayton G. C., Mathis J. S., 1989, ApJ, 345, 245Colbert E. J. M., Baum S. A., Gallimore J. F., O’Dea C. P.,Lehnert M. D., Tsvetanov Z. I., Mulchaey J. S., Caganoff S.,1996, ApJS, 105, 75Cruz-Gonzalez I., Carrasco L., Serrano A., Guichard J., Dultzin-Hacyan D., Bisiacchi G. F., 1994, ApJS, 94, 47Dale D. A., Sheth K., Helou G., Regan M. W., H¨uttemeister S.,2005, AJ, 129, 2197Davies R. I., Maciejewski W., Hicks E. K. S., Tacconi L. J., GenzelR., Engel H., 2009, ApJ, 702, 114Davies R. I., et al., 2014, ApJ, 792, 101MNRAS000 , 1–19 (2019) Astor J. Sch¨onell Jr.
Diniz M. R., Riffel R. A., Storchi-Bergmann T., Winge C., 2015,MNRAS, 453, 1727Diniz M. R., Riffel R. A., Dors O. L., 2018,Research Notes of the American Astronomical Society,2, 3Dors Jr. O. L., Storchi-Bergmann T., Riffel R. A., Schimdt A. A.,2008, A&A, 482, 59Emonts B. H. C., Colina L., Piqueras-L´opez J., Garcia-Burillo S.,Pereira-Santaella M., Arribas S., Labiano A., Alonso-HerreroA., 2017, A&A, 607, A116Evans I. N., Koratkar A. P., Storchi-Bergmann T., KirkpatrickH., Heckman T. M., Wilson A. S., 1996, ApJS, 105, 93Fabian A. C., 2012, ARA&A, 50, 455Falc´on-Barroso J., Ramos Almeida C., B¨oker T., Schinnerer E.,Knapen J. H., Lan¸con A., Ryder S., 2014, MNRAS, 438, 329Ferrarese L., Ford H., 2005, Space Sci. Rev., 116, 523Ferruit P., Wilson A. S., Mulchaey J., 2000, ApJS, 128, 139Goad J. W., Gallagher III J. S., 1987, AJ, 94, 640Goad J. W., Gallagher III J. S., 1988, AJ, 95, 948Hamuy M., Maza J., 1987, A&AS, 68, 383Heckman T. M., Blitz L., Wilson A. S., Armus L., Miley G. K.,1989, ApJ, 342, 735Heckman T. M., Kauffmann G., Brinchmann J., Charlot S.,Tremonti C., White S. D. M., 2004, ApJ, 613, 109Ho L. C., Filippenko A. V., Sargent W. L. W., 1997, ApJS,112, 315Hopkins P. F., Torrey P., Faucher-Gigu`ere C.-A., Quataert E.,Murray N., 2016, MNRAS, 458, 816Huchra J. P., Wyatt W. F., Davis M., 1982, AJ, 87, 1628Ilha G. d. S., Bianchin M., Riffel R. A., 2016, Ap&SS, 361, 178Kay L. E., 1994, ApJ, 430, 196Kennicutt Jr. R. C., 1998, ARA&A, 36, 189Kewley L. J., Heisler C. A., Dopita M. A., Lumsden S., 2001,ApJS, 132, 37Koratkar A., et al., 1996, ApJ, 470, 378Kormendy J., Ho L. C., 2013, ARA&A, 51, 511Kriss G. A., Espey B. R., Krolik J. H., Tsvetanov Z., Zheng W.,Davidsen A. F., 1996, ApJ, 467, 622Maiolino R., Stanga R., Salvati M., Rodriguez Espinosa J. M.,1994, A&A, 290, 40Malkan M. A., Gorjian V., Tam R., 1998, ApJS, 117, 25Markwardt C. B., 2009, in Bohlender D. A., Durand D., DowlerP., eds, Astronomical Society of the Pacific Conference SeriesVol. 411, Astronomical Data Analysis Software and SystemsXVIII. p. 251 ( arXiv:0902.2850 )Martini P., Regan M. W., Mulchaey J. S., Pogge R. W., 2003,ApJS, 146, 353Mazzalay X., et al., 2013, MNRAS, 428, 2389Mazzalay X., et al., 2014, MNRAS, 438, 2036McGregor P. J., et al., 2003, in Iye M., Moorwood A. F. M., eds,Proc. SPIEVol. 4841, Instrument Design and Performancefor Optical/Infrared Ground-based Telescopes. pp 1581–1591,doi:10.1117/12.459448Mulchaey J. S., Tsvetanov Z., Wilson A. S., Perez-Fournon I.,1992, ApJ, 394, 91M¨uller S´anchez F., Davies R. I., Eisenhauer F., Tacconi L. J.,Genzel R., Sternberg A., 2006, A&A, 454, 481M¨uller S´anchez F., Davies R. I., Genzel R., Tacconi L. J., Eisen-hauer F., Hicks E. K. S., Friedrich S., Sternberg A., 2009,ApJ, 691, 749Mundell C. G., Holloway A. J., Pedlar A., Meaburn J., KukulaM. J., Axon D. J., 1995, MNRAS, 275, 67Mundell C. G., James P. A., Loiseau N., Schinnerer E., ForbesD. A., 2004, ApJ, 614, 648Nagar N. M., Wilson A. S., Mulchaey J. S., Gallimore J. F., 1999,ApJS, 120, 209Nagar N. M., Oliva E., Marconi A., Maiolino R., 2002, A&A,391, L21 Osterbrock D. E., Ferland G. J., 2006, Astrophysics of gaseousnebulae and active galactic nucleiPrugniel P., et al., 2001, in Banday A. J., Zaroubi S., BartelmannM., eds, Mining the Sky. p. 683, doi:10.1007/10849171 92Ramos Almeida C., et al., 2009, ApJ, 702, 1127Riffel R. A., 2010, Ap&SS, 327, 239Riffel R. A., Storchi-Bergmann T., 2011, MNRAS, 417, 2752Riffel R. A., Storchi-Bergmann T., Winge C., Barbosa F. K. B.,2006, MNRAS, 373, 2Riffel R. A., Storchi-Bergmann T., Winge C., McGregor P. J.,Beck T., Schmitt H., 2008a, MNRAS, 385, 1129Riffel R., Pastoriza M. G., Rodr´ıguez-Ardila A., Maraston C.,2008b, MNRAS, 388, 803Riffel R. A., Storchi-Bergmann T., Dors O. L., Winge C., 2009a,MNRAS, 393, 783Riffel R., Pastoriza M. G., Rodr´ıguez-Ardila A., Bonatto C.,2009b, MNRAS, 400, 273Riffel R. A., Storchi-Bergmann T., McGregor P. J., 2009c, ApJ,698, 1767Riffel R. A., Storchi-Bergmann T., Nagar N. M., 2010a, MNRAS,404, 166Riffel R. A., Storchi-Bergmann T., Riffel R., Pastoriza M. G.,2010b, ApJ, 713, 469Riffel R., Riffel R. A., Ferrari F., Storchi-Bergmann T., 2011,MNRAS, 416, 493Riffel R. A., Storchi-Bergmann T., Winge C., 2013, MNRAS,430, 2249Riffel R. A., Vale T. B., Storchi-Bergmann T., McGregor P. J.,2014a, MNRAS, 442, 656Riffel R. A., Storchi-Bergmann T., Riffel R., 2014b, ApJ, 780, L24Riffel R. A., Storchi-Bergmann T., Riffel R., 2015, MNRAS,451, 3587Riffel R. A., et al., 2016, MNRAS, 461, 4192Riffel R. A., Storchi-Bergmann T., Riffel R., Dahmer-Hahn L. G.,Diniz M. R., Sch¨onell A. J., Dametto N. Z., 2017, MNRAS,470, 992Riffel R. A., et al., 2018, MNRAS, 474, 1373Roche P. F., Aitken D. K., Smith C. H., Ward M. J., 1991,MNRAS, 248, 606Roche P. F., Packham C., Aitken D. K., Mason R. E., 2007,MNRAS, 375, 99Rodr´ıguez-Ardila A., Contini M., Viegas S. M., 2005, MNRAS,357, 220Schinnerer E., Eckart A., Tacconi L. J., 2000, ApJ, 533, 826Schmidt M., 1959, ApJ, 129, 243Schmitt H. R., Kinney A. L., 2000, ApJS, 128, 479Sch¨onell A. J., Riffel R. A., Storchi-Bergmann T., Winge C., 2014,MNRAS, 445, 414Sch¨onell Jr. A. J., Storchi-Bergmann T., Riffel R. A., Riffel R.,2017, MNRAS, 464, 1771Scoville N. Z., Hall D. N. B., Ridgway S. T., Kleinmann S. G.,1982, ApJ, 253, 136Shi L., Gu Q. S., Peng Z. X., 2006, A&A, 450, 15Siebenmorgen R., Kr¨ugel E., Spoon H. W. W., 2004, A&A,414, 123Somerville R. S., Hopkins P. F., Cox T. J., Robertson B. E.,Hernquist L., 2008, MNRAS, 391, 481Springel V., Di Matteo T., Hernquist L., 2005, MNRAS, 361, 776Storchi-Bergmann T., Schnorr-M¨uller A., 2019,Nature Astronomy, 3, 48Storchi-Bergmann T., McGregor P. J., Riffel R. A., Sim˜oes LopesR., Beck T., Dopita M., 2009, MNRAS, 394, 1148Storchi-Bergmann T., Lopes R. D. S., McGregor P. J., RiffelR. A., Beck T., Martini P., 2010, MNRAS, 402, 819Storchi-Bergmann T., Riffel R. A., Riffel R., Diniz M. R., BorgesVale T., McGregor P. J., 2012, ApJ, 755, 87Terrazas B. A., Bell E. F., Henriques B. M. B., White S. D. M.,Cattaneo A., Woo J., 2016, ApJ, 830, L12MNRAS , 1–19 (2019) eeding and Feedback in AGN Tsai M., Hwang C.-Y., 2015, AJ, 150, 43Ulrich M.-H., Pequignot D., 1980, ApJ, 238, 45Ulvestad J. S., Wilson A. S., 1989, ApJ, 343, 659Veilleux S., Tully R. B., Bland-Hawthorn J., 1993, AJ, 105, 1318Voit G. M., Shull J. M., Begelman M. C., 1987, ApJ, 316, 573Wagner S. J., 1988, PASP, 100, 54Westfall K. B., Bershady M. A., VerheijenM. A. W., Andersen D. R., Swaters R. A., 2007,Astrophysics and Space Science Proceedings, 3, 157Zaw I., Farrar G. R., Greene J. E., 2009, ApJ, 696, 1218de Vaucouleurs G., de Vaucouleurs A., Corwin Jr. H. G., ButaR. J., Paturel G., Fouqu´e P., 1991, Third Reference Catalogueof Bright Galaxies. Volume I: Explanations and references.Volume II: Data for galaxies between 0 h and 12 h . VolumeIII: Data for galaxies between 12 h and 24 h .This paper has been typeset from a TEX/L A TEX file prepared bythe author.MNRAS000