Systematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties
Carlos Lopez Coba, Sebastian F. Sanchez, Joss Bland-Hawthorn, Alexei V. Moiseev, Irene Cruz-Gonzalez, Ruben Garcia-Benito, Jorge K. Barrera-Ballesteros, Lluis Galbany
MMNRAS , 1–16 (2018) Preprint 8 November 2018 Compiled using MNRAS L A TEX style file v3.0
Systematic study of outflows in the Local Universe using CALIFA: I.Sample selection and main properties.
Carlos L ´opez-Cob´a (cid:63) , Sebasti´an F. S´anchez , Joss Bland-Hawthorn , ,Alexei V. Moiseev , , , Irene Cruz-Gonz´alez , Rub´en Garc´ıa-Benito ,Jorge K. Barrera-Ballesteros , Llu´ıs Galbany Instituto de Astronom´ıa, Universidad Nacional Aut´onoma de M´exico, Circuito Exterior, Ciudad Universitaria, Ciudad de M´exico 04510, Mexico Sydney Institute for Astronomy, School of Physics, University of Sydney, NSW 2006, Australia Centre of Excellence for All Sky Astrophysics in 3D (ASTRO-3D), Australia Special Astrophysical Observatory, Russian Academy of Sciences, Nizhny Arkhyz 369167, Russia Lomonosov Moscow State University, Sternberg Astronomical Institute, Universitetsky pr. 13, Moscow 119234, Russia Space Research Institute, Russian Academy of Sciences, Profsoyuznaya ul. 84 /
32, Moscow 117997, Russia Instituto de Astrof´sica de Andaluc´ıa (IAA / CSIC), Glorieta de la Astronom´ıa s / n Apdo. 3004, 18080, Granada, Spain Department of Physics & Astronomy, John Hopkins University, Bloomberg Center, 3400 N. Charles St., Baltimore, MD 21218, USA PITT PACC, Department of Physics and Astronomy, University of Pittsburgh, Pittsburgh, PA 15260, USA
Accepted XXX. Received YYY; in original form ZZZ
ABSTRACT
We present a sample of 17 objects from the CALIFA survey where we find initial evidenceof galactic winds based on their o ff -axis ionization properties. We identify the presence of out-flows using various optical diagnostic diagrams (e.g., EW(H α ), [N ii ] / H α , [S ii ] / H α , [O i ] / H α line-ratio maps). We find that all 17 candidate outflow galaxies lie along the sequence of activestar formation in the M (cid:63) vs. star-formation rate diagram, without a clear excess in the inte-grated SFR. The location of galaxies along the star-formation main sequence (SFMS) doesnot influence strongly the presence or not of outflows. The analysis of the star-formation ratedensity ( Σ SFR ) reveals that the CALIFA sources present higher values when compared withnormal star-forming galaxies. The strength of this relation depends on the calibrator used toestimate the SFR. This excess in Σ SFR is significant within the first e ff ective radius supportingthe idea that most outflows are driven by processes in the inner regions of a galaxy. We findthat the molecular gas mass density ( Σ gas ) is a key parameter that plays an important role in thegeneration of outflows through its association with the local SFR. The canonical threshold re-ported for the generation of outflows – Σ SFR > . (cid:12) yr − kpc − – is only marginally exceededin our sample. Within the Kennicutt-Schmidt diagram we propose a domain for galaxies host-ing starburst-driven outflows defined by Σ SFR > − M (cid:12) yr − kpc − and Σ gas > . M (cid:12) pc − within a central kiloparcec region. Key words: galaxies: ISM — galaxies: star formation — galaxies: structure — ISM: jets andoutflows
Galactic outflows have been invoked in many astrophysical prob-lems to explain some local and global properties of galaxies likethe tight correlation in the stellar mass and metallicity (Tremontiet al. 2004; Heckman 2002), the metal enrichment of intergalac-tic medium (Pettini et al. 1998; Veilleux et al. 2005), and in thecurrent models of galaxy formation, where the amount of feedbackfrom outflows is a key ingredient which is not well constrained (e.g, (cid:63) [email protected]
Aguirre et al. 2001; Springel & Hernquist 2003; Scannapieco et al.2006). Even their global e ff ect, either in preventing or triggeringstar-formation is under discussion (e.g., Silk 2013).Outflows are driven either by supernovae explosions (SN),stellar winds or by active galactic nuclei (AGN), or some combi-nation of these − we refer to such objects collectively as activegalaxies. Nuclear star formation is found to occur in the nuclear re-gions of most Seyfert galaxies (Esquej et al. 2014) and so both mayact in concert to generate outflows, although just how this worksis mysterious (Hopkins & Quataert 2010). The scale of these out-flows depends partly on the escape velocity via the gravitational po- c (cid:13) a r X i v : . [ a s t r o - ph . GA ] N ov C. L ´opez-Cob ´a et al. tential well (Tanner et al. 2017). A high fraction of active galaxieswith lower total mass are expected to host outflows because of theirlower escape velocity (Martin 1998; Bland-Hawthorn et al. 2015).This favours the loss of large fractions of gas and metals in thesegalaxies (e.g., Barrera-Ballesteros et al. 2018). Massive galaxies re-tain their baryons more e ff ectively although considerable recyclingthroughout the halo can take place (Cooper et al. 2008; Tanner et al.2016).Although there has been extensive e ff ort on the theory of out-flows, the observational counterparts are far from being understood,mainly because their multiphase nature makes them hard to detectand interpret. Numerical simulations today are far from capturingthe full complexity of galactic winds too (e.g., Martel 2011). Out-flows have been detected at high redshift (Coil et al. 2011; Genzelet al. 2014), in the nearby universe (e.g., Franx et al. 1997; Fogartyet al. 2012; Ho et al. 2014, 2016), even in the Local Group (Bland-Hawthorn & Cohen 2003; Su et al. 2010; Fox et al. 2015). Galacticwinds have also been detected across most galaxy types (e.g., Axon& Taylor 1978; Bland & Tully 1988; Heckman et al. 1990; Lehnert& Heckman 1996; Martin 1998; Rupke et al. 2005a). Despite ofall these studies, the nature, properties and influence of outflows ingalaxy evolution are still unclear. How, why and where outflows areproduced in a galaxy, as well as the loss rates of mass, metals andenergy that they produce are still open questions that have not beencompletely understood (see Veilleux et al. 2005, for an extensivereview). High quality spatially resolved spectroscopic data couldbring some light in the understanding of these processes.A primary problem when studying outflows is the detectionitself. Outflows are commonly studied in starburst galaxies or inultraluminous infrared galaxies (ULIRG) due to their large star for-mation rates (SFR), making them more prone to develop outflows.This means that studies of outflows are biased towards galaxieswith high star formation rates. Other studies analyze outflows di-rectly associated with strong AGNs, in particular with those onesdirectly pointing towards the observer (e.g., BL Lacs or Blazars,Antonucci & Ulvestad 1985; Scarpa et al. 2000; Celotti & Ghis-ellini 2008), being biased towards these particular kind of objects.Thus, there are few systematic studies of the presence of outflows inan unbiased population of galaxies (cf., Sharp & Bland-Hawthorn2010; Ho et al. 2016).Early studies of galaxies with bona fide outflows have con-structed the basis in the methodology to detect and characterizethem (e.g., Heckman et al. 1990; Lehnert & Heckman 1996; Rupkeet al. 2005a,b). This methodology is based on the study of the spa-tial distribution of certain emission line ratios over the extra-planarregions of disk galaxies, and their comparison with certain kine-matic properties. Although these studies provide moderate samplesof outflows, they do not provide well defined statistics about thefrequency of galaxies hosting outflows and their properties in com-parison with those not hosting them. In the present study, we ad-dress the search and characterization of the statistical propertiesof galaxies hosting outflows. For this purpose, we exploit the CAL-IFA integral field spectroscopic survey (IFS) which achieved a largesample (835) of galaxies observed from the Calar Alto telescope inSpain.Di ff erent optical IFS surveys (e.g., SAMI, MaNGA, AMUS-ING, Allen et al. 2015; Bundy et al. 2015; Galbany et al. 2016),have already taken advantage of this technique to study the spa-tially resolved properties of the warm ionized gas component ofoutflows (e.g., Sharp & Bland-Hawthorn 2010; Rich et al. 2011;Fogarty et al. 2012; Ho et al. 2014; Wild et al. 2014; Rich et al.2015; Ho et al. 2016; Prieto et al. 2016; L´opez-Cob´a et al. 2017; Maiolino et al. 2017). The use of larger samples allows us to per-form statistical analysis not only on the outflowing galaxies, but onthose that do not present outflows, i.e., a properly selected controlsample typically overlooked in resolved galaxy surveys.The layout of this article is as follows: In Sec. 2, we present thedata and physical properties (data products) extracted from themused along this article; describing the analysis of the stellar popula-tion in Section 2.1 and of the ionized gas in 2.2. The outflow sampleanalysis of these data products is presented in Section 3. It includesthe selection of candidates that host outflows in Sec. 3.1, and a de-scription of their distribution along the color-magnitude diagram(CMD), in Sec. 3.2.1, their masses and morphologies, in Sec. 3.2.2and source of the ionization in the central regions, in Sec. 3.2.3. Allof these properties are presented in comparison with those of galax-ies without a host outflow. The comparison of the integrated SFR(Sec. 3.2.4), the radial distribution of the SFR density (Sec. 3.3),and their central values (Sec. 3.3.1), have lead to the main resultsof this investigation, discussed in Section 4. The conclusions andfuture perspectives are presented in Sec. 5. In this work the stan-dard Λ CDM cosmology with H =
70 km s − Mpc − , Ω m = . Ω Λ = . The analysed sample comprises all galaxies in the CALIFA sur-vey (e.g., S´anchez et al. 2012) up to January 2018, i.e., those withgood quality spectroscopic data observed at the 3.5-m of CentroAstron´omico Hispano-Alem´an (CAHA). It includes the 667 galax-ies comprising the 3rd CALIFA Data Release (e.g., S´anchez et al.2016c), and in addition we include those galaxies with good qual-ity data that were excluded from DR3 because either they did nothave SDSS-DR7 imaging data (a primary selection for DR3) orthey were observed after the final sample was closed, as part ofthe CALIFA-extended programs (e.g., Garc´ıa-Benito et al. 2017;PISCO: Galbany et al. 2018). The final sample comprises a totalof 835 galaxies. All galaxies were selected following the same pri-mary selection criteria of the main CALIFA survey, i.e., that theiroptical extent fits within the field-of-view (FoV) of the instrument,relaxing other selection criteria outlined in Walcher et al. (2014)like the redshift range or the absolute magnitude. Thus, this compi-lation is essentially a diameter-selected survey.The details of the CALIFA survey, including the observa-tional strategy and data reduction are explained in S´anchez et al.(2012) and S´anchez et al. (2016c). All galaxies were observed usingPMAS (e.g., Roth et al. 2005) in the PPaK configuration (e.g., Kelzet al. 2006), covering an hexagonal field of view (FoV) of 74 (cid:48)(cid:48) × (cid:48)(cid:48) ,which is su ffi cient to map the full optical extent of the galaxiesup to two to three disk e ff ective radii. This is possible because ofthe diameter selection of the CALIFA sample (e.g., Walcher et al.2014). The observing strategy guarantees complete coverage of theFoV, with a final spatial resolution of FWHM ∼ (cid:48)(cid:48) , correspondingto ∼ λ/ ∆ λ ∼ ffi cient to explore themost prominent ionized gas emission lines from [O ii ] λ ii ] λ http://califa.caha.es MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. Fernandes et al. 2013a, 2014; S´anchez et al. 2013, 2014, 2016a).The current dataset was reduced using version 2.2 of the CALIFApipeline, whose modifications with respect to previous ones (e.g.,S´anchez et al. 2012; Husemann et al. 2013; Garc´ıa-Benito et al.2015) are described in S´anchez et al. (2016c). The final productof the reduction is a datacube comprising the spatial informationin the x and y axis, and the spectral one in the z axis. For furtherdetails of the adopted data-format and the quality of the data seeS´anchez et al. (2016c). The datacubes were analysed using the P ipe
3D pipeline (e.g.,S´anchez et al. 2016a). P ipe
3D performs a combination of multi-ple synthetic stellar population (SSP) templates, extracted from theMILES (e.g., S´anchez-Bl´azquez et al. 2006; Vazdekis et al. 2010;Falc´on-Barroso et al. 2011) and the gsd156 library (e.g., Cid Fer-nandes et al. 2013b), to determine the best stellar model. Thesetemplates cover a wide range in metallicities from sub solar to suprasolar, with di ff erent stellar ages from 1 Myr to 14 Gyr. Before start-ing with the fitting process, a tesselation procedure is performed onthe data cube in order to increase the signal-to-noise (S / N) of thestellar continuum. This segmentation produces tesselas of di ff er-ent sizes to achieve the desired S / N. All the spectra in each spatialbin is co-added and is treated as individual spectra, and at the endof the fitting analysis, a dezonification of the coadded spectra isapplied by taking into account the area of each tessela (see CidFernandes et al. 2013b). A two dimensional set of data-products,described in S´anchez et al. (2016b), are obtained from the SSP fit-ting. One of such data products is the cumulative stellar mass atdi ff erent epochs. The stellar mass (M (cid:63) ) of a galaxy is estimated byadding the mass in each bin from the tesselation procedure, takinginto account the local luminosity of each spectrum and the mass-to-light ratio (see Gonz´alez Delgado et al. 2015). For a given age(that defines a look-back time), the stellar mass is M (cid:63), age = n (cid:88) j = M j (1)where the j index runs over the number of templates in the SSPlibrary up to the considered age . Integrated over all the complete setof SSP-templates it provides the actual stellar mass of the galaxy.As shown in S´anchez et al. (2016b) and Bitsakis et al. (in prep.),this stellar mass is totally consistent with the one provided usingmulti-band photometric data.Having estimated the stellar mass at a certain look-back time,it is straightforward to estimate the star-formation rate at this partic-ular time. The SFR would be the di ff erential mass at two adjacenttimes ( ∆ t age ) over the time range between them ∆ t age : S FR age = ∆ M age ∆ t age (2)In Gonz´alez Delgado et al. (2017) and Sanchez et al. (in prep.) ithas been explicitly shown that this star-formation rate, that we willdefine as SFR SSP , correlates very well with other estimations of theSFR.
Once obtained the best stellar population model for each spectrain the data cube, it is subtracted from the original cube to obtain apure-gas cube, following the procedures described in S´anchez et al. (2016a). Then, we analysed each of the detected emission lines ineach individual spectrum within this cube using the fitting code fit
3D (e.g., S´anchez et al. 2016b). For this particular study it wasperformed a non parametric method based on a moment analysisin the pure-gas cube as described in S´anchez et al. (2016b). We re-cover the main properties of the emission lines, including the inte-grated flux intensity, line velocity and velocity dispersion. For thisanalysis, we assume that all emission lines within a spaxel sharethe same velocity and velocity dispersion. The result of this pro-cedure applied to each data cube is a set of bi-dimensional mapsof the considered parameters, with their corresponding errors, foreach analysed emission line.In addition to these parameters the equivalent width (EW) ofeach emission line is derived. In particular that of H α which willbe used in our scheme of classification of the ionization source. Toderive this quantity the stellar continuum flux density is estimatedprior to the subtraction of the stellar model. Then, the integratedflux of each emission line, derived by the moment analysis is di-vided by this continuum density, at the wavelength of the emissionlines, resulting in the required EW. Highly inclined galaxies are particularly good candidates to detectextraplanar ionized gas and therefore they are more suitable candi-dates to host outflows. We started the selection process by consid-ering only those galaxies with high inclination ( i > ◦ ), in order tominimize the e ff ect of mixing of ionization along the line-of-sightdue to projection e ff ects. Although we cannot preclude for a certainlevel of contamination. Using this criterion results in 203 galaxies.Then, we select those galaxies with an increase in the optical lineratios [N ii ] / H α , [O iii ] / H β , [S ii ] / H α and [O i ] / H α along the semi-minor axis and the disk vertical direction. These increments arecharacteristic of shocks produced by galactic outflows (Veilleux &Rupke 2002; Veilleux et al. 2005), although they are not exclusiveof these processes. Here increase means that the ionization is notcompatible with the typical line-ratios observed in SF regions. Out-flows are favoured to expand in the direction of the lowest gradientof pressure, which is found along the semi-minor axis or in the ex-traplanar region. As larger the inclination, sharper will be observedthe separation between the soft ionization from the SF regions, andthe harder would be the ionization produced by shocks in outflows.Within the high inclined sub-sample we find that 39 galaxiespresent such line ratios enhancement. In Figure 1 we present onegalaxy, NGC 6286, that complies with these criteria. In this figureit is clear how extraplanar gas extends beyond the continuum ex-tension. Hereafter, we will refer to the disk region, regardless ofthe inclination, as the area located within ± extraplanar region to the area located fartherthan 5 (cid:48)(cid:48) of this axis. This transition region varies in each galaxy,although it represents a mean value at the average redshift of CAL-IFA ( ∼ MNRAS , 1–16 (2018)
C. L ´opez-Cob ´a et al. [h] ∆ D E C ( a r c s e c ) NGC6286 V-band[NII][OIII] log([NII]/H α ) log([OIII]/H β ) ∆ RA (arcsec) ∆ D E C ( a r c s e c ) log EW(H α ) ∆ RA (arcsec) log([SII]/H α ) ∆ RA (arcsec) log([OI]/H α ) Figure 1.
Example of a wind galaxy selected from the CALIFA sample with the high inclination and line-ratio criteria. This galaxy, NGC 6286, is part of thecandidates galaxies with a host outflow. The top left panel, shows the RGB image of NGC 6286, where red is [N ii ], green is the V-band and blue is [O iii ].The top central panel, is the spatially resolved [N ii ] / H α line ratio map. The black contour in this, and in the others maps, indicate the continuum level at0.1 and 0.05 × − erg s − cm − , while the intensity color bar is in the right corner of each map. The top right panel, is the spatially resolved [O iii ] / H β line ratio map. The bottom left panel, shows the 2D-Equivalent Width of H α estimated with the SSP fitting analysis. The bottom central panel, shows the[S ii ] λλ , / H α line ratio map; and the bottom right panel, the [O i ] λ / H α line ratio map. Depending on the shock velocities, ionization by shocks cancover a wide range of line ratios making them rather di ffi cult toidentify in the classical diagnostic diagrams (e.g., Baldwin et al.1981; Veilleux & Osterbrock 1987), at di ff erence with other ion-izing sources which are confined to specific regions of these dia-grams. Therefore, it is not possible to define demarcation curves,like the ones used to separate, for example, SF and AGN-like ion-ization zones (K01: Kewley et al. 2001; K03: Kau ff mann et al.2003; S06: Stasi´nska et al. 2006). Even more, shock-like ionizationoverlaps in these diagrams with locations of SF regions of interme-diate / high metallicities, low-luminosity AGNs and / or post-AGBsionizations (e.g., Alatalo et al. 2016). Despite of its complexity,there have been e ff orts trying to constrain the location of shocks inthese diagrams, either with direct observations of bonafide outflows(e.g., Sharp & Bland-Hawthorn 2010), or by the implementation ofshock models such as mappings - iii (e.g., Dopita & Sutherland 1995,1996; Allen et al. 2008).In addition to this problem, some characteristics of outflows, like the enhanced line ratios in the extraplanar regions and theirlocation in the diagnostic diagrams, are shared with the ioniza-tion produced by the so-called hot low-mass evolved stars (e.g.,HOLMES, Flores-Fajardo et al. 2011). These old stars can domi-nate the production of ionizing photons with respect to young mas-sive stars and produce the ionization observed in the extraplanardi ff use ionized gas (eDIG / DIG). Their e ff ect could be particularlyimportant in early-type galaxies (e.g, Binette et al. 1994; Stasi´nskaet al. 2008), and can reproduce the observed ionization in the so-called low-ionization nuclear emission-line regions (LINER, Heck-man 1980), classically associated with low-luminosity AGNs. Thisionization has been recently found to be ubiquitous in the retiredregions of any galaxy (e.g., Singh et al. 2013; Belfiore et al. 2016),and frequently detected in edge-on galaxies (e.g., Jones et al. 2017;Lacerda et al. 2018). In star-forming galaxies, the contaminationby DIG, regardless of the galaxy inclination, is to move the SF re-gions towards the composite or the LINER region in the diagnosticdiagrams (e.g., Zhang et al. 2017). Therefore, if we had adopted MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. [h] log [NII]/H . . . . . l o g [ O III ] / H SF Comp
EW(H ) < 3 Å log [SII]/H log [OI]/H log [NII]/H E W ( H ) SF sAGNwAGNDIG-HOLMES
Figure 2.
Diagnostic diagrams for the galaxy NGC 6286. All the emission lines involved have a S / N > Cyan dots represent spaxels lying in the disk while pink dots represent spaxels lying beyond 5 (cid:48)(cid:48) from the disk, i.e., in the extraplanar region. The dark blue colors in this colormap represent the transition fromdisk to the extraplanar region. This color scheme is adopted in all the candidates regardless of their inclination. The black continuous curve in the first threepanels represent the Kewley et al. (2001) demarcations from SF regions and AGNs. The red and blue lines in the first panel represent the Kau ff mann et al.(2003) and Stasi´nska et al. (2006) curves respectively. The black circle represent those spaxels in which EW(H α ) < α ) > ii ] / H α vs. EW(H α ). a lower value in the inclination angle, or if we had searched foroutflows regardless of their inclination, it would increase the DIGfraction in our sample.A characteristic that share both DIGs and HOLMES is theirlow equivalent width of H α (e.g., Stasi´nska et al. 2008). As amethod to distinguish this ionization from shocks, we adopted theWHAN diagram introduced by Cid Fernandes et al. (2011) whichuses the [N ii ] / H α vs. equivalent width of H α to distinguish be-tween true and fake AGNs (retired galaxies with EW(H α ) < α ) < ii ] / H α , [S ii ] / H α and [O i ] / H α vs.[O iii ] / H β ) along with the WHAN diagram, for the spatially re-solved components, both disk and extraplanar regions, applied tothe archetypal outflow galaxy NGC 6286. Although a fraction ofthe extraplanar gas falls below the SF demarcation line by K01,probably due to projection e ff ects, and a fraction of the disk gasfalls in the sAGN region of the WHAN diagram, it is clear that theextraplanar gas is not compatible with being ionized by old starsbut by a strong source of ionization. We have included in these di-agrams the locus of AGN and shock-excited emission from Sharp& Bland-Hawthorn (2010) to distinguish between these two typesof outflows. The location of the outflowing gas in the diagnosticdiagrams is mainly distributed in the shock-excited region (to theright from the bisector line). We refine the classification proposedby Sharp & Bland-Hawthorn (2010), imposing the condition thatthe ionized gas should have an EW(H α ) >
3Å in the extraplanar re-gion to be classified as an outflow, and below this limit to be classi-fied as DIG, as indicated before. Therefore, those spaxels above theKewley et al. (2001) curve, at the left-size of the Sharp & Bland- Hawthorn (2010) demarcation line, in the extraplanar region, andwith above the indicated EW(H α ) will be classified as AGN-drivenoutflows (or compatible with being ionized by an AGN). On theother hand, all those spaxels following the same criteria, but at theright-side of the Sharp & Bland-Hawthorn (2010) demarcation line,would be classified as shock ionized (or SF-driven outflows in gen-eral). In particular, for NGC 6286, the extraplanar gas is consistentto be ionized in 100% of the spaxels by shocks, 0% by AGN and0% by DIG according to the indicated criteria. The spatial distribu-tion of the line ratios shown in Fig. 1, together with their locationwithin the four diagrams of Fig. 2, allow us to conclude that theobserved outflow in NGC 6286 is most probably a SF-driven andshock-excited wind. Note that the wind outflow in this galaxy wasalready suspected in the paper by Shalyapina et al. (2004) based onscanning Fabry–Perot interferometer observations in [N ii ] and H α emission lines. Now CALIFA diagnostic diagrams give a detailedpicture of the gas ionization in the butterfly–like extended nebulaeof this object.In summary, to select galaxies that host an outflow we adoptthe following selection criteria: (i) high inclined galaxies, (ii) de-tection of extraplanar ionized gas, (iii) identification of an enhancein the line ratios along the semi-minor axis, (iv) EW(H α ) > can-didates with a host outflow. We cannot firmly conclude that theyhost an outflow since we lack of the required high spectral res-olution data to perform a detailed kinematics analysis. Thus, wecannot resolve the asymmetries in the emission line profiles, fre-quently detected in outflows due to the expansion of the gas, oranalyse the known correlation between the velocity dispersion andthe line ratios, a unique signature of shock ionization (e.g., Do- MNRAS000
3Å in the extraplanar re-gion to be classified as an outflow, and below this limit to be classi-fied as DIG, as indicated before. Therefore, those spaxels above theKewley et al. (2001) curve, at the left-size of the Sharp & Bland- Hawthorn (2010) demarcation line, in the extraplanar region, andwith above the indicated EW(H α ) will be classified as AGN-drivenoutflows (or compatible with being ionized by an AGN). On theother hand, all those spaxels following the same criteria, but at theright-side of the Sharp & Bland-Hawthorn (2010) demarcation line,would be classified as shock ionized (or SF-driven outflows in gen-eral). In particular, for NGC 6286, the extraplanar gas is consistentto be ionized in 100% of the spaxels by shocks, 0% by AGN and0% by DIG according to the indicated criteria. The spatial distribu-tion of the line ratios shown in Fig. 1, together with their locationwithin the four diagrams of Fig. 2, allow us to conclude that theobserved outflow in NGC 6286 is most probably a SF-driven andshock-excited wind. Note that the wind outflow in this galaxy wasalready suspected in the paper by Shalyapina et al. (2004) based onscanning Fabry–Perot interferometer observations in [N ii ] and H α emission lines. Now CALIFA diagnostic diagrams give a detailedpicture of the gas ionization in the butterfly–like extended nebulaeof this object.In summary, to select galaxies that host an outflow we adoptthe following selection criteria: (i) high inclined galaxies, (ii) de-tection of extraplanar ionized gas, (iii) identification of an enhancein the line ratios along the semi-minor axis, (iv) EW(H α ) > can-didates with a host outflow. We cannot firmly conclude that theyhost an outflow since we lack of the required high spectral res-olution data to perform a detailed kinematics analysis. Thus, wecannot resolve the asymmetries in the emission line profiles, fre-quently detected in outflows due to the expansion of the gas, oranalyse the known correlation between the velocity dispersion andthe line ratios, a unique signature of shock ionization (e.g., Do- MNRAS000 , 1–16 (2018)
C. L ´opez-Cob ´a et al. pita & Sutherland 1995; Lehnert & Heckman 1996; Monreal-Iberoet al. 2010; L´opez-Cob´a et al. 2017). For instance, in the specialcase of NGC 6286 the line-of-sight ionized gas velocity dispersionsignificantly increases outside the stellar disk (according to Fig. 5in Shalyapina et al. 2004).In Table 1 we summarize the main properties of this sample ofgalaxies and in the Appendix A we present the same plots shownfor NGC 6286 (i.e., equivalent to Figs. 1 and 2), for all the outflowcandidates in the CALIFA sample. In addition, we list in Table B1the remaining 26 galaxies that were not classified as outflows bythe imposed criteria, although they present extraplanar di ff use ion-ized gas (eDIG), and in may cases, show an enhancement in theanalysed line ratios. Following the criteria indicated before we listin this table the fraction of spaxels in the extraplanar region beingcompatible with either DIG, AGN-driven or SF-driven outflows,based on the combination of the classical diagnostic diagrams, thevalue of the EW(H α ), and the Kewley et al. (2001) and Sharp &Bland-Hawthorn (2010) demarcation lines.So far, we finish our classification process with three di ff erentsub-samples: (i) Those galaxies with i < ◦ , which we will denoteas the CALIFA low-inclination sample, or just CALIFA sample forsimplicity, since it dominates the number statistics (615 galaxies),and therefore comprises a representative sub-sample of the originalone, (ii) the high-inclination galaxies (203 galaxies), and (iii) theoutflow candidates (17 galaxies). In this section we explore the global properties of the candidateswith a host outflows in comparison with those of the other two sub-samples of galaxies.
Figure 3 shows the distribution of the three galaxy samples in thecolour magnitude diagram (CMD) for the ( g − r ) colour versusthe g –band absolute magnitude. The galaxies from CALIFA spanover the full CMD from the red sequence to the blue cloud andover the intermediate region known as green valley, populated bytransition galaxies and AGN hosts (e.g., S´anchez et al. 2018). Theglobal properties of the CALIFA sample has been reported in pre-vious papers for the di ff erent data releases (e.g., S´anchez et al.2012; Walcher et al. 2014; Garc´ıa-Benito et al. 2015; S´anchez et al.2016c).Now, we would like to investigate the di ff erences between thethree sub-samples. In order to quantify these di ff erences, we per-formed a two dimensional Kolmogorov-Smirnov test (2D KS, Pea-cock 1983; Fasano & Franceschini 1987; Press et al. 1992). Thistest compares two 2D distributions. The null hypothesis is that theobserved population of galaxies (the high inclined or the candi-dates) is drawn randomly from a parent population (CALIFA or thehighly inclined galaxies). Typically one assumes a critical p–valueto reject the null hypothesis. In our case we will adopt a p–value of0 .
05. As mentioned in Press et al. (1992), the resulting p–value inthe 2D KS is only an approximation, and the test is accurate enoughwhen N ∼
20 and p–value (cid:46) .
20. We applied the 2D KS test forthe galaxies from the CALIFA sub-sample and the highly inclinedgalaxies. The resulting p–value is of the order of 10 − , which ishighly significant. This implies that both samples are statisticallydi ff erent. Thus, the highly inclined galaxies are not a representativesub-sample of the CALIFA galaxies, at least in the space of pa-rameters considered. This is not really surprising, because the full
22 20 18 16Absolute Magnitude, M g . . . . . . ( g - r ) c o l o u r E S0 Sa Sb Sbc Sc Sd Sm Ir N g a l a x i e s N g a l a x i e s CALIFA i > 70 o Candidates
Figure 3.
Top panel, distribution of the three sub-samples in the ( g − r )vs. M g diagram. Blue dots represent the CALIFA sub-sample, red squares represent the high inclination galaxies and green stars the outflow candi-dates. Histograms of the ( g − r ) and M g distributions for each sub-sampleare included in each axis. The colour code of the histograms are the sameas mentioned before. Middle panel: histogram of the morphological distri-bution for three sub-samples.
Bottom panel: histogram of the stellar massdistribution. The morphology histogram includes the 734 galaxies observedby CALIFA up to 2017 (e.g., S´anchez et al. 2017) while the others two his-tograms include the total sample of 835 galaxies (i.e., including the PISCOsample, Galbany et al. 2018). MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. log [NII]/H . . . . . l o g [ O III ] / H log [SII]/H log [OI]/H E W ( H ) Figure 4.
Diagnostic diagrams for the central regions (3 (cid:48)(cid:48) × (cid:48)(cid:48) ) of the galaxies in the three sub-samples: the CALIFA sub-sample (cyan dots), the high inclinedgalaxies (red dots) and the candidates galaxies (green stars). The green colour code in the lower-right of the first panel represents the EW(H α ). We have codedonly the EW(H α ) for the candidates. The continuous black curve in the three panels represent the Kewley et al. (2001) demarcation curves. The blue andred lines in the first panel represent the demarcations from Kau ff mann et al. (2003) and Stasi´nska et al. (2006) respectively. The broken line represents thedemarcation between Seyfert (up-left) and LINER (up-right) from Kewley et al. (2001). CALIFA sample comprises a wide range of morphological types,with a substantial fraction of elliptical galaxies, that by definitionare more roundish and prompt to be rejected from a selection ofhigh-inclined galaxies based on the semi-major to semi-minor ra-tio. We now applied the 2D KS test for the high inclination andcandidate galaxies. The resulting p–value is 0.007, which is alsosignificant at the 0.05 confidence level. In this case it is not obvi-ous why these two samples should be so di ff erent. An inspectionof Fig. 3 shows that the candidates occupy a narrower region in theCMD diagram, − . < M g < − .
0, which would probably re-flect a bias in the luminosity distribution. In other words, we do notfind outflow candidates brighter than − . v ∼ Figure 3, middle panel, shows the distribution in morphology andstellar mass for the three sub-samples. The CALIFA sub-sample isdistributed in a wide variety of morphological types, from early- tolate-types and irregulars, and stellar masses 6 < log M (cid:63) <
12. Thehigh inclination sub-sample is clearly dominated by spiral galaxies(1 elliptical). Their mass distribution seems to follow the same asthe CALIFA sub-sample, but without high / low mass galaxies, i.e.,restricted to 8 . < log M (cid:63) < .
5. Finally, the outflow candidatesample only includes spiral galaxies of types Sa, Sb, Sc and Sd.The masses in this sub-sample are distributed in an even narrowerrange, 9 . < log M (cid:63) < ff erences between the massdistributions of the candidates and the other two sub-samples, we applied the one dimensional Kolmogorov-Smirnov (K-S) test (e.g.,Press et al. 1992). The K-S test compares the maximum di ff erencebetween two cumulative distribution functions. As larger is the dif-ference between two distributions, larger is the probability that thetwo distributions arise from di ff erent samples.With a resulting p–value of 0.02, the K-S reveals a signifi-cant di ff erence, at the level of 0.05, between the high- and low-inclination galaxies. This is in concordance with the result of the2D KS estimated for the CMD in the previous section. On theother hand, the resulting p-value for candidates and the high in-clined galaxies is 0.25. So, the mass distribution of the candidatesub-sample is consistent of being drawn from the same mass distri-bution of the high inclination galaxies. Thus, the candidates presenta similar stellar mass as the highly-inclined galaxies. Therefore, thedi ff erences found in the CMD are most probably due to a di ff erencein colour, rather than in absolute magnitude (or mass). In summary,the galaxies hosting outflows seem to be slightly brighter, with asimilar stellar mass and slightly more evolved stellar populationsor with larger dust attenuations than the average inclined galaxies. In order to investigate the dominant ionization in the nuclear regionof the galaxies from the three sub-samples, we co-add the emissionline fluxes of H α , H β , [O iii ], [N ii ], [S ii ] and [O i ] over an area of3 (cid:48)(cid:48) × (cid:48)(cid:48) at the nucleus of each galaxy. Then, we plot the line ra-tios in the diagnostic diagrams explained before. This is shown inFig. 4. We note that the CALIFA and the highly inclined galaxiesare distributed following the classical seagull shape, which reflectsthe variety of ionizing sources in these samples. A large fraction ofAGN populate these two sub-samples. The number of AGN in eachsub-sample varies depending on which diagnostic is used to clas-sify them. An AGN is classified if it lies above the K01 curve and MNRAS000
5. Finally, the outflow candidatesample only includes spiral galaxies of types Sa, Sb, Sc and Sd.The masses in this sub-sample are distributed in an even narrowerrange, 9 . < log M (cid:63) < ff erences between the massdistributions of the candidates and the other two sub-samples, we applied the one dimensional Kolmogorov-Smirnov (K-S) test (e.g.,Press et al. 1992). The K-S test compares the maximum di ff erencebetween two cumulative distribution functions. As larger is the dif-ference between two distributions, larger is the probability that thetwo distributions arise from di ff erent samples.With a resulting p–value of 0.02, the K-S reveals a signifi-cant di ff erence, at the level of 0.05, between the high- and low-inclination galaxies. This is in concordance with the result of the2D KS estimated for the CMD in the previous section. On theother hand, the resulting p-value for candidates and the high in-clined galaxies is 0.25. So, the mass distribution of the candidatesub-sample is consistent of being drawn from the same mass distri-bution of the high inclination galaxies. Thus, the candidates presenta similar stellar mass as the highly-inclined galaxies. Therefore, thedi ff erences found in the CMD are most probably due to a di ff erencein colour, rather than in absolute magnitude (or mass). In summary,the galaxies hosting outflows seem to be slightly brighter, with asimilar stellar mass and slightly more evolved stellar populationsor with larger dust attenuations than the average inclined galaxies. In order to investigate the dominant ionization in the nuclear regionof the galaxies from the three sub-samples, we co-add the emissionline fluxes of H α , H β , [O iii ], [N ii ], [S ii ] and [O i ] over an area of3 (cid:48)(cid:48) × (cid:48)(cid:48) at the nucleus of each galaxy. Then, we plot the line ra-tios in the diagnostic diagrams explained before. This is shown inFig. 4. We note that the CALIFA and the highly inclined galaxiesare distributed following the classical seagull shape, which reflectsthe variety of ionizing sources in these samples. A large fraction ofAGN populate these two sub-samples. The number of AGN in eachsub-sample varies depending on which diagnostic is used to clas-sify them. An AGN is classified if it lies above the K01 curve and MNRAS000 , 1–16 (2018)
C. L ´opez-Cob ´a et al. presents an EW(H α ) > ii ] / H α diagram, 71 and 28 AGNin the [S ii ] / H α diagram, and finally 40 and 14 AGN in the [O i ] / H α diagram. On the other hand, a large fraction of the outflow candi-dates are grouped in the SF region. Only two galaxies lie above theK01 demarcation, one of them it is notably far away from this de-marcation, in the AGN region (NGC 4388). As pointed in previoussubsections, the classical interpretation of the diagnostic diagramswhich attempt to separate between di ff erent sources of ionization isno longer valid without other extra parameters like the EW(H α ) orany other physical information about the source of ionization. Allcandidates present EW(H α ) > α ) crite-rion, these galaxies are dominated by SF (15 of them) or AGN (1weak and 1 strong). From the total candidates, 3 /
17 galaxies are cat-alogued as X-ray sources, NGC 4676A (log L X = .
2, Gonz´alez-Mart´ın et al. 2009), NGC 4388 (log L X = .
45, Corral et al. 2014)and NGC 6286 (log L X = . L X >
42, the classical limit to be considered as an AGN. Indeed,this is the only target which outflowing material is compatible withbeing ionized by and AGN-driven ionization, based on the schemedescribed in Sec. 3.1 (as indicated in Table 1 and Figure A1). Insummary, our selection of highly inclined candidates to outflowsseem to bias the sample towards outflows driven by star-formationin the vast majority.We should stress out that our selection bias the sample againstearly-type galaxies (as shown in the previous section), and this, byconstruction, excludes the detection of outflows in these galaxiesthat in their vast majority should be dominated by the presence ofan AGN. In particular, we are excluding the detection of the re-cently classified as
Red Geysers , a kind of object first reported byKehrig et al. (2012), and confirmed by Cheung et al. (2016), andmost probably associated with a weak AGN activity.
In the previous section we show that most of our candidates tohost an outflow present ionized gas in their central regions dom-inated by star-formation (15 of 17). We explore in this section ifthis star-formation is more intense than the one of the other twosub-samples.The SFR is a measurement of how much mass in stars isformed during a period of time. Star formation bursts create starsin a wide range of masses following a certain initial mass func-tion (e.g., Salpeter 1955; Kroupa 2001; Chabrier 2003), but onlythe massive ones will dominate the production of ionizing pho-tons ( > ∼ α (SFR = . × − L H α ; Kennicutt 1998). Thismethod requires that the measured H α flux is produced only by SFprocess, which is not necessary true in the presence of an AGN,shocks or other ionization sources (e.g., Catal´an-Torrecilla et al.2015). To derive L H α , we integrate the observed H α flux, and aftercorrection by dust attenuation using the extinction law by Cardelliet al. (1989), assuming the case B of recombination (e.g., Oster-brock 1989) and using the cosmological distance for each galaxy.We applied the Kennicutt (1998) law to transform L H α into SFR H α .In this estimation we ignored the contribution of other sources ofionization. However, as shown by Catal´an-Torrecilla et al. (2015)and S´anchez et al. (2017) their e ff ects are limited. Nevertheless, in retired galaxies dominated by old stellar population, this relationmust be considered just as a linear transformation between the L H α to SFR.In Fig. 5 we show the well known relation between the SFRand the integrated stellar mass (e.g., Brinchmann et al. 2004; Salimet al. 2007; Noeske et al. 2007). In this figure we plot both theSFR H α and SFR SSP (explained in Sec. 2.1) for the outflow can-didates and the other two sub-samples. In both panels the CAL-IFA sub-sample is distributed in a bimodal sequence shown by theblue contours, one comprising active star-forming galaxies, the socalled star formation main sequence (SFMS, e.g., Brinchmann et al.2004; Salim et al. 2007), and the other the passive or retired se-quence of galaxies (RSG). These sequences have been previouslystudied spatially resolved for the CALIFA sample (e.g., Cano-D´ıazet al. 2016). The high inclination galaxies are distributed around theSFMS with some galaxies falling in the RSG and the green valley.On the other hand, the outflow candidates are distributed aroundthe SFMS regardless of the calibrator used to estimate the SFR, i.e.,no excess is evident. This result has been previously noticed by Hoet al. (2016) in a sample of outflows selected from the SAMI survey.Figure 5 also shows evidence that using the full optical extensionof galaxies, no excess in the SFR of the candidates is appreciated asit would be expected if outflows are driven by strong periods of SF.The outflow candidates seem to be part of the normal star-forminggalaxies as revealed by the χ test. Although it seems that outflowsare preferentially located along the SFMS, their location in this di-agram does not seem to define if a galaxy hosts or not an outflow.Recent studies have pointed out that the local concentration of theSFR might play an important role when driving outflows (e.g., Hoet al. 2016). In other words, the SFR surface density may be a bet-ter parameter instead of the integrated SFR to trace or regulate thepresence of outflows. Early studies in local starburst galaxies and high–z Lyman breakgalaxies, have evidenced that outflows are ubiquitous in galaxieswith SFR surface densities ( Σ SFR ) larger than 10 − M (cid:12) yr − kpc − (e.g., Heckman 2001, 2002). Based on these results this value hasbeen adopted in the literature as a canonical threshold for outflows.Motivated by these results, and the results from the previoussection, we proceed to estimate the radial distribution of the Σ SFR .One of the great advantages of IFS is its capability to study the spa-tially resolved properties of galaxies, like Σ SFR , instead of derivingit averaged across the entire optical extension of galaxies like itwas done in previous analysis. For example, Kennicutt (1998) usedthe area within the isophotal radius of the galaxies (D = ) toestimate Σ SFR ( = SFR /π R ). Other authors have adopted the e ff ec-tive radii to estimate the area of the galaxies ∼ π R (e.g., Lundgrenet al. 2012; Ho et al. 2016) or it has been determined by impos-ing the Schmidt-Kennicutt law (SK-law, Kennicutt et al. 1989). Insome cases in which it was possible to estimate the size of the star-burst region (few hundreds of pc) it was adopted as the proper areawhere star formation is detected (e.g., Wood et al. 2015). Thesedi ff erences in the procedure adopted to derive the Σ SFR introduceclear uncertainties in the absolute scale of the proposed canonicalthreshold described before.In our case we estimate the SFR derived from H α and theSSPs at di ff erent galactocentric elliptical rings, following the posi-tion angle and ellipticity of the object. Then we divide each regionby the physical area of the corresponding ring, corrected by the in-clination angle, to finally obtain the radial distribution of Σ SFR for
MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. log M [M ] l o g S F R H [ M y r ] R S G S F M S = log M [M ] l o g S F R SS P [ M y r ] S F M S = Figure 5.
Star formation rate derived from H α ( left panel ) and estimated with the SSPs ( right panel ) vs. the integrated stellar mass for the three sub-samples.Cyan and red contours enclose the 90, 68, and 34 % of the total data in the CALIFA and the high inclination galaxies respectively. Green stars represent theoutflow candidate galaxies. The yellow star represent the strong AGN, NGC 4388, found in the candidates as shown in Fig. 4. The continuous and dashedblack lines in the left panel correspond to the spatially resolved star formation main sequence (SFMS H α ) and the retired sequence of galaxies (RSG) derivedby Cano-D´ıaz et al. (2016). The black line in the right panel correspond to the SFMS SSP derived from the best fit for SF galaxies in the full CALIFA sample(EW(H α ) >
3Å and line ratios below the K01 curve) . The slope and zero point correspond to 0 . ± .
02 and − . ± .
20 respectively. A chi squared testwas applied for the candidates and the theoretical value given by the SMFS. The reduced chi squares is shown in the top left corner in each panel. each galaxy. We selected annular rings of 0.1 R e width, up to 3 R e c.f., Fig. 6. In addition, we estimate the Σ SFR with the SFR derivedfrom the SSP fitting analysis (SFR
SSP ), as described in Sec. 2.1.This method has the great advantage that it does not depend onthe physical properties of the ionized gas. However, the SFR
SSP isonly estimated where stellar continuum is detected. This means thatSFR
SSP traces pure SF with no contamination, at the penalty of alower precision (due to the limitations of the SSP-fitting procedure,S´anchez et al. 2016a). P ipe
3D estimates the SFR
SSP for the assem-bled mass in the last ∆ t =
32 Myr, as described in Sec. 2.1. Weadopted the same annular rings for this complementary estimationof the star-formation density.Figure 6 shows the radial profiles of Σ SFR estimated based onthe H α flux and the SSP fitting analysis. These plots were con-structed by taking the average value of the Σ SFR for each galaxyin each radial bin, for each sub-sample of galaxies. The inner 0.2R e ( ∼
500 pc) are unresolved due to the PSF size (FWHM ∼ . (cid:48)(cid:48) S´anchez et al. 2016c). Therefore, any trend below this inner regionshould be taken with care. In this plot we considered only galax-ies dominated by SF in each sub-sample (i.e., those galaxies withline ratios below the K01 curve in Fig. 4). Adopting this criterionfor the high and low inclined and outflow candidate galaxies, thesub-samples are limited to 412, 158 and 16 galaxies, respectively.The radial profiles show that in both cases the candidate galax-ies present, on average, higher values of the Σ SFR , at least in theinner regions, when compared to the other two sub-samples. The Σ SFR estimated with the SSP method has in general larger valuesin comparison with the estimated with H α because of the di ff erenttime scales of both methods. The SFR based on H α traces the SF in the last ∼ Σ SFR in the innermost regions thatin the outermost. This comparison between both estimators of theSFR exhibit that the observed excess is not due to a possible con-tamination by an extra source of ionization of H α .In order to quantify how significant is the di ff erence in the ra-dial distribution of the Σ SFR of the candidates in comparison thetwo other samples, we performed an Anderson-Darling test (A-D,Press et al. 1992; Feigelson & Babu 2012). In contrast with theK-S test which tends to be more sensitive to di ff erences in the cen-tral regions of the distributions, the A-D test is more sensitive todi ff erences also in the outermost regimes of the distributions. Weapplied the A-D test at each radii, comparing the distributions of Σ SFR for the candidates and the high inclination sub-samples, andthe outflow candidates and the low inclination galaxies. The nullhypothesis in both cases is that at each radius both distributionsof Σ SFR are sub-samples from the same population. The results ofthis statistical tests are shown in the bottom insets of Fig. 6. Byadopting a significance level of 5% we see that in the inner radii,the resulting p-value for the candidates and the high inclined galax-ies is clearly below the significance level. This is still significantat a 1% level. Although the radius at which the null hypothesis isrejected depends on the calibrator used to estimate the SFR, it iscertainly clear that below 1 R e both distributions seems to be di ff er-ent. Indeed at this point the Σ SFR profile presents a clear break andsteepening most probably due to the outflow contribution inside thisregion. As we go to outer regions in the galaxies, the outflow can-
MNRAS000
MNRAS000 , 1–16 (2018) C. L´opez-Cob´a et al.
R/Re . . . . . l o g ( S F R [ M y r k p c ]) ( H ) CALIFA (SF) i > 70 (SF)CANDIDATES (SF)1.0 1.5 2.0 2.5 3.0 R e . . . p - v a l u e AD testKS test
R/Re . . . . . l o g ( S F R [ M y r k p c ]) ( SS P ) R e . . . p - v a l u e Figure 6.
Radial profiles of Σ SFR for the three sub-samples estimated with H α ( left panel ) and the SSP fitting analysis from P ipe
3D ( right panel ). In bothpanels the green line represents the Σ SFR profile for the outflow candidates (excluding the AGNs, see table 1), the red line represents the Σ SFR radial profilefor galaxies with high inclination catalogued as SF. The cyan line represent the radial profile for the low inclined galaxies (CALIFA subsample) cataloguedas SF. We define SF as those galaxies with nuclear line-ratios lying below the K01 curve. Shadow regions represent the standard deviation at each bin of thee ff ective radii. The bottom insets in both panel represent the A-D and K-S tests at each bin of the e ff ective radius. The blue and red connected points representthe p-value obtained at each bin of the e ff ective radius for the candidates and the high inclined galaxies for both statistical test. The black connected pointsrepresent the p-value obtained at each bin of the e ff ective radius for the candidates and the CALIFA sub-sample. We have plotted the resulting p-values upto 0.5, larger p-values result obviously in the acceptance of the null hypothesis. The black arrow in both insets represent the transition point where the nullhypothesis goes from being rejected to being accepted. In the left inset this occur at R = . e while in the right plot occurs at R = e . An example of thering segmentation performed to calculate the radial distribution of Σ SFR is shown in the left panel as an illustration. didates follow the same behaviour of the high-inclination galaxies.On the other hand, the p-values for the outflow candidates and thelow-inclination CALIFA sub-sample show that both distributionsare di ff erent at any radius.It is interesting to note the low values of the Σ SFR that outflowcandidates have, with values always lower log Σ SFR < − .
0, whichis below the canonical value expected for outflows, described be-fore. We will explore this result in more detail in the next section. Σ SFR vs. Σ gas : The K-S law. It has been frequently adopted the density of SFR as the main pa-rameter that controls the production of an outflow (e.g., Kennicutt1998; Heckman 2002). As larger the Σ SFR is, more concentratedwould be the energy released by the SN explosions and thereforethe over pressured cavity would expand until large scale galacticwinds are driven giving rise to an outflow (e.g., Heckman et al.1990). To achieve high values in the Σ SFR it is needed a high SFRconcentrated in small regions (hundreds of pc). A large SFR is re-flected in a large fraction of gas that is transformed into newborn stars, from which only the massive ones (O-B stars) will contributeto the formation of the winds required to produce outflows.It is well known that there is a tight correlation between the Σ SFR and the Σ gas (molecular and atomic) content in galaxies (theso-called K-S law, Kennicutt et al. 1989). Although we are not ableto measure directly the gas fraction, due to the lack of CO and HIobservations for all galaxies in our sample, it is still possible to havean estimation of the molecular gas content via the dust extinction.Following S´anchez et al. (2018), we proceed to estimate the gascontent via the extinction A V : Σ gas = A V mag [M (cid:12) pc − ] (3)This relation presents a scatter of ∼ . α and the SSP analysis, and the molecularmass density estimated from Eq. 3, averaged within the central re-gions (R < . e ) of the individual galaxies of our three sub-samples. If we focus in the left panel, we observe that the can- MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. log ( gas [M pc ]) . . . . . . . . l o g ( S F R [ M y r k p c ]) ( H ) CALIFA (SF) i > 70 (SF)CANDIDATES 0.5 0.0 0.5 1.0 1.5 2.0 log ( gas [M pc ]) . . . . . . . . l o g ( S F R [ M y r k p c ]) ( SS P ) CALIFA (SF) i > 70 (SF)CANDIDATES Figure 7. Σ SFR vs. Σ gas for the individual galaxies in three sub-samples; left panel shows the Σ SFR estimated with H α ; right panel shows the Σ SFR estimatedwith the SSP fitting analysis. The cyan dots represent the SF galaxies in the CALIFA sub-sample. The red squares represent the SF galaxies in the high inclinedsub-sample and the green stars the candidates (excluding the AGNs, see Table 1). The green stars represent the outflow candidate galaxies. The horizontaldashed line represent the canonical threshold expected for outflows (log Σ SFR = didates are basically concentrated in the region of galaxies withSF around the K-S relation. On the other hand, the high inclina-tion and the CALIFA sub-samples are distributed in a cloud, nar-rower for the first, also around the K-S relation. The scatter is largerthan the usually reported for the K-S law, most probably due to therough estimation of the gas density (as already noticed by S´anchezet al. 2017). In this panel we observe that only one of the candidategalaxies surpass the threshold of 10 − M (cid:12) yr − kpc − ( = Σ SFR , threshold ).If we now focus on the right panel, we observe that a large frac-tion of the candidates are concentrated in a small region close thecanonical value. Indeed ∼
95% of the candidates present SFR sur-face densities larger than 10 − . M (cid:12) yr − kpc − . The galaxies departfrom the canonical location of the K-S law in the right panel, mostlydue to the di ff erent time scale sampled by the SFR derived fromthe SSP analysis. As indicated before, the SSP analysis traces theSF in a longer period of time, (32 Myr and 4 Myr, respectively).Starbursts have typical time-scales of <
100 Myr (e.g., Leitherer2001). This means that using H α as calibrator to estimate the SFRwe only measure the recent SF ( ∼ Σ SFR . We have explored the ionized gas properties for all galaxies fromthe full CALIFA sample to investigate the presence of outflows inthe local universe. We imposed a set of criteria in the morphology,on the physical properties of the ionized gas and in the continuumto select a sample of candidate galaxies with a host outflow. The adopted criteria are: (i) highly inclined galaxies, (ii) detection ofextraplanar ionized gas, (iii) identification of an enhanced line ra-tios along the semi-minor axis, (iv) a biconical, bipolar or a sym-metric morphology in the extraplanar gas, (v) EW(H α ) > (cid:63) > .
5. Althoughin low mass galaxies outflows are less frequent, their local impactmight be stronger than in galaxies with higher potentials.The amount of outflows detected in the full CALIFA samplemay be a consequence of the short life-time of these processes. Thedynamical time scale of outflows in starburst galaxies and AGNdriven winds is in the range ∼ −
10 Myr (e.g., Veilleux et al.2005). CALIFA samples galaxies in the Local Universe, in a rangeof redshift between 0.005 and 0.03. This range translates into arange of time of ∆ t age = .
34 Gyr. This means that if the 17 out-flows detected in this sample are representative of the full sample,then it is expected on average one outflow every 20 Myr. So, it isstill possible that all galaxies in the sample have su ff ered an out-flow process in the past, but that these were not observed due totheir short life-time. The detection of outflows with much lower Σ SFR than anticipated and their random location along the SFMSmight also reflect the stochasticity of these processes.Fig. A1 shows that the vast majority of our outflow candidatespresent shock-excited emission lines in the extraplanar gas. This isquantified in Table 1, by the sharing between di ff erent ionizationsources of the extraplanar ionized spaxels: most of them are dom-inated by shock-like SF-driven winds (16 /
17) and only one is con-
MNRAS000
MNRAS000 , 1–16 (2018) C. L´opez-Cob´a et al. sistent to be an AGN-driven wind: NGC 4388. Indeed, this is theonly target which central ionization is clearly compatible with thepresence of an AGN and with a strong X-ray luminosity. From thisanalysis we conclude that most of our selected outflow candidatesare consistent with being driven by star-formation. However, thereis still the possibility that a galaxy hosts both SF and AGN activitycausing a mixing in the ionization (e.g., Davies et al. 2014), whichwill produce a complex distribution of points along the diagram. Inaddition, the high inclination might produce a strong nuclear ob-scuration and blur the signal of a possible AGN. This would a ff ectthe observed optical emission lines, locating them in the SF regionin the considered diagnostic diagrams. So far, we cannot reject nonof both possibilities. Indeed, we find two targets, IC 2247 and NGC4676A, with a fraction of ∼
30% of their extraplanar ionization be-ing compatible with ionization by an AGN based on our criteria.The former one has a mixed / composite ionization in the central re-gion (AGN / SF), while the later has clear X-ray emission although itis not considered to host an AGN (Wild et al. 2014, , and referencestherein).We found that the global SFRs of the outflow candidates putsall of them along the active star formation sequence, and that thereis no significant excess in the SFR. This is contrary to expecta-tion if this parameter was the major driver for the presence of out-flows. Nevertheless, when we explore the spatial concentration ofthe SFR, we observe that on average, the candidates do present anexcess in their Σ SFR , when compared with galaxies with SF activitybut without evidence of outflows. This excess in Σ SFR is statisticallysignificant for the innermost regions and it holds up to ∼ e ( ∼ . − . pc (e.g., Lehnert & Heckman 1996),where the outflows have a significant signature in the properties ofgalaxies. For R > e the outflow candidates behave as normalSF galaxies and within this region the Σ SFR distribution steepens.Although on average the candidates do not seem to surpass theproposed canonical threshold for the star formation surface density Σ SFR > − M (cid:12) yr − kpc − , when we analyse the individual valuesof this parameter we observe that all of them lie close to this canon-ical value. Depending on whether it is used the H α or the SSPs cal-ibration for the SFR, they can surpass the canonical value only ina few cases. In starburst galaxies, the IR luminosity ( L FIR ) is usedas a tracer of the SFR. Indeed, the calibrator used to estimate thecanonical threshold in starburst galaxies adopts the L FIR that tracesthe dust heating due to stars of 10–100 Myr (e.g., Kennicutt 1998).Due to fact that the SFR
SSP comprises larger periods of SF, com-pared with H α , the SFR SSP might be used as a better estimator of theSFR in outflows. The sampled time by this calibrator approaches tothe dust re-emission time scales.Our results suggest that the threshold limit in Σ SFR might bemore flexible and include galaxies with lower values, in a regimewhere normal SF spiral galaxies dominates, rather than extremestarbursts. If we go back to the initial studies in outflows, we seethat this threshold is achieved only for starburst and high-z Lymanbreak galaxies, and not for normal disk galaxies which can presentvalues of the star-formation density as low as 10 − − Σ SFR , threshold (e.g., Kennicutt 1998; Heckman 2001; Kennicutt et al. 2007). Thefact that outflows are ubiquitous in galaxies that exceed the pro-posed limit does not exclude the possibility to find outflows ingalaxies with Σ SFR < Σ SFR , threshold . Indeed, more recent studies havealso pointed that this threshold is quite high for the bulk populationof outflows (e.g., Ho et al. 2016).We have also shown that not only the Σ SFR is a key parameter to generate outflows, but it must be accompanied with large densi-ties of molecular gas (i.e., Σ gas ). We propose a region for galaxieshosting outflows in the K-S diagram: Σ SFR > − M (cid:12) yr − kpc − with Σ gas > . M (cid:12) pc − , in a central region of ∼ i λλ , The main conclusions of the exploration of the presence of outflowsin the complete sample of galaxies observed by the CALIFA surveyare the following ones: • The fraction of galaxies with clear evidence of outflows rangebetween 2% to 8%, depending if we consider the full sample ofgalaxies with any possible evidence or those ones that fulfil all ourselection criteria. • The properties of galaxies hosting outflows are similar to thatof the non hosting ones in terms of their distribution along theCMD, mass, morphology and integrated SFR, when the compar-ison is restricted to galaxies of the same inclination. • Galaxies hosting outflows are distributed in a high mass rangeof 9 . < log M (cid:63) < • Most of our outflow candidates are compatible with beingdriven by star-formation, based both on the dominant ionization inthe central regions and their location in the diagnostic diagramsin comparison with demarcation described by Sharp & Bland-Hawthorn (2010). Only in one case we see clear evidence of AGN-driven outflows (NGC 4388). • The highly-inclined galaxies hosting an outflow present a sig-nificant excess in the star-formation rate surface density in the cen-tral regions (R < e ), when compared with the non hosting out-flow ones, indicating that at least in these galaxies, outflows aremostly driven by a central increase in the SFR. • The galaxies hosting outflows in the CALIFA sample onlymarginally exceed the canonical threshold on the Σ SFR , maybe be-cause they present regular star-formation which yields lower values
MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. in the star-formation surface density, and therefore produce weakeroutflows compared to those of starburst galaxies.Our results indicate that outflows are less restricted to extremestar-formation events, either central or integrated, being more fre-quent events than anticipated. Further studies are needed to explorethe outflows in galaxies with lower inclinations, where data withbetter spatial and spectral resolutions could break the confusion be-tween the di ff erent ionization components (e.g., L´opez-Cob´a et al.2017), and over much larger samples, like the ones provided bythe MaNGA survey (e.g., Bundy et al. 2015), to provide with betterstatistics. Even more, we need to explore in more detail the physicalproperties of the outflows themselves, only outlined in the currentstudy, and focus on the detectability of these events in retired / early-type galaxies, mostly excluded in this analysis due the imposed in-clination selection. CLC and SFS are grateful for the support of a CONACYT (Mex-ico) grant CB-285080, and funding from the PAPIIT-DGAPA-IA101217(UNAM), PAPIIT: IN103318 and CONACYT: 168251projects. ICG, SFS and CLC acknowledge support from DGAPA-UNAM (Mexico) grant IN11341. CLC acknowledges CONACYT(Mexico) Ph. D. scholarship. JBH acknowledges the support of anARC Laureate Fellowship from the Australian Government. LGwas supported in part by the US National Science Foundation underGrant AST-1311862.This study uses data provided by the Calar Alto Legacy Inte-gral Field Area (CALIFA) survey (http: // califa.caha.es / ).CALIFA is the first legacy survey performed at Calar Alto.The CALIFA collaboration would like to thank the IAA-CSIC andMPIA-MPG as major partners of the observatory, and CAHA it-self, for the unique access to telescope time and support in man-power and infrastructures. The CALIFA collaboration also thanksthe CAHA sta ff for the dedication to this project.Based on observations collected at the Centro Astron´omicoHispano Alem´an (CAHA) at Calar Alto, operated jointly by theMax-Planck-Institut f¨ur Astronomie and the Instituto de Astrof´ısicade Andaluc´ıa (CSIC). REFERENCES
Aguirre A., Hernquist L., Schaye J., Katz N., Weinberg D. H., Gardner J.,2001, ApJ, 561, 521Alatalo K., et al., 2016, ApJS, 224, 38Allen M. G., Groves B. A., Dopita M. A., Sutherland R. S., Kewley L. J.,2008, ApJS, 178, 20Allen J. T., et al., 2015, MNRAS, 446, 1567Antonucci R. R. J., Ulvestad J. S., 1985, ApJ, 294, 158Axon D. J., Taylor K., 1978, Nature, 274, 37Baldwin J. A., Phillips M. M., Terlevich R., 1981, PASP, 93, 5Barrera-Ballesteros J. K., et al., 2018, ApJ, 852, 74Belfiore F., et al., 2016, MNRAS, 461, 3111Binette L., Magris C. G., Stasi´nska G., Bruzual A. G., 1994, A&A, 292, 13Bland J., Tully B., 1988, Nature, 334, 43Bland-Hawthorn J., Cohen M., 2003, ApJ, 582, 246Bland-Hawthorn J., Sutherland R., Webster D., 2015, ApJ, 807, 154Bolatto A. D., Wolfire M., Leroy A. K., 2013, ARA&A, 51, 207Brightman M., Nandra K., 2011, MNRAS, 413, 1206Brinchmann J., Charlot S., White S. D. M., Tremonti C., Kau ff mann G.,Heckman T., Brinkmann J., 2004, MNRAS, 351, 1151Bundy K., et al., 2015, ApJ, 798, 7 Cano-D´ıaz M., et al., 2016, ApJ, 821, L26Cardelli J. A., Clayton G. C., Mathis J. S., 1989, ApJ, 345, 245Catal´an-Torrecilla C., et al., 2015, A&A, 584, A87Cecil G., Bland-Hawthorn J., Veilleux S., Filippenko A. V., 2001, ApJ, 555,338Celotti A., Ghisellini G., 2008, MNRAS, 385, 283Chabrier G., 2003, PASP, 115, 763Chen Y.-M., Tremonti C. A., Heckman T. M., Kau ff mann G., Weiner B. J.,Brinchmann J., Wang J., 2010, AJ, 140, 445Cheung E., et al., 2016, Nature, 533, 504Cid Fernandes R., Stasi´nska G., Mateus A., Vale Asari N., 2011, MNRAS,413, 1687Cid Fernandes R., et al., 2013a, A&A, 557, A86Cid Fernandes R., et al., 2013b, A&A, 557, A86Cid Fernandes R., et al., 2014, A&A, 561, A130Coil A. L., Weiner B. J., Holz D. E., Cooper M. C., Yan R., Aird J., 2011,ApJ, 743, 46Cooper J. L., Bicknell G. V., Sutherland R. S., Bland-Hawthorn J., 2008,ApJ, 674, 157Corral A., et al., 2014, A&A, 569, A71Davies R. L., Kewley L. J., Ho I.-T., Dopita M. A., 2014, MNRAS, 444,3961Dopita M. A., Sutherland R. S., 1995, ApJ, 455, 468Dopita M. A., Sutherland R. S., 1996, ApJS, 102, 161Esquej P., et al., 2014, ApJ, 780, 86Falc´on-Barroso J., S´anchez-Bl´azquez P., Vazdekis A., Ricciardelli E.,Cardiel N., Cenarro A. J., Gorgas J., Peletier R. F., 2011, A&A, 532,A95Fasano G., Franceschini A., 1987, MNRAS, 225, 155Feigelson E. D., Babu G. J., 2012, Modern Statistical Methods for Astron-omyFerguson A. M. N., Wyse R. F. G., Gallagher III J. S., Hunter D. A., 1996,AJ, 111, 2265Flores-Fajardo N., Morisset C., Stasi´nska G., Binette L., 2011, MNRAS,415, 2182Fogarty L. M. R., et al., 2012, ApJ, 761, 169Fox A. J., et al., 2015, ApJ, 799, L7Franx M., Illingworth G. D., Kelson D. D., van Dokkum P. G., Tran K.-V.,1997, ApJ, 486, L75Galbany L., et al., 2016, MNRAS, 455, 4087Galbany L., et al., 2017, MNRAS, 468, 628Galbany L., et al., 2018, ApJ, 855, 107Garc´ıa-Benito R., et al., 2015, A&A, 576, A135Garc´ıa-Benito R., et al., 2017, A&A, 608, A27Genzel R., et al., 2014, ApJ, 796, 7Gonz´alez Delgado R. M., et al., 2015, A&A, 581, A103Gonz´alez Delgado R. M., et al., 2017, A&A, 607, A128Gonz´alez-Mart´ın O., Masegosa J., M´arquez I., Guainazzi M., Jim´enez-Bail´on E., 2009, A&A, 506, 1107Heckman T. M., 1980, A&A, 87, 152Heckman T. M., 2001, in Hibbard J. E., Rupen M., van Gorkom J. H., eds,Astronomical Society of the Pacific Conference Series Vol. 240, Gasand Galaxy Evolution. p. 345 ( arXiv:astro-ph/0009075 )Heckman T. M., 2002, in Mulchaey J. S., Stocke J. T., eds, AstronomicalSociety of the Pacific Conference Series Vol. 254, Extragalactic Gas atLow Redshift. p. 292 ( arXiv:astro-ph/0107438 )Heckman T. M., Armus L., Miley G. K., 1990, ApJS, 74, 833Heckman T. M., Lehnert M. D., Strickland D. K., Armus L., 2000, ApJS,129, 493Ho I.-T., et al., 2014, MNRAS, 444, 3894Ho I.-T., et al., 2016, MNRAS, 457, 1257Hoopes C. G., Walterbos R. A. M., 2003, ApJ, 586, 902Hopkins P. F., Quataert E., 2010, MNRAS, 407, 1529Husemann B., et al., 2013, A&A, 549, A87Jones A., et al., 2017, A&A, 599, A141Kau ff mann G., et al., 2003, MNRAS, 346, 1055Kehrig C., et al., 2012, A&A, 540, A11Kelz A., et al., 2006, PASP, 118, 129MNRAS000
Aguirre A., Hernquist L., Schaye J., Katz N., Weinberg D. H., Gardner J.,2001, ApJ, 561, 521Alatalo K., et al., 2016, ApJS, 224, 38Allen M. G., Groves B. A., Dopita M. A., Sutherland R. S., Kewley L. J.,2008, ApJS, 178, 20Allen J. T., et al., 2015, MNRAS, 446, 1567Antonucci R. R. J., Ulvestad J. S., 1985, ApJ, 294, 158Axon D. J., Taylor K., 1978, Nature, 274, 37Baldwin J. A., Phillips M. M., Terlevich R., 1981, PASP, 93, 5Barrera-Ballesteros J. K., et al., 2018, ApJ, 852, 74Belfiore F., et al., 2016, MNRAS, 461, 3111Binette L., Magris C. G., Stasi´nska G., Bruzual A. G., 1994, A&A, 292, 13Bland J., Tully B., 1988, Nature, 334, 43Bland-Hawthorn J., Cohen M., 2003, ApJ, 582, 246Bland-Hawthorn J., Sutherland R., Webster D., 2015, ApJ, 807, 154Bolatto A. D., Wolfire M., Leroy A. K., 2013, ARA&A, 51, 207Brightman M., Nandra K., 2011, MNRAS, 413, 1206Brinchmann J., Charlot S., White S. D. M., Tremonti C., Kau ff mann G.,Heckman T., Brinkmann J., 2004, MNRAS, 351, 1151Bundy K., et al., 2015, ApJ, 798, 7 Cano-D´ıaz M., et al., 2016, ApJ, 821, L26Cardelli J. A., Clayton G. C., Mathis J. S., 1989, ApJ, 345, 245Catal´an-Torrecilla C., et al., 2015, A&A, 584, A87Cecil G., Bland-Hawthorn J., Veilleux S., Filippenko A. V., 2001, ApJ, 555,338Celotti A., Ghisellini G., 2008, MNRAS, 385, 283Chabrier G., 2003, PASP, 115, 763Chen Y.-M., Tremonti C. A., Heckman T. M., Kau ff mann G., Weiner B. J.,Brinchmann J., Wang J., 2010, AJ, 140, 445Cheung E., et al., 2016, Nature, 533, 504Cid Fernandes R., Stasi´nska G., Mateus A., Vale Asari N., 2011, MNRAS,413, 1687Cid Fernandes R., et al., 2013a, A&A, 557, A86Cid Fernandes R., et al., 2013b, A&A, 557, A86Cid Fernandes R., et al., 2014, A&A, 561, A130Coil A. L., Weiner B. J., Holz D. E., Cooper M. C., Yan R., Aird J., 2011,ApJ, 743, 46Cooper J. L., Bicknell G. V., Sutherland R. S., Bland-Hawthorn J., 2008,ApJ, 674, 157Corral A., et al., 2014, A&A, 569, A71Davies R. L., Kewley L. J., Ho I.-T., Dopita M. A., 2014, MNRAS, 444,3961Dopita M. A., Sutherland R. S., 1995, ApJ, 455, 468Dopita M. A., Sutherland R. S., 1996, ApJS, 102, 161Esquej P., et al., 2014, ApJ, 780, 86Falc´on-Barroso J., S´anchez-Bl´azquez P., Vazdekis A., Ricciardelli E.,Cardiel N., Cenarro A. J., Gorgas J., Peletier R. F., 2011, A&A, 532,A95Fasano G., Franceschini A., 1987, MNRAS, 225, 155Feigelson E. D., Babu G. J., 2012, Modern Statistical Methods for Astron-omyFerguson A. M. N., Wyse R. F. G., Gallagher III J. S., Hunter D. A., 1996,AJ, 111, 2265Flores-Fajardo N., Morisset C., Stasi´nska G., Binette L., 2011, MNRAS,415, 2182Fogarty L. M. R., et al., 2012, ApJ, 761, 169Fox A. J., et al., 2015, ApJ, 799, L7Franx M., Illingworth G. D., Kelson D. D., van Dokkum P. G., Tran K.-V.,1997, ApJ, 486, L75Galbany L., et al., 2016, MNRAS, 455, 4087Galbany L., et al., 2017, MNRAS, 468, 628Galbany L., et al., 2018, ApJ, 855, 107Garc´ıa-Benito R., et al., 2015, A&A, 576, A135Garc´ıa-Benito R., et al., 2017, A&A, 608, A27Genzel R., et al., 2014, ApJ, 796, 7Gonz´alez Delgado R. M., et al., 2015, A&A, 581, A103Gonz´alez Delgado R. M., et al., 2017, A&A, 607, A128Gonz´alez-Mart´ın O., Masegosa J., M´arquez I., Guainazzi M., Jim´enez-Bail´on E., 2009, A&A, 506, 1107Heckman T. M., 1980, A&A, 87, 152Heckman T. M., 2001, in Hibbard J. E., Rupen M., van Gorkom J. H., eds,Astronomical Society of the Pacific Conference Series Vol. 240, Gasand Galaxy Evolution. p. 345 ( arXiv:astro-ph/0009075 )Heckman T. M., 2002, in Mulchaey J. S., Stocke J. T., eds, AstronomicalSociety of the Pacific Conference Series Vol. 254, Extragalactic Gas atLow Redshift. p. 292 ( arXiv:astro-ph/0107438 )Heckman T. M., Armus L., Miley G. K., 1990, ApJS, 74, 833Heckman T. M., Lehnert M. D., Strickland D. K., Armus L., 2000, ApJS,129, 493Ho I.-T., et al., 2014, MNRAS, 444, 3894Ho I.-T., et al., 2016, MNRAS, 457, 1257Hoopes C. G., Walterbos R. A. M., 2003, ApJ, 586, 902Hopkins P. F., Quataert E., 2010, MNRAS, 407, 1529Husemann B., et al., 2013, A&A, 549, A87Jones A., et al., 2017, A&A, 599, A141Kau ff mann G., et al., 2003, MNRAS, 346, 1055Kehrig C., et al., 2012, A&A, 540, A11Kelz A., et al., 2006, PASP, 118, 129MNRAS000 , 1–16 (2018) C. L´opez-Cob´a et al.
Kennicutt Jr. R. C., 1998, ApJ, 498, 541Kennicutt Jr. R. C., Keel W. C., Blaha C. A., 1989, AJ, 97, 1022Kennicutt Jr. R. C., et al., 2007, ApJ, 671, 333Kewley L. J., Dopita M. A., Sutherland R. S., Heisler C. A., Trevena J.,2001, ApJ, 556, 121Kroupa P., 2001, MNRAS, 322, 231Lacerda E. A. D., et al., 2018, MNRAS, 474, 3727Lehnert M. D., Heckman T. M., 1996, ApJ, 472, 546Leitherer C., 2001, in von Hippel T., Simpson C., Manset N., eds, Astro-nomical Society of the Pacific Conference Series Vol. 245, Astrophysi-cal Ages and Times Scales. p. 390L´opez-Cob´a C., et al., 2017, MNRAS, 467, 4951Lundgren B. F., et al., 2012, ApJ, 760, 49Maiolino R., et al., 2017, Nature, 544, 202Martel H., 2011, preprint, ( arXiv:1110.1061 )Martin C. L., 1998, ApJ, 506, 222Martin C. L., Kobulnicky H. A., Heckman T. M., 2002, ApJ, 574, 663Martin C. L., Dijkstra M., Henry A., Soto K. T., Danforth C. W., Wong J.,2015, ApJ, 803, 6Monreal-Ibero A., Arribas S., Colina L., Rodr´ıguez-Zaur´ın J., Alonso-Herrero A., Garc´ıa-Mar´ın M., 2010, A&A, 517, A28Noeske K. G., et al., 2007, ApJ, 660, L43Osterbrock D. E., 1989, Astrophysics of gaseous nebulae and active galacticnucleiPeacock J. A., 1983, MNRAS, 202, 615Pettini M., Kellogg M., Steidel C. C., Dickinson M., Adelberger K. L., Gi-avalisco M., 1998, ApJ, 508, 539Press W. H., Teukolsky S. A., Vetterling W. T., Flannery B. P., 1992, Nu-merical Recipes in C (2Nd Ed.): The Art of Scientific Computing. Cam-bridge University Press, New York, NY, USAPrieto J. L., et al., 2016, ApJ, 830, L32Rich J. A., Kewley L. J., Dopita M. A., 2011, ApJ, 734, 87Rich J. A., Kewley L. J., Dopita M. A., 2015, ApJS, 221, 28Roth M. M., et al., 2005, PASP, 117, 620Rupke D. S., Veilleux S., Sanders D. B., 2005a, ApJS, 160, 87Rupke D. S., Veilleux S., Sanders D. B., 2005b, ApJS, 160, 115Salim S., et al., 2007, ApJS, 173, 267Salpeter E. E., 1955, ApJ, 121, 161S´anchez-Bl´azquez P., et al., 2006, MNRAS, 371, 703S´anchez S. F., et al., 2012, A&A, 538, A8S´anchez S. F., et al., 2013, A&A, 554, A58S´anchez S. F., et al., 2014, A&A, 563, A49S´anchez S. F., et al., 2016a, Rev. Mex. Astron. Astrofis., 52, 21S´anchez S. F., et al., 2016b, Rev. Mex. Astron. Astrofis., 52, 171S´anchez S. F., et al., 2016c, A&A, 594, A36S´anchez S. F., et al., 2017, MNRAS, 469, 2121S´anchez S. F., et al., 2018, Rev. Mex. Astron. Astrofis., 54, 217Scannapieco C., Tissera P. B., White S. D. M., Springel V., 2006, MNRAS,371, 1125Scarpa R., Urry C. M., Falomo R., Pesce J. E., Treves A., 2000, ApJ, 532,740Shalyapina L. V., Moiseev A. V., Yakovleva V. A., Hagen-Thorn V. A., Bu-renkov A. N., 2004, Astronomy Letters, 30, 1Sharp R. G., Bland-Hawthorn J., 2010, ApJ, 711, 818Silk J., 2013, ApJ, 772, 112Singh R., et al., 2013, A&A, 558, A43Springel V., Hernquist L., 2003, MNRAS, 339, 289Stasi´nska G., Cid Fernandes R., Mateus A., Sodr´e L., Asari N. V., 2006,MNRAS, 371, 972Stasi´nska G., et al., 2008, MNRAS, 391, L29Su M., Slatyer T. R., Finkbeiner D. P., 2010, ApJ, 724, 1044Tanner R., Cecil G., Heitsch F., 2016, ApJ, 821, 7Tanner R., Cecil G., Heitsch F., 2017, ApJ, 843, 137Tremonti C. A., et al., 2004, ApJ, 613, 898Vazdekis A., S´anchez-Bl´azquez P., Falc´on-Barroso J., Cenarro A. J.,Beasley M. A., Cardiel N., Gorgas J., Peletier R. F., 2010, MNRAS,404, 1639Veilleux S., Osterbrock D. E., 1987, ApJS, 63, 295 Veilleux S., Rupke D. S., 2002, ApJ, 565, L63Veilleux S., Shopbell P. L., Miller S. T., 2001, AJ, 121, 198Veilleux S., Cecil G., Bland-Hawthorn J., 2005, ARA&A, 43, 769Walcher C. J., et al., 2014, A&A, 569, A1Wild V., et al., 2014, A&A, 567, A132Wood K., Hill A. S., Joung M. R., Mac Low M.-M., Benjamin R. A.,Ha ff ner L. M., Reynolds R. J., Madsen G. J., 2010, ApJ, 721, 1397Wood C. M., Tremonti C. A., Calzetti D., Leitherer C., Chisholm J., Gal-lagher J. S., 2015, MNRAS, 452, 2712Zhang K., et al., 2017, MNRAS, 466, 3217 MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. T a b l e . M a i np r op e r ti e s o f t h e g a l a x i e s ca nd i d a t e s t oho s t ou t fl o w s . NA S A / I P A C E x t r a g a l ac ti c D a t a b a s e . H yp e r L e d a . E s ti m a t e d fr o m a n i s opho t a l a n a l y s i s on t h e S D SS r- b a nd i m a g e s a s d e s c r i b e d i n W a l c h e r e t a l . ( ) . E s ti m a t e d fr o m t h e SSPfi tti ng a n a l y s i s . E s ti m a t e dov e r a n a r ea o f (cid:48)(cid:48) × (cid:48)(cid:48) a r ound t h e nu c l ea rr e g i on . A cc o r d i ng t o t h e K e w l e y e t a l . ( ) d e m a r ca ti on . E s ti m a t e d a t . R e . T h e i n t e r ac ti on i s r e f e r ee d a s i f t h e r ea r ec l o s e r c o m p a n i on s a tt h e s a m e r e d s h i f ti n t h e S D SS i m a g e s o r i f p r e s e n t e v i d e n ce o f i n t e r ac ti on . E x c it a ti on m ec h a n i s m f o r t h e ob s e r v e dou t fl o w . T h e fr ac ti on s r e p r e s e n tt h ec on t r i bu ti on i n t h e i on i za ti ono f t h ee x t r a p l a n a r g a s ( s p a x e l s l y i ngb e yond5 (cid:48)(cid:48) fr o m t h e d i s k ) by a n AGN , s ho c k s o r D I G . T h e s e fr ac ti on s t a k e s i n t o acc oun t s i m u lt a n e ou s l y a llt h e po i n t s l y i ng a bov e t h e K e w l e y e t a l . ( ) c u r v e s i n eac hd i a gno s ti c d i a g r a m s . T h e s ho c k a nd AGN fr ac ti on s w a s e s ti m a t e dby t h ea m oun t o f s p a x e l s w it h E W ( H α ) > Å l y i ng a tt h e r i gh t - a nd l e f t - s i d e r e s p ec ti v e l yo f t h e s ho c k / AGN e x c it a ti onb i s ec t o r li n e s fr o m S h a r p & B l a nd - H a w t ho r n ( ) , i n t h e t h r ee d i a gno s ti c d i a g r a m s . T h e D I G fr ac ti on w a s e s ti m a t e d fr o m t h ea m oun t s p a x e l s i n t h ee x t r a p l a n a rr e g i on w it h E W ( H α ) < Å i n t h e t h r ee d i a gno s ti c d i a g r a m s . O b j ec t z ( ) H ubb l e ( ) i ( ) P A ( ) R ( ) e l og M ( ) (cid:63) l og SF R ( ) N u c l ea r ( , ) Σ ( ) H α Σ ( ) SSP Σ ( ) g a s I n t e r ac ti ng ( ) AGN : S ho c k : D I G ( ) T yp e [ ◦ ][ ◦ ][ kp c ][ M (cid:12) ][ M (cid:12) y r − ] i on i za ti on [ M (cid:12) , y r − p c − ][ M (cid:12) y r − p c − ][ M (cid:12) , p c − ]( Y / N ) % I C . S c d79 . . . . . SF - . - . . N : : I C . S a b72 . . . . . AGN / SF - . - . . N : : M C G - - - . S c d76 . - . . . . SF - . - . . N : : NG C A . S d m . - . . . . SF - . - . . Y : : NG C . S d71 . - . . . - . SF - . - . . N : : NG C . S c . . . . - . SF - . - . . N : : UG C . S b72 . - . . . - . SF - . - . . N : : UG C . S a . - . . . . SF - . - . . N : : UG C . S a b77 . - . . . . SF - . - . . N : : UG C . S b79 . . . . . SF - . - . . N : : I C . S c . . . . - . SF - . - . . N : : NG C B . S c . - . . . . SF - . - . . Y : : UG C . S c . . . . - . SF - . - . . N : : UG C . S a b82 . . . . - . SF - . - . . N : : NG C . S B b66 . - . . . . AGN - . - . . N : : NG C . S b75 . - . . . . SF - . - . . Y : : M C G + - - . S a b54 . . . . . SF - . - . . N : : MNRAS000
Kennicutt Jr. R. C., 1998, ApJ, 498, 541Kennicutt Jr. R. C., Keel W. C., Blaha C. A., 1989, AJ, 97, 1022Kennicutt Jr. R. C., et al., 2007, ApJ, 671, 333Kewley L. J., Dopita M. A., Sutherland R. S., Heisler C. A., Trevena J.,2001, ApJ, 556, 121Kroupa P., 2001, MNRAS, 322, 231Lacerda E. A. D., et al., 2018, MNRAS, 474, 3727Lehnert M. D., Heckman T. M., 1996, ApJ, 472, 546Leitherer C., 2001, in von Hippel T., Simpson C., Manset N., eds, Astro-nomical Society of the Pacific Conference Series Vol. 245, Astrophysi-cal Ages and Times Scales. p. 390L´opez-Cob´a C., et al., 2017, MNRAS, 467, 4951Lundgren B. F., et al., 2012, ApJ, 760, 49Maiolino R., et al., 2017, Nature, 544, 202Martel H., 2011, preprint, ( arXiv:1110.1061 )Martin C. L., 1998, ApJ, 506, 222Martin C. L., Kobulnicky H. A., Heckman T. M., 2002, ApJ, 574, 663Martin C. L., Dijkstra M., Henry A., Soto K. T., Danforth C. W., Wong J.,2015, ApJ, 803, 6Monreal-Ibero A., Arribas S., Colina L., Rodr´ıguez-Zaur´ın J., Alonso-Herrero A., Garc´ıa-Mar´ın M., 2010, A&A, 517, A28Noeske K. G., et al., 2007, ApJ, 660, L43Osterbrock D. E., 1989, Astrophysics of gaseous nebulae and active galacticnucleiPeacock J. A., 1983, MNRAS, 202, 615Pettini M., Kellogg M., Steidel C. C., Dickinson M., Adelberger K. L., Gi-avalisco M., 1998, ApJ, 508, 539Press W. H., Teukolsky S. A., Vetterling W. T., Flannery B. P., 1992, Nu-merical Recipes in C (2Nd Ed.): The Art of Scientific Computing. Cam-bridge University Press, New York, NY, USAPrieto J. L., et al., 2016, ApJ, 830, L32Rich J. A., Kewley L. J., Dopita M. A., 2011, ApJ, 734, 87Rich J. A., Kewley L. J., Dopita M. A., 2015, ApJS, 221, 28Roth M. M., et al., 2005, PASP, 117, 620Rupke D. S., Veilleux S., Sanders D. B., 2005a, ApJS, 160, 87Rupke D. S., Veilleux S., Sanders D. B., 2005b, ApJS, 160, 115Salim S., et al., 2007, ApJS, 173, 267Salpeter E. E., 1955, ApJ, 121, 161S´anchez-Bl´azquez P., et al., 2006, MNRAS, 371, 703S´anchez S. F., et al., 2012, A&A, 538, A8S´anchez S. F., et al., 2013, A&A, 554, A58S´anchez S. F., et al., 2014, A&A, 563, A49S´anchez S. F., et al., 2016a, Rev. Mex. Astron. Astrofis., 52, 21S´anchez S. F., et al., 2016b, Rev. Mex. Astron. Astrofis., 52, 171S´anchez S. F., et al., 2016c, A&A, 594, A36S´anchez S. F., et al., 2017, MNRAS, 469, 2121S´anchez S. F., et al., 2018, Rev. Mex. Astron. Astrofis., 54, 217Scannapieco C., Tissera P. B., White S. D. M., Springel V., 2006, MNRAS,371, 1125Scarpa R., Urry C. M., Falomo R., Pesce J. E., Treves A., 2000, ApJ, 532,740Shalyapina L. V., Moiseev A. V., Yakovleva V. A., Hagen-Thorn V. A., Bu-renkov A. N., 2004, Astronomy Letters, 30, 1Sharp R. G., Bland-Hawthorn J., 2010, ApJ, 711, 818Silk J., 2013, ApJ, 772, 112Singh R., et al., 2013, A&A, 558, A43Springel V., Hernquist L., 2003, MNRAS, 339, 289Stasi´nska G., Cid Fernandes R., Mateus A., Sodr´e L., Asari N. V., 2006,MNRAS, 371, 972Stasi´nska G., et al., 2008, MNRAS, 391, L29Su M., Slatyer T. R., Finkbeiner D. P., 2010, ApJ, 724, 1044Tanner R., Cecil G., Heitsch F., 2016, ApJ, 821, 7Tanner R., Cecil G., Heitsch F., 2017, ApJ, 843, 137Tremonti C. A., et al., 2004, ApJ, 613, 898Vazdekis A., S´anchez-Bl´azquez P., Falc´on-Barroso J., Cenarro A. J.,Beasley M. A., Cardiel N., Gorgas J., Peletier R. F., 2010, MNRAS,404, 1639Veilleux S., Osterbrock D. E., 1987, ApJS, 63, 295 Veilleux S., Rupke D. S., 2002, ApJ, 565, L63Veilleux S., Shopbell P. L., Miller S. T., 2001, AJ, 121, 198Veilleux S., Cecil G., Bland-Hawthorn J., 2005, ARA&A, 43, 769Walcher C. J., et al., 2014, A&A, 569, A1Wild V., et al., 2014, A&A, 567, A132Wood K., Hill A. S., Joung M. R., Mac Low M.-M., Benjamin R. A.,Ha ff ner L. M., Reynolds R. J., Madsen G. J., 2010, ApJ, 721, 1397Wood C. M., Tremonti C. A., Calzetti D., Leitherer C., Chisholm J., Gal-lagher J. S., 2015, MNRAS, 452, 2712Zhang K., et al., 2017, MNRAS, 466, 3217 MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. T a b l e . M a i np r op e r ti e s o f t h e g a l a x i e s ca nd i d a t e s t oho s t ou t fl o w s . NA S A / I P A C E x t r a g a l ac ti c D a t a b a s e . H yp e r L e d a . E s ti m a t e d fr o m a n i s opho t a l a n a l y s i s on t h e S D SS r- b a nd i m a g e s a s d e s c r i b e d i n W a l c h e r e t a l . ( ) . E s ti m a t e d fr o m t h e SSPfi tti ng a n a l y s i s . E s ti m a t e dov e r a n a r ea o f (cid:48)(cid:48) × (cid:48)(cid:48) a r ound t h e nu c l ea rr e g i on . A cc o r d i ng t o t h e K e w l e y e t a l . ( ) d e m a r ca ti on . E s ti m a t e d a t . R e . T h e i n t e r ac ti on i s r e f e r ee d a s i f t h e r ea r ec l o s e r c o m p a n i on s a tt h e s a m e r e d s h i f ti n t h e S D SS i m a g e s o r i f p r e s e n t e v i d e n ce o f i n t e r ac ti on . E x c it a ti on m ec h a n i s m f o r t h e ob s e r v e dou t fl o w . T h e fr ac ti on s r e p r e s e n tt h ec on t r i bu ti on i n t h e i on i za ti ono f t h ee x t r a p l a n a r g a s ( s p a x e l s l y i ngb e yond5 (cid:48)(cid:48) fr o m t h e d i s k ) by a n AGN , s ho c k s o r D I G . T h e s e fr ac ti on s t a k e s i n t o acc oun t s i m u lt a n e ou s l y a llt h e po i n t s l y i ng a bov e t h e K e w l e y e t a l . ( ) c u r v e s i n eac hd i a gno s ti c d i a g r a m s . T h e s ho c k a nd AGN fr ac ti on s w a s e s ti m a t e dby t h ea m oun t o f s p a x e l s w it h E W ( H α ) > Å l y i ng a tt h e r i gh t - a nd l e f t - s i d e r e s p ec ti v e l yo f t h e s ho c k / AGN e x c it a ti onb i s ec t o r li n e s fr o m S h a r p & B l a nd - H a w t ho r n ( ) , i n t h e t h r ee d i a gno s ti c d i a g r a m s . T h e D I G fr ac ti on w a s e s ti m a t e d fr o m t h ea m oun t s p a x e l s i n t h ee x t r a p l a n a rr e g i on w it h E W ( H α ) < Å i n t h e t h r ee d i a gno s ti c d i a g r a m s . O b j ec t z ( ) H ubb l e ( ) i ( ) P A ( ) R ( ) e l og M ( ) (cid:63) l og SF R ( ) N u c l ea r ( , ) Σ ( ) H α Σ ( ) SSP Σ ( ) g a s I n t e r ac ti ng ( ) AGN : S ho c k : D I G ( ) T yp e [ ◦ ][ ◦ ][ kp c ][ M (cid:12) ][ M (cid:12) y r − ] i on i za ti on [ M (cid:12) , y r − p c − ][ M (cid:12) y r − p c − ][ M (cid:12) , p c − ]( Y / N ) % I C . S c d79 . . . . . SF - . - . . N : : I C . S a b72 . . . . . AGN / SF - . - . . N : : M C G - - - . S c d76 . - . . . . SF - . - . . N : : NG C A . S d m . - . . . . SF - . - . . Y : : NG C . S d71 . - . . . - . SF - . - . . N : : NG C . S c . . . . - . SF - . - . . N : : UG C . S b72 . - . . . - . SF - . - . . N : : UG C . S a . - . . . . SF - . - . . N : : UG C . S a b77 . - . . . . SF - . - . . N : : UG C . S b79 . . . . . SF - . - . . N : : I C . S c . . . . - . SF - . - . . N : : NG C B . S c . - . . . . SF - . - . . Y : : UG C . S c . . . . - . SF - . - . . N : : UG C . S a b82 . . . . - . SF - . - . . N : : NG C . S B b66 . - . . . . AGN - . - . . N : : NG C . S b75 . - . . . . SF - . - . . Y : : M C G + - - . S a b54 . . . . . SF - . - . . N : : MNRAS000 , 1–16 (2018) C. L´opez-Cob´a et al.
APPENDIX A: CANDIDATES GALAXIES
In Fig. A1 it is shown the spatially resolved line ratio maps anddiagnostic diagrams for all the candidate galaxies with a host out-flow listed in Table 1 comprising the same information shown forNGC 6282 in Figs. 1 and 2.
APPENDIX B: GALAXIES NOT CATALOGUED ASOUTFLOWS
In Table B1 we present the remaining galaxies with detected extra-planar ionized gas or some increase in the line ratios, but that werenot classified as outflow candidates. Some of these galaxies presentextraplanar di ff use ionized gas (eDIG). In some cases the eDIG isdominated by HOLMES or post-AGB. In other cases the extrapla-nar gas presents EW(H α ) > ii regions that scape from the disk.In galaxies with high star-formation rates, or starburst galaxies, afraction of the ionizing photons can scape from the H ii regionswithout been absorbed. It has been suggested that this leaky H ii regions photons may scape to the di ff use ISM and the inter galacticmedium and ionize regions of kilo parsec scales from the disk (e.g.,Ferguson et al. 1996; Hoopes & Walterbos 2003; Wood et al. 2010;Martin et al. 2015).It is important to emphasize that this work is not focused in theexploration of eDIG in general. There are other studies in the CAL-IFA survey that focuses in the analysis of DIG in all type of galaxiesregardles of their inclination (e.g., Singh et al. 2013; Lacerda et al.2018). Nevertheless we list in here those galaxies that might beprobably confused with outflows, which do not imply these galax-ies are the only ones with eDIG in the CALIFA sample. It may bealso possible that some of these galaxies could be re-classified asoutflow candidates with better spatial and spectral resolution data. APPENDIX C: SUPPLEMENTAL MATERIAL
Fig. C1 shows the same plots from Fig. A1 but for a galaxy notclassified as an outflow candidate from Table B1. The remainingplots for the galaxies listed in Table B1 are available in the supple-mentary material for this article.
This paper has been typeset from a TEX / L A TEX file prepared by the author.
Table B1.
Galaxies with extraplanar ionized gas, but not classified as out-flows because they do not fulfil all the required criteria indicated in Sec3.1. The fraction of spaxels in the extraplanar region compatible with beingionized by an AGN, SF-driven shocks, and old-stars, based on the schemedescribed in Sec. 3.1 and Table 1 is included for reference.Object AGN:Shock:DIGNGC 0693 9:91:0PGC 0063016 1:99:0UGC 04730 23:77:0UGC 5392 6:94:0NGC 1677 1:99:0NGC 4149 2:80:18NGC 5908 7:63:30MCG -01-01-012 0:50:50NGC 2480 8:92:0NGC 5402 0:100:0NGC5439 0:100:0NGC 6361 5:88:7UGC 04550 12:88:0UGC 09262 2:98:0UGC 09665 0:100:0IC 2098 0:98:2IC 4582 3:97:0NGC 0681 0:72:28NGC 1056 12:87:1NGC 5145 3:94:2IC 1481 56:43:1IC 0540 30:14.:57 MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. Figure A1.
Spatially resolved line ratios and diagnostic diagrams together the WHAN diagram for the candidates galaxies to host outflows listed in Table 1.In each panel it has included a false colour image of the galaxy ( green: V-band, red: [N ii ] and blue: [O iii ]). The two black contours indicate the continuumlevel at 0.1 and 0.05 × − erg s − . The meaning of the demarcation curves and the symbols are the same from Figs. 1 and 2.MNRAS000
Spatially resolved line ratios and diagnostic diagrams together the WHAN diagram for the candidates galaxies to host outflows listed in Table 1.In each panel it has included a false colour image of the galaxy ( green: V-band, red: [N ii ] and blue: [O iii ]). The two black contours indicate the continuumlevel at 0.1 and 0.05 × − erg s − . The meaning of the demarcation curves and the symbols are the same from Figs. 1 and 2.MNRAS000 , 1–16 (2018) C. L´opez-Cob´a et al.
Figure A1. ( continue ) MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. Figure A1. ( continue )MNRAS000
Figure A1. ( continue ) MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. Figure A1. ( continue )MNRAS000 , 1–16 (2018) C. L´opez-Cob´a et al.
Figure A1. ( continue ) MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. Figure A1. ( continue )MNRAS000
Figure A1. ( continue ) MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. Figure A1. ( continue )MNRAS000 , 1–16 (2018) C. L´opez-Cob´a et al.
Figure A1. ( continue ) MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. Figure A1. ( continue )MNRAS000
Figure A1. ( continue ) MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. Figure A1. ( continue )MNRAS000 , 1–16 (2018) C. L´opez-Cob´a et al.
Figure A1. ( continue ) MNRAS , 1–16 (2018) ystematic study of outflows in the Local Universe using CALIFA: I. Sample selection and main properties. Figure C1.
Galaxy with detected extraplanar emission but not selected as outflow candidate. In each panel it has included a false color image of the galaxy(green: H α red: [N ii ] and blue: [O iii ]). The two black contours indicate the continuum level at 0.1 and 0.05 × − erg s − . Black circles in the diagnosticdiagrams indicate EW(H α ) <000