Diagnostics of ionized gas in galaxies with the "BPT--radial velocity dispersion" relation
aa r X i v : . [ a s t r o - ph . GA ] A ug Astrophysical Bulletin, 2018, vol. 73, No.3, p. 267
August 28, 2018Translated from Astrofizicheskij Byulleten, 2018, vol.73, No.3, p. 315
Diagnostics of ionized gas in galaxies with the “BPT–radialvelocity dispersion” relation
D.V. Oparin ∗ and A.V. Moiseev ∗∗∗ Special Astrophysical Observatory, Russian Academy of Sciences, Nizhnij Arkhyz, 369167 Russia S2Space Research Institute, Moscow, 117997 RussiaMarch 30, 2018/Revised: June 28, 2018
Abstract.
In order to study the state of gas in galaxies, diagrams of the relation of optical emission line fluxesare used allowing one to separate main ionization sources: young stars in the H II regions, active galactic nuclei,and shock waves. In the intermediate cases, when the contributions of radiation from OB stars and from shockwaves mix, identification becomes uncertain, and the issue remains unresolved on what determines the observedstate of the diffuse ionized gas (DIG) including the one on large distances from the galactic plane. Adding ofan extra parameter — the gas line-of-sight velocity dispersion — to classical diagnostic diagrams helps to find asolution. In the present paper, we analyze the observed data for several nearby galaxies: for UGC 10043 with thegalactic wind, for the star forming dwarf galaxies VII Zw 403 and Mrk 35, for the galaxy Arp 212 with a polar ring.The data on the velocity dispersion are obtained at the 6-m SAO RAS telescope with the Fabry-Perot scanninginterferometer, the information on the relation of main emission-line fluxes — from the published results of theintegral-field spectroscopy (the CALIFA survey and the MPFS spectrograph). A positive correlation between theradial velocity dispersion and the contribution of shock excitation to gas ionization are observed. In particular, instudying Arp 212, “BPT– σ relation” allowed us to confirm the assumption on a direct collision of gaseous cloudson the inclined orbits with the main disk of the galaxy. Key words. galaxies: interstellar medium—galaxies: kinematics and dynamics—galaxies: star formation
1. Introduction
Diagrams of ratios of optical emission-line fluxesare widely used for diagnostics of gas-ionizationsources in galaxies. In the classic work by Baldwin,Phillips, & Terlevich (1981) the two-dimensiondiagram of line fluxes of [O III] λ β and[N II] λ α was suggested for separation of ob-jects with different ionization sources. The methodbecame popular due to the use of measurements of linesbright in the visible range that are close in wavelengthsand, consequently, with a weak dependence of theirintensity ratio on the interstellar extinction. Later, thismethod was extended by adding the relations [S II]/H α and [O I]/H α (Veilleux & Osterbrock, 1987; Kewleyet al., 2001) as the second parameter. All the mentioneddiagrams are frequently called in the literature the “BPTdiagrams” after the authors of the method. Using them,it is possible to distinguish the regions, where the largest Send offprint requests to : Dmitry Oparin e-mail: [email protected] Hereinafter, for short we will designate [O I] λ λ λ λ λ contribution to gas ionization is made by young massivestars (hereinafter, the H II type) and the regions ofdominating hard ionizing radiation of the active galacticnucleus (AGN). At the same time, the regions ionizedby shock waves, the asymptotic giant branch (AGB)stars or nuclei of galaxies of the LINER type mix in thediagrams . Various variants of the demarcation lines weresuggested (Monreal-Ibero et al., 2010; Ho et al., 2014),but it is often problematic to separate the contribution ofionizing sources with the soft spectrum.Addition of one more parameter – the velocity disper-sion of the ionized gas along the line of sight ( σ ) – tothe classic diagnostic diagrams allows one to escape un-certainty in the cases, when the increase of σ indicates theincrease of turbulent velocities of gas beyond the front ofa shock wave. However, to estimate σ reliably, the spec-tral resolution is necessary that is noticeably better thanthat usually required for measuring the fluxes of radialvelocities of separate spectral lines. Thus, until recently,the dependence of the relation of line fluxes characteriz- Low-Ionization Narrow Emission-line Region in which theshock ionization of gas can be associated both with a burst ofstar formation and with a weak nuclear activity. Oparin & Moiseev: Diagnostics of ionized gas in galaxies ing the shock ionization from σ was rarely considered andmainly for the objects with σ > − such asgalaxies with intense star formation (Monreal-Ibero et al.,2010; Ho et al., 2014). Such an approach has not previ-ously been used to study the ionization of diffuse gas indwarf galaxies, around separate star-forming regions, orat some distance from the plane of the galactic disk.There is a discussion about the sources of ionizationof this diffuse ionized gas (DIG) in galaxies whose roleis assigned to an old stellar population, leakage of Lymanquanta from H II regions, and also possibly to shock frontscaused by star formation processes (see references anddiscussion in Jones et al., 2017; Egorov et al., 2017).The most effective methods for studying the extendedlow-brightness structures in galaxies are panoramic spec-troscopy also called integral-field, or 3D. In a recent pa-per based on the results of the SDSS MaNGA survey by(Zhang et al., 2017), it was concluded that DIG is asso-ciated mainly with the evolved stellar population (AGBstars, etc.). At the same time, in Section 6.2 of the papercited, it was noted that shock waves can be the cause ofthe observed increase in the flux ratio of the forbidden andBalmer lines. It is difficult to verify, since the spectral res-olution of the MaNGA survey is about two times poorerthan that required to be able see the effects of moderateshock waves (with a velocity of less than 500 km s − ) inthe observed kinematics of the ionized gas. Unfortunately,most of the available observed data on spectrophotometryand kinematics of the gas of nearby galaxies are obtainedwith the spectral resolution F W HM > − in terms ofradial velocity dispersion or greater than 230 km s − interms of the F W HM in the H α line. Observations withsuch resolution are a compulsory compromise in the studyof low-surface-brightness objects.In the SAMI survey of galaxies (Ho et al., 2014)with the 3D spectroscopy at the 3.9-m Anglo-AustralianTelescope (AAT), the “line ratio–velocity dispersion” di-agrams were built for the galaxies with active star for-mation. A positive correlation of the ionized gas σ witha characteristic emission lines ratios was noticed, whichwas interpreted as an increase of shock waves contribu-tion with velocities of about 200-300 km s − accompa-nying a burst of star formation. The spectral resolutionof the SAMI survey is greater than that of MaNGA andequals R ≈ ′′ in SAMI (Croom et al., 2012) and 12 ′′ –32 ′′ in SDSS MaNGA (Bundy et al., 2015). In these sur-veys, relatively distant ( z > .
01) galaxies are studied. Atthe same time, the largest contribution to the kinemat-ics of interstellar medium from motion due to supernovaeand winds of young stars in star-forming regions is madeon considerably smaller spatial scales (from tens to hun-dreds of parsec). Consequently, any observed manifesta- tions of shock fronts in star-forming regions become unev-ident, when averaging over a scale of one kpc or more. Theexamples of decreasing of peak velocity dispersion of theionized gas in dwarf galaxies with degradation of a spatialresolution are presented in Moiseev & Lozinskaya (2012).Vasiliev et al. (2015) considered the same effect in sim-ulations of multiple supernova explosions. Therefore, forobservational studies of the relation between an ionizationstate of gas and dispersion of its radial velocities in galax-ies without an active nucleus and with a moderate starformation rate, 3D spectroscopic data are required simul-taneously with a considerably high spectral and spatialresolution.In this paper, we consider this relation for severalnearby galaxies using a combination of two spectroscopicmethods with similar spatial resolution and quite a largefield of view. Velocity dispersion map are derived from theobservations with a scanning Fabry-Perot interferometer(FPI) at the 6-m SAO RAS telescope. Information on themain emission lines ratios are taken from open data on theintegral-field spectroscopy with low spectral resolution.In order to show the relation between the velocity dis-persion and the lines ratios characterizing the ionizationstate, we use various methods through our paper: color-ing in BPT diagrams, the “ σ –line-flux relation” diagrams,and “ σ –distance from the H II/AGN demarcation line.” As a general name for the dependencies under study, weuse the term “BPT relation– σ ” . Classical BPT–diagramsare two-dimensional plots, where the axes represent rela-tions of line fluxes. The inclusion of the velocity dispersionin the analysis is equivalent to the transition to three-dimensional plots, where a coordinate axis σ is added toeach diagram. The more familiar two-dimensional plotsgiven in our paper and in the above papers are a projec-tion of the BPT– σ common relation to the selected plane.
2. Spectral data and data reduction
We considered the sample of of galaxies with data on a ion-ized gas state obtained with two 3D spectroscopy meth-ods. When compiling the sample, we prepared a list ofnearby galaxies for which, based on observations with thescanning FPI at the 6-m telescope of SAO RAS, the fieldsof velocity dispersion of ionized gas in the H α or [N II]emission lines were constructed. In total, this is about 60objects that were observed in 2002–2015; most of themare presented in Moiseev et al. (2015). For each galaxy inthe list, we checked the presence of data cubes in opensources obtained with the integral-field spectroscopy. Forthree galaxies: Arp 212, Mrk 35, and UGC 10043 such datawere obtained within the framework of the CALIFA sur-vey (The Calar-Alto Legacy Integral Field Area, S´anchezet al., 2011). We used the third data release of CALIFA;the spectra are available on the project website . Let us http://califa.caha.es/ parin & Moiseev: Diagnostics of ionized gas in galaxies 3 Table 1.
Characteristics of the galaxies under study and parameters of their observations with different methodsGalaxy D , M B Integral-field spectroscopy Scanning FPIMpc Instrument ∆ λ , ˚A δλ , ˚A θ ′′ ∆ λ δλ , ˚A θ ′′ UGC 10043 34.9 − . − .
87 MPFS 4250–7200 8 2.0 H α − .
75 PPAK 3750–7500 5–9 2.7 H α − . α notice that UGC 10043 was observed with the 6-m tele-scope specially on request of the CALIFA team. The re-sults were presented in our collaborative paper (L´opez-Cob´a et al., 2017) and triggered our interest for furtherstudy of the BPT– σ relation. Nevertheless, to keep homo-geneity, we repeated the analysis of the UGC 10043 datain this paper using the same methods as for other galaxies.For the galaxy VII Zw 403, there was a data cube obtainedby combining several fields of the MPFS spectrograph inobservations at the 6-m telescope of SAO RAS and pub-lished by Arkhipova et al. (2007).Table 1 briefly presents the objects under study (ac-cepted distance D and absolute magnitude M B based onthe NED data) and on the data used (observation instru-ments, ∆ λ –spectral range or a selected line, δλ –spectralresolution in terms of F W HM , θ ′′ –angular resolution).Figure 1 shows the images of the sample galaxies in the r filter and in the emission lines, and velocity dispersionfields σ of the ionized gas constructed from the scanningFPI data. It also shows the field of view of the spectro-graphs used. For UGC 10043, Mrk 35, and Arp 212, we used theCALIFA data obtained at the 3.5-m telescope of the CalarAlto observatory in the mode of the integral-field spec-troscopy of the PPAK wide field (Kelz et al., 2006) ofthe PMAS spectrograph (Roth et al., 2005). The arrayof PPAK optical fibers comprises 331 elements of the 2 . ′′ ′′ × ′′ hexagonal field. Weused the cubes obtained in the low-resolution mode cover-ing the entire visible range (grating V500, R ∼ ′′ .The galaxy VII Zw 403 was observed with the MPFSmultislit spectrograph (Afanasiev et al., 2001) at the SAORAS 6-m telescope. An array of square lenses combinedwith fiber optics provided a field of view of 16 ×
16 el-ements with a scale of 1 ′′ per lens. The data cube pre-sented in Arkhipova et al. (2007) is a mosaic of the sizeof 49 ′′ × ′′ comprising seven MPFS fields. The spectralrange was 4250–7200 ˚A; the resolution was 8 ˚A.For the analysis, we used the data cubes with the 2 × ′′ per element (see next Section 2.3).Approximation of the lines in the spectra was carriedout by the one-component Gaussian function. The line fluxes ratios were measured only from the spectra in which S/N >
The archival observations with the scanning Fabry-Perotinterferometer installed at the SCORPIO (Afanasiev &Moiseev, 2005) and SCORPIO-2 (Afanasiev & Moiseev,2011) focal reducers in the primary focus of the 6-mtelescope were used to create the velocity dispersionmaps. The emission line (H α or [N II] λ ∼ F W HM = 1 . F W HM = 0 . . ′′
56 px − with the field of view of 4 . ′ . ′′ − and 6 . ′
1, respectively. The re-sult of the reduction of the set of interferograms was adata cube, where each pixel contained the spectrum ofthe selected emission line consisting of 36–40 channels.The details of data reduction and observational logs werepublished earlier (see references in Section 3). Since the in-strumental contour of the interferometer is well describedwith the Lorentz profile, the observed profiles of emis-sion lines were approximated by the Voigt function — theconvolution of the Gauss and Lorentz functions (Moiseev& Egorov, 2008). It is assumed that the initial (withoutany instrumental broadening) profile of an emission line issatisfactorily described by a Gaussian, which is a good ap-proximation for observations of H II regions, with the ex-ception of individual peculiar cases (expanding envelopes,neighborhood of Wolf-Rayet stars, etc. (see Moiseev &Lozinskaya, 2012; Egorov et al., 2017)). Based on the re-sults of approximation, we built the monochromatic im-ages in this line, the distribution of radial velocities ofthe ionized gas, and the radial velocity dispersion mapsfree from the instrumental broadening of the spectral lineprofile (Moiseev & Egorov, 2008).The σ maps in Fig. 1 are shown with the originalsampling of the FPI images (0 . ′′ . ′′
7) which is betterthan in the data used in the integral-field spectroscopy(1 ′′ /spaxel), while the angular resolution θ of both datasets is similar (see Table 1). To account for this effect,first we interpolated the FPI cube to a coarser grid cor-responding to the CALIFA or MPFS data. The accuracyof coincidence of both data sets was controlled from the Oparin & Moiseev: Diagnostics of ionized gas in galaxies
Fig. 1.
The images of the studied galaxies. The left-hand column shows the SDSS r -filter images from Knapen et al. (2014), forUGC 6456 the image in the R filter is given from Gil de Paz et al. (2003). The middle column gives the images in the [N II] λ α line (for others) from the observed data with the FPI at the 6-m BTA telescope. The right-handcolumn shows the velocity dispersion of the ionized gas; the scale in km s − . Image size is 90 ′′ × ′′ . The arrangement of thefields of view of integral-field spectrographs is shown: the CALIFA survey and the MPFS mosaic (for UGC 6456). The velocitydispersion fields are taken from Moiseev et al. (2015) and L´opez-Cob´a et al. (2017) (for UGC 10043) without correction for thethermal line broadening.parin & Moiseev: Diagnostics of ionized gas in galaxies 5 images in the emission lines and continuum and was bet-ter than 0 . ′′ . ′′
5. In order to detect the emission linesin the low-surface-brightness regions in the optimum way,we performed the pixel binning of 2 × ′′ element size. After combining andbinning in the FPI cubes, we built the maps σ . In the nextsection, these maps were used for direct per-pixel compar-ison with the low-resolution spectra.We used masking to highlight the points with the ra-tio S/N ≥ σ . Let us note that, as distinctfrom Moiseev et al. (2015), we did not correct the velocitydispersion maps for thermal line broadening.
3. BPT– σ diagrams UGC 10043 is an edge-on spiral galaxy. Observationsin the H α and [N II] lines carried out with theHST (Matthews & de Grijs, 2004) have shown signs ofstar formation in the galactic core, as well as an extendedemission structure that is perpendicular to the disk and isthe result of the galactic wind influence. L´opez-Cob´a et al.(2017) presented diagnostic diagrams for the central partof the galaxy according to the CALIFA survey. Some ofthe points on the diagram relating to the central region ofstar formation turned out to be located in the region char-acteristic of photoionization by young stars, while othersfell into the region typical of shock excitation. Within theframework of a shock excitation model, a wind velocitywas constrained: no greater than 400 km s − . Analyzingthe gas velocity field in the [N II] line constructed with ascanning FPI allowed us to obtain a more strict limitationon the galactic wind velocity: less than 250 km s − in ac-cordance with the gas shock excitation model. In the samepaper L´opez-Cob´a et al. (2017), it has been shown thatthere is a distinct relation BPT– σ in the wind nebula ofUGC 10043.As it is shown in the diagnostic diagrams presented inFig. 2 (the upper row), the regions with shock excitationof emission lines in the wind nebula are characterized by ahigher velocity dispersion as compared to the regions dom-inated by photoionization. In this case, there is a positivecorrelation between the relations of line fluxes of [S II] toH α , of [N II] to H α , of [O I] to H α and σ (see Fig. 2). Atthe same time, negative correlation is observed betweenthe ratio of the sulfur doublet lines ([S II]6731/[S II]6717)and σ . This means that a higher velocity dispersion ischaracteristic of the diffuse gas with a lower electron den-sity n e .We tried to quantify the BPT– σ relation. For eachpoint in the diagrams of the lines ratios, it is possible todetermine the minimum distance to the curve that boundsthe H II-type ionization region from Kewley et al. (2001) (in the case of the [N II]/H α –[O III]/H β diagram, this isthe boundary between the Comp and AGN regions in ourfigures). We marked this distance as ρ and determined itso that negative values of ρ corresponded to the shift ofthe points from the demarcation line to the side corre-sponding to photoionization by young stars, and positive–towards other ionization mechanisms. Figure 3 shows theexamples of relations involved this parameter. For brevityand convenience of reading, we have designated the value ρ for the [N II]/H α –[O III]/H β diagrams as ρ ([N II]), forthe [S II]/H α –[O III]/H β diagrams as ρ ([S II]), and for the[O I]/H α –[O III]/H β diagrams as ρ ([O I]). It can be seenthat in all the cases presented, the increase in the velocitydispersion along the line of sight correlates with the dis-tance from the region characteristic of ionization by youngstars in the BPT diagram. VII Zw 403 is one of the nearest blue compact dwarf galax-ies with several episodes of recent star formation. The cur-rent outburst is located in the central kpc, where severalcompact OB-stars associations are identified and the as-sociated H II shells that are immersed in the diffuse ion-ized gas (see Egorov & Lozinskaya, 2011, and referencestherein). The fields of velocities and velocity dispersions ofthe ionized gas in this galaxy were previously consideredin Lozinskaya et al. (2006); Moiseev & Lozinskaya (2012);Moiseev et al. (2015), where a sufficiently quiet kinemat-ics of the gas with a low level of peculiar velocities wasnoticed. The value σ is in the range of 15–40 km s − .In the BPT diagrams, most points are located in the re-gion of photoionization (see Fig. 4). A certain number ofpoints with higher dispersion are found near the separa-tion curve. Along with this, the expanding H II shells as-sociated with bright star formation regions are character-ized by a smaller value of σ ∼
20 km s − . One of the re-gions with a higher σ is located between these two shells.Others are located on the periphery of the ionized gasdisk (Moiseev & Lozinskaya, 2012). In the “lines ratio–velocity dispersion” diagrams, there are no noticeably sig-nificant correlations (Fig. 4). Therefore, we can concludethat the contribution of shock excitation to gas ionizationin this galaxy is negligible and even at the boundaries ofthe expanding shells it is noticeably inferior to photoion-ization (the H II type). This is also indicated by the ab-sence of significant correlation between ρ and σ in Fig. 5.Mrk 35 is another example of a blue compact galaxy.The ongoing star formation here is concentrated in severalbright compact regions. Star-forming regions near the op-tical center of the galaxy form a bar-like structure, wherethe population of Wolf–Rayet stars is observed (Cair´oset al., 2007). The radial velocity dispersion of the ionizedgas in the galaxy reaches about 70 km s − , whereas in thecentral regions it lies within the range of 20–35 km s − .The highest dispersion of radial velocities is observed in Oparin & Moiseev: Diagnostics of ionized gas in galaxies
Fig. 2.
Upper row: BPT diagrams for UGC 10043. The ionized gas line-of-sight velocity dispersion in the given pixel accordingto the map shown bottom right is colored. The lines separating the H II regions, the objects with the combined ionization typeactive Seyfert galaxies, and LINER are taken from Kewley et al. (2006). The other diagrams: the dependence between thevelocity dispersion and the emission lines ratios.
Fig. 3.
UGC 10043. Dependence of σ on the distance of the point to the demarcation curve in the BPT diagram separating theH II regions and regions with other ionization mechanisms (accroding Kewley et al., 2001).parin & Moiseev: Diagnostics of ionized gas in galaxies 7 Fig. 4.
Same as in Fig. 2, for VII Zw 403.
Fig. 5.
Diagrams similar to those in Fig. 3, for VII Zw 403. the gas located between three central regions of star for-mation. In the “arms”, the dispersion is several timeslower in comparison with the central regions; and as awhole does not exceed 20 km s − . In the BPT diagrams(see Fig. 6), the points corresponding to the regions with the ongoing star formation are located in the region ofphotoionization. The outer parts of the galaxy, charac-terized by low surface brightness and high dispersion ofradial velocities, appear near the separation curves whichsuggests a certain contribution of shock waves to the gas Oparin & Moiseev: Diagnostics of ionized gas in galaxies
Fig. 6.
Same as in Fig. 2, for Mrk 35.
Fig. 7.
Diagrams similar to those Fig. 3, for Mrk 35. ionization in these regions. As well as in UGC 10043, thesulfur lines ratio in Fig. 6 demonstrates the anticorrela-tion. The σ – ρ diagrams show a positive correlation be-tween the distance to the model curve and the velocitydispersion (Fig. 7). Arp 212 is a peculiar galaxy in which two rotating gassubsystems that are kinematically different have been dis-covered: an internal disk of the 3.5-kpc size and outer H II parin & Moiseev: Diagnostics of ionized gas in galaxies 9
Fig. 8.
Same as in Fig. 2, for Arp 212.
Fig. 9.
Diagrams similar to those Fig. 3, for Arp 212. regions whose orbits are inclined at a significant angle tothe stellar disk (Moiseev, 2008). The observed picture wasexplained in the assumption that the gas (mostly neutral)in the outer regions of the galaxy is located in a wide ringof a diameter of about 20 kpc rotating in the plane almost orthogonal to the disk. As the radii of the gaseous-cloudorbits decrease, their inclination angle decreases too; andat a radius of 2–3 kpc, the gas from the ring begins to fallout onto the plane of the galaxy inducing a burst of starformation. It is the region with the highest observed ve- locity dispersion reaching 80–100 km s − (see Fig. 8). Thepoints belonging to this collision region of the gas subsys-tems are shifted in the BPT diagrams (Fig. 8) from theregions dominated by photoionization towards the domi-nance of shock ionization. At the same time, photoioniza-tion clearly dominates in the central region of the galaxy.As well as in UGC 10043 and Mrk 35, there is a posi-tive correlation between σ and ρ for all the BPT diagrams(Fig. 9). It is important to notice that in all three galaxiesthis dependence can be observed for the velocity disper-sion above 30–40 km s − and practically disappears forsmaller σ . In other words, the correlation between σ and ρ manifests itself in the presence of shock excitation inthe diffuse gas (DIG) and disappears in the H II regionscharacterized by a low level of turbulent motions. Thiscan also be confirmed by the absence of distinct σ – ρ cor-relations in the galaxy VII Zw 403, where in all the points σ <
40 km s − .As distinct from UGC 10043 and Mrk 35, the relationof sulfur lines ratio in Arp 212 does not show any pro-nounced dependence on σ which agrees with the assump-tion that high velocity dispersion is observed not only inDIG with low electron density but also in a denser mediumof colliding gaseous clouds.
4. Conclusion
For an observational study of the relation between tur-bulent motions of the ionized gas in nearby galaxies andthe state of its ionization, it is required to have panoramicspectroscopy data together with a large field of view andquite high spectral resolution. Since it is necessary to ob-serve the low-surface-brightness region with an angularresolution of about 1 ′′ , then an optical telescope of a large( D > σ diagram together with theclassical diagnostic methods based on lines ratios helps usbetter understand of ionization of the galactic interstellarmedium in each specific case. The only galaxy in whichwe did not find a correlation between σ and characteristicline flux relations (or ρ parameter) is VII Zw 403. The on-going star-formation rate here is the lowest in our sample( ∼ . M ⊙ yr − , Lozinskaya et al., 2006). Apparently,for this reason, the contribution of shock waves to gasionization is practically invisible.We plan to conduct further expansion of the sampleof the objects under study in two ways. The first is newobservations with a high spectral resolution of galaxies,for which there are the CALIFA survey data already, withthe scanning FPI. The second is the creation of images inthe emission lines of galaxies, for which we already havemaps of the velocity dispersion of the ionized gas. Here itis proposed to use a tunable–filter photometer, the firstobservations with which are already being conducted byour team . Acknowledgements.
The work was supported by the RussianScience Foundation (project No. 17-12-01335 “Ionized gas ingalactic disks and beyond the optical radius.” The paper usedthe survey data provided by the Calar Alto Legacy IntegralField Area (CALIFA) survey ( http://califa.caha.es/ )based on the observations collected at the Centro Astron´omicoHispano Alem´an (CAHA) at Calar Alto operated jointly withthe Max-Planck-Institut f¨ur Astronomie and the Instituto deAstrof´ısica de Andaluc´ıa (CSIC). This research has made use ofthe NASA/IPAC Extragalactic Database (NED) which is oper-ated by the Jet Propulsion Laboratory, California Institute ofTechnology, under contract with the National Aeronautics andSpace Administration. The authors are grateful to AlexandrinaSmirnova and the reviewer for constructive comments.
References
Afanasiev, V. L., Dodonov, S. N., & Moiseev, A. V. 2001,in Stellar Dynamics: from Classic to Modern, ed. L. P.Ossipkov & I. I. Nikiforov, 103Afanasiev, V. L. & Moiseev, A. V. 2005, AstronomyLetters, 31, 194Afanasiev, V. L. & Moiseev, A. V. 2011, Baltic Astronomy,20, 363Arkhipova, V. P., Lozinskaya, T. A., Moiseev, A. V., &Egorov, O. V. 2007, Astronomy Reports, 51, 871Bacon, R., Accardo, M., Adjali, L., et al. 2010, inProceedings of the SPIE, Vol. 7735, Ground-based andAirborne Instrumentation for Astronomy III, 773508 parin & Moiseev: Diagnostics of ionized gas in galaxies 11parin & Moiseev: Diagnostics of ionized gas in galaxies 11